In the matter of 51Job, Inc - Judgment (Trial)

¶ 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 1 of 257 IN THE GRAND COURT OF THE CAYMAN ISLANDS FINANCIAL SERVICES DIVISION Neutral Citation Number: [2025] CIGC (FSD) 112 CAUSE NO: FSD 155 OF 2022 (DDJ) IN THE MATTER OF SECTION 238 OF THE COMPANIES ACT (AS REVISED) AND IN THE MATTER OF 51JOB, INC. Before: The Hon. Justice David Doyle Appearances: Richard Boulton KC and Mac Imrie KC instructed by Malachi Sweetman, Adrian Davey and Joanne Poland of Maples and Calder (Cayman) LLP on behalf of 51job, Inc. Jonathan Adkin KC, instructed by Rocco Cecere, Zachary Hoskin and Matthew Harders of Collas Crill LLP, Christopher Easdon of Campbells LLP and Rupert Bell and Patrick McConvey of Walkers (Cayman) LLP for the CCCW Dissenters Tom Lowe KC, instructed by Mark Ffrancon Dowds and Kalyani Dixit of Carey Olsen for the Carey Olsen Dissenters Heard: 24 June 2025 to 31 July 2025 (20 sitting days) Page 1 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 Digitally signed by Advance Performance Exponents Inc. Date: 2025.11.24 12:41:25 -05:00 Reason: Apex Certified Location: Apex 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 2 of 257 Competing lists of issues for determination filed: 28 August 2025 Written submissions on Trina Solar [2025] UKPC 48 filed: 31 October 2025 Draft Judgment circulated: 3 November 2025 Judgment delivered: 24 November 2025 Determination of the fair value of shares of a company pursuant to section 238 of the Companies Act (2025 Revision) Index Heading Paragraphs Introduction 1 to 7 Summary of outcome 8 The background 9 to 20 Timeline and some definitions 21 Law - Section 238 The Cayman Islands caselaw - Trina Solar - Shanda Games The Delaware jurisprudence - Cammer - DFC - Dell - Solera - Stillwater - Panera - Fir Tree v Jarden 23 24 to 26 27 to 56 57 to 61 62 to 69 70 to 80 81 to 95 96 to 117 118 to 134 135 to 136 137 to 145 Page 2 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 3 of 257 - Pivotal 146 to 151 The factual evidence - Mr Yan’s evidence - Ms Chien’s evidence - Mr Robert Anthony Bartell’s evidence - Mr Cheng’s evidence - Mr Tianji (Tony) Jiang’s evidence 153 to 212 213 to 280 281 to 334 335 to 396 397 to 440 The expert evidence - General - Professor Lehn’s first report - Mr Taylor’s first report - The joint memorandum - Professor Lehn’s supplemental report - Mr Taylor’s supplemental report - The expert oral evidence - Professor Lehn’s oral evidence - Mr Taylor’s oral evidence 442 to 462 463 to 468 469 to 483 484 485 to 496 497 to 519 520 521 to 572 573 to 702 Submissions 703 to 705 Lists of Issues - The Company’s list of issues - The Dissenters’ list of issues - Additional issues submitted by the Carey Olsen Dissenters 706 707 to 722 723 to 738 739 to 741 Determination - The hierarchy point - Rejection of the DCF analysis in this case - The methodologies - Merger price - Robust public information factors - The Cammer factors - Actively traded market - Analyst coverage - Market makers 743 to 749 750 to 755 756 to 758 759 to 761 762 763 to 764 765 to 766 767 to 774 775 Page 3 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 4 of 257 - Registration statements - Cause and effect - “Controller” theory - Access to non-public information and no discrimination between buyers - A chance to bid and a robust market check - The Special Committee - Involvement of Mr Hsu - Conflict and “heightened scrutiny” - Conclusion on robust market check and merger process - Two merger prices - Disclosure/discovery issues - The acquisition premium - The market price - Semi-strong efficient market - No MNPI - The November 2021 Press Release - The Ocean Link forecasts/projections - Value relevant information? - “Deal stock”? - Affected roll-forward - Co-movement issue - Volatility of the Golden Dragon Index - Affected trading price as start date for AMTP - The AMTP methodology - Standing back and sense-check 776 777 to 778 779 to 781 782 783 to 801 802 803 to 804 805 to 810 811 to 817 818 to 821 822 to 831 832 to 838 839 840 to 850 851 to 853 854 to 858 859 to 862 863 to 867 868 to 871 872 to 879 880 to 885 886 to 891 892 to 919 920 to 927 928 to 933 Conclusions 934 to 939 Summary of determinations 940 to 958 Ancillaries 959 The Order 960 Page 4 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 5 of 257 JUDGMENT Introduction

¶ [1] Under section 238 of the Companies Act (2025 Revision) (“Section 238”) a shareholder is entitled to payment of a fair value of that person’s shares upon dissenting from such a merger. At the hearing it is the duty of the court to determine the fair value of the shares of such dissenting members.

¶ [2] A lay person may think that if independent professional expert valuers are engaged that the margin of difference on a valuation of shares should not be too great. In this case the difference between the expert valuers is about US$1.7 billion. The merger price was US$61.00 per share/American Depository Share (“ADS”). The Company’s valuation expert using the adjusted market trading price (“AMTP”) methodology says that a fair value for the shares is US$31.11 each. The Dissenters’ valuation expert using the discounted cash flow (“DCF”) methodology says that a fair value for the shares is US$111.06 per share. Using different methodologies the experts have “arrived on different planets”, in “different galaxies” as the Delaware judges would say.

¶ [3] In this judgment I use ADS and shares interchangeably and where $ appears it is US$.

¶ [4] The Cayman Islands appears to have created an industry of its own out of Section 238 cases. But they put a huge strain on the legal system, on its judicial administration staff and on its judges. Well- resourced litigants, experts and lawyers raise every conceivable point and sometimes inconceivable points.

¶ [5] The lead-in time to trial is too long. Experts’ reports need to be more concise and in easily understandable language. Experts need to be truly independent and not partisan. Written submissions need to be shorter. It is not ideal that there are piecemeal additions to the bundles throughout the trial. Trials need to be shorter. Oral submissions and cross-examinations need to be shorter. Judgments take too long to write and are themselves too long in content. This one being a prime example. If I had more time at my disposal it would have been much shorter but I have many other cases crying out for scarce judicial time to be spent on them. Judgments need to be shorter and delivered within a reasonable time. The present approach of the parties and their attorneys makes this near impossible. Page 5 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 6 of 257

¶ [6] The parties, the experts and the attorneys must use their resources and skills to reduce the complexity rather than actively seeking to add to it. The parties and their legal advisors need to try much harder, as they are duty bound to do, to assist the court in achieving the overriding objective of dealing with cases justly, expeditiously and economically. I know this is easy for a judge to say and difficult for the parties and attorneys to undertake especially in complex cases where large amounts of money are at stake. Perhaps, however ,the Section 238 Sub-Committee of the Financial Services Division can consider the position and make practical recommendations for improvements in the way Section 238 cases are presently presented and conducted before the court. At the moment the position is far from ideal.

¶ [7] In fairness I should add that the attorneys and experts in this case did, from time to time, attempt to simplify certain issues for the court’s benefit. I also thank the attorneys in this case for enabling the trial to be completed within the time allocated. That no doubt took a lot of cooperation between the legal teams in the background, led by leading counsel and has not gone unnoticed. The fact remains however that insofar as Section 238 cases are concerned all is not well and changes need to be proposed, considered and where appropriate and possible implemented. Summary of outcome

¶ [8] In my judgment the fair value is $31.11 per share. The background

¶ [9] 51job, Inc (the “Company”) is an exempted limited company incorporated under the laws of the Cayman Islands. The Company was described as a leading provider of integrated human resource services in the People’s Republic of China (the “PRC”).

¶ [10] On 21 June 2021 the Company entered into a merger agreement with Garnet Faith Limited (“Garnet”) another Cayman Islands exempted company (the “Original Merger Agreement”). The consideration offered was US$79.05 per share (the “Original Merger Price”).

¶ [11] On 1 March 2022 the Petitioner and Garnet entered an agreement by which the Original Merger Agreement was amended (the “Revised Merger Agreement”). The consideration was US$61.00 per share (the “Revised Merger Price”). Page 6 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 7 of 257

¶ [12] On 27 April 2022 at 9.00am (Shanghai time) there was an extraordinary general meeting at which a special resolution approving the Revised Merger Agreement was passed.

¶ [13] On 6 May 2022 the plan of merger was executed and filed with the Registrar of Companies in the Cayman Islands.

¶ [14] On 14 June 2022 the Company made a fair value offer of US$56.38 per share to the Dissenters for the purposes of section 238(8) of the Companies Act (as revised) and without prejudice to the Company’s position at trial.

¶ [15] The Company presented a petition pursuant to Section 238 on 15 July 2022, and 21 dissenters did likewise.

¶ [16] There was a two days directions’ hearing on 21 and 22 November 2022. On 2 December 2022 I delivered a 14-page judgment. At paragraphs 3 and 5 I again stressed the duty of the parties and their attorneys to help the court to further the overriding objective of dealing with cases justly and at paragraph 4 I complained of the “somewhat hostile atmosphere between the attorneys.” At paragraph 22 I referred to the need for experts to provide independent assistance to the court by way of an objective unbiased opinion in relation to relevant matters within their expertise. I stated that an expert should never assume the role of an advocate or seek to promote a client’s case. At paragraph 28 I stated, “The parties should carefully choose their valuation experts and ensure that they are in a position to give relevant and helpful opinions to the Court and that they deal with the relevant issues in a simple and straightforward way and in accordance with their onerous duties to the Court.”

¶ [17] By Order made by consent on 21 June 2024 the trial of the petition was listed to commence at 10am on 24 June 2025 (20 hearing days allocated with the court not sitting on Fridays or Constitution Day and the trial finishing on 29 July 2025).

¶ [18] By sSmmons dated 28 February 2025 the Company sought to adjourn the trial on the basis that the 20 days previously agreed would not be long enough. The Company sought a 27 day fixture. Various applications for interrogatories were also filed and another two day hearing requested. By letter application dated 25 February 2025 the Company also sought access to various court files spanning over a period of nearly 10 years in respect of 10 other cases brought under Section 238. The hearing Page 7 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 8 of 257 took place on 16 and 17 April 2025 and I delivered judgment on 13 May 2025 in which I gave reasons for my decision on 16 April 2025 refusing the application for an adjournment and dealt with the letter application. As it transpired the hearing was duly completed within 20 sitting days and there were no complaints of unfairness. I am grateful to the attorneys for their cooperation and assistance in that respect.

¶ [19] By Summons dated 6 June 2025 the Collas Crill Dissenters applied for an order declaring that the Company had breached various undertakings and was in contempt of court. By consent order dated 24 June 2025 the Summons was withdrawn.

¶ [20] The trial commenced promptly on 24 June 2025. Timeline and some definitions

¶ [21] I provide the following timeline and some definitions for ease of reference: • 1998 - The Company commenced business operations in PRC • 1999 - Mr Rick Yan (“Mr Yan”) one of the co-founders of the Company appointed Chief Executive Officer (“CEO”) • September 2004 – Company listed on NASDAQ Global Select Market • 2004 - March 2009 – Ms Kathleen Chien (“Ms Chien”) one of the co-founders was the Chief Financial Officer (“CFO”) of the Company • March 2009 - Ms Chien appointed Chief Operating Officer (“COO”) • November 2009 – January 2024 – Ms Chien acting CFO • 7 March 2013 - Mr Li-Lan Cheng (“Mr Cheng”) appointed non-executive director of the Company • 20 June 2020 Mr Yan received an email from Mr Christopher Hsu (“Mr Hsu”) who wanted to explore opportunities between 58.com and the Company • 14 September 2020 – the Company engaged Mr Hsu, through his company Rocketeer Management sometimes known as Kilometre (“RM”) as a strategic consultant to explore potential strategic transactions Page 8 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 9 of 257 • 15 September 2020 – RM approached DCP Capital Partners II L.P. (“DCP”) to explore possible financing or strategic transactions involving the Company. • 16 September 2020 – share price $68.12 • 17 September 2020 – 27 April 2022 – Ms Chien was both COO and CFO • 17 September 2020 – DCP submitted a non-binding proposal (the “Original Proposal” or “DCP Proposal”) to acquire all of the outstanding shares of the Company, including shares represented by ADSs, for US$79.05 in cash per share or per ADS (the “Proposed Transaction”) • 17 September 2020 – take-private announcement (the “Announcement Date”) • 17 September 2020 – as at that date Mr Yan held 17.6% of the shares in the Company and was the second largest shareholder • 18 September 2020 – Ocean Link Partners II GP Limited (“Ocean Link”) (DCP and Ocean Link together the “PE Sponsors”) reached out to Company management and expressed an interest in participating in the Proposed Transaction as a member of a potential buyer consortium or group • 20 September 2020 – The Company set up a special committee (the “Special Committee”) consisting of two independent directors, Mr Cheng and Mr Eric He (“Mr He”) • 27 September 2020 – RM contact Recruit Holdings Co Ltd (“Recruit”) a company incorporated under the laws of Japan which held approximately 34.8% of shares in the Company at the time of the Original Proposal to discuss Recruit’s views regarding the Proposed Transaction. Recruit indicated it had not yet formed a view and would evaluate the Proposed Transaction • 30 September 2020 – the Special Committee engaged Davis Polk & Wardwell LLP (“Davis Polk” or “DPW”) as its independent legal counsel, and Kroll (formerly Duff & Phelps) (“Kroll” or “Duff & Phelps”) as its independent financial adviser • 10 October 2020 – DCP and Ocean Link commenced due diligence on the Company • 19 October 2020 – Recruit expressed its reluctance to support the Proposed Transaction at that time to the Company’s representatives • 21 October 2020 – RM reached out to Mr Yan, in his capacity as a major shareholder of the Company, to discuss his view on the Proposed Transaction. Mr Yan indicated that he had not decided whether to participate in the Proposed Transaction by joining a buyer consortium and/or rolling over all or a portion of his shares Page 9 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 10 of 257 • 21 October 2020 – RM advised the Company that the Proposed Transaction would be unlikely to succeed without the support of either Recruit or Mr Yan, given their respective voting power in the Company • 1-2 May 2021 – Paul Weiss/Kirkland & Ellis reached out to DPW on behalf of DCP/Ocean Link, respectively to seek the Special Committee’s consent to form a buyer consortium that included DCP, Ocean Link and Mr Yan (the “Buyer Consortium” or the “Buyer Group”) • 4 May 2021 – the formation of the Buyer Consortium was announced and the Buyer Consortium jointly delivered an updated non-binding going private proposal letter (the “Updated Proposal”) pursuant to which the parties proposed to, together with their respective affiliates, acquire all of the outstanding shares of the Company for US$79.05 in cash per share (the same purchase price as in the Original Proposal) • 7 May 2021 – Company management sent draft financial projections to the Special Committee, Kroll and DPW (the “2021 Management Projections” or the “Original Management Projections”) • June 2021 – Kroll orally delivered its opinion, and subsequently confirmed in writing that the offer of US$79.05 per share pursuant to the Original Merger Agreement was fair in financial terms to shareholders • 21 June 2021 – the Updated Proposal was accepted following the recommendation of the Special Committee and the Company entered into a merger agreement (the Original Merger Agreement) with a merger consideration of US$79.05 per share (the Original Merger Price) • 21 June 2021 – Kroll commenced the post-signing 30 day “go-shop” process and the 12 potential investors contacted did not express any interest • 21 June 2021 – Mr Yan, Recruit and Ms Chien enter into support agreements (the “Support Agreements”) effectively to vote in favour of the merger and against any competing transactions • August, September and December 2021 – representatives and advisers of DCP/Ocean Link held discussions with representatives and advisers of the Company on recent regulatory changes in PRC • 8 November 2021 – a press release (the “November 2021 Press Release”) was issued • 11 January 2022 – share price $45.83 (the “Pre-Announcement Price”) Page 10 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 11 of 257 • 12 January 2022 – the Buyer Consortium submitted a non-binding proposal letter (the “Revised Proposal”) to reduce the per share merger consideration and the per ADS merger consideration as contemplated in the Original Merger Agreement to US$57.25 • 12 January – 1 March 2022 - the Special Committee obtained financial and legal advice and held negotiations with the Buyer Consortium regarding the Revised Proposal • 28 January 2022 – Company management sent Kroll updated financial projections (the “2022 Management Projections” or the “Revised Management Projections”) • 28 February 2022 – the Buyer Consortium proposed US$59.95 as the per share merger consideration. After further negotiations that day, the Buyer Consortium agreed to US$61 per share • 1 March 2022 – Kroll orally delivered its opinion to the Special Committee, and subsequently confirmed in writing, that the offer of US$61.00 per share was fair in financial terms to shareholders • 1 March 2022 – the Buyer Consortium’s revised proposal was accepted and the Company entered into an amended merger agreement (“Revised Merger Agreement”) which provided that the merger consideration was US$61 per share (“Merger Consideration” or the “Revised Merger Price”) • 2 March 2022 – Kroll commenced the 15 day go-shop process with 11 potential investors none of whom indicated any interest in a potential alternative transaction involving the Company • 26 April 2022 – the Valuation Date • 27 April 2022 – the Company held an Extraordinary General Meeting (the “EGM”) at which a special resolution approving the Amended Merger Agreement was passed. 93.20% of the shares voted were in favour of the merger • 6 May 2022 – the Merger became effective Law

¶ [22] I deal now with the relevant law. Page 11 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 12 of 257 Section 238

¶ [23] Under section 238 of the Companies Act (2025 Revision) (“Section 238”) a shareholder is entitled to payment of a fair value of that person’s shares upon dissenting from a merger. At the hearing the court shall determine the fair value of the shares of such dissenting members together with a fair rate of interest, if any, to be paid by the company upon the amount determined to be the fair value. The Cayman Islands case law

¶ [24] Over the last decade there have been a number of cases in the Cayman Islands where judgments have cast light on section 238. These judgments include the following: (1) Integra Group 2016 (1) CILR 192 (Jones J) (2) Shanda Games Limited (unreported judgment delivered by Segal J on 25 April 2017) (3) Shanda Games Limited 2018 (1) CILR 352 (CICA) (4) Shanda Games Ltd v Maso Capital Investments Ltd [2020] UKPC 2; [2020] B.C.C. 466 (“Shanda Games”) (5) Qunar Cayman Islands Limited v Maso Capital Investments Limited and seven others 2019 (1) CILR 611 (Parker J) (“Qunar”) (6) Nord Anglia Education Inc (unreported judgment delivered by Kawaley J on 17 March 2020) (“Nord Anglia”) (7) Trina Solar Limited (unreported judgment delivered by Segal J on 23 September 2020) (8) Trina Solar Limited 2023 (1) CILR 569 (CICA) (9) Trina Solar Limited [2025] UKPC 48 (10) Changyou.com Limited 2021 (1) CILR 294 (Smellie CJ) (11) Changyou.com Limited (CICA judgment delivered 16 September 2022) (12) Changyou.com Limited [2025] UKPC 12 (13) FGL Holdings (unreported judgment delivered by Parker J on 20 September 2022) (“FGL”) (14) iKang Healthcare Group, Inc (unreported judgment delivered by Segal J 21 June 2023) (“iKang”) (15) Sina Corporation 2023 (2) CILR 400 (CICA) (“Sina”) Page 12 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 13 of 257 (16) Xingxuan Technology Ltd (unreported judgment delivered by Kawaley J on 9 September 2024)

¶ [25] I should add that judgments in 58.com (Margaret Ramsay-Hale CJ) and Sina (Parker J) are presently awaited and will no doubt add to Cayman Islands jurisprudence in this area of law once they are published.

¶ [26] Guidance can be gleaned from these authorities as to the correct approach of the court when determining a Section 238 petition. To fully understand the development and present state of Cayman Islands law in Section 238 cases it is instructive to consider three Cayman Islands judgments, all in Trina Solar: one from Sir Andrew Popplewell; one from Birt JA and one from Segal J. Trina Solar

¶ [27] On 30 September 2025 the Judicial Committee of the Privy Council (the “JCPC”) promptly delivered judgment in Trina Solar [2025] UKPC 48 after the appeal was heard on 8 July 2025. I have considered that judgment in detail together with the judgment of the Cayman Islands Court of Appeal (“CICA”) and the judgment of Segal J at first instance in Trina Solar. Trina Solar in the JCPC

¶ [28] Sir Andrew Popplewell (delivering the judgment of the JCPC in Trina Solar) set out the meaning of “fair value” in Section 238 as follows ([7]): “It is the valuation of the dissenter’s share in the value of the company as a going concern, discounted for their minority interest. It is not necessarily the same as the market price at which shares are trading because, for example, the value may be affected by information which is not publicly available to those conducting the trading, what was referred to as ‘Material Non Public Information’ or ‘MNPI’; or, as in this case, by knowledge of the merger proposal. Nor is the merger price necessarily representative of market value or fair value. For example, it too may be influenced by MNPI; and the merger process may have been conducted in a way which prevented or discouraged a higher offer from a rival bidder emerging. A DCF valuation is also subject to potential weaknesses and uncertainties as a Page 13 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 14 of 257 reliable guide to the company’s value. It is subject to multiple assumptions as to the future, by reference both to the commercial and financial conditions in which the company will be trading, and its own performance in those conditions. As is well known, a small variation in only one input is often capable of producing a large change in the output valuation.”

¶ [29] Sir Andrew Popplewell felt it necessary to refer to three Delaware decisions (DFC, Dell and Solera) which I refer to below together with some other Delaware authorities.

¶ [30] Having briefly referred to DFC, Dell and Solera Sir Andrew Popplewell stated: “15. A number of observations may be made about these authorities and the applicable principles. Transaction price, market price, comparable company valuation and DCF valuation are all valuation measures which may properly be drawn on in seeking to assess fair value. This is not intended to be an exhaustive list. There is no hierarchy between them. There is no presumption, for example, in favour of the transaction price or a market price. All merger transactions are factually unique and the relevant circumstances will often differ very substantially from case to case. In considering each measure, the court will usually be required to take account of extensive factual and expert evidence and consider many variables. Each methodology must be assessed individually to identify strengths and weaknesses which may affect its reliability as a guide to fair value. Reliability is not in this context a binary concept in which the court must conclude that the measure in question is or is not reliable; rather it is a qualitative concept in which the court may conclude that it is more or less reliable on a sliding scale. Often there will be uncertainties inherent in the methodology itself. For example, when considering the transaction price, the court will be concerned with the hypothetical possibility of rival bids; and in a DCF valuation, with future projections and the discount rate. Sometimes uncertainties will arise from paucity in the evidence before the court. Save where it is proper to draw adverse inferences or to treat a party as having failed to discharge an evidential burden which it bears, the Court has to form a view of the effect of these uncertainties. Page 14 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 15 of 257

¶ [16] Moreover, the exercise is not one simply of assessing the reliability of each methodology individually, but one of assessing comparative reliability between all of them. The courts in the Cayman Islands and Delaware have sometimes used one methodology as a cross-check against another; sometimes they have adopted a percentage weighting of a kind the Judge adopted in this case. Where the values reached by the different valuation measures are relatively close, the concept of a cross-check between them has some coherence. Where, however, the values are widely divergent, it makes little sense to talk of using one as a cross-check against another. In this case, for example, the DCF Value reached by the Judge was some 2 ½ times larger than the Market Price (and indeed on the Dissenters’ expert evidence would have been potentially 10 times greater or more). Neither could realistically be said to support the other. Nevertheless in such circumstances the court may properly ascribe some weight to each by a relative weighting. The court may be faced with two valuation measures, neither of which it would consider to be more reliable than not as a guide to fair value if taken individually. Nevertheless it is entitled to ascribe more than 50% weighting to one if it considers it less unreliable than the other. Assessing reliability is not only a qualitative exercise in relation to each methodology individually, but the court is concerned with a qualitative assessment of the reliability of the methodologies relative to each other.

¶ [17] It follows that the task for the trial judge is one which is highly case specific and highly fact dependent. It not only involves findings of primary fact but is peculiarly dependent on nuanced evaluative assessments. In considering the hypothetical questions which often arise, the court is dealing in risks and uncertainties which are not readily quantifiable. The same is true in relation to the conclusory evaluation of the extent to which a methodology provides a reliable guide to fair value, both individually and relative to another methodology.

¶ [18] In this context, the lists of factors set out in cases such as Dell, DFC Global and Solera can be a useful guide to considerations which in any individual case may point to a greater or lesser degree of reliability in using the transaction price. They are not, however, to be treated as checklists, all or most of which must be met before any reliance may be placed on a transaction price. There may, for example, have been substantial failings in Page 15 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 16 of 257 relation to any market check process which is designed to flush out rival bidders at a higher price than the proposed transaction price; but the court may in a particular case be satisfied that there were no realistic rival bidders; or that none would have come forward even with the most exemplary market check process. It would be perfectly proper for the court to attach weight to a transaction price in those circumstances, although the fact that the court has been deprived of evidence of the point being tested may increase the level of uncertainty and consequently reduce the weighting to be attributed to such price. All will depend on the facts of the individual case, including the relative reliability of other methodologies contended for.”

¶ [31] The JCPC considered that the CICA was not entitled to interfere with Segal J’s assessment that a weighting of 45% should be accorded to the Merger Price ([39]). The JCPC added: “The Judge was looking at the position holistically and in the round, whereas the Court of Appeal was focusing on only one component element of the analysis” ([41]).

¶ [32] The JCPC at [42] referred to the way in which the CICA addressed the need for a robust market check and Birt JA’s statement at [142] that “if the breaches are substantial, this is likely to mean that the merger price is not reliable as an indicator of fair value and accordingly little or no weight should be given to the merger price.” Sir Andrew Popplewell felt that: “This is too rigid an approach and unduly elevates the importance of a market check to something approaching a threshold. In a case such as the present the existence of any deficiencies in the market process (which is a more appropriate term than “breaches”) needs to be evaluated alongside a consideration of the impact of such deficiencies; and alongside consideration of other factors which militate in favour of the reliability of the merger price.”

¶ [33] The JCPC at [46] referred to the CICA’s comments in respect of the Delaware jurisprudence and the need for a robust market check and “outreach to all logical buyers” and the CICA’s view that it would Page 16 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 17 of 257 usually be the case that it would be public knowledge, of which potential bidders were aware, that the company was in play; and if that fact alone were sufficient, there would be no need for a market check. The JCPC criticised this approach as: “… it appears to treat the requirement of outreach to all logical buyers as if it is a threshold requirement which must substantially be met before any reliance can be placed on a merger price. That is contrary to the multifactorial and fact specific evaluation which must be undertaken in each case, and which the Judge took conspicuous care to undertake in this case” ([46]).

¶ [34] The JCPC stated at [48] that the CICA “was not justified in giving no effect at all to the factors which the Judge identified as lending support to the reliability of the Merger Price.” The JCPC referred to Birt JA stating that although the majority vote may not always be so large, these were features which were likely to be present, in many if not most cases; that they do not diminish the need for a robust market check; and that he did not see “that the existence of the factors listed by the judge can possibly balance or outweigh the deficiencies in the market check process and the concerns about the impact of Mr Gao’s position.” Sir Andrew Popplewell at [49] stated: “This suggests a number of errors of approach.” He felt it did not “identify all the factors properly relied upon by the Judge …” and “it also unduly elevates the importance of a robust market check to something approaching a threshold requirement, in the absence of which no reliance can be placed on a transaction price.” Sir Andrew Popplewell added: “The Judge expressed his findings about the deficiencies in the market check process in a nuanced way as giving rise to “concern”. He was in the best position properly to assess the degree of that concern and the significance of the deficiencies in the light of countervailing factors. His findings that they did not undermine the Merger Price as one to which some resort could be had was an evaluative assessment which he was very much better placed to make than an appellate court …”.

¶ [35] Sir Andrew Popplewell from [50] onwards deals with the proper approach to “reliability” of a valuation measure. The JCPC took the position that an approach that treats “reliability as a concept in which the supporting factors and undermining factors fall to be balanced so as to produce an all or nothing result, such that if the weaknesses are more potent than the strengths, no weight can be attached”… “erroneously treats reliability as a binary concept and ignores the fact that weight may be attached to a particular valuation measure even where the court concludes that the uncertainties or weaknesses make it less rather than more reliable.” Page 17 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 18 of 257

¶ [36] The JCPC felt that the CICA “also failed to treat reliability as a comparative concept as between different valuation methods, in which they are to be assessed by reference to how reliable one is relative to the reliability of another” ([51]). Sir Andrew Popplewell stated: “52. This error in approach is perhaps most obvious in what the Court of Appeal did by way of redistribution of the 45% Merger Price weighting. Having decided that it would not disturb the weighting of 30% applied to the Market Price on the basis that the Judge could not be shown to have erred in giving it that weighting, and discarded the Merger Price, it added the 45% to the DCF Value so as to fill the gap without any analysis or reasoning. The court simply failed to undertake the required comparative reliability exercise which was necessary if the Merger Price was to be accorded no weight. The surprising result was that despite Ms Glass’ evidence, with detailed reasons, that a market based approach was much more reliable than a DCF Value, and despite the Judge finding that her evidence generally was reliable and was in this specific respect to be taken as his starting point for relative weighting, the ultimate weighting would be 70% in favour of the DCF Value, almost three times that attributed by the Judge and far more than the market based weighting.”

¶ [37] On the issue of management projections Sir Andrew Popplewell had this to say at [57]: “…When considering the reliability of the Management Projections generally, the Judge noted at para 189(b) that the Company had explained how they were compiled and that those involved in their preparation had an in-depth knowledge of the Company’s business. Projections produced by a company form a starting point in a DCF valuation because a company is taken to know its own business; and insider knowledge which informs such management projections may justify according them greater weight than industry evidence based on third party source material. Ms Glass was not giving primary evidence of solar industry experience, but rather assisting in the evaluation of the evidence of Dr Goffri and Mr Russo from a valuation point of view. The Judge was not bound to accept the evidence of any of the three as to the exact figures which should be taken, or those in the Management Projections, but rather had to evaluate for himself the worth and the effect of each. It is important to keep in mind the nature of the exercise here being conducted. It Page 18 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 19 of 257 involved making a prediction at the valuation date of what would have happened over the subsequent seven years … The Court of Appeal’s approach was to treat it as a calculation exercise, rather than, as the Judge correctly treated it, an exercise in estimation which could produce a range of reasonable results…”

¶ [38] The JCPC allowed the appeal and accordingly restored the assessment of fair value to that determined by Segal J ([67]).

¶ [39] I do not treat the judgment of the JCPC as delivered by Sir Andrew Popplewell as a statute but as a first instance judge I list the following main “takeaways” from it: (1) The approach of the Delaware Courts can be of assistance and value in this area ([9]). (2) There is no hierarchy between the valuation measures ([15]). (3) Each methodology must be assessed individually to identify strengths and weaknesses which may affect its reliability as a guide to fair value. Reliability is not in this context a binary concept in which the court must conclude that the measure in question is or is not reliable; rather it is a qualitative concept in which the court may conclude that it is more or less reliable on a sliding scale ([15]). (4) The exercise is not one of simply assessing the reliability of each methodology individually, but one of assessing comparative reliability between all of them ([16]). (5) Where the values reached by the different valuation measures are relatively close, the concept of a cross-check between them has some coherence. Where, however, the values are widely divergent, it makes little sense to talk of using one as a cross-check against another ([16]). (6) Assessing reliability is not only a qualitative exercise in relation to each methodology individually, but the court is concerned with a qualitative assessment of the reliabilities of the methodologies relative to each other ([16]). Page 19 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 20 of 257 (7) The task for the trial judge is one which is highly case specific and highly fact dependent. It not only involves findings of primary fact but is peculiarly dependent on nuanced evaluative assessments ([17]). (8) The list of factors set out in cases such as Dell, DFC and Solera can be a useful guide to considerations which in any particular case may point to a greater or lesser degree of reliability in using the transaction price. They are not, however, to be treated as checklists, all or most of which must be met before any reliance may be placed on a transaction price. All will depend on the facts of the individual case, including the relative reliability of other methodologies contended for ([18]). (9) The trial judge should look at the position holistically and in the round rather than focusing on only one component element of an analysis ([41]). (10) The trial judge should not adopt a rigid approach to the market process and the absence of rival bidders and should not unduly elevate the importance of a market check to something approaching a threshold. Any deficiencies in the market check process need to be evaluated alongside a consideration of the impact of such deficiencies and alongside consideration of other factors which militate in favour of the reliability of the merger price ([42]). (11) It is wrong to treat the requirement of outreach to all logical buyers as if it were a threshold requirement which must substantially be met before any reliance can be placed on a merger price. A multifactorial and fact specific evaluation must be undertaken in each case ([46]). (12) The importance of a robust market check should not be unduly elevated to something approaching a threshold requirement, in the absence of which no reliance can be placed on a transaction price ([49]). (13) Reliability of a valuation measure or methodology should not be treated as a concept in which the supporting factors and undermining factors fall to be balanced so as to produce an all or nothing result, such that if the weaknesses are more potent than the strengths, no weight can be attached. That would erroneously treat reliability as a binary concept and would ignore the fact that weight may be attached to a particular valuation measure even Page 20 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 21 of 257 where the court concludes that the uncertainties or weaknesses make it less rather than more reliable ([50]). (14) Reliability is a comparative concept as between different valuation measures, in which they are to be assessed by reference to how reliable one is relative to the reliability of another ([51]). (15) If a measure or methodology is to be accorded no weight the court must undertake a comparative reliability exercise ([52]). (16) Projections produced by a company form a starting point in a DCF valuation because a company is taken to know its own business and insider knowledge which informs such management projections and may justify according them greater weight than industry evidence based on third party source material. It is important to keep in mind the nature of the exercise being conducted. In a DCF analysis it involves making a prediction at the valuation date of what would have happened over subsequent years (in Trina Solar seven subsequent years). It is not a calculation exercise, it is an exercise in estimation which could produce a range of reasonable results. The judge must evaluate the worth and effect of the evidence before the court ([57]). (17) Figures should not be plucked out of the air and there should be no tinkering ([66]). Trina Solar in the CICA

¶ [40] The judgment of the CICA in Trina Solar must now be read in light of the JCPC’s judgment.

¶ [41] Birt JA at [34] referred to Segal J’s comments at [91] in respect of fair value. In ascertaining fair value, the court must assess and determine a monetary amount which in the circumstances represents (its best estimate of) the worth, the true worth, of the dissenting shareholder’s shares (true worth meaning the actual value to the shareholder of the financial benefits derived and available to him from his shares and by being a shareholder). The reference to fair requires amongst other things that the manner and method of that assessment and determination is fair to the dissenting shareholder by ensuring that all relevant facts and matters are considered and that the sum selected properly reflects the true monetary Page 21 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 22 of 257 worth to the shareholder of what he has lost, undistorted by the limitations and flaws of particular valuation methodologies and fairly balancing, where appropriate, the competing, reasonably reliable alternative approaches to valuation relied on by the parties.

¶ [42] Birt JA in Trina Solar at [13]-[15] stated that in cases concerning fair value under Section 238 there were three main valuation methodologies which fall for consideration: Adjusted trading price or market price “… The first is the adjusted trading price. This is the price at which the shares were trading on the relevant stock market but subject to an estimate of what the price of those shares would have been as at the valuation date in the absence of any proposed merger. In other words an adjustment is made to try and strip out the effect on the trading price of the fact that an offer to acquire the shares has been made at a known price. When speaking in general terms, I shall refer to this as “the adjusted trading price” or “the market price” (at [13]). Merger price “Secondly, there is the price which has been offered by the acquiring entity and which has been accepted by the shareholders other than those who are dissenting. When speaking in general terms, I shall refer to this price as the “merger price” (at [14]). Discounted cash flow “Finally, there is a Discounted Cash Flow (“DCF”) valuation. This was summarised by this court in the case of In Re Shanda Games Limited (13) (2018 (1) CILR 352, at para. 61) in the following terms: “As its name indicates, a DCF analysis contains two main elements: a prediction of future cash flows, and the application to those cash flows of a discount rate so as to translate the future cash flows into a present capital value. In effect, the exercise is designed to identify how much it would have cost at the valuation date Page 22 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 23 of 257 to buy an investment with a rate of return and a risk profile equivalent to that of the company’s business” (at [15]).

¶ [43] Birt JA in Trina Solar at [35] summarised, without criticism, various points made by Segal J at first instance: (1) The 'true worth' of a share can in an appropriate case be assessed by assuming an immediate sale under certain conditions (i.e. a 'market based' approach using the adjusted trading price and/or merger price as a measure of fair value). This will often be the most reliable method of capturing the full monetary worth of the share. But 'true worth' can also, in an appropriate case, be assessed by assuming that the shareholder retains the share and obtains the financial benefits of doing so. This is sometimes referred to as the 'intrinsic value' i.e. establishing a monetary value for the shareholder’s bundle of rights by reference to the financial benefits flowing from the right to participate in profits and obtain distributions in a winding up. (2) A DCF valuation of the company can assist and be relied upon on either approach; either (directly) as a model of what the shareholder will receive if the share is retained or (indirectly) as an estimate of what a purchase would be likely to pay for the share if it is sold. (3) Apart from the fact that market perceptions may affect the market price, the worth of a share is not, as matter of principle, the same as or limited to its market price, but can also be assessed by reference to the financial benefits which can be derived from the retention of the share. As the CICA said in Shanda Games at [22] “… fair value is not necessarily the same as the merger price or the price at which the shares were trading”. (4) The selection of the valuation methodology (or methodologies) is a fact sensitive issue. In some cases it would be appropriate to give particular weight to market based indicia of value and use a DCF valuation as a means of testing those other valuation methodologies. Thus, where the adjusted trading price and merger price may be seen as reliable, they may probably be used as the starting point of the valuation subject to testing by reference to a DCF valuation. Page 23 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 24 of 257 (5) Although Segal J did not specifically say this, Birt JA felt it was implicit and was correct that in other cases it will be appropriate to estimate fair value exclusively or primarily using a DCF valuation. No plumping but in toto possibility

¶ [44] At [36] Birt JA added: “What is clear is that, although it will undoubtedly look to the expert evidence in order to assist it, the court must reach its own decision as to fair value and the constituent elements which go to make up that fair value. It must not simply plump for one expert over another …” Birt JA quoted from Vice-Chancellor Strine in a Delaware case in a passage approved by Kawaley J in Re Nord Anglia Education at [70]: “In making the fair value determination, the court may look to the opinions advanced by the parties’ experts, select one party’s expert opinion as a framework; fashion its own framework, or adopt piecemeal, some portion of an expert’s model methodology or mathematical calculations. But, the court may not adopt an 'either-or' approach and must use its judgment and an independent valuation exercise to reach its conclusion.”

¶ [45] In this context it is interesting also to note footnote 104 of the judgment of the Supreme Court of Delaware in Dell Inc v Magnetar Global Event Driven Master Fund Ltd No. 565, 2016 (Del. Supr. December 14, 2017) at page 35. I shall consider Dell in detail later in this judgment. For the present purposes I simply set out the relevant passage as follows: “As such, “the trial of an appraisal case under the Delaware General Corporation Law presents unique challenges to the judicial factfinder.”102 And this task is complicated by “the clash of contrary, and often antagonistic, expert opinions of value,” prompting the trial Page 24 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 25 of 257 court to wade through “widely divergent views reflecting partisan positions” in arriving at its determination of a single number for fair value.103 In the end, after this analysis of the relevant factors, “[i]n some cases, it may be that a single valuation metric is the most reliable evidence of fair value and that giving weight to another factor will do nothing but distort that best estimate. In other cases, it may be necessary to consider two or more factors.”104 Or, in still others, the court might apportion weight among a variety of methodologies. But, whatever route it chooses, the trial court must justify its methodology (or methodologies) according to the facts of the case and relevant, accepted financial principles.105”

¶ [46] Footnote 104 reads: “104 DFC, 2017 WL 3261190 at *31; see also M.G. Bancorp., 737 A.2d at 525-26 (“[T]he Court of Chancery has the discretion to select one of the parties’ valuation models as its general framework or to fashion its own.”); id. At 526 (“[A]lthough not required to do so, it is entirely proper for the Court of Chancery to adopt any one expert’s model, methodology, and mathematical calculations, in toto, if that valuation is supported by credible evidence and withstands a critical judicial analysis on the record.”).”

¶ [47] Birt JA in Trina Solar referred to the need for the court to reach its own decision having been emphasised by Martin JA in Re Shanda Games (CICA) at [22]: “… Since the fair value is not necessarily the same as the merger price or the price at which shares were trading before the market in them was affected by knowledge of the merger, it is inevitable that the determination will involve an assessment of a substantial quantity of information relating to the financial affairs of the company whose shares are to be valued. It is unlikely in the extreme that the court will be able to make that assessment without expert assistance. In the ordinary case, as in this one, both the company and any dissenting shareholders will appoint experts; but even in a case where no dissenting shareholder is prepared to participate in the litigation, the company, and perhaps also the court itself, will instruct an expert. In every case, the court’s task will be to assess the utility of the expert Page 25 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 26 of 257 evidence to the determination of fair value … the court will determine generally, or on an issue-by-issue basis, whether an expert’s evidence is to be accepted in whole or in part and how conflicts are to be resolved. If necessary, the court is entitled to substitute its own view for that of the experts.” Semi-strong form efficiency

¶ [48] Birt JA in Trina Solar at [41] stated: “(i) Semi-strong form efficiency (a) Introduction 41 The judge correctly accepted that, if weight was to be placed on the adjusted trading place of the ADS when determining fair value, there must be a semi-strong efficient market in those ADS. Thus he said (Cause No. FSD 92 of 2017, at para. 131): “As regards [the efficient market hypothesis] the Dissenting Shareholders’ core proposition is that it could not be assumed that in a given market for a given asset the trading price will approximate fair value, that considerable care has to be used before relying on the trading price and it was wrong to assume that markets get it right all the time. These propositions are unobjectionable but, as a general matter, I am not persuaded that it is inappropriate for the Court to place reliance on the market price where there is sufficient evidence demonstrating that the requirements for a semi-strong version of the efficient market hypothesis are satisfied.” 42 A similar approach was adopted by Kawaley, J. in Nord Anglia (Cause No. FSD 235 of 2017, at para. 6). 43 In this respect, as he said (Cause No. FSD 92 of 2017, at para.128), the judge was adopting the approach followed by the courts of Delaware, particularly in the cases of Dell Inc. v. Magnetar Global Event Driven Master Fund Ltd. (5) and DFC Page 26 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 27 of 257 Global Corp. v. Muirfield Value Partners LP (4), both decisions of the Delaware Supreme Court. 44 In Dell, the Court of Chancery had decided that no weight should be given to the trading price or the merger price and that fair value should be assessed solely by reference to a DCF valuation. The Supreme Court held that, on the facts of that case, this approach was wrong and remitted the case to the Court of Chancery for decision, albeit making clear its view that, on the facts, the merger price should carry very substantial, if not exclusive, weight. In the course of reaching its conclusion, the Supreme Court said this (177 A.3d at 24): “Further, the Court of Chancery’s analysis ignored the efficient market hypothesis long endorsed by this Court. It teaches that the price produced by an efficient market is generally a more reliable assessment of fair value than the view of a single analyst, especially an expert witness who caters her valuation to the litigation imperatives of a well-heeled client.””

¶ [49] As to what is meant by a “semi-strong form efficient market” Birt JA in Trina Solar at [45] commented that Kawaley J had summarised the position concisely in Nord Anglia at [58] in the following terms: “It was agreed that the stock market is not generally considered to be “strong-form efficient” i.e. with market prices reflecting both public and private information. A market is “semi-strong form efficient” if the market fully reflects all public information. [Emphasis added]. It was agreed that in a semi-strong efficient market: (a) stock prices respond to and rapidly incorporate new value-relevant publicly available information (market-industry and/or company specific); and (b) no significant changes in stock price would be expected on days when there was no new value-relevant information.” [Emphasis in original]”.

¶ [50] Birt JA at [46] to [48] in Trina Solar added: Page 27 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 28 of 257 “46 The judge expressed the matter as follows (Cause No. FSD 92 of 2017, at para 111): “The Company submitted that, as was clear from DFC and Dell, the Delaware courts subscribed to the semi-strong version of the efficient market hypothesis … This is that well-traded stocks in a mature market, such as the US, are semi-strong efficient, which means the prices reflect all publicly-available information regarding the stock, the company, the industry, the economy but not private information.” 47 The most detailed and helpful summary of the relevant factors when assessing whether a market is semi-strong efficient is to be found in Dell (5). Thus, the Delaware Supreme Court said this (177 A.3d at 7-8): “Dell had a deep public float and was actively traded as more than 5% of Dell’s shares were traded each week. The stock had a bid-ask spread of approximately 0.08%. It was also widely covered by equity analysts, and its share price quickly reflected the market’s view on breaking developments. Based on these metrics, the record suggests the market for Dell stock was semi-strong efficient, meaning that the market’s digestion and assessment of all publicly available information concerning Dell was quickly impounded into the Company’s stock price.” 48 Later, the Supreme Court said this (ibid., at 28): “A market is more likely efficient, or semi-strong efficient, if it has many stockholders; no controlling stockholder; ‘highly active trading’; and if information about the company is widely available and easily disseminated to the market. In such circumstances, a company’s stock price ‘reflects the judgments of many stockholders about the company’s future prospects, based on public filings, industry information, and research conducted by equity analysts.’ In these circumstances, a mass of Page 28 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 29 of 257 investors quickly digests all publicly available information about a company, and in trading the company’s stock, recalibrates its price to reflect the market’s adjusted, consensus valuation of the company.” In agreement with the judge, I consider that these observations of the Delaware Supreme Court are equally applicable when courts in this jurisdiction have to determine whether a market in a company’s shares is semi-strong efficient for the purposes of deciding whether weight should be placed on the adjusted trading price in assessing fair value.”

¶ [51] In a passage emphasised by counsel in the case presently before me, Birt JA at [65] stated: “65 As to the additional points made by Mr. Salzedo as summarized at para. 64(v) above, when determining the weight to be given to the market price, the judge expressly took into account the fact that Ms. Glass had accepted that the trading price might have been understated by 20%. I do not accept that this of itself means that no weight could be placed on the market price. In the passage referred to earlier, Ms. Glass is talking about markets generally and the fact that prices may be out (as compared with fair value) by up to 20% (or indeed 25%). Despite this possibility generally, it is nevertheless the case that a court may place some weight on the adjusted trading price if satisfied that there is a semi-strong efficient market and no material non-public information. Whether it will be right to do so in any particular case is a matter of evaluation having regard to the particular circumstances.” Merger price and market check

¶ [52] In respect of the merger price and the weight that can be attached to it, Birt JA stated at [137] to [145]: “137 Subject to the qualification at para. 146 below, I consider that the judge accurately summarized (Cause No. FSD 92 of 2017, at para. 156) the circumstances in which weight can be placed on the merger price as evidence of fair value (see para. 113 above). Page 29 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 30 of 257 138 As the judge made clear, his formulation drew upon the decision of the Delaware Supreme Court in Dell (5). The other leading case in Delaware is the decision of the Supreme Court in DFC (4). In the latter case, the court emphasized the need for a “robust market check” (172 A. 3d at 366) and went on to place reliance on the fact that the financial adviser engaged by the company had approached “every logical buyer” (ibid., at 376). 139 In my judgment, a helpful summary of some of the features to be looked for when considering whether weight should be placed on a merger price is to be found in Re Solera Holdings Inc. (14) where Bouchard, C. said ([2018] WL 3625644, at P17): “I now turn to my own independent determination of the fair value of Solera’s shares with the guidance from DFC and Dell in mind. Those decisions teach that deal price is ‘the best evidence of fair value’ when there was an ‘open process,’ meaning that the process is characterised by ‘objective indicia of reliability.’ Such ‘indicia’ include but, consistent with the mandate of the appraisal statute to consider ‘all relevant factors,’ are not limited to: • ‘[R]obust public information,’ comprised of the stock price of a company with ‘a deep base of public shareholders, and highly active trading,’ and the views of ‘equity analysts, equity buyers, debt analysts, debt providers and others.’ • ‘[E]asy access to deeper, non-public information’ where there is no discrimination between potential buyers and cooperation from management helps address any information asymmetries between potential buyers. Page 30 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 31 of 257 • ‘[M]any parties with an incentive to make a profit had a chance to bid,’ meaning that there was a ‘robust market check’ with ‘outreach to all logical buyers’ and a go-shop characterized by ‘low barriers to entry’ such that there is a realistic possibility of a topping bid. • [A] special committee ‘composed of independent, experienced directors and armed with that power to say “no”,’ which is advised by competent legal and financial advisors. • ‘[N]o conflicts related to the transaction,’ with the company purchased by a third party in an arm’s length sale and ‘no hint of self-interest.’” 140 In relation to the last bullet point, I do not consider that the existence of any conflict of interest precludes absolutely any reliance on a merger price, but it certainly militates against such reliance. The present case is in effect a management buyout in that Mr. Gao, as founder, chairman and chief executive of the company, was also part of the buyer group. In such circumstances, there must, in my judgment, be heightened scrutiny in order to ensure that there is a realistic and fair opportunity for alternative bidders. Inevitably, in a management buyout, there is considerable scope for management to acquire the company on the cheap because of their detailed knowledge of the company as compared with the knowledge of ordinary shareholders and other potential bidders. 141 Examples of the sort measures which have been considered as a “robust market check” can be found in some of the cases: (i) The situation in Dell (5) was not wholly dissimilar to the present case in that Mr. Dell, as founder and CEO of Dell, held 15.4% of the shares and it was considered important that he should continue with the company. To that end, he had agreed to explore in good Page 31 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 32 of 257 faith the possibility of working with any potential buyer and to vote his shares in proportion to the unaffiliated shares. The special committee of independent directors did not carry out a market check before signing a merger agreement with a company called Silver Lake. However, there was then a well publicized go-shop period of forty-five days in which the committee had the opportunity to find a better offer. During the go-shop period, the financial advisers to the committee approached some 67 parties, including 20 potential strategic buyers and 17 financial companies. Amongst the potential strategic buyers was Dell’s closest competitor. The Supreme Court reviewed the process which had been followed in considerable detail and concluded that the Court of Chancery below had been wrong not to place weight on the merger price because there was compelling evidence of “market efficiency, fair play, low barriers to entry, outreach to all logical buyers, and the chance for any topping bidder to have the support of Mr Dell’s own votes.” (ii) In DFC (4), the Delaware Supreme Court overturned the decision of the Court of Chancery to give only one-third weighting to the deal price and remitted it to the lower court with clear guidance to give considerable and possibly exclusive weighting to the deal price. This was in circumstances where: (a) the proposed purchaser was a third party acting at arm’s length, (b) the financial adviser to the company had approached every logical buyer (41 financial buyers and 3 strategic buyers) over a two year period before the deal, and (c) there was no hint of self-interest that compromised the market check. The deal did not involve the potential Page 32 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 33 of 257 conflicts of interest in a management buyout or negotiations to retain existing management. (iii) In Nord Anglia (9), Kawaley, J. agreed with the need for a robust sales process as described in Dell and said (Cause No. FSD 235 of 2017, at para. 86): “In re Appraisal of Petsmart, Inc. is also instructive as regards illustrating what type of auction process has been judicially viewed as robust. The merger process was initiated by an ‘activist hedge fund’ which had acquired a 9.9% stake in the company, not by a controlling shareholder. 27 potential bidders were identified between August 2014 and October 2014, and 15 potential bidders signed non-disclosure agreements. A data room including non- public information was made available to the potential bidders and management made presentations to them. Five bids were received by October 31, 2014; on November 3, 2014, four were allowed to continue further with their bids. The Board in early December was still considering alternatives to a sale. Final bids were considered on December 10, 2014. There were three separate competing bidding groups, and one group expanded its membership near the end of the auction process to increase its bidding capacity. The highest bid was ultimately accepted. The merger was approved by over 77% of all shareholders and over 99% over those who voted. The dissenters did not challenge the efficacy of the process as such. The valuation process deployed by the Ad Hoc Committee’s advisers was subjected to apparently strong criticism, as was their impartiality; it was also argued, inter alia, that market conditions limited the range of participants in the auction process, such that the Board did not make a properly informed decision. The Vice- Chancellor crucially concluded: Page 33 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 34 of 257 ‘Importantly, the evidence reveals that the private equity bidders did not know who they were bidding against and whether or not they were competing with strategic bidders. They had every incentive to put their best offer on the table … … Respondent has carried its burden of demonstrating that the Merger Price of $83 per share was the result of a ‘proper transactional process’ comprised of a robust pre-signing auction in which adequately informed bidders were given every incentive to make their best offer in the midst of a ‘well-functioning market.’” [Emphasis added.] 142 Although, as stated above, the judge misunderstood the decision in Norcraft, I reject Mr. Salzedo’s submission, which was effectively that no or minimal weight must be given to the merger price if any of the conditions listed (Cause No. FSD 92 of 2017, at para. 156) are found not to be satisfied. It seems to me that this must depend upon the manner and extent to which the conditions are not satisfied. If there are fairly minor breaches, this may well enable weight still to be given to the merger price on the basis that it can still be regarded as a reliable indicator of fair value; see for example Nord Anglia, where Kawaley, J. found that the sale process, although an arm’s length process, was not as robust as it might have been but concluded that the merger price still provided a reasonable indicator of fair value and gave it a weighting of 60% with 40% being attributed to a DCF valuation. However, if the breaches are substantial, this is likely to mean that the merger price is not reliable as an indicator of fair value and that accordingly little or no weight should be given to the merger price when assessing fair value. 143 In this connection, Mr. Jones submitted that the case of Solera (14) supported the judge’s decision in the present case to give weight to the merger price despite the defects which he had identified in the sale process. Page 34 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 35 of 257 However, the facts of Solera were very different. Bouchard, C. summarized the position as follows ([2018] WL 3625644, at 1): “As discussed below, the record reflects that Solera was sold in an open process that, although not perfect, was characterized by many objective indicia of reliability. The merger was the product of a two-month outreach to large private equity firms followed by a six-week auction conducted by an independent and fully authorized special committee of the board, which contacted eleven financial and seven strategic firms. Public disclosures made clear to the market that the company was for sale. The special committee had competent legal and financial advisors and the power to say no to an underpriced bid, which it did twice, without the safety net of another bid. The merger price of $55.85 proved to be a market-clearing price through a 28-day go-shop that the special committee secured as a condition of the deal with Vista, one which afforded favorable terms to allow a key strategic competitor of Solera to continue to bid for the company.” [Emphasis added.] 144 The reference to “although not perfect” would appear to be a reference to two matters. The first was the court’s finding (ibid., at 22) that the board of Solera could have done a better job of monitoring the CEO and his interaction with potential buyers, particularly after the special committee was in place. However, the court also found that those interactions did not compromise the integrity or effectiveness of the sale process. The second was that two potential buyers were excluded from the sale process. However, the judge found (ibid., at 19) that, when they contacted Solera’s financial adviser, there was no indication of their ability to pay and their introductory emails were perfunctory. They did not express any further interest either before or during the go-shop period and the judge concluded that they were just “kicking the tires.” These were therefore clearly very minor imperfections in the process and it is not surprising that the Chancellor still felt able to place sole weight upon the merger price. Apart from confirming that minor imperfections in the process will not necessarily prevent significant or exclusive weight being Page 35 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 36 of 257 placed on the merger price (with which I agree), I do not think this case assists Mr. Jones because the facts were so different; in this case the deficiencies in the market check and in the conduct of the special committee were very much more serious. 145 Whilst it is always important not to treat words in a judgment as if they were in a statute, I agree with the decisions in the Grand Court that the guidance in Dell (5) and DFC (4) as to the need for a market check and the circumstances in which a merger price may be relied upon is equally applicable in this jurisdiction. The possible qualification referred to at para. 138 above relates to the sentence where the judge says (Cause No. FSD 92 of 2017, at para. 156): “But I do not accept the Dissenting Shareholders’.”

¶ [53] At [146] (v)(a) and (b) Birt JA added: “(v) The judge held (ibid., at para. 163(g)) that the failures in the market check were not significant because it had not been shown that competitors were shut out as, if they had a real interest in making a bid, they would have known that the company was on the market and could therefore have contacted the company or Citi. Mr. Jones submitted that the judge was entitled to reach this view. I consider that the judge erred in this respect for two reasons: (a) It was inconsistent with the judge’s own self-direction (ibid., at para. 156) and with the Delaware jurisprudence which, correctly in my view, emphasizes the need for a robust market check and “outreach to all logical buyers.” No doubt it will usually be the case that the fact that a company is in play will be public knowledge and will be known to potential bidders. If that fact alone were held to be sufficient, there would be no need for a market check. But there is, in my judgment, a considerable difference between knowing that a company is for sale on the one hand and, on the other, being approached by the financial adviser to a company using all its skill to persuade a potential buyer that it would be advantageous for it to bid for the company. I would therefore disagree that the suggestion that Page 36 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 37 of 257 potential buyers could have come forward even though they were not on the list to be approached by Citi was a reason to ignore the deficiencies in the market check procedure. (b) The judge (ibid., at para. 163(g)) found that it had not been shown that the failure to include competitors in the market check resulted in such competitors being shut out from participating in the process or from coming forward to express their interest had they desired to do so. As Mr. Salzedo submitted, this was in effect to place the burden on the dissenting shareholders to show that competitors were prevented from coming forward. But this is something which dissenting shareholders will virtually never be in a position to prove. How can they have been expected to show what would have happened if the company had carried out the market check which it should have? As the judge correctly stated (ibid., at para. 156), the evidential burden is on the company to show that the conditions justifying reliance on the merger price are satisfied. Thus in this case it was for the company to show that the market check carried out by the special committee was adequate to ensure that all potential bidders were approached and given an opportunity to bid. The judge’s decision effectively to place the burden on the dissenting shareholders was inconsistent with what he had correctly stated (ibid.) and was, in my opinion, wrong in law. The whole point of requiring a robust market check is to avoid the need to speculate on what potential bidders would have done if they had been approached.”

¶ [54] In Trina Solar the buyer group held only 5.6% of the shares in the company at the time of the vote, the merger had been approved by 94.4% of the shareholders who were independent of the buyer group, including a large number of institutional investors and the market in the company’s shares was liquid, moved with announcements and was followed by analysts and the buyer group’s proposal was not timed at a particularly bad time for the company so as to take advantage of a particularly low share price (see

¶ [146] viii of the judgment). Birt JA at [146] (viii) of the judgment stated: Page 37 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 38 of 257 “(viii) It seems to me that, although the majority vote may not always be so large, these are all features which are likely to be present in many if not most cases. They do not diminish the need for a robust market check as required by the authorities, so as to maximize the prospects of a competing bid and minimize the chances (particularly in the case of a management buyout) that the company is being purchased on the cheap because of the greater knowledge of the buyers. Accordingly, I do not see that the existence of the factors listed by the judge can possibly balance or outweigh the deficiencies in the market check process and the concerns about the impact of Mr. Gao’s position.” Management projections

¶ [55] From [150] onwards Birt JA gave some helpful guidance on the issue of management projections: “(i) Test for departing from the management projections 150 The judge dealt with this topic in his judgment (Cause No. FSD 92 of 2017, at paras. 186–188). Having (ibid., at para. 186) noted certain observations of Parker, J. in In re Qunar Cayman Islands Ltd. (12) (2019 (1) CILR 611, at paras. 179–182) and of Kawaley, J. in Nord Anglia (9) (Cause No. FSD 235 of 2017, at paras. 150– 153), he summarized his approach as follows in the following terms (Cause No. FSD 92 of 2017, at para. 188): “188. Accordingly, while each case depends on its facts and there are no rigid rules to be mechanically applied, the approach to be followed can be summarised as follows: (a). the Court’s task is to determine the most realistic forecast of the Company’s future performance for the purpose of the DCF valuation. As I said in Shanda, the Court is aiming to find projections that are reasonable and reliable in the circumstances (a case in which I was prepared to require adjustments to be made to the management projections). Page 38 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 39 of 257 (b). the starting point is the general reliability test—have the management projections been prepared in good faith by a competent management team which understands the business and is capable of making informed judgments about future performance? (c). if the management projections satisfy this test, in the absence of some contemporaneous challenge to management’s projections suggesting some serious failure in management’s general approach to the preparation of projections, the Court must consider whether they have been shown to be obviously wrong, careless or tainted by an improper purpose (such as bias). There needs to be serious and material doubts as to the reasonableness and reliability of such projections. (d). this is because the Court should be cautious about allowing expert witnesses to substitute their judgments about the company’s future business prospects for the judgments of those with intimate inside business knowledge—although the level of caution will depend on and vary with the issue in question and challenges made. Where the disputed adjustments are primarily dependent on questions of insider business judgment then the Court will be slow to displace management’s assessment and forecasts. Where the issues in dispute involve questions of expert valuation methodology, then the Court may be less deferential to management’s approach. (e). where there are competing valuation methodologies, the Court may give reduced weight to a DCF valuation based on management projections if it has serious concerns as to the reasonableness and reliability of those projections (either generally or in material respects). (f). where the Court is considering particular items to be included in the DCF valuation (relating to a particular input in the DCF valuation based Page 39 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 40 of 257 on an item in or aspect of a forecast) the Court can limit the weight to be attached to the item derived from the management projections if there are serious and material doubts as to the reasonableness and reliability of that item or aspect. (g). management projections produced for the purpose of or in connection with the merger rather than being produced in the ordinary course of the company’s operations, will need to be carefully scrutinised to ensure that they have not been influenced by, and constructed so as to facilitate, the merger.” 151 Mr. Salzedo submitted that the observation in (c) above to the effect that the court may only intervene if the projections have been shown to be “obviously wrong, careless or tainted by an improper purpose (such as bias)” was inconsistent with the judge’s assertion at (a) above that the court’s task is “to determine the most realistic forecast of the Company’s future performance.” The judge should, therefore, have approached this issue on the basis that his task was to determine the best forecast, as at the valuation date, on the basis of all available evidence including, but not limited to, the management projections; and if, weighing all of that evidence, the best forecast was something other than as stated in the management projections, the judge should have adopted that alternative figure without requiring a demonstration that the management projections were “obviously wrong, careless or tainted by an improper purpose.” 152 Mr. Jones, on the other hand, emphasized that management projections have the advantage of being prepared at the relevant time by those with knowledge and experience of a company’s business. It was therefore correct for the judge to hold that, if the court is satisfied that the general reliability test which the judge articulated (ibid., at para. 188(b)) is satisfied, the evidential burden falls on the dissenting shareholders to show that the projections are obviously wrong, careless or tainted by improper purpose. Page 40 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 41 of 257 153 In my judgment, the judge articulated the test for departing from a forecast in management projections in terms which are too high, with its reference to only interfering if the forecast is shown to be “obviously wrong, careless or tainted by an improper purpose.” 154 I accept that the judge adopted this expression from the judgment of Parker, J. in Qunar (ibid., at para. 181). However, I do not consider that Parker, J. was seeking to lay down a general test which should invariably be applied; he was also considering the matter in a case where there were no industry expert witnesses. 155 I agree with the judge that the starting point is to consider the general reliability test as stated (ibid., at para. 188(b)). Clearly, if that is not satisfied, the court should be very cautious about placing too much weight on management projections. But, even if the general reliability test is satisfied, the court’s duty, as the judge correctly stated (ibid., at para. 188(a)) is to determine the most realistic forecast of the company’s future performance. That means that the court must reach its own decision pursuant to its obligation to determine fair value. 156 I find helpful an observation of Kawaley, J. from Nord Anglia (9) (Cause No. FSD 235 of 2017, at para. 153) as follows: “But, it bears repeating, this [satisfaction of the general reliability test] does not exclude the need in the context of a judicial appraisal proceeding to test the accuracy of the estimates and make appropriate adjustments with a view to determining the most realistic forecast of the Company’s future performance.” 157 There must, of course, be some evidence before the court to raise an issue as to the appropriateness of one or more assumptions or forecasts in the management projections before any question of departing from such projections arises. But once there is such evidence, the court’s duty is to reach its own decision. In doing so, it will of course give due weight to the knowledge and experience which the management team has of the company’s business, but it must consider the evidence of both parties and reach its own decision on the most realistic forecast. It is wrong Page 41 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 42 of 257 to say that the court can only vary a forecast in management projections if it concludes that that forecast is obviously wrong, careless or tainted by an improper purpose. The court may simply conclude that, even after making allowance for the knowledge and expertise of the management team and bearing in mind the cautionary words quoted above about allowing expert witnesses to substitute their views for those of management, an expert witness’s views are to be preferred and that the best forecast is that put forward by the expert witness or lies somewhere between the management projections and that of the expert witness. If the court reaches this conclusion, it must give effect to that conclusion when assessing fair value. 158 Apart from sub-para. (c) of para. 188, the dissenting shareholders did not criticize the judge’s summary in para. 188 and, for my part, I consider it to be a useful summary of the court’s approach.” Trina Solar at first instance

¶ [56] The JCPC neatly summarised the decision of Segal J at [22] to [32] of Sir Andrew Popplewell’s judgment. Shanda Games

¶ [57] Lady Arden in Shanda Games (JCPC) at [49] stated that the Cayman Islands courts could “use the jurisprudence of the Delaware and other courts if placed before them as a source of guidance on some particular issues” and “the jurisprudence of Delaware is of great value in this field.” Lady Arden was however at pains to suggest that the courts should bear in mind that “there may be different rules in related contexts” and “there may be different economic and social policy considerations affecting legislation in Delaware.”

¶ [58] At [50] it was noted that the JCPC took the view that “comparative law whose subject matter is similar to s.238 may provide to varying degrees a useful comparison, but it would be wrong to fail fully to Page 42 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 43 of 257 recognise that the Cayman Islands legislative made the decision in enacting s.238 to use a new and undefined phrase … namely “fair value”.”

¶ [59] Parker J also put it well, if I may say so, in FGL at [264] (footnote omitted): “Delaware jurisprudence is helpful to look to when solving problems perhaps new to Cayman law and practice, but it is not Cayman law and is not to be preferred to Cayman law and practice where it differs.”

¶ [60] Sir Andrew Popplewell in Trina Solar (JCPC) at [8] noted that “The Cayman Islands courts have followed the practice and jurisprudence in Delaware in applying both market based and DCF valuation methodologies when assessing fair value.” At [9] Sir Andrew Popplewell added: “The approach of the Delaware Courts can be of assistance and value in this area because the Companies Act was based partly on legislation in Delaware (which has similarities as well as differences), amongst other jurisdictions; and Delaware has a longer history of applying its law and therefore a more developed jurisprudence: see Shanda, at para 49, and the judgment of Martin JA in the Court of Appeal in that case at para 46 cited at para 21.”

¶ [61] With those comments firmly in mind I now turn to the Delaware jurisprudence. The Delaware jurisprudence Cammer

¶ [62] In Cammer v Bloom, 711 F.Supp.1264, 1287 (D.N.J. 1989) (“Cammer”) the United States District Court for the District of New Jersey (Alfred J Lechner Jr United States District Judge) referred under the heading “The fraud on the market theory” to the oft-used quote from the Supreme Court of the United States of America in Basic v Levinson, 485, U.S. 224: “The fraud on the market theory is based on the hypothesis that, in an open and developed securities market, the price of a company’s stock is determined by the available material information regarding the company and its business.” Page 43 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 44 of 257

¶ [63] The judge also at page 8 referred to definitions offered in Bromberg & Lowenfels, 4 Securities Fraud and Commodities Fraud at 8.6: “An open market is one in which anyone, or at least a large number of persons, can buy or sell. A developed market is one which has a relatively high level of activity and frequency, and for trading information (e.g., price and value) is widely available. It is principally a secondary market in outstanding securities. It usually, but not necessarily, has continuity and liquidity (the ability to absorb a reasonable amount of trading with relatively small price changes). An efficient market is one which rapidly reflects new information in price. These terms are cumulative in the sense that a developed market will always be an open one. And an efficient market will almost invariably be a developed one.”

¶ [64] At page 16 the judge stated that “the existence of Form S-3 status is an important factor weighing in favor of a finding that a market is efficient.”

¶ [65] At page 17 the judge referred to average weekly trading volumes and stated: “The reason the existence of an actively traded market, as evidenced by a large weekly volume of stock trades, suggests there is an efficient market is because it implies ... a likelihood that many investors are executing trades on the basis of newly available or disseminated corporate information.”

¶ [66] At page 18 the judge added: “… It would be persuasive to allege a significant number of securities analysts followed and reported on a company’s stock during the class period. The existence of such analysis would imply, for example, the PMM reports were closely reviewed by investment professionals, who would in turn make buy/sell recommendations to client investors …”. Page 44 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 45 of 257

¶ [67] At page 18 the judge stated: “… it would be helpful to allege the Company was entitled to file an S-3 Registration Statement in connection with public offerings …. Again, it is the number of shares traded and value of shares outstanding that involve facts which imply efficiency.”

¶ [68] At page 18 the judge stated: “… it could be alleged the stock had numerous market makers. The existences of market makers and arbitrageurs would ensure completion of the market mechanism; these individuals would react swiftly to company news and reported financial results by buying or selling stock and driving it to a changed price level.”

¶ [69] The judge continued: “Finally, it would be helpful to a plaintiff seeking to allege an efficient market to allege empirical facts showing a cause and effect relationship between unexpected corporate events or financial releases and an immediate response in the stock price. This, after all, is the essence of an efficient market and the foundation for the fraud on the market theory.” DFC

¶ [70] In DFC Global Corporation v Muirfield Value Partners, L.P No 518, 2016 (Del. Supr. August 1, 2017) (“DFC”) the Supreme Court of Delaware declined to establish a presumption that in certain cases involving arm’s-length mergers the price of the transaction giving rise to appraisal rights is the best estimate of fair value. The Supreme Court of Delaware noted that the respondent had made two convincing case-specific arguments as to why the Court of Chancery’s determination of fair value could not be sustained on appeal. For starters the respondent noted that the Court of Chancery found that (i) the transaction resulted from a robust market search that lasted approximately two years, in which financial and strategic buyers had an open opportunity to buy without inhibition of deal protections; (ii) the company was purchased by a third party in an arm’s length sale; and (iii) there was no hint of self- interest that compromised the market check. Page 45 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 46 of 257

¶ [71] Strine CJ at page 2 stated: “Although there is no presumption in favor of the deal price, under the conditions found by the Court of Chancery, economic principles suggest that the best evidence of fair value was the deal price, as it resulted from an open process, informed by robust public information, and easy access to deeper, non-public information, in which many parties with an incentive to make a profit had a chance to bid. But despite its own findings about the adequacy of the market check, the Court of Chancery determined it would not give more than one-third weight to the deal price for two reasons.”

¶ [72] The first reason was that there were regulatory developments making the market’s assessment of the company’s value not as reliable as under ordinary conditions. The Supreme Court of Delaware agreed that such finding was not rationally supported by the record. The Supreme Court of Delaware stated that they were “unaware of, any academic or empirical basis to conclude that market players like the many who were focused on this company’s value would not have examined the potential for regulatory action and factored it in their assessment of the company’s value. Like any factor relevant to a company’s future performance, the market’s collective judgment of the effect of regulatory risk may turn out to be wrong, but established corporate finance theories suggest that the collective judgment of the many is more likely to be accurate than any individual’s guess …”.

¶ [73] The Supreme Court of Delaware also criticised the Court of Chancery’s decision in respect of the DCF analysis and referred at pages 5-6 to the “lack of any basis to conclude that the company would sustain high growth beyond the projection period, the record does not sustain the Court of Chancery’s decision to substantially increase the company’s perpetuity growth rate in its discounted cash flow model after reargument.”

¶ [74] The Supreme Court of Delaware at page 6 added that “the Court of Chancery’s decision to give one- third weight each to the deal price, the discounted cash flow valuation, and the comparable companies valuation was not explained.” The Supreme Court of Delaware reversed and remanded the Chancery Court’s ruling and at page 7 suggested that the Chancellor might conclude that “the deal price was the most reliable indication of fair value.” Page 46 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 47 of 257

¶ [75] The Supreme Court of Delaware referred to DFC’s growth as the provider of “alternative consumer financial services, predominantly payday loans”. From page 12 onwards reference was made to the regulatory headwinds, from page 14 onwards to the sales process, from page 19 onwards to the appraisal trial, from page 27 onwards to reargument post-trial decision. From page 31 onwards the Supreme Court of Delaware dealt with arguments in respect of a judicial presumption in favour of the deal price.

¶ [76] The Chief Justice at page 45 stated: “Market prices are typically viewed superior to other valuation techniques because, unlike, e.g., a single person’s discounted cash flow model, the market price should distill the collective judgment of the many based on all the publicly available information about a given company and the value of its shares. Indeed, the relationship between market valuation and fundamental valuation has been strong historically … But, a single person’s own estimates of the cash flows are just that, a good faith estimate by a single, reasonably informed person to predict the future. Thus, a singular discounted cash flow model is often most helpful when there isn’t an observable market price.”

¶ [77] At page 47 the Chief Justice stated: “Other realities emphasize why real world transaction prices can be the most probative evidence of fair value even through appraisal’s particular lens. As the preceding discussion emphasizes, fair value is just that, “fair”. It does not mean the highest possible price that a company might have sold for had Warren Buffett negotiated for it on his best day and the Lenape who sold Manhattan at their worst” … Capitalism is rough and ready, and the purpose of an appraisal is not to make sure that the petitioners get the highest conceivable value that might have been procured had every domino fallen out of the company’s way; rather, it is to make sure that they receive fair compensation for their shares in the sense that it reflects what they deserve to receive based on what would fairly be given to them in an arm’s-length transaction.”

¶ [78] In section D the Supreme Court of Delaware deal with issues in respect of the discounted cash flow valuation. Page 47 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 48 of 257

¶ [79] In section E the Supreme Court of Delaware deal with issues in respect of the comparable companies analysis.

¶ [80] In section F at pages 83-85 the Supreme Court of Delaware stated (footnote omitted): “… When faced with briefs and expert reports written by highly-skilled litigators in concert with men and women of valuation science that often come to ridiculously varying positions, the Court of Chancery may well feel tempted to turn its valuation decisions into a more improvisational variation of the old Delaware Block Method, but one in which the court takes every valuation method put in the record, gives each equal weight, and then divides by the number of them. When life is sloppy and unpredictable, the visual appeal of a mathematical formula to create an impression of precision is understandable. But, in keeping with our refusal to establish a “presumption” in favor of the deal price because of the statute’s broad mandate, we also conclude that the Court of Chancery must exercise its considerable discretion while also explaining, with reference to the economic facts before it and corporate finance principles, why it is according a certain weight to a certain indicator of value. In some cases, it may be that a single valuation metric is the most reliable evidence of fair value and that giving weight to another factor will do nothing but distort that best estimate. In other cases, it may be necessary to consider two or more factors. As one appraisal treatise points out, “laying down in advance fixed rules that state how competing approaches are to be weighted is impossible.” What is necessary in any particular case though is for the Court of Chancery to explain its weighting in a manner that is grounded in the record before it. That did not happen here. In this case, the decision to give one-third weight to each metric was unexplained and in tension with the Court of Chancery’s own findings about the robustness of the market check.” Dell

¶ [81] In Dell Inc (14 December 2017) (“Dell”) the Supreme Court of Delaware provided some useful guidance in an American “long-running appraisal saga” in respect of an appeal from decisions made after a four-day appraisal trial in October 2015. Justice Valihura at page 1 stated: Page 48 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 49 of 257 “… The appraisal statute requires the Court of Chancery to assess the “fair value” of such shares and, in doing so, “take into account all relevant factors.” The trial court complied: it took into account all the relevant factors presented by the parties in advocating for their view of fair value – including Dell’s stock price and deal price – and then arrived at its own determination of fair value.”

¶ [82] The trial court had decided to give the “market data” no weight and relied exclusively on its own DCF to reach a fair value calculation which the Supreme Court of Delaware held did “not follow from the court’s key factual finding and from relevant, accepted financial principles.”

¶ [83] The trial court gave no weight to Dell’s stock price because it found its market to be inefficient. The Supreme Court of Delaware held that the evidence suggested that the market for Dell’s shares was actually efficient and, therefore, likely a possible proxy for fair value.

¶ [84] The trial court concluded that several features of management led buyout (“MBO”) transactions rendered the deal prices resulting from such transactions unreliable. The Supreme Court of Delaware held that such features were “largely absent”.

¶ [85] Dell’s stock traded on the NASDAQ with a market capitalisation of more than $20 billion and was actively traded as more than 5% of Dell’s shares were traded each week. It was also widely covered by equity analysts, and its share price reflected the market’s view on breaking developments. Justice Valihura at page 6 stated: “… the record suggests the market for Dell stock was semi-strong efficient, meaning that the market’s digestion and assessment of all publicly available information concerning Dell was quickly impounded into the Company’s stock price.”

¶ [86] The Supreme Court of Delaware at page 35 acknowledged that the trial of an appraisal case “presents unique challenges to the judicial factfinder” and that the task is complicated by the clash of contrary, and often antagonistic expert opinions of value prompting the trial court to wade through “widely divergent views reflecting partisan positions” in arriving at its determinations of a single number for fair value. Page 49 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 50 of 257

¶ [87] From page 37 onwards the Supreme Court of Delaware commented that in the end after the analyses of the relevant factors “[i]n some cases, it may be that a single valuation metric is the most reliable evidence of fair value and that giving weight to another factor will do nothing but distort the best estimate. In other cases, it may be necessary to consider two or more factors.” Footnote 104 of Dell refers to DFC, 2017 WL 3261190 at 31. See also MG Bancorp 737 A.2d (“Bancorp”) at 525-26 “[T]he Court of Chancery has the discretion to select one of the parties’ valuation models as its general framework or to fashion its own.” And “[A]lthough not required to do so, it is entirely proper for the Court of Chancery to adopt any one expert’s model, methodology, and mathematical calculations, in toto, if that valuation is supported by credible evidence and withstands critical judicial analysis on the record.”

¶ [88] In other cases the court might apportion weight amongst a variety of methodologies. But, whatever route it chooses, the trial court must justify its methodology (or methodologies) according to the facts of the case and relevant, accepted financial principles.

¶ [89] At page 40 the Supreme Court of Delaware stated that the trial court had “ignored the efficient market hypothesis long endorsed by this Court. It teaches that the price produced by an efficient market is generally a more reliable assessment of fair value than the view of a single analyst, especially an expert witness who caters her valuation to the litigation imperatives of a well-heeled client.” The Supreme Court of Delaware added (footnotes omitted) at pages 40-41: “A market is more likely efficient or semi-strong efficient, if it has many stockholders; no controlling stockholder; “highly active trading”; and if information about the company is widely available and easily disseminated to the market. In such circumstances, a company’s stock price “reflects the judgments of many stockholders about the company’s future prospects, based on public filings, industry information, and research conducted by equity analysts.” In these circumstances, a mass of investors quickly digests all publicly available information about a company, and in trading the company’s stock, recalibrates its price to reflect the market’s adjusted, consensus valuation of the company.”

¶ [90] At page 46 the Supreme Court of Delaware referred to various objective factors in DFC as including “every logical buyer” being canvassed. Page 50 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 51 of 257

¶ [91] At page 49 the Supreme Court of Delaware stated: “The Court of Chancery ignored an important reality: if a company is one that no strategic buyer is interested in buying it does not suggest a higher value, but a lower one.”

¶ [92] In one of the two passages from Dell referred to by Sir Andrew Popplewell in Trina Solar (JCPC) at

¶ [11] the Supreme Court of Delaware in Dell at pages 60-61 stated: “… There is no requirement that a company prove that the sales process is the most reliable evidence of its going concern value in order for the resulting deal price to be granted any weight. If, as here, the reasoning behind the decision to assign no weight to market data is flawed, then the ultimate conclusion necessarily crumbles as well – especially in light of the less-than-surefire DCF analysis – as demonstrated below. In so holding, we are not saying that the market is always the best indicator of value, or that it should always be granted some weight. We only note that, when the evidence of market efficiency, fair play, low barriers to entry, outreach to all logical buyers, and the chance for any topping bidder to have the support of Mr Dell’s own rates is so compelling, then failure to give the resulting price heavy weight because the trial judge believes there was mispricing by all the Dell stockholders, analysts, and potential buyer abuses even the wide discretion afforded the Court of Chancery in these difficult cases. And, of course, to give no weight to the prices resulting from the actions of Dell’s stockholders and potential buyers presupposes that there is a more plausible basis for determining Dell’s value in the form of expert testimony, such as from the petitioners’ expert, who argued that his DCF analysis showed the fair value of Dell’s stock is $28.61 per share – almost three times higher than the unaffected stock price of $9.96 per share and more than two times higher than the deal price of $13.75 per share.”

¶ [93] At page 61 the Supreme Court of Delaware in a DCF section dealing with analysis commented that where there is a robust sales process the trial court “should be chary about imposing the hazards that always come when a law-trained judge is forced to make a point of estimate of fair value based on widely divergent partisan expert testimony.” The Supreme Court of Delaware commented: Page 51 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 52 of 257 “As is common in appraisal proceedings, [footnote 180 see e.g. DFC, 2017 WL 3261190 at 31 describing “briefs and expert reports written by highly-skilled litigators in concert with men and women of valuation science that often come to ridiculously varying positions”] each party – the petitioners and the Company – enlisted highly paid, well- credentialed experts to produce DCF valuations. But their valuations landed galaxies apart.”

¶ [94] At pages 64-65 the Supreme Court of Delaware (in a passage partially referred to by the JCPC in Trina Solar as referring to the “inherent hazards in using a DCF methodology”) stated: “When an asset has few, or no, buyers at the price selected, that is not a sign that the asset is stronger than believed – it is a sign that it is weaker. This fact should give pause to law- trained judges who might attempt to outguess all of these interested economic players with an actual stake in a company’s future … Given that we have concluded that the trial court’s key reasons for disregarding the market data were erroneous, and given the obvious lack of credibility of the petitioner’s DCF model – as well as legitimate questions about the reliability of the projections upon which all of the various DCF analyses are based – these factors suggest strong reliance upon the deal price and far less weight, if any, on the DCF analyses … Although widely considered the best tool for valuing companies when there is no credible market information and no market check, DCF valuations involve many inputs – all subject to disagreement by well-compensated and highly credentialed experts – and even slight differences of these inputs can produce larger valuation gaps … Here … the experts’ fair value determinations … landed on different planets.”

¶ [95] The Supreme Court of Delaware made reference to alternatives to “gambling on an appraisal’s battle of the experts.” Solera

¶ [96] Solera Holdings Inc C.A. No 12080-CB (Del. Ch. July 30, 2018) (“Solera”) is an example of the treatment of Dell and DFC by the Court of Chancery of the State of Delaware and the judgment also helpfully refers to some other authorities of relevance. Page 52 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 53 of 257

¶ [97] Bouchard C at page 1 noted that “Over the past year, our Supreme Court has heavily endorsed the application of market efficiency principles in appraisal actions”. Bouchard C rejected reliance on a DCF analysis and gave his “independent determination” of fair value based on “the deal price less synergies”. He added that: “… the record reflects that Solera was sold in an open process that, although not perfect, was characterized by many objective indicia of reliability.”

¶ [98] The Chancellor referred to outreach and an auction, public disclosures which made clear to the market that the company was for sale, a special committee that had competent legal and financial advisors and a go-shop adding “the sales process was conducted against the backdrop of an efficient and well- functioning market for Solera’s stock … its shares were actively traded on the New York Stock Exchange and were covered by numerous analysts, and its debt was closely monitored by rating agencies … the sales process delivered for Solera stockholders the value obtainable in a bona fide arm’s-length transaction and provides the most reliable evidence of fair value. Accordingly, I give the deal price, after adjusting for synergies in accordance with longstanding precedent, sole and dispositive weight in determining the fair value of petitioners’ shares as of the date of the merger”.

¶ [99] In outlining the background the Chancellor referred to the five-day trial at first instance. From page 32 onwards the Chancellor analysed the relevant legal standard in Delaware and referred at page 33 to the court when conducting a valuation of a company must take into account the operative reality of the company viewing the company as occupying a particular market position in the light of future prospects. The Chancellor stated that the court has a significant discretion to use the valuation methods it deems appropriate, including the parties proposed valuation frameworks or one of the court’s own fashioning. The Chancellor made reference to an authority to the effect that the court may not adopt at the outset an “either-or” approach, thereby accepting uncritically the valuation of one party, as it is the court’s duty to determine the core issue of fair value on the relevant date. The Chancellor conducted “a brief discussion” of “two [Supreme Court of Delaware] decisions providing important guidance for the Court of Chancery” namely DFC and Dell. The Chancellor emphasised that in DFC the Supreme Court of Delaware cautioned that the market price is not always right “but that one should have little confidence she can be the special one able to outwit the larger universe of equally avid capitalists with an incentive to reap rewards by buying the asset if it is too cheaply priced.” Page 53 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 54 of 257

¶ [100] The Chancellor also referred to Dell and stated at page 38: “Given leaks in the press that Dell was exploring a sale, moreover, the world was put on notice of the possibility of a transaction so that “any interested parties would have approached the Company before the go-shop if serious about pursuing a deal”.”

¶ [101] The Chancellor referred to and commented upon the Court of Chancery’s decisions in Verition Partners Master Fund Ltd v Aruba Networks Inc 2018 WL 922139 (Del. Ch. Feb 15, 2018) (“Aruba”) and In re Appraisal of AOL Inc 2018 WL 1037450 (Del. Ch. Feb. 23, 2018).

¶ [102] In Aruba the court observed that the Supreme Court of Delaware’s decisions in DFC and Dell “caution against relying on discounted cash flow analyses prepared by adversarial experts when reliable market indicators are available”. The court in Aruba determined that fair value was the “unaffected market price” of the Petitioners’ shares which was more than 30% below the transaction price.

¶ [103] The Chancellor in Solera said (at page 41) that the court in AOL similarly construed DFC and Dell to mean that where the “transaction price represents an unhindered, informed, and competitive market valuation, the judge must give particular and serious consideration to transaction price as evidence of fair value.” In doing so the court in AOL coined the phrase “Dell Compliant” to mean a transaction “where (i) information was sufficiently disseminated to potential bidders, so that (ii) an informed sale could take place, (iii) without undue impediments imposed by the deal structure itself.”

¶ [104] In Solera the petitioners, using a discounted cash flow model prepared by their expert, contended that the fair value of their shares was $84.65 per share – approximately 51.6% over the deal price. The respondent’s expert concluded that the “best evidence of Solera’s value is the market-generated Merger price [$55.85], adjusted for synergies [$1.90] to $53.95.”

¶ [105] The Chancellor in Solera from page 44 onwards turned to his own independent determination of the fair value in Solera’s shares “with the guidance from DFC and Dell in mind”. The Chancellor helpfully continued in a passage cited by Sir Andrew Popplewell in Trina Solar in the JCPC: (footnotes omitted) “Those decisions teach that deal price is “the best evidence of fair value” when there was an “open process,” meaning that the process is characterized by “objective indicia of Page 54 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 55 of 257 reliability. Such “indicia” include but, consistent with the mandate of the appraisal statute to consider “all relevant factors”, are not limited to: • “[R]obust public information,” comprised of the stock price of a company with “a deep base of public shareholders, and highly active trading,” and the views of “equity analysts, equity buyers, debt analysts, debt providers and others.” • “[E]asy access to deeper, non-public information,” where there is no discrimination between potential buyers and cooperation from management helps address any information asymmetries between potential buyers. • “[M]any parties with an incentive to make a profit had a chance to bid, meaning that there was a “robust market check” with “outreach to all logical buyers” and a go-shop characterized by “low barriers to entry” such that there is a realistic possibility of a topping bid. • A special committee, “composed of independent, experienced directors and armed with that power to say ‘no,’” which is advised by competent legal and financial advisors. • “[N]o conflicts related to the transaction,” with the company purchased by a third party in an arm’s length sale and “no hint of self-interest.” If the process was open, then “the deal price deserve[s] heavy, if not dispositive weight.” This is not to say that the market is always correct: “In some cases, it may be that a single valuation metric is the most reliable evidence of fair value and that giving weight to another factor will do nothing but distort that best estimate. In other cases, it may be necessary to consider two or more factors.” Whichever route it takes, however, the Court of Chancery is required to “justify its methodology (or methodologies) according to the facts of the case and relevant, accepted financial principles.”

¶ [1] The Deal Price Less Synergies Deserves Dispositive Weight Page 55 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 56 of 257 For the reasons explained below, I find that the Merger was the product of an open process that, although not perfect, has the requisite objective indicia of reliability emphasized in DFC and Dell. Thus, I conclude that the deal price, minus synergies, is the best evidence of fair value and deserves dispositive weight in this case. My consideration of the evidence supporting this conclusion follows in three parts focusing on (i) the opportunity many potential buyers had to bid, (ii) the Special Committee’s role in actively negotiating an arm’s-length transaction, and (iii) the evidence that the market for Solera’s stock was efficient and well-functioning.”

¶ [106] The Chancellor at page 48 referred to various potential bidders being directly contacted and “the whole universe of potential bidders was put on notice”. (Footnote 247 In re Appraisal of PetSmart, 2017 WL 2303599 at *28 (Del. Ch. May 26, 2017); see also Tr. 1036 (Hubbard) (“Once a sales process became public in the Bloomberg story, anyone who wished to bid on this asset could certainly have jumped in.”); Dell, 177 A.3d at 28 (“[G]iven leaks that Dell was exploring strategic alternatives, record testimony suggests that [Dell’s banker] presumed that any interested parties would have approached the company before the go-shop if serious about pursuing a deal”). At page 49 the Chancellor referred to “public disclosures” and the fact that any potential bidder knew that Solera was exploring a sale “Yet no one else ever seriously showed up to make a topping bid.” In Solera the record reflected that “the Company provided all seriously interested bidders access to deeper, non-public information after they signed non-disclosure agreements” (page 51). In Solera the record demonstrated that the merger price resulted from an open process informed by robust public information, and easy access to deeper, non- public information, in which many parties with an incentive to make a profit had a chance to bid. The Chancellor concluded at page 55: “Thus, consistent with our high court’s recent teachings, economic principles suggest that the Merger price is what petitioners “deserve to receive” for their shares.”

¶ [107] The Chancellor at page 55 added: “Reliance on the deal price as evidence of fair value is strengthened when independent representatives of a target company actively negotiate with potential buyers and demonstrates a real willingness to reject inadequate bids.” Page 56 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 57 of 257

¶ [108] The Chancellor at page 63 stated: “… petitioners contend that MBOs should be subject to “heightened security” but fail to explain why. As the Supreme Court stated in Dell, even though there may be “theoretical characteristics” of an MBO that could “detract [ ] from the reliability of the deal price”, the deal price that results from an MBO is not inherently suspect or unreliable per se …”. Insofar as such comment goes against the comments of Birt JA in Trina Solar (CICA) on MBOs and heightened scrutiny I ignore it and follow the CICA’s existing guidance, which on this point was left untouched by the JCPC.

¶ [109] From page 75 onwards the Chancellor dealt with the DCF models and cited Union Ill., 847 A.2d at 359 and stated that compared with a market-generated transaction price: “The use of alternative valuation techniques like a DCF analysis is necessarily a second- best method to derive value.”

¶ [110] At page 77 the Chancellor cited PetSmart, 2017 WL 2303599 at *40 n 439: “Delaware courts must remain mindful that the DCF method is [ ] subject to manipulation and guesswork [and that] the valuation results that it generates in the setting of a litigation [can be] volatile.”

¶ [111] The Chancellor felt that the petitioner’s DCF analysis was “facially unbelievable as it suggests that, in a transaction with an equity value of approximately $3.85 billion at the deal price, potential buyers left almost $2 billion on the table by not outbidding Vista”.

¶ [112] Citing from Dell at footnote 373 the Chancellor added that even slight differences in a DCF’s inputs can produce large valuation gaps.

¶ [113] The Chancellor commented at page 77 that: Page 57 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 58 of 257 “A number of factors explain the gaping difference between the petitioners’ and respondent’s DCF analyses, and, notably, many of these disagreements relate to how to value Solera into perpetuity. Such assumptions about Solera’s business in the terminal period, i.e. ten years plus into the future, are unavoidably tinged with a heavy dose of speculation.”

¶ [114] The Chancellor referred to the major areas of disagreement between the parties: “All of these disagreements predictably result in a higher asserted valuation by petitioners and a lower asserted valuation by respondents”.

¶ [115] The Chancellor’s reference at page 77 to certain assumptions of the DCF as being “unavoidably tinged with a heavy dose of speculation” and at page 80 to a candid response from the petitioner’s expert as to how he landed at a particular percentage as “It’s a little bit of a finger in the wind”, reminded me of Mr Taylor’s understatements that there are degrees of speculation in the DCF process.

¶ [116] At pages 82-83 the Chancellor concluded: “Given the huge gap between petitioners’ DCF valuation and the Merger price, which I have found to be a reliable indicator of value in accordance with the teachings of DFC and Dell, I find Petitioners’ DCF valuation not to be credible on its face and accord it no weight … I also give no weight to respondent’s DCF valuation …”.

¶ [117] At page 89 the Chancellor referred to DFC and Dell being “transformative decisions” and “their full- throated endorsement of applying market efficiency principles in appraisal actions” Stillwater

¶ [118] In Stillwater Mining Company CA No 2017 – 0385 – JTL (Del. Ch. August 21, 2019) Laster VC also considered Dell and DFC. In Stillwater (another four day trial) the petitioners contended a fair value of $25.91 per share relying on a DCF model. The other side (Sibanye) contended for a fair value of $17.63 per share relying on a combination of metrics including the deal price, the unaffected trading price with an adjustment and a DCF model. Sibanye proved that the sales process was sufficiently reliable to make the deal price a persuasive indicator of fair value. The parties engaged in a lengthy Page 58 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 59 of 257 debate over whether Stillwater’s adjusted trading price could provide a persuasive indicator of fair value. The reliability of the adjusted trading price depended on the reliability of the unaffected trading price. The Vice-Chancellor at page 2 illuminatingly stated: “Neither side proved that its DCF valuation provided a persuasive indicator of fair value. The experts disagreed over many inputs, and the resulting valuation swings were too great, for this decision to rely on a model when a market-tested indicator is available.”

¶ [119] At section L of his judgment the Vice-Chancellor refers to Sibanye revising its price downwards from $18.25 to $15.75 per share.

¶ [120] Under section II the Vice-Chancellor conducted a legal analysis and at page 40 quoted from M.G. Bancorp., 737 A.2d at 525-26: “In discharging its statutory mandate, the Court of Chancery has discretion to select one of the parties’ valuation models as its general framework to fashion its own.” “[I]t is entirely proper for the Court of Chancery to adopt any one expert’s model, methodology, and mathematical calculations in toto, if that valuation is supported by credible evidence and withstands a critical judicial analysis on the record.”

¶ [121] The Vice-Chancellor continued at page 40-41: “Or the court “may evaluate the valuations opinions submitted by the parties, select the most representative analysis, and then make appropriate adjustments to the resulting valuation” … The court may also “make its own independent valuation calculation by … adopting or blending the factual assumptions of the parties’ experts… But the court must also be cautious when adopting an approach that deviates from the parties’ positions …”

¶ [122] From page 45 onwards the Vice-Chancellor considered the deal price of $18.00 per share and the standard for evaluating a sale process noting that Dell confirmed that there was no presumption that the deal price reflects fair value adding at page 46: Page 59 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 60 of 257 “As a general matter, the persuasiveness of the deal price depends on the reliability of the sales process that generated it …” [T]he key inquiry is whether the dissenters got fair value and were not exploited.” Dell, 177 A.130 at 33.

¶ [123] The Vice-Chancellor at page 47 disagreed with the petitioners that the deal price deserves weight only if the merger is the product of a “robust market search” and an arm’s length third party transaction with no hint of self-interest that compromised the market check adding “That is not what DFC held.” The Vice-Chancellor stated: “The decision did not identify minimum characteristics that a sale process must have before a trial court can give it weight. The decision also did not address what makes a sale process sufficiently bad that a trial court cannot give it weight … The same is true for the Delaware Supreme Court’s comments about the sale process in Dell… The Delaware Supreme Court was not identifying the minimum requirements for a sale process to generate reliable information about fair value, nor was it enumerating qualities which, if absent, would render the outcome of a sale process so unreliable as to provide no insight into fair value … The decisions in DCF, Dell and Aruba are highly informative because they analyze fact patterns in which the Delaware Supreme Court viewed the sales process as sufficiently reliable to use the deal price as either (i) the exclusive basis for its own fair value determination (Aruba), (ii) as a valuation indicator that “deserved heavy, if not dispositive weight” (Dell), or (iii) as a valuation indicator that provided “the best evidence of fair value” (DFC). But Aruba, Dell, and DFC do not establish legal requirements for a sales process. Whether a sale process is sufficiently good that the deal price should be regarded as persuasive evidence of fair value, or whether a sale process is sufficiently bad that the deal price should not be regarded as persuasive evidence of fair value are invariably fact- specific questions, and the answers depend on the arguments made and the evidence presented in a given case.”

¶ [124] From page 49 onwards the Vice-Chancellor referred to the objective indica of reliability of the sales process as including in Stillwater: Page 60 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 61 of 257 (1) the merger was an arm’s-length transaction with a third party. It was not a transaction involving a controlling stakeholder; (2) the Board dbd not labour under any conflicts of interest; (3) Sibanye conducted due diligence and received confidential information about Stillwater’s value; (4) Stillwater negotiated with Sibanye and extracted multiple price increases; (5) No bidders emerged during the post-signing phase, which is a factor that the Delaware Supreme Court of Delaware has stressed when evaluating a sale process. The Vice-Chancellor felt that these objective indicia “provide a cogent foundation for relying on the deal price as a persuasive indicator of fair value, subject to further view of the evidence.” (page 52)

¶ [125] The Vice-Chancellor at page 100 referred to the sales process not being “perfect” and to the petitioners highlighting “its flaws” but nevertheless concluded that the “sales process made the deal price a persuasive indicator of fair value”: “The sales process that led to the Merger bore objective indicia of fairness that rendered the deal price a reliable indicator of fair value …”.

¶ [126] In Section 6 from page 102 onwards the Vice-Chancellor dealt with the adjustment for value arising from the merger pursuant to section 262 of the Delaware General Corporation Law.

¶ [127] In Section 7 from page 104 onwards the Vice-Chancellor dealt with the adjustment for changes in value between signing and closing pursuant to section 262 of the Delaware General Corporation Law.

¶ [128] At page 115 the Vice-Chancellor dealt with the adjusted trading price as a reliable indicator of the fair value of a company: Page 61 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 62 of 257 “Sibanye generates the adjusted trading price by making adjustments to the unaffected trading price, so the reliability of the adjusted trading price depends on the reliability of the unaffected trading price. As the proponent of using this valuation indicator, Sibanye bore the burden of establishing its reliability and persuasiveness … Assessing the reliability of the trading price for Stillwater’s common stock means getting “deep into the weeds of economics and corporate finance” … The thicket of market efficiency is one such place where “law-trained judges should not go without the guidance of experts trained in these disciplines” … In this case, both sides retained financial experts who tried to lead the court through the undergrowth …”.

¶ [129] From page 116 onwards the Vice-Chancellor dealt with informational efficiency and fundamental value efficiency.

¶ [130] At page 117 the Vice-Chancellor stated: “Whether called fundamental value, true value, intrinsic value, or fair value, the really-real value of something is always an unobservable concept. No valuation methodology provides direct access to it. Fundamental value is like a Platonic form, and the various valuation methodologies only cutouts casting shadows on the wall of the cave. The real issue is not whether a particular method generates a shadow (they all do), but rather whether the shadow is more or less distinct than what other methods produce.”

¶ [131] At pages 117-118 the Vice-Chancellor added that for the purposes of determining fair value in an appraisal proceeding “….the trading price has a lot going for it. Like democracy, the trading price may be imperfect, but it will often serve better than the other metrics that have been tried … The petitioners’ admittedly valid objection that it is impossible to prove that a trading price reflects fundamental value is thus not one that automatically disqualifies the use of the trading price as a valuation indicator in an appraisal.”

¶ [132] On page 120 the Vice-Chancellor added: Page 62 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 63 of 257 “A market that evidences directionality and proportionality is more likely to be fundamental-value efficient. A market that lacks evidence of directionality and proportionality is less likely to be fundamental-value efficient.”

¶ [133] From page 134 onwards the Vice-Chancellor deals with the DCF models and acknowledges that the “DCF method is a technique that is generally accepted in the financial community” adding at page 135: “In Dell and DCF, the Delaware Supreme Court cautioned against the DCF methodology when market-based indicators are available.”

¶ [134] At page 135 onwards the Vice-Chancellor referred to the disagreement of the experts in respect of the DCF model: “The experts disagreed over many inputs, with small changes producing large swings in value. The briefing focused on eight inputs, with four generating the bulk of the difference. First, the experts debated whether to apply a small-company risk premium, otherwise known as a size premium … Second, the experts debated the size of the equity risk premium … Third, the experts disputed which set of commodity price forecasts to use to generate cash flows … Fourth, the experts diverged in their treatment of Stillwater’s exploration areas … Four other disputes account for the remaining valuation swing … The legitimate debates over these inputs and the large swings in value they create undercut the reliability of the DCF model as a valuation indicator. If this were a case where a reliable market-based metric was not available, then the court might have to parse through the valuation inputs and hazard semi-informed guesses about which expert’s view was closer to the truth. In this case, there is a persuasive market-based metric: the deal price that resulted from a reliable sale process. Dell and DFC teach that a trial court should have greater confidence in market indicators and less confidence in divergent expert determinations. See Dell, 177 A.3d at 35-38; DFC, 172 A.3d at 368-70 & n.118. Compared to the deal-price metric, the DCF technique “is necessarily a second-best method to derive Page 63 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 64 of 257 value.” Union Illinois, 847 A.2d at 359. This decision therefore does not use it. See In re Appraisal of Solera Hldgs., Inc., 2018 WL 3625644, at *32 (Del. Ch. July 30, 2018).” Panera

¶ [135] In Panera Bread Company (Del. Ch. January 13, 2020) at page 116 the court referred to a “persuasive deal price” and concluded that the petitioners had not proved their DCF model’s reliability. Quoting from Pinson “While the particular assumptions underlying its application may always be challenged in any particular case, the validity of the [DCF] technique qua valuation methodology is no longer open to question”. Nevertheless the Supreme Court of Delaware “cautioned against using the DCF methodology when market-based indicators are available.”

¶ [136] In Panera, the court citing Stillwater added that compared to a persuasive, market-based deal price metric the DCF is necessarily a second-best method to derive value. The court added that because the petitioners were urging the court to give significant weight to the DCF model “they bear the burden of convincing the court that the model is sufficiently reliable to merit weight in the face of Panera’s reliable deal process.” Fir Tree v Jarden

¶ [137] In Fir Tree Master Fund LP v Jarden Corporation 236 A 3d. 313 (decided 9 July 2020) (“Fir Tree v Jarden”) the Supreme Court of Delaware dealt with another appraisal appeal and held that fair value is a jurisprudential, rather than purely economic construct. Chief Justice Seitz referred to the 4-day trial.

¶ [138] The headnote also reflects the fact that it was held, amongst other things, that (a) both sides have the burden of proving their respective valuation positions by a preponderance of the evidence; (b) the trial judge must determine fair value, which just means fair and not the highest possible price that a company might have sold for; (c) the Delaware Court of Chancery was not barred from using the company's unaffected market price to supports its determination of fair value; (d) the Court of Chancery was within its discretion in finding that the market did not lack material non-public information about the company’s financial prospects, in relying on the company’s unaffected market price to determine that Page 64 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 65 of 257 $48.31 per share was fair value of its stock at the relevant date even though the negotiated deal price was $59.21 per share, where a semi-strong efficient market quickly assimilated all publicly available information into its stock price, and differences in stock prices was not attributable to undisclosed information; (e) the Court of Chancery was within its discretion in declining to treat the negotiated deal price as a floor for fair value of the Company’s stock as at the relevant date even if the deal negotiation process was flawed and seller’s negotiator capped value under what might be achieved in true arm’s length negotiations, where there was evidence that the merger price exceeded fair value and that shareholders received value of synergy from the merger that was created by the deal; (f) the Delaware Court of Chancery was within its discretion in finding that the market value evidence supported its conclusion that the company’s unaffected market price of $48.31 per share fair value of its stock as of the relevant date even though the negotiated deal price was $59.21 per share where the record contained a mix of valuation ranges supporting either side’s valuation position, including higher price targets from some market analysts and financial advisors, there was evidence that the acquiring corporation paid premium for control and anticipated synergies from the merger, and the acquired corporation capped its stock buyback at $49 per share following the market’s negative reaction.

¶ [139] The Supreme Court of Delaware in an admirably concise 19 page judgment in Fir Tree v Jarden referred to Jarden Corporation C.A. No 12456-VCS (Del.Ch. July 19, 2019) (“Jarden”) where the Delaware Court of Chancery decided not to use the deal price less synergies as a reliable indicator of fair value. Slights, Vice-Chancellor noted at page 1 that at the close of trial he had observed “[w]e are in the classic case where … very-well credentialed experts are miles apart … There’s some explaining that is required here to understand how it is that two very well-credentialed, I think, well-intended experts view this company so fundamentally differently.”

¶ [140] At page 5 of his judgment the Vice-Chancellor in Jarden offered an apology for “an unfortunately long opinion” adding: “Appraisal litigation can be unwieldy. This is one of those cases. Apologies in advance to those who read on.” Slights, Vice-Chancellor, began where he thought he must “with the market evidence.” Page 65 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 66 of 257

¶ [141] At page 8 of his judgment the Vice-Chancellor stated “… not surprisingly, the experts’ DCF analyses yielded results that were solar systems apart.” The Vice-Chancellor was “satisfied that both experts utilized inputs in their DCF models that were not justified and that skewed the results.” At footnote 12 the Vice-Chancellor made an important point with which I concur, that it was not “proper, in my view, for parties to an appraisal case to cite decisions of this court, or our Supreme Court, for the proposition that a particular valuation methodology should be applied to value the target company. While legal authority may support the contention that a valuation methodology has been accepted by Delaware courts as generally reliable, I see no value in referring to the factual conclusions of another court in another case while appraising fair value of another company when attempting to fulfill the statutory mandate that I determine the fair value of this Company.”

¶ [142] The Supreme Court of Delaware in Fir Tree v Jarden noted that the Delaware Court of Chancery in Jarden was mindful of the Supreme Court of Delaware’s guidance in Dell that a “robust sale process” can discover a corporation’s fair value but held that the Jarden sale process raised concerns and “left much to be desired.” It held that Jarden’s CEO acted with “little to no oversight by the Board.” There was a lack of pre-signing or post-signing market check. There was no dispute in respect of synergies. After reviewing the evidence, the Delaware Court of Chancery decided to “place less weight on this market-based valuation approach in this case because the sales process was not well-conceived or well- executed and the expert analysis of the transaction synergies raised more questions than it answered.” The Delaware Court of Chancery rejected the comparable companies analysis of $71.35 per share. The Delaware Court of Chancery “refused to adopt either parties’ DCF model because of the wildly divergent fair value conclusion” (page 7). They had varied from a high of $70.36 to a low of $48.01 per share. The experts’ inputs to populate the DCF models were similarly divergent. One expert advocated a 4.9% terminal investment rate, another 33.9% rate. The Delaware Court of Chancery selected its own inputs that led to a DCF fair value of $48.13 per share – a result it ultimately chose not to rely on as a primary source for its fair value award. The Delaware Court of Chancery expressed “significant discomfort with the comparable companies and DCF valuation models”, the Delaware Court of Chancery decided to use Jarden’s unaffected stock price ($48.31 per share) as the best evidence of Jarden’s fair value. Relying in part on an event study performed by Jarden’s expert, the court found that the market for Jarden stock was informationally efficient, meaning the corporation’s stock price quickly reflected publicly available information. Further, according to the Delaware Court of Chancery, the parties did not have material non-public information that would not be reflected in Jarden’s unaffected stock price. The court also found that minority and conglomerate discounts should not be Page 66 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 67 of 257 applied. And the unaffected market price was not stale as of the merger date because the evidence showed that Jarden’s financial prospects worsened between the unaffected trading date and closing the merger.” (page 8)

¶ [143] In Fir Tree v Jarden the Supreme Court of Delaware commented (at page 9) that neither Aruba which followed closely on the heels of two other important Supreme Court of Delaware appraisal decisions, DFC Global Corp and Dell, nor their other recent appraisal decisions ruled out using any recognised valuation methods to support fair value. The Supreme Court of Delaware (at page 10) said that the thrust of the DFC decision took issue with the Delaware Court of Chancery’s reasoning for rejecting deal price as relevant to fair value. The Supreme Court of Delaware also questioned inputs made by the trial court in its DCF analysis. The Supreme Court of Delaware had refused to adopt a presumption in appraisal proceedings favour deal price for fair value and they noted that “our refused to craft a statutory presumption … does not in any way signal our ignorance to the economic reality that the sale value resulting from a robust market check will often be the most reliable evidence of fair value” and “second-guessing the value arrived upon by the collective views of many sophisticated parties with a real stake in the matter is hazardous.” The Supreme Court of Delaware in Fir Tree v Jarden commented that in Dell they found that the Delaware Court of Chancery erred when it assigned no weight to market value or deal price as part of its valuation analysis, but recognised that the market is not always the best indicator of value and need not always be accorded some weight. The Supreme Court of Delaware stated that in Aruba they explained the “considerable weight” a court should give the deal price “absent deficiencies in the deal process” because, for example, “a buyer in possession of material non public information about the seller is in a strong position (and is uniquely incentivized) to properly value the seller when agreeing to buy the company at a particular deal price”. The Supreme Court of Delaware also recognised that “when a market was informationally efficient in the sense that ‘the market’s digestion and assessment of all publicy available information concerning [the Company] [is] quickly impounded into the Company’s stock price’, the market price is likely to be more informative of fundamental value”. The Supreme Court of Delaware in Fir Tree v Jarden (at page 10) continued: “And how informative of fundamental value an informationally efficient market is depends, at least in part, on the extent of material nonpublic information. We also noted that it is a “traditional Delaware view” that in some cases “the price a stock trades at in an Page 67 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 68 of 257 efficient market is an important indicator of its economic value” and “should be given weight”.”

¶ [144] The Supreme Court of Delaware at page 18 noted that the “wide swing in value attributed to one input in the DCF model supports the reason the court did not rely on it – a lack of confidence that the experts were providing reliable economic advice on the inputs driving the DCF model”.

¶ [145] The Supreme Court of Delaware affirmed the Delaware Court of Chancery’s judgment. Pivotal

¶ [146] In HBK Master Fund L.P. v Pivotal Software Inc CA 2020-0165-KSJM (Del. Ch. August 14, 2023, as corrected March 12, 2024) (another case in which Professor Lehn provided expert evidence) McCormick, C in a 133 page judgment dealt with the legal analysis from page 41 onwards. At pages 54-55 the Chancellor stated (footnotes omitted): “Although the Delaware Supreme Court has expressly declined to adopt a presumption in favor of any one valuation methodology over another, recent decisions of Delaware Courts suggest a pecking order of methodologies for determining fair value. “In the aftermath of Dell and DFC, … the fair value analysis should ‘begin with the market evidence’”. Among the market based indicators, the deal price (minus synergies) approach is the “first amongst equals”. “[A] more subjective valuation technique, like DCF methodology or comparable company analysis, “is necessarily a second-best method” when ‘market-based indicators are available’. The structure of this decision follows this hierarchy ascribed to valuation methodologies by the Delaware Supreme Court.”

¶ [147] At pages 58 the Chancellor commented that the non-exhaustive list of objective criteria in respect of the sales process “does not map neatly onto a controller squeeze-out.”

¶ [148] At page 63 the Chancellor stated: “Delaware courts recognize that a stock’s unaffected trading price” can be a proxy for fair value “where the market is efficient” .At page 64 the Chancellor commented that a market is semi-strong efficient if all public information is fully reflected in the securities price and the release of new public information is quickly reflected in the security price. At page 67 the Page 68 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 69 of 257 Chancellor commented that “the presence of a controlling stockholder provides reason to be skeptical of arguments touting market efficiency”. At page 68 the Chancellor refers to a controller being able to veto any transaction it disfavours. At page 69 the Chancellor added: “The result is that the presence of a controlling stockholder should weigh against a finding of efficiency. In any event, “efficiency is a matter of degree”.

¶ [149] From page 69 onwards the Chancellor refers to two DCF analyses conducted by Professor Lehn. It is clear from page 70 that the petitioners did not dispute Professor Lehn’s selected discount rates or his approval of arranging two parallel DCF analyses that use different discount rates. Instead the petitioners attacked Professor Lehn’s free cash flow projections and terminal value on the grounds that the management projections from which Professor Lehn derived these inputs were unreliable. Footnote 367 records that the Petitioners also challenge Professor Lehn for merely seeking to corroborate his opinion that the merger price reflected fair value and that he performed the analysis “with bias.” It is noted that “The court does not find this argument persuasive, as it merely shows that Respondent’s expert had a motive to produce a result favoring Respondent. That much is apparent from the adversarial nature of litigation.” It appears that the rules in respect of expert evidence in Delaware are different to the rules in the Cayman Islands.

¶ [150] The Chancellor refers to a quote from Stillwater “The DCF method is a technique that is generally accepted in the financial community” and to quotes from other cases: “[T]he reliability of a DCF analysis depends on the reliability of the inputs to the model”, “Although widely considered the best tool for valuing companies when there is no credible market information and no market check, DCF valuations involve many inputs – all subject to disagreement by well-compensated and highly credentialed experts – and even slight differences in these inputs can produce large valuation gaps.”

¶ [151] From page 101 onwards the Chancellor refers to “Lehn’s Market Adjustment” to his DCF analyses. On the particular facts and in the particular circumstances of that case the Chancellor declined “to adopt a similar market adjustment” (page 102). The factual evidence

¶ [152] I now turn to the factual evidence placed before the court. Page 69 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 70 of 257 Mr Yan’s evidence

¶ [153] Mr Yan is the President and CEO of the Company and has held the position of CEO since 2000.

¶ [154] In his affirmation affirmed on 5 August 2024 Mr Yan provides the following evidence.

¶ [155] The Company’s business commenced operations in the PRC in 1998. He is one of the co-founders along with Mr Michael Lei Feng, Mr Noman Lui and Ms Kathleen Chien. The Company completed an initial public offering (“IPO”) on the NASDAQ Global Select Market in September 2004. He beneficially owned about 30.6% of the Company’s shares immediately following the IPO and was its largest shareholder. He has retained a substantial shareholding in the Company since and as at 17 September 2020 (the date of the Original Proposal) his shareholding was about 17.6%.

¶ [156] Mr Yan says that by mid-2020 the Company was faced with the following challenges: (1) the COVID-19 pandemic; (2) increasing competition most notably from Khanzun Limited (“BOSS”); and (3) geo-political risks from the trade dispute between the United States of America (“USA”) and the PRC and a real risk that the Company’s listing could be forcibly terminated at short notice which would likely be disruptive and costly to the Company.

¶ [157] Mr Yan had said that he came to believe that taking the Company private would likely be one of the best ways to address the challenges it was facing, basically for three reasons: (1) the Company needed to make substantial investments in its business, particular in its sales and marketing and its product and technology teams to respond to the competitive threats. Such investments were expected to depress the Company’s profit margins in the near term and be unpopular with public investors. Privatisation would help create a company which could be more nimble and dynamic in its decision making and better prioritise addressing longer term threats over short-term margins or changes in share price; (2) a response to rising geo-political tensions between the PRC and the USA; and Page 70 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 71 of 257 (3) the Company required a renewed entrepreneurial spirit. It needed to attract and retain talent, particularly in its sales and its product and technology teams, that would enable it to respond to new technology threats and to competitors poaching members of its sales team. Mr Yan says that as a public company it is difficult to convince shareholders to agree to incentive schemes comparable to those offered by private start-up companies. Mr Yan adds that privatisation would allow the Company the chance to offer employees stock option programmes competitive with those offered by private technology start-up companies.

¶ [158] Mr Yan describes the events leading to the Original Proposal and the Original Merger Agreement.

¶ [159] Mr Yan says on 20 June 2020 he received an email from Mr Hsu who he describes as a Hong Kong based financial advisor and private equity specialist who wanted to explore opportunities between 58.com and the Company. Mr Yan asked Mr Hsu to liaise with Ms Chien and he says he had limited direct contact with Mr Hsu. Mr Yan says he was a member of a WeChat group with Mr Hsu and he also “downloaded the Signal messaging at the request of Mr Hsu in September 2020 but I rarely communicated directly with Mr Hsu.”

¶ [160] On 14 September 2020 the Company engaged Mr Hsu, through his company RM, as a strategic consultant to explore potential strategic transactions.

¶ [161] There was a conference call between Mr David Liu, the executive chairman of DCP, a private equity firm, Mr Yan and Ms Chien on 16 September 2020. The following day on 17 September 2020, the board of directors of the Company (the “Board”) received DCP’s preliminary non-binding proposal letter (the Original Proposal) to acquire all of the outstanding shares for US$79.05 in cash per share (the Proposed Transaction). Mr Yan says that Ms Chien told him that Mr Alex Zhang of Ocean Link had reached out to her and expressed an interest in participating in the Proposed Transaction as a member of a potential buyer consortium.

¶ [162] On 20 September 2020 the Board resolved to establish a Special Committee to investigate the Proposed Transaction and negotiate with any potential investors. Page 71 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 72 of 257

¶ [163] Recruit, a Japanese company, was the largest shareholder in the Company at the time of the Original Proposal and held approximately 34.8% of its shares. Mr Yan was of the view that no deal was possible without Recruit’s support.

¶ [164] In the Original Proposal, DCP expressed their interest for Recruit and Mr Yan to rollover all, or a portion of their shares in the Company. If Recruit simply rolled over all of its shares then it would become the majority owner of the Company post-merger holding about 55% of the shares. Mr Yan believed this would have been “very problematic for the Company, because the Company would likely have come to be seen within the PRC as a foreign, and specifically a Japanese company”. Mr Yan says that when Recruit first became a shareholder in 2006 he and Ms Chien negotiated a 40% standstill clause to set a limit on how large a shareholder Recruit could become. It was critical that the Company not be seen as a foreign company.

¶ [165] Mr Yan says that Recruit’s position was key to him in deciding whether to seriously consider the Original Proposal. Mr Yan says that he was only willing to participate in the Proposed Transaction with Recruit rolling over if Recruit’s shareholding was at least 1% less than his. Mr Yan believed that it was in the best interests of the Company for there to be a majority shareholder and for that shareholder to be a Chinese person as this would give the Company the best opportunity to navigate the changing regulatory landscape in the PRC.

¶ [166] By the end of March 2021 Mr Yan and Recruit had agreed in principle that (a) Mr Yan would provide a commitment of a least US$300 million to increase his post-merger shareholding to 38% (b) Recruit would retain 37% of the common equity of the post-merger company and another 10% in the form of a convertible bond and sell enough shares so that its shareholding post-merger would be no more than 37% and (c) the A and B class shares would carry different voting rights such that Mr Yan would obtain majority voting control notwithstanding that his economic interest in the Company would be less than 50%.

¶ [167] Following the alignment reached with Recruit, Mr Yan agreed that he would be willing to roll over his shares and participate in the Proposed Transaction. In early April 2021 Mr Yan turned his focus to PE Sponsors who started to communicate their intention to seek a price reduction. Mr Yan says he “pushed back on any reduction, because I believed any changes in the key parameters of the deal structure would undermine the alignment I had reached with Recruit, which had taken some 6 months to achieve”. Mr Page 72 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 73 of 257 Yan says that the PE Sponsors relented and agreed not to pursue a price reduction so on 4 May 2021 he joined with the PE Sponsors to form a buyer consortium. On 4 May 2021 the Buyer Consortium delivered an updated non-binding privatisation proposal letter to the Special Committee (the Updated Proposal) at a price of US$79.05 per share i.e. the same price as in the Original Proposal. Recruit declined to join the Buyer Consortium at this time. Mr Yan said he joined the Buyer Consortium and agreed to work exclusively with the PE Sponsors as he had reached alignment with them and Recruit and privatisation was the best path forward for the Company.

¶ [168] The Original Merger Agreement was executed on 21 June 2021 and the parties were planning to make a joint filing with the US Securities and Exchange Commission in early August 2021. However on 4 August 2021 the PE Sponsors informed him they would not be ready to file on time as they were concerned that the transaction may require regulatory approval from the National Development and Reform Commission (“NDRC”). Mr Yan said he pushed back as he was concerned over delay. Over his and the Company’s objections the PE Sponsors unilaterally initiated an informal consultation with the NDRC in mid-August 2021.

¶ [169] Mr Yan says that the Company was coming under increasing pressure from the market to provide an update on the status of the merger and its share price was dropping due to the uncertainty of the transaction as perceived by the public. Mr Yan says that the Company was reluctant to publicise the fact of the consultation with the NDRC without the NDRC first confirming that it did not object to the consultation being published and the PE Sponsors shared similar concerns.

¶ [170] However the Company issued a press release on 8 November 2021 which Mr Yan says stated “that certain members of the consortium were in consultation with regulators”. Mr Yan adds: “The Company’s share price promptly dropped almost 20% that week … The Company’s share price dropped to around US$50 by mid-November and it would drop further to around US$45 in December 2021.”

¶ [171] Mr Yan says that during November 2021, the PE Sponsors started discussing the idea of revising the deal terms with him to structure it to avoid their need for regulatory approval. In addition on 26 November 2021 Mr Wolhardt of DCP sent Mr Yan an email attaching a revised proposal in which the Page 73 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 74 of 257 PE Sponsors proposed to offer a lower price of US$65.14 per share. Mr Yan was not in favour of any reopening of negotiations over the deal terms.

¶ [172] On 23 December 2021 Mr Wolhardt sent Mr Yan an email attaching another revised proposal which offered an even lower price of US$57.25 per share. It also reduced the PE Sponsors’ post-merger shareholding to just below 10%, which Mr Wolhardt said would allow in Mr Yan’s words, the PE Sponsors “to avoid the alleged requirement to obtain the NDRC’s approval to the Merger.” Mr Yan replied on 24 December 2021 and said he was doubtful that the parties would be able to make significant changes to the deal structure and close before the longstop date of 21 March 2022. Mr Yan refused to consent to the revised proposal but did not object to the PE Sponsors discussing it with Recruit and the Special Committee. Mr Yan informed Mr Wolhardt that if the PE Sponsors were able to reach agreement with the Special Committee and Recruit then he would not be the one to stand in the way.

¶ [173] On 12 January 2022 a revised proposal was submitted to the Special Committee (the “Revised Proposal”) under which the PE Sponsors proposed to reduce (a) the offer price to US$57.25 per share; and (b) their combined ownership to 9.9%. The Special Committee negotiated a raise in the share price to US$61.00 per share.

¶ [174] During his cross-examination Mr Yan came across as a confident, honest individual with a clear passion, verging on an obsession, for the long-term success of the Company. He was plainly proud of the Company and what it and he had achieved over many years.

¶ [175] Mr Yan was keen to portray himself as an operations person rather than a person of finance.

¶ [176] Mr Yan stated: “The most important thing is the competitive threat. We had a very serious competitive threat that we have to address. If we don’t address it well, the long-term success of the company will be in doubt. And, secondly … the geopolitical tension between China and the US is becoming very complicated, and with the Holding Foreign Companies Accountable Act we … there’s a possibility we could be delisted from NASDAQ. Page 74 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 75 of 257 And, finally, the company has been around for more than 20 years. I felt that we need to re-ignite the entrepreneurial spirit of the company. Particularly in light of the competition from BOSS, we need to attract the best talent and we need to have a better incentive scheme, including stock options, so that we can attract, retain and motivate the best talent …”

¶ [177] Mr Yan stressed these 3 main factors; (1a the competitive threat; (b) the USA/PRC geopolitical situation; and (c) the need to reignite entrepreneurial spirit. He was to return to these 3 important factors on other occasions during his evidence and I have no doubt that they were genuinely in his mind at the relevant times. Mr Yan did not strike me as an individual who was simply hoping to get the Company “on the cheap”. He came across as an individual who genuinely cared for the long-term successful future of the Company. I believed him when he spoke of his thinking behind the merger.

¶ [178] Mr Yan stated that “We knew very little about what [take private] meant” and its implications so they engaged Mr Hsu because “[h]e has experience in take-private processes as well as in the recruitment industry in China.”

¶ [179] Mr Yan admitted that taking the Company private could result in very considerable gains but “only if we can maintain the long-term success of the company after privatisation. If we cannot ensure the long- term success of the company, the value of the company can go … can decrease meaningfully, and it is not about… a financial engineering exercise. It’s about how we combine our assets and compete more effectively with BOSS, given their trajectory.”

¶ [180] It was put to Mr Yan that Mr Hsu was suggesting it was a good time for a take private because the market price was at a low point in the cycle with the consequence that anybody taking it private could get the company on highly attractive terms. Mr Yan responded “No, I disagree” and then gave the reasons for the lower price and when pressed by Mr Adkin for the third time to answer the question responded “I disagree because share price can go lower if the economy continued to deteriorate and our performance continued to be affected by BOSS. I never thought that was a good point because share price changes over time as a result of the economy, our performance and technology shifts.” Mr Yan maintained his position when further challenged by Mr Adkin: “I disagree with the fact that that is a good point because performance changes over time.” Page 75 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 76 of 257

¶ [181] Mr Yan openly and without hesitation accepted that “this was an MBO” [a management buy out] and accepted that the idea was for the Company to be relisted in the PRC but “relisting only is possible if we deliver performance and long-term success.”

¶ [182] Mr Yan, who was plainly worried about the competitive threat from BOSS, with apparent genuine frustration in his face and voice stated: “… we were outspent by BOSS around three to five times in advertising and promotions. I was very frustrated because we have US$1.5 billion cash. I could have fought back directly and compete with them. I could not because we were a public company. We were listed for more than 17 years. We’ve done 63 earnings call, we’ve given 63 guidance, we only missed five … we exceeded slightly the top range of the guidance like more than half the time. So we have been a very consistent and profitable company on NASDAQ. I have a lot of cash but if I use that cash in a meaningful way to compete with BOSS, my earnings will go down. The character or personality of being a consistently profitable company will be gone. So I was very frustrated. I had the cash, I could have fought back BOSS, but I couldn’t because I was listed.”

¶ [183] It was put to Mr Yan that he could together with Recruit (over 50% of the shareholders) have agreed to dramatically increase the spending to compete with BOSS and there would be nothing the minority shareholders could do. Mr Yan said as a CEO of a US listed company he had a fiduciary duty to look after the interests of the minority shareholders.

¶ [184] I accept that Mr Yan, a person of business, was rather unconvincing in his response to Mr Adkins’s legally-based question to the effect whether he was suggesting that if he felt it was in the interests of the minority shareholders to pump up the spending in order to compete with a competitor that would be in breach of his fiduciary duties to them. Mr Yan, clearly in a legal area outside his operational comfort zone, responded “Yes, depending on the scale … if we increase our sales and marketing spend by three to five times, I believe we’ll get a lot of questions … the smaller minority shareholders … sell out. Our share price will tank … minority shareholders are important …. I have limited financial Page 76 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 77 of 257 knowledge but I know that being a public company the majority shareholder cannot just take advantage of the minority shareholders. We do things in the best interests of the Company.”

¶ [185] Mr Yan, with justification, proudly stated: “My focus is the long-term success of the company. Over the 17 years being public … we created a lot of value for our shareholders, not only Recruit but also the minority shareholders, as well as our employees through their stock options. I have a track record that I’m very proud of, that I will tell my children, which I have already done, have already told my children, and I want to keep that track record and being a very solid, consistent, well-performing company in China.”

¶ [186] Mr Adkin directly suggested to Mr Yan that the motivations Mr Yan mentioned for going private namely renewed entrepreneurial spirit and the need to see off a perceived competitive threat from BOSS were not actually the real motivations at the time. Mr Yan firmly maintained his unshaken consistent position and I believed his response: “No, that’s totally incorrect. The most important motivation is to stay being the market leader and ensure the long-term success of the company. And that will mean that we need to effectively address the competitive threat from BOSS as well as creating an organisation … we can attract, retain and motivate the best talents.”

¶ [187] Mr Yan felt that it was important that Mr Hsu of RM had “skin in the game … that our interests are aligned.”

¶ [188] Mr Yan was cross-examined at some length in connection with his use of the Signal app: “I did use Signal for the transaction as advised by Mr Hsu, our adviser … I also talked to Ms Chien on Signal because she didn’t have any other instant messaging tools and she was in the US. She was stuck in the US, couldn’t come back to China. We had a … 15/16 – hour time zone difference … I did not use Signal for any other purposes, and certainly not in our ordinary course of business.” Page 77 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 78 of 257

¶ [189] A Maples letter dated 5 December 2024 (LA/112/1) was put to Mr Yan:-“Have you ever used your personal WeChat/Ding Ding/What’s App/Signal for any work related communications” and Mr Yan’s answer is recorded as “No”. Mr Yan said he did not use Signal for work related purposes. He used emails as he needed a trace. “I always used email so that I can keep a record and I can trace what I have communicated to my team … I already apologised earlier on that I misunderstood the question as to whether that is in the ordinary course of business.”

¶ [190] Mr Yan accepted that there were Signal communications “for the transaction” and also for the period “when Ms Chien was stuck in the US and the only IM tool that she had was Signal.”

¶ [191] I have carefully considered the detailed cross-examination on the answer recorded in the Maples letter but do not find that Mr Yan provided knowingly false information in that respect. I find that it was a genuine layman’s misunderstanding. The lawyers for the Dissenters are throwing it out of all reasonable proportion.

¶ [192] Mr Yan stated that he had made available all his emails, all his Signal messages and all of his WeChat messages: “I have not blocked any information or channels of communication … Or refused to give access … to any communication or channels of communication.”

¶ [193] Mr Adkin suggested to Mr Yan that he did not make a mistake in his answer in the Maples letter but he had made a choice. Mr Yan replied “I disagree”.

¶ [194] I considered the detailed cross-examination on the Signal auto delete function and do not make a finding that Mr Yan deliberately enabled that feature to prevent material evidence coming to light in these proceedings.

¶ [195] Mr Yan stated “I do not recall setting that auto delete function.”

¶ [196] The Maples letter says the disappearing messages feature in the conversation between Mr Yan and Kat C was activated on 10 December 2021.

¶ [197] Mr Yan stated “I do not recall who actually activated the function. It must be either me or Ms Chien, probably at the advice of our consultant [Mr Hsu]. But I felt it was not relevant because I’ve been talking to Ms Chien on WeChat by November 2021 [when Ms Chien came back to the PRC]”. Page 78 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 79 of 257

¶ [198] Mr Yan agreed that the experienced Mr Hsu had a “playbook for taking things private” and when it was put to him that “part of his playbook was getting people to set their Signal to autodelete, wasn’t it?” Mr Yan replied “It appears so”. That was an open and honest answer. I did not get the impression that Mr Yan was seeking to deliberately conceal any relevant evidence from the court.

¶ [199] On a separate topic, Mr Yan said that “[t]here are scenarios that I could exit the company … We had no idea how Recruit will react … I am aware there’s a possibility that Recruit may make a counter-offer to buy out the company, or they might want to take this opportunity to gain majority control or consolidation of the company. If they want to do that, I would rather sell out and walk away. I do not want to be dealing with a company that’s owned by the Japanese. It will make our life very difficult … Perfectly okay for me to retire and walk away … It’s always something that I had in mind that this is a possibility, although it never surfaced.”

¶ [200] Mr Yan disagreed that the take-private had nothing to do with the Company’s fundamentals.

¶ [201] Mr Yan was asked who looked after his “personal financing”. Mr Yan initially responded: “I don’t have anyone looking at my financing.” Mr Yan accepted he was very familiar with financial concepts. Mr Yan was then specifically asked who went to the banks and made enquiries to see if they could get personal financing for him. Mr Yan responded “Kilometre helped me to talk to some of the banks … They’ve done it before, they know who are the banks that will play in these Chinese take-private deals.” Mr Yan at first appeared reluctant to immediately admit that Kilometre acted for him personally in respect of his borrowing to finance his extension of equity. Perhaps he was simply being naturally and understandably coy about his personal finances. “They were giving me an idea who might be out there.” Having been pressed a few more times Mr Yan eventually responded “Yes … in some part” to the question “They were helping you with your personal financing”. Again I do not throw this point out of all reasonable proportion.

¶ [202] Mr Yan was asked about a document entitled “Project Elevate” (DA/162/1), to which he stated “I’m an operating manager. I’m not a financier … I have advisers because they’re expert and professionals in different fields. I don’t think they understand how to run a business. I have to run the business. I have to deliver results … I’ve been very consistent with the public about the three reasons why we go for Page 79 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 80 of 257 take-private. Whether the adviser got it, whether they add additional points for various reasons, I don’t know. I have been very consistent and clear about what our objectives are.”

¶ [203] Mr Adkin put it to Mr Yan that his “evidence about the position with BOSS being the driver for the transaction is something that you have created afterwards to support your case in this litigation”. Mr Yan’s firm and clear response was “No, absolutely incorrect … we were all concerned about BOSS. That was my top of my issue in the industry.”

¶ [204] On stock options, Mr Yan stated “[t]here were limits about how much public company can issue options. I remember we were issuing less than 2% of our outstanding shares every year to employees but if you look at a start-up company normally there will be a 10%-15% option pool given to employees …”.

¶ [205] Mr Yan agreed that the conduct of DCP was damaging the Company. Mr Yan agreed that the Company was being damaged by the delay and the share price was dropping.

¶ [206] It was put to Mr Yan that the November 2021 Press Release was a compromise. He responded: “It was a discussion between the counsels that we all felt comfortable what to be released. We have to tell the market what’s going on because we’re getting a lot of enquiries, but then we have to take the advice from counsel what can be said and what cannot be said.”

¶ [207] Mr Yan agreed that the press release was at least capable of causing a reasonable investor or potential investor to conclude that there had been regulatory changes which were concerned with the operation of the Company itself and which would negatively impact its business. I will return to the press release later in this judgment. Suffice to say I do not find the subjective comments of individuals of much assistance in construing its terms and impact.

¶ [208] A communication from Churchill Capital (DA/423/1) was put to Mr Yan who accepted that the market was reaching conclusions about the press release which were wrong and the Company did not feel able to correct them. A letter dated 14 November 2021 from Palliser (DA/424/1) was also referred to. Page 80 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 81 of 257

¶ [209] Mr Yan agreed that it was unsurprising that the press release had a negative impact on investor sentiment and the Company’s share price but “not completely accurate because … there’s certainly signs that the deal is not going smoothly. We know that if the deal doesn’t go through, it will be detrimental to the long-term success of the company because we will not be able to complete the merger and be able to compete – invest and compete and catch up with BOSS.”

¶ [210] It was put to Mr Yan that the press release caused the share price crash. Mr Yan responded: “… my understanding is the press release caused doubt whether the deal can be done … Yes the press release has an impact on our share price because it casts doubt on whether the deal will be completed, which took away the support for the share price at … 79.05. But the general trend in terms of value of the Chinese companies decline specifically for us that we lost our online market leadership position to BOSS, we were overspent not only on A & P but also in R & D and technology investment. There are just many things going on, unfortunately I think, nothing good. So it was a trigger. I agree it was a trigger, but I would not suggest the … press release itself caused a decline on the value of the company.”

¶ [211] Mr Yan said he did not consent to the DCP and Ocean Link proposal to a revision to the deal. Mr Yan agreed that he was prepared to push forward with the existing merger at the original price but would not block the way forward if DCP and Ocean Link got everybody else in line.

¶ [212] Mr Yan also agreed that the shareholder base changed from long-term investors to arbitragers. Ms Chien’s evidence

¶ [213] Ms Chien is the COO of the Company and she has held that position since March 2009. She says prior to that she was the CFO from 2004 to March 2009. In November 2009 she stepped in as acting CFO a position she held until January 2024. During the relevant merger period (17 September 2020 to 27 April 2022) she was both COO and acting CFO.

¶ [214] Ms Chien led the Company’s finance team and her responsibilities included overseeing the preparation of all financial reporting, managing the processes for financial forecasting and budgets and advising Page 81 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 82 of 257 Mr Yan and the Board on long-term business and financial planning and providing strategic recommendations.

¶ [215] In April and May 2021, and in January 2022 the following finance team members helped her prepare the management projections: (1) Mr Chunyan (Winner) Xu, who had joined the Company in February 2006 (2) Ms Iris Ma who had joined the Company in September 2002.

¶ [216] Ms Chien says she also had discussions with senior management (including Mr Yan, various Vice- Presidents and Ms Linda Chien, her sister, who had joined the Company in March 2004 and was a Vice- President and manager of investor relations) who provided their views on the outlook of the business.

¶ [217] Ms Chien described the Company’s range of human resource services in two main categories (1) online recruitment services and (2) other human resource related services including (a) Human Resources (“HR”) business outsourcing; (b) campus recruitment; (c) training; (d) placement and executive search services also known as recruitment process outsourcing and (e) other HR related services such as assessment tools, salary surveys, conferences and awards. These two categories are reflected in the “Revenues” section of the income statement in both sets of management projections. As at 31 December 2021 the Company employed about 8,400 people across its 25 offices in the PRC and generated US$693,661,000 in net revenues.

¶ [218] Ms Chien refers to various PRC subsidiaries and affiliated PRC entities and describes the Company’s long-term investments.

¶ [219] Ms Chien provides detailed evidence in respect of the regular budgeting and planning activities undertaken by the finance team including historical budgeting and planning, quarterly market guidance, annual budget process and monthly internal business unit reports.

¶ [220] Ms Chien gives detailed evidence in respect of the preparation of the 2021 Management Projections in paragraphs 57 to 72 of her affirmation.

¶ [221] In section E of her affirmation Ms Chien details the challenges the Company was facing when she was preparing the 2021 Management Projections (paragraphs 73 to 89 of her affirmation). Page 82 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 83 of 257

¶ [222] In section F of her affirmation Ms Chien provides detailed evidence in respect of the 2022 Management Projections (paragraphs 90 to 94 of her affirmation).

¶ [223] In section G of her affirmation Ms Chien refers to the key differences between the 2021 Management Projections and the 2022 Management Projections (paragraphs 95 to 108 of her affirmation).

¶ [224] In section H of her affirmation Ms Chien deals with the impact of COVID lockdowns on the Company after the finalisation of the 2022 Management Projections (paragraphs 109 to 112 of her affirmation).

¶ [225] In section I of her affirmation Ms Chien refers to the reasons for the merger. She says she believed that the escalating US-China tensions and the increasingly onerous regulatory environment would likely create significant challenges that would come to outweigh the advantages to the Company of remaining a US-listed company. It was therefore worthwhile to explore strategic options, including the possibility of delisting in such an evolving environment.

¶ [226] In section J of her affirmation (paragraphs 117 to 135) Ms Chien outlines her involvement in the merger process, including engagement of RM, involvement in the discussions with the PE Sponsors, discussion with Recruit and the frustrating NDRC issue. Ms Chien stated: “Although Mr Hsu had suggested at the outset for the Merger Process that we communicate by the Signal messaging application and I did download it to my phone, it was primarily used as a communication channel to set up call times. I have found the app to be unreliable in the PRC. I principally communicated with Mr Hsu over the phone or by email.”

¶ [227] In section K of her affirmation Ms Chien outlines her financial interest in the merger. On 27 April 2022 she held 3.4% of the shares in the Company and rolled over all of her common shares in the merger and held 3.86% of the shares in the surviving company following the merger. In September 2020 she had not thought about what to do with her shares if the transaction proceeded. Over time she was of the mind that she would follow Mr Yan’s decision as to whether or not to roll over her shares. She had worked with Mr Yan for more than twenty years and she decided if he was going to continue with the Company then she would continue to support him. Likewise, if he had decided to leave the Company then she would follow suit as she was not keen on working with someone new. Page 83 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 84 of 257

¶ [228] Ms Chien says that to avoid the administrative hassle of having to convert her ADSs into common shares she sold them at the merger consideration and automatically exercised certain options. She also elected to participate in the management special purpose vehicle and purchased some shares. She specifies the amount she received on the closing of the merger (US$7,109,520).

¶ [229] During her cross-examination Ms Chien came across as a confident, honest individual who carefully listened to the questions and endeavoured to provide precise considered answers. At times during her cross-examination she also displayed a refreshing sense of humour and remained calm and unruffled at all times. She displayed an impressive attention to detail and strived to provide precise responses. On occasions she became a little more defensive but I gained the impression that she was honestly endeavouring generally to provide the court with accurate evidence. She plainly knew the business of the Company very well.

¶ [230] Under cross-examination Ms Chien agreed that she had a significant role to play in the provision of information about the Company to the market before September 2020 and Mr Adkin took her through the various ways in which the Company provided that information including (a) quarterly earnings calls involving discussions on results of the previous quarter and the provision of forward looking guidance. By way of example there was a reference to a transcript of the 10 August 2020 earnings call; (b) quarterly press releases with the prior quarter’s results, commentary on the results and forward looking guidance; (c) the market being able to ask questions of the Company’s investor relations department; (d) the Company would answer questions at various virtual investor events hosted by financial institutions; (e) the Company also answered investors’ questions by email; (f) the annual filing of Form 20-F with the US Securities and Exchange Commission; (g) an opportunity at annual general meetings for shareholders to participate in a meeting with the Company’s senior management; and (i) the Company was covered by a number of analysts including the Bank of China International.

¶ [231] Mr Adkin referred Ms Chien to an email dated 21 November 2020 (DA 398/2) and she accepted that there were no quarterly earnings calls after the original offer in September 2020 and the Company also stopped providing forward guidance.

¶ [232] Mr Adkin also referred Ms Chien to an email 27 September 2021 from her sister to the Special Committee referring to the “no communication policy” and asking in effect how investor queries should be dealt with in the future. Ms Chien referred to the Company having taken guidance and “we would Page 84 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 85 of 257 not make forward-looking statements, … we would be responding to emails for communication to ensure that the pipeline of communication was open, and that we were still releasing press releases for results to make sure that public shareholders were updated on our financial progress … we were not responding to comments about the transaction in general … we were not publishing forward guidance because we had consulted our attorneys and they thought that there is a potential to be considered maybe tipping the market or maybe market manipulation, there’s more general caution around that when there is a transaction on the table … we would respond in written form to make sure that we were clear and that we stay within the boundaries of not communicating or commenting on certain issues.”

¶ [233] Ms Chien agreed that the 2021 AGM was cancelled and no Form 20-F was produced for the financial year 2020.

¶ [234] Ms Chien was also referred to evidence indicating that certain analysts had discontinued their coverage of the Company as it was undergoing a planned privatisation and delisting.

¶ [235] Ms Chien referred to her understanding that some of the longer-term shareholders had exited or sold out some of their positions. Ms Chien openly accepted that long-term investors had gone out and arbitragers had come in.

¶ [236] Ms Chien was cross-examined at length in respect of the management projections she prepared in April and May 2021. Ms Chien says that the members of the finance team “did satisfactory work”. Ms Chien said that the purpose of the projections was to “be used as an evaluation related to the transaction”. In respect of the 2021 Management Projections “five years of data” were gathered to produce a starting point and then they extrapolated from that baseline what was their best estimate of where they thought the business would go absent the merger.

¶ [237] Ms Chien accepted that she prepared a single set of projections rather than different cases based on different assumptions.

¶ [238] Ms Chien was also taken to the proxy statement (DA 331/1). Ms Chien agreed that the projections were prepared based on management’s best estimates and judgments. Page 85 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 86 of 257

¶ [239] Ms Chien was also taken to the management representation letter provided to Duff & Phelps (DA 280.01/1). She was asked if it was an important letter and responded “I think it’s a standard rep letter, yes”. She was pressed “And those are important, aren’t they?” and she answered “Yes, I suppose, yes.”

¶ [240] Ms Chien said that “the projections at the time were projections that I felt we could achieve if we worked hard and that nothing bad happened … that’s the best case which I thought were stretch targets that we would have to work hard to achieve, that’s what I believed … if something bad happened we may not achieve it”. Ms Chien was pressed if something good happened it would be better and responded “It would be easier for us to achieve it, yes.” It was put to her that “you might achieve more” to which she responded “I don’t know. Anything is possible. It’s the future. I can’t predict it.”

¶ [241] Ms Chien said that when the 2021 Management Projections were sent to Duff & Phelps she had not personally committed to vote in favour of the merger. That did not happen until she signed a support agreement on 21 June 2021. There was a gap from the 2021 Management Projections of about eight months before the 2022 Management Projections were prepared in January 2022.

¶ [242] Ms Chien was asked about her use of Signal both before and during the period. Ms Chien recalled using the Signal messaging app “with a private chat with Rick and myself” and using Signal was suggested by Mr Hsu. It appears that Ms Chien downloaded the Signal app on 11 September 2020. Reference was made to a Signal communication with Mr Yan on 21 April 2021 where he says he would have to restart his Signal account. Ms Chien was referred to the various Signal chat groups. Some involved Kilometre. Ms Chien accepted that she had produced no Signal chats with Mr Yan “I would produce a lot of email communications he had, so we did produce information.” “Signal was not working well at all in China because I think there is some sort of firewall issue when you couldn’t get the message for the, what do you call it? The verification code.”

¶ [243] Ms Chien said that she did not recall activating the disappearing message feature on Signal on 10 December 2021 “I’m not actually a very good user of these kind of products because I historically didn’t use messaging services, which is why Mr Hsu had asked if I would actually use something, and I didn’t have that before … I am actually not very good at these kind of messages, but I don’t recall, maybe I did or I didn’t, I don’t know. It may have been a default feature, I don’t know. I’m not very good at these things. I did delete the app in 2022 early.” She later added in effect that she was not IT savvy. Page 86 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 87 of 257

¶ [244] Ms Chien was cross-examined at length as to the differences between the 2021 Management Projections and the 2022 Management Projections. In respect of the increasing impact of the PRC regulatory challenges Ms Chien was referred to the dates of the regulations and provided the following evidence which had a ring of truth about it: “Just in general when Chinese regulations are enacted, usually you don’t have a sense of how that’s going to impact you directly because usually when it’s actually then first promulgated, then it actually has to be interpreted by the so-called government regulators. So there is some time between the issuance of certain guidance and regulations until they’re implemented, and you actually only get a better sense of how it’s going to be implemented and how it affects you so that a lot of times you may not know that from day one.”

¶ [245] Ms Chien added: “We had to have people to look at the regulations. We actually had to change the way some of our product was provisioned, and so it did actually require quite a lot of technical workaround, and then there was actually additional requests for us to monitor or validate users to the platform and other things, so it did change quite a lot of way things were done or how the product was designed, if you will.”

¶ [246] In a particularly memorable exchange, Mr Adkin asked Ms Chien if she was saying that “the way Chinese law works is that some Delphic utterance is produced and everyone has to try and work out what it means?” Ms Chien responded “Actually, yes. I am suggesting that … I guess we have a different view on that.” In relation to the new laws which were in force or in the pipeline before the management projections were produced in 2021 Ms Chien accepted that they knew the laws were going to be implemented and “we understood that there may be increased costs, but it was not clear to us how much it would be and how it would be again interpreted and implemented …”. Ms Chien added that in 2021 “I don’t think we were in a position to clearly understand the cost of new regulation, and so that only became apparent once they were implemented and once there’s been some time for interpretation to happen, such that enforcement made clear what the related costs would be … I took account of what I knew or what I could clearly understand to be the case but, as it turned out, I underestimated and I did not fully appreciate what that meant in terms of the implementation of the Page 87 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 88 of 257 new regulations.” Ms Chien accepts that she did not in her affirmation put a figure on the underestimation.

¶ [247] Ms Chien was asked about the Double Reduction policy which effectively barred foreign investors from owning or operating for-profit education on tutoring companies in the PRC and responded “it did not directly impact the Company, I agree, but there is actually negative investor sentiment which would then reduce market demand for our services, so there is some indirect impact to that … it actually impacted the after-school tutoring market in China, which is actually a very substantial industry … there was a lot of hiring in that industry … so if this particular industry is going to be heavily impacted, that means it would actually impact the hiring demand in the industry, so technically investor sentiment in that segment is reduced, but that also means that it would ultimately translate to reduce kind of demand for our services, which is recruitment … I’m sorry if I didn’t make that clear enough in my statement before. I guess I didn’t explain furthermore about the double-reduction policy.”

¶ [248] Ms Chien was cross-examined in respect of the removal of the Company’s apps from various app stores.

¶ [249] It was put to Ms Chien that her statement at paragraph 81 of her affirmation that the removal of the Company’s apps from the Android and Apple stores on 21 July came as a complete surprise was not true. Ms Chien politely but firmly responded “No, it’s absolutely true. I disagree with you … yes, our apps were removed from a number of the Chinese phone providers … Apple made an enquiry but did not take action on us. But the removal only lasted a short time, so within days, because I said we worked night and day to try and understand and find out what’s happened, we were then quickly restored to the shelves so we thought the issue was resolved. So when you actually said earlier that July incident and this, these are not related incidents. In fact, we had thought that this issue was closed because in March we were reinstated very quickly, even though it was a very difficult few days … we could not have foreseen this. In fact we corrected or remediated our product … as requested by the regulators in March. We also complied with the request in May. We again complied with the request in June … and there was no feedback to say we were not making the right remediations. So going through March, May and June, and taking steps that we felt appropriate and aligned to regulatory request and not getting any other feedback to be very honest we were completely shocked and not understanding what happened in July … the July removal was a shock to us because it was never actually one where, unlike the May and June issue when they told us what the so-called infractions were and what we needed to do to remediate, we actually did not receive any official notices from the Page 88 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 89 of 257 regulators. It was simply “Remove from the shelf”. There was no explanation for that. So, similarly in November, after nearly four months of trying to communicate with regulators to no avail, it was restored, but we still don’t really understand what happened. So in the back of our mind we were always very concerned because we don’t know – if you don’t know what the problem is and … it’s difficult to anticipate whether or not there would be another problem down the line, so that’s the situation … in March it was very clear what the issue was because they saw a report on the CCTV segment, they specifically told us what they wanted us to change and we were able to do that … Similarly, in May and June we were given very specific information on what they wanted fixed. So I think that the July incident we would have to say is a very different nature because we just did not receive information.”

¶ [250] Ms Chien was cross-examined as to the increase in market competitiveness and the position of BOSS as being a reason for the difference between the 2021 and 2022 projections. It was put to her that BOSS was not spending three or four or five times the Company on sales and marketing (para 87 of her affirmation) but was spending less than twice in absolute terms. Ms Chien referred to the distinction between marketing and sales and apples as oranges.

¶ [251] Ms Chien thought that Mr Yan was aware of the threat but she thought he underestimated the impact of BOSS as a competitive threat. “I think that we were all very conscious that we would have to play catch-up on the technology and that he would have to invest more in marketing to keep pace with what was going on with BOSS. But having said that, even with some of those changes already in mind, I still felt … it was even a greater threat than what Rick was thinking … He was underestimating that … I think he was very aware that technology and marketing … and in terms of the recruitment business where we go head to head with BOSS that’s almost everything.”

¶ [252] In respect of the 2022 Management Projections Ms Chien accepted that they were prepared in January 2022 after the original merger agreement was entered into on 21 June 2021 at the price of US$79.05 a share.

¶ [253] The 21 June 2021 support agreement was put to Ms Chien whereby she agreed to vote in favour of the merger and against any competing transactions – “I guess I supported it [the merger]. I’m not sure at the time I was keen, but I supported it.” Page 89 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 90 of 257

¶ [254] The Duff & Phelps chart called by counsel the “football field” (DA 285.01) was put to Ms Chien with the lowest range US$70.57 from the DCF. There were references to the original deal price of $57.25 per share and then the $61 figure for the eventual recut deal price. Ms Chien agreed that she had been asked to produce updated projections to enable a new Duff & Phelps fairness opinion to be produced.

¶ [255] Mr Adkin put it to Ms Chien: “You would have realised at the time that if you didn’t make quite significant changes to the 2021 projections, then there was no way in which Duff & Phelps would be able to produce a fairness opinion, saying that $57.25 or anything near that low was fair from a financial point of view to the minority shareholders”. and she responded: “I wasn’t in charge of producing those fairness opinions, so what I did was just focus on the projections.” When pressed she added “… if nothing changed then the outcome will be the same but in fact time has passed and things have changed.” She was asked “If Duff & Phelps weren’t able to produce such an opinion, then you would have realised there would be no deal” and responded “Do you know, I don’t know. This is a decision by the Special Committee … All I could do was just to produce updated projections based on my understanding of the market and the Company’s circumstances and all that had transpired between May ’21 and January of 2022 which … was less favourable than I would have liked to see.” She ultimately agreed that the Special Committee would “probably not have taken it” if the opinion was not fair “But that decision lies with them and the fairness opinion lies with Kroll … I had no participation in that.”

¶ [256] Ms Chien seemed somewhat reluctant to accept that the Special Committee would be “heavily guided by what Duff & Phelps / Kroll would have advised” initially responding “they would have been advised by Kroll yes” and when further pressed “I think they would have taken into consideration what Kroll advise.” She eventually accepted that if Kroll advised that the proposed price on the recut deal would not be fair it is “most likely” that the Special Committee would have said “we’re not going to do the deal.” Ms Chien agreed that Kroll were going to rely on her projections: “Yes, they asked for an updated Page 90 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 91 of 257 set of projections.” She was asked “So you would have realised that if you’d produced the same set of projections with the same results, then the deal would be off.” She responded “My task was to be providing updated projections, so my task was not just to give them the same set of projections as I understood it.” Ms Chien acknowledged that if she came back with the same projections Kroll could come back with a similar response and there would be no deal. Ms Chien agreed that she did not have a specific conversation with anybody in management about the 2022 Management Projections but she had conversations with everybody about the operational state of things: “those discussions were ongoing all the time”.

¶ [257] Ms Chien said she had a meeting in May 2021 when Kroll interrogated the projections but not in January 2022.

¶ [258] Ms Chien said she had ten or eleven months’ data when preparing the January 2022 projections. Ms Chien was cross-examined on her figures and said “What I was saying is that if you look at the growth rate of the two sets of projections, the growth in the second set of projections is lower because I had a different view on our growth rate going forward, even though we had achieved higher revenues in 2021. That was my point … the 2022 estimate was still my best estimate at the time of the projections being prepared … I had underestimated what revenue we ultimately achieved in 2021 when in January ’22 I was looking back, yes, that’s correct.”

¶ [259] The brief suspension of the Company’s apps and the longer suspension of BOSS’s apps was also the subject of questioning.

¶ [260] Ms Chien was asked what she meant in that message when she said “If we want to jack up EBITDA, we can always ‘shut the spigot’ on marketing” and whether she meant “you could shut off the spigot of marketing spend to increase cash flows, yes, in the preliminary estimates that you give, yes?” Ms Chien responded: “No, I think that’s a general comment I was making that the … it looks like we were talking about the PEs being actually skittish because I think they had done their diligence and they were concerned about the price of the deal because they thought it was expensive, and I think, based on time line, it’s because they were probably hearing a lot of market rumours about BOSS was coming to the market … I thought that if they were really worried about it the thing that they would look at very closely was the marketing spend, so in order for them to actually see that we could, let’s say, improve Page 91 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 92 of 257 EBITDA one of the ideas probably they would say is “Hey, maybe you should actually reduce marketing” because in that same quarter that we just said that we’re spending a lot on market.”

¶ [261] It was suggested to Ms Chien that what she was talking about was being able to “manipulate the data by shutting off a particular expenditure, yes?”. Ms Chien firmly and confidently responded: “No, I don’t agree with your use of the word. I don’t think we were manipulating data. I’m just saying that in order to change the outcome, there are factors you can use or leverage you can pull, and so marketing obviously is an area. There could be other … you know, reducing headcount … This is just an example of what can be done if you want to improve your bottom line”.

¶ [262] It was put to her and “just as you could shut off the spigot to increase cash flow, you could open the spigot to reduce cash flows; yes?” to which she replied “Technically it works both ways, right? You can spend more money or less money, yes, of course.”

¶ [263] Ms Chien stated: “Lagou’s headcount, generally speaking, would probably be around 10% of 51 job.”

¶ [264] Ms Chien agreed that in the 2021 and 2022 Management Projections the Company was consolidated with Lagou and Zhiding and in the 20-F filing Zhiding had no headcount and its revenue and operating costs were minimal compared to the Company and Lagou.

¶ [265] It was put to Ms Chien that the difference between what they were budgeting for advertising and promotion expenses for 2022 in the Management Projections in 2022 and the budget that they were forecasting for the Company advertising and promotion expenses for 2022, plus the Lagou advertising and promotion expenses budget for 2022 there was a difference of RMB 252,128,000 that’s 914,731,000 minus 662,603,000 and she agreed that there was “that difference”. It was put to her that the difference was 38% higher in the projections than the internal budget which in the case of the Company she produced and in the case of Lagou she received at exactly the same time. She responded: “Yes, the input I received and provided at the time is different than what is actually in the ’22 projections, yes.” Page 92 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 93 of 257 Ms Chien agreed it was “apples to apples”. “We have agreed that the basis is the same”. Ms Chien was asked to explain the “enormous difference, 38% higher, from what you’re doing at the time, as it were, internally, to what you’re putting in the 2022 projections?”. The following appears in the transcript at pages 39-45 of Day 4: “A. Yes, I would like to explain that, if I can. Q: Right. A: Can I explain that? JUSTICE DAVID DOYLE: Yes, please. A. Yes, I think the key here is actually if you look at the 2021 numbers, as we've already discussed, as well as the 2022 inputs, there are always going to be some difference in terms of what the number is that we provide in January and what we ultimately end up using. So -- and sometimes, as we've said, we do update the budget, sometimes we don't. So -- but in terms of the basis, I agree that it's the same as we've actually already agreed, but the key here is -- very importantly is that (a) these numbers are absent of looking at the actual final revenue we're trying to achieve – JUSTICE DAVID DOYLE: Can you take this very slowly [for] me? A. I'm sorry. JUSTICE DAVID DOYLE: Could you take this very slowly [for] me? A. Okay. As I said, if we look at the 2021 inputs we had been shown, as well as the 2022 inputs that we had been shown, ultimately that doesn't necessarily tie exactly with what goes into the budget over the year, because there may be some changes in terms of what we decide to spend based on competitive situation. And here I just really want to highlight that marketing, unlike any other items almost within the company, is very much a competitive situation we have to assess. So at the beginning of the year, our team may think that, okay, last year we spent $10 and this year maybe adjust a little bit for inflation or some external factor maybe we'll spend, let's say, 11 or 12. That would be kind of a normal situation that they would start with. But in fact in marketing we really have to look at what our competitors are doing because we were in the very difficult situation where we had very intense competition. So a lot of what we're looking at -- you can start with a so-called Page 93 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 94 of 257 internal budget or internal input as a starting point, or almost like a placeholder, but at the end of the day when you really think about what you're trying to achieve is you're trying to actually win back customers, you're trying to make sure you maintain a share voice in the marketplace, so, therefore, you then need to go external to look at what your competitors are doing. So even though that, yes, the starting point that we had been giving to the team in January may look lower than what we end up spending, but in fact we actually really have to recalibrate and so, again, the difference you can see that, there is -- I acknowledge, that there is a difference between what was in January when we've shown these numbers when we said it was about 690,000 million -- 690 million, sorry, not thousand. 690 million, and what went into the 2022 budget, which is about 900 million, and that indeed is exactly what was important in terms of when we actually went back and looked at what our competitors are doing. So we can look at what BOSS was spending in 2021, and that was a really an important guideline for us to consider as we're trying to tinker with what we ultimately think that we will need to spend, because, again, this is a competitive situation. So I think our sort of starting point may not be what we end up with and we need to really think about that because the backdrop is that we were in a very intense competition in the marketplace, and in 2021 we had been losing market share and we had our apps pulled from the shelves for about four months' time. So in fact the second half of 2021 we were already underspending versus what we had wanted to spend. So we go into 2022, again we had to look for kind of signs on what our competitors were doing. And, yes, we put in a number that's a placeholder that we start with, which is roughly equal to what we actually spent in 2021, but then we had to recalibrate to think about what we actually may have to spend to achieve a right share of mine so that we weren't going to be again completely beaten by our competitors, which was our situation in 2021 and earlier when we said we were three to five X lower than our competitors. We didn't want that situation to happen again. MR ADKIN: Ms Chien – JUSTICE DAVID DOYLE: Can -- just. A. Sorry. Page 94 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 95 of 257 JUSTICE DAVID DOYLE: I'm a very simple soul but I need to understand your evidence, and I think what you've been asked for is for an explanation -- I'm going to put this in really simple terms -- this is what is coming across to for me -- in really simple terms there seems to be a 38% difference between the internal budget figures and the management projections. A. Yes. JUSTICE DAVID DOYLE: The evidence you've just given, do you say that's the explanation for that difference? A. Yes, as I said, we kind of used the -- the four line number I put in this actually a very rough kind of a starting point, because you'll see that -- you know, I only had, like, two days. I respond very quickly to these requests from the finance team, because they need just something to start with as a placeholder to kind of get the process going. I didn't want to hold up that process. But in terms of when we look at and said what would we need to actually spend, we were then trying to get market intelligence and think about what our competitors are doing. And then when we actually put into the final projections, we did put in a number where we were calibrating versus market versus using a placeholder number that I was basically lifting from the 2021 actuals. MR ADKIN: Ms Chien, with respect, that isn't an answer to the question. Let's take it in stages. You were producing both sets of documents at exactly the same time. We've agreed that; yes? A. I agree with that. Q. You would want your 2022 internal budget to be accurate as they could be, wouldn't you; yes? A. For the internal budget, I really wasn't paying as much attention to some of these because I was just asking to do four items, rather than review the budget in its entirety. So when I was providing the four inputs, the four lines, if you will -- again, this is in isolation -- I hadn't looked at or confirmed what the revenues levels would be. I hadn't looked at what our other cost items would be. I hadn't had a chance to look at the headcounts. So, yes, I provided these numbers and they looked lower, I agree. I'm not disputing that. I'm just saying that the four numbers were provided Page 95 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 96 of 257 in isolation without looking at the whole budget, and it didn't really tie to anything else. Q. You -- forgive me. You never went back to change the budget numbers, did you? A. No, we did not. Q. No. Never took the management projections numbers that you put in for advertising and promotion spend and put those into your internal budgets; yes? A. We did not, and we did not do that for 2021 either. Q. Even though, as we've seen earlier, you did adjust the budgets as you went along; yes? A. Sorry, I didn't hear -- Q. Even though -- A. -- the question? Q. -- as we saw earlier, you did adjust the budgets to produce latest estimates; yes? A. I think that when we looked at it was one quarter that we made an adjustment. Q. Yeah. A. I don't think we did it for any other -- Q. No, and -- A. -- so we didn't have a practice of doing that. Q. -- it's also part of your evidence, isn't it, in your witness statement that you adjusted the budgets as you went along? A. No, I said we don't usually adjust our budget as we went along. We may but we don't usually.”

¶ [266] Mr Adkin suggested to Ms Chien that “you have adjusted the spigot and manipulated – whether deliberately or unconsciously, you have manipulated the figures in the ’22 projections because you realised that if the ’22 projections remained the same as 2021 projections the deal would not go through. That’s the position, isn’t it”. Ms Chien held her ground and with apparent calm responded “No. I do not agree with that.”

¶ [267] Ms Chien was asked whether in section K of her affirmation she had set out all of her financial interest in the merger and she responded “I believe so, yes.” Mr Adkin suggested to Ms Chien that the 3.4% of the common shares in the Company was wrong and it was 2.11%. Ms Chien responded “Yes, that’s Page 96 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 97 of 257 correct, but I think that calculation may have been that it includes some beneficial shares from the options.” Ms Chien agrees that it might be that she included in the 3.4% the 20,000 ADS and 820,000 vested exercisable options. Ms Chien appeared to agree that she rolled over roughly 1.4 million common shares resulting in her owning 3.86% of the surviving company.

¶ [268] Ms Chien says that she did not mention in her affirmation the new incentive programme that was implemented as part of the merger “because that hadn’t happened yet”. Ms Chien said that potentially it was possible for her to double her shareholding and was asked whether if you start with the correct 2.11% figure more than triple it “Yes, potentially, it’s all possible.” Ms Chien also accepted that she stood to receive 1% of the restricted stock units “Yes, these are all possible scenarios … they’re tied to the service conditions as prescribed in the agreement … if the deal didn’t close and the company continues to be a public company, I would not get these shares, but I would continue to earn options as company executive as per the original 51 job agreement as a public company.” It was put to her that when the RSU entitlement vested “that’s $22 million worth of free shares” she responded “No, again I think we have to look at the service conditions.” She accepted that provided she can complete the so- called service conditions she gets “the RSU’s for free”.

¶ [269] Ms Chien was also cross-examined as to the dual class share structure being created post merger (DA 673/66) and on the class B shares she was going to be entitled to 2.2 votes per share as opposed to one vote for the class A shareholders and she agreed that was the case.

¶ [270] It was put to her that even if the shares she held had been the same her shares would be worth more than double in terms of voting percentage and she responded “Yes, they would have a higher voting percentage if you will.” She agreed that such improvement would not have been available had the merger not closed but “I’m still a minority shareholder”. She agreed that she had not mentioned this or the fact her percentage as a shareholder was increasing from 2.11% to 3.86%, or the incentives package, or the increase in voting control rights associated with her shares would be significantly increased, in her affirmation. “I do not mention that because they’re actually not something that has actually happened as a matter of fact, so hypothetically it’s possible but, again, hasn’t happened”. She was asked whether she accepted that if a potential contract is going to happen under which she may obtain valuable options which may generate large amounts of money that that is a financial interest even though the ultimate event may not happen. Ms Chien responded: Page 97 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 98 of 257 “Well, I’m not sure I can fully agree with that because, again, these are things that may happen. I wouldn’t be able to describe exactly how it would happen or if it would happen, so I guess you could say that if the Company had stayed public I would have continued to receive options or be paid bonus as well”. Ms Chien said it did not occur to her that when she was drafting the financial interest part of her statement she would be giving a misleading impression by omitting these things “because all of this information is publicly available. I don’t seek to hide it or omit it. Again, I think what I was trying to focus on was actually stated on the facts as I knew it and which are certain. So other events, which are probable perhaps or hypothetically possible is not something that I had put into that statement.”

¶ [271] Ms Chien was cross-examined as to the COVID lockdowns in PRC.

¶ [272] Ms Chien agreed that the plan was to reduce the headcount at Lagou. It was put to her that she had blended the Company’s headcount and Lagou’s and if she had used the specific figures for each company “the overall employment costs would have come in lower”. Ms Chien responded “Yes and no at the time” as “there are actually severance costs to be paid”. It was put to her that in the relevant email exchange (DA 504TC/3) she was talking about “natural wastage” to which she responded “Yes, that is true if they could go according to my plan … in order to minimise our costs, although that may not be something we can achieve … I’m just saying that we didn’t actually take that approach to projections.”

¶ [273] In a WeChat message (DA/2/1) Ms Chien had referred to her view that DCP and Ocean Link who had initiated the NDRC review were hurting the Company: “I am increasingly disappointed and angry with their irresponsible behaviour”. Ms Chien agreed that the review delayed the take-private process and “in general I think the fact that the process was drawn out, it hurt the Company, yes.”

¶ [274] Ms Chien was cross-examined as to the “disclosure’ or rather discovery process and I note all her evidence in that respect.

¶ [275] In re-examination Ms Chien was taken by Mr Imrie to form 20-F (DA 225/23) and the risks related to doing business in the PRC. Page 98 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 99 of 257

¶ [276] Ms Chien said in respect of the 2022 Management Projections “[w]e took into … account some of the laws that came in into 2021, we had more visibility so we knew that that would actually change some of our cost structure in terms of adding people for compliance reasons or whatnot. But this is an ongoing exercise, so that’s something that we’ll continue to monitor, but some of that information was actually accounted for in the ’22 projections based on increased headcount or some of the extra additional costs that would be required to meet new regulations that were introduced.”

¶ [277] There was reference to DA 356/17 and the assessment of the potential impact of new regulations (DA 356/18).

¶ [278] Ms Chien was asked as to the approach to the internal budget on the one hand and the projections on the other and responded: “the internal budget … is more of a bottoms-up process but we just have team leaders input certain information based on the way they understand it and that includes, as we said, you know, the revenue definition which is different. And then for the projections, this is more of an external recognised revenue, so we have to make changes to how we account for everything so that there would be some differences in terms of how the information is collected, the basis of some of the inputs and definitionally speaking the way people understand internally would be different.”

¶ [279] Ms Chien was asked whether the incentive programme was playing in her mind when she was preparing the 2022 Management Projections and she responded: “No, I wasn’t thinking about that at all.”

¶ [280] On the whole I found Ms Chien’s evidence believable. I gained the impression that she was a human being of integrity and competence and was endeavouring to assist the court rather than mislead it. Her evidence in respect of the auto-delete function and the reasons for the differences in the 2022 Management Projections and the internal budget was not ideal but human beings rarely if ever achieve perfection when giving evidence in a court of law. Page 99 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 100 of 257 Mr Robert Anthony Bartell’s evidence

¶ [281] Robert Anthony Bartell (“Mr Bartell”) from Kroll, LLC was called by the Company to give evidence. I have to say that during his cross-examination Mr Bartell came across as an uncomfortable and unimpressive witness but in fairness that may in part be because he was being cross-examined in respect of matters of which he had no direct knowledge. His evidence in respect of the Kroll document retention policy was however deeply unsatisfactory and he provided little assistance to the court as to the real reasons behind such a policy. I think Mr Bartell wisely recognised he was in difficulties as at one stage he made the unusual request part way through his evidence for permission to speak to his lawyers if it became necessary.

¶ [282] Mr Bartell appears to have been fielded by the Company as the person who signed the fairness opinion on behalf of Kroll but it transpired he had no direct contact with representatives of the Company and was unable to give direct evidence in respect of many of the questions put to him in cross-examination by Mr Lowe.

¶ [283] Mr Bartell says that he is “a Vice Chairman and Managing Director of Kroll, LLC (formerly Duff & Phelps LLC) and CEO of Kroll Securities, LLC (formerly Duff & Phelps Securities, LLC), a FINRA registered dealer (“Kroll”)” and he says he makes his affirmation on behalf of Kroll as the independent financial advisor to the Special Committee of independent directors of the Company. He says he is a Chartered Financial Analyst and holds FINRA licenses but is not a qualified accountant. Mr Bartell appears to be based in Chicago. He is providing factual evidence rather than expert opinion evidence.

¶ [284] He says that “Kroll does not create its own projections. Rather, Kroll obtains projections created by the subject company’s management. Per the engagement letter, Kroll does not independently verify or audit that information. In addition, Kroll requires a management representation letter to evidence our understanding of the information provided.” Kroll’s proposed fee to act as financial adviser and for the fairness opinion was US$1.25 million which they subsequently adjusted down to US$1 million. Mr Bartell says that no part of Kroll’s fee was contingent on the conclusions reached in the fairness opinion. The Kroll team was led “by David Lu on a day-to-day basis and ultimately as me as the most senior Managing Director on the engagement.” Matthew Schwartz a director in the Chicago office was also part of the Kroll project team and Mr David Lu, based in the PRC, served as the local client contact. Page 100 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 101 of 257 Mr Bartell says that as “the most senior Managing Director on the engagement my role was to represent the firm in executing the engagement letter and reviewing the analysis and opinion letter before signing the opinion letter on behalf of the firm.”

¶ [285] Mr Bartell says that on 7 May 2021 Kroll received the Company’s 2021 Management Projections and the project team received them “and considered them in light of historical performance, equity research analysts forecasts, and comparable company forecasted growth rates” and “[a]s is our normal procedure, following our initial review of the 2021 Management Projections, members of the project team discussed [them] with the Company”.

¶ [286] Mr Bartell provides evidence as to the scope of Kroll’s 2021 fairness analysis and the 2021 fairness opinion in connection with the Original Proposal. Mr Bartell at paragraph 45 of his affirmation set out some of the documents the project team reviewed including “other internal documents relating to the history, current operations, and probable future outlook of the Company, as provided by the management of the Company.”

¶ [287] From paragraph 49 of his affirmation onwards Mr Bartell refers to Kroll’s “2021 Discounted Cash Flow (“DCF”) Analysis.”

¶ [288] From paragraph 55 of his affirmation onwards Mr Bartell refers to Kroll’s “2021 Selected Public Companies and Selected M&A Transactions Analysis.”

¶ [289] Mr Bartell from paragraph of his affirmation 63 onwards gives a summary of the Original Proposal Valuation and at paragraph 66 of his affirmation says “the result was that each ADS was worth in the range of US$73.18 – US$86.94 and at paragraph 67 of his affirmation says “As the Original Proposal (US$79.05 per share/ADS) was well within this range, Kroll opined that the Original Proposal was fair from a financial point of view.”

¶ [290] In section J of his affirmation Mr Bartell covers “The 2021 go-shop process in connection with the Original Proposal.”

¶ [291] In section K of his affirmation Mr Bartell covers “Kroll’s 2022 amended engagement by the Special Committee in connection with the revised proposal and the final proposal.” Page 101 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 102 of 257

¶ [292] Mr Bartell says that on 12 January 2022 the Special Committee’s counsel emailed Kroll attaching a non-binding proposal letter dated 12 January 2022 “for the sum of US$57.25 per share or ADS (the “Revised Proposal”). Kroll ultimately analysed a final proposal “for the sum of US$61.00”. The Special Committee asked Mr Lu for an updated fairness opinion. Kroll was to be paid an opinion fee of US$1.5 million and an incentive fee. Kroll provided a financial analysis and an opinion letter opining that the final proposal at US$61.00 “was fair from a financial perspective.”

¶ [293] Mr Bartell at section L of his affirmation covers Management’s 2022 Management Projections for the Company. Access to them was provided on 28 January 2022. Mr Bartell says that the 2022 Management Projections were reviewed and considered in light of historical performance, equity research, analysts forecasts, and comparable company forecasted growth rates.

¶ [294] Mr Bartell says that although he was not involved in the discussions David Lu informed him that the “project team discussed with management the changes between the 2021 Management Projections and the 2022 Management Projections and that there were clear market-driven reasons for negative changes between them”. Mr Bartell adds: “For example, Mr Lu informed me that management’s view was that: 98.1 they needed to spend more just to stay at the same revenue level; 98.2 there was increased competition in their sector; 98.3 the economy and job market in China was very uncertain given the slow pace of the post-COVID reopening.”

¶ [295] Mr Bartell referred to the 2022 Management Representation Letter.

¶ [296] In section M of his affirmation Mr Bartell deals with “Kroll’s 2022 fairness analysis and the 2022 fairness opinion letter in connection with the final proposal.” Mr Bartell says that the project team, amongst other matters, “took into account their assessment of general economic, market and financial conditions, as well as their experience in securities and business valuation in general, and with respect to similar transactions in particular.” Mr Bartell says: Page 102 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 103 of 257 “104 Unlike the 2021 Fairness Analysis, rather than concluding an enterprise value range from the two methodologies and then calculating the equity range based on the concluded enterprise value range, we presented the enterprise value/equity value/per ADS value ranges separately for each methodology. This change in presentation from the 2021 Fairness Analysis was the result of suggestions from our internal review process (in part, in order to make the presentation of the implied ADS value ranges resulting from each methodology clearer to the Special Committee). 105 In addition, the project team discussed the information referred to above and the background and other elements of the 2022 Management Projections with the management of the Company, as well as its plans and intentions with respect to the management and operation of the Company’s business. Mr Lu also discussed at a high level the 2022 Management Projections with the Special Committee at a meeting on 9 February 2022 [51JOB-00094641].”

¶ [297] From paragraph 106 of his affirmation Mr Bartell deals with “2022 Discounted Cash Flow (“DCF”) Analysis.” At paragraph 110 Mr Bartell said that “Kroll estimated the value of each ADS to range from US$56.38 to US$66.90”.

¶ [298] From paragraph 111 of his affirmation Mr Bartell deals with the “2022 Selected Public Companies and M&A Transactions Analysis.” At paragraph 121 Mr Bartell says that “Kroll estimated the value of each ADS to range from US$57.73 to US$68.37.”

¶ [299] At paragraph 127 of his affirmation Mr Bartell notes that “the Buyer Group subsequently accepted the Special Committee’s offer of US$61.00 per share or ADS (the Final Proposal).”

¶ [300] In section O of his affirmation “Presentation of the 2022 fairness opinion in connection with the final proposal” Mr Bartell says that on 1 March 2022 Mr Lu attended the final meeting of the Special Committee and made a presentation using the 2022 Fairness Analysis and on 3 March 2022 Kroll provided its executed 2022 Fairness Opinion Letter to the Special Committee and opined that “the Final Proposal was fair from a financial point of view.” Page 103 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 104 of 257

¶ [301] At section P of his affirmation Mr Bartell deals with: “The 2022 go-shop process in connection with the final proposal.”

¶ [302] Mr Bartell accepted that he had not known what David Lu “was doing day on day on this particular engagement” and he did not recall what any of the people who worked on the team were doing and when they were doing it.

¶ [303] Mr Bartell said he was “unaware of any documents that were not provided to the court.”

¶ [304] Mr Bartell appeared to accept that he did not have any personal knowledge of anything that was said at any Special Committee meeting “only through what was discussed in the proxy in the 13-E-3”. He did not attend any of the meetings with the management of the Company “I don’t speak Mandarin or Chinese, so no” and “We don’t have records or transcripts of interviews with our clients” and they do not have “notes of those interviews”; “We don’t refer to them as interviews. It’s part of our normal due diligence investigations”; and “I don’t know what goes on at a meeting that I wasn’t part of”. Mr Bartell said he did not have notes of any conversations with David Lu, “if I did, they’d be written in the emails.”

¶ [305] Mr Bartell said he did not know Mr Hsu.

¶ [306] Mr Bartell said that his hourly fees were being paid for his attendance as a witness: “It’s customary to support our opinions and our clients …”. “I’m not being paid personally, the firm is being reimbursed.” Mr Bartell accepted that “Kroll had to fulfil its obligation under this engagement in order to support its opinion …”.

¶ [307] Mr Bartell said that he has given testimony before and he has “been involved in appraisal rights matters similar to this one …”

¶ [308] Mr Bartell said that “David and I co-led this engagement” but accepted that the Kroll people who would interact with the Company would logically be Mr Lu and those who worked with him on the assignment. Mr Bartell accepted that Mr Lu was the most senior person on the team in the PRC. Mr Bartell accepted that he did not get any communications from the Company and neither did Mr Page 104 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 105 of 257 Schwartz. Mr Bartell accepted that Davis Polk was the legal adviser to the Special Committee. Mr Bartell says that he did not meet with the Company’s representatives and he did not communicate with them directly but he communicated indirectly “through analysing the company’s financial information and business information.” Mr Bartell accepted that he did not ask others to communicate with the Company on his behalf. Mr Bartell accepted that he was not “copied in on emails with the Company.”

¶ [309] Mr Bartell indicated that he did not recall from four years ago specific conversations: “I do not take notes”. It was put to Mr Bartell whether he would “accept that the PRC team would have – when they went to these meetings, would have made notes of what happened at the meeting [of the Special Committee] and he responded: “No, they don’t make notes pursuant to our company policy. I also don’t refer to my teammates as the PRC team, I refer to them as my teammates”. It was put to Mr Bartell “Well, you say it’s company policy not to take notes. Is that really the – is that really true?” The following exchange occurred: “A: We don’t preserve any notes and … that’s our company policy on opinions. Q: That’s not the question … we’ll come to your preservation policy – do you have a policy that people shouldn’t take notes at meetings? A: No. Q: No. And in fact it would be bad practice, wouldn’t it, to go to a meeting and just sit there and try and absorb everything and not take a note? Bad professional practice, isn’t it, not to take notes? Do you agree? A: I think that’s someone’s personal opinion .. It’s nothing to do with good professional practice …” “Q: What’s the point of having meetings with management to discuss projections if you’re not taking down information that tells you what the underlying assumptions are and projections, for example, if you don’t get new information? What’s the point of the meetings if there’s no information for you to absorb? A: There’s … information in every client discussion we have. Q: Yes, and wouldn’t you expect notes to be taken of those discussions? A: Yes.”

¶ [310] Mr Bartell said he didn’t know that RM was the same as Kilometre and that Mr Hsu was the principal of RM (Kilometre) and said he did not have interactions with them. Mr Lu was liaising with Mr Hsu. Page 105 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 106 of 257

¶ [311] Mr Bartell said attending a meeting in the PRC was not his role on the engagement. Mr Bartell was referred to paragraphs 41 and 42 of his affirmation and accepted that he could not tell the court what happened at a meeting he did not attend. Mr Bartell accepted that he did not know what was said at the meeting about the management projections.

¶ [312] In respect of the power point analysis Mr Bartell accepted that there was nothing “specific to the Company that [was his] work in this presentation”. Mr Bartell did not seem to know who had inputted the information “The template of the document is substantially similar to all of our fairness opinions, so the actual slides and format in deliverables are not unique to 51.”

¶ [313] Mr Bartell said that a “concluded range” “For a fairness opinion it is typical.”

¶ [314] Mr Bartell says that he does not have the draft of the fairness opinion: “Q. … you would have got a draft from Mr Lu and … you would have made some comments on it and marked it up. Do you remember what you did? A: No, I don’t remember exactly what I did.”

¶ [315] In respect of paragraph 45.5 of his affirmation and the reference to “internal documents” Mr Bartell was asked what “internal documents” he meant. His response was “I don’t know”. There was a similar exchange in respect of paragraph 101.4 of his affirmation.

¶ [316] Mr Lowe referred to Mr Bartell’s statements in his affirmation such as the project team did this, Kroll estimated, Kroll concluded, Kroll derived and “throughout that description of activities and analyses, review and so on, you’re not able to tell us, on any instance, who in particular did any particular exercise, because you have no notes and you have no records of particular documents that were marked up or sent out; is that correct?” Mr Bartell responded “That’s correct.”

¶ [317] Mr Bartell said that “Internal documents’ is a very ambiguous term, so I’m not sure what it relates to.” “Q. Not my language, your language. A: Yes.” Page 106 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 107 of 257

¶ [318] Mr Bartell initially said he recalled reviewing the 2022 Management Projections and when pressed stated: “I don’t recall reviewing – if I reviewed the company’s projections or Kroll’s discounted cash flow analysis which reflected the company’s projections. I don’t recall which one I looked at or if I looked at both … I don’t have any recollection of examining the company’s Excel spreadsheet but I may have …”.

¶ [319] In respect of the go-shop process Mr Bartell accepted that he was not personally contacting the buyers and he was not getting the responses from the buyers. “Q. … you don’t know why in your statement at paragraph 98 you say “Mr Lu informed me” because you say you might have got it from the 13E. You don’t know where you got this from now: is that correct? A: That is correct. I think probably a better description would be “Mr Lu” replaced with “the Kroll team”.

¶ [320] Mr Bartell accepted that it was not Kroll’s practice to verify or audit management projections “that’s not just our firm, that’s the market.”

¶ [321] Mr Bartell said that members of the Kroll team had discussions with management about the change in the projections. Mr Bartell was referred to Ms Chien’s evidence that there was no additional discussion with Kroll and it was put to Mr Bartell that Ms Chien’s evidence on this point was not consistent with what he says in his affirmation at paragraph 98 with respect to the 2022 Management Projections that there were discussions with management, whereas the Company’s CFO said that there were no such discussions. "A: I don’t believe that’s accurate. Q: But you don’t have any knowledge yourself of what discussions took place with management about the projections; is that correct? A: I don’t have any knowledge …” Page 107 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 108 of 257

¶ [322] Mr Bartell said it was very standard for projections to be compared but accepted that it was not “very standard to have a recut of a deal with such a big price change …”. Mr Bartell accepted that “this was an unusual situation where Kroll was concerned to find out why the price had changed to such an extent …”. Mr Bartell said he had a role in ensuring that the comparison took place and he remembers discussing it with Mr Schwartz or Mr Lu. Mr Bartell was referred to the eight or nine management explanations. Mr Bartell did not know who in the team collected these explanations.

¶ [323] Mr Bartell was referred to Ms Wu’s (of Davis Polk) question “Does [Kroll] think these management explanations are reasonable?” “Q. Did you see the concern that was being expressed by the lawyers on this engagement? Are you aware of this? A: Yes, but I would call it a question, not a concern … I’m aware that they said that. It’s highly unusual that they would ask that question.”

¶ [324] Mr Bartell accepted that there was a “rigorous analysis” about the revised projections rather than a “fairly fierce debate” internally.

¶ [325] Mr Bartell accepted that for this transaction overview they did not have a concluded range: “Q. And the reason is because what you’re going to do is you’re going to suggest a floor, rather than a range. And we see that Mr Lu, when he goes to the Special Committee, as you can probably remember, says it mustn’t be less than $61. A: Correct.”

¶ [326] Mr Bartell agreed that he was aware that discovery is important in Cayman Islands proceedings. There was reference to a communication from Davis Polk asking Kroll to ensure that notes were taken at Special Committee meetings. Mr Bartell accepted that Kroll kept an electronic file of all documents that was uploaded on the server.

¶ [327] Mr Bartell said that keeping manuscript notes “would be against our procedures and policy … we have a policy to not save notes in the electronic files.” Page 108 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 109 of 257 “Q. … do you have a policy against people taking manuscript notes? A: No … do I have a kind of a personal notebook? Well … some months I do and some months I don’t … I’m very careful on any assignment not to take notes which would be misunderstood or manipulated or misconstrued later on … Notes are taken in all engagements … but they’re properly discarded and destroyed … I think we’ve been advised by legal counsel, as our peers have been to not take notes [during meetings] because the deliverable is the opinion letter … that’s the only thing that the Special Committee is hiring us for …” Mr Bartell seemed at times to be unable to distinguish between taking notes, preserving notes and destroying notes.

¶ [328] Mr Bartell (in respect of DA 650.01.1) stated “this email is typically forwarded straight to our legal department’s office that then sends out a hold notice on the files and they also electronically lock the files at that time.” “I don’t remember receiving this or not.”

¶ [329] In respect of Kroll’s role Mr Bartell said: “Our interest is to protect the public minority shareholders”. “We’re acting for the Special Committee of independent directors” “the Company … pays the fees ….”.

¶ [330] Mr Bartell says “because the opinion is given by a FINRA broker dealer, you can’t delete or destroy anything in the servers, even if you wanted to … we don’t keep any hard copy notes on engagements …”. Mr Bartell accepted “that’s company policy”. “You mentioned earlier that notes would be destroyed, can you tell me how that works? Answer …. Put them in the shredder.” (FA/1/55)

¶ [331] Mr Bartell is asked how long the document retention checklist has been in existence: “To the best of my knowledge, as long as I can remember, but at least ten to 15 years, and it’s updated or kept the same annually in our fairness opinion compliance training.”

¶ [332] Mr Bartell said it was the policy to delete all notes that are saved on the file by the end of the project. Mr Bartell is referred to the words “Due diligence notes. Our official policy is NOT to keep notes for opinions. If you must keep them, review for accuracy and type them if necessary.” Mr Bartell was Page 109 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 110 of 257 asked “So that is pursuant to that policy you don’t keep any notes.” He responded: “Yes, with an important caveat to clarify. For the court, unless we’re under legal compulsion with a hold, then nothing can be destroyed.”

¶ [333] Mr Bartell was fairly asked what the purpose was of destroying due diligence notes and drafts of documents. The rather unsatisfactory answer to legal ears was: “… drafts are drafts, so there’s no reason to keep them because they’re irrelevant, and other notes are reflected in the analysis itself … No one’s hiring us to be a scribe … I don’t really know why there’s a policy … It’s all electronic now.”

¶ [334] “Q. Shall I suggest to you … in case anybody sued Kroll or Duff & Phelps for negligence, they would be unable to establish any sort of chain of documents that showed how you got to an end result. Is that why you destroy documents? A: No … Q: Do you tell your clients that you … will not keep any notes or drafts. Do they know that you’re going to do this or do you keep it a secret from them? A: We don’t keep it secret but we don’t tell them and I don’t remember ever being asked.” Mr Cheng’s evidence

¶ [335] Mr Cheng says that during the period 7 March 2013 to 6 May 2022 he was an independent non- executive director of the Company. Ms Chien recruited him to the Company. He was also a member of the Special Committee. He had not met Mr Yan or any other members of the Board or anyone from Recruit prior to joining the Board in 2013. Mr Cheng was also a member of the audit committee.

¶ [336] Throughout his time on the Board Mr Cheng’s view was that “the Company’s management prioritised the Company’s business and long-term strategy … That long-term focus was something I liked about working with the Company.”

¶ [337] On 17 September 2020 Mr Cheng received a text from Ms Chien asking for a phone call with her and Mr Yan which duly took place. He was told that the Company “had received an unsolicited offer from DCP to take the Company private.” Although no member of the Company’s management was involved Page 110 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 111 of 257 in DCP’s proposal Mr Cheng learned at the Board meeting on 20 September 2020 that DCP had indicated to RM that it intended to invite Mr Yan and Recruit to roll over all or part of their existing equity in the Company and therefore it was resolved that a Special Committee would be formed to consider the proposed transactions and possible alternatives and report to the Board with its recommendations. Mr He and Mr Cheng were appointed to the Special Committee on 20 September

¶ [338] Mr Cheng had experience with take-private transactions. E-House (China) Holdings Limited was taken private (having been listed on the NYSE) in 2016 while he was its COO and he says he was heavily involved in the process. He also participated as an independent board member and special committee member in the take-private transactions of Le Gaga Holdings Limited in 2014 and Country Style Cooking Chain Co., Ltd in 2016.

¶ [339] At paragraph 26 of his affirmation Mr Cheng says: “I understood … that the Special Committee’s main role in this context was to protect the interests of the independent or minority shareholders who were not rolling over their shares and were not part of the buyer consortium. This could mean, for example, negotiating specific terms to protect those shareholders’ interests or negotiating a higher price to maximise their returns. I understood that the Special Committee could, and had a duty to, reject a deal if we believed that the final offer price and key terms were not fair to the independent/minority shareholders.”

¶ [340] Mr Cheng said that he and Mr He each received compensation of US$50,000 per month for their services on the Special Committee and the payment was not contingent upon the completion of the Merger nor on their recommendation of the Merger.

¶ [341] Mr Cheng was not involved in, or aware of any early discussions with Mr Hsu or his company RM. Mr Cheng recalls asking Ms Chien about Mr Hsu’s role and who he represented: “She informed me that Mr Hsu represented the Company, not the buyer consortium or any member of it, nor did he represent the Special Committee …”. Page 111 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 112 of 257

¶ [342] Mr Cheng says the Special Committee were aware of the terms pursuant to which RM was engaged by the Company and added: “Although Mr Hsu or other representatives of Rocketeer Management frequently joined Special Committee meetings at our invitation in order to update us on recent developments and answer questions, they were never present during our privileged discussions with our advisors …”.

¶ [343] Mr Cheng said that the Special Committee had very little direct contact with the Buyer Consortium, save in relation to regulatory matters. Almost all of their communications were conducted through their respective advisers and through RM.

¶ [344] On 30 September 2020 the Special Committee formally engaged Davis Polk as its independent legal adviser and Kroll (formerly Duff & Phelps) as its independent financial adviser. Mr Hsu assisted in the process of appointing them by sharing contact details and making introductions.

¶ [345] Mr Cheng says that on 7 May 2021 the Company’s management sent a draft of financial projections prepared by them to the Special Committee, Kroll and Davis Polk. Mr Cheng says “We discussed the Original Management Projections with Kroll and Davis Polk at a Special Committee meeting on 11 May 2021 and resolved to approve (i) the distribution of the Original Management Projections to the Buyer Consortium and (ii) the use of the Original Management Projections by Kroll for purposes of its financial analysis of the Company.”

¶ [346] Mr Cheng said he reviewed the Original Management Projections but he says he left it to Kroll to conduct a close review. He says he was comfortable doing so because, based on his review there was nothing in the projections that appeared incorrect or unreasonable. He says he knew that Kroll would prepare a fairness opinion that would contain a range for fair value, which could therefore potentially be supported by a range of projections, and that they would not rely solely on the projections prepared by the Company’s management to do so, but would utilise multiple methodologies.

¶ [347] Mr Cheng said that the only item in the Original Management Projections that caught his attention was the increased marketing spend from the first quarter of 2021, which was more than double what it had Page 112 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 113 of 257 been prior to that quarter. He says that stood out to him but he says he was aware of the following points: (1) prior to 2020 the Company had enjoyed a market-leader position but during 2020 competitors such as BOSS began to spend large amounts of money on advertising and this was threatening the Company’s market share. He says those companies were backed by private equity and not publicly listed so they had an advantage in the sense that they could spend more on advertising (which may have been loss making) without having to account to public shareholders. Mr Cheng says he was aware that the Board believed that the Company needed to increase its advertising spend so as not to lose market share; (2) BOSS were investing aggressively in advertising and he could see the logic in the Company increasing its own advertising spend in response; and (3) Mr Cheng said he also took comfort from the fact that Kroll was aware of the increased advertising spending and considered it reasonable to incorporate this projection in its analysis.

¶ [348] Mr Cheng says that the Special Committee received a draft merger agreement, which would give effect to the Original Proposal, on 31 May 2021. They held a number of discussions with their advisers and RM during June 2021 and the key points in dispute between the Special Committee and the Buyer Consortium during the negotiations were: (1) the Company’s ability to undertake a “go-shop” process, post-signing; (2) the inclusion of a provision requiring a “majority of the minority” (“MoM”) vote; (3) a condition relating to the maximum percentage of shares for which holders could exercise dissenters’ rights; and (4) the inclusion of a provision regarding the Board’s power to change its recommendation regarding the Proposed Transaction. Page 113 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 114 of 257

¶ [349] Mr Cheng says that he and Mr He generally relied on Davis Polk’s advice in respect of these matters. However they had both been involved in several similar transactions, so they made decisions based on their experience and what they understood to be generally accepted practice.

¶ [350] Mr Cheng says that the Special Committee’s position initially was to insist on the inclusion of provisions providing for a go-shop period, a MoM vote condition, and the power of the Board to change its recommendation, and to remove the condition relating to the maximum dissent. The Buyer Consortium was initially resistant to these points.

¶ [351] Mr Cheng says at a meeting on 16 June 2021, the Special Committee decided to continue to advocate for these amendments to the draft merger agreement and they also instructed Kroll to ask the Buyer Consortium for an increased price. Mr Cheng says that his expectation was that the Buyer Consortium would be unlikely to agree to this, given the prevailing market conditions, such as a generally declining market valuation for “China concepts stock” (companies operating in the PRC, but listed elsewhere overseas), and competitive pressure the Company was facing, and the length of time it had taken the Buyer Consortium to form. Mr Cheng says that he and Mr He believed that it would be a good result if the price remained as it was and was not decreased, given the market conditions.

¶ [352] The Buyer Consortium did not agree to a higher price and did not agree to a MoM vote requirement. Mr Cheng says that the Special Committee ultimately decided to forego this requirement in view of the fact that all of the other amendments were accepted, and after considering advice that it was not required as a matter of law and considering that such a provision was not market standard. Mr Cheng says that in his experience such provision is not typically included, and the few cases where it is included are instances where the buyer group's percentage of the shareholder vote is so small that the provision is unlikely ever to have an impact and is symbolic only. Mr Cheng says that he expected from the beginning of the negotiations that the Special Committee could end up not pressing for a majority of the minority clause to be included.

¶ [353] On 21 June 2021 the Special Committee and Davis Polk met with Kroll to discuss, amongst other things, a detailed set of slides which Kroll had prepared to present its valuation analysis. Mr Cheng says it was clear from the slides that US$79.05 per share was, in Kroll’s opinion, well within the fair value range. Page 114 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 115 of 257

¶ [354] Mr Cheng says that Kroll’s analysis adopted a combination of methodologies, namely a DCF analysis and a comparison of comparable companies and merger/acquisition transactions. Mr Cheng says that this is consistent with a sound valuation exercise which should be based on a combination of approaches where each can be used to check the reasonableness of the others.

¶ [355] On 21 June 2021 the Special Committee resolved unanimously to recommend to the Board that the Company take the necessary steps to proceed with the Merger at a purchase price of US$79.05 per ADS.

¶ [356] From paragraphs 69 to 72 of his affirmation Mr Cheng deals with the go-shop process. It appears that the Special Committee instructed Kroll to deal with this process. Of the 12 potential investors contacted by Kroll, half did not respond and the other half declined to make an offer. Mr Cheng said this did not surprise him.

¶ [357] In section VIII of his affirmation Mr Cheng deals with the regulatory issues raised in the second half of 2021. Mr Cheng did not believe that the concerns raised by DCP and Ocean Link in respect of the NDRC were genuine and he had the impression that “they were using the regulatory changes as an excuse to justify their not proceeding with the Proposed Transaction, or to negotiate a lower price.”

¶ [358] Mr Cheng says he was also aware that some of the Company’s shareholders were raising concerns around the time about the lack of progress and updates on the Proposed Transaction. For example, on 27 September 2021 Linda Chien forwarded on an email to him and Mr He from MDS and sought guidance on how the Company might manage shareholder communications regarding the Proposed Transaction and recorded her sense that the public shareholders were becoming increasingly frustrated at the flack of information.

¶ [359] Mr Cheng noted a letter dated 21 September 2021 from the Company to the Buyer Consortium advising them that the Company believed the delay in consenting to the filing of the 13-E-3 Statement constituted a breach of the Merger Agreement and Davis Polk sent a further notice to the Buyer Consortium on 30 September 2021 reiterating their position and recorded the Company’s disagreement that the Proposed Transaction gave rise to any requirement to notify or consult with the NDRC.

¶ [360] Mr Cheng stated that the Special Committee spoke with representatives of DCP and Ocean Link on 13 December 2021 but the call was not productive and no progress was made on the regulatory issues Page 115 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 116 of 257 point. This was the only instance Mr Cheng can recall of the Special Committee having direct contact with the Buyer Consortium.

¶ [361] The Special Committee obtained legal advice from Davis Polk as to whether the Buyer Consortium could be compelled to complete the Original Agreement and they considered taking action in this regard but were concerned that this would not be a good use of time or money (and therefore was not in the interests of the Company’s shareholders) as they felt it would be difficult to prove that the regulatory approval was not required. This was discussed at Special Committee meetings on 19 and 26 November

¶ [2021] Mr Cheng says that ultimately this was not necessary as an alternative proposal was prepared in December 2021 and settled in January 2022. In general terms this involved an amendment to the structure whereby the stakes of DCP and Ocean Link were lowered from around 20% to 9.9% such that they no longer appeared to be “driving” the Proposed Transaction. Mr Cheng says they apparently then considered that they did not need to initiate any further engagement with the PRC regulators, and agreed to withdraw the informal consultation request that they had made.

¶ [362] In section IX of his affirmation Mr Cheng deals with the Revised Proposal. This was received on 12 January 2022 with a reduced purchase price of US$57.25 per share. The relevant draft documentation was provided on 20 January 2022. The Special Committee held a number of discussions with their advisers during January 2022 in respect of the Revised Proposal and the terms of the draft transaction document. In addition to price the key points that the Special Committee raised during negotiations were (a) a go-shop process; (b) the termination fees and (c) the closing conditions and covenants in respect of regulatory filings.

¶ [363] Mr Cheng says the Special Committee relied on Davis Polk’s advice and where relevant, their own experience of similar transactions. They were successful in having a go-shop right added, increasing the termination fee and a tightened regulatory covenant. Mr Cheng says, based on the prevailing market conditions, their focus was “on closing the deal expeditiously whilst ensuring fair value.”

¶ [364] Mr Cheng says that the Special Committee instructed Kroll to conduct a further financial analysis of the Company and to provide a further fairness opinion. Mr Cheng recalls that, during 2022, he and Mr He discussed the possibility of appointing a new financial advisor “as we knew it would likely be difficult for Kroll to issue a second opinion after they had already expressed the opinion that a price of US$79.05 per share was fair. We also discussed this with Ms K Chien. However, we believed that we Page 116 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 117 of 257 had no choice but to continue to retain Kroll’s services as we did not want there to be any suggestion that we had “shopped around” for a financial advisor who would give an opinion that would be more likely to support the Revised Proposal.”

¶ [365] On 28 January 2022 the Company’s management sent Kroll updated financial projections.

¶ [366] The Special Committee held a meeting on 9 February 2022 with Kroll and Davis Polk at which Kroll reported its analysis of the Revised Management Projections and expressed its preliminary view on the purchase price but did not express a view on what a fair price might be at that stage. Mr Cheng reviewed the projections but left it to Kroll to conduct a close review. At the meeting the Special Committee instructed Kroll to communicate with the Buyer Consortium to indicate that the purchase price should be increased which Kroll did on 11 February 2022 but on 15 February 2022 the Buyer Consortium declined to offer an increased purchase price.

¶ [367] The Special Committee had a meeting with “our advisers (Kroll and Phelps and Davis Polk) and [RM] on 16 February 2022, at which Kroll provided an update on its financial analysis of the Company.” Kroll still did not provide any indication as to what a fair price might be, other than to express an equivocal view that it would need to be at least US$60 per share. As there appeared to be time- sensitivity with the relevant documentation due to be signed on 24 February 2022, the Special Committee instructed RM to propose a purchase price of US$63 per share to the Buyer Consortium.

¶ [368] The Special Committee considered that there was a real risk that there would be no deal if matters were not progressed quickly. The Special Committee did not think a no-deal outcome would be in the interests of the minority shareholders, provided they could get a purchase price of more than US$60 per share, because a no-deal option was not desirable given the worsening market conditions. The Special Committee believed that they could not accept a price of less than US$60 per share because, otherwise, regardless of Kroll’s opinion they considered the drop in price to be too severe. Mr Cheng said that several investors had also contacted the Company expressing concerns about the reduced offer price and urging the Company and the Special Committee to push back on the price and protect the interests of the minority shareholders. The Special Committee did not feel able to advocate for the purchase price contained in the Original Proposal. Page 117 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 118 of 257

¶ [369] Mr Cheng says that on 28 February 2022 “we held a meeting with our advisors and with [RM]” and Kroll expressed a view that the price should not be less than US$61 per share. The upshot was a Special Committee instruction to RM “to propose a purchase price of US$61 per share or per ADS to the Buyer Consortium on a ‘take it or leave it’ basis”. The Buyer Consortium initially rejected this but ultimately later that day it agreed to US$61 subject to certain conditions.

¶ [370] The Special Committee held its final meeting on 1 March 2022 and representatives of Kroll, Davis Polk and RM were present at the meeting at the invitation of the Special Committee. Kroll presented its fairness opinion and provided “a range of fair value, based on a DCF valuation, of US$56.38 to US$66.90 per share, and, based on an analysis of comparable companies and merger transactions, of US$57.73 to US$68.37 per share”. At the meeting Mr Hsu reported that the Buyer Consortium had agreed to a purchase price of US$61 per ADS subject to certain conditions.

¶ [371] Following the Special Committee meeting a Board meeting was held on the same day and Davis Polk, on behalf of the Special Committee, reported to the Board that the Special Committee had determined that the Revised Proposal was fair and in the best interests of the Company and the unaffiliated shareholders and had resolved to recommend that the Board take the necessary steps to proceed with the Merger at a purchase price of US$61 per ADS.

¶ [372] Mr Cheng says that there was not one single factor that led him to conclude that the purchase price was fair but the factors on which he relied included: (1) the price was within Kroll’s fair value range and Kroll had confirmed that a price of US$61 per share would in its opinion be fair; (2) no alternative expressions of interest for a topping bid had been received; (3) the challenges the Company was facing including (a) increased competition in the PRC human resources industry; (b) challenges arising from macro-economic conditions and general market conditions for PRC companies in the US; and (c) the adverse impact of COVID-19 and the measures imposed by the PRC government in respect of it; (4) the trading price of the Company’s ADSs which had substantially decreased and historic trading prices; and Page 118 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 119 of 257 (5) the Special Committee had tried to obtain a higher price but was not able to do so.

¶ [373] Mr Cheng said that all these factors together suggested to him that the Buyer Consortium agreement of the price of US$61 per share was fair and he would not have recommended the deal if he thought otherwise.

¶ [374] A go-shop process was conducted from 2 to 16 March 2022 by Kroll on behalf of the Special Committee. Kroll contacted 11 potential investors, nine of which did not respond and two of which declined to make an offer. Mr Cheng says he was not surprised at the outcome.

¶ [375] Mr Adkin cross-examined Mr Cheng at length.

¶ [376] Mr Cheng appeared to be an experienced, competent, intelligent and honest individual with no real personal axe to grind. He struck me as an individual who was aware of his duties and was not simply going through the motions or acting simply to please Mr Yan and Ms Chien or simply doing Mr Yan’s bidding. He gave evidence in a precise and considered way. To take just one example see Day 5; page 155, line 21 of the transcript where he corrected Mr Adkin for putting it to him he learnt of the offer in a text when his evidence was he got a text from Ms Chien asking for a phone call and he learnt of the offer in that phone call. He paid attention to detail.

¶ [377] Mr Cheng assumed that Mr Yan would join the Buyer Consortium “because in China you almost never see an outside financial investor without involvement – without the support of the company’s founder and large shareholder … to attempt or completely take over a company. That just – no, it almost never happens.”

¶ [378] Mr Cheng said he would also have been highly surprised if Ms Chien did not participate. Mr Cheng was not concerned that his prior contact with Ms Chien and acquaintance with Mr Yan might become problematic if he ended up negotiating against a consortium of which he assumed they would become members. Mr Cheng accepted that he had known Ms Chien for several years mostly “in a work-related environment” and occasionally socialized with her. He did not know Mr Yan before he joined the Board. In response to Mr Adkin’s question as to a potential conflict Mr Cheng stated: Page 119 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 120 of 257 “I was an independent director, and in my capacity as independent director it was my job to do such negotiations. And the fact I had been a board member for a number of years has no impact on that. It’s the same role I played in the other companies, take-private transactions in which I was an independent director – I had been an independent director for a number of years.”

¶ [379] Mr Adkin put to Mr Cheng various communications from shareholders (including Surveyor Capital (DA 73/1), Hel Ved Capital (DA Tab 74), C Bus (to Recruit) (DA 116/1), JP Morgan Asset Management (DA 164/1), Aberdeen Standard (DA 169/1) and MFS (DA 152/1)) complaining that the original offer undervalued the Company. It appeared that Mr Cheng was not aware of these communications and did not respond to them. Mr Cheng accepted that he did not know the individual names of the “major unaffiliated shareholders [he was] representing” “I knew approximately how … many percentage … but I did not know the individual shareholders”.

¶ [380] The Linda Chien email of 20 November 2020 to the Special Committee in respect of a communication from a shareholder (DA 153/2) was put to Mr Cheng. Mr Cheng commented “As the privatisation process moves along, you may get more such emails. If it’s from an individual shareholder, my suggestion is not to respond or offer to forward their email to the board.” Mr Cheng was asked whether he advised LiMsnda Chien just to ignore it: “Yes, because we could not, you know, provide any information to individual shareholders without, you know, making the same disclosure to everyone else.” Mr Adkin put it to Mr Cheng that “[y]ou could at least invite representations, the provision of information, the provision of analyses, the provision of advice from the people that you are representing.” It was put to Mr Cheng that his attitude was “not to engage with them or even invite representations from the people you were representing but to stonewall, to ignore it.” Mr Cheng responded: “I wouldn’t characterise that as a stonewall or ignore, because we … were engaging professional advisers to analyse the merits of the transaction, and we were all waiting for that trial result of that analysis, and we would use that as a basis for – to evaluate.” Page 120 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 121 of 257 Mr Cheng denied that he was not interested in the opinions of the people whose interests he was meant to be representing: “It’s not that I wasn’t interested. It was that at any time you have a transaction proposal like this you will get … you will hear from shareholders saying that they think the price is too low. They always say that. But whether it is too low or … far too high, you know should not depend on individual investors’ view. It should depend on professional detailed analysis, evaluation … their response at this juncture of the transaction was typical. As long as an offer was not higher than the highest historical price of the … Company shares, there would always be some shareholders who said it was too low.” Mr Cheng said he was representing the unaffiliated shareholders’ interests.

¶ [381] Mr Cheng agreed that Mr Hsu was engaged by the Company and not by the Special Committee and that Mr Hsu or his advisory company stood to earn a fee tied to the success or completion of a transaction. Mr Cheng accepted that it would have been “against Mr Hsu’s interests for the Company or the Special Committee to walk away … That’s something which would be financially – he had 22 million reasons to avoid.” Mr Cheng classified Mr Hsu’s role as follows: “I viewed him as a – more of an intermediary or a go-between for negotiations between Special Committee and the buyer group.”

¶ [382] Mr Cheng accepted that he was not having discussions with the Buyer Group: “In some cases it was Duff & Phelps. For instance, the price demand it was passed on sometimes through an email from Duff & Phelps … to the buyer group’s counsel … and sometimes it was through Kilometre, through Mr Hsu … On legal terms, it was Davis Polk … some of the discussions were done between the two … sides’ legal advisers.” Mr Adkin said “Let’s confine ourselves to the price … that wasn’t discussed by the lawyers. Was there anybody other than Mr Hsu who was discussing the price with the buyer group on behalf of the Special Committee?” A: “No.” Q: “Was there anybody other than Mr Hsu from whom the Special Committee received information about the position of the buyer group in relation to price, for example?” A: “No.” Q: “No. So you were very much dependent for your information and very Page 121 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 122 of 257 much dependent for the negotiations on Mr Hsu: yes?” A: “I thought having him was facilitating the negotiation better than in … previous take-private transactions I was involved in …” Q: “And the problem with that, of course, was that Mr Hsu was not somebody who was truly independent. He had an enormous financial incentive in this deal, yes, to get a deal done at any price, yes?” A: “He has incentive to get a deal done, but that doesn’t make him not independent: right? … his role was no different from just any other financial adviser in another potential transaction … Their job was to advise and to help negotiation. Yeah, they stood to gain a fee – to earn a fee in the event of successful transaction …”. Q: “He’s not independent is he? He’s not somebody who should be negotiating because the absolute last thing he is ever going to do is say “no deal” …” A: “The decision whether there would be a deal or no deal would be made by the principals, namely the buyer consortium and the Special Committee … so Mr Hsu doesn’t get to decide whether there’s a deal or not.” Q: “No, but he does negotiate and he is a source of information … Will you accept that that was an inappropriate role for him to play in circumstances where you … the Special Committee was meant to be representing interests of the unaffiliated shareholders? Would you accept that?” A: “No, I don’t accept that that … was an inappropriate role for him to play.”

¶ [383] Mr Cheng was referred to a message on 20 September 2020 with Mr He (DA670 T/4 and 5). There was reference to Mr Hsu and Mr Cheng’s comment “[b]ut not sure if it’s a good idea to have too much communication with him. Potentially this is subject to discovery.” “Q: Now, what you were referring to there was that if you communicated with Mr Hsu in writing that would potentially be subject to discovery and litigation later; yes? A: I think – I need to see this to – to recall there was such a communication. This was September 20 right? So at the time I was – still wasn’t clear exactly who Mr Hsu was and what his role would be in the transaction, so I wasn’t sure whether I should talk to him at all.” Q: And would you accept that it does look like your instinct at that time was to try and avoid the generation of documents that could be subject to discovery and litigation later? A: No, my thought was any communication should be, you know, through email, which was more official and, you know, can more easily be kept as records, rather than through, you know, social media platforms and chats, stuff like that”.

¶ [384] Mr Cheng was closely cross-examined in respect of the issue of a MoM vote requirement: Page 122 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 123 of 257 “Q. … Now, a majority of the minority clause is a provision which makes the commission of a deal conditional on the deal achieving a vote in favour of more than 50% of the unaffiliated shares; yes? A: Yes. Q: And, in other words, it enables the unaffiliated shareholders to decide by majority vote amongst them whether or not they wish the deal to proceed; yes? A: Yes. Q: It is the surest protection, is it not, for the interests of unaffiliated shareholders in a deal such as this because it places the fate of the deal in their hands by majority vote; yes? A: In theory, yes, but in practice … it’s extremely rare to have that term, and to the extent that the majority of the take-private transactions I’m aware of did not happen at the price that was higher than the recent historical high … there will always be some shareholders who are not happy and to have that clause … in theory to have that clause in would … raise the hurdle for a deal to consummate, and the vast majority of the … take-private cases not end up having that clause.”

¶ [385] Mr Cheng asked Davis Polk about a MoM clause and their answer was “this clause is extremely rare among all the … take-private cases and it happened only … couple of cases where such a clause – the existence of such a clause would not have made a difference one way or the other.” Mr Adkin put to Mr Cheng that the buyer group in this case had a substantial majority of the voting stock “So surely this was a case where, if a protection was going to be afforded to the unaffiliated shareholders, i.e. people who weren’t in the buyer group on the buyer side, they would need a majority of the minority vote because otherwise they were very likely to be outvoted by the buyers”. Mr Cheng responded: “… this is the more typical shareholding structure among the dozens of take-private transactions. In other words, the buyer group includes the founder and the largest shareholder, together with some other participants, yet the buyer group very often exceed 50% of the total voting stock. But I was advised by my lawyer in none of the take-private transactions with this type of shareholding structure you ended up having the – majority of minority condition. Page 123 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 124 of 257 Q: And so acting on that advice, you just didn’t … press for it? A: We demanded it but we … expected the buyer group to resist that, to push back vigorously and it ended up using that as a negotiation tool to get the other conditions … it just didn’t happen in the market as a general market practice.”

¶ [386] Mr Cheng accepted that in the event the deal was not done at the original price and the PE sponsors (DCP and Ocean Link) delayed and by November 2021 the market was desperate for information. Mr Cheng was referred to the November 2021 Press Release. There was reference to the drops in the share price thereafter: “Q. … the reason the market was sensitive to the press release was because the merger had been stalled for months and the market had been given no information as to why; yes? A: It was more complicated than that.” Mr Cheng sought to explain the “bigger background” and made the following points, amongst others, in a lengthy response: - the overall market sentiment in the PRC internet company had deteriorated in a big way; - there was the “crackdown on the education company” that wiped out 80% of the market cap of two … very reputable listed companies; - Alibaba got into trouble with the PRC government and their share price fell by 40%; - Didi, the largest ride hailing company in the PRC did an IPO against the advice of the government and their share price fell by half; - on 51job the fact that there was total silence after June or July when the market was expecting just a routine announcement of a 13E-3 filing and then a company EGM caused investors great anxiety; and - difficulties arose with the NDRC point raised by the PE funds “So we, on the one hand, wanted them to drop this consultation altogether and push the transaction forward, but we were also nervous, if you … if we push them too hard, pushed the fund managers into a corner and they really decided to back out of the transaction altogether” or they could lodge a formal application Page 124 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 125 of 257 with the PRC regulator “the whole process would have been brought to a complete stop until and unless you get an affirmative response from the Chinese regulator.” “That was … the last thing we wanted to happen because we knew the market was … in much worse shape than before … this potential transaction at $79 was looking increasingly expensive in the eyes of investors – in the eyes of the fund managers … we said the market needs something, we need to tell the market something. So we drafted a press release and they came back … their lawyer edited it down. So the final version that went out was just completely vaguely worded …”.

¶ [387] Mr Cheng appeared to agree that the November 2021 Press Release did not make it clear to market whether the problem was that the Company was in trouble with the PRC government or whether it was the transaction that was in trouble with the PRC government.

¶ [388] It was put to Mr Cheng that his first instinct when the Revised Proposal was received by the Special Committee on 12 January 2022 with the much reduced price was not to say “[w]ell, if Kroll are unlikely to say the new price is fair, then we shouldn’t do the deal” but the first instinct was to instruct a new fairness opinion provider. Mr Cheng responded: “A. No, our instinct was – we needed a fairness opinion to support the transaction. Q. … Why didn’t you need a fairness opinion to tell you whether or not you should do the deal? A. That was why – that was the reason we had financial advisers to give us professional opinion, analysis – based on analysis. Q. No, Mr Cheng you said: “We needed a fairness opinion to support the transaction.” Is that your approach? Is that what you though you were getting a fairness opinion for? A. Yes, without a fairness opinion we could – we would not proceed with the transaction. Q. Did you understand that the function of obtaining a fairness opinion was not simply to allow the transaction to proceed or to support the transaction but to inform you as to whether, having regard to the interests of the people you were representing you should enter into the transaction at all. Was that not your understanding, Mr Cheng? Page 125 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 126 of 257 A. It was. The fairness opinion was a necessary condition for a transaction to be approved. It’s – it may not be a sufficient condition, but that was the minimum we needed. Q. And what you’re saying at paragraph 87 of your affirmation, Mr Cheng, is that your first instinct when you got the new proposal at a lower price was to try and find a new financial adviser because you thought it would be difficult for the old one to bless the new price? A. No … I didn’t suggest we find a new financial adviser … I don’t remember who … made that suggestion … But that idea was quickly shot down because … we felt we needed Duff & Phelps, the same party to … do … the same analysis … to see if the new revised price was fair … Q. So you didn’t sit there thinking, “Well, we certainly should use the previous adviser, because they’ve done all the work, they’re very familiar with the case, they’re very familiar with the facts and we want consistent advice.” No, your instinct is you shouldn’t do that because it would later look bad. Is that the position, Mr Cheng? A. Appearance does matter, which is why we decided to stick with … Duff & Phelps or Kroll. Q. Appearance matters. Why? What was more important to you, Mr Cheng, the appearances of getting a fairness opinion to support the deal or actually getting a reliable opinion that would help you to decide whether or not to say yes to the deal? A. No, we decided to …. we needed to have Duff & Phelps do an analysis and explain to the extent that the fairness, you know, level or range changed, why the change? If you hired a new one they will do a brand-new analysis, and there would be likely no … explanation between Duff & Phelps’ opinion and the new opinion performed by someone else. And the market could ask that’s how – you have two different opinions from two different parties within a short period of time … the markets may … have suspicion that you … we were shopping around … so we wanted to avoid that, that perception. Q. The market – and what you really mean by that is the unaffiliated shareholders who you were representing … A. Yes. Page 126 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 127 of 257 Q. They would have thought that you had shopped around to find a financial adviser who would give you the answer you wanted; yes? A. That’s why we didn’t – we decided against it. Q. Yes. But the point I’m putting to you, and I’m quite sure you understand, Mr Cheng, is that your first instinct – you, your colleague on the Special Committee, Mr He, and also, it seems Ms Kathleen Chien, your first instinct was to do exactly that, because you were concerned that your existing financial adviser would not give you the answer you wanted. A. It’s not our first instinct. It was … one of several possibilities we discussed. Q. And if you were doing your job properly, as a Special Committee, that should not even have crossed anybody’s mind, should it? A. Why shouldn’t we consider all possibilities?”

¶ [389] It was put to Mr Cheng that: “Rather than attempting to negotiate the best possible price and have Kroll opine on whether that was fair and, therefore, whether you should do the deal, you are having Kroll tell you the lowest price floor and then you’re proposing to take that to the buyer group and agree it; yes? A. It happened that was because Kroll’s work was delayed and we didn’t know what the fairness range was … what the minimum acceptable price was, so … we were negotiating on the fly … . Q. You are doing the polar opposite of what a Special Committee should have done, aren’t you? You’re looking for an opinion from Kroll not to tell you whether to do the deal or not but simply to enable you to get it done, and you’re going to put whatever their floor figure is to the buyer group without attempting to get anything more. That’s the position, isn’t it? A. That’s not true. That’s not true.”

¶ [390] It was put to Mr Cheng that in their negotiations with the Buyer Consortium the Special Committee did not attempt to extract any value from the fact that a recut would be a breach of contract. Mr Cheng said Page 127 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 128 of 257 they discussed it with Davis Polk but “concluded that the risk of not succeeding in that venue [a Hong Kong arbitration] was very high because it would come down to one Chinese lawyer’s opinion against the other.” “Q. … Why did you not attempt to extract some value from the potential claim that the company had against the buyer group in your negotiations with the buyer group? A. Because it would not have gotten anywhere.”

¶ [391] Mr Adkin revisited the go-shop point: “Q. The go-shops were window dressing really, weren’t they, because it was clear to the market … that a majority of shareholders, Mr Yan and Recruit, were committed to the deal and committed to oppose any other deal; yes? A. Yes. Q. And before that time the market, as we agreed, would have known that you needed to join with Mr Yan and Recruit to get into the deal; yes? A. … in theory anyone could come out and you know, make a bid for the whole company … with or without the support of the founders or Recruit. But in reality, it would be very difficult, very unlikely for anyone to propose anything like that, so the presumption is that if you want to … participate in this transaction you would need to work with or join force with … Mr Yan and Recruit.”

¶ [392] In re-examination Mr Cheng gave evidence as to his perception of the changing market conditions at the time the revised offer was received. Mr Cheng said that action taken by the PRC government “gave … investors and the market in general a big chill … so the overall share price of all these listed companies was trending down.”

¶ [393] Mr Cheng was asked about his reaction to the revised offer price of US$57.25. He said it came as a relief “that they were not trying to pull out”. It “obviously was much lower than 79.5 … but relative to where the overall market was and relative to where 51job’s own share price was … a market reaction following the November 8, 2021, company announcement, however vaguely worded that was, the share price crashed by something like 19/20% and it continued to trend down further until the revised offer Page 128 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 129 of 257 came … So in that context, the revised offer of $57, in my view, was a better deal relative to all these conditions compared to the $79 … So, in my view, the fact that they made an offer was a relief because we had something to work with, and the price, frankly, was actually reasonably good. I mean, obviously we would try and negotiate somewhere higher, but relative to where the market was and relative to where the company might have traded had there not been an original deal, it was not bad.”

¶ [394] Mr Imrie, in a well-focused re-examination, took Mr Cheng to various parts of the second proxy statement (GB/71/74) and Mr Cheng confirmed that they reflected his recollection of the price negotiations which took place.

¶ [395] Mr Cheng said they did try to negotiate the price: “I talked to Eric and … we thought, okay 57, at least it’s a price, you know, that’s attractive, given the market conditions, but optically it looked bad because it was a drop from 79 to the 50s. The first thing I said we should try to get it above 60 – somewhere above 60, if nothing else, just for, you know, for the optics, and – though we knew that would be a challenge because in all the take-private cases I was involved in and I was aware of, usually the negotiation – price negotiation at most could get – could increase the offer price by a couple of percentage points, two to three percentage points. That was usually the result. So to get it from 57 to 60 we thought was a challenge but something we should try very hard … started by negotiating at $63 … then they came back and said 61 minimum, and I just said “Okay, let’s just go with that”. Tell the buyer group not a penny less. Take it or leave it. And we ended up with just that.”

¶ [396] I gained the impression that Mr Cheng was not a puppet whose strings Mr Yan pulled. Could he have done more for those he was representing? Perhaps. Did he do enough to discharge his duties and to ensure that the Special Committee adequately discharged its duties? Yes. Mr Cheng was well aware of his duties as a member of the Special Committee. Overall he gave me the impression that he did his adequate best, with the help of the Special Committee’s professional financial and legal advisers, to get the best price he could for the unaffiliated shareholders. At times during his cross-examination it appeared that with hindsight counsel was trying to hold Mr Cheng to unrealistic and unattainable standards of perfection. Page 129 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 130 of 257 Mr Tianyi (Tony) Jiang’s evidence

¶ [397] Mr Tianyi (Tony) Jiang (“Mr Jiang”) says that he is a co-founder and partner of Ocean Link stated to be a private equity firm which focuses on the consumer, travel and telecon, media, and technology companies in the PRC. Ocean Link was a member of the Buyer Consortium which took the Company private pursuant to a merger agreement as amended on 1 March 2022. Mr Jiang says that he has been a director of the Company since 6 May 2022.

¶ [398] Mr Jiang says that he makes his affirmation to explain Ocean Link’s reasons for investing in the Company and to explain the position of the PE Sponsors in relation to the NDRC consultation where they seemed to take a different view from the Company.

¶ [399] Mr Jiang refers to the Company’s Proxy Statement filed in accordance with Section 13 (e) of the Securities Exchange Act of 1934 dated 29 March 2022 which he says sets forth a description of negotiations between Ocean Link, DCP, Mr Yan, the Company and the Special Committee leading to the completion of the transaction which, insofar as the narrative concerns events involving him he confirms is materially accurate.

¶ [400] At section C of his affirmation Mr Jiang outlines his relationship with what he describes as the “key actors” prior to the transaction. Prior to the take-private process he did not have a relationship with the Company and he had never met Mr Yan or Ms Chien. He had worked on a number of deals with Mr Hsu of RM.

¶ [401] At section D of his affirmation Mr Jiang refers to the initial discussions. On 17 September 2020 Alex Zheng of Ocean Link sent an email to Ms Chien expressing an interest in potentially participating in a privatisation of the Company. Mr Jiang says that shortly afterwards Mr Hsu reached out to him and said he was representing the Company. Mr Jiang told Mr Hsu that Ocean Link was interested but he thought that the valuation implied by the DCP Proposal was high and therefore Ocean Link did not intend to put in any topping bid.

¶ [402] Mr Jiang had contact with DCP and he expressed an interest in conducting due diligence on the Company at the same time DCP would be doing so. Page 130 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 131 of 257

¶ [403] At section E of his affirmation Mr Jiang deals with due diligence and views on the value of the Company. The PE Sponsors jointly engaged Meritco Group to conduct commercial due diligence and market research.

¶ [404] Mr Jiang said that as the due diligence progressed his concerns about the Company’s business were validated. The Company’s online products were increasingly becoming outdated due to years of under- investment in technology and marketing and the Company had been losing market share along the way. The Company was being significantly outspent by its competitors in sales and marketing. BOSS (also known as Kanzhun Limited) emerged as a major disruptor and rapidly gained market share.

¶ [405] At paragraph 33 of his affirmation Mr Jiang says that it became clear to him that: (1) the Company’s historical profit margins were not sustainable in the future in light of the significant under-investment in R&D and sales and marketing; (2) the Company needed to significantly ramp up its R&D and sales and marketing spending in order to catch up with its competitors (mostly notably BOSS) on algorithms and user experience, and more aggressively attract both job seekers and employers; and (3) such changes would represent a strategic pivot from what the Company had been doing over the previous few years. The execution risk was significant and during the transformation period, the Company’s profitability would be negatively impacted by the higher R&D and sales and marketing spend.

¶ [406] Mr Jiang says, based on the due diligence findings, his view was that privatisation was the right direction for the Company in order to give effect to the necessary changes. However “given the high initial price of the DCP Proposal and the overall risks of the Company’s business and potential transformation, I felt that Ocean Link should look to be a partner in the DCP Proposal rather than submit a competing bid which would have to be at a higher price.”

¶ [407] Mr Jiang said by around late April 2021 his view on the deal structure was as follows: Page 131 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 132 of 257 (1) the initial price offered by DCP was too high. The PE Sponsors should demand a lower price considering the Company’s competitive market position was worse-than-expected based on due diligence. Accordingly on 29 April 2021 Ocean Link and DCP expressed a desire to reduce the price, via Mr Hsu to the Company’s Special Committee, although such request was rejected; (2) the necessary transformation of the Company would require management commitment and Mr Yan and Ms Chien should roll over significant stakes to ensure they had enough “skin- in-the-game” and interest alignment with financial investors; (3) post take-private the PE Sponsors should be able to influence the Company’s key decision making given the complexity of the transformation; (4) the PE Sponsors should seek more downside protection.

¶ [408] Mr Jiang says that at the time he believed if the management team and shareholders executed the strategic initiatives referred to in his evidence (at paragraph 33 of his affirmation) and if there were no other macro-economic challenges to the PRC economy and the Company’s sector, the Company had a chance to leverage its established brand recognition and platform scale to win back job seekers and employers and improve its competitive position. Mr Jiang believed that if this was achieved, the Transaction would potentially yield a reasonable investment return for Ocean Link and its investors, justifying the investment risks Ocean Link was to undertake. Mr Jiang says that in reality “the investment did not play out as Ocean Link hoped for …”.

¶ [409] At section F of his affirmation Mr Jiang provides detailed evidence as to what he describes as “Developments adverse to the Chinese internet and technology sectors”. He says that during the course of 2021 there was a significant increase in regulatory and enforcement activities by the PRC government as it became more concerned with data management by both PRC and foreign companies, especially relating to national security, cybersecurity and data privacy, and it enacted a series of new laws and regulations to deal with these issues. Mr Jiang refers to the need to obtain approval from the NDRC in certain instances. Page 132 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 133 of 257

¶ [410] In section G of his affirmation Mr Jiang refers at length to the decision to consult the NDRC and the advice received in that respect. On 17 August 2021 the PE Sponsors submitted an “informal national security consultation application to the NDRC …”. Mr Jiang refers to the delays in making progress with this consultation application.

¶ [411] Mr Jiang says that Ocean Link considered revising the deal structure such that the likelihood of the Transaction being notifiable under the PRC national security review could be significantly reduced.

¶ [412] Mr Jiang says that on 23 December 2021 the PE Sponsors shared via email more detailed discussion material with Mr Yan and Ms Chien regarding a potential alternative transaction: (a) the PE Sponsors’ combined shareholding would be reduced from 20.9% to 9.99%; and (b) the original offer price would be reduced from $79.05 to $57.25 per ADS. Mr Jiang says that between 25 December 2021 and 12 January 2022 the discussions continued. The Buyer Consortium circulated an updated non-binding proposal via email to the Company and the Special Committee on 12 January 2022. On the same day the NDRC was informed that the PE Sponsors “would like to withdraw the application to the NDRC for informal consultation” and the Special Committee issued a press release announcing the receipt of the Revised Proposal of $57.25 per ADS.

¶ [413] Mr Jiang explains that in the discussions Ocean Link had with DCP regarding changing the deal structure, DCP suggested the revised price of $57.25 per ADS. Mr Jiang adds that Ocean Link evaluated the proposed revised price and considered that it would be an attractive proposal to the Company’s shareholders because: (1) the PRC economy continued to deteriorate because of, amongst other factors, COVID restrictions, worsened geopolitical tensions and the PRC’s property market downturn and as a result recruitment demand by businesses was negatively impacted; (2) the tightening of regulatory policies over the PRC internet companies; (3) the Company’s active users and viewership also deteriorated which suggested that the Company continued to lose market share. Page 133 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 134 of 257

¶ [414] Mr Jiang said that accordingly the Company’s share price declined significantly. The shares traded at $76.39 on 21 June 2021 and on 21 December 2021 at $45.54, which represented a decline of 40% and the NASDAQ Golden Dragon Index declined by 37% over the same period.

¶ [415] Mr Jiang said that the PE Sponsors concluded that the Company was worth significantly less than it was at the time of the Original Merger Agreement and whilst this $57.25 per ADS price was lower than the price proposed in June 2021 it was still a significant premium (25.7%) to the Company’s trading price on 21 December 2021, when the draft discussion material was prepared.

¶ [416] Mr Jiang says that on 28 January 2022 the PE Sponsors were provided with updated management projections. Mr Jiang considered the downward revision of projected profitability was warranted, although he says he thought the Company’s management still remained too optimistic about competition and the PRC’s economic outlook. Mr Jiang says that his concern was validated by the Company’s actual performance in 2022 and 2023.

¶ [417] In section I of his affirmation Mr Jiang sets out his evidence in respect of what he describes as negotiating the amendment to the Merger Agreement. He says on 10 February 2022 Kroll stated it was the opinion of the Special Committee that the revised offer price of $57.25 was too low. On 15 February 2022 on behalf of the Buyer Consortium it was indicated that the price of $57.25 was fair. On 16 February 2022 the Special Committee requested the Buyer Consortium to raise the price to $63 and the PE Sponsors quickly decided to reject such request. Mr Jiang says that he thought the $57.25 per ADS was already too high, as the Company’s trading price was inevitably inflated by the fact of the ongoing take-private process. Mr Jiang said in his view the revised price represented an even greater premium for the shareholders over what would likely have been the traded price had no take-private process been in motion.

¶ [418] Mr Jiang says that on 28 February 2022 representatives of the Buyer Consortium including himself held a call with Mr Hsu and a price of $61.00 per ADS was proposed on behalf of the Special Committee. The Buyer Consortium counter-proposed a price of US$59.95 which was rejected. After further negotiations on the same day the Buyer Consortium agreed to the $61.00 price. The amendments to the Original Merger Agreement reflecting the agreed terms (including the price of $61 per ADS) was executed on 1 March 2022. Page 134 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 135 of 257

¶ [419] At section J of his affirmation Mr Jiang deals with co-investment solicitation.

¶ [420] At section K of his affirmation Mr Jiang briefly deals with the Company’s “Actual Performance Since Completion” stating that the Company’s performance since the completion of the transaction has not achieved the targets in either of the management projections. Mr Jiang says that (a) the Company’s actual total revenue was 22% and 43% lower than in the revised management projections respectively in 2022 and 2023; and (b) the Company’s 2022-2023 combined actual operating profit was 63% lower than the revised management projections.

¶ [421] Mr Jiang was cross-examined by Mr Adkin and accepted that Ocean Link is a private equity firm and that Ocean Link had a financial interest in the outcome of the litigation and he also had “a small interest.”

¶ [422] Mr Jiang disagreed that he would have known, through Ocean Link that Mr Hsu had been working with the Company on a potential take-private back in June 2020. Mr Jiang said he was not aware of the emails [at DA/34/1] of 23 June 2020 between Mr Yan and Mr Hsu and also Ms Chien. Mr Jiang was referred to the references by Mr Hsu to Mr Yan that the new investors had authorised them to advance potential discussion and Mr Jiang replied “That’s what’s said in the email but I never gave that authorisation … at least Ocean Link never gave such authorisation.” Mr Jiang disagreed that the email was right and further disagreed that his evidence that he was not aware of Mr Hsu’s involvement until after the Company had announced the DCP proposal was wrong.

¶ [423] Mr Jiang accepted that the DCP proposal was sent on 17 September 2020 and Ocean Link signed the original merger agreement at the US$79.05 price on 21 June 2021 and that between those times Ocean Link conducted very considerable due diligence on the Company.

¶ [424] Mr Jiang was cross-examined in respect of the engagement of third-party due diligence providers including Meritco for commercial due diligence who produced a very thorough 248 page report. Mr Jiang agreed that Meritco was providing very extensive well-researched commercial due diligence. Mr Jiang agreed that Ocean Link and DCP also engaged Ernst & Young to conduct financial and tax due diligence, Fangda Partners to conduct Chinese legal due diligence, Paul Weiss to conduct US legal due diligence, and Ogier to conduct offshore legal due diligence. Page 135 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 136 of 257

¶ [425] Mr Jiang accepted that Ocean Link interviewed the CEO Mr Yan, the COO and acting CFO Ms Chien and the heads of various business units of the Company and all the interviews had occurred by the beginning of June 2021.

¶ [426] Mr Jiang was referred to Ocean Link’s four-member investment committee which made decisions about how Ocean Link would invest its funds and he agreed that the approval of this committee was required before Ocean Link could enter into the original merger agreement at $79.05 and as part of the process the results of the due diligence were presented to the committee in early – mid June 2021. Mr Jiang confirmed that the committee provided unanimous agreement to enter into the original merger agreement.

¶ [427] Mr Jiang was referred to the proxy statement filed with the SEC as part of the original merger agreement in July 2021 setting out the amounts to be invested by DCP and Ocean Link.

¶ [428] Mr Jiang was referred to messages of 22 June 2021 (DA 666T/68) between Mr Yan and Ms Yvonne Yan of RM / Kilometre and agreed that DCP and Ocean Link were trying to get more shares and added as DCP was “The launching party, they started a deal, so I felt as out of courtesy to them they should own a slightly bigger share.”

¶ [429] Mr Jiang agreed that Ocean Link received the Company’s May 2021 Management Projections on 11 May 2021 and Ocean Link also prepared its own projections. Mr Jiang said that only Ocean Link’s projections were presented to its investment committee. Mr Jiang agreed that the evidence he gave on deposition about Ocean Link’s projections was true and the projections went from 2022 to 2027.

¶ [430] Mr Jiang was referred to a shareholders agreement dated 1 March 2022 between the buyers which included Ocean Link and agreed that the plan at that point and under the original merger agreement was that the listing priority of the Company would be a PRC A-share listing or a listing in Hong Kong. Mr Jiang agreed that that was “one of the expectations” “that’s the goal, yes” within 3 years.

¶ [431] Mr Jiang agreed that Ocean Link considered Mr Yan, Ms Kathleen Chien “and other existing management members” to be “key to the future performance of the company” and Ocean Link intended that these executives would remain in post after the closing of the deal. Page 136 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 137 of 257

¶ [432] Mr Jiang disagreed that in his affirmation he had sought to overplay what he says were the problems with the Company and underplay the belief that the Company was an attractive proposal and accepted that the Company “was for almost two decades the leading provider of online recruiting services in China, but then there was at a very critical point … they were [over] taken by another competitor, BOSS” and “it now turns out that a lot of my concerns actually turned out to be actually quite true … that the company … continued to lose market share … we felt that the management team … I think their age was a little old … so it’s about hiring younger managers, people who actually, you know, know how to do sales and marketing, social media and things like that … keeping existing management doesn’t mean we don’t have a plan to actually change the management team.”

¶ [433] Mr Jiang was referred to the final proxy statement (GD 39/1) which he signed and the reference to the Company continuing to conduct its operations substantially as they were currently being conducted.

¶ [434] Mr Jiang was referred to the proposal in late 2021 of an alternative transaction by Ocean Link and DCP with the PE Sponsors’ combined shareholding being reduced from 20.9% to 9.9% and the price being reduced from $79.05 to $57.25 per ADS. The lower price “reflected the developments related to the Company’s business and the general market since the signing of the original merger agreement on 21 June 2021”. Mr Jiang was cross-examined on the “draft discussion material” and the issue with the NDRC. It was put to Mr Jiang that the revised offer price was because of the funding gap created by the NDRC issue. Mr Jiang responded “No, I disagree, because we have to take things into context. The entire document was a proposal to resolve the impasse that we had for months and then we run the risk of not being able to complete the original transaction … So this proposal in its entirety is about how do we resolve this issue … Price is obviously a very important element of it.” Mr Jiang agreed that the “NDRC issue was one of the key consideration[s] that we proposed a revised deal, correct.” Mr Jiang accepted that dropping the percentage ownership below 10% would create a funding gap “But there’s also more to it … the reduction of price obviously as it feeds into the deal parameters help us to lead to a deal – an alternative deal that can resolve the impasse, but also a very important context of this was our view at that point in time, Ocean Link’s view and DCP’s view, on the valuation of the business, and in that point in time when we proposed the proposal we actually felt that the fair value of the business had deteriorated, and then proceeding to this proposal there was already a discussion from DCP to Rick [Yan] about a reduction in price, which didn’t go very well. And then at that point in time the relationship among the buyer consortium was a little tense. So this is we basically trying to focus less on the difference and focus more on what is an alternative solution that is better for everybody else Page 137 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 138 of 257 to resolve a solution. So price has to do with the entire deal but also as a separate matter reflecting our view in the business as well.”

¶ [435] Mr Jiang was referred to a document (DA/458/1) and cross-examined at length on the “funding gap” point and the “financial burden”. Mr Jiang said: “… there’s the context to this document, which is where we will all focus. We wanted to focus less on the differences and focus more on a solution to the problem, the difference being that Rick Yan, being a very proud CEO of the company, Recruit being an existing shareholder of the business, keep arguing the business is worth less, it’s not going to work very well with them, especially when the relationship was in a very tense situation for about four or five months due to the NDRC issue. So we focus more on a solution that is going to help everybody and which is make everybody better off … Mr Rick Yan has been a very proud CEO … in the sense of his love for the company and how proud he was – he was not always, you know, completely rational about these sort of economic calculations.”

¶ [436] It was put to Mr Jiang that nowhere in the document did they explain as part of the rationale for the revised proposal or for reducing the price that they thought the value of the business was worth less than the original proposal. Mr Jiang responded “Except for where I just pointed out”, referring to his point that the reference to “financial burden” was a reference to “the fact that there is a very big premium that we had to pay over the fair value of the business, which is reflected by the market value.” Mr Adkin retorted “Well, we’ll have to agree to disagree on whether that means what you’re now suggesting it means.” [Day 6/166:2-3]. It was put to Mr Jiang that nowhere in the document was there any “mention of any worsening problems with COVID” to which Mr Jiang stated “… the CEO, because of proudness, didn’t want to engage on a price reduction because of the deterioration of the business …”. It was put to Mr Jiang that there was not mention of “tightening regulatory policies in this document”. Mr Jiang responded “that was public knowledge at that point in time”. The following exchange ensued: “Q. … just let me be clear, Mr Jiang, what I’m putting to you is that your affirmation explanations for why you decided to propose a lower deal are not reflected in the contemporaneous document which explains why you’re proposing a lower deal price … Page 138 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 139 of 257 A. I disagree because there are many other discussions and email exchanges, for example the email exchange from us, our legal counsel to the Special Committee in terms of explaining on the different various reasons for the price reduction … This is just one document out of many documents … the context of this document was not to confront the CEO and Recruit about how bad their business is performing but rather trying to focus on a solution than argument points. Q. … you want a lower price; yes? A. Yes. Q. For buyers of course, when you’re presenting a negotiating position to the Special Committee you say “Well, the business has gone to the dogs, gone to the wall” but when we look behind the curtain at what you are saying to the rest of your consortium, it isn’t what you’re saying. It’s a different rationale … A. I disagree … Q. And the suggestion that you had undertaken further due diligence, which suggested that the value of the company had decreased since the original Merger Agreement is also not correct, because no further due diligence was taken, was it? A. That’s a completely inaccurate statement. There was diligence done by our us – our own together with Meritco as well as BOSS become, you know, a publicly listed company in June, the largest competitor. Q. Well, that was before the original Merger Agreement wasn’t it, and you knew that was in the offing; yes? You knew that when the original Merger Agreement was signed; yes? A. Yes. BOSS continued to publish their quarterly results since they became public, and we got more insights through public … research reports and our own diligence around BOSS as well … Q. No, no, Mr Jiang, they published their IPO statement on 11 May … 2021, that’s month and a bit before you sign the original Merger Agreement; yes? A. That is correct.”

¶ [437] Mr Jiang was asked whether it came as a surprise to him that the proxy statement (DA/331/57) which mentioned RM did not mention Mr Hsu by name or mention Kilometre and Mr Jiang responded “That Page 139 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 140 of 257 was not a surprise to me because his company has been named here … Q. Nothing in the proxy statement to tell you that it’s his company; yes? A. … that’s not something I looked into.”

¶ [438] Mr Jiang was cross-examined as to why he did not produce documents to support his essential point that the Company was overpriced “Q … you could have provided to support your affirmation documents which back up what you say in it, but you have chosen not to. That’s what I’m putting to you. A. No. I did not make a choice not to submit documents. I was not under the impression that I need to provide documents … I think I’m a witness to the case. I provide my affirmation with reference to certain documents … Q. … you have made a deliberate choice not to provide documents which would enable the veracity of your evidence to be tested. That’s the case, isn’t it? A. That I disagree. I think I’ve been very forthcoming to the court.”

¶ [439] Mr Imrie conducted a well-focused re-examination with admirable precision and laudable substantial adherence (with a couple of minor exceptions) to the rules on re-examination and asked Mr Jiang about the plan to re-list the Company upon which he had been cross-examined. Mr Jiang stated: “That plan couldn’t be carried out because, for example, the Company’s revenue has been declining since 2021, because the EBITDA of the business was negative last year, in 2024, and this year at best slightly above break-even. So because of the poor operating performance of the business, I think it’s a very far-fetched goal that it will get listed any time soon.”

¶ [440] Mr Jiang also described the additional due diligence conducted by Ocean Link subsequent to the original Merger Agreement: “Subsequent to the original Merger Agreement and in parallel with the discussion around NDRC process, we continued to monitor the market on our own, using our team as well as together with the assistance of Meritco who conduct due diligence for us based on what’s publicly available information, and that includes the data tracking using Quest Mobile and EBIT, many other externally available resources. At the same time, there are research reports, industry experts, that Meritco – we and – ourselves and Meritco speak to, to help us gain continuous understanding of industry dynamics.” Page 140 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 141 of 257 The expert evidence

¶ [441] I now turn to some of the relevant expert evidence placed before the court. General

¶ [442] In addition to the factual evidence, the court was provided with a great deal of written and oral evidence from the two experts instructed in respect of this case. The experts provided their first reports (Professor Kenneth M. Lehn’s first report with attachments ran to some 256 pages with 318 footnotes to the main body of his report, Mr Travis Taylor’s first report ran to 273 pages with 1286 footnotes), a joint memorandum running to 96 pages and then their supplementary reports (Professor Lehn’s 145 pages with 354 footnotes, Mr Taylor’s 158 pages with 884 footnotes). The experts were cross-examined over 5 and a half days each. To be honest they did not make it easy for me. They added rather than subtracted complexity.

¶ [443] I deal with the experts' written reports below and their oral evidence and set out my conclusions on their evidence in the determination section of this judgment.

¶ [444] At the beginning of the oral evidence of each expert I gave the following direction: “… please do not hesitate to dumb down any of these technical subjects for my benefit. I won’t regard that as disrespectful. The real quality of an expert is to make difficult and complex issues very simple for a simple soul like a judge, like me. So please don’t hesitate to simplify issues for my assistance.”

¶ [445] The experts were no doubt very intelligent human beings with skills in corporate valuations but neither expert fully complied with that direction.

¶ [446] At times both experts came dangerously close to discarding their clothes of independence and slipping into the gowns of attorneys (if not actually putting them on) and engaging in advocacy on behalf of those who had engaged them. In answering some of the questions put to them they both could not resist Page 141 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 142 of 257 the temptation of making adverse comments about the other in a thinly disguised attempt to score some points in their client’s favour and to persuade the court to accept their opinion.

¶ [447] Perhaps I was expecting far too much from the experts. The difficulties presented by expert evidence are not new. As long ago as 1873, the formidable and wise Sir George Jessel MR in Abinger v Ashton (1873) 17 LR Eq 358 at 374 stated: “Now it is natural that [an expert’s] mind, however honest he may be, should be biased in favour of the person employing him … Undoubtedly there is a natural bias to do something serviceable for those who employ you and adequately remunerate you. It is very natural, and it is so effectual that we constantly see persons, instead of considering themselves witnesses, rather consider themselves as the paid agents of the person who employs them … Each set of valuers values high or low, according to the requirements of the person who employs them.”

¶ [448] Some years later, David Neuberger speaking extrajudicially in a speech Science and Law: Contrasts and Cooperation to the Royal Society on 24 November 2015 stated at paragraph 29: “It is very hard to find an expert who is genuinely independent; yet that is what the law expects from every [expert] witness … [who] often have many meetings with the lawyers acting for their client, where they discuss the issues and the evidence, and such discussions will almost always be directed to how to advance the client’s case and do down the opposition. All this will almost inevitably cause the expert witness to develop feelings of loyalty and commitment to his client’s case.”

¶ [449] Lord Neuberger at paragraph 31 referred to courtroom “oral gladiatorial combats”: “Many informed spectators may think that in cross-examination of experts, the forensic skills of the advocate and the quick-wittedness of the expert are more in play than the right answer to the question they are discussing.”

¶ [450] In this complex area of valuation, we must not lose sight of the basis. Page 142 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 143 of 257

¶ [451] It is worth revisiting the first few sub-paragraphs of B5.2(b) of the Financial Services Division Guide Second Edition: “(i) It is the duty of an expert to help the Court on the matters within his expertise. This duty is paramount and overrides any obligation to the party from whom the expert has received instructions or by whom he is paid. (ii) Expert evidence presented to the Court should be, and should be seen to be, the independent product of the expert uninfluenced by the pressures of litigation or any party. (iii) An expert witness should provide independent assistance to the Court by way of objective unbiased opinion in relation to matters within his expertise. An expert witness should never assume the role of an advocate or seek to promote his client’s case.”

¶ [452] In Professor Lehn’s opinion the fair value is US$31.11 per share. His valuation is derived from his estimate of the AMTP. The Company stresses that regardless of what methodology the court adopts it will need to consider the fact that the merger took place against a declining market. The merger price was $61 and the closing price of the shares on 11 January 2022 was $45.83.

¶ [453] Professor Lehn’s AMTP starts from a trading price of $45.83 that was affected upwards by the Original Proposal and then rolls that price forwards to the Valuation Date using the mathematically estimated relationship between the returns of the Company’s shares and the returns of the Golden Dragon Index with a result of a fair value on the Valuation Date of $31.11.

¶ [454] In support of this, and as cross-checks on his conclusion, Professor Lehn also performed the same roll- forward analysis from the Unaffected Price on 16 September 2020 (which arrived at the price of $30.88), further roll-forwards using different estimation periods and a one-factor model (arriving at prices from $30.15 at $31.91) and a comparable companies analysis (which provides a range of $19.74 to $34.77) although he places limited reliance on the latter analysis. Page 143 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 144 of 257

¶ [455] The Company says that Professor Lehn’s estimate of fair value is also consistent with the break prices disclosed by some dissenters.

¶ [456] Using a DCF model Mr Taylor says that the fair value is US$111.06 per share.

¶ [457] The CCCW Dissenters (and the Carey Olsen Dissenters adopt their submissions) stress that Professor Lehn is an academic by profession and has never given a valuation opinion for a real-life buyer or seller, apart from “when my children have bought houses I’ve given them my opinion about the value …” (Day 7/154). They suggest that his academic approach and lack of real-world advisory experience came across in his evidence and in effect tainted it.

¶ [458] The CCCW Dissenters stress that in contrast Mr Taylor is a professional valuer with significant experience of advising real-world commercial buyers and sellers and gave a recent example of a non- contentious valuation for a business relating to luxury watches.

¶ [459] The Company says that in addition to his extensive and highly-credited academic career spanning more than 30 years Professor Lehn was the Chief Economist at the United States Securities and Exchange Commission from 1987-1991 and in addition to teaching at the University of Pittsburgh is a senior consultant at Compass Lexecon.

¶ [460] Mr Taylor is a managing director at Secretariat and the Company criticises him as having lacked expertise in many of the subjects upon which he was purporting to give expert evidence including the efficient market hypothesis, event studies and co-movement. The Company did however accept that Mr Taylor had expertise in the mechanics of the DCF Methodology. Mr Taylor accepted that he could not claim to understand the Company’s business better than management.

¶ [461] Under the directions order (see paragraph 43 of the Order made on 19 January 2023) there was a clear requirement on the experts to prepare a joint memorandum recording various issues including (a) the issues on which they agree; (b) the issues on which they disagree; and (c) a brief summary of the reasons for any such disagreement. There was also provision for supplemental expert reports to be exchanged following the issuance of the joint memorandum. Page 144 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 145 of 257

¶ [462] There are a number of issues identified in the joint memorandum where both experts purport to in effect reserve their positions on such issues until they have filed their supplemental reports. This is not helpful. Experts should not keep cards up their sleeves. There should be no attempts at ambushes. The cards on what is agreed and what is not agreed and why should be put down on the table in the joint memorandum. They should not be left to be played in the supplemental expert reports or in oral evidence at trial. Professor Lehn’s first report

¶ [463] In his first report dated 22 January 2025 Professor Lehn says that he considered four methods to estimate the fair value of a share as of the Valuation Date; (a) the fair value implied by the market price and other relevant market evidence; (b) the per share merger consideration; (c) DCF analysis; and (d) the Comparative Company Analysis ( the “CCA”). He concluded that the best estimate of the fair value of a share on the Valuation Date was US$31.11. He says that his conclusion is based on the following considerations: (1) there was a “semi-strong form” efficient market during the relevant period indicating that all public information was reflected in the price including information about the value of the Company as a going concern. Professor Lehn says that as a result the market price serves as a reliable basis, with some adjustments, for estimating the fair value of a share as of the Valuation Date; (2) the closing price on 11 January 2022, the day before the revised proposal letter was $45.83 (the Pre-Announcement Price). As of 11 January 2022, the Original Merger Agreement, which provided for an acquisition at $79.05 per share, had not been terminated or amended, which meant that it was still possible that the Original Merger Agreement would be consummated. Professor Lehn opines that the market price of a company’s stock will exceed the market’s assessment of its fair value when market participants believe that there is a possibility that the company will be acquired at a premium – therefore, the Pre- Announcement Price likely overstates the fair value of a share on 11 January 2022; and (3) there was a substantial deterioration in both market and industry conditions over the period from 11 January 2022 to 26 April 2022 (the Valuation Date) (the “Interim Period” or the Page 145 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 146 of 257 “Adjustment Period”) due to several factors including (a) regulatory changes in both the US and the PRC; (b) the COVID situation in the PRC; and (c) the adverse effects of the war in Ukraine on the global economy. During this period, the value of US listed stocks of companies that operated primarily in the PRC declined by 32.7% and the stock prices of the three comparable companies that operated primarily in the PRC (Tongdao Liepin Group and Kanzhun Limited (i.e Boss)), declined in value by 25.9% and 46.4%, respectively, with an average decline of 36.2%. Professor Lehn says he adjusted the Pre- Announcement Price to account for changes in market and industry conditions based on the historic relation between the performance of the Company, a market index and an industry index. Professor Lehn says that this resulted in an adjusted Pre-Announcement Price of $31.11 on the Valuation Date which, in his opinion, is the most reliable estimate of the fair value of a share on the Valuation Date.

¶ [464] Professor Lehn says that the three other valuation methods he considered suffered from limitations that he felt made them considerably less reliable than the market-based estimate of the fair value of a share.

¶ [465] Professor Lehn summarised his conclusions on the per share merger consideration valuation method as follows: (1) the per share merger consideration of US$61 is likely to substantially overstate the fair value of a share as of the Valuation Date as it included a substantial acquisition premium and the fair value declined substantially after 1 March 2022; and (2) the consideration that the per share merger consideration exceeded the fair value of a share is also supported by the fact that (a) no competing bids emerged; and (b) more than 75% of the shares owned by shareholders that attended the EGM and which were not affiliated with the Buyer Consortium or the Company voted in favour of the merger proposal.

¶ [466] Professor Lehn summarised his conclusions on the DCF valuation method as follows: (1) market evidence provides more reliable estimates of fair value than DCF analysis. Market prices reflect a consensus view about the value of an asset, reflecting the actions of many willing buyers and willing sellers who are putting actual money on the line when they make Page 146 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 147 of 257 their decisions to buy or sell the asset. In contrast, in the context of appraisal proceedings, a DCF analysis reflects numerous assumptions and judgments made by a single person who is not making investment decisions based on his DCF analysis. Furthermore, DCF analyses are highly sensitive to the choice of inputs, where seemingly small changes in inputs, such as cash flow projections, discount rates, and expected growth rates, among others, can cause large changes in DCF valuations. Hence, disagreements between reasonable experts can give rise to large differences in DCF valuations; (2) DCF analysis depends critically on the cash flow projections used in the analysis. The available cash flow projections for the Company are not reliable because (a) they are not unbiased estimates of expected cash flows; (b) they are stale as of the Valuation Date; and (c) they do not project the Company to be in “steady state” as at the end of the forecast period, which is a critical assumption in DCF analysis; and (3) formal and informal PRC regulations restrict the Company’s ability to distribute cash to its shareholders. Professor Lehn says that he is not aware of a methodology that would provide a reliable estimate of the effect of these regulations on the value of the Company’s substantial cash holdings and other assets.

¶ [467] For these reasons, Professor Lehn concluded in his first report that DCF analysis would not provide a reliable estimate of fair value.

¶ [468] Professor Lehn summarises his conclusions on the CCA valuation method as follows: (1) CCA suffers from two substantial limitations. First, there are no publicly traded recruitment companies that are truly comparable to the Company in terms of key variables that affect business valuations. A standard CCA will not provide a reliable estimate of fair value in this matter. Secondly, a more sophisticated CCA in which regression analysis is used to control for salient differences between the comparable companies and the subject company is not feasible in this matter because there are many salient differences but only seven publicly traded recruitment companies; and Page 147 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 148 of 257 (2) it is possible to use a “Modified Multiples” analysis to account for differences in expected growth rates and in this case it results in estimates of fair value that range from US$31.28 to $41.51 per share. Professor Lehn says that this range likely overstates the fair value of a share because it does not account for the adverse effect of regulatory restraints on the Company’s ability to distribute cash or the proceeds from asset sales on the value of the Company’s cash holdings and non-operating assets. He adds that nevertheless the Modified Multiples analysis supports his conclusion that the fair value of a share on the Valuation Date was substantially lower than the Pre-Announcement Date. He concludes however that CCA provides a considerably less reliable estimate of the fair value than the market evidence in this matter. Mr Taylor’s first report

¶ [469] Mr Taylor in his first report dated 22 January 2025 concluded that the fair value as at 26 April 2022 was US$111.06 per ADS.

¶ [470] Mr Taylor says he considered a range of valuation methodologies and ultimately concluded that the DCF method (which he says is a form of an Income Approach) was the most reliable. Accordingly he relied on the DCF method for his fair value assessment.

¶ [471] Mr Taylor says he considered whether it would be appropriate to rely on (a) the Revised Merger Price; and (b) the Company’s listed share price but concluded that these were unreliable indicators of fair value as at the Valuation Date.

¶ [472] In respect of the Revised Merger Price Mr Taylor felt that it was the result of a process with little to no effective protections to unaffiliated minority shareholders. In his opinion the Revised Merger Price could only be relied upon as an indicator of fair value if there was a robust sales process creating a competitive bidding environment in order to ensure that the deal was negotiated on arm’s-length, competitive terms.

¶ [473] Mr Taylor says he has identified several deficiencies relating to the merger: Page 148 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 149 of 257 (1) the Original Merger Price of US$79.05 per ADS was revised downwards to US$61.00 per ADS despite legal advice that the price in the Original Merger Agreement was enforceable. The DCP presentation stated “All parties are better off from lower price.” Unaffiliated minority shareholders were significantly worse off; (2) unequal terms given to unaffiliated minority shareholders compared to employees; (3) an inefficient market check. There was no pre-signing market check and the post-signing go-shop was poorly executed. Furthermore, the members of the Buyer Group that were rolling over their shares had signed an agreement obligating them to vote for the proposal and against a competing bid. This rendered the go-shop meaningless; (4) the absence of effective protections for minority shareholders. There was no Majority of the Minority (MoM) clause. The ability of minority shareholders to challenge the terms of the transaction was very limited; (5) the absence of proper oversight from the Special Committee. The Special Committee did not appear to intervene proactively to limit management’s control over insider information. It was not involved in the process of preparing the 2022 Management Projections nor did it closely review or interrogate them. The 2022 Management Projections were prepared by a member of the Buyer Group who had a financial incentive to conclude the take-private transaction at a lower price; and (6) the role played by Kilometre Capital. The Special Committee seemingly allowed itself to be influenced by an advisor of the Company with ties to company management and the Buyer Group. The Special Committee delegated important business to Kilometre Capital, including critical negotiations. This raises questions over the arm’s length nature of the transaction.

¶ [474] Mr Taylor says that given the inherent difficulties and conflicts of interest associated with an MBO, along with significant procedural deficiencies, he does not consider that the Revised Merger Price can be treated as a reliable indicator of fair value. Page 149 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 150 of 257

¶ [475] In respect of the Company’s listed share price Mr Taylor says that this can be a reliable indicator of fair value if: (1) there is an efficient market for the shares; and (2) there is no evidence of insider or material non-public information (MNPI). He says this factor is particularly relevant for MBOs where management can control the flow of information to the market.

¶ [476] Mr Taylor adds that immediately prior to the Announcement Date there was a general decline in the market due to the COVID-19 pandemic. He says that academic literature confirms that markets can make significant pricing errors in times of economic turmoil or market displacement and so the question of efficiency is particularly important in this case, given that the transaction was announced in September 2020. On the Unaffected Date of 16 September 2020, the Company’s ADSs were priced at US$68.12, having traded at US$91.75 on 16 January 2020, two weeks before the World Health Organization declared a public health emergency.

¶ [477] Mr Taylor states that once the transaction was announced the Company’s share price was driven by transaction-related developments rather than issues of intrinsic value. He therefore considers the Company’s share price in the Post-Announcement Period to be an unreliable indicator of fair value.

¶ [478] Mr Taylor does not rely upon the Company’s historical traded price as a proxy for fair value because: (1) based on his event study analysis, in particular he does not consider the market for the Company’s ADSs to be efficient; (2) the existence of value-relevant information available to management but not the market in general; (3) the proximity of the Announcement Date to the outbreak of the COVID-19 pandemic and the associated impact on equity markets; and Page 150 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 151 of 257 (4) he did not consider it reasonable to roll forward the Company’s unaffected share price to the Valuation Date in light of the protracted 19-month take-private process.

¶ [479] Mr Taylor says that the DCF method is his preferred valuation approach because it captures the underlying fundamental drivers of a business and he has not found other valuation approaches, such as the transaction price or the Company’s share price, to be reliable indicators of fair value. Mr Taylor says that the DCF method is forward-looking, considers the time value of money, and focuses on the business’s ability to generate cash. Mr Taylor’s view is that in this case it should provide a more accurate estimate of a company’s intrinsic value other than valuation methodologies.

¶ [480] Following his review of the process involved in the production of the 2022 Management Projections and the projections themselves Mr Taylor lists what he regards as deficiencies: (1) the 2022 Management Projections were prepared by a member of the Buyer Group with little oversight or review. Ms Chien was financially incentivized and legally bound to support the proposed transaction. Further, the Special Committee delegated responsibility for a thorough review to Kroll, who had no obligation to verify the accuracy or completeness of the projections and, according to Ms Chien, never met with her to discuss the heavily revised projections; (2) advertising and promotional costs were materially over-estimated compared to contemporaneous budgets; (3) the Company performed better than expected in 2021 and the competitive threat of BOSS was well-known when the 2021 Management Projections were prepared; and (4) reducing the merger price seems to have been driven by the PE Sponsors (DCP and Ocean Link) who expressed concerns unrelated to fair value. The PE Sponsors cited regulatory concerns related to their ownership as a reason for the reduction in merger. These concerns did not warrant a reduction in the merger price. Page 151 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 152 of 257

¶ [481] For the purposes of his DCF analysis, Mr Taylor adopts the 2021 Management Projections prepared by Ms Chien as part of the take-private process. Subject to what he describes as “minor” adjustments he considers the 2021 Management Projections to be reliable due to the following factors: (1) when preparing them Ms Chien discussed the business outlook with multiple members of the Company’s senior management; (2) Ms Chien and Mr Yan were questioned by Kroll on the 2021 Management Projections and key assumptions; (3) they were reviewed and discussed by the Special Committee; and (4) Ms Chien prepared them before she signed the Support Agreements which bound her to vote in favour of the proposed transaction.

¶ [482] Mr Taylor says that because the explicit forecast period in the 2021 Management Projections is just 3.7 years (i.e. the period between the Valuation Date of 26 April 2022 and 31 December 2025), in accordance with what he describes as “accepted practice” he has extended the explicit forecast period by seven years to 2032. In doing so he says he tapers down the high growth rates in 2025 to his selected terminal growth rate of 4.37%.

¶ [483] Mr Taylor considers the appropriate discount rate to be the Company’s WACC which he assesses to be 9.39%. He assumes an EBITDA margin of 32.5% in the Terminal Year, which he considers to be reasonable given historical performance and contemporaneous Buyer Group/Ocean Link Projections. The joint-memorandum

¶ [484] I have noted the contents of the joint memorandum of the experts dated 2 April 2025, which to be frank I did not find of much assistance. It was a very dense document and did not evidence any real intention on behalf of the experts to limit the areas of disagreement. Moreover, it did not state the issues, the disagreements and the reasons for them in a straightforward easy to follow way. The main intent on behalf of both experts appears to have been to state nothing and to concede nothing that would in their Page 152 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 153 of 257 minds prejudicially impact on their respective clients’ position at trial. That was not the purpose behind my direction for a joint memorandum. Professor Lehn’s supplemental report

¶ [485] In his supplemental report dated 5 May 2025 Professor Lehn endeavours to demonstrate that Mr Taylor’s conclusion that the fair value of a share is US$111.06 is implausible on its face because it is inconsistent with the market evidence. He seeks to identify flaws in Mr Taylor’s analysis which are said to further undermine his conclusions concerning the fair value of a share. Professor Lehn also addresses Mr Taylor’s analysis of the market trading price and the Per Share Merger Consideration, and his DCF analysis and his CCA. Professor Lehn also describes a revision to his CCA that results in revised estimate of the equity value per share that ranges from US$19.74 to $34.77 on the Valuation Date.

¶ [486] Professor Lehn says that Mr Taylor’s valuation of US$111.06 per share is 63% higher than the closing price of the ADS on 16 September 2020 of US$68.12 (i.e. the Unaffected Price) whereas the NASDAQ Golden Dragon China Index declined by 52.48% between 16 September 2020 and the Valuation Date. In a much debated footnote 10 Professor Lehn adds “[i]f the Unaffected Price were adjusted using the methodology described in Lehn Report ¶ 28 the adjusted Unaffected Price on the Valuation Date would be $30.88, which is 54.67% lower than the Unaffected Price.” Professor Lehn details various other implausibilities.

¶ [487] In respect of the market price Professor Lehn says that in fact Mr Taylor’s event study demonstrates market efficiency. Professor Lehn says that Mr Taylor’s analysis of directionality and magnitude is flawed.

¶ [488] Professor Lehn says that Mr Taylor’s analysis of unexplained share price movement is flawed and that Mr Taylor ignores other indicators of market efficiency. Professor Lehn says that Mr Taylor’s analysis of MNPI is flawed as is Mr Taylor’s analysis of the impact of COVID-19 on equity markets. Professor Lehn says that Mr Taylor’s analysis of the Post-Announcement Share Price is flawed.

¶ [489] In respect of Mr Taylor’s analysis of the merger conclusions, Professor Lehn says it is flawed as is Mr Taylor’s analysis of the renegotiation of the Original Merger Agreement. Page 153 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 154 of 257

¶ [490] Professor Lehn says that Mr Taylor’s analysis of the consequences of the reduction in the merger consideration is flawed.

¶ [491] Professor Lehn says that Mr Taylor’s analysis of the sales process is flawed and deals with each of Mr Taylor’s criticisms under the following headings: - The Company’s response to unsolicited expressions of interest - The Special Committee’s tools to encourage a competitive environment - The purportedly “defective market check”.

¶ [492] Professor Lehn says that Mr Taylor’s analysis of Ms Chien’s purported conflict of interest is flawed and that Mr Taylor’s concerns about the Special Committee’s independence lack merit as do Mr Taylor’s concerns about the role of RM.

¶ [493] Professor Lehn adds that Mr Taylor’s analysis of MoM provisions is flawed as is Mr Taylor’s analysis of the Special Committee’s response to shareholder concerns.

¶ [494] Professor Lehn says that Mr Taylor’s analysis of the likelihood of the MBO’s completion is flawed.

¶ [495] In a lengthy, section Professor Lehn seeks to persuade the court that Mr Taylor’s DCF analysis is fundamentally flawed under the following sub-headings: - Mr Taylor’s DCF analysis is based on unreliable projections of the Company’s expected cash flow - Mr Taylor’s extension of the prior company projections is speculative and implausible - Mr Taylor’s estimate of terminal value is speculative and implausible - Mr Taylor’s analysis of 51job WACC is flawed - Mr Taylor’s analysis of the PRC withholding tax is flawed Page 154 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 155 of 257 - Mr Taylor ignores other PRC restrictions on distributions of cash - Mr Taylor provides no reliable basis for excluding a minority discount.

¶ [496] Professor Lehn says that Mr Taylor’s CCA does not support his DCF valuation. Mr Taylor’s supplemental report

¶ [497] In his supplemental report dated 6 May 2025 Mr Taylor says that Professor Lehn’s conclusion that the fair value as at 26 April 2022 was US$31.11 per ADS is unreasonably low against any reasonable benchmark of the Company’s value.

¶ [498] Mr Taylor says that Professor Lehn’s valuation is divorced from the Company’s historical trading data and is so low that it amounts to little more than the value of its cash balance.

¶ [499] Mr Taylor says that as Professor Lehn adopts what he describes as an AMTP approach, the reliability and accuracy of his valuation is dependent upon (a) the Company’s share price being a reasonable proxy for fair value on 11 January 2022 (his starting point); and (b) the integrity of the roll-forward to the Valuation Date (26 April 2022). Mr Taylor considers both components of Professor Lehn’s calculation to be faulty.

¶ [500] Mr Taylor says that the Company’s share price as at 11 January 2022, 16 months after the transaction was announced on 17 September 2020, did not represent its fair value and this therefore invalidates Professor Lehn’s entire valuation conclusion. Mr Taylor says that there was considerable confusion in the market as to the Company’s value and the progress of the take-private transaction.

¶ [501] Mr Taylor says that in any case, trading in the Company’s ADSs after the Announcement Date was driven by idiosyncratic strategies of certain types of shareholders (merger arbitrageurs) rather than issues related to the Company’s intrinsic value. The shareholder base shifted away from funds investing on the basis of the Company’s intrinsic value and growth over the long term. In Mr Taylor’s opinion the Company’s share price after the Announcement Date has limited relevance to assessing its fair value. Page 155 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 156 of 257

¶ [502] Further, Mr Taylor does not consider the Company’s share price to be a reliable indicator of fair value as the stock did not trade efficiently. Mr Taylor says that no reliance can be placed upon the Company’s trading price as a measure of fair value, particularly after the Announcement Date.

¶ [503] Mr Taylor says that Professor Lehn’s methodology for adjusting the share price forward from 11 January 2022 to the Valuation Date during a period of significant COVID-19 driven market volatility is unreliable and would not accurately reflect the fair value of the Company as at the Valuation Date, even if he used an accurate starting point.

¶ [504] Mr Taylor says that Professor Lehn’s analysis is reliant on exceptionally volatile data and therefore carries an excessive risk of significant pricing errors that are common in times of market displacement.

¶ [505] Mr Taylor says that Professor Lehn did not conduct a valuation of the Company using the DCF method which Professor Lehn says cannot provide a reliable estimate of the Company’s fair value for three principal reasons: (1) Professor Lehn considers the management projections have an upward bias; (2) Professor Lehn considers the management projections to be stale as at the Valuation Date; and (3) the Management Projections do not forecast the Company’s performance out to steady state.

¶ [506] Mr Taylor disagrees with Professor Lehn and says that if the restriction he places on circumstances in which the DCF method were imposed it would be impossible to apply the DCF method to most, if not all, companies.

¶ [507] Mr Taylor says that Professor Lehn’s conclusion that the 2021 Management Projections have an upwards bias is based on a selective reading of the Proxy Statement which sets out various assumptions underlying them. Mr Taylor says that Professor Lehn highlights certain assumptions, whilst ignoring other assumptions with a downward bias relating to, for example, material increases in certain costs. Page 156 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 157 of 257 Mr Taylor says that even the assumptions Professor Lehn highlights, such as the assumption that there would be no major changes to existing political, legal, fiscal and economy conditions in China, do not reflect an upwards bias; such changes, if they did occur, could be either positive or negative for the Company.

¶ [508] Mr Taylor adds that the Company’s inability (if any) to predict changes to the prevailing political, legal or economic conditions, for example, does not make the DCF method inapplicable. Mr Taylor says that such factors are commonly taken into account and if Professor Lehn were correct the DCF method would not be applicable to any company in the world, and that is simply not the case.

¶ [509] Mr Taylor says that any valuation using the 2021 Management Projections can be updated to the Valuation Date. For example, he says he decreased the 2021 Management Projections’ forecast of the Company performance in 2022 to reflect the resurgence of COVID-19 and he says his discount rate reflects inputs as at the Valuation Date.

¶ [510] Mr Taylor agrees with Professor Lehn that the 2021 Management Projections do not forecast the Company’s performance out to a steady state but says that it is standard practice and uncomplicated to extend them such that the Company reaches steady state. Mr Taylor adds that the need to extend projections, which he says is a commonplace practice in valuations, is not a reason to disregard the DCF.

¶ [511] Mr Taylor says that even if one did not extend the projections, that is not a justification to discard the DCF method. He says it is possible to use the 2021 Management Projections “as-is”, although doing so would undervalue the Company unless a higher Terminal Growth Rate were applied.

¶ [512] Mr Taylor also refers to Professor Lehn’s position that alleged restrictions on distributing the Company’s cash and non-operating assets prevent him from applying a DCF. Mr Taylor says that the matters Professor Lehn cites (such as the Company’s USD 0.04 per ADS statutory reserves) are neither material nor insurmountable. Mr Taylor adds that if they truly rendered standard valuation methods and value of cash distributions unreliable, overseas investors would not have committed US$1.5 trillion to Chinese assets by April 2022. Mr Taylor says that Professor Lehn’s logic implies the DCF could effectively never be used for a Chinese company, which he says is simply not the case. Page 157 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 158 of 257

¶ [513] Mr Taylor says that Professor Lehn’s view that his conclusion that the Revised Merger Price exceeded the fair value of the Dissenters’ ADSs is supported by the merger process is inconsistent with the facts of the case.

¶ [514] Mr Taylor says that the negotiation process in an MBO requires higher scrutiny, due to management’s inherent conflict of interests in such transactions. Mr Taylor says that management as net buyers are highly incentivised to pay as low a price as possible for the target and they have superior access to value-relevant information concerning the target and are able to control whether and when this information is revealed to outsiders. Mr Taylor says that using the information, management can opportunistically take a company private when its trading price diverges from its intrinsic value. Mr Taylor refers to the comments of Kilometre in June 2020 (DA/50) stating that the “timing is good” for taking the Company private given its “relatively low-valuation-point in the cycle” that “provide[d] an opportunity for execution at highly attractive terms”.

¶ [515] Mr Taylor says in such circumstances it is vital that the Special Committee act as a check on management and address conflicts of interest, as well as ensuring a competitive bidding process. Mr Taylor says this did not occur in this case. He says that the members of the Special Committee, both long-standing members of the Board, outsourced various functions to Kilometre who, given that the greater part of its compensation was conditional upon a merger completing, was incentivised to see the merger succeed. He adds that the Special Committee also failed to enforce the Original Merger Agreement and agreed to the Revised Merger Price representing a 22.8% price cut to the Original Merger Price.

¶ [516] Mr Taylor says that the Special Committee did not conduct a close review of the 2022 Management Projections prepared by Ms Chien, who Mr Taylor says was a member of the Buyer Group and close associate of Mr Yan. Mr Taylor says this task was left to Kroll who had no obligation to review the projections and did not do so. Mr Taylor says that Kroll instead assumed that the 2022 Management Projections were prepared in “good faith”.

¶ [517] Mr Taylor said that the Special Committee was not able to secure a competitive bidding process and it conceded minority shareholder protection such as a MoM voting condition. Mr Taylor says that the go-shops (conducted over short periods of 30 and 15 days) were not effective and neither Kroll, the Special Committee or the third parties contacted believed that this was an effective means of outreach Page 158 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 159 of 257 or soliciting competition. Mr Taylor says that any competing bids were unlikely to be successful, given that both Mr Yan and Recruit (collectively controlling more than 50% of the Company’s shares) had committed to support the merger and the Buyer Group had publicly stated that Mr Yan did not intend to sell his shares in any other transaction involving the Company.

¶ [518] Mr Taylor says that neither he nor Professor Lehn assigns CCA any explicit weight in their conclusions. He says they both acknowledge the “substantial limitations in this matter” and Professor Lehn describes his own results as “at best, indicative.”

¶ [519] Mr Taylor says that the DCF method is the only approach of reliably assessing fair value at the Valuation Date. The expert oral evidence

¶ [520] In addition to considering the lengthy written reports of the experts I have carefully considered their oral evidence. They were both subjected to protracted cross-examinations and mercifully brief re- examinations over 11 intense days. It should be clear from what follows and in particular from the determination section of this judgment which expert evidence on the disputed issues I have accepted and which expert evidence I have rejected. Professor Lehn’s oral evidence

¶ [521] Professor Lehn was an impressive expert witness. It was put to him by Mr Adkin early on in the cross- examination that the genesis for the conclusions in his report was his “a priori views about the different methodologies for valuing companies”. Professor Lehn answered the question directly, with authority and balance in the following way: “… I don’t think you can divorce the methodologies from the facts and circumstances of an individual case. So I certainly don’t have an ideological view that, for example, market evidence is always preferable to a DCF, because that depends on the facts and circumstances of the case. As a general matter, all else equal, and based on an enormous amount of academic research over the past 50 years, my view, I think, is consistent with Page 159 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 160 of 257 the peer-review literature that, all else equal, market evidence is far preferable to the other methodologies.” (Day 7/50:11-21).

¶ [522] Professor Lehn was challenged as to his conclusion that a DCF analysis in this case was unreliable and fundamentally flawed and the fact that he had not conducted a DCF analysis. Professor Lehn stated: “I came into this matter with an open mind as to the methodologies, so I considered the possibility of a DCF early on in this matter, and then upon reviewing facts and circumstances it became readily apparent to me that a DCF in this matter would provide the court with an unreliable estimate of fair value.” (Day 7/90:12-17). “… There’s nothing inherently wrong with the [DCF] methodology, but if you don’t have the proper inputs, reliable inputs to use the methodology, then you end up with garbage in, garbage out, and you can go through the motions of a DCF but it gives you then an estimate of fair value that, in my opinion, is unreliable. So hence, I did not attempt to provide an affirmative valuation based on DCF.” (Day 7/91:6-13).

¶ [523] Professor Lehn confirmed that on 11 January 2022 the closing price was $45.83 (Day 7/92:10-11) and that was the price he started with in his AMTP approach as it provided the most reliable estimate of the fair value of a share on that date.

¶ [524] Professor Lehn accepted that the second step in his AMTP exercise involved him performing an adjustment to adjust the actual observed trading price as of 11 January 2022 forward to the valuation date of 26 April 2022 (Day 7/93:25).

¶ [525] Professor Lehn accepted that there were two parts of his adjustment method. The first part is a backward-looking regression analysis where he looks at the historic relationship between how the Company’s share price moved each day with how two indices that he had selected (a market value and an industry index) moved on those days (Day 7/94:9).

¶ [526] Professor Lehn did not appear to take serious issue with the description of the second part of his adjustment method as put to him by Mr Adkin, namely a “forward-looking roll-forward” where he uses the historical relationship identified between the Company’s share price and two indices to predict how Page 160 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 161 of 257 the Company’s shares would have moved during the roll-forward period from 11 January 2022 to 26 April 2022 (the Valuation Date). The Professor added “And when you say how it would have moved, it means how it would have moved but for the information related to the revised merger proposal … and the subsequent information about the merger.” (Day 7/95:14-25)

¶ [527] Professor Lehn added: “… what was the fair value of a share as of the valuation date? The reason for selecting the January 11 stock price is that it was the day immediately before receiving the revised merger proposal, and we know that on January 12, the day before the market learned of the revised merger proposal, there was a significant increase in 51job’s stock price. So that’s why I selected the day before that announcement. I discuss in my report that that is a conservative number for the purposes of this litigation, in the sense that the original merger proposal at $79.05 was still on the table, but the market had become skeptical as to whether or not that transaction would proceed, because, again, the stock price of 51job on January 16 had dropped to $45.83. Notwithstanding that, that $45.83 still likely reflects to some extent some expected premium of the $79.05 deal went through, and a basic concept in economics and finance is that as a general matter a company’s stock price on any given day is equal to the market assessment of its stand- alone value plus an expected premium in some takeover. And, as I point out in my report, 51job’s stock price on January 16 of $45.83 probably overstates its fair value on that date insofar as it reflects an expected premium.” (Day 7/99:24-25; 100:1-22) The reference to “January 16” was plainly an error and should have been to “January 11”. See Day 7/131:10-23.

¶ [528] Professor Lehn stated that over the three-month period his estimated price moves downwards from $45.83 to $31.11 on the Valuation Date “reflecting the fact that both the market index and the industry index had been declining during that period.” (Day 7/101:18-23).

¶ [529] Professor Lehn added: Page 161 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 162 of 257 “… I think $31.11 is the best estimate. And in science generally, and certainly in the science of financial economics, we estimate. And with estimation, you may have to provide certain caveats and the caveat here is that I think $31.11 is an upper bound on fair value, insofar that there’s was my expected premium in the starting price of $45.83. So I think it does inform an opinion on fair value.” (Day 7/107:13-20)

¶ [530] It was put to Professor Lehn that his approach did not seek to strip out the effect of the original proposal of 17 September 2020 or any of the effects of the developments in the merger process from 16 September 2020 up until 11 January 2022 and his response was as follows: “I don’t explicitly attempt to account for that, Mr Adkin. But again, if we go back to the footnote [footnote 10 second report GA/5/8] that I am referring to, which is if you take the September 16, 2020, price … and if you roll that forward in exactly the same way as I roll forward the January 11 price, you end up with a predicted price on January 11 of, roughly, very close to the $45.83, which is the pre-announcement price that I use for estimating fair value …” (Day 7/118: 12-21) “… if you do that same adjustment, using the same methodology, on the September 16 price, rather than the January 11 price, you end up, number 1, at roughly the same place on the valuation date with an estimate of fair value of $30.88 as opposed to my estimate of $31.11. Furthermore … if you use that same roll-forward methodology on the September 16 price, when you roll it to the January 11 price, which is my starting point, it gets you a price very close to $45.83.” (Day 7/121:18-25 and 122:1-3)

¶ [531] Professor Lehn stated that “the price presumably fell for a number of reasons. One is related to the likelihood of the deal going through … and if you look at the operating performance of 51job over that period, it had been declining. The company was facing headwind, a perfect storm …” (Day 7/132:19- 20 and 25 and 133/1-2)

¶ [532] Professor Lehn accepted as a general proposition that one of the critical questions for the court was whether the actual trading price at the start date of 11 January 2022 was a reliable proxy for fair value, with “the caveat that, if anything, it might overstate fair value.” (Day 7/133:23-24) Page 162 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 163 of 257

¶ [533] Professor Lehn stated: “Now, when you talk about the arbitrage spread, sir, with all due respect, you’re incorrect when you say that it’s driven by the merger price. An arbitrage spread consists of two components. One is the offer price assessed by the probability that the offer will go through, and then the merger arbitrageurs refer to the break price, which is the stand-alone value of the target company if the deal does not go through, which means the arbitrage spread is being determined by both the stand-alone value, ie fair value, and by the offer price, each weighted by probabilities … I think there’s a common understanding of the phrase “arbitrage spread” in the finance profession – to indicate that it is a function of both stand-alone value and the offer price … Their business is arbitrage, but I think it’s misleading to say that they’re not focused in the long-term value of the company, because an important component of that spread, which is something that is prominent in their minds is the stand-alone value of the company, and the stand-alone value of the company is based upon long-term prospects for the company. That’s the so-called break price in the arbitrage spread … even through that holding period may be short they’re still interested in the long- term prospects of the company in which they’ve bought the stock … Manolo Advisors … talk about how the break price of 51job was around $33. And, again the break price to merger arbitrageurs, is the stand-alone value if the deal falls through …” (Day 7/135:19 to 139:16)

¶ [534] Professor Lehn accepted that he had never given a valuation opinion for a real buyer or seller apart from “when my children have bought houses I’ve given them my opinion about the value … they should buy or sell it” (Day 7/154:17-24, 155:1)

¶ [535] Professor Lehn did not accept that “what had happened after the original offer is that trading in the company’s shares was being driven by the merger process and not by fundamental value.” (Day 7/160:14-17)

¶ [536] Mr Adkin provided a helpful recap at the beginning of Day 8 of the trial (the day on which the JCPC were considering Trina Solar) which Professor Lehn did not disagree with: Page 163 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 164 of 257 “… we were considering your AMTP valuation approach. That has two steps, a start date and a roll-forward to the valuation date. The roll-forward itself, we established, involves two steps, a backward-looking regression to establish a historical relationship between the company’s shares and certain indices, and the use of that historical relationship to estimate how the company’s stock price during the roll-forward period to the valuation date would perform, and we were considering your choice of start date, and I was suggesting that the time of your start date, 11 January 2022, the company’s share price was not being driven by matters relating to the merger.” (Day 8/2:15-25, 3:1-2)

¶ [537] Professor Lehn was cross-examined on event studies generally and his event study in particular.

¶ [538] The November 2021 Press Release was put to Professor Lehn who stated: “If you look at the actual wording of the press release, there’s no information in the press release that talks about how the regulations might affect the operations of 51job. There is discussion about the transaction … So I’m not aware of any hard evidence that would allow one to conclude that the stock price decline on November 8 was due to the fact that the company misled investors into thinking that the consultation had to do with factors related to the fair value of the company. There’s simply no evidence in the case on that.” (Day 8/60:3-25)

¶ [539] Professor Lehn provided detail as to the “null hypothesis” and added by way of implicit criticism of Mr Taylor’s approach: “… for an individual company you can’t – it is to me the height of arrogance, and I’m not saying your (sic) arrogant or Mr Taylor, but the methodology that somehow we can predict direction when there’s both good news and bad news I think gives us more credit than we can get. All we can say is whether or not the market reacted to the information … unless you can explain how you can directionally predict when there’s both good news and bad news combined in an earnings announcement, this notion of directionality for one company is something that just cannot be accommodated in the test.” (Day 8/148:15-25, 149:1-4) Page 164 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 165 of 257

¶ [540] Professor Lehn said it was “incorrect” to say that all he had done in his test was to demonstrate that the market was “not perfectly inefficient”. (Day 8/149:5-22). Adding “I’ve never seen the phrase “perfectly inefficient” used in the academic literature in the context of a test. The test again is whether or not you can reject the null hypothesis. It’s a well-accepted test for market efficiency. I’ve followed accepted practice and I’ve rejected the null hypothesis, which establishes, according to the scientific literature on market efficiency, that the market for this stock was efficient.” (Day 8/155:109)

¶ [541] It was put to Professor Lehn that he was “uninterested in directionality and magnitude and simply testing to see whether [he] can reject what [he calls] the null hypothesis is not the accepted standard approach” (Day 8/158:24-25, 159:1-2). Professor Lehn responded: “I disagree again, sir. Again, when you say my indifference towards directionality, it’s not a matter of indifference, it’s a matter that we – neither, Mr Taylor, I or, frankly, I believe, anyone can sit before the fact and predict the directionality when a company simultaneously releases good news and bad news. It’s rather heroic to think that we could and it’s the height of arrogance to think that we can.” (Day 8/159:4-11)

¶ [542] Professor Lehn was challenged but maintained in cross-examination that the Golden Dragon Index was a reliable index to use for the roll-forward. (Day 8/176:2-12)

¶ [543] Professor Lehn was challenged in respect of his relatively limited use of an industry index. It was put to Professor Lehn (Day 9/5:3-12) that there was no rational basis for assuming that the Company’s shares would behave in the same way as the three companies Professor Lehn chose. Professor Lehn disagreed and referred to the “coefficient” (Day 9/5:13-25, 6:1-7). Professor Lehn also referred to the “t-statistic” indicating “that that coefficient is not statistically significant, but’s a real coefficient.” (Day 9/6:9-19). Professor Lehn added “… even if you dismiss the industry index, it would have very little effect on my overall conclusion.” (Day 9/6:20-22). Professor Lehn disagreed that his “roll-forward methodology and its application is misconceived” (Day 9/6:23-25). Professor Lehn added: “I followed accepted practice and I would disagree with that.” (Day 9/7:8-9)

¶ [544] Professor Lehn was also cross-examined at length in respect of DCFs. Professor Lehn was asked whether he agreed with the Company’s position in its written opening that in most cases a DCF is not a reliable valuation approach. Professor Lehn responded: Page 165 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 166 of 257 “… it depends on the facts and circumstances of the case. It’s certainly been my experience that more often than not the DCF will provide a far less reliable estimate of value than market evidence, but I wouldn’t want to leave the impression that it’s never the case that a DCF might be appropriate. The facts and circumstances of this case led me to the conclusion that a DCF analysis would provide the court with an … unreliable estimate of fair value … and far less reliable than the market evidence.” (Day 9/34; 3-13).

¶ [545] Professor Lehn was cross-examined as to the projections and the budget. Professor Lehn added important evidence to place matters “in proper perspective”: “… the company’s revenue had been declining in prior years. Its EBITDA had been dropping. The projected EBITDA was declining. The company was facing headwinds, a perfect storm of deteriorating economic conditions, COVID, increased competition, more regulation, and you can’t just make this a mathematical exercise, as Mr Taylor does, without considering the full backdrop for these projections, and with that backdrop these are inordinately optimistic projections.” (Day 9/62:16-25, 63:1).

¶ [546] It was put to Professor Lehn that he was “looking through this glass very darkly” (Day 9/71:10) and what had been produced was a “central case”. Professor Lehn responded: “And when you say that, you know, there’s a downside and an upside case, this is where this becomes more than simply a mathematical exercise. With the backdrop that the company’s performance had been declining and with the backdrop that there were four or five headwinds that the company was facing, with the backdrop that the stock prices of Chinese companies listed in the United States had been falling, with the backdrop that competitors in the online recruitment industry had suffered stock price declines, when you put all of that together to have one set of assumptions that success will be achieved, in my opinion, is wholly unrealistic, and if that’s the set of projections that is to be used in a discounted cash flow model, which is extremely sensitive to the assumptions you’re making, if you assume success you’re going to get an inflated discounted cash flow valuation.” (Day 9/72:2-17) Page 166 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 167 of 257

¶ [547] Professor Lehn added: “… I do point out that – I mean, frankly, I was shocked when I received Mr Taylor’s initial report. I, frankly, did not expect him to use the 2021 projections, which were almost one year stale as of the valuation date. I don’t think, in my experience in litigation, I have ever seen an expert use projections that were a year old. I mean – and in some sense, I mean, not to over-exaggerate here, but it’s almost like Nero fiddling while Rome burns. The world had changed a lot in that one year and to just make a few minor adjustments that are more of a mathematical exercise than reflecting the economic reality I think is something – is something that is inappropriate. You know, stock prices of companies were dropping dramatically during that period. You know, the Golden Dragon, as I pointed out, which consists of Chinese companies listed in the US, had declined dramatically during that period. Online recruitment companies, their stock prices had been falling dramatically, reflecting the fact that the market had changed its expectations about the future prosperity of these companies, and none of that is reflected in any adjustment that Mr Taylor made to the 2021 projections, and I do – I do discuss some of that in my report.” (Day 9/86:13-25, 87:1-10)

¶ [548] Professor Lehn stated: “… the context I think for looking at the accounting data, the revenue and the EBITDA and so forth, is to present a full picture. But ultimately what matters in the valuation of these companies, and the valuation is based on prospects that go beyond simply the near term. And the two companies that are closest to 51job, Kanzhun and Tongdao Liepin, from January 11, 2022, to the valuation date suffered very large stock price declines. And, again, the backdrop is I think there was a lockdown in Shanghai. There were other issues related to COVID in China during that period. You know, the economic environment was not strong. It was deteriorating. So when you look at the valuation of these companies, which is the ultimate test, you get a very different picture than what is being presented by Mr Taylor here.” (Day 9/92: 17-25, 93:1-7) Page 167 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 168 of 257

¶ [549] It was put to Professor Lehn that extending projections out to steady state is not unorthodox (Day 9/95:16-17). Professor Lehn responded: “If one has reliable data or one has a reliable basis for making those projections, but it should not be a simple mathematical exercise, which is the way Mr Taylor does it.” (Day 9/95:19-22)

¶ [550] It was put to Professor Lehn that adopting a transitional period was “all perfectly logical and orthodox.” (Day 9/99:16). Professor Lehn responded: “If you have the data. But if you don’t have reliable data, then – there’s nothing inherently wrong with the transitional period, but if you don’t have the proper inputs, it’s going to be garbage in, garbage out.” (Day 9/99:20-23)

¶ [551] Professor Lehn also stated that what should be recognised is that “all companies, no matter how successful they are, will at some point run out of new projects that can create value.” (Day 9/104:15- 18). The Professor added: “So, I mean, all companies are limited in projects that will create value. And one of the principles of a sound DCF is that your projections get you to that point where any growth beyond that point is value-neutral. So you’re still making profits on the products to that date, but you can’t perpetually keep finding new things that create new value. When you look at the projections that Mr Taylor has in that intermediate period, even at the end of that period he has his company, 51 job, operating returns on capital that are essentially above the weighted average cost of capital. And that’s something that no company can do, and so he’s modelled it in a way that doesn’t reflect economic reality. And what it’s done is it’s vastly inflated his DCF valuation. He gets to $111, or something, which to my knowledge, you know, no other market participants had come close to you. You know, one benchmark is if you look at his projections versus the Morgan Stanley projections, his are substantially above Morgan Stanley I think by 10 percentage points in terms of EBITDA margins, and he exacerbates that in his second stage, so he’s projecting enormous optimism into his DCF model.” (Day 9/105:2-24) Page 168 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 169 of 257

¶ [552] Professor Lehn was cross-examined in detail on the topic of terminal growth and large chunks of a textbook were read to him and questions asked in respect of return on invested capital (“ROIC”) and weighted average cost of capital (“WACC”). It was put to Professor Lehn that the Company had “a strong and well-established brand in a market with few players” and he responded: “It was not translating into large differences between its ROIC and WACC. In other words, the words on paper I hear them, I see them, but the ultimate test is whether or not they had an advantage that would allow them to generate in perpetuity returns on capital that were substantially greater than their cost of capital. And, again, there’s no evidence that they were doing that at the time, let alone that they would be able to perpetually do this in the future.” (Day 9/140:13-21)

¶ [553] Professor Lehn referred to the convergence theory and cases where “the notion that you can just cavalierly take a perpetual growth rate without checking to make sure you have the appropriate amount of investment” was rejected. (Day 9/173:14-19). The Professor continued: “… if one, again, selects a perpetual growth rate, you should forensically solve for what the implied RONIC is, and for every perpetual growth rate and every given level of investment in a DCF model, once you select a perpetual growth rate you can forensically solve for what the implied RONIC is. The problem with some of these valuations out there, including Mr Taylor’s, is they don’t check the implied RONIC. They just cavalierly take a perpetual growth rate. They don’t check to make sure that you’ve got realistically the amount of investment that would support that growth. So I’m not saying it’s a waste of time to pick a perpetual growth rate. What I am saying is if you do select a perpetual growth rate, as Mr Taylor has, it’s incumbent on you to then check to make sure that that makes sense in the context of your DCF model, namely you have enough investment built into that model to support the growth. Otherwise you end up having to debate between experts is the perpetual growth rate 4% or 6%? But if it’s not tethered to the investment, it becomes a purely mathematical exercise and it violates some basic valuation principles.” (Day 9/174:14-25, 175:1-10) Page 169 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 170 of 257

¶ [554] Professor Lehn continued: “… if one wants to select a perpetual growth rate and if Mr Taylor believes the perpetual growth rate is 4.6% or whatever it is, then that’s fine. But you can’t stop there. You then have to check to make sure that you have built assumptions into your model that would support 4.67% growth. Otherwise, you’re left with an expert perhaps saying, “I think 7% perpetual growth would work” without checking whether that has enough investment, and then it becomes a purely mathematical exercise. And this point, Mr Adkin, is one of the reasons why so many of us feel that if you can establish for the market for a stock is efficient and there’s no significant MNPI, that market prices are far superior to a DCF, because if you get this wrong, and we’ve seen there are intricate assumptions that go into the calculations of these numbers, and these are numbers that have a big effect on the DCF value, the DCF is – is very sensitive to these types of issues. And if you get it wrong, your DCF is going to be wrong. And, in my opinion, Mr Taylor has gotten it wrong. He has a value in his DCF that no one seems to have. I mean, market participants didn’t have it. Morgan Stanley didn’t have it. The shareholders who voted in favour of the merger certainly didn’t believe it. Otherwise, why would they have voted in favour of it? And it’s largely an artefact of bad projections, a low discount rate, namely WACC, and a highly inflated terminal value for the reasons that we’ve discussed.” (Day 9/176:7-25, 177:1-10)

¶ [555] Professor Lehn was also cross-examined in detail on the discount rate and other issues in respect of a DCF approach.

¶ [556] Professor Lehn stated: “… again, discounted cash flow valuations can be very sensitive to seemingly small differences and assumptions … so many of us in finance, especially academic finance, place much more credence on market prices than on the DCF model of one person because if that person is wrong and there are so many moving parts in the DCF model, if they’re wrong about a lot of things, that can have a huge impact on the DCF valuation, and those are things you don’t worry about with respect to market evidence.” (Day 11/27:2-17)

¶ [557] Professor Lehn continued: Page 170 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 171 of 257 “But the more important point is that this again shows the difficulty of doing what Mr Taylor did, which is just abstract away from reality and make a single assumption with no sensitivity, no base case, no bull case, no bear case, just one assumption and derive his entire fair value estimate based on that one assumption because the potential risk of an error is extremely high here, and that can obviously have a big impact on the DCF valuation.” (Day 11/28:19-25, 29:1-2)

¶ [558] Professor Lehn was cross-examined about his evidence in another case in Delaware (Pivotal) and his use of a DCF approach and responded to the effect that each case relies on its own facts and circumstances: “… given the facts and circumstances of that case, which differs substantially from the facts and circumstances of this case” (Day 11/35:11-13)

¶ [559] Professor Lehn added: “There was only one set of projections, and we discussed I believe on the first day the assumptions that went into those projections, which, in my opinion, are unduly optimistic, and I think Ms Chien may have said the "best available estimates" to Kroll. They might have been the best available but, in my opinion, they're not the expected values based on the inherent optimism in the underlying assumptions. The second major factor, Mr Adkin, that I think distinguishes Pivotal from this case, as I recall, because this Pivotal case I think was three years ago, is the staleness issue. Here, you know, we have, you know, certainly a major change in the business environment facing 51job from the 2021 projections to the 2022 projections, so the ones that Mr Taylor is using are very, very stale, but then the 2022 projections were developed, I believe, in late January of 2022 and the world changed dramatically between that time and the valuation date in ways that were adverse to 51job and other Chinese technology companies, especially those listed in the US. You know, we know the Golden Dragon Index declined dramatically during that period. We know that Kanzhun fell by 46% at stock price. Liepin I think fell by 25/30%. So the world was changing. The Ukraine war I think began in February of that -- shortly after those Page 171 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 172 of 257 projections were made. There were lockdowns in Shanghai. COVID was having greater travel -- impact on travel restrictions in China. So the world changed dramatically, and I don't recall any of that being operative in the Pivotal matter.” (Day 11/38:4-25, 39:1-9)

¶ [560] Professor Lehn was also cross-examined on the merger price. Professor Lehn stated: “… I discuss merger price and in effect as an upper bound on fair value, that fair value would be less than merger price.” (Day 11/43:11-13)

¶ [561] Professor Lehn added: “… go back to facts and circumstances. So, for example, if the initial bid is made by an outside private equity firm, such as DCP, and it’s publicly announced that any party could now look at this company and choose to make a bid or not make a bid that, I think, is sort of a fairly important factor to consider, because it wasn’t management, if you will, that initially timed that first proposal.” (Day 11/48:1-8)

¶ [562] It was put to Professor Lehn that he had not considered in his reports whether there was any evidence of “opportunistic timing in this case” and the Professor responded: “… I have not come across anything that would cause me to believe or conclude that that’s the case. Certainly no evidence that would be reliable to conclude that that was the case.” (Day 11/48:17-20)

¶ [563] Professor Lehn was asked if he was suggesting that it was a realistic prospect on 1 March 2022 for somebody else to come and bid for the Company and he responded: “Well, I believe the go-shop started, I believe, on March 1, it was a 15-day go-shop. The company had previously announced as far back as September 2020 the original merger proposal, and then, you know, there are several other takeover announcements that were publicly released, so market participants had had, you know, 18/10 months to kick the tyres and see if they felt a higher bid would be in their economic interest. So the answer is, yes, Page 172 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 173 of 257 I mean, if in fact fair value exceeded that $57.25 I would have fully expected that someone would have come forward and indicated an interest or made a competing bid… If in fact -- and, again, I think it's important to keep in mind that, you know, if Mr Taylor is right and fair value was $111 a share, and he bases his fair value estimate on public information that was available to -- for the most part to every other market participant, not only individual investors and institutional investors but strategic buyers, other private equity firms, and if he could arrive at a value that is so much above $57.25 a share, and if he was right, I fully would have expected somebody would step forward and say, "Look, we want to buy this company because we could make a huge profit by buying it above 57.25 and then capturing the fair value of … “$111 a share”.” (Day 11 59:11-22, 60:6-20)

¶ [564] It was put to Professor Lehn that none of the analysts’ target prices could provide any evidence that the merger price was above fair value and he responded: “Well, I disagree in the sense that, I mean, if Morgan Stanley believed that fair value, ie stand-alone value, if you will, was greater than the offer price of $57.25 I would expect to see some commentary such as, "This offer price is way below what we estimate stand- alone value to be." And, you know, certainly I've seen cases over the years where analysts would look at an offer price and they might say, "This appears to be sort of a low price not necessarily low relative to stand-alone value but low relative to what we think the company might fetch if another bidder came in." So it's not as if, again, they just robotically take an offer and say, “We're going to make that our price target."…. They could go above or they could go below.” (Day 11/64:1-15)

¶ [565] Professor Lehn was cross-examined as to the role and activities of the Special Committee.

¶ [566] It was also put to Professor Lehn that on the facts and circumstances of this case the go-shops were never likely to produce a competing bid. Professor Lehn said he did not have “first-hand knowledge of that” (Day 11/126:12) and added: “Again, I go back to the notion that if the issue is fair value was substantially above that $79.05 a share bid, I would find it not very credible that active market participants would Page 173 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 174 of 257 leave as much money on the table as would be suggested by Mr Taylor’s estimate of fair value.” (Day 11/126:16-21)

¶ [567] Professor Lehn added: “… there is an enormous amount of value being left on the table by these market participants, and it’s generally not been my experience that in competitive capital markets that much money is left on the table.” (Day 8/11/128:18-22)

¶ [568] Professor Lehn stated: “… I’ve been studying mergers and acquisitions for 40 years and it’s certainly been my experience, not just anecdotally but from looking at data, that it’s a very rare case where fair value, ie stand-alone value of the company exceeds merger consideration. For a publicly traded company, that’s very, very rare because we know that mergers entail on average premiums of 30 to 40% over the stand-alone value of the company as measured in the markets prior to the first merger proposal …” (Day 11/153:2-11) “… I’m telling you that the general presumption in financial markets is that for publicly traded companies that trade, for example, in the NASDAQ Global Select Index as 51job did, which has the strictest listing standards among that group in NASDAQ, that the presumption – and I’m not saying that the court need – can rely on presumption, but the presumption before you even look at the data is that the market for those stocks is efficient. And then, of course, we do our tests, and the tests, in my opinion, verify what is the presumption, the general presumption … in the scientific literature the general presumption among academics is for these publicly traded companies it’s safe to assume if they trade on NASDAQ, the NYC, that the market for those stocks is efficient and, therefore, unless there’s MNPI, positive MNPI … the general presumption is that pre-announcement price reflects either fair value or fair value plus an expected premium.” (Day 11/155:5-25, 156)

¶ [569] Professor Lehn added: Page 174 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 175 of 257 “… suppose that fair value was $111 a share. That’s $50 a share over the merger consideration. That $111 that Mr Taylor arrives at was based on, for the most part, public information that the merger arbitragers, among others, would have been able to vote, and I can almost assure you that they’re not going to leave $50 a share, the spread between 111 and 61, they are not going to leave that on the table, and if they really believe that fair value was as high as Mr Taylor would have us believe, they would not have voted in favour of a $61 a share merger.” (Day 11/156:14-25)

¶ [570] Professor Lehn added: “… in this case we’re talking about a hypothetical spread of $50 a share between what Mr Taylor claims is fair value and the merger consideration of $61 a share … it is naïve to think that that much money would be left on the table by a group of investors who are highly informed and have the capacity to do these types of valuations, certainly to do what Mr Taylor did if they believe that was valid.” (Day 11/157:10-17)

¶ [571] Professor Lehn added: “… I think it’s incredibly naïve to think that both categories of investors are so highly irrational such that they’re missing the fact that fair value is substantially above the merger consideration. It’s assuming that those so-called long-term investors are highly irrational because they’re selling their shares at a value that is less than the fair value they truly are supposed to be aware of. By definition, you said that Mr Taylor claims that these long- term investors are focused on intrinsic value, then why are they selling? They must feel the price they’re getting is higher than their assessment of fair value. They certainly wouldn’t sell if they felt fair value was higher than the merger consideration. And then the second group that we’re claiming is highly irrational are the merger arbitragers, who are highly well-informed investors who want the deal to close because, again, most of the time fair value is less than the merger consideration. So, of course, they want the deal to close. But if they felt fair value was substantially above the merger consideration they – it’s just incredibly naïve to think that they would blindly vote in favour of the deal.” (Day 11/158:19-25, 160:1-16) Page 175 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 176 of 257

¶ [572] Professor Lehn was also cross-examined in respect of the comparable companies analysis part of the case and the Professor spoke of his objective approach and criticised Mr Taylor’s “incredibly subjective” criteria. (Day 11/173:13-14) Mr Taylor’s oral evidence

¶ [573] Mr Taylor, despite a lengthy, rigorous and penetrating cross-examination, stuck to his guns in respect of his DCF approach in this case.

¶ [574] Mr Taylor said they “considered the threat of BOSS very carefully.” (Day 12/169:16)

¶ [575] Mr Taylor stated: “To the best of my understanding, there hadn’t been any restriction on any documents that were provided to me and I’m satisfied that I have sufficient information to render a wholly independent opinion.” (Day 12/176:15-18)

¶ [576] Mr Taylor accepted that “it’s always helpful to perform a cross-check” (Day 12/183:24, 184:1). Mr Taylor appeared to accept that with a DCF “quite small changes to assumptions and inputs make a big difference to the figure that comes out of a DCF” (Day 12/184:13-15) and stated: “Yes, and I think it’s important to obviously sense check the inputs into the DCF itself as well.” (Day 12/184:16-17)

¶ [577] Mr Taylor accepted that 16 September 2020 was “an unaffected date” “Unaffected by the merger, yes” (Day 12/184:18-22). Mr Taylor stated: “I think certainly in the year before September ’20 in particular, I found no evidence of efficiency looking at earnings announcements. But then I think to do a meaningful roll- forward analysis or some adjustments to that so-called unaffected number, the 19 months just precludes anything to be done with any level of confidence – any sufficient level of confidence.” (Day 12/186:1-7) Page 176 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 177 of 257

¶ [578] Mr Taylor accepted that on 16 September 2020 the unaffected market price was $68.12 per share (Day 12/187:9-16)

¶ [579] Mr Taylor sought to remind the court “about the change in composition of … the constituents of the Golden Dragon Index.” (Day 12/194:24-25, 195:1-4)

¶ [580] Mr Taylor stated that he did not claim “to have expertise on co-movement” (Day 12/198:11) and added “there’s certainly no consensus around co-movement” (Day 12/198:22-23).

¶ [581] Mr Taylor said “I think we provided a fairly comprehensive critique in the supplemental report with respect to the shortcomings of the … Golden Dragon Index and particularly around volatility and the composition as well.” (Day 12/199:17-21)

¶ [582] Mr Taylor was asked whether through the period September 2020 through to the valuation date the situation was improving or getting worse for the PRC stocks listed on US exchanges (Day 12/201:7- 15). Mr Taylor candidly responded: “Yes, I would say there was – there was challenges with – particularly in March, with the outbreak of Omicron in Shanghai in particular, there was the Ukraine war as well, and then there was HFCAA considerations as well.” (Day 12/201:16-20)

¶ [583] Mr Taylor did not deny “there were some challenges with China, I’ve just highlighted some of these.” (Day 12/203:1-2)

¶ [584] Mr Taylor stated, “We need to make clear that these projections put forward by Morgan Stanley in March and April 2022 are simply not fit for purpose … so that I do not place any importance on those Morgan Stanley projections.” (Day 12/205:6-8 and 7-19).

¶ [585] Mr Boulton helpfully finished off Day 12 of the trial as follows: “And so to round off the day by putting this all together Mr Taylor, I’ve gone through over the last 45 minutes examples about the company’s own releases, most of which you saw as being negative, the increased expenditure on sales and marketing, the app removal, the BOSS competition, the regulatory and compliance burden, the increase in headcount and Page 177 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 178 of 257 labour costs, what happened to the indices, what happened to Kanzhun, what happened to Liepen, the geographical trends, the resurgence in COVID and Morgan Stanley’s view on EBITDA margins, and my suggestion to you it’s inescapable that between September 2020 and April 2022 the value of 51job would have fallen and probably very significantly?” (Day 12/206:25, 207:1-12)

¶ [586] Mr Taylor did not give a direct response to this important question. He simply responded “the indications that you’ve summarised would – would indicate that the share price would go down” (Day 12/207:13-14). He did not comment on the level of that reduction or whether it would have been a very significant fall. Mr Taylor however added “a couple of things”. He believed that “the starting point was not a fair indication of fair value” (Day 12/207:19-21) and he said that “we need to be careful about the margins in 2021 and the impact of the sales and marketing expense because I don’t believe that that’s necessarily reflective of a normal time period. Advertising and promotion tripled from 2010 levels, and I understand the Company has some reasons for that. There’s also reason to believe that that would not be sustained going forward.” (Day 12/207:24-25, 208:1-6)

¶ [587] Mr Taylor was cross-examined at length in respect of the market price approach to assessing fair value.

¶ [588] In respect of the share price of a listed company Mr Taylor was asked “If the listed company is actively and efficiently traded … and if there is no MNPI, then it’s the best evidence of value isn’t it?” (Day 13/2:21-23). Mr Taylor could not quite bring himself to agree that it was the best evidence of value and responded: “Yes, I think it would be a good indication of value.” (Day 13/2:24-25)

¶ [589] Mr Taylor accepted that an income approach refers to the DCF approach (Day 13/2:2-4). It was put to Mr Taylor that “before you move to the income approach for any listed company you would want to start with the market approach and establish whether or not it was reliable on the facts of that case?” (Day 13/3:13-17). Mr Taylor responded: “Yes, I think that’s fair” (Day 13/3:18).

¶ [590] Mr Taylor accepted that he did not say anywhere in his first report that “the volume of trading in 51job’s ADS is inadequate to be able to rely on market price” (Day 13/8:3-6).

¶ [591] In respect of his event study, it was put to Mr Taylor that for his analysis he decided to focus on 51job’s quarterly earnings releases and to ignore all other corporate events (Day 13/6:15-17). Mr Taylor Page 178 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 179 of 257 responded: “I think for efficiency testing, I think adopting quarterly earnings approaches – quarterly earnings announcements is a correct approach.”

¶ [592] Mr Taylor was cross-examined as to his approach on magnitude and directionality and mixed news events. Mr Taylor stated: “What the literature says, my reading of the literature, starting from Cammer v Bloom says that when you’re testing for efficiency, cause and effect, you have to take into account directionality, and that’s supported by Damodaran, Bowman and MacKinlay, and what I’m saying is that to ensure that we treat the event in a fair way, it’s appropriate to put in place thresholds, and I’ve taken MacKinlay’s suggested thresholds in his aggregate analysis to assist me in coming up with a fair determination of directionality in each event”.

¶ [593] Mr Boulton referred to “the hypothesis that every over or under achievement greater than 2.5% should produce a statistically significant price response. That’s a Travis Taylor hypothesis that is nowhere in the literature.” (Day 13/56:10-14).

¶ [594] Mr Taylor responded “That is my assumption in my event study” (Day 13/56:15). The following illuminating exchange then took place and it is just one illustration of the unreliability of Mr Taylor’s conclusion on his event study having used his subjective decision rules: “Q. The next hypothesis we’ve already discussed, that’s where surprises on revenue and EPS conflict. You conclude on a neutral reaction. I think I’ve put a variant of this one to you, but just to confirm, if a company has a positive revenue surprise of 3%, slightly up and a negative earnings surprise of 50%, so revenues just a bit above the market, and EPS in half the market, your decision rules say the expected price reaction would be neutral. A. Under my decision rules, yes. Q. And that’s obviously on my example a nonsense, isn’t it? Let me make that good and you can then answer. If the price – if the market price dropped in response to 3% revenues up, 50% EPS down, you couldn’t conclude that that was inconsistent with market efficiency. A: So what I’ve done in every case, as you know, is look at each announcement, each quarterly announcement, and also take into account, you know, the contemporaneous Page 179 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 180 of 257 analyst reports. I don’t recall a situation of – of that extreme example and if – if there were, I would be reviewing it. Q: Well, if you were going to have to review your decision rules, that means they’re not scientific, are they? They’re not hard and fast rules, that you’re always going to follow. You’re saying “Well, I follow them until I find a problem and then I’ll subjectively change them”? A: Well, I’ve done to my decision rules. I don’t recall finding an example that – to the extremity that you’re talking about.” (Day 13/56:16-25, 57:1-21)

¶ [595] A further exchange damaging for Mr Taylor took place at Day 13/58:22-25, 59:1-25, 6:1-25 with Mr Taylor, mastering the understatement, finally, properly and inevitably concluding that under his decision rules and based on his personal views of directionality “I would acknowledge that there is an element of subjectivity to this and I could get directionality wrong.” (Day 13/61:9-11)

¶ [596] Mr Taylor was cross-examined at length as to his event study. Suffice to say Mr Taylor’s event study did not survive the cross-examination.

¶ [597] Mr Taylor sought to maintain that his “approach taking into account directionality is the same approach as Professor Lehn. He just does not take into account directionality.” (Day 13/81:24-25, 82:1). Mr Taylor was cross-examined as to his statement that both approaches yield a confidence level of 95% and it was put to him that his approach yielded a confidence level of 97.5%. Frankly I found Mr Taylor’s evidence in this area difficult to follow and his explanations difficult to accept. My difficulties are apparent in the following exchange: “Q. Do you accept that the second sentence that I just read to you is wrong, your approach yields a confidence level of 97.5%. A. I accept that my supplemental report is far clearer and -- and I do agree that it is confusing in -- in my first report, but it's cleared up, I adopt the same approach as -- as in the first report as I do in my supplemental report. JUSTICE DAVID DOYLE: Sorry, just -- I'm a very simple soul. I've listened to this exchange with interest, but just limit yourself to the first report. We'll be coming on to the supplemental, your second report. What confidence level were you adopting in your first report? Page 180 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 181 of 257 A. So the confidence level I was adopting was the same as Professor Lehn, a 95% confidence level, but I was -- I was concerned with directionality. So if it was the wrong direction, I excluded it as being efficient. So that's the difference between a so-called one-tailed test and a two-tailed test. JUSTICE DAVID DOYLE: Because I thought in parts of this first report I did see 97.5. A. Yes. Yes, I agree it's confusing, because part of the distribution curve you've got 2.5% on each side of the distribution curve which relates to positive and negative earnings surprises, and so taking into account directionality, I exclude events which are contrary to my opinion on directionality. But, as I said, Professor Lehn agrees that the confidence level is the same, it's just a matter of directionality. MR BOULTON: A third attempt to get an answer, Mr Taylor. What you have said twice is, well, you tried to clear it up in your second report. What I'm putting to you is what you wrote in your first report was wrong as a matter of statistics and that's why you felt the need to clear it up. A. Yes, I did feel the need to clear it up and it was cleared up in appendix C to my supplemental report. JUSTICE DAVID DOYLE: What about the first part of that question, it was wrong as a matter of statistics? A. No, I don't agree it's wrong as a matter of statistics. The statistical hurdle was the same that I applied, the critical value of 1.97 was consistent in both reports. MR BOULTON: Yes, I know you use the -- well, we can go round the houses but you can't keep falling back on the fact that the size of the t-statistic was the same when you are performing a fundamentally different test from Professor Lehn. Professor Lehn is testing for price responses in either direction. You are testing for price responses in one direction, and that is a fundamental difference. A. I agree it's a fundamental difference because I take into account directionality and Professor Lehn doesn't …” (Day 13/82:2-25, 83:1-25, 84:1-3)

¶ [598] Mr Taylor accepted that he did not have a better understanding of statistics than Professor Lehn adding “I respect Professor Lehn’s views …”. (Day 13/88:19-22) Page 181 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 182 of 257

¶ [599] Mr Taylor was referred to his Excel event study and accepted that he had changed “the formulae that is used to produce 1.97, so although the figure is the same, it is the figure for the two-tailed test” (Day 13/92:19-23). It was put to Mr Taylor that because he continued to predict direction he was not performing a “two-tailed test.” (Day 13/92:19-23). Mr Taylor responded: “Well, it’s a two-tailed test with a 95% confidence level and then I ignore – I take into account directionality which, as I say, I explain in appendix C to my supplemental report.” (Day 13/93:1-4)

¶ [600] Mr Boulton methodically went through the specific examples in appendix 6.10 from Day 13/93:19 onwards and challenged the totality of Mr Taylor’s event study. I do not set out the totality of the cross- examination on this topic in this judgment but indicate my conclusion upon it in the determination section. Suffice to say I do not accept as reliable the conclusions Mr Taylor draws from his event study.

¶ [601] Mr Taylor found it very difficult to immediately concede that Professor Lehn’s conclusion that 51job’s negative but not statistically significant abnormal return following the Q4 2018 earnings release could be attributed to the mixed news in 51job’s earnings release. Mr Taylor initially sought to avoid giving a direct answer to the question and responded “that’s as subjective as mine is” (Day 13/135:24). The following exchange took place: “JUSTICE DAVID DOYLE: "Is it a reasonable conclusion?" was the question. A. Of Professor Lehn's? JUSTICE DAVID DOYLE: Where Professor Lehn concludes that it's mixed news, is that a reasonable conclusion? I think was the question. A. Well, on my -- on my reading of the analyst reports, I came to a conclusion that I shouldn't overturn the positive response as a result of the -- you know, the earnings and the revenue announcement on that date. Professor Lehn's looked at some news on that particular date and reached a conclusion that it was mixed and there was confounding events, and so overall there wouldn't be a statistically significant movement. I mean, that's his opinion. JUSTICE DAVID DOYLE: Is it a reasonable conclusion? A: It’s not an unreasonable conclusion, but - - Page 182 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 183 of 257 JUSTICE DAVID DOYLE: So if it's not -- it's not an unreasonable conclusion, it's a reasonable conclusion, is it? A. Yes, it's a reasonable conclusion, but our opinions differ – JUSTICE DAVID DOYLE: Yes, understood. A. – and I think reasonable minds can diverge. JUSTICE DAVID DOYLE: Yes.” (Day 13/136:3-25, 138:1-2)

¶ [602] Various guidance and estimates were put to Mr Taylor. To take just one example, a report from Macquarie (GC/5/1) and a reference to an announcement missing a Citi estimate by 5% and also missing the Company’s revenue guidance. Mr Taylor agreed that was “fair” (Day 13/141:5). It was put to Mr Taylor that his overall takeaway from the whole report was “a neutral to positive response” (Day 13/141:25, 142:1). Mr Taylor said “… it needs to be read in its entirety but I don’t disagree that the headline figures not encouraging.” (Day 13/142:10-11). Mr Taylor agreed that if he had “read those reports and determined the price response should be negative, then this way would be consistent with market efficiency and get a green tick” (Day 13/142:17-21). There were many other examples put to Mr Taylor in cross-examination that support market efficiency.

¶ [603] Mr Taylor was cross-examined on his use of the NASDAQ market index and Professor Lehn’s use of the Golden Dragon Index in respect of the event study. Mr Taylor said that “2019 is… is safe" (Day 13/156:1). The following exchange took place: “JUSTICE DAVID DOYLE: When you say – I’m sorry for being slow but we come from different backgrounds and we’re dealing with a subject that’s within your expertise and not mine, that’s why I’m receiving this evidence. A: Yes. JUSTICE DAVID DOYLE: When you say “2019 is safe”, in my simplistic terms, what went through my head then was it’s safe to use the Golden Dragon Index; is that right? A: Yes, in – I mean, I believe that it’s fine to use the Golden Dragon Index within the event study during that period of time, and I acknowledge that the – that the R-squared or – that is calculated by Professor Lehn versus that in the NASDAQ had a higher R-squared, better goodness of fit, as Mr Boulton said. So I acknowledge that. But then, my Lord, might recall we have had some back and forth about the nature of the Golden Dragon Index. Page 183 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 184 of 257 JUSTICE DAVID DOYLE: And the composition changes. A: Exactly, and the volatility later on. But I think Mr Boulton makes a fair point that within 2019 it’s – it would be acceptable to use that.” (Day 13/156:13-25, 157:1-8)

¶ [604] In finishing off with Q3 2019 Mr Boulton put to Mr Taylor that if he “had used the NASDAQ Golden Dragon Index on this date, then you would have had a conclusion consistent with efficiency; yes?” Mr Taylor responded “That’s my understanding …”. (Day 13/158:2)

¶ [605] Mr Taylor stated: “I’ve always said that there’s a degree of subjectivity with respect to the view of the analyst reports.” (Day 13/170:16-17)

¶ [606] Mr Taylor disagreed with what he described as Mr Boulton’s “conclusion” when it was put to him that: “… when one takes into account the analyst reports, when one discounts what I say are your arbitrary decision rules on directionality, and when I correct what I say is an inappropriate one – tailed t-statistic, I can change almost every day within your event study from inefficient to efficient … At the end of that cross-examination you are left with no days or maybe one day which you can say is clearly inconsistent with market efficiency.” (Day 13/174: 22-25, 175:1-11)

¶ [607] Mr Taylor accepted that “randomly” “you would expect unexplained price movements on some days. That’s in the nature of what you’re testing for via your t-statistic.” (Day 13/175:18-21)

¶ [608] Mr Taylor stated that “You would expect to find some random events, yes” (Day 13/176:5). By way of example there was reference to a $4.38 drop on 25 June 2019 and Mr Taylor agreed he remembered that (Day 13/176:10-12). The following exchange took place: “Q. You hadn’t tested to see if that drop was statistically significant, such that you could reject the null hypothesis that it was random, had you? A: I don’t believe so.” (Day 13/176:13-16) Page 184 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 185 of 257

¶ [609] Mr Taylor said that he had no reason to doubt Professor Lehn’s evidence as to the effect in respect of the price fall on 25 June “we can’t be confident that we’re observing on that day is nothing more than a random movement in the stock price of the Company, after you adjust for the market and the industry.” (Day 13/177:19-25)

¶ [610] It was put to Mr Taylor that “In principle you can adjust the traded price to take into account the value of the MNPI, in practice that may be difficult” and he responded “I think that’s fair.” (Day 13/180:24- 25, 181:1-2)

¶ [611] Mr Taylor was cross-examined on his definition of MNPI.

¶ [612] In respect of the management projections and the internal budget, Mr Taylor accepted that he had not seen any evidence of manipulation. (Day 14/11:12)

¶ [613] Mr Taylor agreed that in what Mr Boulton had shown him there was no suggestion of private information passing to the Buyer Group. (Day 14/27:7-10)

¶ [614] Mr Taylor also agreed that he could not “point to an example of Mr Yan behaving anything other than entirely properly throughout this whole process” (Day 14/27:12-15).

¶ [615] It was put to Mr Taylor that in December 2021 there had been a deterioration in the stock market, at least as far as it applied to US-listed PRC stocks and of the industry in which 51job operated and according to Mr Kirchner a drop in the share price below the announcement price is likely. (Day 14/38:6-11). Mr Taylor stated: “What we saw in December ’21 – well I think we have to go back to – from July 2021, and what we saw was a lack of communication from the Company to the shareholder base, and I think, as you go through 2021, there’s - - there’s more and more outreach from shareholders to the Company, and you can see the concern building around whether the transaction will go ahead at $79, and you can see the delta between the share price and 79 start to widen, and that would indicate that potentially the market believes that the chance of a deal going through at $79 had lessened, and then, of course, you get to 8 November, and that press release caused all sorts of issues with respect to people concerned about the fundamentals of the company, whether the price would be recut, and all sorts of things. So Page 185 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 186 of 257 I don’t think you can infer that the - - the drop in the share price is related to the general economy.” (Day 14/38:14-25, 39:1-3)

¶ [616] It was put to Mr Taylor that if the Company “was following the same trend as the US-listed Chinese companies or if it was following the same trend of other companies in its sector, then that would reflect a fall in the fundamental value of the company’s stock …” (Day 14/39:5-9). Mr Taylor responded: “But I don’t believe – as I said before, 51job is not trading on – on fundamentals at – at that time. It was – and to the extent that it was, it had no information. The analyst coverage had dried up and there was no information or forward guidance being provided by the company, and then shareholders were reaching out, asking for information, particularly after the 8 November press release, and they weren’t being satisfied.” (Day 14/39:10-17)

¶ [617] It was put to Mr Taylor that in November 2020 the Company was “continuing to provide detailed information to investors as they enquire” (Day14/45:10-11). Mr Taylor responded “To individuals who enquire at that time, yes.” (Day 14/45:12)

¶ [618] The following exchange took place: “JUSTICE DAVID DOYLE: Sorry, counsel is asking you quite a straightforward question and I just need to get my head round it. This may be oversimplifying matters, but you've got -- 11 January 2022 you have a share price of, say, $45, and then there's a deal on the table at 79 and it was put to you, how could that depress the share price? Can you answer that question? A. Yes, no, so I don't think it's depressing it but what I'm saying is that -- so one option is -- is that the deal goes ahead at 79 but there's obviously other factors at play which result in the -- in the share price being lower. MR BOULTON: But what follows from that, Mr Taylor, is that if Professor Lehn is rolling forward the price on 11 January, he's either rolling forward an unaffected price, if there's no impact of the $79, or he's rolling forward a price that is affected upwards by the remaining possibility of a deal at $79. A. So I think what he's saying is that the -- the 79 deal is at a premium to the -- to the fair value, which is holding the price up, but what I'm saying to you is that the -- Page 186 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 187 of 257 the value -- it's very difficult to determine the fair value of the underlying stock because it's not trading on fundamentals. That's -- that's the problem. That's the problem with the whole analysis is that trying to determine the fair value at that point is extremely difficult. I mean, there's arguments that it could be higher than 79 but, according to and engaging with Professor Lehn's approach, is that, you know, you're trying to assign probabilities and you're trying to put yourself in a position of a myriad investors who have got different views.” (Day 14/56:19-25, 57:1-25; 58:1)

¶ [619] Mr Taylor was asked if he had looked at “cases in Cayman to understand whether the courts here have ever rolled forward from an affected price” (Day 14/60:15-17). Mr Taylor responded “Yes, I understand that it – that it is considered in 238 cases” (Day 14/60:18-19). Mr Taylor continued: “I accept and I note in section 3 of my first report that that – that it has been used in Cayman courses – cases previously, and obviously it needs to be fact-specific as to whether it is appropriate. And, in my view, what Professor Lehn is trying to do here is – is overly speculative.” (Day 14/60:23-25, 61:1-3)

¶ [620] Mr Taylor accepted that Mr Yan himself is not a controller: “Not on his own, no.” (Day 14/63:6-7)

¶ [621] Mr Taylor referred to Mr Hsu’s comment that the Company in August 2020 is at a “relatively-low- valuation point in the cycle” and it was put to Mr Taylor: “But he’s an individual, Mr Taylor. I suggest this is the same problem as with your DCF. Single individuals can have idiosyncratic views that might be right or might be wrong, but what that share price chart or the multiple chart is showing is what the market as a whole, all of the thousands of investors, think is the right price for 51job to trade.” (Day 14/68:16- 22)

¶ [622] Mr Taylor responded “I think as a general point that’s right.” (Day 14/68:23) and added that “we should not lose sight of the fact that this happened, a company involved in the recruitment section was – the offer for this take-private was taken in the midst of a pandemic …” (Day 14/68:24-25, 69:1-3) Page 187 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 188 of 257

¶ [623] It was put to Mr Taylor that “this was not an MBO until Mr Yan became a member of the buyer group on 4 May 2021” (Day 14/74:5-7). Mr Taylor responded: “I think technically that’s – that’s right but it would never have happened without Mr Yan.” (Day 14/74:8-9)

¶ [624] Mr Taylor was referred to the “turn off the spigot comment” in a Signal message but agreed that he was not suggesting “any conscious manipulation of figures” (Day 14/81:7-13).

¶ [625] Mr Taylor said that the November 2021 Press Release “caused serious confusion” (Day 14/85:19-20). In his opinion, if there was a “fuller disclosure” the market would have reacted “less severely because … the market would be reassured, I assume, that there wasn’t any problem with the fundamental value of the company, that the insiders were still committed to the transaction …” (Day 14/85:23-25, 86:1-4)

¶ [626] It was suggested to Mr Taylor that “any uncertainty that there might have been in the market on 8 November can clearly be cleared up quite swiftly and certainly by 11 January. There’s going to be no basis to suggest that on 11 January shareholders or investors or the market in the whole are in the dark about what that press release had referred.” (Day 14/92:14-20). Mr Taylor responded “I disagree completely …”. (Day 14/92:21)

¶ [627] Mr Boulton put it to Mr Taylor that “it’s not unheard of for companies with large cash to trade at a very significant discount to that cash” (Day 14/98:8-10) and Mr Taylor responded “Right. Well, I think I’ve agreed with you there are circumstances where that does happen.” (Day 14/98:11-12)

¶ [628] Mr Taylor accepted that every company has non-public information which is not material (Day 14/105:12-17). Mr Taylor accepted that he did not do a test for materiality. (Day 14/105:18-23)

¶ [629] Mr Taylor said “… trying to infer that the 45.83 in any way reflects fair value is – I mean, I just cannot see how that can be done with any precision whatsoever.” (Day 14/112:9-11) Mr Taylor added: “… it just cannot be done with any precision, as Professor Lehn has tried to do it.” (Day 14/113:25, 114:1)

¶ [630] Mr Taylor stated “it’s not trading on fundamentals, it’s trading according to whether or not the transaction will … complete or not …” (Day 14/116:1-3). Mr Taylor added that “trying to speculate as to every shareholder’s investment strategy is difficult, but it’s reasonable to assume that several Page 188 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 189 of 257 fundamental investors will leave as a result of … the so-called ceiling that I put on it …”. (Day 14/116:12-15)

¶ [631] Mr Taylor was asked “how are your clients going on to realise anything approaching $111 from the value of the shares that they hold?” (Day 14/126:9-11) and he responded: “Because I’m valuing it as of the valuation date, ignoring the shareholders base as of that date. I’m assuming that the merger is a but-for scenario.” (Day 14/126:12-14)

¶ [632] Mr Taylor was cross-examined as to the break price analysis including the Manalo Advisors break price of $32.33 (Day 14/131:2-14). The following exchange took place: “Q. … And just to summarise where we've got to over the last hour and a half or so on break prices, I've shown you contemporaneous -- contemporaneous speaking broadly over a period from about July 2021 up to March/April 2022, but from what we have in dissenter disclosure I've shown you assessments of the break price from actors at the time. I've not deliberately excluded any higher values, Mr Taylor. I'm sure if there are any, you'll be directed to them later. But everything that I've shown you has concluded that the break price or the stand-alone price if the deal doesn't go through is somewhere way less than half your fair value determination. Do you accept that? A. Certainly the numbers that you've taken me to would indicate that. I would also say that obviously the dissenters are doing their own fair value exercise, so they're -- I assume this -- these break prices to be modelling worst-case scenarios. I think what's also clear from the contemporaneous documents that I've seen is that there's a real lack of information that the dissenters have access to that -- given, you know, the lack of disclosure from the companies and so forth. And, finally, I would say that in terms of the way they go about calculating the break price, and it's consistent with the Manalo report that we saw, you know, the Kirchner papers and other papers talk about these chartist-type approaches about rolling forward prices with comparable companies or indices, and they talk about it being largely guesswork, and I would agree with that. I don't think it's any replacement for a fundamental valuation exercise”. (Day 15/8:21-25, 9:25, 1:1-3) Page 189 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 190 of 257

¶ [633] Mr Taylor was cross-examined as “a prelude to the DCF” (Day 15/12:1) as to his background section about the Company and it was suggested to him that “in common with many of the Chinese internet companies, 51job enjoyed growth for nearly two decades driven by three things, and see if this is something you can agree with. First of all, population growth in China, second internet use, and, third, the prevalence of mobile phone technology, you would accept that all three of those were contributors of growth, not only for 51job but actually internet companies in China general?” (Day 15/14:20-25, 15:1-3). The following exchanges followed: “A. I wouldn't -- maybe, generally speaking, that's true, but I don't think population growth is necessarily important for 51job. The other one is urbanisation I think is another factor for companies generally. I think with population growth I think what a benefit that recruitment companies have seen is that the -- the wider economy has outgrown population growth because population growth in China has slowed remarkably and when the economy is growing at a faster rate than the population, what you see is a rush for resources for companies that need -- need personnel, and that provides a big opportunity for companies like 51job, and that was -- that was certainly a factor that Mr Hsu I think referred to in -- originally as well. Q. Once mobile phone use reaches something like saturation point, where you have greater number of mobile phone than there are -- than the size of the population, then that can't continue to grow, can it? A. Well, I don't -- I don't know the specific metrics on mobile phone usage but what we see is -- in China in particular is the -- as people -- as urbanization increases, that is a real driver of -- of recruitment services. Q. And we also saw that 51job had had a market-leading position over much, if not all, of this period, certainly until 2021. A. It -- it did, yes. It had -- it was a pioneer in the industry, Q. But going forward there is going to be quite a battle for market share, isn't there, standing at the valuation date? A. I think what's interesting about the market is that if you refer to Kanzhun's China- listing document from 2022, there's a very good industry section there, and what Page 190 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 191 of 257 amazed me about that was if you stand back and -- because you think about this big threat from BOSS, then you realise the top four companies in the recruitment space only account for 14 or 15% of the whole industry. BOSS -- BOSS as market leader has 6%, I think 51job has 3.5%, and this is in an industry that's growing at a compound annual growth rate of about 28% through to 2026. I mean, the title of that document is "Massive and Fast-Growing Industry". I mean, the online recruitment industry is -- is huge, and although 51job only has a 3.5% market share, it's clearly a fragmented industry and there's huge opportunities for all players.” (Day 15/15:4-25, 16:1-24)

¶ [634] Mr Taylor was cross-examined in respect of BOSS and Liepin. He agreed that on revenue growth in 2020 BOSS grew 95% and Liepin grew 24% and the Company contracted 8% (Day 15/25:24-5). It was put to Mr Taylor that the Company was “very dramatically losing market share to two of its biggest competitors” (Day 15/26:8-9). Mr Taylor stated that “Kanzhun in particular is at a completely different stage to 51job” (Day 15/26:10-11). Mr Taylor added “… these companies don’t have the same business model. I mean, what’s interesting about 51job is almost half their revenue comes from HR services, which is a completely different market to online recruitment…”. (Day 15/28:4-8)

¶ [635] Mr Taylor was cross-examined at length in respect of his DCF and various topics in particular “which projections do you use if you’re trying to do a DCF? How do you extend the projections if you believe you have to extend them? The discount rate, terminal growth rate and all of those inputs. And then, finally, some reasonableness checks.” (Day 15/58:20-24)

¶ [636] Mr Taylor agreed that: (1) “The DCF method is based on the principle that the value of a financial asset is equal to the present value of the future cash flows the asset is expected to generate.” (Day 15/59:2-6); (2) “… for a company’s operating business, the expected future free cash flows are discounted at a discount rate that reflects the company’s cost of capital.” (Day 15/59:7- 10); Page 191 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 192 of 257 (3) and “you focus on free cash flows to the firm.” (Day 15/59:11-13); (4) and “the premise behind free cash flow to the firm is that it is the cash flow available to all funding providers, so depending on the company debt, preferred stock, common stock, corporate bonds” (Day 15/59:14-18); (5) and “the premise is, therefore, that that free cash flow is available to those funding providers.” (Day 15/59:19-21); and (6) “beyond whatever your explicit forecast period is, you then calculate a terminal value based on some kind of assumed terminal growth rate or multiple”. (Day 15/60:15-19)

¶ [637] Mr Taylor thought it “fair” that “It’s normal to start with cash flows for an explicit period.” (Day 15/60:1-3)

¶ [638] Mr Taylor thought it “fair” that “a valuer relying on a single projection needs to ensure that it reflects potential bad outcomes as well as what is hoped for by the preparer of the projection.” (Day 15/64:8- 12)

¶ [639] Mr Taylor accepted that management are “often best placed to pull the projections together.” (Day 15/65:5-6)

¶ [640] Mr Taylor was asked whether he accepted that “one of the problems with a discounted cash flow method is that it’s particularly susceptible to manipulation” (Day 15/69:18-20). Mr Taylor responded: “I was no - -; I don’t accept that. I mean, I accept that - - I accept that certain factors will – changing certain factors will make a meaningful difference to the outcome…” (Day 15/69:21-24). Mr Taylor added “I think all valuation approaches … are susceptible to manipulation. I think all valuation approaches are but I think the beauty of the DCF approach is its transparency and being able to pull together a fundamental analysis of the valuation date without a minefield of MNPI and all these other sorts of complications that we have with AMTP and other things.” Page 192 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 193 of 257

¶ [641] It was put to Mr Taylor in respect of a roll-forward from a market trading price “if you get the right starting point you can’t manipulate it. It’s a consensus view of thousands of investors in the market who are putting their own capital at risk” (Day 15/70:22-5, 71:1-3). Mr Taylor responded: “If you had an efficient, fully informed market and you were satisfied with the – with the veracity of the starting point in principle that is true, and in principle if you had a robust roll-forward, that is implementable, but it’s not the case here.” (Day 15/71:4-8)

¶ [642] Mr Boulton highlighted the “lots of assumptions and inputs” (Day 15/71:11) of Mr Taylor’s DCF model. In respect of the management projections, Mr Taylor stated “… my starting point was to look at the ’22 projections, and I think that would have been my preference, given the proximity to the valuation date, but for the reasons I list in my report, I concluded that the 2021 projections were a more robust starting point.” (Day 15/73:18-23)

¶ [643] Mr Taylor was cross-examined on his reasons for preferring the 2021 Management Projections.

¶ [644] Mr Taylor said he had “not seen or heard a reasonable explanation for the difference between the figure in the internal budget versus what was presented in the ’22 management projections for advertising and promotion, and ultimately that it is – I recognise that is a matter for his Lordship.” (Day 15/83:9-14). Mr Taylor added “I’ve not seen any evidence of wrongdoing. What I’ve seen is a completely inconsistent advertising and promotion forecast put forward in a public document versus what was prepared in an internal budget in the ordinary course of business …”. (Day 15/96:19-23)

¶ [645] Mr Taylor said: “I should say just generally that … the DCF valuation and model is a dynamic model and can be updated as his Lordship wishes if he disagrees with any of the … assumptions that we’re talking about.” (Day 15/136:19-23)

¶ [646] Mr Taylor was referred to Ms Chien’s evidence that not only did the environment get worse between June 2021 and January 2022 but it then worsened again and general comments “on competitive pressure, rising costs, regulatory challenges, regulatory environment more difficult” and the “complete Page 193 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 194 of 257 lockdown in Shanghai and tightened COVID control policies starting in March 2022.” (Day 15/136:24- 25, 137:1-24). Mr Taylor responded: “That’s what she’s saying, and I think this is the sort of beauty of the DCF approach is that the discount rate that I’ve taken is taken on the valuation date, which reflects the environment at that date, but also allowed me to make an adjustment in 2022 for – to reduce revenue by 6.8% at the valuation date for the whole of 2022.” (Day 15/137:25, 138:1-6)

¶ [647] It was put to Mr Taylor he had emphasised the word “whole” but actually all he had done was “assume that COVID would have some impact on just 2022, and then your DCF projections climb back to exactly where they would have been” (Day 15/138:7-10). Mr Taylor responded “I think that’s a perfectly reasonable assumption.” (Day 15/138:11-12)

¶ [648] Mr Taylor accepted that he did not know “with a high degree of precision” when the Company would reach “steady state” (Day 15/160:8-10). He was asked: “If you were standing in 2017 and you had projected growth of 31.3% in 2018, then applying your steady state mentality you would have assumed that that would continue, 25%, 20%, 15%, but in the real world 31% growth in one year was followed by 5.8% in another year” (Day 15/166:8-13). Mr Taylor, displaying again his skill in the art of understatement, responded: “But, yeah, I mean, obviously there’s an element of subjectivity in all this. I will admit to that. But what you’re trying to do is capture what you believe is - - reasonable in the circumstances.” (Day 15/166:14-17)

¶ [649] Mr Taylor stated “… there’s clearly a trend there, but I recognise that beyond a level, there will be a slowing and then you have to recognise that you have to get to a steady state. So I recognise the slowing and then got to 32.5, which, based on my cross-checks, appeared reasonable. So I’m recognising the increase in the trend over the explicit forecast period.” (Day 15/174:17-24)

¶ [650] Mr Taylor accepted that he did not claim to understand the business better than management and had no experience of managing an advertising and promotion budget for an internet services business. (Day 15/175:6-11) Page 194 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 195 of 257

¶ [651] It was put to Mr Taylor that his predictions “are hugely optimistic, there’s absolutely no allowance for a new competitor, a bad year, a collapse in the economy, anything of that nature” (Day 15/177:5-8). Mr Taylor responded: “Well, we reflect all of the known risks in the cash flows at the valuation date … I make sure that the discount rate that is applied is appropriate, including adding in a country risk premium to account for the economic and political risks specific to China.” (Day 15/177:9-14)

¶ [652] It was put to Mr Taylor that “standing in April 2022 you would have to be blindly optimistic to assume that the Company would generate free cash flow of over RMB 2 billion, not only every single year from 2026 onwards, but once it reaches near RMB 3 billion for every single year forever in perpetuity.” (Day 15/177:22-25, 178:1-2). Mr Taylor responded: “I think looking at what the Company itself projected for free cash flow at the end of 2025, and bearing in mind the once-in-a-generation impact of COVID and the nature of the industry itself, I think this is – this is reasonable and achievable.” (Day 15/178:3-7)

¶ [653] It was put to Mr Taylor that “there’s absolutely no account taken in these projections of things that could go badly wrong.” (Day 15/178:17-19). Mr Taylor responded “I believe that they are …” (Day 15/178:20)

¶ [654] Mr Taylor was cross-examined as to the tax rates. Mr Taylor’s assumption was that “15% holds throughout the DCF” (Day 15/190:8-9) but he agreed “it may change in the future. It may be up or down” (Day 15/191:11-12) but he accepted “I don’t have an insight into Chinese tax law ...” (Day 15/191:19-20)

¶ [655] Mr Taylor was cross-examined about his “calculation of FCFF” (Day 15/194:18-19)

¶ [656] In respect of the terminal period and his “terminal calculation” Mr Taylor agreed that what he was saying was “that the company will generate RMB 216 million growing at 4.37% every year for the rest of time just from working capital.” (Day 15/198:4-8)

¶ [657] Mr Taylor stated “I’m not a tax expert by any means … I’ve modelled a scenario where the money is dividended back to shareholders and my understanding is that would attract withholding tax … I guess Page 195 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 196 of 257 the least efficient way to return it, but whether there’s more efficient ways I’m not sure.” (Day 16/18:14- 20)

¶ [658] Mr Taylor accepted that in his conclusion he did “not apply any kind of withholding tax” (Day 16/20:2- 5)

¶ [659] Mr Taylor did not dispute that he was not a PRC or tax expert but maintained that his understanding was that the 10% PRC withholding tax and the PRC government’s currency controls would have a “limited impact” (Day 16/20:12-25). Mr Taylor said he had seen nothing “that would indicate that there is significant problems with people extracting money out of China from overseas.” (Day 16/29:13-15)

¶ [660] Mr Taylor was cross-examined in respect of operating and non-operating assets.

¶ [661] Mr Taylor accepted that almost 90% of the value he attributes to the Company’s operating assets is derived from his extension of the Company’s projections beyond 2025: “Yes … that’s a normal point of running a DCF, in my experience” (Day 16/41:6-10). It was suggested to Mr Taylor that “it’s heroic to think that you actually know what the company will look like or what cash flows it will generate even in 2026, let alone in 2033 and forever.” (Day 16/42:1-4). Mr Taylor simply and unconvincingly responded “Well, this is just a standard part of undertaking a DCF analysis.” (Day 16/42:5-6)

¶ [662] Mr Taylor was cross-examined as to terminal growth.

¶ [663] Mr Taylor said he was “taking out the projections to a steady state” (Day 16/66:17-18) and again agreed “there’s a level of subjectivity.” (Day 16/66:25)

¶ [664] Mr Taylor was cross-examined as to the discount rate.

¶ [665] Mr Taylor accepted that the measurement of beta (the measurement of the extent to which a company is exposed to market-wide systematic risk) upon which he was cross-examined at some length, is subject to “significant uncertainty” (Day 16/74:10)

¶ [666] Mr Taylor accepted that there were “many ways at which you can calculate beta and they will give rise to different results” (Day 16/92:21-24). Page 196 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 197 of 257

¶ [667] Mr Taylor was cross-examined on the equity risk premium and accepted that “different people come up with different estimates of the equity risk premium. Your equity risk premium is the third line of this table [GA/1/118] and you use 5%.” (Day 16/94:12-16). Mr Taylor accepted that “there are many ways to estimate the equity risk premium” (Day 16/94:23-25). Mr Taylor “absolutely” agreed that “reasonable people can disagree and there is a range of figures that reasonable valuers might use for the equity market risk premium” (Day 16/95:9-12). Mr Boulton made it plain that he was “not challenging 5% as being indefensible” (Day 16/96:20-21).

¶ [668] Mr Taylor was cross-examined as to risk premiums including small company risk premium and a country risk premium. Mr Taylor stated “I can’t remember the last time an expert used a small company premium who wasn’t from Kroll. It just seems to be accepted now that it is not applied.” (Day 16/100:6- 9)

¶ [669] It was put to Mr Taylor that in Section 238 cases “experts on both sides frequently thought that a small company size premium was justified” (Day 16/106:7-11). Mr Taylor boldly responded: “I’m aware that they have been applied in – 238 cases historically, and in my view, that is – that is wrong.” (Day 16/106:15-16)

¶ [670] It was put to Mr Taylor that “what the last day and a half has shown is that your DCF model with all of its subjective assumptions is not a reliable basis for estimating the fair value of 51job’s shares” (Day 16/128:11-15). Mr Taylor responded: “And I disagree. I think it’s clearly the best approach to apply in the circumstances, and I believe it to be accurate.” (Day 16/128:16-18)

¶ [671] Mr Taylor accepted that his position was that no weight can be placed on the revised merger price when assessing fair value. (Day 16/128:19-21)

¶ [672] Mr Taylor accepted that there was “an independent special committee” (Day 16/129:8). Mr Taylor stated “I don’t think there was an independent market check” (Day 16/129:10). Mr Taylor stated “I think there were disincentives for third parties to – to bid.” (Day 16/129:14-15)

¶ [673] Mr Taylor did not criticise the appointment of Davis Polk and Kroll (Day 16/133:6-8). Page 197 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 198 of 257

¶ [674] Mr Taylor stated “… I haven’t done an exercise of what fair value was in September 2020. I know that 79 represented a premium to the trading price. I don’t believe that necessarily reflects a premium to fair value …” (Day 16/143:5-8)

¶ [675] Mr Taylor stated “…the company is not trading on fundamentals following the announcement in September 2020 and we know … what happened on 8 November 2021 where it subsequently declined to $45 around the time of Mr Lehn – Professor Lehn’s roll-forward.” (Day 16/152:16-19)

¶ [676] Mr Taylor stated “… it is unusual for … these deals to be reduced in the fashion they were, but I accept your point that it’s not unheard of.” (Day 16/152:16-19)

¶ [677] Mr Taylor was asked if he was suggesting that Kilometre was playing an unusual role (Day 16/176:17- 19). Mr Taylor responded “I think it was unusual, yes. I mean it is clear people were confused as to who they were acting for, and to be charged with responsibility of … negotiating on behalf of the Special Committee I found strange in particular.” (Day 16/176:20-24)

¶ [678] Mr Taylor stated “… Mr Hsu is incentivised to get a deal done. You could argue he’s incentivised to get a higher price, but at the end of the day he is going to have a pay day if the deal is done, and this Special Committee should be acting in the best interest of the unaffiliated shareholders, and they should be – they should be tasked with negotiation, rather than someone who is engaged by the … company” (Day 16/178:7-15).

¶ [679] Mr Taylor was cross-examined as to the possibility of competing bids and the go-shops and I note his evidence in that respect.

¶ [680] Mr Taylor was also cross-examined in respect of the CIC report in the listing document (GC/34/1) for BOSS.

¶ [681] There was reference to the involvement of Kilometre and Mr Taylor was asked whether an “incentive fee on closing is entirely normal for banks and others acting in this capacity on transactions.” (Day 17/10:23-24). He responded: “I would imagine so, yes.” (Day 17/11:1)

¶ [682] Mr Taylor said that he was not sure that the Special Committee charging Mr Hsu with the responsibility of negotiating merger prices was appropriate (Day 17/14:11-13). Page 198 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 199 of 257

¶ [683] Mr Taylor was cross-examined as to a majority of the minority provision and accepted that he had never “been involved in a transaction personally where there has been a majority of the minority provision.” (Day 17/15:7-10)

¶ [684] Mr Taylor was also cross-examined on the Support Agreements and stated that he was not a lawyer “so I wouldn’t be relying upon my interpretation of … those provisions.” (Day 17/20:4-6)

¶ [685] Mr Taylor was cross-examined as to the votes of various shareholders and he said “[i]t’s difficult … to speak for each of the shareholders, as I said, what’s going on and the motivation.” (Day 17/29:13-15)

¶ [686] Mr Taylor was cross-examined on the market price. Mr Taylor stated: “So in the context of late 2021, where you’ve got a price that is not trading on fundamentals, I think Professor Lehn’s proposed probability adjusted sum of the parts assessment is – is simplistic and flawed because we know that, as a result of 8 November in particular … the shares are not trading on … fundamentals and … there’s concern in the market about the prospect of deal cuts and, you know, people walking away from .. insiders walking away from the merger entirely … so it just doesn’t work. That’s the bottom line.” (Day 17/53:19-25, 54:1-5)

¶ [687] Mr Boulton referred to value-relevant information and referred Mr Taylor to Mr Taylor’s comment that the Company seemingly continued to undertake value-accretive activities of which the market was not aware in the period after the Announcement Date and this included the Company creating a new subsidiary in Xuzhou under Lagou to conduct operations in the area of executive search and placement and Mr Taylor was asked whether he remembered this (Day 17/55:17-25, 56:1-3). The following exchanges took place: “A. I do remember this. I’m not sure exactly how material it would be but it’s an example. Q. Well, indeed. You describe it as “value-accretive” and as soon as ask you about it you start backing off whether it’s material. If it’s not material, it’s not MNPI. Page 199 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 200 of 257 A. I – but I agree in this – I was looking at this recently in my review and I would agree that it’s – it’s not material. Q. They were setting up a subsidiary with two employees; was that right? A. Something along those lines. Q. And that’s – the underlying document you cite is at {DA/399/1}. It’s an email chain. A. Yes. Q. But my reading of that – let me see if this is your memory – is that the regional government was proposing to offer support such as rent-free office space, but that would require a new company to be established locally. A. That is – that is my recollection, yes. Q. So to – in order to get rent-free office space for a loss-making subsidiary of Lagou, this company was being established. A. That is my understanding, yes. Q. So I think from what you’re fairly saying, Mr Taylor, you’re accepting that this is not material and, therefore, not MNPI? A. I agree on this occasion, yes.” (Day 17/56:13-15, 57:1-16)

¶ [688] Mr Taylor was cross-examined on the event studies.

¶ [689] Mr Taylor was cross-examined as to the roll-forward of a market price. At one stage during the cross- examination the following exchange took place: “Q. …Now, if you are to do a roll-forward of a market price, would you accept that you would normally – fundamentally you would use the same approach that Professor Lehn has adopted? If you like, I’ll summarise that and then you can say what you would object to in principle. You would identify relevant markets and industry indices, you would test their explanatory power by considering their historical correlation to the - - to the subject stock, here 51job, and then, having worked out your regression, you would roll forward the market price of the subject company by reference to those indices. In principle, that is how one would seek to do such a roll-forward. Page 200 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 201 of 257 A. In principle, I would agree. You would also take into account other factors like the - - the length of time and the events that have occurred in the interim and so forth, but in principle, I agree that that’s how you would go about doing it if you wanted to do it.” (Day 17/73:8-25, 74:1)

¶ [690] Mr Taylor was cross-examined in respect of his assertion that Professor Lehn’s use of the Golden Dragon Index was inappropriate because of its volatility and the risk of pricing errors. Mr Taylor did not suggest the use of another index. (Day 17/76:10)

¶ [691] Mr Taylor was cross-examined as to his criticism that Professor Lehn’s roll forward “fails to consider the possible impacts of co-movement. That’s {GA/4/40}, section 3.3.C” (Day 17/78:14-17). Certain recent literature was put to Mr Taylor and a comment that when “using robust univariate regressions and match control samples almost all evidence of excess co-movement disappears.” (Day 17/79:20-25, 80:1-2). Another comment from an article “there is no longer any evidence of meaningful excess co- movement from either an economic or statistical standpoint.” (Day 17/80:7-20). Mr Taylor accepted that he had not performed any analysis that demonstrates that the Company’s stock had excessive co- movement relative to the NASDAQ Golden Dragon Index (Day 17/82:11-15). The following further exchange took place: “Q. … So the up-to-date literature, I’m going to suggest to you, Mr Taylor, is that any co-movement identified has dwindled to almost nothing, and we can see from the abstract that this article, published in April 2025 - - my Lord, you may be the first judge to cite this article, in fact, if it makes its way into your judgment - - is: “The abnormal return associated with a stock being added to the S&P 500 has fallen from an average of 7.4% in the 1990s to less than 1% over the past decade.” So although you were relying on Professor Greenwood’s analysis, it appears that he thinks that the index effect, he now says on page 39, is near zero. A. He - - this updated article says the - - when looking at S&P 500, the - - the effect has lessened, yes.”

¶ [692] One exchange on the composition of the Golden Dragon Index went as follows: Page 201 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 202 of 257 “Q. … The point I suppose I’m making, Mr Taylor, is you talk about the changes in the Golden Dragon Index, but Professor Lehn’s sensitivities show that using different estimation periods doesn’t change the outcome. And then you criticise his industry index and Professor Lehn shows that if you don’t include industry as a factor at all, it doesn’t materially change his outcome. So please do comment. A. Yes, so I take what he says at face value. So I think he would acknowledge - - I think he did acknowledge under cross-examination that the industry index was not that important to his overall analysis so, therefore, the - - and I accept he did some sensitivity analyses but I think obviously the - - the criticism around Golden Dragon obviously - - obviously stands.” (Day 17/84:23-25, 85:1-12)

¶ [693] In respect of the event studies, Mr Taylor accepted that both he and Professor Lehn “had a similar null hypothesis.” (Day 17/88:18-19)

¶ [694] Mr Taylor was cross-examined at length in respect of one-tail and two-tail tests and directionality.

¶ [695] Mr Taylor was briefly cross-examined on comparable companies and valuation multiples. Mr Taylor accepted that he could not find a company which was sufficiently similar to the Company as a whole to be used as a comparable. (Day 17/104:18-24)

¶ [696] Mr Taylor was cross-examined as to his independence “in light of the Secretariat Singapore report.” (Day 17/112:15-16)

¶ [697] In re-examination Mr Taylor referred to the Company having a significant cash balance and added: “I think if you’re going to entertain a roll-forward with reference to the Golden Dragon Index … what you’re looking at is to separate the non-operating assets, being the large cash balance principally, and then the operating business …”. (Day 17/121:14-25) “Logically, you wouldn’t expect a cash balance of $2 billion to move in line with an index of that nature.” (Day 17/122:3-4) Page 202 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 203 of 257

¶ [698] Mr Taylor said “in terms of an efficiency test” he had never “before come across an event study performed in the way that Professor Lehn performed his event study in this case.” (Day 17/144:6-9)

¶ [699] There was a sting in the tail of Mr Boulton’s cross-examination of Mr Taylor with about 15 minutes left of his 5 and a half days plus cross-examination of time left. Mr Boulton put to Mr Taylor a conflict of interest point. I do not think that the alleged conflict adversely impacted on Mr Taylor’s expert opinion. Mr Taylor responded to the conflict point as follows (Day 17/118): “The entire valuation exercise that I’ve been involved in has been entirely independent, and I stand by all of my independent declarations that I’ve made on each – each valuation. It has had no bearing on my analyses or conclusions whatsoever.”

¶ [700] It would have been better if Mr Taylor had raised the issue in his first report and I note the English Court of Appeal’s criticism of Secretariat entities in A Company v Secretariat Consulting Pte Ltd and other companies [2021] EWCA Civ 6, [2021] 4 ALL ER 348 but on this occasion I have concluded that the alleged conflict does not disqualify Mr Taylor from giving expert evidence in this case. In my judgment, Mr Taylor’s opinions are not tainted by conflict in this case. Whether they should be relied upon is another matter, but I do not exclude them on grounds of conflict of interest.

¶ [701] I should add that it will be clear from the above review of Mr Taylor’s evidence that from timeto-time Mr Taylor referred to Ms Chien as part of the Buyer Group. She was never a member of the Buyer Group although she was a party to the Support Agreements. The repeated mis-references by Mr Taylor to Ms Chien being a member of the Buyer Group, when she plainly never was, taints Mr Taylor’s approach and opinions in this case.

¶ [702] I deal in the Determination section of the judgment with the wildly contrasting positions adopted by the experts in this case and come to conclusions on the various relevant disputed issues raised in their evidence. Generally, I found Professor Lehn’s approach and evidence far more helpful than Mr Taylor’s approach and evidence. Page 203 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 204 of 257 Submissions

¶ [703] The attorneys (generally admitted and those benefiting from limited admission) showed great skill and commitment but again did not make my task an easy one. Their written submissions ran to many hundreds of pages (with the Company’s closing written submissions running to 277 pages with 1148 footnotes, the CCCW Dissenters 271 pages with 583 footnotes and the Carey Olsen Dissenters a mere 134 pages with 503 footnotes), their oral submissions occupied 4 days of scarce court time. In Delaware appraisal hearings appear often to be limited to 4 days in total with live evidence but I note the reference to the procedure of taking many depositions in advance.

¶ [704] I have to say that the closing oral submissions were delivered at great speed and I was not given the luxury of time to reflect upon them during their delivery. To be frank, I gained the impression that at certain stages the attorneys rather than seeking to assist or persuade me were simply getting submissions on the record to assist them in any subsequent appeal. The speed with which some of the closing oral submissions were presented to the court brought to mind the comments of Judge Pelling KC sitting as a judge of the English High Court in The Libyan Investment Authority v Credit Suisse International

¶ [2021] EWHC 2684 (Comm) at [139]-[140] but in the case before me the attorneys were allocated sufficient time for their oral closing submissions and the length of the hearing was more than adequate.

¶ [705] In addition to considering the voluminous oral and written opening and closing submissions I should record that I also considered the written submissions of the parties filed on 31 October 2025 in respect of Trina Solar [2025] UKPC 48 where judgment had been delivered by the JCPC on 30 September 2025, after the date of the conclusion of the hearing in this case but before I had finalised my judgment. Lists of Issues

¶ [706] Sadly and unhelpfully the parties could not even agree on the list of issues for determination by the court and so it was that on 28 August 2025 I received separate competing lists of issues from the attorneys acting for the Company and the Dissenters. Page 204 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 205 of 257 The Company’s list of issues

¶ [707] The Company’s list of issues (which it stated was “not exhaustive” and it “may not be necessary for the Court to determine each of the issues identified, depending on the approach which is adopted”) included the following: Overarching issues

¶ [708] As a matter of the statutory interpretation of Section 238, is there a hierarchy of valuation methodologies?

¶ [709] For each valuation methodology, in the circumstances of this case: (a) does giving 100% weight to the methodology produce a better estimate of fair value than placing weight on any other methodology or methodologies? And (b) if not, what weight, if any, should be placed on the methodology and why would placing that weight on the methodology improve the estimate of fair value compared to the alternative weightings (including zero)? Adjusted Market Trading Price Methodology

¶ [710] Has the Company established that using an AMTP to estimate fair value as Professor Lehn contends is possible in the circumstances of this case? If so, is it more reliable than any other available methodologies?

¶ [711] Was the market price on 11 January 2022 at or above fair value on that date, given the original merger agreement, the November 2022 Press Release and the Ocean Link Projections?

¶ [712] Was the market for the Company’s shares semi-strong form efficient on 11 January 2022, in particular with regard to: (a) the experts’ event studies; (b) the availability and dissemination of information about the Company; (c) analyst coverage; (d) volatility testing; (e) trading volumes and increases in those volumes following news disclosure; and (f) other Cammer factors considered by Professor Lehn? Page 205 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 206 of 257

¶ [713] If there was any MNPI, was it value positive or value negative and can it be reliably quantified? Does the existence of any MNPI preclude reliance on the AMTP?

¶ [714] What, if any, method is the best way of rolling forward the share price on 11 January 2022 to the Valuation Date and what, if any cross-checks, support this? In particular: (a) were the Golden Dragon Index and/or the industries index meaningfully comparable indices correlated with the Company’s shares? (b) did the volatility of the Golden Dragon Index during the roll-forward period cause the roll- forward adjustment to be inappropriate or unreliable? (c) to the extent that any existed, did Professor Lehn’s roll-forward model fail to account for excess co-movement, and if so does this render his model unreliable? (d) does the evidence about the direction and magnitude of the movement in the wider market provide a cross-check and/or alternative to Professor Lehn’s roll forward model?

¶ [715] Was there any value-relevant information which came into existence during the roll-forward period which would render the rolled-forward figure less reliable than the figure estimated via any other valuation methodology?

¶ [716] What is the significance (if any) of the “break prices” modelled by the Dissenters and/or third parties? Do they serve as a relevant cross-check to the AMTP on the Valuation Date? Discounted Cash Flow methodology

¶ [717] Have the Dissenters established that a DCF analysis is possible in the circumstances of this case? If so, is it more reliable than any other available methodologies?

¶ [718] The Company raises detailed sub-issues in respect of the DCF methodology under the following sub- headings: Projections (paragraphs 10-12) Extension of explicit forecast period to steady stage (paragraph 13) Terminal Growth Rate (paragraphs 14-16) WACC (paragraph 17) Regarding beta (paragraph 18) Withholding tax (paragraph 19) Page 206 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 207 of 257 Tax rate (paragraph 20) Cash and non-operating assets (paragraphs 21-23) Minority discount (paragraph 24) Merger Price

¶ [719] Can the Company rely on the merger price methodology as a cross-check?

¶ [720] Is there any reason to believe that any material which was not disclosed would have indicated material failings in the merger process?

¶ [721] Do any of the following prove that fair value is at or below the merger price?: (a) the parties’ stated reasons for the merger; (b) the fact of the Merger Agreement; (c) any acquisition premium embedded in the merger price; (d) the change in the Company’s prospects and/or the estimated market price for the Company’s shares but for the merger between the date the merger agreement was signed and the Valuation Date; (e) the vote of shareholders unaffiliated to the Buyer Consortium at the EGM; (f) the vote of shareholders unaffiliated to the Buyer Consortium or the Dissenters at the EGM; (h) the fact that many long term shareholders sold their shares to arbitrageurs below the merger price after the merger announced; (i) the absence of topping bids; (j) the fact that the Special Committee, the Board, Duff & Phelps, Glass Lewis and ISS each endorsed the merger; (k) the fact that a majority of the Company’s vice-presidents declined Mr Yan’s invitation to invest in the merger?; and (l) is it reasonable to assume that merger arbitrageurs invest in a firm with no regard for the fundamental value of the target? Comparable Companies Analysis

¶ [722] Do the comparable companies analysis provide a relevant cross-check on Professor Lehn’s or Mr Taylor’s fair value estimates? Is Mr Taylor’s assertion that the fact that his DCF conclusion is at the top end of the range of his CCA multiples reasonable given the disparity in the growth rates as presented in his report? Page 207 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 208 of 257 The Dissenters’ list of issues

¶ [723] The Dissenters’ list of issues (which it is stated “is not intended to be exhaustive, but instead to give a broad indication of the issues which arise for consideration following trial. Further, it may not be necessary for the Court to determine each of the issues identified, depending on the approach which is adopted”) included the following: Adjusted Market Trading Price Approach

¶ [724] Has the Company established that using an adjusted trading price as Professor Lehn contends is possible and/or appropriate in the circumstances of this case? In particular has the Company established that the market traded price on 11 January 2022 was a reliable indicator of, or was at a premium to, fair value?

¶ [725] Given that it is common ground that the market traded price on 11 January 2022 was affected by the Original Offer on 17 September 2020 and subsequent merger process, has the Company established that it is appropriate to assume that the market traded price on 11 January 2022 reflected, or was at a premium to, fair value notwithstanding the effect (e.g. any changes to the nature of the shareholders of the Company) of the Original Offer and subsequent merger process?

¶ [726] Has the Company in any event established that the market for the shares was semi-strong form efficient on 11 January 2022, having regard to: (a) the expert’s event studies; and (b) other matters, including the availability and dissemination of information about the Company, and analyst coverage, by that time?

¶ [727] Has the Company in any event established that the 11 January 2022 market traded price is reliable as an indicator of fair value notwithstanding: (a) the November 2022 Press Release (which, it is common ground, was not concerned with the fundamental value of the Company) and its effect on the market traded price; and/or (b) the existence of the Ocean Link Projections (which were not released to the market)?

¶ [728] In all the circumstances, has the Company established that the market traded price on 11 January 2022 represents an appropriate and fair starting point for an AMTP analysis? Page 208 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 209 of 257

¶ [729] Has the Company in addition established that Professor Lehn’s roll-forward adjustment produces a figure at the Valuation Date which is a reliable indicator of fair value?

¶ [730] Were the Golden Dragon Index and/or the industry index meaningfully comparable indices correlated with the Company’s shares and, if not, did the use of those indices in the regression model cause the roll-forward adjustment to be unreliable?

¶ [731] Did the volatility of the Golden Dragon Index during the roll-forward period cause the roll-forward adjustment to be inappropriate or unfair for the same reasons as the roll-forward-adjustment in FGL.

¶ [732] Were there structural flaws in Professor Lehn’s roll-forward model (as identified by Mr Taylor and referred to in the Dissenters’ submission) which caused the roll-forward adjustment to be unreliable?

¶ [733] Should Professor Lehn’s roll-forward model have been rejected as unreliable for failing to account for excess co-movement?

¶ [734] Has the Company established that there was no value-relevant information (public or non-public) which came into existence during the roll-forward period (i.e. between 11 January 2022 and the Valuation Date) the existence of which would render the rolled-forward figure unreliable, including in particular: (i) the 2022 Management Projections and/or (ii) the internal 2022 budgets?

¶ [735] In all the circumstances, has the Company established that the roll-forward adjustment performed by Professor Lehn produces a reliable estimate of fair value as at the Valuation Date? Discounted Cash Flow Approach

¶ [736] Have the Dissenters established that a DCF analysis is possible and/or appropriate on the facts of this case?

¶ [737] The Dissenters raise detailed sub-issues in respect of the DCF approach under the following sub- headings: Projections (paragraphs 13-16) Page 209 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 210 of 257 Extension of explicit forecast period to steady state (paragraph 17) Terminal growth rate (paragraphs 18-19) WACC (paragraph 20) Withholding tax (paragraph 21) Income tax rate (paragraph 22) Cash and non-operating assets (paragraph 23) Minority discount (paragraph 24) Merger Price

¶ [738] Can the Company rely on the merger price as a cross-check? In particular: (1) Did the Company adduce the necessary evidence to satisfy the conditions justifying reliance on the merger price (summarised at Trina Solar CICA at [112], [137] to

¶ [141] and iKang at [35]) ? (2) Have there been any failings of discovery by the Company and, if so, can the Company rely on the merger price as a cross-check notwithstanding such failures? (3) Does the EGM vote enable or assist the Company’s reliance on the merger price as a cross-check? Additional issues submitted by the Carey Olsen Dissenters

¶ [739] The Carey Olsen Dissenters specified the following additional issues. Adjusted Market Trading Approach

¶ [740] By 11 January 2022 is the Court’s ability to rely on the market price in any way affected by the character of the Buyer Group as a buyout by a controlling shareholder group which included management who had agreed to vote as a group? Should the Court follow the reasoning in HBK v Pivotal? Page 210 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 211 of 257 Merger Price

¶ [741] What is the merger or deal price in this case when there were two merger agreements both of which were valid and binding in Q1 of 2022, the subject of Special Committee consideration, fairness opinions, deliberation and recommendation? If the Court is minded to place weight in the merger price, does the price in the Original Merger Agreement satisfy more of the conditions justifying reliance on it and constitute more reliable evidence of fair value than the price in the Revised Merger Agreement? Determination

¶ [742] Having dealt with the relevant law and the evidence in some detail I can take the determination section of this judgment relatively briefly, conscious that I must deal with the main disputed issues and provide adequate reasons for my decisions on the relevant material disputed issues before the court for determination. The hierarchy point

¶ [743] On behalf of the Company it was boldly submitted (I think for the first time in a Cayman Islands court) that the rule of law, certainty and a proper construction of the statutory provisions lead to the conclusion that in Cayman Islands law there was a hierarchy of valuation methodologies and “at the top of the hierarchy of valuation methodologies sits the Market Price Methodology followed by the Merger Price Methodology” and the DCF Methodology “is vastly inferior to the preferred methodologies (where they are available) for clear economic and public policy reasons.” Even after the judgment of the JCPC in Trina Solar the Company did not abandon this “hierarchy” point. At paragraph 1 of its written submissions dated 31 October 2025 it maintained “its position as stated in its previous submissions and does not intend to waive any of its arguments as expressed during the trial.” I should therefore refer to some of the relevant law on this new “hierarchy” point and reach a determination upon it.

¶ [744] At [15] of Trina Solar Sir Andrew Popplewell in the JCPC stated “… there is no hierarchy between [the various valuation measures] … Each methodology must be assessed individually to identify strengths and weaknesses which may affect its reliability as a guide to fair value. Reliability is not in Page 211 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 212 of 257 the context a binary concept in which the court must conclude that the measure in question is or is not reliable; rather it is a qualitative concept in which the court may conclude that it is more or less reliable on a sliding scale.”

¶ [745] Sir Andrew Popplewell at [16] added: “Moreover, the exercise is not one of simply the reliability of each methodology individually, but one of assessing comparative reliability between all of them … The court may be faced with two valuation measures, neither of which it would consider to be more reliable than not as a guide to fair value if taken individually. Nevertheless it is entitled to ascribe more than a 50% weighting to one if it considers it less unreliable than the other. Assessing reliability is not only a qualitative exercise in relation to each methodology individually, but the court is concerned with a qualitative assessment of the reliability of the methodologies relative to each other.”

¶ [746] Segal J in iKang at [37] stated: “There is no hierarchy with respect to the valuation approaches which may be applied to determine fair value.”

¶ [747] Furthermore the word “hierarchy” did not appear in the judgment of Birt JA in Trina Solar.

¶ [748] Mr Boulton made reference to certain Delaware authorities. In fairness to him he took care to set them in the context of Lady Arden’s comments in Shanda as to the relevance of Delaware jurisprudence in Cayman Islands law. Suffice to say, I accept Mr Adkin’s submission to the effect that the hierarchy position is clear under local Cayman Islands law (there is no hierarchy) and the Delaware judgments on this point may be of interest but little more than that.

¶ [749] Whether by way of statutory interpretation or consideration of previous case law, I am not persuaded that there is a strict formal hierarchy of valuation methodologies under the laws of the Cayman Islands. Page 212 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 213 of 257 Rejection of the DCF analysis in this case

¶ [750] In my judgment, in the particular circumstances of this case I have concluded that no weight whatsoever can be placed on Mr Taylor’s DCF analysis. It is fundamentally flawed and totally unreliable for the reasons essentially put forward by Professor Lehn and the Company (including optimistic projections, staleness and steady state issues). I have considered all the evidence before the court in respect of the projections and the proposed adjustments but am unable to come to a reliable conclusion as to an appropriate forecast. I have also descended into the other DCF factors which are disputed between the parties (such as Terminal Growth Rate, WACC, beta, withholding tax, tax rate, cash and non-operating assets and minority discount) but try as I might I have been unable to come to a conclusion that this court, on the basis of the evidence before it, could produce its own DCF analysis upon which any weight could properly be attached.

¶ [751] I take fully on board the JCPC’s wise advice that reliability in this context is not a binary concept. It is a qualitative concept in which the court may conclude that a methodology is more or less reliable on a sliding scale. The problem in this case is that on a sliding scale of 0-10 I place Mr Taylor’s DCF analysis at a 0 and try as I might I cannot confidently budge it any further up the scale. The court is unable on the evidence placed before it to conduct an appropriate DCF analysis. The DCF’s relative reliability, in this case, when compared with the AMTP is zero.

¶ [752] I accept that I should not treat reliability as a concept in which the supporting factors and undermining factors fall to be balanced so as to produce an all or nothing result, such that if the weaknesses are more potent than the strengths, no weight can be attached. I note the JCPC’s advice that reliability is not a binary concept and that the court should not ignore the fact that weight may be attached to a particular valuation measure even where the court concludes that the uncertainties or weaknesses make it less rather than more reliable.

¶ [753] The fact remains that in this case, having undertaken the required comparative reliability exercise, I have concluded that no weight can properly be accorded to the DCF analysis and 100% weight can and should be accorded to the AMTP analysis.

¶ [754] Although I take on board the CICA’s guidance in Trina Solar to the effect that mere subjectivity should not prevent reliance on a robust DCF and that trial judges should not simply plump for one expert over Page 213 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 214 of 257 another I also note the guidance from Delaware (for example in Dell, Bancorp and Stillwater) that although not required to do so, it is entirely proper for the first instance court to adopt any one expert’s methodology in toto if that valuation is supported by credible evidence and withstands a critical judicial analysis on the record.

¶ [755] For reasons which follow I have concluded that Professor Lehn’s AMTP methodology is reliable and supported by credible evidence. It withstands a critical judicial analysis on the record. Having considered the position independently and taking into account all the evidence and submissions before the court I adopt it in toto together with the fair value figure of $31.11 per share. I do not seek to tinker with it for the sake of tinkering and so as not to be perceived as plumping for one expert over another contrary to the guidance of the CICA in Trina Solar. In my judgment as a matter of law I am entitled, if I so conclude, to agree with and adopt Professor Lehn’s market-based approach and his figure of $31.11 per share as fair value. I so conclude for the reasons outlined in this judgment. The methodologies

¶ [756] As Birt JA stated at [35 (iv)] in his judgment in Trina Solar (CICA) the selection of a valuation methodology or valuation methodologies is “a fact sensitive issue”. As Sir Andrew Popplewell emphasised in Trina Solar (JCPC) “All merger transactions are factually unique and the relevant circumstances will often differ very substantially from case to case” ([15]) and “All will depend on the facts of the individual case, including the relative reliability of other methodologies contended for” ([18]).

¶ [757] In this case Professor Lehn relies on the AMTP approach and conducts no DCF analysis of his own. Mr Taylor relies on a DCF approach. Professor Lehn in effect says that the court can obtain no assistance whatsoever from a DCF approach in this case. Neither the experts nor the attorneys for the parties suggested that the court should adopt a blended approach. They all advocated in effect an all or nothing approach. The Company suggesting the court adopts the AMTP approach adopted by Professor Lehn. The Dissenters suggesting the court should adopt Mr Taylor’s DCF approach, with if necessary any appropriate adjustments the court sees fit to make.

¶ [758] For Professor Lehn’s AMTP approach to get off the ground he relies to an extent on the trading price, the merger price and, his opinion that an acquisition premium would have been in existence. He also Page 214 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 215 of 257 rolls forward the trading price of US$45.83 on 11 January 2022 to the Valuation Date (26 April 2022). Although when questioned, Mr Adkin accepted that with the AMTP approach there was less scope for adjustments than with a DCF approach he did not suggest any specific adjustments to the AMTP. The Dissenters’ main case was that the AMTP was simply unreliable and should not be used. Merger price

¶ [759] The first issue I deal with is therefore whether any reliance can be placed on the merger price. I have set out the law in some detail above.

¶ [760] During his oral closing submissions I asked Mr Boulton to summarise “the relevance of the merger price on his main case on the AMTP.” Mr Boulton responded: “… the merger price gives you an upper band but [Professor Lehn’s] conclusion is that it overstates fair value … my Lord can be confident or can conclude that the merger process was robust. That it must overstate fair value … twofold reasons … first of all, the embedding of that acquisition premium to which the dissenters are not entitled and, secondly, because there is a decline in value between the date of the Merger Agreement, 1 March 2022, and the valuation date … on 12 January … the market price was $45.83, that’s the figure that Professor Lehn rolls forward … The final offer ended up at $61, so that’s a 15 to $16 premium over the original market price. Now, of course a couple of months have gone by in the intervening period but, given that the market generally is falling, and one can certainly argue that the acquisition premium embedded in the $61 price is at least $15, and so we say one would have to reverse that out because it’s something to which the dissenters are not entitled because they have dissented from the transaction.” (Day 19/30:13-25; 31:1-25; 32:1-21)

¶ [761] In very short summary the court must consider whether the merger process in this case was an “open process” characterised by “objective indicia of reliability”. I consider these matters on the basis of the evidence provided in this case. Page 215 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 216 of 257 Robust public information

¶ [762] In my judgment there was robust public information which comprised of a share price with many shareholders and active trading. From time to time the stock was regularly followed by analysts and others. The Cammer factors

¶ [763] The Cammer factors are intended to be an analytical tool and not a checklist. This is a similar point to the point made by Sir Andrew Popplewell in Trina Solar at [18]. The Cammer factors however provide useful evidence from which market efficiency may be inferred and are therefore relevant to a court’s determination of whether a market is efficient.

¶ [764] The Cammer factors include (a) the existence of an actively traded market, as evidenced by a large weekly volume of stock trades; (b) the number of securities analysts that followed and report on the stock; (c) the number of market makers; (d) the company’s eligibility to file a Form S-3 Registration Statement; and (e) a cause-and-effect relationship, over time, between “unexpected corporate events or financial releases and an immediate response in the stock price”. Professor Lehn says that commentators generally regard these factors (apart from (b) and (e)) as indicators of whether the market is “open and developed” not indicators of whether the market is efficient. Professor Lehn considered all five Cammer factors. Actively traded market

¶ [765] The ADSs were listed on the NASDAQ Global Select Market which Professor Lehn says is the tier of the NASDAQ Stock Market with the highest listing standards. Professor Lehn says that Exhibit D-3 shows the average weekly trading volume during the period of 17 September 2015 to 1 March 2022 was 3.7% of the ADSs outstanding which substantially exceeds the 2% level that an authority cited favourably by the Cammer court stated “would justify a strong presumption that the market for a security is an efficient one.” Professor Lehn says Exhibit D-4 shows the average weekly volume was 1,075,614 ADSs during this period and Exhibit D-5 shows the average weekly dollar volume of trading of the ADSs during this period was approximately $70.8 million. The data shows that there was “significant investor interest” in the Company which “in turn implies a likelihood that many investors Page 216 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 217 of 257 [were] executing trades on the basis of newly available or disseminated corporate information” (Cammer at 1286).

¶ [766] Professor Lehn says that Exhibit D-6 shows that most outstanding ADSs were held by institutional investors (i.e., mutual funds, money managers and banks). Institutional investors are presumed to be well-informed, professional investors. The assumption is that market professionals predict earnings more accurately than other investors. Furthermore, Professor Lehn highlights that at least 75 different institutional investors held, bought or sold ADSs during each quarter. Analyst coverage

¶ [767] The Company in its written closing submissions in a section entitled “Decline in analyst coverage” beginning at paragraph 236 on page 99 accepts that there were fewer public-facing analysts covering the Company but adds that the chart from Mr Taylor’s supplemental report which was put to Ms Chien and Professor Lehn in cross-examination was misleading as it cuts off before the public analyst reports issued in 2022, being Tavira’s on 11 January 2022 and the Morgan Stanley reports on 1 March 2022 and 4 April 2022. The Company adds that it is also misleading because it self-servingly excludes reports that Mr Taylor considered were not relevant to efficiency because he wrongly claimed that they do not relate to Company-value issues (focusing on merger issues). The Company says that this is demonstrably wrong and cites the example of both Citi and Morgan Stanley published reports on 17 September 2020 which contained some Company-performance related information. Mr Taylor also failed to take into account reports by other companies which contained information about the Company, nor any industry reports.

¶ [768] The Company is right to remind the court as Professor Lehn pointed out, that while the number of analysts covering a company is considered a relevant factor to efficiency under the Cammer factors, some of the literature which Mr Taylor cites proposes a presumption of efficiency when a company is covered by simply one analyst (with a stronger presumption at two or more). Morgan Stanley continued to issue reports into April 2022 and Tavira had issued a report in January 2022.

¶ [769] Professor Lehn at Exhibit D-7 produces data to support his comment that analysts from at least nine different brokerage firms covered the Company and these analysts issued at least 103 reports during the period September 2015 to March 2022. Page 217 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 218 of 257

¶ [770] The Company says that an absence of analyst reports does not mean that analysts are not continuing to cover a company. It is Professor Lehn’s view, based on his experience and expertise, that analysts will tend to publish a report when there is more value relevant information to publish and publish less when there is less to report on: “… you can’t say that the absence of reports means the absence of coverage. We know factually Morgan Stanley issued two reports in the spring of 2022: one in March, one in April.” (Day 7/194: 11-14)

¶ [771] I note also that Citi by email notification dated 28 June 2021 stated that they were discontinuing coverage of the Company as the Company was “undergoing a planned privatization and delisting is expected to close during 2021.” Professor Lehn said that this was “not inconsistent with what I’ve said” (Day 7/195:10). Professor Lehn pointed out that Citi were not saying that the Company “is no longer trading on fundamentals.” (Day 7/196: 6-7)

¶ [772] The Company says that documents from the Dissenters’ disclosure show that Eddie Wang, the analyst from Morgan Stanley, continued to closely monitor the Company’s business and industry between Morgan Stanley’s last public report in July 2021 and the next in March 2022, and certain dissenters relied heavily on him for insight into the Company and the merger.

¶ [773] It appears from the evidence that Citi ceased coverage because of the take-private and the anticipated de-listing but the evidence also indicates that other analysts continued to cover the Company.

¶ [774] I also note that buy-side investors had their own analysts (for example Churchill Capital’s 17 November 2021 deal update and Manalo monthly update April 2022). Market makers

¶ [775] Professor Lehn at Exhibit D-8 produces data to support his comment that 102 different market makers traded the ADSs during the period of September 2015 to March 2022. Page 218 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 219 of 257 Registration statements

¶ [776] The Company filed a registration statement on Form F-1 which is a full registration statement issued by foreign private issuers when they first want to sell securities on a US stock exchange. The evidence indicates that the Form S-3 for domestic US companies is shorter. Cause and effect

¶ [777] Professor Lehn refers to ample evidence of a cause-and-effect relationship between information disclosure and price movements.

¶ [778] Mr Boulton in his oral closing submissions stressed that Professor Lehn reaches his conclusion on semi- strong efficiency by reference to his event study but also considers other indicia of efficiency based on the Cammer factors which Mr Boulton says are all supportive. They are not necessary conditions but they are helpful. Mr Boulton says that the only one of these that received any real challenge was the analyst coverage. I am satisfied that the Cammer factors have been satisfied in this case. “Controller” theory

¶ [779] The Carey Olsen Dissenters in the Dissenters list of issues, filed on 28 August 2025, raise an issue as to whether the court’s ability to rely on the market price is in any way affected by the character of the Buyer Group as a buyout by a controlling shareholder group which included management who had agreed to vote as a group and whether the court should follow the reasoning in HBK v Pivotal.

¶ [780] I have considered HBK v Pivotal, the other authorities and the limited submissions that were put before the court on the “controller” point. I note in particular the way in which this point was developed in the Carey Olsen Dissenters’ written closing submissions. I note the references to HBK v Pivotal from paragraph 74 onwards and the submissions in section M on the effect of control on reliability of merger value. I note the submission at paragraph 124 that “Delaware courts are especially cautious in controller scenarios like HBK v Privotal (sic)”.

¶ [781] I have to say that I was not addressed in any meaningful way as to the concept of a controlling shareholder as a matter of Cayman Islands law. If you construe that as having 75% or more of the Page 219 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 220 of 257 shares or just 50% either way Mr Yan would not fall within that definition. He was not a “controller”, either by himself or with Recruit. I accept that any potential buyers would want Mr Yan on board but he alone did not have “control”. Mr Yan’s evidence was to the effect that his support for a merger did not automatically mean that he would lead or even participate in a transaction. The Company is right at paragraph 237.2 of its written closing submissions to refer to Mr Yan’s evidence that he was not averse to retiring depending on the circumstances. I also agree that there is no basis on the evidence to join Mr Yan and Recruit together. The evidence indicates that there were frictions between Mr Yan and Recruit. Mr Yan could not rely on the cooperation of Recruit. Recruit did not agree to support the Original Merger Proposal until 21 June 2021 and, like Ms Chien, was never a member of the Buyer Consortium. I do not regard the presence of Mr Yan or Recruit and the Support Agreements as preventing the court from relying on the merger price. Access to non-public information and no discrimination between buyers

¶ [782] In my judgment there was access to non-public information by the buyer and no discrimination between potential buyers. There were no competing bids in this case and there is nothing to suggest that if there had been there would not be access to non-public information or that there would have been discrimination between potential buyers. In this case this is not a live point. A chance to bid and a robust market check

¶ [783] In my judgment others had a chance to bid. There was an outreach to other potential bidders and there were adequate “go-shops”. For a well-resourced and determined bidder there was a realistic possibility of a topping bid. Furthermore, the market was well aware that the Company was “in play”. The merger process in this case was not conducted in a way which prevented or discouraged a higher offer from a rival bidder emerging. There was a realistic and fair opportunity for alternative bidders. Mr Yan was not out to acquire the Company “on the cheap”.

¶ [784] One objective criterion of reliability of the merger price is whether there was a robust market check with “outreach to all logical buyers” and go-shop characterised by “low barriers to entry” such that there is a realistic possibility of a topping bid. In my judgment there was an adequate market check. Page 220 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 221 of 257

¶ [785] The Company says that where alternative transactions and/or topping bids are possible (and the Company adds that they were in this case) then the fact that none were received during a suitable merger process should be taken into account by the court. The Company says that the absence of any topping bid or alternative transaction supports the proposition that the true value of the Company was less than the merger price.

¶ [786] The Company says that the Special Committee could have entertained competing bids at any stage prior to the EGM. The Company adds that there were several periods during which the circumstances were particularly opportune for a competing bid: (1) between 17 September 2020 (the date when the DCP offer was announced) and 4 May 2021 (the day on which the formation of the Buyer Consortium was announced) when it was clear to market participants that the Company was in play and the only competition was an external equity fund. In reality there were no barriers preventing other parties from proposing an alternative transaction or putting forward a similar proposal with a higher bid during this period; (2) between 4 May 2021 and 21 June 2021 (the date on which the Original Merger Agreement was executed) when Recruit had no legal obligation to support the Buyer Consortium bid; (3) between 8 November 2021 (the date on which the Company announced that certain members of the Buyer Consortium had been consulting with PRC regulators) and 12 January 2022 (the date on which the revised offer was announced) when against the backdrop of the PRC regulatory consultation and with the Original Merger Agreement’s long-stop date of 21 March 2022 fast approaching, it was clear that neither Recruit nor Mr Yan were involved in the regulatory consultation and were instead hoping to push ahead with the Original Merger Agreement; and (4) between 12 January 2022 and 1 March 2022 (the date on which the Merger Agreement was executed) market participants knew that the market value of the Company was falling and that the Buyer Consortium was seeking to renegotiate the terms of the Original Merger Agreement, thus presenting clear opportunities for a competing higher bid. Page 221 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 222 of 257

¶ [787] The Company points out that at no point in the above timeline were there any structural impediments to competing bids. The merger agreements conferred upon the Special Committee the ability to engage with other prospective bidders if they thought that such discussions were likely to elicit a “Superior Proposal” (as defined in section 9.03 of the Original Merger Agreement) and prospective parties could have utilised the go-shop provisions to displace the proposals by the Buyer Consortium.

¶ [788] The Company also submits that there was no reason to suppose that a competing bid could not have come in on 1 March 2022 or even after.

¶ [789] The Company submits that the inclusion of “match rights” was not a deterrent to competing bidders.

¶ [790] The Company refers to the Dissenters’ argument that the prospect of any competing bid was foreclosed entirely from the outset, because it was obvious to market participants that Mr Yan supported DCP’s offer and because no transaction could have proceeded without Mr Yan’s support and in rebuttal refer to the evidence that Mr Yan did not support DCP’s offer from the outset.

¶ [791] In respect of the Dissenters’ argument that the Support Agreements precluded a competing bid arising the Company repeats the point that the Support Agreements would have terminated had the Company elected to terminate the Original Merger Agreement in connection with a Superior Proposal.

¶ [792] The Company submits that the Special Committee undertook a market check in the form of: (1) a pre-signing market check (i.e. the early announcement of the DCP proposal well before the execution of any definitive merger agreement, signaling that the Company was “in play” or “on the market”); and (2) a post-signing market check (i.e. the two go-shop processes, neither of which elicited a competing bidder or indeed any expressions of interest).

¶ [793] The Company, with some considerable force, stresses that the 9-month period between the September 2020 announcement of DCP’s offer and the June 2021 Original Merger Agreement meant that anyone seriously interested in the Company had the opportunity to prepare a competing bid. Page 222 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 223 of 257

¶ [794] The Company refers to the Dissenters’ criticism of the Special Committee’s alleged failures to explore unsolicited expressions of interest and to conduct effective go-shop processes.

¶ [795] In respect of the unsolicited expressions of interest the Company refers to early approaches from and discussions with various entities (such as Blackstone, Hel Ved and Sequoia) and evidence (email 25 September 2020 from Tomomasa Matsumoto of Indeed/Recruit) that suggests DCP received over 100 inquiries to potentially co-invest along with DCP in the Company. The Company says that with the exception of Blackstone’s presentation, provided in response to a request from the Company (which was framed as “illustrative” and “for discussion purposes only” and which pre-dated DCP’s offer) the early outreach, unsurprisingly, was exclusively premised on prospective investors joining a buyout group.

¶ [796] The Company refers to Mr Taylor making a criticism that he had seen “only limited evidence” that these early approaches were not followed up on. Mr Taylor at 5.4.21 of his first report in footnote 337 does recognise that Mr Yan says he spoke with Sequoia following its expression of interest (paragraph 28 of Mr Yan’s affirmation from which it appears Sequoia were also interested in BOSS and were concerned over a potential conflict).

¶ [797] The Company at paragraph 426 of its closing written submissions submits that “the contemporaneous evidence does not support Mr Taylor’s sweeping statement that the Company did not follow up expressions of interest”. Mr Taylor however at 5.4.21 does not make that “sweeping statement”. Mr Taylor simply says that he has “seen only limited evidence that these approaches were followed up on”. Be that as it may, the Company submits that the key point to be taken from the approaches is that prospective investors were willing to contact the Company and express their interest without an express invitation to do so.

¶ [798] The Company also emphasises that the Special Committee ran two go-shop processes (the first, in connection with the Original Merger Agreement, lasting 30 days and involving outreach to 12 potential investors and the second, in connection with the Amended Merger Agreement, lasting 15 days and involving outreach to 11 potential investors). Neither process elicited any bids. The Company submits that the test of an effective go-shop is not whether it elicits any bids. The Company submits that the go-shops in this case “enabled the Special Committee to satisfy itself that there were no competing bidders out there” (paragraph 431 of the Company’s written closing submissions). Mr Taylor accepted Page 223 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 224 of 257 that it was “possibly fair” to suggest that the test of a successful go-shop is not only whether it produces a high bid but also whether it allows the Special Committee to satisfy itself that there is not somebody out there who would be willing to pay more (Day 16/193; 21-25 and 194;1).

¶ [799] The Dissenters submit that the evidential burden is on a company contending that the merger price is indicative of fair value to show that the conditions justifying reliance on the merger price are satisfied and in particular, sufficient evidence of market efficiency, fair play, low barriers to entry, outreach to all logical buyers and a well-designed sales process must be adduced. The Dissenters say that the critical question is whether there was ever a realistic and fair opportunity for alternative bidders (paragraph 564 of other CCCW Dissenters’ written closing submissions). The Dissenters emphasise that in the context of a management buyout there must be heightened scrutiny. The Dissenters submit that there was never any real prospect of an alternative bidder on the facts of this case and as such no reliance can be placed on the merger price as being an indication of fair value (whether as proxy for or a cap on fair value). The Dissenters stress the following points: (1) there was a strong disincentive for any competing bids as there was a high likelihood that the merger would complete as a result of Mr Yan’s and Recruit’s “controlling stake”. The Dissenters say that the evidence was that given the nature of PRC companies no other rational bidder would even have considered competing against Mr Yan. (2) there was a failure to explore expressions of interest; (3) the go-shop process was conducted in a defective manner; and (4) there was a lack of topping bids.

¶ [800] I do not accept the submissions put forward by the Dissenters in respect of reliance on the merger price.

¶ [801] The evidence and submissions presented by the Company have persuaded me that the merger price can be relied upon by the Company for the purpose upon which the Company relies upon it. Page 224 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 225 of 257 The Special Committee

¶ [802] There was a special committee comprised of independent and experienced directors. Even Mr Taylor accepted that there was an independent Special Committee. They were advised by top rate competent legal and financial advisers. Involvement of Mr Hsu

¶ [803] I accept that the involvement of Mr Hsu was not ideal but it was not such as to persuade me that there was a lack of independent decision making process within the Special Committee or that Mr Hsu’s involvement was such that the Company should be prevented from placing reliance on the merger price for the purposes on which it relies upon it.

¶ [804] The Dissenters submitted that it was wholly inappropriate for the Special Committee to allow Mr Hsu/RM to play an integral part in the merger process. I have considered Mr Cheng’s evidence as to the involvement of Mr Hsu and his company and in light of that evidence and the position generally do not conclude that the Dissenters’ criticisms should lead this court to determine that such involvement prevents the Company from relying on the merger price. Conflicts and “heightened scrutiny”

¶ [805] There were plainly people on the management side and within the Buyer Consortium who had an interest in the merger proceeding.

¶ [806] I have no hesitation in concluding however that the merger was not driven by Mr Yan wanting to “acquire the company on the cheap”. I accept that Mr Yan gave honest and believable evidence in respect of his motivations for allowing the merger to progress. I accept the Company’s submission that Mr Yan was the “anti-thesis of a scheming, hard-driving corporate raider.” I accept the Company’s submission that Mr Yan was not focused on his personal economic interests but the survival of “his” company which he took great pride in (See for example Day 2/43:12-21). Mr Yan was plainly motivated by factors besides money. Page 225 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 226 of 257

¶ [807] Furthermore, although Ms Chien had loyalty towards Mr Yan I have concluded that she was an honest and competent individual who did nothing improper to facilitate the progression of the merger. Mr Yan was part of the Buyer Consortium and although Mr Taylor littered his reports with references to Ms Chien being part of the Buyer Consortium the evidence is plainly to the contrary. Ms Chien although she was a party to the Support Agreements was never a party to the Buyer Consortium and it was inaccurate and wrong of Mr Taylor to repeatedly describe her as such.

¶ [808] I note the Dissenters’ submissions in respect of the financial interest of Ms Chien in the merger but balance those out taking into account the Company’s submissions.

¶ [809] I accept that the involvement of Mr Yan and Ms Chien (even though she was not part of the Buyer Consortium) as a matter of good sense and caution, requires that the court applies “heightened scrutiny” to the merger process in this case. In my judgment having applied that “heightened scrutiny” the merger process although not totally flawless (and I doubt a flawless merger process has ever existed or will ever exist in reality) was adequate enough to permit the reliance Professor Lehn places upon it.

¶ [810] As Vice-Chancellor Slights stated in PetSmart Inc (Del. Ch. May 26, 2017) at page 82 (footnotes omitted): “… perfection in human endeavors is not attainable. Even in the best case, a process to facilitate the sale of a company, constructed as it must be by the humans that manage the company and their human advisors, will not be perfect.” Conclusion on robust market check and merger process

¶ [811] Having considered the evidence, the arguments and the law I have concluded that there was an adequate market check in this case.

¶ [812] The merger process was not perfect but it was characterised by many objective indicia of reliability. The minor imperfections or deficiencies in this case included: Page 226 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 227 of 257 (1) the involvement of Mr Hsu but such involvement had pluses (identified by Mr Cheng) as well as minuses (identified by the Dissenters); (2) the Special Committee were unsuccessful in obtaining an inclusion of a MoM clause but Mr Cheng adequately explained that; (3) the Special Committee could have been more aggressive in maintaining reliance on the original merger deal (but I note Mr Cheng’s evidence of the discussions with Davis Polk in this respect); (4) the Special Committee could have been more aggressive in negotiating a deal at a higher price but I note Mr Cheng’s evidence to the effect that a deal at $60 or above was a good deal and it was in the best interests of all shareholders to go ahead with that deal; (5) Mr Cheng was an acquaintance of Ms Chien but that is of very little significance and did not unduly influence Mr Cheng in undertaking his duties on the Special Committee; and (6) there were issues in respect of lack of communication and discovery but again they were not significant.

¶ [813] I agree that there were plenty of opportunities for topping bids to be made. I do not accept the Dissenters' submission that no other rational bidders would even have considered competing against Mr Yan. There were plenty of opportunities for “rational bidders” to get into bed with Mr Yan if they wanted to progress a competing bid. The evidence indicates that Mr Yan’s mind was not closed to competing bids. Furthermore there was a proper market check and a competitive sales process.

¶ [814] In my judgment there were no substantial deficiencies in this case and I am satisfied that there was an adequate market check and that the Company should be permitted to rely on the merger price as it seeks to do so for the purposes of and in the context of this case.

¶ [815] There was no significant failure to explore expressions of interest to the extent that the Company should be prevented from relying on the merger price to the extent that it seeks to do so. The go-shops were not conducted in a defective manner. There was a lack of topping bids but that is not indicative of there Page 227 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 228 of 257 being no robust market check. I agree with the Company that the go-shops in this case were sufficient to enable “the Special Committee to satisfy itself that there were no competing bidders out there.”

¶ [816] Rising above all the detail, it is clear that if at the end of the hearing having considered all of the evidence the court is of the view that the merger price is a reliable indicator for the purpose in which a company seeks to engage it then the court should permit such engagement and I do so in this case.

¶ [817] I accept that the evidential burden was on the Company to show that the conditions justifying reliance on the merger price were satisfied. On the facts and in the circumstances of this case the Company satisfied that burden. Two merger prices

¶ [818] The Carey Olsen Dissenters in the Dissenters' list of issues filed on 28 August 2025 raise issues in respect of two merger agreements and in effect ask whether the merger price of the Original Merger Agreement satisfies more of the conditions justifying reliance on it and constitutes more reliable evidence of fair value than the price in the Revised Merger Agreement.

¶ [819] I note the conclusion of the Carey Olsen Dissenters at paragraph 161 of their written submissions to the effect that the Original Merger Price is in principle a fair and equitable proxy for the fair value of shares. They say that in contrast the Revised Merger Price was made when the market was already affected by the Original Merger Price. They add that the shareholder base changed, analysts stopped reporting and the market price cannot support the merger price because it needs to be fundamentally adjusted.

¶ [820] I note also footnote 10 of Professor Lehn’s supplementary report (GA/5/8). Mr Boulton (Day 18/40:3- 4) submitted that we have two roll-forwards from Professor Lehn. His primary one starts from 11 January 2022 and it is recognised that it may still have some lingering effect from the fact that there is a $79 offer on the table. The Company says that if there was any effect on the price at 11 January it would have been that $79 was holding the price higher than it would otherwise have been as some investors still hoped that the $79 deal would complete. The Company say they have a cross-check which is if you go back to September 2020 and do the same roll-forward with the same indices, the same regression, over the whole period to April 2022 you arrive at $30.88. Page 228 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 229 of 257

¶ [821] On this two merger prices point I accept the evidence of Professor Lehn and his AMTP approach and much prefer his approach to the position adopted by the Carey Olsen Dissenters on this point. Disclosure/discovery issues

¶ [822] In its list of issues the Company at paragraph 26 asks the question “Is there any reason to believe that that (sic) any material which was not disclosed would have indicated material failings in the merger process?”

¶ [823] In their list of issues the Dissenters at paragraph 25.2 ask the question “Have there been any failings of discovery by the Company and, if so, can the Company rely on the merger price as a cross-check notwithstanding such failures?”

¶ [824] In my judgment there have been no material failings by the Company in respect of disclosure/discovery issues which would prevent the Company from relying on the merger price to the limited extent it does so or which would necessitate the court making adverse inferences against the Company.

¶ [825] In particular I record that, although not entirely satisfactory, I accept the explanations offered by Mr Yan and Ms Chien in respect of issues connected to their mobile phones and the Signal application.

¶ [826] In their written closing submissions the Dissenters under the heading “Non-Disclosure and Missing Evidence” beginning at paragraph 572 on page 235 make the point that where the merger price is being considered, there must be a particularly strong focus on the integrity of the evidence and discovery given by the Company relating to the management buyout. The Dissenters submit that in this case there were serious shortcomings especially in relation to the main actors concerned in the discussions around the merger price. The Dissenters refer to Trina Solar and Simetra Global Assets Ltd v Ikon Finance

¶ [2019] 4 WLR 122 (CA) at [48] and submit that a key element of the company being “entirely open” is the disclosure of internal communications by reference to the particular importance of “a party’s internal documents including e-mails and instant messaging” which “tend to be the documents where a witness’ guard is down and their true thoughts are plain to see”. The Dissenters, relying on Trina Solar, make the point that the obligation is on the company to fulfill its discovery obligations in Section 238 cases and not on the dissenters to seek specific discovery. The Dissenters say that there are Page 229 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 230 of 257 consequences if the company fails in its obligations including possible costs orders and adverse inferences being drawn where appropriate.

¶ [827] The Dissenters in their written closing submissions relied on what they described as “four key issues”: (1) first, during the course of the proceedings the Company has misrepresented the use of the “Signal” messaging application; (2) second, a number of custodians erased relevant instant message data or otherwise failed to preserve personal devices once the merger had been completed and before discovery of that data had been given; and (3) fhird, the Company failed to take adequate steps to preserve, retrieve and produce data from instant messaging communications.

¶ [828] The Dissenters say that they will invite the court to draw adverse inferences.

¶ [829] During Mr Adkin’s oral submissions I asked him to indicate at some stage during the course of his submissions the “specific inferences” he was asking the court to draw. Mr Adkin said he would do so but he also made a general point to the effect that if a company comes to the court seeking to rely, even just as a cross-check, on the merger price it is going to run into difficulties, even without any inference being drawn, if it is unable to produce all of the documents and all of the relevant witnesses which bear on the merger process (Day 19/123: 16-23; 124: 1-19). Mr Adkin said it was not necessary for the court to draw inferences; all the court has to do is to conclude that the company “has not handed over the goods” and therefore cannot make out its case. Mr Adkin added that “the unhappy story of the missing Signal messages” revealed more than “a hint of suspicion that the court has not received the full unvarnished contemporaneous documentary record”. Mr Adkin said that the inference he is asking the court to draw is “quite a simple one.” The inference is that that was not an accident, it was deliberate, and it was done in order to prevent correspondence and communications of relevance being subject to scrutiny. Mr Adkin asked me (Day 19/126:23-25; 127:1-14) to bear in mind 3 points: (a) there were documents that should be here and are not because of choices made by the Company; (b2) to infer since those choices were made that these documents would have been of assistance to the court; and (c) to infer since those choices were made that the people making those choices did not want them to be seen Page 230 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 231 of 257 by the court or the Dissenters and therefore they would be unlikely to have been of assistance to the Company. Mr Adkin (Day 19/127:24) agreed that the next stage is that the inference must be that the evidence would have indicated that the merger process was not robust. Mr Adkin (Day 19/128; 23-25) submitted relying on Trina that the court is entitled to say “Well, you’ve not produced the goods and, therefore, you can’t rely on this particular methodology because there are gaps.” In my judgment that is unreasonably stretching what Birt J.A. stated in Trina Solar on disclosure to breaking point.

¶ [830] Mr Adkin (Day 19/139: 20-25; 140:1-4) later added that the “mass and nature of what’s missing is, of necessity, a matter of speculation” but he submitted that “it’s not difficult to conclude that it would have been of very great assistance to your Lordship in the case.”

¶ [831] Having considered the evidence and the submissions on this disclosure/discovery point I have concluded that I should not shut out the Company from relying on the merger price for the purposes for which it relies upon it nor should I draw any adverse inferences against the Company. In Trina Solar (CICA) Birt JA at [146 (iv)(f)] stated that “the company’s efforts in relation to discovery had been lamentable”. The company in Trina Solar had also failed to call relevant witnesses. The same cannot be said in the case presently before me. This case is a very different case to Trina Solar. In the case before me the Company made relevant witnesses available for cross-examination, positively and constructively engaged in the proceedings (as outlined by Mr Imrie in his closing oral submissions) and provided more than adequate discovery of relevant documents. The acquisition premium

¶ [832] Professor Lehn says that the per share merger consideration of US$61 is likely to substantially overstate the fair value of a share as of the Valuation Date as it included a substantial acquisition premium and the fair value declined substantially after 1 March 2022. I accept Professor Lehn’s evidence in this respect and the evidence of the Company’s management in respect of the deteriorating market environment.

¶ [833] Put simply, Professor Lehn’s AMTP starts from a trading price of $45.83 that was affected upwards by the Original Proposal and then rolls that price forwards to the Valuation Date using the mathematically estimated relationship between the returns on the Company’s shares and the returns of the Golden Dragon Index. The result is an estimate of fair value on the Valuation Date of $31.11. Professor Lehn’s Page 231 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 232 of 257 estimate of fair value is consistent with the break prices disclosed by a number of the Dissenters. I do no place undue reliance on the evidence in respect of break prices but it is evidence that the court can take into account.

¶ [834] At paragraph 93 of his first report Professor Lehn in a section entitled “The relation between the merger consideration and the fair value of a share” states: “… A large body of academic literature finds that, on average, merger and acquisitions increase value, with most of the gains accruing to shareholders of the acquired companies [Holthausen & Zmijewski (2020) at 802]. Hence, the value of the merger consideration provided to target shareholders typically exceeds the fair value of the target company’s shares (i.e., the value of the Company’s shares under the assumption that the company is operated as a going concern, exclusive of any increase in value resulting from the acquisition) …”

¶ [835] Put in simple terms Professor Lehn’s evidence, which I accept, was that mergers generally increase value and so merger prices generally exceed the fundamental value of the target. Sir Andrew Popplewell in Trina Solar (JCPC) at [7] acknowledged, albeit in very general terms, that the market price at which shares are trading may be affected “by knowledge of the merger proposal.”

¶ [836] I also accept the submission made on behalf of the Company that in Section 238 cases the court is concerned with a share’s true or intrinsic value (i.e. its fundamental value excluding the benefits to be derived from the merger transaction, including the price offered for the shares in the merger). If a transaction does not proceed, any merger premium will fall away, and a company’s market-trading price will generally revert to fundamental value, in due course.

¶ [837] I accept that an offer to take-private is invariably made at a premium to the market trading price. This is supported by some of the many academic texts put before the court. For example T Kirchner Merger Arbitrage: How to Profit from Global Event-Driven Arbitrage (2nd Edition, John Wiley & Sons, Inc 2016). At Chapter 4 “Incorporating Risk into the Arbitrage Decision” at page 103: “Most acquisitions are made at a significant premium to the most recent market price”. Page 232 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 233 of 257 and at page 146: “Most mergers are done at a significant premium to the last trading price of the stock.”

¶ [838] Mr Kirchner is said to be a Portfolio Manager at Camelot Event Driven Fund, a series of Frank Fund Trust, a dissenter in this case holding 15,400 shares. The market price

¶ [839] As Trina Solar (CICA) made clear if weight is to be placed on the adjusted trading price (the word “place” at [41] of Trina Solar in the CICA appears to be a typo for “price”) of the shares there must be a semi-strong efficient market in those shares. Semi-strong efficient market

¶ [840] I am satisfied in this case that there was a semi-strong efficient market in the shares. The shares were well-traded stocks in a mature market.

¶ [841] There is academic literature that suggests that public stock markets in the USA such as the NASDAQ exchange are generally considered to be semi-strong form efficient.

¶ [842] In Parrino Fundamentals of Corporate Finance (2022) at 2-12: “The concept of semi-strong-form efficiency is a reasonable representation of the public stock markets in developed countries such as the United States.”

¶ [843] In the Joint Memorandum at 3.3.6 Professor Lehn’s position is that “[p]ublic stock markets in developed countries such as the USA are generally considered to be semi-strong form efficient.” Mr Taylor’s position in the Joint Memorandum in the lengthy Matters Not Agreed section was “[t]here is no presumption of efficiency for the purposes of assessing Fair Value.” However in cross-examination Mr Taylor accepted that it was “generally right” that most economists would say that there is a presumption that shares traded on major exchanges, such as NASDAQ, are semi-strong efficient. Page 233 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 234 of 257

¶ [844] I do not however lose sight of Parker J’s comment at [265] of FGL that: “There is no presumption in Cayman Islands law that markets are efficient, or that a market for a share at a given time produces a trading price that approximates to fair value. Those matters must be proven on the facts of the case and on the balance of probabilities”.

¶ [845] In the case presently before me having considered the evidence I have concluded that the market was semi-strong efficient and the trading price at the relevant time was approximate to fair value and indeed may well have exceeded it. I accept Professor Lehn’s AMTP approach and his evidence in this respect.

¶ [846] The authorities make it plain (and both experts accept) that event studies can be of great assistance in determining whether there is a semi-strong efficient market. Professor Lehn has exceptional credentials and experience in carrying out event studies. Mr Taylor less so.

¶ [847] I accept that Professor Lehn’s event study was a well-designed study backed by years of experience and supported by some of the academic literature including Bowman Understanding and Conducting Event Studies (1983) at page 577 where the author makes it plain that “the expectations model of security price reactions may be faulty … and … the methods of calculating and/or aggregating excess returns may be improper” and Johnston & Zhao Contrarian Share Price Reactions to Earnings Surprises (Journal of Accounts Auditing & Finance (2012) at 238-239) (methodology and sample construction) “… the direction and magnitude of the earnings surprise is not a reliable indicator of the market reaction to earnings announcements.” I found Mr Adkin's cross-examination of Professor Lehn to the effect that Professor Lehn was only able to conclude that the market was not “perfectly inefficient” difficult to follow and unpersuasive. Professor Lehn’s experience and expertise in event studies and the reliability of the event study he had undertaken in this case and his interpretation of Mr Taylor’s event study and the proper conclusion that should be drawn from it were undamaged during his cross-examination in this important area.

¶ [848] Professor Lehn’s event study was easy to follow and I had no hesitation in accepting its conclusions. The conclusion of Mr Taylor’s event study and his subjective self-imposed decision rules and subjective views of the potential impact of certain events especially those with “mixed news” were seriously undermined in a well-focused and comprehensive cross-examination on these points. I can quite understand why the Company was so overjoyed with that cross-examination that it devoted a 7 page Appendix to the Company’s written closing submissions (Appendix 2). I accept how this is described Page 234 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 235 of 257 at paragraph 223 of the Company’s closing written submissions. It is a summary of the cross- examination of Mr Taylor on his event study where it was successfully shown that for every event which Mr Taylor concluded demonstrated inefficiency (or was inconsistent with efficiency) that result was driven by, amongst other things, his subjective decision rules as to directionality and magnitude, his inappropriate extraction of the 2.5% threshold from MacKinlay and his subjective reading of the analyst reports (which never caused him to change his view on the expected price response derived solely from the quarterly revenue and EPS figures). I agree that by modifying or removing the subjective criteria around each event it can be shown that Mr Taylor should, or at least reasonably could, have concluded that each event he studied was evidence of market efficiency, not inefficiency.

¶ [849] Professor Lehn’s evidence in respect of his event study to establish market efficiency withstood a prolonged attack upon it during cross-examination. The conclusions made by Mr Taylor in respect of his event study were demolished in a methodical and effective cross-examination.

¶ [850] Mr Adkin said that the court was in “interesting territory” as “Professor Lehn says that he has never seen an event study done in the way Mr Taylor has done his, and Mr Taylor says that he has never seen an event study done in the way Professor Lehn does his.” Professor Lehn was crystal clear that the design and implementation of his event study was entirely standard (and Mr Taylor’s was not). I accept Professor Lehn’s evidence in this respect. Professor Lehn put the position well when he stated (Day 8/152:18-155:2): “… the common denominator to a sound event study is that you have a well-defined null hypothesis … it’s a scientific test, allowed us to reject that null hypothesis, which is the ultimate test of market efficiency, and that I have established … there is a science for testing for market efficiency… we have rejected that null hypothesis, which is the ultimate test of market efficiency. And that I have established. And as a matter of fact, if you take Mr Taylor’s event study analysis – and he invents a methodology I’ve never seen before in the academic literature, where he arbitrarily says “I think these days are positive news and these days are negative news”, all of which is arbitrary because he has arbitrary decision rules, so he misinterprets his own results, but if he goes back to the null hypothesis that he sets out in his initial report, his own very event study would allow one to reject the null hypothesis, which means that his event study actually establishes market efficiency contrary to the inference that he draws from his data.” Page 235 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 236 of 257 No MNPI

¶ [851] There was no significant material non-public information in this case.

¶ [852] Mr Taylor in his report adopted a somewhat scattergun/kitchen sink approach in respect of MNPI. He made a long list but did not focus on the M in MNPI.

¶ [853] During his oral closing submissions I asked Mr Adkin to list the MNPI the Dissenters relied on. He said the information behind the November 2021 Press Release was “in effect a thumping great piece of MNPI”. He also made fleeting reference to the Ocean Link projections. The November 2021 Press Release

¶ [854] There was a lot of evidence and a lot of argument presented to the court in respect of the November 2021 Press Release. The essence of the Dissenters’ argument was that the Company’s management knew, but the market did not, that the discussions with the regulator announced in the November 2021 Press Release concerned the merger and not the company’s operations.

¶ [855] In view of the emphasis placed on the press release by the Dissenters it is worthwhile noting it in its entirety. It read as follows: “(“51job”, or the “Company”), a leading provider of integrated human resource services in China, announced today that certain members of the buyer consortium that formed Garnet Faith Limited to acquire the Company (the “Proposed Transaction”) pursuant to the previously announced agreement and plan of merger dated as of June 21, 2021, have been in consultation with Chinese regulators on recent regulatory changes in China that may be applicable to the Company and the Proposed Transaction. The consultation process by these buyer consortium members is currently ongoing and a clear timeline to its completion cannot be provided at this time. The Company intends to file an updated Schedule 13E-3 Transaction Statement (including the Company’s proxy statement) as soon as possible, subject to the progress of the Consultation.” Page 236 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 237 of 257

¶ [856] Although I accept that the wording of the press release was far from ideal, I do not accept that the market was misled by it or that it constituted MNPI within the definition of that term in this context.

¶ [857] Construed objectively and reasonably it appears from the terms of the press release that the consultation was to do with the Merger and not the Company or its operations. Furthermore, the fall in the Company’s share price subsequent to the issuing of the press release is consistent with the market interpreting the press release as negative news about the likelihood of the Merger completing.

¶ [858] The November 2021 Press Release is not indicative of MNPI. The Ocean Link forecasts/projections

¶ [859] The Dissenters further allege that the revenue forecasts/projections produced by Ocean Link in June 2021, which apparently projected higher revenue growth and profitability than the 2021 Management Projections were MNPI because they were not disclosed to the market. The Dissenters say that it is clear there was MNPI in existence at the relevant time (11 January 2022) that was potentially relevant to fair value and which in consequence meant that the shares were potentially undervalued by the market.

¶ [860] The main difficulty with the Dissenters’ reliance on the Ocean Link forecasts as evidence of MNPI is that they have never been in the possession of the Company and they are not even in evidence before this court in their entirety. There is some evidence of selected extracts from them in the transcript of Mr Jiang’s section 1782 deposition transcript but the entire projections document has not been put into evidence. There is insufficient evidence before the court to safely conclude that the Ocean Link projections amounted to MNPI and I do not so conclude.

¶ [861] There was some debate between the Company and the Dissenters as to the burden of proof of proving the existence or lack of the existence of MNPI. It was an interesting debate but it is not a debate I need to resolve. Whether the burden is on the Company to negatively prove the lack of existence of MNPI or whether it is on the Dissenters to positively prove the existence of MNPI I have concluded, based on the relevant law, evidence and submissions before the court, that there was no MNPI in this case. Page 237 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 238 of 257

¶ [862] In my judgment there was a semi-strong efficient market and no MNPI. In such circumstances the Company can properly rely on the AMTP. Value relevant information?

¶ [863] The Dissenters raised the issue as to whether the Company has established that there was no value- relevant information which came into existence during the roll-forward period the existence of which would render the rolled-forward figure unreliable.

¶ [864] I agree with the Company that overall Professor Lehn was clear and compelling that his regression analysis was a sophisticated method to get to a precise estimate of the fair value during the Adjustment Period (Day 8/181:1-25, 182:1-16).

¶ [865] I agree with the Company that it is not correct to suggest that any “value-relevant information” that came into existence after 11 January 2022 undermined the reliability of Professor Lehn’s regression analysis. Professor Lehn was plainly acutely aware of this point and took it into account. It was put to Professor Lehn that if there was any value-relevant information that came into existence after 11 January 2022 that was specific to the Company that would not be captured in his AMTP methodology (Day 8/83:16-19). Professor Lehn was asked to take it in stages and it was further put to him that any value-relevant information would not be captured because he took the 11 January 2022 traded price and rolled it forward (Day 8/83:20-23).

¶ [866] The following exchange then took place (Day 8/83:24-84:11). : “A. Correct, and, you know, I’m aware of that, and when I have analysed similar situations we are cognisant of the fact that the roll-forward is accounting for market movements and industry movements but it’s not accounting for firm-specific information. Q. Yes. A: So what I have done in other cases and what I’ve done in this case is look to see if there were additional information disclosed to the market that would be deemed to be firm-specific that - - that would elicit a stock price reaction. So there are ways Page 238 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 239 of 257 that one can accommodate that with the roll-forward but it was unnecessary in this case.”

¶ [867] I accept Professor Lehn’s evidence and the Company’s submissions on this point. I am satisfied that the rolled-forward figure was reliable. “Deal stock”?

¶ [868] The Dissenters further attack reliance on an AMTP approach by contending that the market price after the Original Proposal was a so-called “deal stock” affected only by trading based on the likelihood of the merger completing due to the existence of merger arbitrageurs and therefore the market price did not reflect the Company’s fundamental value.

¶ [869] In short I accept Professor Lehn’s evidence on this point to the effect that those selling the shares and the merger arbitrageurs do have regard to the Company’s fundamental value.

¶ [870] Professor Lehn (Day 7/pages 136-137) was clear and persuasive in his expert opinion that an arbitrage spread consists of two components. First, the offer price assessed by the probability that the deal will go through and then what the merger arbitrageurs refer to as the “break price” which Professor Lehn said was the stand-alone value of the target company if the deal does not go through “which means the arbitrage spread is being determined by both the stand-alone value, i.e. fair value, and by the offer price, each weighted by probabilities.” Professor Lehn added: “Their business is arbitrage, but I think it’s misleading to say that they’re not focused in the long-term value of the company, because an important component of that spread, which is something that is prominent in their minds is the stand-alone value of the company, and the stand-alone value of the company is based upon long-term prospects for the company. That’s the so-called break price in the arbitrage spread.”

¶ [871] It is interesting and informative to note in this case that in the evidence before the court is contemporaneous evidence of some of the Dissenters’ internal assessments of what they regarded as the break price in this case. I agree with the Company that this evidence provides some support for Professor Lehn’s expert opinion on fair value but do not place undue reliance upon it. To take but one Page 239 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 240 of 257 example in “Monthly Update – April 2022” “1 April 2022”, Manalo Advisors arrived at an estimated break price of $32.33 per share if the merger did not complete. The Dissenters’ downward trend in their estimated value of the share price if the deal breaks is also, as the Company submits, consistent with the fall in the market generally and the decline in the standalone value of the Company. Affected roll-forward

¶ [872] The Company’s position is that the market price on the Valuation Date exceeded fair value because of the possibility that the merger would be approved at the EGM and stockholders would receive the merger price of $61. Accordingly to calculate the fair value of the Dissenters’ shares on the Valuation Date, the Pre-Announcement Price needs to be adjusted (or rolled forward) to the Valuation Date to calculate the Adjusted Pre-Announcement Price.

¶ [873] The Company says that Professor Lehn uses a commonly accepted method to estimate the Adjusted Market Price on the Valuation Date. First he finds the last date before the Valuation Date on which the market price was unaffected by the Amended Merger Agreement which was 11 January 2022, the day before the announcement of the Revised Proposal, which was priced at $57.50. The market price on this date was $45.83 i.e. the Pre-Announcement Price. Second, Professor Lehn performed a regression analysis to estimate the relation between the daily returns on JOBS, the corresponding daily returns on the NASDAQ Golden Dragon China Index , and the corresponding daily returns on an equal-weighted portfolio of the US-traded Comparable Company Stocks (the “Industry Index”) over the two-year period ending the day before the announcement of the Original Proposal. Professor Lehn then used this estimated relation to predict what the percentage change in the price of JOBS would have been on each day based on market and industry factors, absent the Amended Merger Agreement, and adjusted the Pre-Announcement Price for these predicted price changes during the period between 11 January 2022 and the Valuation Date (the “Adjustment Period”). The Pre-Announcement Price, as adjusted, decreased from $45.83 to $31.11 (i.e. by 32.12%) during the Adjustment Period.

¶ [874] The Company say that independent of the regression method used by Professor Lehn the 32.12% decline in the unaffected value of the Company’s shares appears reasonable in light of three economically adverse events that occurred during the Adjustment Period of 11 January 2022 to 26 April Page 240 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 241 of 257 2022 namely (a) Significant COVID-related disruption of the PRC economy; (b) Ongoing regulatory changes in the PRC and the USA; and (c) the Russian invasion of Ukraine.

¶ [875] The Company adds that a calculated 32.12% decline in the unaffected value of the Company’s shares is also consistent with quantifiable changes in the value of stock indices related to the Company and of companies comparable to the Company that declined during the Adjustment Period.

¶ [876] In cross-examination Mr Taylor accepted (Day 17/73:22-25, 74:1) that if you were to do a roll-forward you would “in principle” conduct it in the way Professor Lehn conducted his, adding: “You would also take into account other factors, like … the length of time and the events that have occurred in the interim and so forth, but in principle I agree that’s how you would go about doing it if you wanted to do it.”

¶ [877] I agree that the Dissenters’ criticisms of the regression analysis do not withstand scrutiny. In particular the Dissenters’ suggestion that the Golden Dragon Index is not an appropriate index to use is unjustifiable and unhelpfully they do not suggest an alternative index. I have no hesitation in accepting that Professor Lehn’s use of the Golden Dragon Index was appropriate and justified.

¶ [878] I agree that overall Professor Lehn’s evidence was clear and compelling to the effect that his regression analysis was an appropriate method to arrive at a “more precise estimate for fair value during the Adjustment Period.” (Day 8/181:13)

¶ [879] I also accept that a cruder method (as opposed to Professor Lehn’s sophisticated method) would have been to simply estimate the likely decline in the Company’s share price based on the general decline in other indices or companies. The Company says that this would come to roughly the same answer and provides a “confirming cross-check.” Co-movement issue

¶ [880] The Dissenters raised the issue as to whether Professor Lehn’s roll-forward model should be rejected as unreliable for failing to account for what they describe as “excess co-movement”. Page 241 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 242 of 257

¶ [881] I have to say that I found Mr Taylor’s evidence on “co-movement” to be unsatisfactory. I have set some of it out above. The up-to-date literature was against Mr Taylor. Mr Taylor cited an article from Professor Greenwood from 2008 but a more recent article by Professor Greenwood published in 2025 (GY/8) concluded that the co-movement effect was no longer meaningfully observed. Furthermore Mr Taylor accepted that he had not performed any analysis that demonstrated that the Company’s stock had excessive co-movement relative to the NASDAQ Golden Dragon Index (Day 17/82:11-15).

¶ [882] Mr Boulton put it to Mr Taylor that the up to date literature “is that any co-movement identified has dwindled to almost nothing; and we can see from the abstract that this article published in April 2025 … has fallen from an average of 7.9% in the 1990s to less than 1% over the past decade” and “So, although you were relying on Professor Greenwood’s analysis, it appears that he thinks that the index effect he now says on page 39 is near zero.” (Day 17/83:1-14). Faced with that recent article the best Mr Taylor could come up with was “[h]e – this updated article says the – when looking at S&P500, the – effect has lessened, yes” (Day 17/83:15-16).

¶ [883] I pause to record that at paragraph 245 of the Company’s closing written submissions it is stated “Mr Taylor conceded under cross-examination that the article suggested that the co-movement effect had “lessened”, if it exists at all” (Day 17/83:16). Mr Taylor did not use the additional words “if it exists at all”, although he did accept that “the effect has lessened”. The first sentence of paragraph 245 had the potential to mislead.

¶ [884] I do however entirely agree with the second sentence of paragraph 245, namely Mr Taylor’s position on co-movement is not a credible critique of Professor Lehn’s valuation approach. Mr Taylor appears to have raised it simply as a potential theoretical possibility without a proper and balanced consideration of the recent academic literature on the topic.

¶ [885] There is nothing in the Dissenters’ co-movement issue. Volatility of Golden Dragon Index

¶ [886] The Dissenters raise the issue as to whether “the volatility of the Golden Dragon Index during the roll- forward period cause the roll-forward adjustment to be inappropriate or unfair for the same reasons as the roll-forward adjustment in FGL.” Page 242 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 243 of 257

¶ [887] In FGL Parker J rejected Professor Lehn’s Adjusted Unaffected Market Price approach as the sole determinant of fair value. Parker J accepted Mr Davidson’s view that an adjusted price methodology was too speculative in light of dislocation of market prices from drivers of value on the valuation date ([592]). Parker J stated that the adjusted unaffected market price which reflected the economic impact of the COVID-19 pandemic also did not produce a fair value. Parker J therefore rejected Professor Lehn’s view that an adjusted price methodology, including a regression analysis to assess the notional market trading price of an ordinary FGL share at the valuation date should be $8.60, which represented a fall of over 15% ([593]). Parker J felt that the intrinsic value of FGL as at 29 May 2020 could be better and more fairly assessed as being US$11.06 defined by Parker J as the Transaction Price ([594]). I note Parker J’s conclusions at [378] to [387] in respect of the market price at the valuation date in that case.

¶ [888] Each case, of course, must be decided in its own context, on its own facts and circumstances, its own evidence and the submissions put before the court for consideration and determination. In this particular case I am not persuaded that Professor Lehn’s reliance on the Golden Dragon Index was misplaced. I have considered the arguments on composition and volatility. I have not concluded that the volatility of the Golden Dragon Index causes Professor Lehn’s roll-forward adjustment to be inappropriate or unfair.

¶ [889] I noted the Company’s position that Mr Taylor’s criticism that the Golden Dragon Index was particularly volatile during the Adjustment Period conflates the issue of volatility during the estimation period i.e. the period up to 16 September 2020 (which was not a high-volatility period save for the months of February to April 2020) with volatility during the Adjustment Period. The Company also referred to Professor Lehn’s evidence (Day 8/202:5-19) and in particular the words “Mr Taylor and my estimation period, which ends on September 16, 2020, which Judge Parker was alluding to in the FGL case. I’ve actually replicated my results by excluding those three months …”.

¶ [890] I agree with the Company’s submission at paragraph 241.8 of their closing written submissions that the concerns which Parker J was expressing in FGL were different concerns in a different case relating to a different period with different expert evidence. I have noted the evidence at Day 8/202:5-19 and the Company’s point that it was in fact a different use of the word “volatility”. The Company also referred to Professor Lehn’s explanation at Day 8/203:2 to Day 8/205:10 and submitted that it was wholly Page 243 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 244 of 257 inappropriate to try and draw conclusions of fact from the facts in a different case to the present case. I agree that each case must be decided on its own facts.

¶ [891] The Company has satisfied this court that Professor Lehn’s reliance on the Golden Dragon Index was well placed. Affected trading price as start date for AMTP

¶ [892] The Dissenters say there is no previous Cayman Islands Section 238 case in which a merger-affected trading price has ever been used as the start date for an AMTP analysis and refer in some detail to the first instance decisions in iKang and Trina Solar. I do not think that the first instance judgments in these cases are authority for the legal proposition that in no case can you use a merger affected trading price as the start date for an AMTP analysis. Each case, of course, must be decided on its own facts and circumstances. In the case presently before me I have been persuaded by the Company that Professor Lehn’s AMTP approach is a reliable indicator of fair value as at the Valuation Date.

¶ [893] The Dissenters place considerable emphasis on the first instance judgments in iKang and Trina Solar.

¶ [894] The Dissenters refer to [396] of iKang and the identification by the dissenters in that case of the “essential problem” namely that “… for almost 3 years the market trading price of the Company was primarily determined by market participants’ assessment of the likelihood of a deal being concluded and the likely price paid under such a deal. It followed that there is no relevant “unaffected price” (probably after the first take private announcement on 31 August 2015) and that the market trading price was not an indicator of fair value.”

¶ [895] As is apparent from [397] of iKang the dissenters in that case submitted that the Company had failed to establish (and that there was no evidence) that the market trading price had shifted to being driven by fundamentals after June 2016.

¶ [896] The Dissenters place significant reliance on Segal J’s acceptance of the dissenters’ submissions regarding the “essential problem” in iKang and the need for “convincing evidence” to be provided to the effect that buyers and sellers would likely have made decisions as to price in the period up to the valuation date in iKang (20 August 2018) “without the distorting effect of the Merger process and Page 244 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 245 of 257 without giving material weight to the effect of a Merger on price and on value” ([223]). The Dissenters say that Segal J rejected the contention that any reliance could be placed on an AMTP in iKang. The Dissenters add that Professor Lehn has not even sought to suggest that the Company was trading on fundamentals at his AMTP start-date.

¶ [897] The Dissenters say that Segal J also considered a start date for an AMTP analysis which post-dated the relevant merger offer in Trina Solar. The Dissenters say that it was wrong to suggest (as Mr Boulton had done in his oral opening submissions for the Company) that an affected trading price had been used as the starting point for the AMTP analysis in that case. Mr Boulton had relied on [115](d) of Segal J’s judgment in Trina Solar where one of the main points relied on by Ms Glass (the company’s expert in that case) was that “after the announcement of the approval of the merger by the Company’s board on the Acceptance Date, ADS trading was driven, at least in part, by the Merger Price, as opposed to the company fundamentals or intrinsic value” and [117](b) where Segal J was summarising the submission of the dissenters.

¶ [898] The Dissenters say that the start date for the AMTP roll forward in iKang had ceased to be affected by the merger on that date. The Dissenters say that Segal J in Trina Solar held that Ms Glass’ “methodology for assessing the unaffected and adjusted price” was reasonable ([139]) and that this was consistent with his subsequent decision in iKang.

¶ [899] The Dissenters say that it is necessary for an expert wishing to rely on an AMTP start-date after the announcement of a merger to produce convincing evidence that trading had reverted to being driven by financial fundamentals and no such evidence has been produced in this case.

¶ [900] In this case I accept that the market price of $45.83 on 11 January 2022 the day before the announcement of the revised merger, is a reliable indicator of fair value on that date. Furthermore, on the balance of probabilities the market price on that date was affected upwards by the Original Offer. I am content, in the circumstances of this case, with Professor Lehn’s roll-forward from 11 January 2022 to the valuation date of 26 April 2022 and do not think that Segal J’s approach in iKang or Trina Solar precludes the court’s acceptance of this method by Professor Lehn on the facts and in the particular circumstances of this case. Page 245 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 246 of 257

¶ [901] I accept the Company’s submission that the court can start with the price on 11 January 2022 even though it is likely “affected”. The AMTP method starts with the trading price in a semi-strong efficient market before the announcement of the relevant proposal to take the company private. I accept that thereafter the share price is affected by the relevant merger price and includes a premium to the publicly traded price which is not part of the fair value to which the Dissenters are entitled.

¶ [902] The proposed roll-forward and affected or unaffected price arguments in Trina and iKang were all determined on the specific evidence and arguments presented in those cases. I accept the Company’s submission that contrary to what is suggested by the Dissenters, iKang is not an authority for the proposition that the court cannot roll-forward from an affected price. I accept the Company’s submission that while the court in iKang declined to place any weight on the market trading price in the circumstances of that case, a close and objective reading of [406] to [416] reveals that this was due to the specific evidence and finding made in that case which are not applicable in the case presently before the court.

¶ [903] In this case Professor Lehn, an impressive individual with considerable expert experience in this area, was persuasively consistent in his evidence that the regression analysis he conducted was methodologically sound and the most accurate and reliable basis for the necessary adjusting of the pre announcement price for probable changes in value in the period to the Valuation Date (Day 8/181:11 - 182:16). I agree that for the reasons put forward by Professor Lehn in his evidence and by the Company in its submissions that the Dissenters’ submission that the share price of a company under offer becomes “divorced” or “detached” from “fundamentals” is misconceived. I accept the Company’s submission that Professor Lehn’s selection of the Pre-Announcement Price for his justifiable AMTP approach is appropriate because it selects a price which reflects all market and industry developments that occurred over the relevant period. Although heavily criticised by the Dissenters I have also found some comfort Professor Lehn’s cross-check contained in the much mentioned footnote 10 in Lehn 2.

¶ [904] In a section where Professor Lehn says that Mr Taylor’s fair value is implausible on its face because it is inconsistent with relevant economic evidence, one example of this justifiable criticism reads as follows: “Mr Taylor’s valuation of $111.06 per share is 63.0% higher than the closing price of the ADSs on September 16, 2020, of $68.12 (i.e., the Unaffected Price), whereas the NASDAQ Page 246 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 247 of 257 Golden Dragon China Index declined by 52.48% between September 16, 2020, and the Valuation Date.10”

¶ [905] Footnote 10 reads as follows: “If the Unaffected Price were adjusted using the methodology described in Lehn Report [paragraph] 28, the adjusted Unaffected Price on the Valuation Date would be $30.88, which is 54.67% lower than the Unaffected Price.”

¶ [906] The Dissenters and their expert Mr Taylor make frequent reference (the Company says at least thirty times) to the price of $45.83 at the starting point on 11 January 2022 as an “affected” price. Mr Taylor initially suggested it was affected down because the $79.05 Original Merger Price acted as a “ceiling on the value in line with the merger price” (Day 14 108/9-10). I sought clarification as follows: “Justice David Doyle: Sorry, counsel is asking you quite a straightforward question, and I just need to get my head around it. This may be oversimplifying matters, but you’ve got – 11 January 2022 you have a share price of, say, $45 and then there’s a deal on the table at 79 and it was put to you, how could that depress the share price? Can you answer that question? A: Yes, no, so I don’t think it’s depressing it but what I’m saying is that – so one option – is that the deal goes ahead at 79 but there’s obviously other factors at play which result in the – in the share price being lower.”

¶ [907] Mr Taylor also repeated his oft-repeated mantra that it was “difficult to determine the fair value of underlying stock because it’s not trading on fundamentals.”

¶ [908] With some repeated judicial prompting, Mr Boulton eventually gracefully and helpfully descended into the detail and met head on the Dissenters’ submission that in Trina Solar the roll-forward was from an unaffected date and took me to passages in the transcript of the evidence of Ms Glass in the judgment of the Court of Appeal and then to what he described as the critical paragraph namely [65]. Page 247 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 248 of 257

¶ [909] In my judgment, the fact that the roll-forward may be from an affected price does not prevent reliance on the AMTP method. Taking Birt JA’s guidance at [65] into account in the context presently before the court I conclude that whether it is right to give 100% weight to the adjusted trading price in “any particular case is a matter of evaluation having regard to the particular circumstances.” Having regard to the particular circumstances of this case I have concluded that the AMTP method is the only reliable indicator of fair value.

¶ [910] Mr Boulton says that iKang must be viewed in context and stresses that the market evidence in that case was much more rudimentary than the evidence presented to me in this case. In iKang the company’s expert did not do an event study ([375]) nor did he do a roll-forward ([372]) and appears to have had reservations about whether the market was efficient ([404]) and hence he only placed 10% reliance on the market price. Professor Lehn’s evidence is very different to the expert evidence presented to the court by the company in iKang. In iKang trading volumes had declined ([409]) and the CEO had commented that the company’s price was not driven by fundamentals while the results of the privatisation process were unknown ([407]). I accept Mr Boulton’s submission that the approach of the Company in iKang is “a million miles away from Professor Lehn’s approach where he deals explicitly with the fact that the start for a roll-over is likely, if anything, to be affected upwards by the offer at $79.05, and we say that iKang cannot be authority for the proposition that in undertaking a roll- forward one cannot use a market price which is affected by the ongoing merger process.” (Day 19/29:4- 11)

¶ [911] In conclusion on this point, there is no principled basis to say that there is in Cayman Islands law a prohibition of using an affected price or put another way a requirement to use an unaffected price in respect of the roll-forward part of an AMTP approach.

¶ [912] I dismiss the Dissenters’ arguments in this area and their submissions that an AMTP approach is the wrong approach to take in this case.

¶ [913] In addition to Trina Solar and iKang the Dissenters also placed reliance on Parker J’s judgment in FGL and his conclusion at [592] that “Professor Lehn’s Adjusted Unaffected Market Price approach should not be adopted as the sole determinant of fair value” in the case before him as “an adjusted price methodology is too speculative in this case in light of dislocation of market prices from the drivers of value on the valuation date.” At [594] in the particular circumstances of FGL Parker J concluded that Page 248 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 249 of 257 the intrinsic value of the company could “be better and more fairly assessed as being the Transaction Price.” Again I stress, as many judges have done so in the past, each case must be determined on its own facts and circumstances.

¶ [914] There is nothing in the evidence or submissions of the Dissenters that persuade me that the approach adopted by Professor Lehn in the case before me was the wrong approach to adopt or produces an unfair result. The Company has persuaded me that Professor Lehn’s approach is the only reliable indicator of fair value in this case.

¶ [915] Mr Taylor’s DCF approach in this case has been totally discredited in cross-examination and I accept Professor Lehn’s informed criticism of it. It is not a reliable indicator of fair value. I place no reliance upon it.

¶ [916] In Trina Solar (CICA) Birt JA at [36] stressed that the court must reach its own decision as to fair value and the constituent elements which go to make up that fair value, adding “it must not simply plump for one expert over another …”.

¶ [917] Birt JA in Trina Solar (CICA) also quoted from Vice-Chancellor Strine in a Delaware case in a passage approved by Kawaley J in Re Nord Anglia Education at [70] to the effect that the court may look at the opinions advanced by the parties’ experts, select one party’s expert opinion as a framework, fashion its own framework, or adopt piecemeal some of an expert’s model methodology or mathematical calculations but “the court may not adopt an “either or” approach and must use its judgment and an independent valuation exercise to reach its conclusion.”

¶ [918] However the Supreme Court of Delaware in Dell (14 December 2017) at page 35 stated that in some cases “a single valuation metric is the most reliable evidence of fair value and that giving weight to another factor will do nothing but distort the best estimate. In other cases, it may be necessary to consider two or more factors. Or, in still others, the court might apportion weight among a variety of methodologies. But, whatever route it chooses, the trial court must justify its methodology (or methodologies) according to the facts of the case and relevant, accepted financial, principles.” The Supreme Court of Delaware cited other leading authorities at footnote 104 to the effect that the court at first instance “has the discretion to select one of the parties’ valuation models as its general framework or to fashion its own: and although not required to do so: Page 249 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 250 of 257 “… it is entirely proper for the Court of Chancery to adopt any one expert’s model methodology, and mathematical calculations, in toto, if that valuation is supported by credible evidence and withstands a critical judicial analysis on the record.”

¶ [919] In an illuminating passage quoted by Birt JA in Trina Solar (CICA) at [37] Martin JA at [22] of Shanda Games (CICA) stated that: “In every case, the court’s task will be to assess the utility of the expert evidence to the determination of fair value … the court will determine generally, or on an issue-by-issue basis, whether an expert’s evidence is to be accepted in whole or in part and how conflicts are to be resolved …”. The AMTP methodology

¶ [920] In this case I have, after considering all the relevant evidence, arguments, and the law arrived at the conclusion that the only reliable method of valuation is the AMTP method. I have considered Professor Lehn’s methodology and calculations. I see no need to “tinker” with them. His approach is supported by credible evidence and has withstood an attack upon it by the Dissenters and a critical judicial analysis. The experts and the parties did not seek to persuade the court to use a blended or weighted approach. As I say, the Company and the Dissenters appear to have adopted an all or nothing approach.

¶ [921] In this case a reliable market-based indicator of fair value is available. I do not adopt a blended or weighted approach as I think the DCF method would be a very unreliable indicator of fair value in this case. To use it in any way would be to drastically distort the fair value. I do not think it appropriate to construe [35] (v) of Birt JA’s judgment in Trina Solar (CICA) as a statutory provision requiring a DCF valuation cross-check in all cases. As Sir Andrew Popplewell put it in Trina Solar (JCPC) at [16]: “Where the values reached by different valuation measures are relatively close, the concept of a cross- check between them has some coherence. Where, however, the values are widely divergent, it makes little sense to talk of using one as a cross-check against another.” In this case in my judgment the AMTP method is both the starting and finishing point. It is reliably available in the circumstances of this case and a DCF cross check would be totally inappropriate. Having had the benefit of submissions and evidence over a 20 day trial I can now see why Professor Lehn felt it totally inappropriate to Page 250 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 251 of 257 undertake a DCF cross-check on his AMTP approach and in my judgment the CICA and JCPC judgments in Trina Solar do not require a DCF cross-check in this case. No reliance can be placed on a DCF approach in this case.

¶ [922] The Company’s points in respect of optimistic projections, staleness and steady state were all well made. I was also concerned (as the Vice-Chancellor was in Stillwater at page 135 onwards) that the experts disagreed with so many aspects of the DCF model and the inputs where small changes would produce large savings in values. These genuine disputes can be seen from the competing lists of issues in respect of the DCF model. I also took on board Professor Lehn’s wise cautionary words “Garbage in, garbage out.” Nothing could be usefully recycled or adjusted. The DCF model was simply not fit for purpose in the circumstances of this case.

¶ [923] If any further support or test is required in respect of the reliability of Professor Lehn’s AMTP approach this can be found to an extent by his cross-check roll-forward from the totally unaffected date of 16 September 2020 and the decline in the relevant periods of comparable companies and indices.

¶ [924] At [35(v)] Birt JA in Trina Solar (CICA) stated that in certain cases “it will be appropriate to estimate fair value exclusively or primarily using a DCF valuation.” It must follow, as the reverse side of that coin, that if there is no reliable DCF valuation available the court may determine fair value exclusively using a reliable AMTP approach. There is no reliable DCF approach available in this case. I determine fair value by using exclusively the AMTP approach.

¶ [925] I do not consider any adjustments are necessary to Professor Lehn’s calculations. I do not see any need to tinker with them.

¶ [926] I specifically asked Mr Adkin to provide the court with his position in respect of any suggested adjustments to the AMTP approach if I were to favour it. Mr Adkin (Day 19/111: 2-13) stated: “Now, so far as concerns the court making adjustments to the constituent elements that go to making up fair value, the principle is perhaps rather less heavily engaged when one is concerned with an AMTP valuation, which tends to be somewhat binary in the sense that the traded price is what it is, subject only to the nature of the roll-forward adjustment …”. Page 251 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 252 of 257

¶ [927] Having considered the evidence, arguments and law in this case I see no necessity to make any adjustments in respect of Professor Lehn’s roll-forward or his AMTP approach generally. Standing back and sense-check

¶ [928] The Company and the Dissenters both engaged “sense-check” arguments and reminded me of the words of Briggs LJ (as he then was) at [52] of Chilukuri v RP Explorer Master Fund [2013] EWCA Civ 1307 to the effect that in any complicated process of valuation it is important, after looking at each element of the process, to stand back and ask yourself whether your “provisional valuation makes commercial or business sense, viewed in the round.”

¶ [929] I have duly stood back and although criticisms and adverse “sense check” points can be and are made against the figures arrived at by both experts in my judgment the $31.11 figure arrived at by Professor Lehn makes far more sense in the real world than the $111.06 arrived at by Mr Taylor. Mr Boulton (Day 12/206:25, 207:1-12) helpfully rounded off Day 12 by putting to Mr Taylor a summary of the last 45 minutes of his cross-examination including: “examples about the company’s own releases, most of which you saw as being negative, the increased expenditure on sales and market, the app removal, the BOSS, the regulatory and compliance burden, the increase in headcount and labour costs, what happened to the indices, what happened to Kanzhun, what happened to Liepin, the geopolitical trends, the resurgence in COVID and Morgan Stanley’s view on EBITDA margins, and my suggestion to you it’s inescapable that between September 2020 and April 2022 the value of 51job would have fallen and probably very significantly?”

¶ [930] Mr Taylor realistically responded (Day 12 207; 13-14): “… the indications that you’ve summarised would – would indicate that the share price would go down …”.

¶ [931] Mr Tayor added that he did not believe the starting point was a fair indication of fair value as Mr Hsu commented that it was at a low point and care needed to be taken about the margins in 2021. There can be no reasonable doubt however, on the evidence before the court, that the indicators were that between Page 252 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 253 of 257 September 2020 and April 2022 the value of the shares was falling significantly. I agree with the Company’s submissions based on the evidence before the court that in the near 2 year time period from when the Company’s exploration of strategic options in response to the serious challenges the Company was facing started to the completion of the merger, the world in which the Company had to survive had changed considerably and the challenges that initially prompted the review of the strategic alternatives only became more acute. I agree that the overall trend during that timeline makes sense of Professor Lehn’s considered opinion on fair value.

¶ [932] Although I do not place primary reliance on the evidence in respect of break prices I do accept the Company’s submissions that in reality and behind the scenes some of the Dissenters’ own modelling and estimates of break prices show a clear downwards trend on their illuminating estimate of what a share would be worth if the Merger failed. To a limited extent this evidence also supports Professor Lehn’s opinion on fair value.

¶ [933] I fully appreciate that it is unusual to have a fair value figure that is in fact significantly lower than the price at which the transaction was closed. Birt JA in Sina stated “dissenters are not entitled to a one- way bet … They are only entitled to fair value and that means a value which is fair to both the dissenters and the company … Dissenters are not gamblers. They are investors who have taken a considered decision that, on the basis of the position at the time of their decision to dissent, fair value is likely to be greater than the merger price. They of course take the risk that the court will not agree and will decide that fair value is in fact less than they thought …”. It is also unusual to have a “recut” merger price. However, each case must be considered on its own facts and circumstances and in this case the figure of $31.11 makes much more sense than the figure of $111.13, taking into account all the evidence and the prevailing circumstances. Conclusions

¶ [934] In my judgment the fair value is $31.11 per share.

¶ [935] On the facts and in the particular circumstances of this case I have concluded that 100% should be placed on the AMTP as the only reliable indicator of fair value to the exclusion of the DCF approach proposed by the Dissenters. The criticisms of the Company against the use of the DCF approach even as a cross-check are extremely persuasive and well-founded in this case and I agree that the AMTP Page 253 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 254 of 257 should be used to the exclusion of the DCF approach. No reliable DCF analysis is possible on the facts and in the circumstances of this case.

¶ [936] The Dissenters failed to persuade me that their DCF approach produced a reliable indicator of fair value or could assist in any way in determining fair value.

¶ [937] The foundations of the Dissenters’ ambitious and “heroic” DCF approach, to put it mildly, were totally undermined during a penetrating, wide ranging and well-focused cross-examination of Mr Taylor who on occasions was too ready to defend the indefensible. Furthermore, in short, as Professor Lehn pithily put it “garbage in, garbage out”. This court, driven by Professor Lehn’s balanced and well considered objective expert evidence and the focused attack by Mr Boulton on the Dissenters’ DCF approach has had little hesitation in coming to the conclusion that no weight whatsoever should be placed on the DCF methodology in the circumstances of this particular case.

¶ [938] In this case the court is not able to make any adjustments to the DCF factors or produce a DCF analysis of its own which would give it comfort that a reliable DCF analysis could be duly completed to produce a reliable or partially reliable DCF figure to enable a fair value to be properly arrived at, whether by weighting, blending or otherwise using the product of a DCF analysis. Having considered all the evidence and submissions I have reached the conclusion that a DCF analysis is of no assistance whatsoever to the court in this case.

¶ [939] The $111.06 figure Mr Taylor has come up with is from another planet in another galaxy totally removed from the real world the Company occupied and would occupy at the relevant times. It does not exist in reality. It only exists in the Dissenters’ speculative, distorted, and unreliable DCF spreadsheet. Summary of determinations

¶ [940] Having considered all the evidence, arguments and the relevant law in this case I determine the issues specified in the lists of issues field by the parties as follows for the reasons stated in this judgment.

¶ [941] As a matter of statutory interpretation and caselaw there is no formal strict hierarchy of valuation methodologies. In this case a reliable market methodology is available. I give the AMTP 100% weight. Page 254 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 255 of 257 The DCF methodology in this case would produce an unreliable indication of fair value. I give it no weight.

¶ [942] The Company has established that using an AMTP to estimate fair value as Professor Lehn contends is possible and appropriate in the circumstances of this case. It is far more reliable than Mr Taylor’s DCF.

¶ [943] The Company has established that the market price on 11 January 2022 was at or above fair value on that date.

¶ [944] In all the circumstances, the Company has established that the market traded price on 11 January 2022 represents an appropriate and fair starting point for an AMTP analysis.

¶ [945] The Company has established that the market for the Company’s share was semi-strong form efficient on 11 January 2022.

¶ [946] There was no MNPI.

¶ [947] Professor Lehn’s roll-forward method, together with his cross-checks, was the best method to use in the circumstances of this case. The Golden Dragon Index and the industry index were meaningfully comparable indices correlated to the Company’s shares. The volatility of the Golden Dragon Index during the roll-forward did not cause the roll-forward to be inappropriate or unreliable. The issue of excess co-movement did not render Professor Lehn’s model unreliable. There were no structural flaws which caused the roll-forward adjustment to be unreliable.

¶ [948] There was no value-relevant information which came into existence during the roll-forward period which would render the roll-forward figure less reliable than a figure estimated via any other valuation methodology.

¶ [949] The break prices do serve as a limited cross-check to the AMTP on the Valuation Date.

¶ [950] The court’s ability to rely on the market price is not adversely affected by the character of the Buyer Group or by the Support Agreements.

¶ [951] In all the circumstances the Company has established that the roll-forward adjustment performed by Professor Lehn produces a reliable estimate of fair value as at the Valuation Date. Page 255 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 256 of 257

¶ [952] The Company can rely on the merger price methodology as a cross-check. The Company adduced the necessary evidence to satisfy the conditions justifying reliance on the merger price. There has been no significant failings of discovery by the Company such that the Company cannot rely on the merger price as a cross-check.

¶ [953] There is no reason to believe that any material which was not disclosed would have indicated material significant failings in the merger process.

¶ [954] The following assists in proving that the fair value is at or below the merger price: (a) the stated reasons for the merger; (b) the fact of the merger agreement; (c) the acquisition premium embedded in the merger price; (d) the changes in the Company’s prospects and the estimated market price for the Company’s shares but for the merger between the date the merger agreement was signed and the Valuation Date; (e) the votes of the shareholders at the EGM; (f) the absence of topping bids; and (g) the fact that the merger was endorsed by the Special Committee, the Board, Duff & Phelps and others. I am satisfied that there was an adequate market check in this case.

¶ [955] I accept Professor Lehn’s evidence and the academic literature that merger arbitrageurs invest in a firm with, amongst other things, a regard for the fundamental value of the target.

¶ [956] In respect of the two merger prices point raised belatedly by the Carey Olsen Dissenters I accept the evidence of Professor Lehn on his AMTP approach and much prefer his approach to the position adopted by the Carey Olsen Dissenters on this point.

¶ [957] The Dissenters have not established that a reliable DCF analysis is possible or appropriate in the circumstances of this case. The projections were not unbiased central estimates. They were stale as the Valuation Date and the staleness cannot be adequately addressed by Mr Taylor’s proposed adjustments. The projections cannot be reliably adjusted to reach a steady state. The court is unable to come to a reliable conclusion as to an appropriate forecast. The court is unable to produce its own DCF analysis upon which any weight could properly be attached.

¶ [958] Using the AMTP methodology, the fair value is $31.11 per share. Page 256 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24 251124 In the matter of 51Job, Inc. – FSD 155 of 2022 (DDJ) – Judgment (Trial) Page 257 of 257 Ancillaries

¶ [959] Subject to consideration of any concise (no more than 5 pages) written submissions to the contrary filed and served within the next 21 days with concise (no more than 5 pages) written submissions in reply within 14 days thereafter, I am minded to order that the Dissenters pay the costs of the Company such costs to be taxed in default of agreement on the standard basis. If there are any written submissions to the contrary I am minded to decide any disputed costs issues on the papers. The Order

¶ [960] Counsel to file a draft order agreed as to form and content to reflect the above determinations within 14 days from the delivery of this judgment. _________________________________ THE HON. JUSTICE DAVID DOYLE JUDGE OF THE GRAND COURT Page 257 of 257 FSD2022-0155 2025-11-24 FSD2022-0155 2025-11-24