ACL Netherlands B.V. (as successor to Autonomy Corporation Limited) & Ors v Michael Richard Lynch & Anor

Fraud on a grand scale; or relentless witch-hunt? That is the question in this claim against two individuals for many billions arising out of a corporate acquisition more than a decade ago. The Claimants’ case is that they were fundamentally misled and are victims of fraud. The Defendants’ case is that the claim is “manufactured” to cover and justify a change of corporate mind, and to cast them as scapegoats for what in reality is buyer’s remorse coupled with management failings.

The acquisition in question (“the Acquisition”) was the purchase for approximately $11.1 billion in cash of the entire issued share capital of Autonomy Corporation plc (“Autonomy”) by a special purpose vehicle (“Bidco”). Bidco was incorporated by Hewlett-Packard Company (“HP”) for the purpose of the Acquisition. HP was the ultimate holding company in the Hewlett-Packard Group. The Acquisition was declared wholly unconditional on 3 October 2011. It was finally completed on 5 January 2012. The fall-out from it has spawned proceedings on both sides of the Atlantic, including at least two sets of criminal proceedings in the Northern District of California, USA.

The principal claim is brought by the Claimants under Schedule 10A of the Financial Services and Markets Act 2000 (“FSMA”). There are further claims in misrepresentation, in the tort of deceit, and for breaches of fiduciary and employment duties. The gist of the claims under the FSMA, and in misrepresentation and deceit, is that the Defendants dishonestly and deliberately misrepresented the financial performance of Autonomy in the period (“the Relevant Period”) from at least (Footnote: ¹) the first quarter of 2009 (“Q1 2009”) (Footnote: ²) until the second quarter of 2011 (“Q2 2011”), so that it was in truth a very different and less attractive proposition than was presented.

Put shortly: the Claimants’ main case is that Autonomy was mis-sold, and they were deceived into paying for it much more than it was worth. That is by far the largest claim. The Claimants have also asserted other claims seeking recovery of transaction-based losses in respect of transactions which were entered into in breach of duty.

The First Defendant, Dr Michael Richard Lynch (“Dr Lynch”) has vigorously denied both the fact of and his involvement in any impropriety. His case is that Autonomy at all times complied with applicable laws, regulations and International Financial Reporting Standards (“IFRS”) and followed the advice of its auditors, Deloitte LLP (“Deloitte”); and that he never came to know of any breach of such laws, regulations or accounting standards. He has consistently maintained that HP’s case is the culmination of sustained effort on the part of HP to shift on to the Defendants the blame for its change of mind about the transformation of HP which the Acquisition was intended to promote, and its subsequent failure to facilitate the efficient integration of Autonomy on which the strategy had depended. He has brought a counterclaim against Autonomy alleging a breach of a duty of trust and confidence and/or duty of care, as well as a breach of the Data Protection Act 1998.

The Second Defendant, Mr Sushovan Tareque Hussain (“Mr Hussain”), has sought to defend himself through solicitors and counsel; but he did not attend in person. He is presently serving a 5-year jail sentence in Pennsylvania, having been convicted by a jury in earlier criminal proceedings in the USA of 14 counts of “wire fraud” and for securities fraud in respect of the Acquisition. In the proceedings in this court, he filed a witness statement, but withdrew it when it became clear that he could not attend court. Having filed no evidence, he relied to a considerable extent on the defence to the first limb of the case (the allegations of impropriety) presented by Dr Lynch (who himself may now face criminal proceedings in Northern California, USA brought on broadly the same basis, for which the US prosecuting authorities are presently seeking his extradition). On the issue as to his personal knowledge of impropriety, he offered no positive case, though the Claimants must still of course prove theirs.

Those are the barest bones of claims tested in a 93-day trial which I believe may rank amongst the longest and most complex in English legal history. The First Defendant was cross-examined for 21 days; there were some 45 factual witnesses, many hundreds of pages of hearsay evidence largely comprised of transcripts from the US criminal proceedings, expert witnesses in three fields whose reports with appendices ranged over eight lever-arch files and who were cross-examined over (in total) 11 days, a trial bundle containing more than 36,000 documents and a ‘corpus' of many millions of electronical documents ranging over myriad transactions undertaken by Autonomy in various different fields between early 2009 and late 2011. The parties’ written closing submissions were together almost 5,000 pages long (excluding schedules and other annexed material), including in aggregate some 10,000 footnotes. There were bundles of 183 authorities and other references. The Claimants and the First Defendant were each represented by two Leading Counsel and five junior counsel (two of whom, happily, became QCs during the course of the trial); the Second Defendant was represented by junior counsel; all had instructing solicitors of the highest quality fielding large teams. Although I should record my thanks to the parties for providing me with a judicial assistant (Mr William Paris of Counsel) during the trial and its immediate aftermath, it was tried before me as a judge sitting alone.

Autonomy was incorporated under the laws of England and Wales in March 1996. Autonomy acted as a holding company for a group of companies all in the business of infrastructure software. Pursuant to a cross-border merger completed on 26 September 2017, long after its acquisition and some time after the commencement of these proceedings, all of the assets and liabilities of Autonomy were transferred to the First Claimant, ACL Netherlands BV.

Hewlett-Packard Vision BV (“Bidco”) was incorporated in the Netherlands on 15 August 2011. It was an indirect wholly owned subsidiary of HP until on 2 November 2015 it became an indirect and wholly owned subsidiary of Hewlett-Packard Enterprise Company. Pursuant to a merger which took effect on 27 October 2018, all of the assets and liabilities of Bidco were transferred to the Second Claimant, Hewlett-Packard The Hague BV.

The Third Claimant, Autonomy Systems Limited (“ASL”), was incorporated in England in 1995 and was an indirect wholly owned subsidiary of Autonomy. ASL operated as a licensor of Autonomy software to other Autonomy entities. Additionally, pursuant to transfer pricing agreements between ASL and some other Autonomy group companies, including the Fourth Claimant, costs and revenues of those other Autonomy group companies were transferred to ASL.

The Fourth Claimant, Hewlett-Packard Enterprise New Jersey Inc, was incorporated in New Jersey in 1996 and was formerly known as Autonomy Inc. I shall refer to it as Autonomy Inc in this judgment. At all relevant times Autonomy Inc was a wholly owned subsidiary of Autonomy and was Autonomy’s main operating company in the USA, based in Palo Alto, California.

The First Defendant (hereafter, “Dr Lynch”) was a director and the Chief Executive Officer of Autonomy from the time of its incorporation in 1996 up until 30 November 2011.

The Second Defendant (hereafter, “Mr Hussain”) was the Autonomy group’s Chief Financial Officer from June 2001 until 30 November 2011 and was a director of Autonomy from 01 June 2003 until 30 November 2011.

It is not disputed that both Defendants were, for the purposes of the FSMA claim, “persons discharging managerial responsibilities within the issuer” (“PDMRs”) within the meaning of Schedule 10A of FSMA (and previously s. 90A(4) before its amendment). The basis for the issuer’s liability is fraud on the part of at least one PDMR.

The fraud alleged consisted of the publication of information to the market which was known by the Defendants to be false. The allegation was based on (a) the allegedly dishonest description of Autonomy as being a “pure software company” when in fact it undertook and had become accustomed to inflating its apparent revenues by undertaking substantial hardware sales and (b) the allegedly dishonest presentation of its financial performance, which did not disclose and instead disguised improper practices which Autonomy adopted to boost and accelerate revenue. The Claimants contended that all this resulted in Autonomy being in fact an enterprise of considerably less value than it appeared to be on the basis of its published information. These improper practices included:

artificially inflating and accelerating Autonomy’s reported revenues;

understating Autonomy’s costs of goods sold so as to inflate gross margins;

misrepresenting Autonomy’s rate of organic growth; and

misrepresenting the nature and quality of Autonomy’s revenues, as well as overstating its gross and net profits.

In a little more detail, the claims relate to six areas of Autonomy’s business and accounting:

The “hardware case” relates to the purchase and resale by Autonomy (usually at a loss) of “pure” hardware (in broad terms, hardware unaccompanied by any Autonomy software) (Footnote: ³)in quantities (of approximately $200 million over the Relevant Period) which the Claimants allege were never disclosed to the market and which, by boosting apparent revenue, gave a false impression of the performance of Autonomy’s business and belied its presentation in its published information as a “pure software company”. The hardware case also raises issues as to (a) whether a proportion of the costs of the sales were improperly accounted for as sales and marketing expenses so as artificially to increase gross margins, and (b) whether Deloitte, who approved Autonomy’s accounting treatment of the sales, were misled as to the true purpose of the hardware sales.

The “reseller” or “VAR” case relates to 37 transactions between Autonomy (or in some cases, Autonomy Inc or another subsidiary, Zantaz Inc, “Zantaz”) and a small group of Value Added Resellers, which the Defendants treated as sales giving rise to revenue which could be and was recognised immediately in Autonomy’s accounts, but which the Claimants contended simply interposed a reseller between Autonomy and the true customer and were not in substance sales at all. The Claimants’ case is that in each VAR sale the VAR was only a passive placeholder with no further participation expected or permitted of it after the VAR sale. Thus, the VAR sales were, in effect, devices to accelerate recognition of revenue in Autonomy’s accounts, with the intended effect of misrepresenting its performance.

The “reciprocal transactions” case relates to what the Claimants alleged were back-to-back transactions with friendly counterparties, in which Autonomy purchased from the counterparty software or other goods or services that Autonomy did not need in order to fund the purchase by that counterparty of high margin software from Autonomy. The Claimants contended that these reciprocal or “round-trip” transactions also were contrived with the dishonest purpose of artificially boosting apparent high-margin software sales, with the effect of giving an exaggerated depiction of the success of Autonomy’s core business.

The “hosting case” relates to transactions between Autonomy (or one of its subsidiaries, Zantaz Inc, Autonomy Inc and ASL) and new or existing customers under which Autonomy agreed to forego future recurring revenue from the provision of hosted archiving and e-Discovery services (which was a substantial and lucrative part of Autonomy’s business) (Footnote: ⁴) for monthly (or other periodic) fees in return for the customer paying a one-off capital sum for a licence to use Autonomy’s software outside the hosted environment and whether in-house or in another provider’s data centre. The licence was alleged to be illusory, and its issue and sale was said to be for the dishonest purpose of treating it as akin to a sale of goods so as to justify the immediate (that is at the transaction date) recognition of the sale proceeds as revenue. Again, it was alleged that the intended effect was artificially to boost apparent revenue in the period in question.

The “OEM case” relates to transactions presented in Autonomy’s published information as generating “OEM” and “OEM derived revenue”. The Claimants’ case is that revenue so presented would be taken in the market to have been generated by a transaction with an Original Equipment Manufacturer (“OEM”) for Autonomy software to be embedded in the OEM’s software in return for royalty payments to Autonomy on all the OEM’s sales of the combined product (and thus a recurring revenue stream); but that in fact Autonomy included in what was compendiously described as the “OEM Metric” revenues from one-off sales of software licences to customers which were not OEMs and did not give rise to royalties or any other recurring revenue. The Claimants did not impugn the transactions themselves but contended that it was misleading and dishonest to include the latter revenues within the OEM metric because it gave the false impression of a valuable recurring category of revenues and thereby dishonestly misrepresented the quality and reliability of Autonomy’s revenue and earnings.

The “Other Transactions” case relates to four transactions entered into in late 2010 and early 2011 by ASL, Autonomy Spain SL and Autonomy Inc which the Claimants allege were also falsely accounted for in Autonomy’s published information as being licence sales but which were in truth the provision of a service.

The FSMA claim has a dog-leg nature: the gist of the claim is fraud on the part of the issuer (Autonomy) for which the PDMRs are alleged to be liable. That claim thus depends on establishing first, that Autonomy was liable (as issuer) to Bidco and that secondly, the Defendants were liable to Autonomy.

It is common ground that the Claimants need to make good their case at each stage. That is so even though in fact, both being under the control of HP, Autonomy has admitted liability to Bidco: the Claimants have accepted that this admission does not bind the court. They have accepted also that Dr Lynch and Mr Hussain will not be liable except in respect of misstatements or omissions about which they themselves knew. It is not sufficient for the Claimants to demonstrate that the transactions or the way they were accounted for was improper (the first limb); they need also to prove personal knowledge and dishonesty in respect of the false accounting on the part of the Defendants as PDMRs (the second limb).

Thus, for example, the “Other Transactions” case must fail against Dr Lynch, since he was not cross-examined on those transactions and his personal knowledge and involvement cannot be established: as the Claimants fairly and properly recognised in their closing submissions. By contrast, in the case of Mr Hussain, his decision not to give evidence means that it is difficult for him to resist a finding against him in all these claims at that second stage if the Claimants have succeeded at the first stage.

The claims under FSMA are by far the largest: the pleaded quantum of loss is at least $4.55 billion (being the amount for which Autonomy has accepted it is liable and for which it is alleged the Defendants are liable as PDMRs).

The claims for fraudulent misrepresentation and/or under s. 2(1) of the Misrepresentation Act 1967 are direct claims against the Defendants: they are based on personal liability, not on liability of the issuer. The quantum of the claims is much lower than the FSMA claims: the damages sought relate only to loss attributable to the shares and share options which the Defendants themselves each held and sold to Bidco. The pleaded quantum of loss is approximately $420 million.

The claims for transactional losses based on breaches of fiduciary and employee duties stand on a different footing. They do not arise in consequence of the Acquisition (except in the sense that they would almost certainly not have been brought if the Defendants still directed Autonomy). They are claims for direct losses suffered by ASL, Autonomy Inc and another group subsidiary called Zantaz as a result of the Defendants’ breaches of duty in causing the relevant subsidiary to enter into the impugned transactions without regard to the interests of that subsidiary. In that context, the Claimants’ primary case is that ASL is the proper claimant for these losses as a result of transfer pricing arrangements between ASL and each of Autonomy Inc, Zantaz and Verity Inc. That is disputed by the Defendants on the basis that the effect of the transfer pricing arrangements was only to shift costs and that in any event, whatever the effect of those arrangements in contract, the fact remains that no duties were owed to ASL in respect of transactions undertaken by different entities, variously Autonomy Inc and Zantaz. The pleaded quantum of loss is in excess of $76.1 million.

The Claimants accept that in the ultimate quantification of loss they must give credit for a recovery of $45 million made in a settlement of a related claim (against Autonomy’s auditors), after deducting the costs of such claim and any tax payable in respect of the settlement sum.

A ‘Dramatis Personae’ and a more detailed Schedule of Key Persons are attached as Appendix 1 and Appendix 2 respectively. A chart prepared by the Claimants setting out their various claims in diagrammatic form (at a very high level) is attached as Appendix 3. The Claimants also provided charts which they put forward as illustrating in graphic form (i) Autonomy’s reported revenue in each quarter of the Relevant Period; (ii) the alleged impact of undisclosed pure hardware revenue and other revenue alleged to have been recognised improperly; (iii) how (according to the Claimants) the recognition of revenue from impugned transactions helped Autonomy create the appearance of meeting market expectations and (iv) an overview of the position in each full year of the Relevant Period, and the half year to the end of June 2011. These are attached in Appendix 4, but I have slightly modified their headings. Chart 3, in particular, provided an arresting overview of the Claimants’ case that Autonomy used ‘pure’ hardware sales which were not disclosed and sales to VARs which had no real substance to give the appearance of meeting market expectations of revenue. Appendix 5 is a graphic depiction (again prepared by the Claimants) of the contribution of impugned VAR transactions to Autonomy’s revenue in each quarter of the Relevant Period. Appendix 6 comprises a Summary of Conclusions which I read and made available in court on 28 January 2022 (and see paragraph 4120 below). These Appendices are attached to this judgment at the end of Part B; but any reader may be assisted by reference to them at this juncture. I have also prepared a separate (and independently page and paragraph numbered) Schedule analysing in more detail each of the impugned VAR transactions: this too appears at the end of Part B.

As already indicated, Mr Hussain adopted Dr Lynch’s submissions. References to the Defendants should be taken to be references to each of them unless otherwise stated.

This is obviously an extremely long judgment. The factual detail provided to me in respect of each of the six main areas of the case is in my experience unprecedented. Whereas in many cases, only a few documents ever come to be relied on otherwise than in passing, here the relevant record and the number of documents cited was very considerable indeed. The process has involved a detailed examination, by reference to a myriad of transactions, of the way Autonomy’s business was carried on at a fairly granular level.

I am very conscious that the length, detail and complexity of this judgment, and its focus in parts on individual transactions to determine whether some or all of the means used to generate and /or accelerate revenue were acceptable commercial strategies or dishonest devices to cover shortfall in software revenues, may obscure the ultimate question in both the FSMA and the deceit and/or misrepresentation claims which is whether the acquirer was dishonestly misled. In particular, it is necessary to bear in mind that, except as regards the claims for transactional losses, the question is not whether the impugned transactions were wrongful, but whether their purpose (and effect) was to enable Autonomy to cover shortfalls in its revenue from software sales, so as to appear to meet market revenue expectations and (especially when combined with giving an appearance of growth in its valuable OEM and Cloud business lines) present itself as a larger and more successful company than in reality it was.

Nevertheless, careful and detailed analysis of the impugned transactions is necessary in providing the answers. It is worth remembering that Autonomy’s business appeared to most to be well run and growing fast because of the phenomenal capabilities and utility of its main product, called IDOL (its acronym for “Intelligent Data Operating Layer”). Autonomy became a FTSE 100 company on that basis. A few analysts were doubtful whether all was as it seemed; but most were not. The process of peeling back what was happening beneath the presentation is a tortuous one.

Furthermore, in respect of each of the six areas of business which the Claimants impugned, they asserted that there was a ‘pattern’ which revealed dishonesty. It might be tempting to assume that if a ‘pattern’ was revealed by, say five to 10 of the 37 impugned VAR transactions, that should be taken to demonstrate a pattern across the board. But such an assumption would, in my view, be unsafe. In short, though the phrase may have become over-used, the devil is in the detail.

I should acknowledge also that it has weighed heavily with me throughout that this judgment, whether logically or not, may affect Dr Lynch in other battles, on which his long-term freedom may depend. I have not wished to leave a stone unturned which might have yielded some different perspective or ‘pattern’.

I feel I should also emphasise, in light of the decision of the Court of Appeal in Bank of St Petersburg PJSC and others v Vitaly Arkhangelsky and others [2020] EWCA Civ 408, that (a) although of necessity I have divided up separate sections of the judgment, I have sought at all stages to reflect on how the parts, and especially my views as to the state of mind of the Defendants, might impact on the whole or more generally; and (b) I have throughout borne in mind and applied the standard of proof applicable in all civil proceedings, that of the balance of probabilities, whilst taking into account in determining the probabilities that other things being equal, it is a fair starting point that people do not usually act dishonestly (and I would add with particular reference to this case) especially when any dishonesty is almost bound quickly to be revealed. Any “infelicities” in my expression should not be taken to connote that I have departed from adjudicating the case on the basis of the balance of probabilities: I have not.

The Claimants and the Defendants adopted very different approaches in terms of the structure and approach of their respective written closing submissions. But all parties dealt separately with the six areas of claim I have identified in paragraphs 16(1) to 16(6) above, and this judgment does also. This (and the extreme detail) has further increased the need for a considerable amount of cross-referencing. I regret that this makes the judgment more unwieldy.

The judgment is divided into two main Parts (Part A and Part B), a division made necessary because of its length and website requirements. In addition to this Introduction and an analysis of relevant issues and the tests of liability under FSMA, this first Part (which ends at paragraph 2336) contains an Introduction describing the Claimants’ allegations in respect of Autonomy’s sales of hardware and the impugned VAR sales. With reference to the latter, and as mentioned above, I have also attached as a separate document a detailed Schedule analysing each of the impugned VAR transactions, which is also of considerable length. In Part B, I address the four remaining areas of the Claimants’ FSMA case (“the reciprocals transactions”, “the hosting case”, “the OEM case” and “the Other transactions”) and also the Claimants’ claims in deceit and/or misrepresentation, the Claimants’ various claims for direct loss, and issues relating to reliance and loss. I address Dr Lynch’s Counterclaim in paragraphs 4106 to 4115 below of Part B. I summarise my conclusions in paragraphs 4116 to 4135. Lastly, I have included a postscript in paragraphs 4116 to 4155. This judgment does not address issues of quantum. That will be dealt with in a separate judgment in due course.

An overview of the business of the two principal commercial entities, HP and Autonomy, and the way they were respectively structured and managed, is necessary in order to place in context the Claimants’ allegations (which in numerical terms relate to only a small part of Autonomy’s business but which impugn transactions of considerable value (Footnote: ⁵)). It is also necessary for an understanding of (a) the Claimants’ overall case that the Defendants, and especially Dr Lynch, were in a position to and did conceive and direct all the impugned transactions with a view to presenting Autonomy as far more successful than it was, and thereby dishonestly misled HP into paying far more than its true value and (b) the Defendants’ overall case that the reality is that HP’s management had no stomach for the transformational change that was the objective of the Acquisition, and contrived this extraordinary series of proceedings to shift the blame away from themselves.

Autonomy was founded in 1996. It was spun out from a company called Cambridge Neurodynamics, which was an early venture into using “machine learning” to develop software, techniques which Dr Lynch had explored in his PhD thesis at Cambridge University and his subsequent research fellowship in “adaptive pattern recognition”.

By the beginning of the Relevant Period, Autonomy had grown from a small start-up into a market leader in enterprise technology, especially in the field of unstructured data analysis. It went public in 1998, with an initial listing on the EASDAQ. It was admitted to the official list of the LSE in November 2000. It re-joined the FTSE 100 in 2008.

The Autonomy group was highly profitable, generated reported annual revenue in 2010 of $870 million, and (as an illustration) held cash reserves of $1.1 billion at the close of 2010 (though it is fair to qualify this by noting that this figure included proceeds of approximately $762 million raised by an issue of convertible loan notes in March 2010). Its customers included blue-chip companies in every sector. (Footnote: ⁶)In 2011, it was the UK’s largest software company based on market capitalisation.

Autonomy was headquartered in the UK but operated in global markets, with (by the time of the Acquisition) more than 2,500 employees, over 25,000 customers and operations in more than 19 countries (including the USA, where it had set up Autonomy Inc with its own management, sales and legal teams in San Francisco).

Its worldwide development and product spread was partly in consequence of successful acquisitions before the Relevant Period. Most significant were the following (all but the last of which were completed before the Relevant Period):

The acquisition of Verity Inc (its then nearest competitor in enterprise search technology) in December 2005. Verity also had US federal security clearance, which until a change of the rules in Autumn 2009 enabled Autonomy to conduct business with the US Federal Government.

The acquisition of Zantaz in July 2007, through which it acquired an archiving software solution, commonly sold under the name Digital Safe. Digital Safe enabled data to be processed and stored more efficiently, thus cutting down on storage costs and search times; and when integrated with Autonomy’s own technology (especially IDOL), the combined solution enabled data to be processed once into a consolidated archive and used with multiple software products, thus eliminating the need to store and search for data across multiple repositories. Especially at a time of increased competition and falling data storage rates (Footnote: ⁷), this efficiency gave Autonomy/Zantaz a commercial advantage which was the springboard for the development of hosting as a major part of Autonomy’s business. Also as part of the Zantaz acquisition, Autonomy acquired an e-Discovery or “EDD” product called Introspect, which it also offered as a hosted solution.

The acquisition in early 2009 of Interwoven Inc which specialised in the provision of enterprise content management software (including a solution called iManage), in e-commerce (Footnote: ⁸), and also had a large law firm customer base.

The acquisition of the Iron Mountain Digital business in May 2011, and through this a large archive product called LiveVault which handled large amounts of stored data and suited customers with especially large structured data storage requirements, and which could be integrated with a product called StorHouse which Autonomy acquired from a company called FileTek for further efficiencies (Footnote: ⁹).

Nevertheless, Autonomy’s success was principally based on its own market-leading core product called IDOL, an acronym for Intelligent Data Operating Layer. IDOL technology, focused on the analysis of unstructured data, was the core technology at the heart of Autonomy’s software.

Some explanation of this technology is appropriate, not least because (a) its extraordinary and world-beating capabilities tended not to be acknowledged by HP in their complaints about the Acquisition, and (b) it is important for an understanding of a central issue in the context of the hardware case, which is as to what the Defendants meant by the sale of an “appliance”. The following is based on Dr Lynch’s explanation in his first witness statement.

There are two types of data: structured and unstructured. Structured data is found in spreadsheets or in prescribed fields in a database. When data is entered into a database it is easily searchable. For example, as Dr Lynch explained, a computer will know that the data in an “address” field of a database is an address. Unstructured data is data that is not contained in prescribed fields. Most data is unstructured. Books, newspaper articles, websites, pictures and indeed, most forms of communication, comprise unstructured data.

Unstructured data is obviously much more difficult for computers to interpret and analyse. In 2009, the vast majority of computer software could only process structured information. It was Autonomy’s ability, using IDOL technology, to handle unstructured information that set it apart.

Dr Lynch explained that IDOL is an engine that can form a conceptual and contextual understanding of unstructured data using probabilistic theory based on the occurrence of certain word patterns, enabling the computer to infer meaning. (Footnote: ¹⁰)

Put shortly, using IDOL technology computers could make sense of unstructured data, and analyse and process raw information in the form of emails, voicemail, websites, telephone conversations and video recordings: a vast universe. It could analyse any digital document, speech or video, independent of its language, to identify and prioritise the main concepts within that document. IDOL users were enabled to automate a broad range of otherwise labour-intensive, iterative tasks, ranging from categorising information by subject matter, to inserting hypertext links to related material, to profiling users based on ideas in the text they read or wrote, to delivering information to those users most likely to be interested. Furthermore, IDOL was flexible and scalable: the fundamental capabilities of its core technology allowed it to be embedded in a wide range of applications.

Between 1996 and 2012, IDOL technology was continuously developed and refined by Autonomy’s team of software engineers. Dr Lynch (whose fondness for analogies became very evident in his 22 days in the witness box) explained that:

In addition to functions, the IDOL platform included (a) connectors, which enabled the IDOL platform to connect with many different systems, such as a bank’s trading system or a law firm’s document management system, and (b) interfaces, where a set of functions and connectors were put together to solve a problem (such as the Introspect e-Discovery interface which was a market-leading document processing, review and production application, used widely in the legal sector).

Autonomy’s corporate governance structure comprised (in 2010-2011) an experienced board of directors chaired by Mr Robert Webb QC (as non-executive chairman) and a three-man Audit Committee, with its external auditors being Deloitte.

Appointed Queen’s Counsel in 1988 and a Recorder of the Crown Court in 1993, Mr Webb had previously served as General Counsel at British Airways, and as at 2010 continued to serve as non-executive director of the London Stock Exchange, the BBC and two other companies.

In addition to Dr Lynch and Mr Hussain (the only executive members) the other board members were Mr Jonathan Bloomer (non-executive) who was also permanent chair of the Audit Committee, and who had been Group Chief Executive of Prudential Plc from March 2000 until May 2005, and before that senior partner in Arthur Andersen’s financial markets division; Mr Richard Gaunt (non-executive), who was the co-founder of Autonomy; Dr Frank Kelly FRS (non-executive), Professor of Mathematics of Systems in the Statistical Laboratory, University of Cambridge and Master of Christ’s College, Cambridge; and Mr John McMonigall (non-executive) a partner of Apax Partners Worldwide LLP and previously a member of the Management Board of BT.

Until September 2010, the members of the Audit Committee had been Mr Barry Ariko (a former senior executive at Oracle Corporation and of other software and hardware companies), Mr Richard Perle (a former US Assistant Secretary of Defense and a former director of various pharmaceutical and aerospace companies) and Mr McMonigall. There had not been a permanent Chair of the Audit Committee (so that its members served as Chair on an ad hoc basis) and none of the members was an accountant or had any formal accountancy training. Mr Bloomer told me when cross-examined that it would be normal to expect that the Chair would be “financially literate” but not necessarily that he or she would have an accountancy qualification (though he thought probably most did).

Mr Ariko and Mr Perle retired in September 2010 and from then until the Acquisition the Audit Committee comprised, in addition to its Chair (Mr Bloomer), Mr Kelly and Mr McMonigall. As noted later by Ernst & Young (“E&Y”) (HP’s auditor), both the Board and the Audit Committee had been expanded over the course of 2010 following analyst pressure, leading to what E&Y described as “the formalisation of the Audit Committee process and a more robust independent challenge to the board.”

Before every meeting members of the Audit Committee would receive a pack including, amongst other things, a draft press release, a quarterly results analysis by the finance department, and Deloitte’s report to the Audit Committee. Deloitte’s report was a substantial document – between 20 and 40 pages – identifying the key risks identified by Deloitte in the quarter. Prior to each Audit Committee meeting, Mr Bloomer would read these documents, and would discuss Deloitte’s findings at his meeting with the audit partner. The meeting of the Audit Committee itself would generally last around 2 hours.

Dr Lynch did not attend Audit Committee meetings. This accorded with good corporate governance. Deloitte did attend, and the Audit Committee could and did ask questions of them. A portion of every Audit Committee meeting took place without any executives present, to ensure that Deloitte had an opportunity to raise any concerns they may have had with management. This was an opportunity for completely frank discussion without any of the management team present. If there were any matters of importance, they would be escalated by Deloitte to the Audit Committee. Mr Lee Welham (“Mr Welham”), the senior manager on the Deloitte audit team from 2008 to August 2011, who had since 2005 been a member of the audit team (and was the only Deloitte witness the Claimants chose to call), never had a sense that there were any areas where the Audit Committee could not or would not go. Deloitte never raised any concerns about management with the Audit Committee.

Mr Hussain’s predicament should not obscure the fact he was highly regarded. He had been a winner of the London Stock Exchange CFO of the year award. The finance department beneath him was substantial and was also led by experienced individuals (see below).

Apart from the Defendants at the apex, the members of the management team were:

Mr Andrew Kanter (“Mr Kanter”), Autonomy’s Chief Operating Officer (or “COO”) and General Counsel throughout the Relevant Period (who provided a witness statement, which was withdrawn when Dr Lynch did not call him, see paragraph 426(4) below). He was mainly based in Cambridge. At the time of the trial and closing submissions Mr Kanter remained subject to continuing investigation by the US Department of Justice (“US DoJ”).

Dr Peter Menell, Autonomy’s Chief Technology Officer (or “CTO”) throughout the Relevant Period, who did not give evidence, see paragraph 428 below. He was based in Cambridge.

Ms Nicole Eagan (“Ms Eagan”), Autonomy’s Chief Marketing Officer throughout the Relevant Period, who like Mr Kanter provided a witness statement which was withdrawn when Dr Lynch did not call her (see paragraphs 426(4) and 1139 below). She was US-based. Like Mr Kanter, at the time of the trial and closing submissions she remained subject to continuing investigation by the US DoJ.

The above individuals were members of the “MRL Leadership” group, which was listed in an organisational chart dated August 2011. Another central figure was Mr Stephen Chamberlain (“Mr Chamberlain”). He was in the finance department, reporting to Mr Hussain: see paragraph 67 below. There were 26 senior individuals listed on that chart, which also shows their location in the various different global locations. (Footnote: ¹²) Thus, Autonomy’s management would often be in different locations to each other.

Autonomy Inc had its own management team, which consisted of:

Mr Christopher ‘Stouffer’ Egan (“Mr Egan”), who was its CEO throughout the Relevant Period, who was responsible for sales activities in North and South America. He gave evidence in Mr Hussain’s criminal trial in the US and provided a witness statement in these proceedings but declined to attend in the UK and was cross-examined over video link. He was based in San Francisco.

Mr Joel Scott (“Mr Scott”), who was its COO and General Counsel throughout the Relevant Period. He too was based in San Francisco. He gave evidence in the criminal trial in the US also, but none in these proceedings.

Mr James Crumbacher (“Mr Crumbacher”), who worked with Mr Scott in Autonomy Inc’s legal department.

Mr Michael Sullivan (“Mr Sullivan”), who was CEO of the Protect business, which was the archiving and litigation discovery division of Autonomy from 2009, carrying on business which before 2009 was carried on through Zantaz. Mr Sullivan had been Senior Vice president of Operations and Services of Zantaz, responsible for delivery of all products and services to customers, before its acquisition by Autonomy in 2007. After that acquisition, he became the CEO of Zantaz from 2007 (when it was acquired by Autonomy). He was based in Boston, Massachusetts. He gave oral evidence in the US criminal proceedings and a witness statement (which was admitted by a hearsay notice) in these, but declined to attend for cross-examination.

Mr Eloy Avila (“Mr Avila”), who was Chief Corporate Architect from September 2009 to May 2010, when he became CTO for the Americas. He became worldwide CTO in November 2010 and was based in San Francisco. He provided a witness statement for Dr Lynch and was cross-examined in these proceedings.

Mr Andrew Joiner (“Mr Joiner”), who was CEO of Autonomy’s Promote/eTalk business and was based in the US. He did not give evidence in these proceedings, and was not central to the matters in issue.

Mr Michael Mooney (“Mr Mooney”), who was a senior salesman based in Pleasanton. He did not give evidence.

Except for Mr Scott and Mr Crumbacher, all the individuals named in the preceding paragraph were members of the “MRL Leadership” group.

There were different sales teams for Autonomy’s various business areas and locations. Mr Egan was effectively head of sales in the US. Mr Egan’s reporting line was to Dr Lynch and Mr Hussain; but in practice he reported to Mr Hussain, whom he described in his witness statement as “de facto Head of Sales for the entire company”. Mr Sullivan headed up the Zantaz business.

Senior salesmen in the US included Mr Mooney (in California) and Mr Robert Sass (“Mr Sass”) (in New York). Autonomy’s federal business had its own sales team. There were different sales teams in the EMEA regions. The number of salesmen increased substantially in the period 2009-2011.

Autonomy’s sales function was run using the Sales Management System or “SMS”. This was a database of pipeline prospects. Salesmen were required to input details into SMS, detailing the opportunity, quotes, notes, call logs and the like. They were supposed to do so every day (although Mr Egan thought that was the hope rather than reality). There were weekly SMS sales calls which were broken up into multiple calls over several hours. US sales were the majority of Autonomy’s sales. The US part of the call was generally led by one of US management, such as Mr Egan or Mr Mooney. Very occasionally Mr Hussain might lead the call if none of US management was available. While Dr Lynch joined these calls in the very early days of the company, his evidence was that he very rarely did so in the Relevant Period.

Autonomy set quarterly revenue targets. Mr Egan accepted in cross-examination that these were an essential management tool. But they were also deployed in another way which is very central to this case. They were reported to the market quarterly and became the reference point for analysts and the market in assessing Autonomy’s performance.

It is clear from their use, from documentary evidence of analyst and investor outlook, and from emails from Mr Marc Geall (“Mr Geall”) in Autonomy’s Investor Relations department that both Autonomy’s success and to some extent its problem was that by mid-2009 Autonomy was being perceived in the market as a “beat and raise” growth stock.

A “beat and raise” stock was a well-known market expression connoting a company, often a technology company, with a track record or reputation of beating consensus estimates of revenue or earnings in one quarter and raising its own forward revenue or earnings forecasts for the next, and whose share price typically reacted according to its success or not in each quarter in that regard. In cross-examination, Dr Lynch sought to downplay this. He told me that although he did not think it “a point worth labouring”, and “if some people looked it that way, then fair enough”, he did not remember or consider Autonomy “being a paradigm of beating and raising through this period”. But the evidence was, and I find, that it was. Mr Hussain and Dr Lynch became obsessed with this performance marker, as will emerge.

That perhaps explains why the Defendants, who might have been expected to trumpet their successes, were surprisingly keen during the trial to emphasise that Autonomy did in fact miss its consensus revenue targets from time to time (five times in the period between 2009-2011). During the Relevant Period, they also sought to depict the increase of its overall revenue targets as the result of an increasing sales force, and launching new products continually, increasing market share, and operating in markets that were expanding rapidly. There was no mention ever of substantial sales of hardware separately from the provision of Autonomy software, or what the Claimants called “pure hardware sales”.

Autonomy had a substantial finance department, headed by Mr Hussain as CFO. His right-hand man was Mr Chamberlain, an ex-Deloitte Chartered Accountant. Mr Chamberlain, whom Mr Welham described as his “main point of contact throughout the audit and review processes” and the “first point of contact for many requests from the Deloitte audit team for information” did not give evidence in these proceedings. He has been indicted in US criminal proceedings, which I believe may now have commenced.

Mr Matthew Paul Stephan (“Mr Stephan”) and Ms Poppy Gustafsson (nee Prentis, and whom I refer to as “Ms Gustafsson”) and (from 2011 onwards) Ms Antonia Anderson (“Ms Anderson”) also worked in the department, reporting to Mr Chamberlain. All three were based in Cambridge: they were all ex-Deloitte Chartered Accountants. Mr Stephan had been part of Deloitte’s Autonomy audit team until December 2008 and joined Autonomy in March 2009, where he worked until early 2011. Ms Elizabeth “Lisa” Jane Harris (“Ms Harris”), another Chartered Accountant (ex-KPMG) also worked there. Mr Stephan and Ms Anderson gave evidence for the US Government in the US criminal proceedings, and Mr Stephan’s evidence was admitted into these proceedings by hearsay notice served by the Claimants.

Ms Gustafsson (who is now Chief Executive Officer of Darktrace, a cyber security company of which Dr Lynch was a founder and remains a substantial shareholder) and Ms Harris (who is referred to as a “co-conspirator” in the most recent form of the Indictment in respect of Dr Lynch in US criminal proceedings for her role in the alleged theft of confidential information uploaded to a USB device/pen drive which seemed to her to evidence knowledge on the part of HP of the fact of Autonomy’s hardware sales) gave evidence and each was cross-examined in these proceedings.

The finance team in Cambridge sat in an open plan office, with everyone within earshot of one another. Ms Harris explained that that team were able to share ideas and have discussions freely; and Mr Chamberlain would have discussions with Mr Hussain and with Deloitte at a table in this office, in earshot of everyone.

The UK finance team was responsible for all the accounting for the non-American countries and consolidation of all global accounts. There was also a team based in the Netherlands who did sales order processing for non-American sales. Accounting for Autonomy's American-based subsidiaries involved work by the US accounting teams, with consolidation taking place in the UK. There were several teams in the US from Autonomy Inc and legacy Autonomy acquisitions, including eTalk, Interwoven, Zantaz and Iron Mountain. All finance teams reported to Mr Chamberlain.

Autonomy’s auditors were Deloitte, a Big Four accounting firm. Deloitte’s work for Autonomy was handled through its Cambridge office. Deloitte spent approximately 12 weeks per year on-site in the Cambridge offices of Autonomy. Deloitte had unfettered access to Autonomy’s accounting records, as was clear in the evidence of Ms Gustafsson and Ms Harris referred to below.

Deloitte had day-to-day contact with members of Autonomy’s finance team in Cambridge, including Mr Chamberlain, Ms Harris, Ms Gustafsson, Mr Stephan and (from 2011) Ms Anderson. Deloitte also had a reasonable amount of contact with Mr Hussain, although he was not based in Cambridge full time. Deloitte could and did also ask questions of Autonomy’s technical staff (including Dr Menell, Mr Lucini, Mr Goodfellow, Mr Gallagher and Dr Blanchflower, as to whom see below), as well as the sales staff, such as Mr Egan.

Deloitte’s audit team numbered around 12 to 15 people. The audit team included:

A lead audit partner/engagement partner. This was Mr Richard Knights (“Mr Knights”) for 2009 and Mr Nigel Mercer (“Mr Mercer”) for 2010 and until August 2011. In the transition period between those two individuals in Q1 2010 Mr Knights and Mr Robertson were the engagement partners. The Defendants invited me to note that although Mr Welham gave evidence as to judgements that Deloitte might have reached in a counterfactual world, he was a senior manager at the time, and these would ultimately not have been his judgements but those of his superiors (as indeed Mr Welham accepted in cross-examination).

There were numerous accountants below the partner level. In the Deloitte hierarchy directors were below partners. Managers were below that. The professionals in the audit team included Mr Rob Knight (a director), Mr Welham (then a senior manager), Ms Anderson and Mr Murray (also managers). (Ms Anderson moved from Deloitte to Autonomy in early 2011.)

Deloitte’s engagement quality assurance review (“EQAR”) partners. These provided second partner review and consultation, and were able to meet and ask questions of the client.

Separate to the audit team there was an independent review partner (“IRP”) who reviewed the work of the audit team and the EQAR partners. The IRP carried out an objective evaluation of the significant judgements made by the engagement team and the conclusions reached in formulating the auditors’ report. The IRP could not meet or have any interaction with the client, so as to preserve their independence.

There was also a professional standards reviewer (“PSR”), who together with the IRP and EQAR also reviewed the audit partner’s work.

Deloitte had an internal technical IT expert, Mr Johnstone, who was an accountant but also an IT specialist.

Assistance on request (by either the engagement or EQAR partners) from Deloitte’s National Accounting and Auditing technical team (“NAA”). Mr Philip Barden (“Mr Barden”) was a partner in that team. He was a member of the Institute of Chartered Accountants’ financial reporting committee, a senior person within Deloitte and one of the senior authors of Deloitte’s 2009 guidance on IFRS.

Around six people from the Deloitte team worked on site in Cambridge during the audit process. The audit team would always sit in an office close to the finance team, and had free rein to go about and talk to anyone. Ms Harris’s unchallenged evidence was that they were entitled to see anything and everything they wanted in any of the files, and the “audit team would take away revenue and any other files of working papers and review them at their leisure”. In preparation for the audit or quarterly review, the finance team would put together bundles for every revenue contract in excess of $100k, including the contract, invoice, proof of delivery and payment history, together with the details of any accounting adjustments. If Deloitte had any questions regarding the wider context of a deal, its commercial rationale, or the features of the relevant products or services, they had ready access to Autonomy’s sales or technical teams. They appear to have become almost an adjunct of the finance department.

Ms Gustafsson had been at Deloitte, working on the Autonomy audit, before she joined Autonomy. She said in cross-examination that “the auditors were never isolated from the sort of general finance function and they were able to come and see all of the original documentation that I myself relied upon.”

As the audit team worked from the Autonomy offices, most requests were made in person, as the auditors could simply walk over to the desk of someone in the finance function, who could in most instances immediately pull the information required from Autonomy’s electronic systems.

Mr Welham agreed that the Deloitte team were of high calibre and properly trained and they acted independently, competently and with integrity.

As Chairman of the Audit Committee (from September 2010), Mr Bloomer met Mr Mercer, the audit partner, on a regular basis. In unchallenged evidence, he said that he “found Deloitte to be thorough and diligent in their audits”; their reports to the Audit Committee were generally more detailed than he had come across in other audit committee roles, and struck him as “detailed, open and direct”.

Deloitte undertook detailed work not only in relation to Autonomy’s annual accounts, but also in relation to its quarterly reporting. They also tested all of Autonomy’s revenue transactions over $1m, and a sample of smaller revenue transactions. This covered the bulk of the transactions impugned in these proceedings.

Deloitte’s audit work for the 2009 and 2010 annual audits was performed in accordance with the International Standards on Auditing. The general requirements of an audit were summarised at §5 of ISA 200:

The requirement that the auditor exercise professional judgement and maintain professional scepticism was explained at §7. Professional scepticism was defined as: “An attitude that includes a questioning mind, being alert to conditions which may indicate possible misstatement due to error or fraud, and a critical assessment of audit evidence.” The requirement to exercise professional scepticism was restated at §15 of the standards. The concept of professional scepticism was at the heart of the audit function, as Mr Welham agreed, and was drilled into the Deloitte team as part of their training.

Ms Gustafsson, speaking about her time at Deloitte, also stressed the importance of professional scepticism. She said in practice that this meant that Deloitte could not simply rely on information provided to it by the finance department in isolation, but would need to undertake its own information gathering process, checking contractual documents, making checks directly with third parties (suppliers, customers and technical experts), validating commercial justifications with technical teams, and seeing product demonstrations.

Mr Welham agreed that Deloitte did not simply rely on management representations. They looked for other audit evidence where possible. The auditors also had to be alive to anything that might suggest risk of misstatement or fraud, and to exercise a testing and enquiring frame of mind, not unthinkingly accepting what was put in front of them. Mr Welham considered that Deloitte’s audit team for Autonomy fully complied with these requirements. Deloitte’s audit team also approached the audits of Autonomy with the specific risks of fraud in mind, including risks of management being under pressure to meet various outcomes including financial targets.

None of the issues raised in this case was identified by Deloitte at the time as issues of fraud. Nor did Mr Welham consider at any stage that there had been any fraud. He saw nothing that caused him to be concerned that there was fraud at a senior management level as now alleged in this case. Mr Welham subsequently applied for the role as Autonomy CFO, after Mr Chamberlain’s departure in December 2011. This would have involved working with Mr Hussain. Mr Welham had no concerns when he applied for that job about the accuracy or integrity of Autonomy’s accounting functions. Nor did Mr Welham have any concerns about the ability or honesty of the technical people Deloitte were given access to, including Dr Menell, Mr Lucini and Dr Blanchflower.

Deloitte also confirmed to HP during due diligence that Deloitte did not have any disagreements with management regarding accounting policies or conclusions. (Footnote: ¹³)

Autonomy had Slaughter and May as its external lawyers. They were appointed as Autonomy's takeover defence counsel in November 2010 and provided advice on the acquisition process, including the NDA which set out Autonomy’s limited obligations to provide information to HP during the due diligence process. (Footnote: ¹⁴)

Inevitably, given the dispute as to Dr Lynch’s involvement in and knowledge of the impugned transactions and their accounting treatment, there was a dispute as to his role within the Autonomy group.

The role of a CEO may of course vary from company to company. But its principal focus usually is to direct the strategy of the company concerned, present and explain that strategy, and create and build the context in which that strategy may be achieved (including by using high level contacts), leaving it to line management to implement the strategy and report on its progress. It is not ordinarily the responsibility of a CEO in a larger enterprise to negotiate or execute transactions or account for them.

Dr Lynch maintained that his role in fact conformed to this usual template. Thus, according to his evidence and submissions:

Perhaps the only more unusual aspect of his role was that as what he called a “true technologist” he continued to play an active role in product development strategy. This was probably atypical in an enterprise software company (as perhaps illustrated by Mr Léo Apotheker’s (“Mr Apotheker”) more standard role when CEO of HP).

By the mid-2000s he was not generally involved in sales transactions, and had not been since Autonomy’s fledgling days. He did from time to time meet at a high level with the most senior people in the large customer institutions. But he did not deal with the procurement people or with the negotiation of the transactions themselves (as the negotiation and contractual documents confirm in the sense that his name is not on them).

From time to time (and especially towards quarter ends) Dr Lynch was sent lists by Mr Hussain of the large deals which were in the running, although their prospects fluctuated over time. (Footnote: ¹⁵) Mr Hussain provided summary information to Dr Lynch showing expected revenue, and progress towards revenue targets. Dr Lynch was keen to stress in his closing submissions that his interest in having Autonomy meet its revenue and earnings targets was unsurprising for the CEO of a listed company. In general terms, I accept that.

He was not involved in documenting the sales (or purchase) transactions. That was dealt with by the legal department, with some involvement from sales or from the finance department. The emails showed involvement from people such as Mr Scott, Mr Crumbacher and Mr Guiao (legal, in the US). Some involvement from Mr Chamberlain and Mr Stephan (finance, in Cambridge in the UK) is evident. The legal department also interacted with sales personnel such as Mr Egan, Mr Sass, Mr Mooney (all in the US) or more junior sales people. Dr Lynch was nowhere near this process.

Likewise, the accounting for sales transactions was undertaken by the finance department. Decisions on collectability, for example, were taken by them. Their work was closely scrutinised by Deloitte.

With respect to the VAR transactions complained of, Mr Egan had no communications with Dr Lynch about the details of the individual deals. Dr Lynch’s evidence was that he had no dealings with any of the resellers themselves.

Dr Lynch did have a role in purchases. For large purchases (which were not repeat business or sell throughs) he was generally one of the people approving. When he did give his approval, it was on the basis that he considered the purchase was a good idea for Autonomy. The Defendants made the point, which I accept, that if the purchase opportunity had been brought by a salesman such as Mr Egan, Dr Lynch’s views and motivations in respect of the transaction might not necessarily have been the same as those of the salesman. Moreover, Dr Lynch was not privy to the discussions between the negotiating counterparties; Mr Egan could not give evidence of any discussion or interaction with Dr Lynch in relation to impugned purchase deals that he negotiated (EDD and StorHouse). In each case, Dr Lynch contended that his approval reflected his understanding that the purchase was in Autonomy’s legitimate commercial interests.

Dr Lynch did not negotiate the purchases, and was not usually involved in the detail of them. He relied to a large extent on the judgement of others, such as Dr Menell, when approving the transaction. In particular, the technical evaluation of the products was led by Dr Menell, who had a team working under him. Dr Lynch was entitled to rely on their assessments as confirming that the purchase was a sound idea from a technical perspective.

In the latter context, while he is highly able technologically, Dr Lynch was not generally involved in the detail of the technical matters being addressed by the Autonomy technical staff.

Dr Lynch summarised the process for purchases, and his role in it, in his first witness statement:

Dr Lynch stressed that he did not involve himself in accounting for the purchase transactions which was, again, a matter for the finance department, scrutinised by Deloitte.

Dr Lynch did have a high-level understanding and approval of the strategy behind the hybrid sales in the hosting business. He identified and saw it as a good business move in Autonomy’s interest. He maintained that he was not generally involved in the individual hybrid transactions, whether new sales or restructurings, although from a revenue perspective he was made aware at a high level of the progress of some of the large deals that were being negotiated, and sometimes sought to use his contacts in higher levels of management to facilitate them. His evidence was that the detail and content of those deals was the province of the sales personnel, largely operating in the US, and that he was not involved in that nor in the accounting for them, which again was dealt with by the finance department, closely scrutinised by Deloitte.

Dr Lynch summarised his role, and his other commitments outside Autonomy, in his witness statement as follows (the detail of which was not challenged):

According to Ms Emily Margaret Orton (“Ms Orton”) (Footnote: ¹⁶), a Cambridge University graduate who worked as Dr Lynch’s assistant from early 2010 for around 18 months, Dr Lynch was primarily based in the London office (as was Ms Orton), and spent only about ten to twenty per cent of his time in the Cambridge office. When in London, Dr Lynch shared an office with Mr Hussain and Mr Kanter (though the latter was usually in Cambridge). In addition to meetings with external parties such as investors, bankers and journalists, and some non-executive responsibilities (including at the BBC), on which Ms Orton reckoned he spent some five to ten per cent of his time, Ms Orton estimated Dr Lynch spent 50% of his time on what she called “public facing matters”, rather than operational matters.

Dr Lynch accepted that, as CEO, he bore responsibility for Autonomy’s financial reporting. But he was not an accountant and was not generally involved in the detail of the accounting or the presentation of the financial reports. His case was that he was entitled to rely on his large and experienced finance function to do that, and was also entitled to take comfort from the scrutiny of Deloitte. He submitted that in this respect the position was no different to any other CEO of a large enterprise: this was illustrated in the cross-examination of Mr Apotheker who described his very similar role as CEO of SAP, in which he also relied on others:

This picture of him operating much as any CEO would ordinarily operate does not, however, capture the full extent of his role. Dr Lynch does appear to have concerned himself in certain contexts at a more operational level than would be typical of a CEO. It seems to me to be clear, for example, that his careful eye on whether Autonomy was going to meet its revenue targets in the last days of every quarter in the Relevant Period did result in him becoming aware of and sometimes involved in the progress of individual transactions to a greater extent and degree than I sensed that, to take the same example as previously, Mr Apotheker would likely have done.

In that connection, the Claimants placed considerable emphasis on a particular comment in a memo dated 23 January 2011 prepared by Deloitte which noted that:

Deloitte also noted in the same memo that Dr Lynch’s approval was required for all purchase orders over $30,000 within the group, denoting a granular approach untypical for a CEO in a large enterprise, and the Claimants pointed to email exchanges with Ms Orton suggesting that Dr Lynch was involved in approving other expenditure of amounts as low as £300.

These approval requirements of themselves do not demonstrate involvement in granular detail. Dr Lynch’s insistence on strict controls of expenditure was, more likely, a matter of corporate discipline: the need for approval is likely, and was probably intended, to ensure control of spending: but the details of the transactions probably seldom registered with him.

I would accept also that the context of the Deloitte memo has to be taken into account. Its accounting purpose was to establish that Autonomy, though it was a conglomeration of a number of acquisitions, could be treated as a company with a single Operating Segment subject to overall central control and direction by Dr Lynch as its Chief Operating Decision Maker (see IFRS 8), and was not stratified or compartmentalised in a way that precluded this. I agree with the Defendants that Deloitte were not saying that Dr Lynch was involved in all the details: indeed, they also noted that:

Nevertheless, the impression I have gained from the evidence I have seen and heard is that Dr Lynch exercised very personal overall control. He was very definitely and insistently at the apex of an unusual management structure in which the non-executive members of the board of directors (from 2010 onwards, Mr Robert Webb QC and Mr Bloomer) were his appointees. He was a very dominant personality. He expected to get his way, and did so. He was resourceful and determined; and he did not expect or tolerate doubts from others as to his chosen strategy, and he expected his strategies to be implemented.

IDOL was substantially Dr Lynch’s brainchild and Autonomy was, in his perception, his company. Having grown rapidly from being a start-up company, and made a number of acquisitions and launched many new products based on its innovative technology over a relatively short time, Autonomy seems to me to have been run with entrepreneurial disdain for bureaucracy, and with Dr Lynch in active and personal control of strategic direction.

There appeared to be no formal committee structure, and the board of directors was not the engine of decision-making. Indeed, it appears to have largely been side-lined, as the dearth of board minutes and the lack of any substantial sign of management direction illustrates (Footnote: ¹⁷).It would be consulted about major corporate acquisitions: but there is no record of it being substantially involved in anything else. None of the Autonomy group companies acted on the basis of resolutions passed by its board of directors. I stress in this connection that nothing in this judgment is intended to suggest that any of the non-executive directors were implicated in any dishonesty: the strategies developed and now alleged to have been dishonest were not theirs.

Dr Lynch ran the Autonomy group informally and through small cliques of loyal lieutenants within the “MRL Leadership Group”, namely Mr Hussain, Mr Kanter and Mr Chamberlain, with lesser input from Dr Menell and Ms Eagan, implementing decisions at the sales level, and through Mr Egan, Mr Scott and Mr Sullivan in the US. I describe them later as a cabal, of which Dr Lynch, when not ostensibly involved, was nevertheless the éminence grise. The impression I have formed is that his lieutenants would not have done anything of which they thought he might disapprove: and in practice, there was no higher authority within the group to whom they could look to justify what they did. As will be seen in relation to an episode which acquired some importance, when Autonomy’s CFO in the USA, Mr Hogenson, questioned certain VAR transactions and sought to refer the matter to the Audit Committee, Dr Lynch was well able to control the process.

One consequence of this personal control is that there are limited evidential records of the decision making with regard, say, to certain purchases made, or strategies being contemplated. From time to time, the Claimants sought to rely on this as suggestive of wrongdoing, or as casting doubt on the explanations that Dr Lynch has given for various decisions. The Defendants contended that that is simply the way the group was run. In my view, the absence of a written record in a company run in this way (as I accept that, in general, it was) is not, of itself, indicative of impropriety or desire to leave no record. Something more has to be demonstrated in the particular context. Put another way, the vast bulk of its business is not impugned but was run no differently, and the suggestion that the way of running the business of itself suggests impropriety is untenable accordingly. I would add also that some aspects of the decision-making process were well, even if informally, documented. For example, the wealth of email exchanges illuminate both the hardware and the VAR transactions, and Mr Hussain’s end of quarter schedules reporting with increasing focus on any shortfalls in expected software revenue are also illuminating, even if Dr Lynch professed to finding them hard to follow.

However, whereas in a more typically run large company, dishonest corporate strategy (as distinct from individual dishonesty) is unlikely because dishonesty is unusual and a board of directors unlikely to be unanimous in its pursuit, in a closely run company it is much more feasible, and much less unlikely. The fact of overall control is also, obviously, relevant to the issue of the Defendants’ knowledge.

The emails also show that to a large extent Dr Lynch was happy to allow those under him to get on with their roles. The emails also show that it was relatively rare for Dr Lynch to involve himself in the detail of the company’s day-to-day activities. When he did become involved this was usually because there was a strategic angle, such as a customer sensitivity.

Mr Webb QC gave unchallenged evidence on his own impression of Autonomy’s culture. As Chairman, he visited the office from time to time and asked staff about their role and what it was like to work at the company. He said in his witness statement that:

This section provides an overview of Autonomy’s approach to its reporting, both quarterly and annual. Some of the matters discussed in this section will be considered in further detail when dealing with the impugned transactions later in this judgment.

In addition to its annual accounts Autonomy reported on a quarterly basis. All Autonomy’s reports, both annual and quarterly, were subject to detailed review by Deloitte.

Deloitte reviewed and approved the financial statements in the annual accounts.

Deloitte also reviewed the narrative portion (the front end) of the reports. When reviewing the front end, Deloitte was required to identify:

whether any of the information in the front half was materially inconsistent with the financial information; and

whether there was any matter which came to their attention which caused them to believe that the narrative portion appeared to include a material misstatement of fact.

Deloitte’s review of the front end followed the ISA issued by the Auditing Practices Board in April 2006 entitled “Other information in documents containing audited financial statements”. This required the auditor to read the front end of the report to identify material inconsistencies with the audited financial statements or material misstatements of fact. If any were identified, the auditor was required to resolve them, and if an amendment was necessary and the entity refused to make the amendment, the auditor was required to express a qualified or adverse opinion. See §§2, 11, 12 and 14 of the ISA. At no stage did Deloitte ever identify (other than internally) any material inconsistencies with the financial statements or any material misstatements of fact.

Mr Welham confirmed in cross-examination that if Deloitte had seen something in the narrative portion of either a press release or an annual report that they considered was misleading in any way they would have said something.

The consequence of this is that many of the Claimants’ criticisms in these proceedings relate to presentations which were considered and approved by Deloitte. By way of example, the Claimants criticised the disclosures made and explanations given in the 2010 accounts. However, that document was reviewed by Deloitte in detail prior to sign off:

A copy of the 2010 report is marked by Mr Murray (manager from Deloitte), on 22 February 2011: “Reviewed throughout prior to sign off”.

That document is replete with tick marks showing Deloitte’s review and approval of the statements made in the 2010 accounts.

This extended to the front end as well as the financial statements themselves. For example, and as developed later, the working papers in evidence show Deloitte’s review of the data given for various non-IFRS metrics provided under categories now targeted by the Claimants as misleading and another means whereby to conceal the hardware sales: IDOL Product, IDOL Cloud, and IDOL OEM. Deloitte’s review involved an analysis of spreadsheets provided by management. Their review involved qualitative as well as quantitative analysis. (Footnote: ¹⁸)

Mr Welham confirmed that this ticking exercise was part of the normal process. A similar exercise was undertaken for the quarterly reports, as discussed below.

Deloitte’s review of quarterly reports

Formally Deloitte’s work on the quarterly reports constituted a voluntary review. However, as Mr Welham explained at §20 of his witness statement, Deloitte’s quarterly review work was substantially similar to an interim (i.e. H1) review. Deloitte undertook to conduct their reviews with an attitude of professional scepticism and:

Mr Welham also explained that in practice Deloitte’s work went beyond what was required for the purposes of a quarterly or half-year review, and Deloitte used the opportunity to undertake audit work in relation to Autonomy’s large transactions, and a sample of smaller ones:

The process of a quarterly review took around six weeks from start to finish, including the planning stage. (Footnote: ¹⁹)

Following the preparation of their review, Deloitte would prepare a report for the Audit Committee. That was the Audit Committee’s principal, and often in practice only, source of detailed information apart from public accounts.

Deloitte’s review of revenue

When conducting its review work, one of the audit risks that Deloitte had in mind was that revenue might be incorrectly recognised. According to their defence in regulatory proceedings brought against them, which Mr Welham stood by, Deloitte considered that they went beyond what was required of them by testing all licence transactions over $1m and a sample of smaller transactions below that level.

Deloitte’s revenue testing process involved the audit team specifically considering each of the IAS 18.14 revenue recognition criteria for each transaction reviewed.

Press releases and earnings calls

Deloitte reviewed Autonomy’s press releases. If Deloitte saw something in the narrative portion of either a press release or any annual report which was inconsistent with the financial statements or misleading in any way, Deloitte would have said something.

Following each quarter, the finance department would prepare a draft press release summarising and commenting on the quarterly results. Deloitte would check through the press release including the non-financial information and the non-IFRS metrics, and would “ tick them back ”. The press release would then be provided to the Audit Committee which would consider and then approve the release.

An example of this ticking back process is a draft of the Q3 2009 press release. Deloitte ticked off all the financial information given in the document, including (as Mr Welham confirmed) the non-IFRS supplemental metrics which Autonomy chose to provide.

Deloitte also listened to Autonomy’s earnings calls, and Mr Welham accepted that he often listened in himself. The documents show that someone from Deloitte listened in to the earnings call for each of the calls in the Relevant Period; Mr Welham listened in to almost all of them. The calls that Mr Welham listened to included the Q3 2009 earnings call. If Mr Welham had heard anything that he thought was misleading he would have raised it with his audit partner in the first instance. He did not, and no one else within Deloitte raised any concern arising out of any of the earnings calls, to Mr Welham’s knowledge.

The relevance of Deloitte’s review and approval

The Defendants placed much emphasis on Deloitte’s review work in meeting allegations in this case. Dr Lynch’s case, in very broad terms, is first that Autonomy’s accounting was fair and reasonable, and the various accounting judgements and decisions inherently involved fell within the reasonable range that was available; secondly that he was unaware of and did not participate in any false accounting, and was entitled to believe, and did believe that the accounting was fair and reasonable; and thirdly, that although the narrative portion or ‘front end’ of the accounts was not audited, Deloitte were satisfied with both its factual accuracy and its consistency with the accounting information provided in the ‘back end’ of the accounts (which, of course, was audited). Mr Hussain’s case was to the same effect; and in relation to the ‘front end’ of the accounts, his written closing submissions also relied specifically on the evidence of Mr Welham in cross-examination that Deloitte were satisfied that the narrative portion of the various quarterly reports was not misleading.

The Claimants’ answers to this were that the Defendants could not rely on Deloitte in circumstances where Deloitte had been misled; nor where they knew Deloitte to be wrong for whatever reason. If, as the Claimants contended was the case, the Defendants knew that the published information provided was deficient and/or inaccurate, they could not rely on Deloitte’s mistaken imprimatur. They acknowledged that if Deloitte were not misled it would be more difficult for the Claimants to prove that the Defendants nevertheless knew them to be mistaken: but on the facts, they contended, that was proved.

An important aspect of Autonomy’s accounting, and one which (especially in the context of the hardware claims) had, in my view, a marked effect on the approach of both Deloitte and the Audit Committee, was that all its business was treated as a single operating segment for the purposes of IFRS 8. The Claimants did not dispute this.

IFRS 8 (Operating Segments) was a new provision in the “Relevant Period”. It was applicable for accounting periods beginning on or after 1 January 2009 and was therefore applicable to Autonomy’s 2009 (and subsequent) accounts.

In January 2010, Deloitte’s audit team consulted with Mr Barden of Deloitte’s NAA in response to a (generic) call by the FRRP encouraging Boards of Directors to test their initial conclusions about their segmental reporting under the (then) new standard. The consultation involved a detailed review of the conclusions that Autonomy’s “chief operating decision-maker” was Dr Lynch and that Autonomy had a single operating segment for the purposes of IFRS 8. Deloitte and Mr Barden for the NAA confirmed those conclusions. The conclusions were re-confirmed on further review by Deloitte in November 2011.

I shall return to IFRS 8, and to the various further paragraphs within the Standard (and especially IFRS 8.32) when addressing the hardware case (to which it is most relevant).

The following synopsis of Deloitte’s approach is intended to illustrate Deloitte’s perception of the way Autonomy was managed (and of Dr Lynch’s role in particular) as well as the rationale of the decision to confirm that Autonomy had a single Operating Segment:

In January 2010, Deloitte concluded that Dr Lynch was the “Chief Operating Decision Maker” (which describes a function to allocate resources to and assess the performance of the operating segments of an entity). The basis for this conclusion was that they considered that he had an “unusual level of control” and that it was:

Deloitte accepted Dr Lynch and management’s central assertion, which was that IDOL technology was at the heart of all the Autonomy group’s products, and that it was this technology’s ability “to extract meaning from unstructured information which allows Autonomy to grow at such a fast pace.” They accepted that IDOL was at the core of the entire business, and that this feature of being a “one technology business”, together with Dr Lynch’s highly centralised management of the entire group which was not based on divisions or disaggregated data relating to the business activities of the group, distinguished Autonomy from all other businesses in the FTSE 100.

Deloitte’s memo of 23 January 2011 (referred to previously at paragraph 94 above) revisited this conclusion in light of various changes including acquisitions and (of particular note):

The January 2011 memo also identified three new categories of software offerings which might affect the previous analysis, being (a) “Power” (described as being “an infrastructure platform for understanding human-friendly data” ; (b) “Protect” (described as being a “selection of legal, regulatory and compliance solutions” ); and (c) “Promote” (described as being “a number of multichannel, customer interaction and revenue optimisation solutions”).

The memo noted that whilst this might suggest to a reader of the financial statements and the Autonomy website that there could be three distinct segments of the business, nevertheless all:

Deloitte noted as further support for the analysis and their conclusion that Autonomy continued to have a single Operating Segment that:

Revenue was not broken down between or into products, divisions or brands;

Cash flow forecasts and revenue figures were all undertaken at a group level, which:

Functions within the business such as sales and marketing and R&D were managed centrally, each by a single member of the senior executive;

The overall organisational hierarchy was very flat, with few middle managers;

All purchases over $30,000 had to be approved by the CEO, very few transactions were processed:

Of particular interest to the question of disclosure:

Deloitte’s review also covered §32 of IFRS 8 as it applied to Autonomy. IFRS 8.32 is part of the guidance dealing with “entity-wide disclosures” at §§31-34. The effect of these was to require separate reporting for certain revenues from products, services, geographical areas or customers, even where an entity had a single operating segment. As part of their review work Deloitte specifically considered whether notwithstanding that Autonomy had a single operating segment it was necessary to make separate disclosure of any products under IFRS 8.32. Deloitte considered the question in detail for the 2009 annual accounts (and revisited their conclusion for the 2010 accounts). Their conclusion, approved by Mr Barden, was that Autonomy’s approach of not making separate disclosure of different products, or groups of products, was reasonable. (Footnote: ²⁰)

Deloitte revisited this conclusion in 2010. As part of this consideration Deloitte specifically considered the strategic hardware sales in that year. Deloitte knew that hardware sales were in the region of 12% of Autonomy’s group revenues for 2010. Deloitte concluded, again, that there was only one operating segment, and did not require any separate disclosure under §32 of IFRS 8.

Thus , in summary, Deloitte’s analysis and the conclusions reached in respect of IFRS 8 confirm that Autonomy was run in a centralised and in some ways idiosyncratic way, with Dr Lynch very much at the apex of the management structure, and the Board of Directors and the Audit Committee being more spectators than active participants.

The last topic in this introduction to Autonomy and its business which I need to address before turning to describe HP in more detail is the Claimants’ contention that (a) by October 2010 Autonomy’s financial performance and share price had slipped so much that (i) it was likely to become a takeover target and (ii) Dr Lynch was put in mind to promote its sale by Mr Frank Quattrone and his company Qatalyst Partners (a boutique investment bank well known for its success in achieving high bid prices (Footnote: ²¹) ); and that thereafter, (b) following meetings between Dr Lynch and Mr Quattrone in late November/early December 2010, Mr Quattrone began approaching the “acquiror universe” he had identified (including Cisco, Oracle and HP) and (c) in January 2011 approached HP via HP’s then Chairman, Mr Raymond Lane (“Mr Lane”), to encourage interest.

The Claimants contended, in effect, that Autonomy was fattened up and marketed for sale. Dr Lynch, however, consistently insisted that he did not want a sale of Autonomy, and would resist a bid if made unless at a price which made resistance futile because it was at a premium that shareholders would find irresistible.

The following summary shows how Dr Lynch moved or was persuaded towards a sale:

On 6 October 2010, Autonomy released its trading statement ahead of its Q3 2010 results. Although the statement highlighted “record third-quarter 2010 revenues” , in fact, it slightly missed consensus. (Footnote: ²²) Further, the statement included a forecast that Autonomy expected “to review our internal model for the full year with a revenue reduction of around 3%” , implying full year revenue growth of 17%. The market had been expecting 21% growth: Autonomy shares promptly fell by 16% and never fully recovered. The headline in the Financial Times on the day of the trading statement read: “ Autonomy slides on reduced revenue forecast” and the headline the following day (7 October 2010) in the Guardian read “Autonomy shares plunge on fears of weak US orders”.

Almost immediately after this Mr Quattrone sent an email to Dr Lynch dated 6 October 2010 headed “Checking in” in which he wrote:

When Dr Lynch asked who would “come knocking” Mr Quattrone identified the most likely candidates as “Cisco, Oracle, IBM, HP (more likely with Leo and Ray Lane)” together with some less likely “wild cards” .

Dr Lynch said that it would all come down to whether someone could offer “a level the shareholders can’t resist” . In that context, he asked Mr Quattrone “what kind of prices do you think they can get to?” ; to which Mr Quattrone replied on 8 October 2010 that “top buyers could pay at least 70pct premium with no dilution even before considering any synergies”.

After receiving some further marketing material from Mr Quattrone, on 10 October 2010 Dr Lynch sent an email to Mr Quattrone, now with the heading “CONFIDENTIAL” , as follows:

This was of course good news for Mr Quattrone, and Autonomy engaged Qatalyst’s services to obtain such a price. Although terms were not finally signed off until 16 August 2011 it seems clear from the available documentation (including exchanges between Mr Kanter and Qatalyst) that Qatalyst’s engagement had been informally agreed by late 2010/early 2011.

Mr Quattrone began approaching the “acquiror universe” he had identified. He arranged, and attended, meetings between Dr Lynch and key M&A contacts identified in his October analysis, including contacts at Cisco, Oracle, and Mr Shane Robison (“Mr Robison”) and Mr Johnson of HP. Meetings were also pursued with some less likely potential acquirers identified by Qatalyst, such as Intel, Adobe and Dell. Other than HP, there seems to be no evidence that any of these potential buyers pursued Autonomy as a potential acquisition (Footnote: ²³) .

Qatalyst approached HP on 25 January 2011 by Mr Quattrone sending an email to Mr Lane asking to “discuss… confidentially” a “specific [situation] where we have some unique insight that things may be changing” . The next day, Mr Quattrone emailed Mr Lane slide decks titled “Autonomy Overview” (“the January Slides”) and “Autonomy Trading and Financial Statistics”.

Shortly after sending HP the January Slides, Mr Quattrone spoke to Mr Robison and the two agreed to a “fact-finding discussion.” Thereafter, Autonomy gave a presentation accompanied by further slides (“the February Slides”) which updated the January Slides to take account of the 2010 financial year end accounts which trumpeted “record full year revenues of $870 million, up 18% from 2009”.

On 4 March 2011, there was a further meeting between HP and Autonomy at HP’s headquarters in Palo Alto. Dr Lynch attended by video link whilst Mr Hussain and Mr Quattrone attended in person. A further slide deck was presented (the “March Slides”). By then HP and Autonomy had entered into a nondisclosure agreement. I discuss later how HP’s focus gradually narrowed to Autonomy as its bid target: see paragraphs 153 to 157 below.

It seems to me that this short chronological summary demonstrates that Dr Lynch, sometime in October 2010, began actively to contemplate a sale, and through Mr Quattrone, to sound out the “acquiror universe” and the possibility of a cash bid at a premium price (of say £26) in light of Mr Quattrone’s advice that a very substantial premium might be payable despite the reverses in the previous quarter. In my view, Dr Lynch’s insistence in his evidence that he made it clear to Mr Quattrone that “Autonomy was not for sale” smacks of protesting too much. I agree with the Claimants that the suggestion that the aim of Qatalyst’s approaches was to line up “white knights” to ride to Autonomy’s rescue if an unwanted suitor bid first is implausible.

I have taken into account the evidence of what the Defendants presented as a different perception held by Mr Robert Webb QC (“Mr Webb QC”), Autonomy’s non-executive Chairman at the time. Mr Webb QC’s evidence in his witness statement, on which he was not cross-examined, was that:

I have no reason to doubt that that was Mr Webb’s perception. I should also make clear that I do not think sale was really on Dr Lynch’s mind before October 2010. But neither that nor Mr Webb’s perception precludes the possibility that Dr Lynch, once he understood from Mr Quattrone the interest that might be generated, also understood that shareholders (of whom he was one, and a major one at that) would ultimately vote with their wallets, as indeed Mr Webb himself there identified. In my view the truth is, to quote Dr Lynch’s own words, that the strategy became that of “coaxing the right buyer”. I turn to describe the buyer that was found, HP, in more detail.

HP was an icon from the first days of the tech revolution in Silicon Valley, Northern California. It too had been a start-up, first housed in a garage which was the birthplace of ‘Silicon Valley’. HP had become a household name, some of its products having become almost synonymous with their function.

By the time of the Acquisition, however, HP was in the doldrums. Its heritage was in the hardware sector. This had become a highly competitive and low margin business. In the new computer world, software sales are the source of far greater profitability than hardware and the Cloud is King. HP had a Software division of considerable size; but it was lagging behind market leaders.

Furthermore, HP (and by that I include the group) had developed into a bureaucratic inward-thinking institution, which had lost its way and its confidence. Business units were siloed and complex, with infighting and little cooperation between them and poor systems of overall management and strategic direction, as Ms Whitman acknowledged in cross-examination.

Appreciating the need for change, HP had instigated ‘Project Cielo’ in October 2010 to assess its overall strategy. This was led by HP’s Strategy and Corporate Development (“SCD”) group (Footnote: ²⁴) and Bain & Company (“Bain & Co”), a well-known strategy consulting firm.

Bain & Co identified multiple concerns about HP as at November 2010. These included:

Mr Apotheker’s first task as HP’s CEO, which he embarked on even before his formal appointment, was to become engaged in Project Cielo. He quickly agreed with Bain & Co’s diagnosis, and its advice identifying what he also agreed were “real problems for HP at that time”.

The view Mr Apotheker formed, in line with that advice and as a principal conclusion of Project Cielo , was that HP needed to use its size, reputation and financial strength to evolve from being a producer of low-margin products and services to become (in his words):

Further, within the software ‘space’, Mr Apotheker’s view was that HP’s focus should be on “analytics for big data – which is the combination of structured data…and the much faster growing unstructured data set” : as explained at paragraph 42 above “structured data” being data maintained in a traditional database, and “unstructured data” being all other data including emails, word-processing, spreadsheet documents, video and sound recordings. This was, Mr Apotheker agreed in cross-examination, a “potentially huge market”.

As Mr Apotheker put it:

The dominant technology for data at this particular moment in time was to store data in what is called relational databases and for sake of simplicity, consider those to be a Excel spreadsheet, sometimes of monumental size. Unstructured data is everything else, everything covering voice, pictures and what-have-you, everything that you could not normally store in an Excel spreadsheet so to speak. Trying to combine the two would be a natural thing to do because that's how we all live and work and that's what is meant by the integration of these two things.”

Thus, Mr Apotheker envisaged HP becoming a one-stop shop for data. Operating across the entire IT stack would enable HP very considerably to increase its overall operating margins. Bain & Co’s figures suggested that operating margins should increase from 10 to 15% and achieve 10% CAGR (Compounded Annual Growth Rate) revenue growth. The then recent acquisition of Vertica Systems Inc (“Vertica”), a structured data company that HP had acquired in early 2011, gave further shape to Mr Apotheker’s plans for a “complete stack” with a combination of structured and unstructured data.

Unstructured data (sometimes referred to as “Big data” ) formed one of the core areas of Mr Apotheker’s plan. Bain & Co also identified “ enhancing portfolio for Enterprise Information Management ” as one of the primary “ potential areas of improvement” , in line with Mr Apotheker’s view regarding an integrated “stack” of layers of technology providing a complete service for enterprise customers.

During 2011 Mr Apotheker and Mr Robison, the chief strategy and technology officer and head of HP’s SCD group, identified and analysed a number of possible software targets with a view to assembling the complete technology “stack”.

Mr Apotheker and Mr Robison initially focused on the possibility of acquiring TIBCO, which was a specialist in “middleware” (Footnote: ²⁵) . A presentation to HP’s finance and investment committee, and board, in relation to TIBCO (code named “Tacoma”) showed that TIBCO’s margins were regarded as being “strong”, based on it having 30% operating margins by full year 2015 (current margins being around 25%, as per p.5 of the document). TIBCO’s gross margins (actual and predicted) were around 76-77%. HP’s discounted cash flow (“DCF”) analysis gave a standalone enterprise value of $4.46bn, and a value with synergies of $8.5bn. (Footnote: ²⁶)

The board authorised Mr Apotheker to pay up to $36 per share for TIBCO. This would have valued the company at about $6.6bn. It would have involved HP ceding approximately 53% of its anticipated synergy value. In the event HP offered $31 per share, valuing the company at about $5.7bn, approximately $1.2bn over HP’s assessment of TIBCO’s standalone value, and 30% of the total synergy value. (Footnote: ²⁷) HP subsequently raised its offer (as Mr Apotheker recalled it, to a price of $33 or $34 per share), but could not reach agreement.

Having walked away from TIBCO in early April 2011, Mr Apotheker asked the SCD group to increase its focus on other potential targets, including Autonomy and another entity, Software AG. Mr Apotheker confirmed, however, that these were regarded then as alternatives. It was not at the time envisaged that HP could acquire both at the same time.

Autonomy had for some time been recognised by HP as one of the market leaders in unstructured data. In a part of a Project Cielo presentation dated 13 November 2010, entitled “HP Play: Technology Feasibility” the highest end of an analysis of what were termed the “key technology pain points by criticality” identified the fact that there were “limited insights from unstructured data” . The document envisaged that the gaps would be closed by acquisitions in “BI/Analytics (e.g. Autonomy, Endeca) and middleware (e.g. Tibco)”. This objective was put at high feasibility albeit “based on assumption that target companies indicated can be successfully acquired”.

In May 2011 HP’s strategy group developed some preliminary analyses of Autonomy and Software AG. These were presented both to the finance and investment committee of HP’s board and to the technology committee. The presentation to the finance and investment committee was on 25 May 2011. That committee’s key job was to review acquisitions and asset allocation.

According to their analysis of Autonomy (code named “Atlantis”):

Autonomy provided a “ broad range of capabilities across all layers of the Enterprise Performance System stack, focused on unstructured content ”. These capabilities supplemented HP’s own in all layers of the enterprise “stack”. Autonomy was described in the executive summary as “ Anchor asset to secure leading position in the Enterprise Information Management segment ”. The executive summary stated, under “strategic”:

The penultimate point regarding bringing structured and unstructured information across business processes chimed with a major part of Mr Apotheker’s strategic thinking.

As the “Financial” section of the executive summary explained, Autonomy was considered to have “ strong growth and margins ”. HP came up with a stand-alone DCF valuation of $9.574bn based on 10.6% revenue CAGR and 46% operating margin by full year 2016 (i.e. higher operating margins than for the earlier TIBCO analysis).

The presentation set out the synergy assumptions arrived at by HP’s SCD group. Mr Apotheker reviewed them and considered them reasonable.

A DCF analysis was also provided. Adding the various categories of synergy to the $9.5bn stand-alone valuation gave a stand-alone plus synergies value at that stage of $15.223bn. The terminal growth rate assumed in the stand-alone value was 4%.

The presentation gave information about the current and predicted scale of Autonomy’s markets, as well as market share. This information came from IDC and Gartner, respected independent industry sources.

The presentation also identified “ other potential interested parties ” at p.28. These potential bidders included Oracle.

There had been some contact between HP and Autonomy prior to April 2011. This included initial contact from Mr Quattrone of Qatalyst, a video-conference in February 2011 and a further meeting in early March. During this period HP were provided with slides (the “January Slides”, the “February Slides” and the “March Slides”), which (as part of their misrepresentation case) the Claimants asserted were misleading (and see paragraphs 3831 to 3872 below).

From April 2011 onwards, Mr Apotheker and Mr Robison became focused on the acquisition of Autonomy (Footnote: ²⁸) . Mr Apotheker confirmed that he did not personally really focus on Autonomy as an acquisition target until April 2011, and after the attempted acquisition of TIBCO had failed.

Before describing in more detail the development of HP’s interest in and eventual offer for Autonomy, a short description of each of the individuals who were principally involved in the identification of Autonomy as a target, the development of plans for the Acquisition, the negotiation of an offer and/or the due diligence exercise undertaken by KPMG on the instructions of HP may assist an understanding of this section.

Mr Apotheker became President and Chief Executive Officer of HP in November 2010. He had a strong background in the software sector. He was head-hunted from a company called SAP, one of the largest software companies in the world at the time, where he had (latterly) been CEO. He had a track-record of successful acquisitions for SAP, having led more than a dozen, including SAP’s successful $7 billion purchase of a company called Business Objects in 2007.

Before his appointment by HP he was asked by HP’s Search Committee to prepare a paper that would provide an outsider’s perspective of HP. When at SAP he had been the executive in charge of SAP’s relationship with HP and had been an admiring observer for many years. His overall view, as expressed in his paper, was that although HP was financially sound, rapid changes in the technology industry represented a significant danger to HP, unless the company could develop and articulate an overarching growth and profit strategy, and rediscover its culture of innovation which had inspired its founders.

In Mr Apotheker’s assessment:

Personal computer sales (which were the main business of HP’s “Personal Systems Group” or “PSG”) had become a commoditized, low margin business: PSG should focus on the provision of mobile devices (such as smartphones, tablets and notebooks).

HP’s Imaging and Printing Group which was responsible for producing printers and inks, needed to adapt to deal with the possibility that printing of documents would become less common.

The most important issues facing HP, however, were in its Enterprise Business, which provided solutions to businesses and enterprises. His prescription was that HP should evolve this business from low-margin products and services to become what he called in his witness statement “a full services and solutions partner for businesses providing the essential/strategic parts of the technology ‘stack’” . (The ‘stack’, he explained, comprises “the hardware, software, and network layers that stand between the basic infrastructure of data creation, storage and distribution and the end-user of information.” ) An important element would be to expand HP’s software offering to make it a more prominent part of HP’s technology ‘stack’.

Mr Apotheker was appointed on this ‘ticket’. He was offered the role of Chairman also, but he refused this (preferring the European model of keeping the position separate) and recommended that Mr Lane be appointed to that position, as he was (Footnote: ²⁹) . Their appointments were announced on the same day, though in fact Mr Apotheker began work before then.

Mr Apotheker gave evidence in these proceedings, and was cross-examined for two days. He came across to me as honest and professional, though his witness statement was over-lawyered. He was palpably hurt by his treatment by HP, which terminated his employment less than a year after his appointment in circumstances I elaborate later. His oral evidence was conspicuously moderate in tone, but he nevertheless communicated to me his experience of HP as an “intensely political organisation” with inter-departmental rivalries, set in its ways and difficult to change.

Mr Shane Robison was Chief Strategy & Technology Officer of HP and head of HP’s SCD group from mid-2002 until November 2011. Before that he had been Chief Technology Officer at Compaq (which was acquired by HP in 2002). Mr Robison agreed with and supported the strategy to move further into enterprise software.

As head of SCD, and consistently with Mr Apotheker’s prescription for HP, Mr Robison spent substantial time through 2011 analysing (with the assistance of Bain & Co) the software market in general and, in particular, the businesses and companies that had sufficient scale and profitability to be considered a potential strategic fit for HP. Mr Robison was Mr Apotheker’s principal assistant in identifying and analysing potential acquisition targets, and they worked closely together in all the preliminary work culminating in the Acquisition (which he strongly supported). He was also involved in direct discussions and negotiations with Dr Lynch. He was selected by Mr Apotheker to be one of the three persons (the others being Mr Apotheker and Dr Lynch) who were to be joint heads of an integration committee which was to be established to steer the Acquisition and realise its objectives and the synergies that were its primary rationale.

Mr Robison provided a witness statement which was admitted as hearsay, but, on the basis of medical problems, he did not give evidence otherwise. His witness statement also bore hallmarks of being over-lawyered, especially when stating what his reaction would have been to various hypotheses put to him.

Ms Catherine Anne Lesjak (“Ms Lesjak”) had been appointed as an interim or caretaker CEO of HP for three months during the search for a replacement for Mr Hurd who had been removed in connection with allegations about expenses. At the time of the Acquisition she was Executive Vice President and Chief Financial Officer (“CFO”) of HP. She was not on HP’s board. When HP separated into Hewlett-Packard Enterprise Company and HP Inc in November 2015 she became CFO of HP Inc. She stepped down from that role in July 2018 to become Interim Chief Operating Officer. She has worked for HP companies for more than 30 years.

Ms Lesjak’s evidence in her witness statement was that she agreed with HP’s strategy of expanding its software business. But she was never persuaded by Mr Apotheker’s approach. In particular, she had “reservations about the timing and price of the proposed acquisition, particularly with respect to how stockholders might react and other challenges facing the Company.” There was considerable antipathy between them. Mr Apotheker wanted to remove her, and her job ultimately depended on his failure as CEO.

Ms Lesjak was too invested in and loyal to HP to be a reliable witness. She was over-defensive and had a variable recollection of events, especially when they appeared to reveal inconsistency on her part, or to suggest that she must have known about matters that HP claimed had never been disclosed such as Autonomy’s hardware sales. She tended to speculate on documents but almost invariably in favour of HP’s position.

Mr Manish Sarin (“Mr Sarin”) was a Senior Director in HP’s SCD team from 2010 to 2012 (when he left HP). Mr Sarin had a background in banking (beginning in 1999 when he worked for JP Morgan as an Associate, with employment thereafter by Wells Fargo from 2003 to 2005, and then Merill Lynch & Co where he was promoted to Director in its Technology Investment Banking group). He is no longer employed by HP and he sought to present himself as an independent witness.

He was on the software side of the Corporate Development team. He worked with Mr Andy Johnson (whom Ms Lesjak routinely referred to as “Andy” and who was head of Corporate Development), Mr Bill Veghte (who was prior to the Acquisition Executive Vice President of HP’s own Software division) and Mr Brian Humphries (Mr Johnson’s predecessor as head of Corporate Development, until March 2011).

In that role, Mr Sarin was heavily involved in the due diligence exercise prior to the Acquisition, and the evaluation of the potential benefits it could bring HP. He considered that Autonomy “appeared to be an attractive asset and a great fit for HP at all stages of the acquisition process” . That view was based, he said, on Autonomy’s published information, oral information provided by Autonomy’s management directly to HP, and the material disclosed to HP during the acquisition process. His evidence was that he was not aware of any of the matters now impugned in these proceedings.

His evidence covered a number of areas of importance, including (a) the processes followed by HP for acquisitions in general and in respect of the Acquisition in particular, (b) how HP came to focus on Autonomy as a target, (c) HP’s earlier consideration of a number of other targets, (d) HP’s valuation methodology and use of a DCF model both generally and in the particular case of Autonomy, (e) how the DCF model was built up in the case of Autonomy, (f) how synergies were valued, (g) the planning for and the various meetings prior to and during the due diligence exercise, and (h) the process of deliberation about the proposed acquisition and its approval. His witness statement also included paragraphs setting out his view as to how knowledge of the matters alleged by the Claimants would have affected the valuation of Autonomy, and the decision to proceed with it.

Mr Sarin attended trial and was cross-examined. He struck me as intelligent and confident, but over-defensive of his role in due diligence and the limitations of it he accepted at the time.

Mr Andrew Keir Markham Gersh (“Mr Gersh”) is a partner in KPMG LLP in the USA. He joined that firm in 1999, and became a US Certified Public Accountant in 2002. He has retained his licence as a UK accountant but cannot perform audits or accounting services in the UK. He was familiar with software revenue recognition and US GAAP. In 2011-2012, he was also familiar with IFRS (including its revenue recognition principles) but his primary experience is with US GAAP. Between 2004 and 2011 he worked on at least 50 financial due diligence engagements for HP, working on all of them in conjunction with HP’s SCD team and Enterprise Financial Reporting (“EFR”) team.

He was first contacted to perform financial due diligence on Autonomy on 22 July 2011 by Mr John Blank in HP’s EFR team and was told the next day that they would likely have until 15 August to perform financial due diligence procedures. He did not regard that as unduly rushed or unusual. His evidence addressed (a) whether he, his team or HP had any knowledge before the Acquisition that Autonomy was making substantial sales of ‘pure’ hardware (as distinct from hardware pre-loaded with Autonomy software), which he insisted none did; (b) whether Autonomy’s published information disclosed such ‘pure’ hardware sales; (c) how the due diligence exercise was undertaken, and particular points of focus in the course of it; (d) Autonomy’s other business lines, including sales through resellers (“VARs”), hosting arrangements and OEM transactions; (e) the constituents of Autonomy’s revenue and (f) why KPMG never provided a final report.

He was engaged also in a post-acquisition closing balance sheet project for HP to assist it in understanding Autonomy’s closing balance sheet so that Autonomy’s financials could be incorporated into HP’s October 2011 financial reporting.

Mr Gersh was in the difficult position of having to explain why, despite being aware that Autonomy sold some hardware (and having been shown three contracts referencing hardware), KPMG (for whom he was the lead partner in the due diligence exercise) never asked Autonomy how much hardware it was selling. After much prevarication, and what the Defendants described as close to “nit-picking” he eventually had to admit that they did know that Autonomy was selling hardware and had simply assumed that the hardware sold was “a component” of sales of software and were thus sales of appliances. (Yet, for example, none of the three contracts appeared to be such sales of appliances.) Both his witness evidence and his oral evidence when cross-examined seemed to me to reflect over-lawyering, and the fact that (so he acknowledged) he had previously given evidence (on 6 and 13 April 2018) in the US criminal trial, had had 5 to 10 meetings with US government lawyers, and had met with HP’s lawyers at least 5 times also.

Mr Apotheker first met with Dr Lynch on 12 April 2011. This was the first of a very small number of direct meetings or conversations between Dr Lynch and Mr Apotheker before the acquisition was announced on 18 August 2011. These comprised:

This initial meeting between Dr Lynch and Mr Apotheker on 12 April 2011.

A meeting on 16 June 2011, which according to Mr Apotheker lasted 1½ - 2 hours.

A meeting lasting a few hours in France in late July 2011.

A call on 14 August 2011 when there was final discussion of the price.

The meeting on 12 April 2011 lasted about an hour. Dr Lynch was in the US on other business. This was really a chance for two CEOs of technology companies to come together and exchange views. Mr Apotheker formed a favourable opinion of Dr Lynch, considering him very knowledgeable and enthusiastic about Autonomy and its products. He could not remember the meeting in any detail, but according to his evidence they discussed trends in the industry, including unstructured data, amongst other things. (Footnote: ³⁰) Mr Apotheker thought that nothing really concrete was discussed in relation to a possible partnership between HP and Autonomy.

The SCD group also provided a presentation about Software AG (code named “Singapore”). The graphic showed that Software AG offered very different capabilities from those offered by Autonomy, with hardly any overlap. (Footnote: ³¹) Mr Apotheker said that he would have regarded Software AG as a “pure software” or “pure play software” company, although it had a more significant services business. Its anticipated operating margins, at 28% by full year 2016, were far lower than Autonomy’s, but still regarded as “strong” in the executive summary at p.3. (In a similar presentation on TIBCO, it was estimated that operating margins would reach 30% by 2015 and that was also regarded as “strong”).

Present at HP’s finance and investment committee were Mr Apotheker, Ms Lesjak, Mr Johnson and Mr Robison. They supported the continued assessment of the potential acquisitions. The minutes of the 25 May meeting record the following, under “M&A update”:

The view was, however, that HP should make the Autonomy acquisition first, and undertake the Software AG acquisition after the Autonomy acquisition. See Mr Johnson’s email of 28 May 2011 to Mr Veghte and others:

By the end of May 2011 HP’s thinking was accordingly as follows, as Mr Apotheker expressly accepted:

First, HP had reached the view that unstructured data was important and a growing part of the software market, and that HP had a limited number of capabilities in that field.

Secondly, HP considered that Autonomy had cutting edge and widely-accepted software which had become an industry standard for analysing and managing structured and unstructured data.

Thirdly, Autonomy was also attractive in view of the fact that HP owned Vertica, and had SQL database technology, a data management system used primarily by stock exchanges, telecommunications companies and banks for online transaction processing.

Fourthly, HP thought acquiring Autonomy would enable HP to combine the unstructured software expertise with its own products and offer an attractive product to the market (the integrated or unified information stack).

Fifthly, HP saw the acquisition as potentially transformative for HP.

The deal sponsors at HP were particularly excited about the potential synergies that could be achieved. Mr Apotheker’s evidence in cross-examination confirmed this:

HP produced various valuations of these synergies, ranging as high as $30bn. The Defendants submitted that emails in July through September 2011 demonstrate that Mr Apotheker’s focus at all times was on the synergies that could be created, and the considerably increased revenues and margins which would result were what drove the deal, rather than the stand-alone value. (This is of considerable relevance to the quantum of loss if liability is established).

The Defendants suggested that this may in part explain Mr Apotheker’s evidence (to which I must also return later) that although he read Autonomy’s 2010 annual report, he did not read Autonomy’s 2011 quarterly or half-yearly releases at all. By July 2011, the 2010 annual report was already 7 months out of date. The Defendants posited that the fact that he did not read the 2011 quarterly or half-yearly releases demonstrates that the detail of Autonomy’s trading results and metrics were not of great interest to him: what most mattered to him was Autonomy’s technology and its potential as a transformative business.

During the summer of 2011, HP’s SCD group worked on various DCF valuations of Autonomy. These (which the Claimants called “Deal Models” ) were prepared by Mr Sarin and his team and their objective was to assess the standalone value of Autonomy. An initial version had been prepared as early as February 2011, and Mr Sarin explained that the valuation went through various editions until 18 August 2011. The elaboration of the valuations is addressed later.

By the end of June 2011, both Mr Apotheker and Mr Robison were plainly very enthusiastic about Autonomy as a business. Their focus was on the technology and how it could be combined with HP’s assets. Mr Robison wrote to Mr Apotheker on 29 June 2011 referring to a “REALLY good meeting today ”. Mr Robison wrote:

Dr Lynch had also concluded by this time that Autonomy was a good fit for HP’s vision. The combination of HP, their sales force and the Vertica database with Autonomy and its IDOL technology was a sound strategy.

Nonetheless, at this stage Dr Lynch still considered it was only possible, rather than probable, that HP would make an offer for Autonomy: he did not consider an acquisition as a real possibility until the meeting with Mr Apotheker in Deauville on 28 July. As he said in cross-examination,

One of the threats perceived by HP and its advisers was the emergence of a rival bidder. Throughout the process of analysis, and until the announcement of the bid, HP was concerned that another bidder might emerge, which would turn any deal between HP and Autonomy into a public takeover battle. This risk was referenced in Mr Apotheker’s email to Mr Robison of 18 June 2011:

On 24 June 2011, BarCap gave HP advice on how to avoid a contested takeover, in a presentation headed “Project Plato – Deal Protection Considerations". This was circulated by Andy Johnson, whose covering email stressed the importance of working closely with Dr Lynch:

The presentation itself noted the need for HP to move rapidly and offer a strong price. It recommended that “ With full Target CEO backing and Target Board support Hercules should move rapidly to formalise an offer and try to avoid a leak ”, and again stressed that the “ Support of CEO is key to securing Atlantis ”.

There was a further meeting between HP and Autonomy personnel on 29 June 2011. Dr Lynch, together with Mr Hussain, Mr Kanter and Dr Menell met with representatives from HP’s corporate development, software and cloud services teams to discuss how HP and Autonomy might be able to work together. The 29 June 2011 meeting was a high-level strategic discussion about how HP and Autonomy could work together; it was not the forum for the provision of non-public information and there was no specific discussion at the meeting about price.

After the meeting, Dr Lynch had a private dinner with Mr Robison. The dinner was an opportunity for the two individuals to get to know each other and to discuss technology and the future of the technology industry. The meeting was around three hours long. It was attended also by (amongst others) Mr Sarin and Mr Jerome Levadoux (a Vice President in HP’s Products, Information Management and Analytics group, and a former employee of SAP), both of whom made a note. Nevertheless, there was a dispute as to what was said:

Mr Sarin accepted that this was a general introductory meeting, and that Mr Robison had been very positive.

While Mr Sarin’s notes record some discussion about the level of premium normally payable on a UK acquisition, he accepted that it was possible that was an internal HP discussion after the meeting.

The point most disputed was Mr Sarin’s evidence in his witness statement, that he thought Dr Lynch had described Autonomy as a “ pure software” company. This was not something reflected either in his or Mr Levadoux’s notes, and Mr Sarin’s memory about what happened at the meeting outside the notes was poor. He said in evidence that he took the phrase “pure software company” to mean that Autonomy “ was a company that was predominantly software, sold a small number of appliances and was one of the more scaled businesses in the software industry ”. The Claimants relied on this as support for their claim that Autonomy was sold on that basis; the Defendants rejected this as contrived and pointed out that Mr Sarin gave no basis for this very elaborate meaning. Although that elaboration does appear to me to be lawyer-crafted, it will be seen later in this judgment (see paragraphs 3878 ff ) that I have concluded that it is more likely than not that Dr Lynch did describe Autonomy (including to Mr Sarin) as a “pure software company” or “pure play software company” with the intention and effect of conveying a special selling point, the success and self-sufficiency of Autonomy’s software business without the need for other revenue streams.

On 7 July 2011 Mr Apotheker asked the corporate development team to think “ aggressively and creatively ” about synergies. Mr Apotheker sought in evidence to underplay this, suggesting that this was “ just one exercise ”, that he wanted to “ see how far we could stretch it ” and there was also a very pessimistic scenario. However, the contemporaneous documentation does not support this: Mr Apotheker was asking for more synergies because he wanted the FY14 revenue to be “ much higher than $2.18 ” billion dollars. Dr Lynch cited an instant messaging conversation between Mr Sarin and Mr Levadoux, which ran as follows:

Consistently with that, Mr Apotheker also asked Mr Risau (his Chief of Staff who had been COO at SAP, and CFO of a division but who was not part of the SCD group within HP) to look at synergies.

Mr Apotheker was not the only one pushing for higher synergies. A little earlier, according to a message from Mr Levadoux sent on 2 June 2011, Mr Bill Veghte (then Executive Vice President, HP Software) had “ challenged us to think how we can turn this into a $5B business (versus current assumption of $2B-$2.5B by 2015 with synergies) ”. (Footnote: ³²)

The SCD group prepared a further valuation in July with new figures for synergies. A tab named “Waterfall” in this spreadsheet posited a standalone value of $10.324bn. The identified synergies brought the value with synergies to $27.835bn. Mr Sarin described this in his witness statement (reflecting the filename of the relevant valuation model) as “ the “aggressive” synergy case ”, and said that he thought it was not included in any of the decks subsequently provided to the board. In fact, certain revenue projections (of revenue, gross and operating profit, gross and operating margin, net income and net margin and EPS) which were modelled on “Divesting Poseidon [HP’s Personal Systems Group] and winning in the Big Data Analytics market” were included in the deck presented to the board at its meeting in July 2011, as Mr Sarin accepted in cross-examination, although he also maintained that he only subsequently became aware of this when it was pointed out to him when he was preparing for his cross-examination. This was known either as the “aggressive synergy case” or as the “Transformational case”

Mr Apotheker’s presentation to the Board did not address valuation; but it did refer to the possibility, under what he referred to as the “transformational case” , of total revenues of $6 billion in 2014. That compared to a base case projection of revenues in 2014 of $2.2 billion. Mr Apotheker had asked the SCD to model what the total value of Autonomy inclusive of synergies would have been on the basis of that “transformational case” and revenues of $6 billion, which had led to a DCF valuation of $46.589 billion (including synergies of $30 billion). Mr Robison had commented that he thought that this figure, which was premised on HP “winning in the Big Data Analytics market” , “may be a bit much” , and that more work would be required “to make sure this is credible.” I accept that the $46 billion valuation was never shown to the Board. (Footnote: ³³)

HP’s growing enthusiasm for the acquisition exacerbated apprehension about the threat of a rival take-over bid from an “interloper”, in particular, Oracle. On 15 July Mr Robison provided Mr Apotheker, Ms Lesjak and others with summary points drawn up by Mr Johnson from an “Interloper Analysis” prepared by BarCap. The analysis had addressed how Oracle might interfere with HP’s process. One of the points made related to due diligence. Mr Johnson wrote:

Later, on 27 July, Mr Robison provided Mr Apotheker with a presentation from BarCap on Oracle’s usual M&A processes, again reflecting HP’s concern about a competitive bid from Oracle.

HP’s board held a strategy meeting on 19 to 21 July 2011. This was a crucial meeting for HP because it was deciding on its future strategy. Mr Apotheker and Mr Robison were recommending not only the acquisition of Autonomy, but also the hiving off of the PC business (referred to as the “PSG” personal services group), amongst other things. The PC business had been part of HP’s DNA.

As to Mr Apotheker’s presentation to the board:

In line with earlier documents, his proposal was that: “ HP will be the leading provider of information solutions on-premise and in the cloud .” The value proposition included “ Disruptive business intelligence and big data analytics stack. ”

He advised the board that decisions needed to be made urgently, or immediately:

Shift portfolio mix towards higher growth, higher margin businesses (e.g., Software)

Decrease exposure to commoditizing, low-margin end-user devices

Decrease dependency on consumer markets; focus on commercial…”

The first bullet point related to the Acquisition; the second and third to the hiving-off of the PSG business. The rationale for “ acquiring [Autonomy] now ” was set out in the document. The aim was to build a leading position for HP in enterprise software with the acquisition of Autonomy. Mr Apotheker’s presentation explained about the creation of a data stack involving structured and unstructured data, and combining Autonomy’s business with Vertica and HP’s assets:

The analytics market that HP was targeting was estimated as being a $31bn market.

The board were told that Autonomy was the best option, having been tested against other potential candidates: Software AG and TIBCO were specifically identified (by code name) as other candidates which had been evaluated.

The board were provided with Mr Apotheker’s financial projections for the “transformational case” (including both the Autonomy acquisition and the PSG divestiture). The 2014 projections were based on the $6.1bn figure for Autonomy revenue (with synergies) that Mr Apotheker had previously targeted and which the SCD group had provided for him. When cross-examined, Mr Apotheker accepted that he thought that this was realistic, and that a statement in his witness statement (which he had repeated earlier in his testimony) that the “more aggressive scenarios” formed no part of the business case presented to the board were “obviously” inaccurate. However, he stressed that this was “the theoretical potential highest number that we could achieve” and assumed “winning in the Big Data Analytics market” : the “base case” forecast $2.2 billion of revenues in 2014 (taking into account expected synergies). (Footnote: ³⁴)

Mr Apotheker’s proposal was to announce the intention to divest PSG simultaneous with the announcement of the Autonomy acquisition in mid-August. This was intended to coincide with the announcement of HP’s Q3 results.

The proposal was that in the following 6-12 month period HP would “ evaluate additional software targets such as [Software AG] and [TIBCO] ”.

Perella Weinberg Partners were one of the two banks advising HP at the time. Mr Apotheker considered them competent and good at their job. They also made a presentation to the board. Perella Weinberg endorsed the strategy, and agreed that the intention to acquire additional software capabilities was critical to fulfil HP’s vision and drive margin and growth. In their executive summary (dated 20 July 2011) their advice began as follows:

Company's stated intention to acquire additional software capabilities is critical to fulfill its vision and drive margin and growth”.

Mr Apotheker told me that he agreed “essentially” with this assessment.

Perella Weinberg also noted that HP’s management had reviewed other M&A targets but few were “truly transformational”, unlike the Autonomy acquisition.

Perella Weinberg also noted HP’s current difficulties, which put it in a “ precarious position ”:

According to Mr Apotheker, the reference to “ uncertainty regarding Q3/4 and FY/2012 ” arose because “ it appears that HP’s forecasting capabilities and abilities under the CFO were not very good ”. He told me that he considered that “reality” had proven the CFO to be “very bad at forecasting revenues”. It appeared that there were tensions even at this stage between Mr Apotheker and Ms Lesjak (the CFO); and a little later, she and Mr Apotheker had a serious falling out with pivotal repercussions, and he considered her to be trying to undermine him and the transformational project from the beginning. Mr Apotheker also complained about the inaccuracy of Ms Lesjak’s “flashes” (which he explained are forecasts of an immediately expected result).

Mr Apotheker agreed that the business would take a considerable time to turn around. The margins of the PC business were as high as they could ever get and could only deteriorate. The services business was also low margin and HP’s corrective measures would take a long time. More than 70% of HP’s revenues were accordingly either deteriorating or only very slowly improving.

Perella Weinberg also explained why the release of the Q3 results was an important “catalyst”, and it was advisable to announce the acquisition at the same time as releasing the results, which were anticipated to be poor. Mr Apotheker also thought this was a good idea. The executive summary made the following points. Mr Apotheker agreed with all of them at the time (save that he did not confirm his agreement to the last point):

Perella Weinberg considered that HP would find it hard to “execute out” of this predicament on a timely basis, a point that Mr Apotheker agreed with. They also considered that in the near-term HP could not organically or incrementally address its fundamental challenges in a meaningful way. Mr Apotheker thought that this might have been possible, if HP had been given enough time, although this was something that the markets would react negatively to. Thus it was never anything he suggested to the board and as he said in cross-examination he “ shelved that idea” in his own mind.

Perella Weinberg made further recommendations with which Mr Apotheker also agreed, including the divestiture of the PSG group, and the announcement of that with the Q3 release. They also recommended that HP pursue the Autonomy acquisition expeditiously as a “ critical offensive move ” and a “ unique opportunity to reposition [HP] as the undisputed leader in context-aware computing ”.

Perella Weinberg also adverted to the interloper risk, stating:

The Defendants suggested that this was a reference to HP being in a position to knock out the competition with the size of its bid. Mr Apotheker thought that was one way of looking at it, but that another way was that another interloper (such as Oracle) with “overlapping interests” might have to divest or restructure and would thus not be able to make the same kind of offer.

However, Perella Weinberg also warned of the impact on the investor community and the real possibility of a hostile reception.

First, they warned that the markets could react badly to the announcement:

Secondly, they advised that the changes could “ catalyze fundamental re-rating ” and shift HP’s shareholder base towards a more growth oriented profile i.e. moving away from value investors to growth investors, which as Mr Apotheker acknowledged would be good for HP.

Perella Weinberg in addition stressed the importance of flawless execution and management devotion to the integration and transformation:

As Mr Apotheker put it:

A proper integration plan, a proper extraction plan of the PC business, execute this to close to perfection and while at the same time continuing to run the existing business.

Mr Apotheker thought that he was the man to carry out this dedicated integration work, along with Mr Robison and the team. As elaborated below, in the event, shortly after the bid, and before the Acquisition had completed, HP jettisoned Mr Apotheker and Mr Robison, and according to the Defendants, did not give the integration anything like the priority and focus that Mr Apotheker and Perella Weinberg had envisaged.

After Mr Apotheker and Perella Weinberg had made their presentations on 20 July 2011 there was a board discussion. The board authorised Mr Apotheker and other members of management to pursue the Autonomy acquisition (code-named Project Tesla) with a maximum price of $11.7 billion. This was substantially in excess of HP’s stand-alone valuation of Autonomy of $10.324 billion. (The same valuation showed an aggregate value plus synergies of $17.596 billion.) It was also at a significant premium to Autonomy’s market capitalisation. The board also authorised Mr Apotheker and Mr Robison “and other members of HP management” to assess the potential divestiture of the PSG group (code-named Project Hermes), with a view to announcing this (if approved by the Board) at the Q3 earnings announcement.

There were two phases of the discussions about price. The first was in July 2011. The second was in August 2011. Certain parameters framed HP’s approach:

First, HP realised that they needed to pay a substantial premium over market capitalisation in order to secure the asset. They also perceived that they needed Dr Lynch’s support for the acquisition to go ahead: they had no appetite for a contested bid, and they wanted him to stay on after the acquisition.

Secondly, HP knew that (assuming Autonomy’s financial position, performance and prospects to be as depicted in its published information) Dr Lynch would not go below £25 per share. He had made that clear, and seemed in no hurry to sell.

Thirdly, HP had determined that Autonomy “uniquely” offered the transformational opportunity that HP needed, and that HP could not delay. HP were prepared to pay a price above Dr Lynch’s floor even after Autonomy’s share price had declined, such that the implied premium increased from that implied by the range of values originally agreed with Dr Lynch at the end of July.

Fourthly, HP were not wedded to obtaining any particular proportion of the expected synergies on an acquisition, and had been prepared to cede 53% of their anticipated synergies in the proposed acquisition of TIBCO.

Fifthly, HP wanted to avoid competitive bids. They were worried that there would be interlopers, including Oracle, and were prepared to pay a “compelling price” to forestall a bidding war.

Mr Apotheker recognised that he would have to pay a “control premium” , which would in the ordinary course be “ at a bare minimum ” 30%, and could be significantly higher “depending on the quality of the asset”.

He emphasised that he looked at two things: (a) present value as a stand-alone business and (b) synergies to be expected in the future. The acquiring company would wish, of course, to retain as much of the synergy value (which equally obviously depended on its own assets and effort) as it could. The company proposed to be acquired would regard synergy value as justifying a higher price (i) because of its value to the acquiring company, for which it ought to be willing to pay and (ii) because the value that it can provide is a valuable attribute of the business which should also be reflected in the price.

Mr Apotheker met with Dr Lynch in Deauville on 28 July 2011. It is common ground that they discussed the following:

The strategic rationale of the deal.

The integration of Autonomy, which would involve Autonomy being a semi-autonomous company within the group. Mr Robison had made it clear that if Autonomy was integrated into the greater HP group, the “ immune system ” of HP would kill Autonomy off and the value would be lost, as in other acquisitions.

Dr Lynch’s role, which would involve him being in charge of the software assets of HP, including the Vertica business. This was appealing to Dr Lynch, given his passion for Autonomy’s software and the technological synergies that could be created in combination with HP.

Mr Apotheker confirmed that at this meeting they agreed that Dr Lynch would become head of HP’s software business in place of Mr Veghte.

The position from Dr Lynch’s perspective, based on his first witness statement, can be summarised as follows:

He had not been looking for a sale (Footnote: ³⁵) , but HP’s proposition was exciting. As regards his personal position Dr Lynch was already a successful and wealthy man. He had already sold a large part of his shareholding in Autonomy over time.

However, he thought that Autonomy was a potential takeover target in any event, and had been advised as such by Mr Quattrone towards the end of 2010. If HP were to offer a premium of 60% or more, Dr Lynch knew that resistance to an offer would be very difficult.

Having met with Mr Apotheker and Mr Robison, Dr Lynch saw HP as an attractive bidder. He could see the merit in the potential combination with HP and the integration of Vertica, and access to HP’s much larger sales channels could take Autonomy to a higher level: HP could provide the firepower for Autonomy to take on competitors such as Oracle.

After Mr Apotheker and Mr Robison explained to Dr Lynch that they would want him to stay on, and to lead Autonomy as a semi-independent company under HP’s aegis, Dr Lynch expected to be working with HP during the years after the takeover, continuing to develop Autonomy’s business as part of a larger group.

Price was discussed at this meeting for the first time. Dr Lynch had been pressing for £27 per share. It was agreed that if HP was to acquire Autonomy, HP would need to offer a price in an agreed range. Dr Lynch’s recollection was that this range was between £25.50 and £26.50 per share. Dr Lynch considered that management would not have the ability to resist a deal within that range and Dr Lynch would recommend such an offer to shareholders.

Dr Lynch’s unchallenged evidence is that:

First, he was not willing to recommend the sale to Autonomy’s shareholders unless they were getting what he believed they deserved, which needed to be an exceptional offer which accounted for the bright future of the company.

Second, that as Mr Robison and Mr Apotheker knew, Dr Lynch would not be willing to support a bid below £25 per share.

Moreover, from HP’s perspective, it was a condition of any offer that Dr Lynch would recommend the bid: they were not willing to embark on a hostile takeover without him; it would have been impossible to put Dr Lynch in charge of the business if he had not recommended it; and the whole thing required his agreement. This is borne out by events in August, when HP were considering the renegotiation of the acquisition price in light of falls in Autonomy’s share price.

As previously explained, HP were at all times conscious of the interloper risk. Part of their strategy for dealing with that risk was to make a compelling offer, which was the “ highest logical price that still made sense in order to discourage everybody else” .

On behalf of HP, BarCap prepared a draft Letter of Intent dated 28 July 2011 which set out the strategic rationale for the sale as follows:

The draft letter stated: “ HP would be prepared to make an all cash offer to acquire the entire issued and to be issued share capital of Autonomy at a price of between £24.94 and £26.94 per share .”

Dr Lynch commented that:

HP’s agreement to the deal was subject to due diligence. HP’s due diligence took place from 1 to 18 August 2011. HP was assisted by KPMG and had advice from two investment banks. When asked about this process, Mr Sarin described it as “ confirmatory due diligence ”, that is to say, a process that took place after the agreement on price:

It was agreed to use oral discussions (by telephone) as much as possible to minimise the generation of hard-copy materials that might have to be provided to competing bidders under the Takeover Code. There was a dispute as to whether it was HP or Autonomy who asked for this. Mr Apotheker insisted in cross-examination that the request came from Dr Lynch and that HP agreed only “because we were respectful of Autonomy’s desire to keep sensitive data out of the hands of competition”. However, an internal HP email dated 15 July 2011 from Mr Johnson to Mr Robison headed “Interloper Analysis” suggests that, amongst other concerns as to “how the O company might interfere with our process” HP was much concerned about the possibility that Oracle might get access to “all the diligence we have received so the takeaway is to be careful about our diligence so we don’t enable O to get a deep dive on sensitive data”. It seems to me likely, and I find, that it was a matter of joint concern and common interests, that HP raised the issue and Autonomy readily agreed.

Dr Lynch played a very limited role in the process. The Claimants do not advance any claim against him in respect of the due diligence process, or anything said to HP during it: although various misrepresentations were allegedly made by Mr Hussain in the process, none is alleged against Dr Lynch. It is, however, alleged that he made a representation about HP’s ability to rely on Autonomy’s accounts at a meeting shortly before due diligence began, on 29 July. Later in this judgment I address broader issues relating to inducement and reliance in the various contexts in which they require to be considered (and see, especially, paragraphs 478 to 522 on the tests of reliance applicable in FSMA claims; paragraphs 3236 to 3252 in respect of the misrepresentations concerning Autonomy’s OEM business; paragraphs 3979 to 3988 in respect of the direct deceit/misrepresentation claims; and paragraphs 3944 to 4055 in respect of the FSMA claims); but the following should be noted in relation to the due diligence process itself.

First, the process was, in Dr Lynch’s view rushed; but the timetable was set by HP. Dr Lynch’s position was that at no stage did he attempt to inhibit the due diligence process. It was not disputed that Autonomy refused to make available Deloitte’s working papers to HP: but Mr Gersh of KPMG advised and Perella Weinberg confirmed to HP that it would have been unusual if working papers had been provided in a transaction of this type. Further, Dr Lynch was content for KPMG to speak directly to Deloitte.

To the extent that HP did limit its exercise the Defendants suggested that this was their decision and it reflected two related facts:

First, that HP was (at least until it had second thoughts) committed to the bid. The HP decision makers had relatively little interest in the outcome of the due diligence (at least in relation to financial aspects: (Footnote: ³⁶) by contrast, technical due diligence, the solidity of the IP rights and the like were closely scrutinised).

Secondly, HP was concerned about an interloper having access to Autonomy’s information, which given HP’s commitment to the bid was undesirable: that militated against a long process, especially since (as Mr Cooke explained at §10 of his (unchallenged) statement), due diligence in a public company takeover in the UK is relatively limited (in contrast to the acquisition of a private business) because of the Takeover Code requirement of equal access to all bidders.

HP was able to ask for all the information it wanted to during the due diligence process:

HP was the buyer. It could ask any questions, or insist on any information, that it felt was necessary if it wanted to proceed with the bid.

HP and KPMG had access to Deloitte. They were able to ask Deloitte about all aspects of Autonomy’s accounts.

From HP’s perspective it was satisfied that it and KPMG had carried out all the due diligence they wanted to carry out.

HP was able to take advice from two investments banks, who were satisfied that the deal was at fair value.

Secondly, the Defendants made three forensic points as follows:

It was always obvious to Dr Lynch that any acquiror would want to do due diligence, and would insist on all the due diligence that it considered that it needed, and ask any questions that it liked (which is exactly what HP did). They submitted that this makes it most improbable that Dr Lynch could have been involved in, or aware of, any fraud.

Similarly, it was always obvious that as the acquiror HP would have complete access to Autonomy’s books and records after the acquisition. (Further, one would expect any competent acquiror to check its own books to see what business it did with the target, as well as talk to joint customers about what they bought from the target.) The Defendants submitted that in these circumstances, it would have been futile deliberately to conceal any of Autonomy's transactions.

In any event, it could never have been assumed in advance by Dr Lynch that due diligence would necessarily be limited. If, as the Claimants contend, he had been consciously involved in a misreporting of Autonomy’s business, the correct position would have been expected to be revealed.

An internal HP preliminary due diligence timetable circulated on 30 July 2011 envisaged due diligence being conducted between 29 July 2011 and 18 August 2011, when there would be signing of the definitive agreement. Mr Apotheker’s evidence was that “the plan was to do proper due diligence, if possible, to finish it by the 18th”. 18 August 2011 was the day on which the Q3 results were to be announced. In the event, due diligence did not begin until 1 August.

Dr Lynch regarded this as an extremely rushed timetable: as he said, he had “never seen a deal done at this speed”. But in cross-examination he explained that he thought the timetable was simply to motivate the team to get on with it: “I thought they absolutely knew it wasn’t going to happen…”.

On 3 August 2011, Autonomy and HP entered into a letter agreement by which Autonomy agreed not to solicit, initiate or encourage an offer from anyone other than HP, until midnight on 17 August.

The Claimants’ closing submissions depicted the period between 1 August 2011, when formal due diligence began, and 18 August 2011, when the Boards of both HP and Autonomy approved the deal – that is, HP’s Board agreed that HP should make a cash offer for Autonomy’s shares, and Autonomy’s Board agreed to recommend the offer to its shareholders – as marked by intense activity by HP, Autonomy, and their respective advisers: for HP, BarCap, Perella, KPMG, Freshfields, and Gibson Dunn & Crutcher; for Autonomy, Qatalyst and Slaughter and May. Documents were uploaded to a virtual data room for HP’s review; there were scheduled due diligence calls on particular topics; legal documents were negotiated; and HP planned its communications strategy. HP held daily internal due diligence update calls from 4 August.

HP’s approach was plainly influenced by its perception of and anxiety to reduce interloper risk. This was further informed by advice in a paper (dated 15 July 2011) from BarCap on ‘Potential Interloper Spoiling Tactics’ , which was attached to the internal HP email which I have referred to in paragraph 202 above, and from which Mr Johnson took the message to be careful about diligence so as not to enable Oracle to do a deep dive.

To that end, the exercise was largely undertaken in a series of calls, between 1 August 2011 and 17 August 2011.

Mr Apotheker explained (Footnote: ³⁷) :

He added that while HP agreed to structure due diligence so as to address this risk, this did not mean that there was not proper due diligence:

In re-examination Mr Apotheker also explained that due diligence was structured so that meetings would take place where the team could review documents and discuss them, without them being stored in a traditional data room.

On Autonomy’s side, the calls were largely with Mr Hussain and Mr Kanter. Dr Lynch’s evidence was that he did not participate in any of the financial or legal due diligence calls, though he did join a few technical calls to discuss Autonomy’s products. This was broadly confirmed by Mr Sarin in cross-examination:

In a call on 4 August 2011, Mr Sarin (together with Mr Johnson and various others from HP) discussed some parts of HP’s valuation model with Mr Hussain and Mr Chamberlain. Only a small part of the model was shown to them, and nothing was provided in advance of the call. Revenue projections were shown by product line, but only up until 2016. To the extent these were commented on by Mr Hussain, those comments were not relied on by HP, as shown by the following exchange with Mr Sarin:

The conversation was a short one (perhaps an hour); and Mr Sarin’s note is very brief. Certain changes were made to HP’s model following the call, but these appear not to have been made in reliance on any comment of Mr Hussain, but rather on HP’s own assumptions. In any event, substantial changes were made to the model at a later stage between 4 and 18 August 2011, as Mr Sarin accepted, though he did not accept that the changes would have had a “substantial impact” on valuation. (In fact, the changes made actually reduced the estimated enterprise value by half a billion dollars from $10.0117 billion to $9.5021 billion.)

In addition to their financial due diligence, HP conducted a technical due diligence of Autonomy’s product portfolio. This was of primary importance to HP. When the due diligence findings were reported to HP’s board, “R&D/Products/Technology” was first on the list of the due diligence areas identified: HP was satisfied with what they had found, which aligned with their synergy assumptions.

The proposal to provide a selection of customer contracts arose at the meeting on 29 July 2011. According to Mr Sarin’s email summarising the meeting:

The proposal evolved, and the number of contracts sought was reduced (to 40). Mr Sarin gave the following explanation of why HP and KPMG wanted the lists:

On 4 August 2011, Autonomy (through Slaughter and May) placed in the virtual data room lists (with names redacted) of what were presented as Autonomy’s top 40 customers, and top 40 contracts, by revenue. Mr Kanter informed HP that they were now in the data room. Redacted copies of the contracts followed. They were subsequently reviewed by KPMG. The top 40 contracts ranged in value from $3 million to $22.5 million.

Dr Lynch did not dispute, when cross-examined about it, that a number of contracts for the resale by Autonomy of third-party hardware within that value range (and worth up to $7 million) were omitted from the list. The Claimants ascribed this to deception; the Defendants ascribed it to Autonomy’s understanding that this was not required, and that what HP was seeking was a list that would enable it to look at the concentration of Autonomy’s business by industry, and a set of contracts so as to understand Autonomy’s contract terms.

Three points may be noted:

Both Mr Sarin and Mr Gersh muddled up the two lists when giving evidence in the US criminal proceedings and neither could really remember which they had seen. The importance they have attached to the top 40 lists in these proceedings appears to be considerably greater than they ascribed to them at the time.

The Claimants did not seriously challenge Dr Lynch’s evidence that he was not involved in the collation of the contracts or the compilation of the lists, and their cross-examination of him proceeded on the basis that he was unable to give evidence in this regard.

Some of the top 40 contracts showed that Autonomy sold hardware, not restricted to appliances, and some of KPMG’s draft questions were about this; but Mr Sarin and HP never questioned how much hardware Autonomy sold.

As mentioned previously, although production of Deloitte’s working papers had been denied to them, KPMG were not prevented from discussing matters with Deloitte, and on 17 August 2011 there was a call between them when they discussed specific questions which KPMG had prepared.

Mr Gersh of KPMG did most of the talking on HP’s side, but Mr Johnson, Mr Sarin and others from HP also attended. The Deloitte representatives were Messrs Mercer, Knights and Welham. Mr Kanter was present in the room with the Deloitte personnel but did not speak. Mr Hussain dialled in. Autonomy management had provided suggested responses to KPMG’s written questions ahead of the call. This apparently upset Mr Sarin when he learned about it subsequently (he did not know it at the time) since (he stated in his witness statement) the “whole purpose of the call was to obtain Deloitte’s independent view” , adding “Had I known then what I now know, this would have been a matter of concern to me.” (I think this was a comment plainly informed by hindsight. I doubt very much that he would have thought anything of the episode at the time.)

Mr Sarin noted the responses given to the questions as follows:

Mr Sarin confirmed, looking at the note in the margin, that Deloitte had explained that they had reviewed all revenue contracts over $1m and a sample of smaller ones as well. He accepted that on the call Deloitte had mentioned somebody who had made a whistleblowing query or complaint (see paragraphs 2232 to 2289 below).

Mr Welham also gave evidence about the call on 17 August 2011, which he remembered. He confirmed some of the answers noted by Mr Sarin. Asked whether there were any disagreements between management regarding accounting policies or conclusions, he said “No major disagreements to my knowledge, no”. He confirmed that the comment about the size and quality of Autonomy’s accounting/finance team was a fair reflection of how he saw things at the time.

The call with Deloitte was the last material stage in the due diligence exercise undertaken, though (see paragraphs 269 to 271 below) that exercise was incomplete. (Footnote: ³⁸)

On 9 August, some time before the call with Deloitte on 17 August 2011, KPMG provided HP with a draft due diligence report. Although the Claimants contended that Dr Lynch was actively engaged behind the scenes, in particular in discussing behind the scenes with Mr Hussain how best to respond to requests from HP, they had to accept that his name is on very few of the emails or other documents connected with the process. The draft due diligence report itself did not suggest that Dr Lynch had been involved in the due diligence process; rather:

The executive summary indicated the limitations of the work that had been done, noting that:

Mr Sarin confirmed that this reflected his understanding.

The report noted that Autonomy recognised revenue in accordance with IFRS, and stated that there were “ some policy differences [with US GAAP] related to extended payment terms, sales to VARs, and potentially fair value analysis ”. It highlighted the fact that Autonomy “ recognizes revenue for license sales upon sell-in to its VARs rather than on a sell-through basis to end customers ”. It identified the fact that under some of Autonomy’s hosting arrangements, there was a licence element under which customers had the right to take the software on premise – a point which Mr Sarin confirmed he understood.

The report stated that the due diligence process was not complete, stating that:

The items in KPMG’s statement of work not yet completed were significant. These included inquiries about Autonomy’s historical revenues and revenue composition; enquiries about trends in revenues; enquiries about expenses and trends in costs, gross margins and operating margins. KPMG never completed these tasks.

Mr Sarin was unable to say who had decided that the deal should proceed despite the fact that the due diligence processes had not been completed, save to say that it was not his own decision. He accepted that the presentations that were put before the board never explained that KPMG had not carried out all of this work. It is to be noted, however, that the strategy decided at the July board meeting was that the acquisition (and the divestment) should be announced as part of the Q3 2011 report.

Mr Apotheker did not read the KPMG draft report (of 9 August 2011) on the due diligence process. He was not aware that KPMG had never finalised the report. He had relied on the “very professional people…to make sure that things were in order…if somebody would have thought there was something for me to read, they would have told me.” Mr Apotheker referred to an overview of analysts, most of whom he said, “were on the buy side”, though he was aware that “there were a couple of analysts who held a contrary opinion”. He said he read “a couple” of analyst reports, but did not remember at which stage in the process. He did not involve himself in the top 40 customer list. His line overall was “You have to trust your people…”.

A report to the HP board of the “key due diligence findings” was provided in advance of the board meeting of 12 August 2011. It gave the ‘green light’:

Technical due diligence: Technical due diligence was the first item on the list. The findings recorded, amongst other things that Autonomy’s product capabilities and features aligned with HP’s synergy assumptions:

VSOE: The report noted that Autonomy’s approach to VSOE was different from HP’s. It stated: “ Post-closing need to align Tesla VSOE with HP ”.

Top 40 customers: All that was reported to the board was that the review of the top 40 customers list was ongoing and no red flags had been found.

The report was updated for the 16 August 2011 board meeting:

There were ‘green lights’ for all the functional areas identified.

The technical due diligence was still first on the list. The updated section was very positive, and read:

In respect of legal/IP, the report made it clear that the review of the Top 40 contracts was in connection with the analysis of contractual documentation and terms (and not part of the financial due diligence, below). The section read:

There was a new section dealing with the financial due diligence, which had not been in the report for the 12 August 2011 meeting. It read:

In the meantime, and in parallel with the due diligence exercise, in the first half of August 2011 HP finalised their decision to go ahead with the acquisition, which they announced on 18 August 2011. The following account of the process is taken very largely from Dr Lynch’s written closing submissions, which unless otherwise recorded, I did not understand to be contradicted.

HP held a board meeting (by telephone) on 5 August 2011. The board presentation:

summarised the strategic context as being (a) “‘Big Data’ explosion” bringing “huge growth in unstructured data not suitable for traditional relational database analysis”, (b) “opportunity to integrate structured + unstructured data” and (c) “shift from historical analysis to real-time, predictive analytics” ;

summarised the strategic rationale of the choice of Autonomy as being its “platform for unstructured + structured data analytics” using IDOL (described later in the document as “massively scalable” ); its ability to “process all content types (e.g. Structured, Text, Audio, Video)” and its successful transition to “cloud product offerings and business model” ;

summarised the strategic opportunity as being that “HP can build a leading position in enterprise software via [Autonomy] acquisition, leveraging [Autonomy’s] analytic capabilities with HP’s brand, market reach and other technology assets” ;

described the focus of due diligence as being “on validating synergies, understanding key product capabilities and retaining key executives” ;

gave a DCF stand-alone value for Autonomy of $10.012bn and implied stock price of $38.61 per share, compared with a current EV (on the basis of the shares’ market price) of $6.229bn and $25.19 (£15.44) per share. The value with synergies was $17.376bn, yielding an implied stock price of $64.74;

identified “Potential revenue synergies from (a) Information Management (b) Unified Analytics (c) Document Processing Solution (d) Data Security Solution and (e) Channel geo-expansion strategy” ; and

summarised the analysts’ recommendations. Most were buy recommendations, although there were three holds and two sells. Mr Apotheker accepted in cross-examination that he would have paid more attention to the bigger houses such as Goldman Sachs and BNP Paribas than houses such as Peel Hunt (a sell recommendation), of which he had not heard.

The 2011-2016 projections in the board presentation were in part based on “Wall Street Research”. HP used a combination of techniques in estimating future revenues, including analysts’ forecasts.

The board presentation document highlighted the concern about the interloper risk. This impacted significantly on HP’s approach to the deal:

Potential interlopers were identified. Of those, Mr Apotheker confirmed that “two companies were looked at in a little bit more detail, that was Oracle and IBM.”

The document reiterated the requirement from Rule 20.2 of the Takeover Code for Autonomy to share information with other bidders. As Mr Apotheker accepted, that had influenced the structuring of the due diligence process.

The document summarised the strategic considerations arising from the interloper risk. It read:

As Mr Apotheker explained, a “compelling offer” meant “ basically an offer where you would pay the highest logical price that still made sense in order to discourage everybody else ”.

The final price was negotiated at a time of generally increased market volatility and against a steep decline in Autonomy’s share value (Footnote: ³⁹) (and that of the market generally). Whereas the price range discussed in Deauville (a range of £24.94 to £26.94) represented a premium of roughly 44% to 56% at the time it was agreed, by the time of HP’s Board Meeting on 5 August 2011, the premium had grown to a range of 61.5% to 74.5%. Although 30-, 60-, 90- and 180-day calculations of implicit premium were lower, and it needs also to be remembered that the calculations reflect gross premiums that did not take into account Autonomy’s significant available cash, the decline in Autonomy’s share price inevitably triggered concern, and prompted consideration by Perella Weinberg and the board as to how best to approach the possibility of a reduction in price.

On 7 August 2011, Perella Weinberg wrote to Mr Apotheker and Mr Robison to raise the possibility of seeking to reduce the acquisition price by £1 per share in light of the declining macroeconomic environment, the 10.1% decline in Autonomy’s share price since 28 July 2011, and the 11.2% decline in the infrastructure software comparable group and 9.9% decline in HP’s share price over the same period:

Perella Weinberg reiterated that they believed in HP’s strategy of transforming HP through the separation of PSG and the Autonomy acquisition and also believed that “ Autonomy was a unique asset and is also uniquely capable of transforming [HP’s] business ” (as Mr Apotheker agreed).

Perella Weinberg also noted that the acquisition was likely to be controversial regardless of the price paid, which also reflected what had been said at the HP July board meeting.

Mr Apotheker and Mr Robison expected an extreme reaction from Dr Lynch if they were to try to reduce the price significantly. In an unchallenged passage in his witness statement, Dr Lynch stated that he now understands “ that HP contemplated offering a price decrease in August, but that Messrs Robison and Apotheker both knew I would be unwilling to support a bid below £25 per share, and thus did not amend their offer ”.

Perella Weinberg followed up with a further email to Mr Apotheker on the same day, 7 August 2011. The email noted the risks of the investor reaction. Mr Weinberg’s central paragraph essentially suggested that the status quo was not a practical option, and despite the difficulties, proceeding with both the PSG divestiture and the Autonomy acquisition was the way forward:

Mr Apotheker agreed with this analysis. He also wanted to announce the two transactions together “if feasible and possible”.

Mr Robison was sent to London to have further discussions on price and seek to close the deal. Mr Robison agreed with the Perella Weinberg advice, like Mr Apotheker. Mr Robison also thought that Dr Lynch would not go below £25 per share; and he also thought it was “ very possibly now or never” :

On 12 August 2011 there was a further HP board meeting. The revised DCF is at p.15 of that presentation. It gave a revised stand-alone value for Autonomy of $9.502bn (revised from $10.012bn from the 5 August 2011 presentation), and a value with synergies of $17.098bn (revised from $17.376bn).

At that board meeting the board unanimously agreed that HP should continue to pursue Autonomy but in a price range of £25-£25.50, which was towards the bottom quarter of the £24.94 - £26.94 range agreed between Mr Apotheker and Dr Lynch before the start of due diligence. This was after a board discussion, and in the light of the falling share price. The Defendants stressed that:

First, this range (at least at the top end) was still within the range previously agreed with Dr Lynch.

Secondly, the implied premium was now considerably higher than that which was implied when the range was agreed with Dr Lynch. Autonomy’s shares had declined in price. According to the updated DCF valuation presented to the board on 12 August 2011, Autonomy’s market price implied that its enterprise value was now $6.35bn.

Mr Apotheker had a final negotiation with Dr Lynch on 14 August 2011. Dr Lynch wanted £26 per share. HP countered with £25. The parties settled on a price of £25.50. This was still within the range previously agreed with Dr Lynch, albeit at the bottom of it; and it was within the range given by HP’s board. Mr Apotheker confirmed with Mr Lane directly afterwards that he was content with this, though of course that was subject to board approval.

Bidco was incorporated on 15 August 2011. HP’s 18 August 2011 offer to Autonomy’s shareholders stated that:

By the time Bidco was incorporated, the due diligence process was almost complete, and the price for the acquisition had already been agreed between Dr Lynch and Mr Apotheker. Bidco had played no part in the process.

Bidco’s directors at the time of the acquisition were Ms Lesjak, Mr Paul Porrini (HP’s Deputy General Counsel for Corporate Securities and M&A, who led the legal due diligence team) and Mr Sergio Letelier (an in-house HP lawyer). There was no evidence from any of these as to the matters that influenced the board of Bidco in deciding to proceed with the acquisition. Mr Porrini and Mr Letelier were not called by the Claimants. Ms Lesjak gave no evidence about what she did or thought in her capacity as a director of Bidco. As the CFO of HP, she was actually opposed to the transaction (see further below). She gave no evidence of having relied on any of the published information or any misrepresentations.

That has led to a dispute (“the Bidco point”) as to whether the FSMA claim must fail in any event: see paragraphs 484 to 500 below.

HP had a series of board meetings in the lead up to the final approval of the acquisition on 18 August 2011. There were meetings (by telephone) on 12 August 2011 (referred to above), and further meetings on 16 August 2011 and 18 August 2011 (both by telephone).

The minutes of the meeting on 16 August record that both Mr Frank of Joele Frank, Wilkinson Brimmer Katcher (a PR consultancy in New York) and Mr Weinberg of Perella Weinberg warned that:

The minutes record that after discussion the board’s consensus was to proceed with the divestiture and acquisition. Although this is unclear from the minutes, this consensus was reached despite Ms Lesjak speaking against the acquisition at the meeting, supported by Mr Holston, HP's General Counsel. Ms Lesjak’s unchallenged evidence in her witness statements was that she expressed reservations about the timing and price of the acquisition (but not, she emphasised, to the expansion of HP’s software business, nor the projected synergies). Mr Apotheker’s evidence was that Ms Lesjak said that “HP should be deploying its capital elsewhere and could not afford the proposed transaction at that time” but contrary to a suggestion made by Dr Lynch (who was not, of course, there) he had no recollection of her having expressed concerns about the synergies or the prospect of their realisation.

The depiction which appears to be suggested by Mr Apotheker and Ms Lesjak of friendly and constructive debate is not supported by the documentation, nor, on closer analysis, by Mr Apotheker’s evidence. There were increasing tensions between Mr Apotheker (the CEO) and Ms Lesjak (CFO, though not a director of HP) suggestive of real personal animosity.

In cross-examination, Ms Lesjak (who had worked for HP for nearly 30 years and plainly feared she was about to lose her job) confirmed her evidence in the US criminal proceedings that she had “pushed back” at Mr Apotheker’s suggestions of multiple acquisitions (TIBCO, Software AG and Autonomy) and told Mr Apotheker that HP “could not afford all three”. She said their relationship was “strained” from at least May/June 2011, and “a little war” had developed between them especially after a row between them about the inaccuracy of her “flash” forecasts. At the board meeting on 16 August 2011 both Ms Lesjak and HP’s General Counsel spoke out against his plans without prior warning: when asked whether he was furious with her, he told me that that was a “fair way of putting it.” That was even before he was shown emails she sent to the Chairman during the meeting itself (so she told me), complaining in the first about his lack of interest in financial detail and the lack of any “real plan” or “financial discipline” and in the second sending a one-lined message:

Despite this sniping, and what Mr Apotheker described as “a poisonous internal environment between Cathie Lesjak and myself” , Mr Lane at that time continued to support the acquisition of Autonomy and so did the board.

However, after the meeting on 16 August 2011 Mr Lane was concerned about the acquisition. He arranged for a call between the non-executive directors on 17 August 2011 to reconsider matters. On 17 August 2011 Mr Lane also emailed the non-executives to report a conversation he had had with Mr Apotheker. The email was marked ‘For your eyes only’ and related that he had told Mr Apotheker that the Board “were shocked by Cathie’s statements (and Mike’s)”, were “very concerned about how poorly this announcement will be perceived and how it will reflect on Leo and BOD” and would be “relieved” if Mr Apotheker decided to defer the Autonomy announcement. It added that “many of us believe it will still be there in a month or two.”

The same email described Mr Apotheker’s response, to the effect that the Autonomy acquisition was critical, he did not think he had an alternative to begin the change to higher value businesses, and he did not believe that the opportunity would still be there if HP did not take it now. After stating that Mr Apotheker was “furious with Cathie and believes she can’t be counted on…” he went on as follows:

There was a further board meeting on 18 August 2011.

In the board presentation:

The situation update gave the new agreed offer price of £25.50 per share. This implied a 64% one day and 58% 30 day average premium. At this price, the diluted equity value was £7.1bn ($11.7bn). The implied enterprise value was £6.7bn ($11.0bn).

The DCF page gave more details of these figures and the valuation:

The latest valuation of Autonomy gave a stand-alone value of $9.502bn, and $17.080bn with synergies.

The current market price implied enterprise value was at $6.365bn (at £15.58 per share). HP’s offer price of £25.50 accordingly represented a 64% premium (enterprise value of $10.99bn).

The board were reminded that once a firm intention to make an offer had been announced, HP would be obliged to make an offer on the terms described in the announcement.

The board were also told that there was a high potential that the rating agencies would put HP on negative outlook or downgrade HP, based on the expected leverage, strategic evaluation and near-term benefit to company performance. This followed from the advice at the previous meeting as to the prospects of an adverse shareholder reaction. Mr Apotheker confirmed that the board were aware of these risks, and that the board understood that it would have to hold its nerve in the face of these negative reactions.

On the same day:

HP’s Technology Committee confirmed their recommendation with respect to the technology aspects of the transaction.

BarCap provided a fairness opinion and a supporting presentation. BarCap explained under the “strategic rationale” that Autonomy was the best-in class asset to address the market opportunity in enterprise information:

BarCap’s DCF was the same as HP’s. They also set out a trading analysis of comparable software companies. All the companies featured had considerably lower operating margins than Autonomy’s adjusted operating margins (shown at 42.7%), Software AG was stated at 27%, TIBCO was 26.3%, the mean operating margin was 23.7% and the median was 26.3%. Mr Apotheker regarded anything upward of 30% as high margin. (Footnote: ⁴⁰)

Perella Weinberg also gave a presentation and a supporting fairness opinion. Their DCF again replicated HP’s. Perella Weinberg set out the anticipated growth of the stored information market, which represented a “massive market opportunity”. Perella Weinberg noted Autonomy’s unique capabilities and its centrality to HP’s strategy:

The board unanimously approved the acquisition at £25.50 per share.

HP announced the takeover for $11.1 billion on 18 August 2011. HP’s shareholders had to come to terms not only with the acquisition of Autonomy for $11.1bn, but also with the potential spinning off of HP’s personal computer business (in effect a legacy business), and HP’s disappointing Q3 results.

The announcement was not well received. A large and vocal section of the shareholders disapproved and the visible reaction was overwhelmingly negative. Moody’s downgraded HP’s outlook from stable to negative. HP’s share price fell sharply (20% in a day). Shareholders lobbied the Chairman, directors and management to see whether it was possible to get out of the bid before it closed. Mr Lane’s nervousness turned to fright and HP’s board instead suffered a collective loss of nerve. Mr Apotheker broadly accepted this:

Mr Apotheker was still optimistic that he could steer a course through the difficulties and wanted to get on with the integration. He remained “completely convinced about the validity of our approach.” He exchanged emails with Dr Lynch who encouraged him with nautical analogies to weather the storm. Mr Lane, however, had in effect abandoned ship.

On Sunday 4 September 2011 Mr Lane emailed Mr Apotheker with obvious signs of a revisionist approach, claiming “still” to be “haunted by Autonomy itself” and adding:

He concluded by asking Mr Apotheker to “analyze for the board” :

First, “ whether there is any way to get out of the Autonomy deal ”.

Secondly, what size of buyback HP could launch to try to recreate lost value for shareholders.

Mr Apotheker was sceptical about the buyback idea:

As for the idea of getting out of the deal, Mr Apotheker responded by email the same day to Mr Lane that he would ask HP’s advisers to look at the questions but also robustly defended the merits of the acquisition, which Mr Apotheker still believed in. Mr Apotheker reminded Mr Lane that the Autonomy acquisition was the only strategic solution that HP had:

Mr Apotheker followed up the next day (5 September 2011), seeking to clarify some of the “noise around Autonomy” and reminding and stressing to Mr Lane that HP’s business case hinged more on the synergies that the combined companies could generate than on Autonomy’s stand-alone capabilities:

There was little in the evidence I was shown to illustrate and explain what I assume was a further deterioration of support for Mr Apotheker and the Acquisition in the period between 5 September and 21 September 2011, except that between 31 August 2011 and 15 September 2011, a series of three emails were posted to some 50 analysts and the press by Mr Harald Collet (“Mr Collet”, who had been Head of OEM Sales in North America from May 2008 to June 2010) and his former Autonomy colleague, Mr Marshall, under the pseudonym “Joe Bloggs”. The analysts selected appear all to have been Autonomy sceptics already. Mr Collet and his colleague “suggested a number of key questions HP should be asking about Autonomy’s OEM revenues .”

These emails (“the Joe Bloggs emails”, which were sent from a specially created “Joe Bloggs” email account) followed an earlier email dated 23 August 2011 from Mr Collet personally to Ms Leslie Owens at Forrester, an analyst who had written a blog post critical of Autonomy on 19 August 2011, in which he set out a case that Autonomy’s OEM revenues were too good to be true, arguing that there was an “ almost ‘Madoff-like’ consistency in their quarterly number of new OEM signings ”. The Joe Bloggs emails were to similar effect. They appear to have encouraged further opposition to the Acquisition from analysts already sceptical of it, but their impact on sentiment and on HP is difficult to gauge. (I address a communication sent by Dr Lynch to HP following the Joe Bloggs emails later when dealing with the OEM case.)

In any event, on 21 September 2011, a story circulated on Reuters wire service that HP was considering removing Mr Apotheker and replacing him as CEO with Ms Meg Whitman (who had run eBay). On 22 September 2011, well before the acquisition had completed, Mr Apotheker was removed summarily by HP’s board. His witness statement gave an anodyne version of events. The first that he heard of this was from a Bloomberg journalist; Mr Apotheker was shocked by the news. He was permitted to speak to the board for an hour, but that was pointless, as he was told at the end of the presentation that the decision was already made to appoint Ms Whitman as the new CEO. Soon after her appointment, Ms Whitman also asked Mr Robison to leave.

Ms Whitman’s background was not in software, a business which in Mr Apotheker’s opinion took a few years to understand. She described herself to me as “more a consumer technology executive than an enterprise technology executive” . Ms Lesjak, whose future had depended on Mr Apotheker’s failure as CEO, continued as CFO.

Ms Whitman was cross-examined about the reasons for the sudden decision to remove Mr Apotheker. She suggested that it was not because, or certainly not only because, of the “adverse market reaction” . She said it was because Mr Lane had interviewed a number of HP executives “who were very concerned about his leadership style” and because he had spent the first two months of his appointment out of the USA. She told me “so what it seemed like was that we needed new leadership for HP. So he was asked to leave.”

This seemed to me to be rather jejune. It was plain and obvious that Mr Apotheker and Mr Robison were sacked because the strategy of which they were the architects had proved unpopular amongst investors, and HP needed scapegoats and to demonstrate that the strategy had been abandoned. As she stated in an email a little later (on 14 December 2011), when Mr Apotheker had dared publicly to say that, at the time, the Acquisition had been supported by the Chairman and the whole board:

Although in cross-examination she described the changes as “largely tactical…a bit more of a stabilisation” following her appointment Mr Apotheker’s strategy was reversed. HP determined not to divest the consumer PC business and reaffirmed its primary commitment to hardware, which she described to analysts as the “DNA of this company” . Ms Whitman told the markets that HP were instead going to refocus on the existing core assets of the company, and she regarded her main and most immediate task as being to “stabilise the core assets” .

Ms Whitman stuck to the line that she remained committed to the Autonomy acquisition. She denied saying to an executive committee meeting that Autonomy was an “unwanted stepchild” and suggested that it was a phrase in fact used by Dr Lynch. (Footnote: ⁴¹) But for all her cavilling, the impression I formed was that she fell in with HP’s corporate antipathy to any culture other than its own, and that the reference to Autonomy as an “unwanted stepchild” is how she regarded it, even if it was forensically not something to be admitted to; and although the Claimants stressed that she had initially maintained the Integration Steering Committee established by Mr Apotheker and Mr Robison and appointed Mr Brossard (whom she regarded as one of her most capable lieutenants) to focus on the process of integration, the low priority she set for Autonomy’s welfare, and her predisposition to require Autonomy to conform to HP’s way of doing things, was vividly illustrated by the abandonment at her direction of Mr Apotheker’s plans for HP Software to be headed by Dr Lynch who would under those have had key oversight of the integration process, with Autonomy at the centre.

The change of strategy is addressed further below. Understandably, Mr Apotheker disagreed with the change of strategy:

As regards the Autonomy acquisition, under the UK takeover rules HP were stuck with the bid they had announced, which became unconditional on 3 October 2011.

After the Acquisition, as would always have been anticipated, HP had unrestricted access to Autonomy’s full books and records. HP’s advisors also had access to Deloitte’s working papers. Ernst & Young (“EY”) reviewed them.

In the event, HP and its advisors did in fact discern many of the matters now complained of. They did not cause any great concern at that time.

Of note in relation to HP’s hardware case, it does seem that from early on in the aftermath of the Acquisition, they were aware of Autonomy’s hardware sales. This is a point explored in more detail later (see paragraphs 1814 to 1854 below), but in brief summary:

Following the acquisition, KPMG were engaged to produce a closing balance sheet in respect of Autonomy for HP. This resulted in a draft report of 24 October 2011. At page 7, KPMG identified $41m payable to Dell, and stated:

KPMG also conveyed the same point in its final draft of the report in March 2012, noting:

Separately, EY in their capacity as HP’s auditors conducted a review of Autonomy’s 2010 audit, including Deloitte’s working papers. That work was complete by 4 November 2011, and EY specifically noted that Autonomy had about $100m in hardware revenue.

EY also produced a slide presentation entitled “Q4 FY’11 CFO Update” for discussion at a pre-meeting with Ms Lesjak on 11 November 2011 as part of the preparatory work in advance of an HP Audit Committee meeting the following week. (Footnote: ⁴²) There were only two substantive pages to the document. One of the four Q4 areas of focus identified in the Executive summary (on page 2) was the Autonomy acquisition. That item was covered on page 3 of the document. Four points were identified in respect of the Autonomy acquisition. One of them was that “ Revenue includes $115M of hardware ”.

Ms Lesjak was defensive about this when cross-examined. I return to consider her evidence as to what she made of the point in paragraphs 1835 to 1837A and 1848 to 1852 below.

On 16 November 2011, EY presented their results to HP’s Audit Committee: Page 7 of that document gave a revenue breakdown (in the form of a bar chart) of the $1bn “Autonomy portfolio”. It broke revenue into Services, Support, Cloud, Licences and Hardware, attributing 11% to hardware. That information does not appear to have given rise to any question or surprise, nor created an issue for anyone at HP. Ms Lesjak accepted that it is likely that she did “flip through” the document but she insisted that she “didn’t spend any time in the audit committee deck on the Autonomy section”. She did not deny that EY went over the relevant page with the Audit Committee but she suggested that the Audit Committee’s focus was on the “purchase accounting” and that the committee “don’t go over every item on the page”. In light of its prominence, I cannot accept it went unnoticed: again, however, it caused neither surprise nor concern.

Furthermore, hardware purchases from Dell for resale continued through to at least March 2012. There was no secret about the purchases or the sales. Upon the appointment of Mr Christopher Yelland (“Mr Yelland”) as CFO, but whilst he was still working in another division of HP, he was asked to approve a purchase of hardware. He was told explicitly in that email, which attached a spreadsheet of the relevant figures, that the hardware would be sold through to customers at a loss of 10%.

Mr Yelland had not yet taken up his post, but he acknowledged the email and asked to be copied in on requests going forward, as he was. He accepted in cross-examination that he had read them and told the Court that he was interested to see the sorts of emails for which his approval was required, even prior to his formal arrival in post. In April 2012, Mr Yelland approved another purchase. That email again stated in terms that Autonomy would “sell-through” the hardware to customers at a loss of approximately 10%. Mr Yelland did not raise any objection to the “sell-through” hardware sales, or suggest that there was any problem with the practice at the time.

There was a similar situation with hybrid hosting deals. HP was aware of the hybrid hosting deals and its initial approach, far from putting a stop to them, was to seek to establish VSOE under US GAAP so that revenue from licences could be recognised up front (i.e. in the same way as Autonomy had always done). The VSOE workstream started in early 2012 and continued into at least May 2012. This too is addressed further below.

The integration of Autonomy into HP’s business was bound to be difficult. I have already noted that HP’s business was fractured in ‘silos’ with intra-division jealousies, poor co-operation and overall direction, and short-term focus. The Defendants described it, with (as it seems to me from the available documentation) good reason, as “a huge, enervated, company, beset with systemic failings” . Autonomy was, as it seemed to me, almost the opposite in terms of the way it was managed: centralised, energetic and driven by entrepreneurial zeal, with little formality or adherence to strict process. HP had become an ailing institution. Ms Whitman’s own report to the HP Board dated 22 March 2012 (entitled “The Road Ahead…”) spoke of a “crisis of confidence among all constituents”, “significant business challenges across every Business Group, HP Labs and Global Sales”, “Suboptimal business processes and lack of focused strategies…unsustainable cost structure” and acknowledged that “HP culture is in tough shape, must be revitalised/realigned to be successful”. Autonomy, on the other hand, retained the outlook of a start-up still under the direct control of its founder.

HP’s earlier efforts to expand by acquisition had been problematic and unsuccessful:

HP bought Electronic Data Systems for $13.9bn. It was rebranded as HP Enterprise Services in 2008. Its value was written down in Q3 2012 by some $8bn.

Other acquisitions are listed in HP’s M&A Scorecards, such as that for Q2 2012. One well known one was the Palm hand-held computer business which it had acquired for $1.2 billion in 2010. The acquired businesses being tracked had under-performed against budgeted revenue by $1.73bn or 26%.

The difficulties of integrating Autonomy were exacerbated by (a) the abandonment of the predicate of its acquisition as envisaged by Mr Apotheker, being the strategic reorientation of HP towards software, (b) the removal of the architects of that strategy, Mr Apotheker and Mr Robison, who were invested in its success, (c) the outlook of their replacements, whose focus was on rebuilding the old HP and its core hardware business, and who had little time to devote to Autonomy, (d) the lack of any properly considered integration plan and the abandonment of Mr Apotheker’s plan for Dr Lynch to head HP Software and drive forward the constructive integration of Autonomy and (e) a culture clash between the two companies.

Perella Weinberg had consistently stressed the importance of flawless execution and management devotion to the integration and transformation:

As Mr Apotheker put it:

The Defendants emphasised the problems which arose without oversight and planning and in a context where (as they saw it) Autonomy was regarded as an unwanted distraction foisted on a group already in difficulty, including (again, as they perceived it):

The failure to establish proper processes for other business groups to sell Autonomy products and worse, infighting between business groups and the continuation of incentives to other business groups to sell third party software and not to sell Autonomy products;

The difficulties which Autonomy experienced in acquiring hardware from HP group’s ESSN unit (Footnote: ⁴³) to sell with Autonomy products because the ESSN was remunerated according to sales quotas, and supply to Autonomy did not count towards quota. Dr Lynch acknowledged in cross-examination that Ms Whitman, in this instance and others, stepped in to fix the supply: but the fact that her intervention repeatedly had to be sought illustrated that Autonomy was not regarded or treated within HP as part of the family;

The removal from Autonomy of its control over long-term relationships with established customers when HP Enterprise Services (“HPES”) successfully lobbied HP management to take control, which in turn led to HPES adding a 35% mark-up on Autonomy prices to boost the reported revenues (and thus the bonuses) of the HPES team. (Ms Whitman did eventually intervene to stop this, but in the meantime Autonomy lost customers.)

Autonomy’s ostracization from HP marketing initiatives such as company gatherings and trade shows, whilst HP continued to insist that all marketing should be centralised and that Autonomy should not have any marketing strategy of its own;

Autonomy’s increasing problems with staff retention, and resulting loss of institutional knowledge and industry experience in consequence of (a) HP’s bureaucratic process-obsessed outlook and stifling corporate culture; (b) HP’s failure to recognise the experience and talent of Autonomy’s key legacy employees, and corresponding failure to put in place promised staff incentive plans although it is obvious that the key asset of any software business is the talent of its employees. The Claimants countered this with evidence from Mr Youngjohns that Autonomy’s attrition rate was in fact lower after the Acquisition than it had been before it, and that various key Autonomy employees including Mr David Jones (CEO and General Manager of Data Protection), Mr Rafiq Mohammadi (CEO and General Manager of Promote), Mr Mike Sullivan (CEO and General Manager of Protect, Enterprise Markets) and Mr Neil Araujo (CEO and General Manager of Protect, Professional Markets) all stayed on for a number of years after the Acquisition and (according to Mr Youngjohns) “only left to take up CEO positions elsewhere.” The fact remains, as I find, that the loss of senior staff in May 2012 is well documented; and Ms Whitman’s approach was less than constructive: and see also paragraph [360] below;

What Dr Lynch described as “flip-flopping” by HP in first agreeing to facilitate VSOE (see paragraphs 273(2) and 328 above), which is a technical accounting exercise required under US GAAP for the purposes of a profitable business product of Autonomy called “hybrid hosting” , but then changing its mind, causing customer confusion and dissatisfaction, because it would have constrained hardware discounting;

Other integration failures including (a) changes to the structure of Autonomy without consultation, warning or explanation (b) bureaucratic practices such as refusing to supply Autonomy with HP hardware for sale to Autonomy customers using its Cloud offering until Autonomy was an “approved vendor” and (c) denial to Autonomy of crucial sales and services facilities and personnel. As to (c), the Claimants drew to my attention evidence to the contrary, showing that HP did arrange the transfer of more than the target number of 42 HP employees to work at Autonomy: but I accept and find that the problem of retaining skilled staff remained a difficulty which was exacerbated by HP’s treatment of Autonomy leading to a sense that jobs were at risk, and its own stifling corporate culture. Mr Youngjohns seemed to me largely to accept this when cross-examined, though he maintained that he did his best to reassure staff that HP was going to build a successful business;

What Dr Lynch saw as a lack of co-operation from HP management hindering the integration within Autonomy of the newly acquired Vertica and the Information Management business unit, which had been an essential part of Mr Apotheker’s concept of a unified information stack;

The imposition of Ms Whitman’s appointee, Mr Bill Veghte, as head of strategy and software (including cloud and e-security) with a view to the eventual absorption of Autonomy within HP Software, in a reversal of the understanding before the acquisition that Mr Veghte would be leaving imminently and Dr Lynch would replace him and be put in charge of HP Software (as well as Autonomy);

What Dr Lynch in cross-examination called “a horrifically political environment”.

On 14 February 2012, Dr Lynch emailed Mr Hussain and Mr Kanter stating “Meeting meg 2 morrow, need all craziness examples”. Mr Kanter replied listing problems that Autonomy was facing in its relationship with HP, including incentives which encouraged HP employees to sell competitors’ products; difficulties in procuring hardware internally; difficulties in accessing HP systems; and a sense of HP seeking to overwhelm Autonomy at internal meetings, with more than 25 employees of HP for every one from Autonomy; and the lack of any overall direction for integration.

Ms Whitman sought in cross-examination to brush all this off as Dr Lynch presenting himself as a “victim” without trying to be “part of the solution” and questioned why he did not try to work with the person she had appointed to help on integration, Mr Gerard Brossard (“Mr Brossard”). She depicted it as Dr Lynch’s “failure to adapt to his new role at HP” and symptomatic of someone used only to “leading a small group of dedicated start-up employees . ” But again, I regret to say that I found this formulaic and (given Autonomy’s past growth and success and HP’s own problems) condescending.

The truth as I see it and find is that Ms Whitman had some time previously, and before Autonomy began to show worrying performance shortfalls, begun to look at a reversal of the model envisaged by Mr Apotheker, and instead of having Autonomy, led by Dr Lynch, at the apex, the absorption of Autonomy within HP Software under the direction of Mr Veghte (whom she had appointed head of strategy in January 2012 and had ultimately retained as head of Software): and see also paragraphs 364 to 366 below.

By February 2012 (4 months after the Acquisition went unconditional on 3 October 2011) Autonomy had fallen materially behind forecast revenue: HP’s February “Flash” results suggested a revenue shortfall of 15% compared to budget, whilst noting that “Pipeline and underlying business remain strong” and also that there were “Very strong cloud signings leading to deferred recognition.”

A series of emails from Mr Hussain to Dr Lynch in April 2012 reveal a deteriorating position. By 27 April, Mr Hussain was reporting “Revenue has collapsed on us. I am very sorry Mike.” On the day before the quarter end, Sunday 29 April 2012, Mr Hussain sent an email to Dr Lynch, timed at 12.07pm, with the subject “ revenue ”, stating as follows:

Thus, Mr Hussain identified the underlying cause of the revenue shortfall as being Autonomy issues: “ poor sales rep productivity ” and the “ sms process not picking it up ”. Dr Lynch took a different view of the causes. In cross-examination, Dr Lynch accepted that he instructed Mr Hussain to produce a further email which he (Dr Lynch) could share with HP’s senior management:

Mr Hussain duly complied with that instruction. At 1.15pm, just over an hour after sending his first email on 29 April 2012 referred to above, Mr Hussain sent a second, freestanding email to Dr Lynch, with the subject “ Updates – not good ”, stating as follows:

The Claimants emphasised the difference between Mr Hussain’s first email (which did not seek to place any blame on HP) and his second email (which did); and they criticised Dr Lynch for having both “instructed” Mr Hussain to write it, and for (they implied) ensuring that it was drafted in such a way as to suggest that Dr Lynch was until then unaware of Autonomy’s second quarter revenue shortfall. They also criticised him for amending the first sentence in the version he sent on to Ms Whitman and Ms Lesjak by deleting the words “very sizeable” and for then dissembling as to the reason for the amendment.

I note, but do not attach great weight to, these criticisms. I do not accept that either of the Defendants was thereby intentionally “papering the record” (which was what Mr Rabinowitz suggested in his cross-examination of Dr Lynch). I can understand why Dr Lynch wished Mr Hussain to reflect the difficulties which he considered were of HP’s making as well as those which Mr Hussain had earlier fixated on. Mr Hussain was self-critical to a fault, and apt to lose the broader picture. I can also understand the reason for the surreptitious amendment made by Dr Lynch to Mr Hussain’s email when forwarding it to Ms Whitman, even if he should not have done it (as I readily accept).

More important was the effect on Ms Whitman and the repercussions of the large shortfall. In her evidence, Ms Whitman described herself as being “ completely blindsided ” by Dr Lynch’s email. She emailed Dr Lynch to seek details of the scale of the shortfall (“ How big is the miss? ”), which in response he estimated at anywhere between “ 7-15% depending on individual deals ”. He suggested waiting a day, because some “could be holding out for last minute negs.”

Ms Whitman expressed herself “shocked at Dr Lynch’s lack of urgency: it was Sunday, April 29, and the quarter was scheduled to close the next day.” She felt, ignorant perhaps of the last-minute nature of many such deals in the past, that it was “ completely unacceptable conduct ”. She arranged for a conference call to take place later that same day.

The conference call took place at around 10.15pm UK time on the Sunday evening, which both Dr Lynch and Mr Hussain joined. (Footnote: ⁴⁴) On the call, Dr Lynch indicated that Autonomy was likely to miss its revenue target by around $30 million to $50 million. He put this down to two reasons, both of which, perhaps predictably, he said were HP’s fault: first, he said that HP’s accounting rules prevented Autonomy from being able to recognise all of the revenue on the sales it was making as quickly as he had expected; and secondly, many of Autonomy’s customers were dissatisfied with HP and were refusing to engage.

It appears that having given this explanation, Dr Lynch was then keen to ensure that others within Autonomy should provide a consistent explanation for the shortfall if asked. Before the call had even finished, Dr Lynch emailed Mr Hussain, with a high importance marking, with the subject “ RING STOF ASAP AFTER CALL ”, stating “ Need tog et [sic] stories straight ”. Mr Hussain immediately responded “ ok ”. In other words, Dr Lynch instructed Mr Hussain to call Mr Egan immediately following the call so as to ensure that, if he was asked about the revenue shortfall, Mr Egan would give the same explanations as Dr Lynch had provided during the conference call. Dr Lynch did not dispute that in cross-examination.

The Claimants submitted that patently, if the explanations that Dr Lynch had given during the conference call accurately reflected the reasons why the revenue shortfall had occurred, there would have been no need for Dr Lynch to have been so concerned to ensure that Mr Egan, Mr Hussain and he got their “ stories straight ”. The Claimants added to this that it was notable that no copies of Dr Lynch’s “ Need tog et [sic] stories straight ” email (which was disclosed by Dr Lynch, seemingly from the server which he used for his own personal storage purposes) were retained on Autonomy’s systems – neither the copy sent by Dr Lynch from his Autonomy email address, nor the copy received by Mr Hussain at his Autonomy email address. Nor were any copies of Mr Hussain’s response to that email (which was again disclosed by Dr Lynch) retained on Autonomy’s systems – neither the copy sent by Mr Hussain from his Autonomy email address, nor the copy received by Dr Lynch at his Autonomy email address.

The Claimants submitted that the only plausible explanation for this state of affairs is that Dr Lynch and Mr Hussain must both have deleted their respective copies of both emails from Autonomy’s system, and that cannot have been a coincidence, so they must have agreed to do so. In cross-examination, Dr Lynch suggested an explanation was that his emails were not kept on Autonomy’s system. The Claimants submitted that this fails to account for the fact that Mr Hussain’s copies of the emails were not retained on Autonomy’s systems either, nor for the fact that (a) Dr Lynch was the custodian for more than 15,000 emails or documents on Autonomy’s systems and (b) a number of other emails between Dr Lynch and Mr Hussain from around this time were retrieved by the Claimants.

The position is certainly suspicious; but I prefer to make no finding in respect of it, given that the question of where and how Dr Lynch’s internal emails were stored was not further pursued with him, and it seems inherently unlikely that either of the Defendants would have thought it important enough to remove traces of the particular exchange from the system, not least bearing in mind their equanimity in respect of other much more damaging records to which HP had full access.

In any event, I am not persuaded by what I take to be the Claimants’ suggestion that their wish to present a consistent explanation of itself shows that the Defendants perceived it necessary to and did combine to concoct a reason for the shortfall which they knew belied the truth. The Defendants perceived themselves to be beleaguered in a highly political and antagonistic corporate environment, and the revenue shortfall was something for which they would need to have a consistent explanation; but that does not necessarily mean that the explanation was manufactured or untrue.

Indeed, at the time, there was in reality little between HP and Autonomy as to the reasons for the shortfall, even though it transpired that Dr Lynch had seriously underestimated its size in his reports to Ms Whitman, and at $136 million it amounted to nearly three times the top end of the range Dr Lynch had indicated to her on the Sunday conference call.

Broadly they were agreed that the main problem was not lack of demand nor any overall market or sales funnel problem (Footnote: ⁴⁵) , but failure to convert orders for Autonomy’s proprietary software into completed sales, and that the effect was magnified by the fact that Autonomy had pinned their hopes on a smaller than usual number of larger than usual deals, which they had expected to close, but did not.

Where they disagreed was as to the reason for that failure to close licence deals, for which each blamed the other; and although Mr Kanter’s report dated 7 May 2012 placed the blame heavily on (a) VSOE issues (b) HP paperwork demands (c) Autonomy having to cede control of deals to other parts of HP and (d) “wider HP customer issues”, even HP’s own reports acknowledged that revenues were adversely affected by the challenges with operating Autonomy in the HP environment (as Ms Whitman also admitted when cross-examined) and that there was an urgent need to address these issues and improve integration of processes and sales function across all business units.

Certainly, Ms Whitman emphasised that at the time there was no suggestion of demand for Autonomy software, or its competitiveness, having been exaggerated. When, at HP’s Earnings Conference Call on 23 May 2012 (at which Dr Lynch’s dismissal was announced, see below) she was asked whether Autonomy’s “weakness in revenue…changes your outlook on the business at all and how it integrates with HP” she answered:

She elaborated in cross-examination:

However, what Ms Whitman later described as the “gigantic financial miss in Q2” , and its late disclosure by Dr Lynch, had important repercussions. Dr Lynch was placed on garden leave on the same day that the Q2 2012 results were announced.

Earlier, on 3 May 2012, Mr Hussain had sent a formal email of resignation to Dr Lynch. This stated:

The Claimants undertook a considerable exegesis in their closing submissions to demonstrate that this final version of Mr Hussain’s retirement email reflected input from Dr Lynch, as well as Mr Hussain’s wife. It seemed that the punchline was that Dr Lynch leant on Mr Hussain not to accept responsibility for the earnings miss, as in his first draft he had suggested he did. It was not clear to me where this point really took the Claimants. What is clear is that, after the earnings miss, there was no real prospect of HP countenancing Autonomy continuing under the pre-acquisition management, and that HP were intent on accumulating reasons why they should be removed entirely as soon as possible.

For example, when Mr Hussain sent an email to “senior leaders” of HP on 24 May 2012 drawing attention to the fact that some 157 staff had left or resigned from Autonomy in the previous weeks, which was then forwarded to Ms Whitman under the heading “Unacceptable email from Sushovan” , Ms Whitman’s immediate response to this was to send an email to Mr Schultz stating “ let’s garden leave this guy.” Mr Schultz actioned this immediately and branded what Mr Hussain had said as “highly unusual and… quite misleading.” In fact, as Mr Hussain demonstrated in his reply, the true total was 163. Mr Hussain also explained that his update reflected his practice previously of updating management on staff departures. The Defendants submitted, and I accept, that by this time, HP wanted rid of Autonomy’s old management, and dismissed as self-serving anything they had to say.

Despite Ms Whitman’s acknowledgement that many, if not most, of the difficulties lay in the problems of integration (Footnote: ⁴⁶) , late in the evening on 23 May 2012 HP’s general counsel, Mr John Schultz, sent a formal notice of termination by email to Dr Lynch, confirming Dr Lynch’s discussion with Ms Whitman earlier that day, and placing Dr Lynch on immediate garden leave for the remainder of his notice period. The reasons given for his termination were as follows:

Ms Whitman told me that by this time she had “ completely lost faith ” in Dr Lynch. She stated in her witness statement:

To this she added in her witness statement that what she had seen of Autonomy at the close of Q2 2012 was:

As I shall elaborate later, the wider truth is that Dr Lynch did not fit in with Ms Whitman’s concept of Autonomy’s place in the group. In this concept, neither Autonomy nor Dr Lynch and his colleagues were seen as providing a new and invigorating direction for a transformative change; and for some time before Dr Lynch’s removal, Ms Whitman had had in mind to “adopt” (her word) or (in reality) absorb Autonomy under the umbrella of and to make Mr Veghte (then head of HP’s own software business) the CEO of that part of the HP group. Thus, whereas before the Acquisition, Mr Apotheker had told Dr Lynch that Mr Veghte would leave and that he, Dr Lynch, would be put in charge of HP Software, by January 2012 Mr Veghte had been appointed as head of strategy and by March 2012 Ms Whitman had determined and made it clear that Mr Veghte was in charge of driving both the cloud and security businesses also. In an email to him dated 26 March 2012 Ms Whitman spoke of there being no one better suited than he for the role of providing “Strategic vision for HP and CEO partnership” and with particular reference to the Software part of the business she wrote:

Ms Whitman claimed, when cross-examined on the contents of the email, that she had not yet made a decision to bring Autonomy under the software unit. She spoke at length of the need for HP Software “to adopt and put their arms around this division to help, to help it grow and prosper and I had not made a decision at all about whether this business should actually report to Bill.” But the longer her answers became (and they were often in the nature of prepared pitches) the clearer it was that the direction of travel was to be rid of Dr Lynch, absorb Autonomy into the software division, and (instead of Dr Lynch becoming head of HP’s software business in place of Mr Veghte) retain Mr Veghte in place and through him ensure central control; and that the only real hesitation in her mind, as she came close to accepting, was whether Mr Veghte (whom Ms Whitman had also appointed as Chief Strategy Officer) would agree to the role.

The removal of Dr Lynch (who had himself begun to think, following the retirement of Mr Hussain, the resignation of Ms Eagan and the departure of a raft of Autonomy personnel, that “rapid assimilation back to the HP model” might be the only remaining option), was the concluding stage of the decapitation of the project of ‘transformation’ envisaged by Mr Apotheker, which had started with his own dismissal and that of Mr Robison. It was probably more symbolic than substantive in effect. It seems to me likely that Dr Lynch had already determined to resign before his termination, as demonstrated by emails on the same day (23 May 2012) discussing a new venture (Invoke Capital) that he was to join, and which Mr Hussain, Mr Kanter, Dr Menell, Ms Eagan, Ms Colomar and Ms Orton had already joined or were about to join (each with a percentage of its equity).

On 25 May 2012, two days after Dr Lynch had been placed on garden leave, Mr Scott, Autonomy’s Chief Operating Officer and General Counsel for the US, asked to see Mr Schultz to discuss concerns he (Mr Scott) had about Autonomy’s hardware reselling strategy which (according to Mr Scott’s evidence in the US criminal proceedings) an earlier telephone conversation between them had suggested Mr Schultz knew nothing about.

According to Mr Scott’s evidence in the US criminal proceedings, he had been “curious” and had felt uncomfortable about the sales; and, after Autonomy’s core management team had left in circumstances Mr Schultz told him suggested some impropriety, he felt he had to raise his concerns about them, and also about the structuring of hosting deals. (Footnote: ⁴⁷)

Be that as it may, after his conversations with Mr Schultz, HP commenced an investigation into the concerns Mr Scott had expressed, and retained for that purpose the law firm, Morgan Lewis & Bockius (“Morgan Lewis”). They have been continually involved since that time. PwC were also retained.

In parallel with the Morgan Lewis investigation, in June and July 2012, a “rebasing exercise” was conducted by Mr Yelland. Mr Yelland had become CFO of Autonomy in April 2012 after Mr Chamberlain (who had replaced Mr Hussain as CFO shortly after the Acquisition) handed in his notice.

In his evidence in these proceedings, Mr Yelland suggested that the exercise required him to review the “ economic substance ” of Autonomy’s business, (Footnote: ⁴⁸) and sought to use this as a platform from which in cross-examination to cast doubt on the “substance” of a wide slate of impugned transactions. However, it was clarified in the course of his cross-examination that the exercise was not intended to assess and determine whether past accounting was right or wrong under IFRS. The primary purpose was to establish what Autonomy’s business actually comprised and to strip out from the accounts revenues referable to business which HP would be unlikely or had decided not to continue, although Mr Yelland also maintained that he stripped out Autonomy’s hardware sales on the further basis that:

Mr Yelland’s team was supported by PwC. PwC’s email of 3 July 2012 (headed “Autonomy review – privileged and confidential” (and redacted in part for privilege)) noted that it understood that the analysis would be used in an acquisition valuation model, but both the analysis and the model was the “ responsibility of HP alone ”. (Footnote: ⁴⁹)

Mr Yelland’s exercise took a broad-brush approach. It did not look at the detail of individual transactions, nor did it attempt to consider the information that had been available at the time the accounting judgements had been made. Ms Meeta Sunderwala (“Ms Sunderwala”), a senior director in HP’s EFR team (Footnote: ⁵⁰) , noted on 18 July 2012 that Mr Yelland’s rebasing exercise involved “ a lot of estimation and judgement (not all built up from specific deals etc) ”.

Mr Yelland concluded the “rebasing” (sometimes “rebaselining”) exercise on 17 July 2012 when he presented his results in a spreadsheet and PowerPoint. These were summarised (in an Exec Summary ) as follows:

Resale of hardware accounted for $108m of the 2010 adjustments, $85m of 2011 and all $38m of FY12

The review of deals >$1m, a review of the acquisition balance sheet and a review of ‘exceptional’ costs booked in Sept 2011 identified the other half of the 2010 and 2011 impacts

Rebaselining required against the FY12 revenue is driven by hardware sales for which the accounting treatment is under review by GRRO. These are now stopped and other impacts are minimal

Licence rebased growth was a decline of 2% in FY11 and a decline of 16% for FY12 forecast

The OP (Footnote: ⁵¹) rate rebaseline did not change the 2010 OP of 36%, reduced the 2011 first 9mnths reported OP by 3pts to 21%, and increased the FY12 OP by 1pt to 21%

Note: the FY11 reported OP for the first 6mths was 34% but dropped in Q3 2011 to 5%.

The study does not consider the IFRS appropriateness of the accounting. This is under review by PwC. High risk areas include the appropriateness and disclosure of hardware accounting, acceleration of deals using channel partners and balance sheet adequacy especially w.r.t. bad and doutful debt provisions

The study has to make certain assumptions related to deals and balance sheet impacts on prior period P&Ls, which have not been auIAdited”.

No further work was done on the rebasing exercise after July 2012: that was the evidence Mr Yelland gave in the US criminal proceedings, and he confirmed it to me. (Footnote: ⁵²)

In mid-July 2012, Mr Yelland was also asked to give an initial view on whether the revenue that had been removed under his rebasing exercise had been recognised in accordance with IFRS. The Claimants emphasised that this was a separate exercise. He delivered this on 22 July 2012. Mr Yelland confirmed that this view was not purporting to be definitive. His experience was primarily in US GAAP, not IFRS. He was not an IFRS expert, felt uncomfortable in doing the exercise and did not want any statements he made about IFRS “ to be over-interpreted ”. He told the Court:

Ms Sunderwala’s view was that it was “ truly an estimation exercise at this point because we have not verified whether these are truly incorrect under IFRS AND there is a significant amount of judgment (and even speculation) around what period these items really relate to ”. This was a fair summary of the exercise, as Mr Yelland confirmed.

That initial view did not raise any red flags. For 2010, reported operating profit and the “rebased” operating profit was the same, at 36%. For 2011, it remained comparable (24%, as compared with 21%). Of the $155.9m revenue adjustment Mr Yelland proposed for FY11, he concluded just $2m of this was “ Not IFRS compliant confirmed ” (though a further $114.3m, which included $84.6m “Hardware resale” and $11.6m relating to “Accelerated rev rec (channel and solution deals)” , was described as “Not IFRS compliant probable” ). No revenue relating to licence hosting deals or purchases from customers was judged to be “ Not IFRS compliant confirmed ”. Licence revenue on hosting deals and 25% of revenue relating to ‘reciprocal’ deals was classed as “ Management judgement/US GAAP difference ”.

In cross-examination, Mr Yelland sought to stress that (a) the need for further investigations was recognised and he had included reference to hardware revenues because although non-compliance had not been confirmed, his own view was that they were not IFRS-compliant and (b) revenue recognition was only one part of what it was necessary to consider, the other part being disclosure. He told me that he had reached the conclusion that hardware revenues should have been disclosed separately.

In August 2012 HP carried out an impairment analysis of its business units, including Autonomy. Mr Yelland’s rebasing exercise had, of course, by then already been concluded. HP accordingly had Mr Yelland’s rebasing conclusions when they considered whether an impairment was necessary in August 2012. HP concluded that no impairment was required, and that the carrying value of Autonomy should be maintained at the $11bn that HP had paid in October 2011. That was so notwithstanding that Ms Lesjak, HP’s CFO and the person who ultimately had to determine whether or not to recommend an impairment charge, stated (in her witness statement) that she had been informed that the exercise conducted by Mr Yelland had:

Why that was the view taken in August 2012, whereas some three months later an impairment charge of $8.8 billion was suddenly announced (see below), was a matter of considerable dispute but the justification advanced at the time seems reasonably clear on the available documentary evidence. The quarterly goodwill memo from Ms Lesjak’s department stated that the:

In an internal email dated 28 June 2012, Ms Lesjak herself was recorded as having told EY (though, as explained in footnote [51] above, she said she could not recall this) that the Autonomy issues were related to “short-term challenges with working in the HP environment and not an issue with the overall market or sales funnel.”

By contrast with HP’s approach to Autonomy, an impairment was taken on the Enterprise Services division, which had been acquired in 2008 for $13.9bn; the write-down announced in August 2012 was $8bn, as Ms Lesjak confirmed. Ms Sunderwala also confirmed in June 2012 that, based on Ms Lesjak’s comments, no impairment analysis was planned for Q3 2012, though one would ordinarily be undertaken at year end.

Despite this, on 20 November 2012, HP publicly announced (as part of its Full Year 2012 Results announcement) that it was writing down the value of Autonomy by $8.8 billion and attributed $5 billion of that amount to fraud. The Claimants and the Defendants disputed the reasons for this change of tack, and the basis for this serious allegation.

HP’s position was heralded in its press release on the same day, which stated that the majority of the impairment charge of more than $5 billion was:

The press release continued:

The mischaracterization of revenue from negative-margin, low-end hardware sales with little or no associated software content as “IDOL product”, and the improper inclusion of such revenue as “license revenue” for purposes of the organic and IDOL growth calculations.

The use of licensing transactions with value-added resellers to inappropriately accelerate revenue recognition, or worse, create revenue where no end-user customer existed at the time of sale.

However, and as the Defendants repeatedly stressed, it was unclear quite what had changed since August 2012, after the conclusion of Mr Yelland’s report on 17 July and his tentative expression of views on IFRS a little later.

Ms Whitman gave the following explanation during the media call which followed the announcement:

No doubt conscious that the sudden announcement a year after completion of the Acquisition required some further explanation, Mr Schultz also suggested at the same press conference that Autonomy did not have a proper set of accounting records and that:

Thus, it appears that HP sought to justify the write-down and accompanying accusations of fraud in November 2012 by reference to the whistleblowing in May, the matters identified and considered in the “intense internal investigation” and “forensic review” , and the piecing together of disparate and partly incomplete information after that.

The curiosity of this is not only that the earlier decision had been against impairment, but also the lack of any documentary evidence of any calculation of what amounts were to be attributed to the wrongdoing. Ms Lesjak (Footnote: ⁵⁵) herself seemed unable to explain this: see further as to this paragraph [396] below.

Furthermore, the direct and contemporaneous documentary evidence might suggest a different and more benign explanation of the reasons for the write-down. Thus, according to the documents:

$3.6bn of the $8.8bn impairment was referable to the effect of changing the WACC from 10% to 16%. I discuss this change further below; but for the present the point to note is that when asked whether this change was anything to do with the allegations of fraud, Ms Lesjak, after some equivocation, agreed that she did not believe that this was ever quantified.

After the impact of the market capitalisation requirements, the remaining $5.2bn of the $8.8bn impairment was split into three categories, as set out in an infographic created internally, which corresponded substantially with the analysis set out in the presentation on the impairment given to HP’s Board on 24 and 26 October 2012:

$2.9bn said to be attributable to the effect of lower margins;

$1.8bn said to be attributable to the effect of lower growth rates; and

$1.3bn said to be attributable to the effect of re-baselining revenue.

The reasons for lower margins were articulated in an email to, amongst others, Ms Lesjak on 19 November 2012 (i.e. the day before the announcement). That email, which was sent to Mr Schultz, stated there were three “buckets” that explained the lower margins:

None of those factors identifying an impact on margin were tied in or shown to relate to the alleged accounting improprieties announced by HP the following day. This was reluctantly accepted, after two deflecting answers, by Ms Lesjak in cross-examination. (Footnote: ⁵⁶) The same was true of the other factors identified in the board presentation and infographic. Ms Lesjak accepted that the points about “lower growth rates” were similar to the points about lower margins.

Further:

Although it is not disputed that in May 2012, HP had retained Morgan Lewis and PwC, no written “forensic review” by PwC is apparent from the evidence until later (in 2013, long after this announcement), and none was disclosed;

Indeed, there is no evidence of any review, investigative report or analysis by PwC dating from before the announcement (or even a short period thereafter), except a commentary on the methodology adopted by Mr Yelland for which PwC disclaimed responsibility. (Footnote: ⁵⁷)

The original purpose of the rebasing exercise had nothing to do with possible mis-accounting under IFRS: it looked at the differences under US GAAP and how HP would have run the business. When Mr Yelland was asked to give a view on IFRS, he was uncomfortable doing so, his conclusions in any event did not support a conclusion that there had been wilful non-compliance with IFRS, and Ms Sunderwala thought that it was “truly an estimation exercise” , involving “a lot of judgment (even speculation)” .

Mr Yelland’s evidence was that his own rebasing exercise (produced in late July) did not materially change thereafter.

Autonomy had a full and orderly set of books, accounts and ledgers which amongst other things identified all Autonomy’s hardware transactions.

The Defendants supported their contention that there had been no such analysis as might be expected before making public claims of fraud by referring to events after the press release. Thus:

On 29 November 2012, 9 days after the announcement, there was an email exchange involving Ms Lesjak, Mr Gomez (HP’s Chief Communications Officer), and Mr Levine (HP’s Corporate Controller). Ms Lesjak was told:

Ms Lesjak plainly had not seen any analysis of this kind. She asked Mr Levine what he had. All Mr Levine had was some PowerPoint slides substantially the same as the board presentation referred to at paragraph 392(2) above. Mr Levine stated in an email to Ms Lesjak dated 30 November 2012 that:

Unable otherwise to explain how, on that basis, the need for impact on margins was announced as relating to fraud, Ms Lesjak first suggested that HP had had input from PwC: but when shown that PwC did not report until 2013 and that in 2012 the work that PwC did to support Mr Yelland consisted of commentary on Mr Yelland’s own work, for which, moreover, PwC disclaimed responsibility, she then told me in cross-examination that despite what appeared from that email, she had nevertheless been told by Mr Andy Johnson, who was then Head of HP’s Corporate Development sub-team within the SCD group (Footnote: ⁵⁸) , that all the adjustments “were based on new information that came out of the forensic investigation.”

Reliance on a conversation with “Andy” (whom she described, and said she relied on, as “keeper” of the original DCF model used by HP in its acquisition of Autonomy) became a refrain of her answers in this regard. The following answer was typical of a number of iterations:

That is what Andy told me. He told me that…looking at the information that they learned in the forensic investigation, that now the stand-alone value was 3.5, and he said that he believed that all of that was as a result of either accounting improprieties, disclosure failures or misrepresentations. That is what he said.”

When it was directly put to Ms Lesjak that she had no idea under any of the heads identified of how much could fairly be attributed to any wrongdoing, and that the announcement (for which she was jointly responsible) of a write-down on the grounds of a major fraud was reckless because she had never commissioned a proper calculation to correlate the allegations with the changes in the DCF she answered:

There were references in the record to some conversations between Ms Lesjak and Mr Johnson likely to have related to the impairment analysis: for example, in an email from Mr Sarin dated 28 September 2012, in an email from Ms Sunderwala dated 17 October 2012 (referring to a meeting with an electronic link that evening), in an email chain on 23 October (referring to a “deck” for discussion) and in a short email arranging a telephone call between them on 15 November 2012. But there was no documentary evidence produced to show how it was that Ms Lesjak was satisfied that the factors resulting in the write-off were attributable to accounting improprieties, disclosure failings and misrepresentation; and Ms Lesjak made no mention of having relied on Mr Johnson in her witness statement (in which she made no mention of him at all).

The Claimants sought to explain this lack of documentary support on the basis that (a) at the time the impairment charge was announced to the market, HP was in the midst of an internal forensic investigation “communications in respect of which have been withheld for privilege” and (b) it is “unsurprising that there is a limit to the volume of documents that have been disclosed by the Claimants”. But neither was or amounted to a direct assertion that there was any such documentary evidence; and the email exchanges after the event already quoted suggest there was not.

Further, as to the suggestion that such an investigation had uncovered impropriety of which HP was unaware, the Defendants’ position was that:

HP had been aware of hardware sales for some time. As explained at paragraphs 324 to 327 above, some were apparent from (and at any rate not hidden in) the due diligence.

HP and its advisors were made aware of hardware sales in some detail following completion in the Autumn of 2011. Yet hardware purchases by Autonomy from Dell for resale continued to at least April 2012: Mr Yelland was asked on at least two documented occasions (one in March 2012 in anticipation of his appointment as CFO, and another after it, in April 2012) to approve such a purchase on the basis that the hardware would be sold through to customers at a loss of approximately 10%. In each case he gave the approval without objection. He sought to argue in cross-examination that he drew a distinction between “sell-through” and “pass through” which he told me “sometimes is used to describe sales that are not connected to the rest of the business” ; but he provided no support for this, accepted that he had not focused on the difference at the time, and I find that this was an afterthought. The hardware revenues were included in spreadsheets provided by Autonomy to HP and were subsequently openly incorporated into HP’s own financial reporting.

As to the second alleged impropriety, the Defendants contended that there was no requirement at all under IFRS for an end-user to exist at the time of sale (even if there was under US GAAP). Moreover, HP had been made aware during the due diligence that Autonomy recognised revenue from resellers on “sell-in” rather than “sell-through” .

On 20 November 2012, HP announced that it was writing down the value of Autonomy by US $8.8 billion. Simultaneously, HP attributed a substantial part of the write down to fraud on the part of Autonomy prior to the Acquisition stating that there appeared to be a wilful effort on the part of certain former Autonomy employees to inflate the underlying financial metrics of the company in order to mislead investors and potential buyers.

In the circumstances described above, the Defendants’ case is that HP had no basis for suggesting that the value written down (or any part of that) was attributable to fraud by the Defendants: the real reason for the write-down was the need to reduce the carrying value of some of HP’s assets in order to take account of the diminution in HP’s market capitalisation following a sustained fall in HP’s share price. According to the Defendants, the excessive write-down of the value of Autonomy was the means of avoiding a write-down of HP’s own software business.

HP’s share price first began to decline significantly in Q4 2011. It continued to fall in 2012. In June 2012, HP’s market capitalisation fell below its book value. By 1 August 2012, its share price was $17.66, around half of what it had been 12 months previously. (Footnote: ⁵⁹) By 24 October 2012, it was $14.47.

Although the Claimants were apt to blame this on the Acquisition, the Defendants pointed out that the decline had much longer origin, and reflected very disappointing performance. Year-on-year revenues were falling by about 7%, year on year EPS had fallen by about 33%, and year on year cash flow from operations was down something over 60%. Ms Whitman accepted that these were pretty dire numbers.

The truth was that in 2012, HP was in crisis and performing poorly across its business. Ms Whitman agreed in cross-examination that when she looked more carefully after her appointment it was in a far worse state than she had initially thought and that there was a crisis of confidence around senior management, board and company strategy. I have already noted that the business was fractured in ‘silos’ with intra-division jealousies, poor co-operation and overall direction, and short-term focus. The Defendants described it, with (as it seems to me from the available documentation) good reason, as a huge, enervated, company, beset with systemic failings. It had (as Dean Acheson famously remarked of Great Britain now nearly 60 years ago) “lost an empire but not yet found a role” , having abandoned Mr Apotheker’s plan and fallen back onto retrenchment to its core business, which was low margin and an outmoded model.

As previously noted, and quite apart from Autonomy, HP’s efforts to expand by acquisition had been astonishingly problematic and unsuccessful:

HP bought Electronic Data Systems for $13.9bn. It was rebranded as HP Enterprise Services in 2008. Its value was written down in Q3 2012 by some $8bn.

Other acquisitions are listed in HP’s M&A Scorecards, such as that for Q2 2012. One well known one was the Palm hand-held computer business which it had acquired for $1.2 billion in 2010 (Footnote: ⁶⁰) . The acquired businesses being tracked had under-performed against budgeted revenue by $1.73bn or 26%.

The sustained fall in its share price had consequences for HP under US accounting rules. As Ms Lesjak explained in her witness statement, the fall in its share price meant that HP needed, in reconciling the value of its business units to market capitalisation, to (a) increase the discount rate (i.e. the premium to the weighted average cost of capital (“WACC”)), which further decreased the value of the business units and (b) revalue its goodwill and intangible assets across all its business units: the sum of the parts of the business plus a control premium of 30% could not exceed HP’s market capitalisation. Ms Lesjak considered this an arbitrary accounting construct. She accepted that it involved reverse engineering the value of segments of the business to get down to the required level.

The process of reverse engineering involved ascribing discount rates to various segments of the business. Through this, the reduction was loaded disproportionately onto certain parts of business, notably Autonomy. HP’s Enterprise Services division was also ascribed a heavy discount rate, but had no goodwill left to impair following the $8bn impairment recorded in Q3 2012.

It appears that HP was keen to avoid its own Software division (other than Autonomy) being impaired. Ms Sunderwala noted in mid-August 2012: “ I feel that even at a $16 price, we could support no impairment in HP Software (without Autonomy) if we took a hit in Autonomy stand-alone .” By October 2012, this was being fed into further models, with the discount being loaded onto Autonomy and HP Enterprise Services as compared with Software, TS and ESSN. (Footnote: ⁶¹) It seems that this was motivated, at least in part, by tax reasons:

A large write-down of Autonomy also mitigated the need to write down HP’s hardware business, which Ms Whitman had affirmed to the market as the core business. Ms Sunderwala drew another advantage to the attention of HP’s auditors, EY, quoting an article that mused:

It was observed internally that to get down to the 30% control premium (see paragraph 409 above), discount rates would have to be “ crazy ” and that the result would be “ nonsensical ”. The result was that, while a WACC of 9.5% had been deemed appropriate for valuing Autonomy in August 2012, by October 2012, HP had increased it to 15%. This was increased further to 16% after Ms Whitman asked, very late in the process (overnight between 23-24 October 2012, with the board meeting to discuss the impairment being scheduled for 24 October 2012) that the final value include some synergies. (Footnote: ⁶³) The effect of changing the WACC from 10% to 16% by itself caused an impairment of $3.6bn.

Although the Claimants sought to dismiss as “not credible” the suggestion that the erosion of Autonomy’s performance flowed from innocent causes, the fact is that remarkably little was provided by the Claimants to support their case that the impairment was the consequence of the revelation of fraud in the course of an in-depth forensic inquiry. Equally little was provided by the Claimants to substantiate the effect of the alleged fraud on the various “buckets” they had identified. On the other hand, the Defendants did provide a plausible account of the “reverse engineering” adopted to achieve a result which shifted blame away from under performance in HP’s other enterprises at Autonomy’s expense. Ultimately, I am not required to determine the reason for this erosion in Autonomy’s performance. The fact it occurred is consistent with and provides some general support for the Claimants’ case but the failure to connect the alleged fraud with the impairment means the establishment of the fraud depends on a detailed consideration of each head of claim.

By the time of the announcement HP had not given Dr Lynch an opportunity to address the allegations. HP went ahead with the announcement and accompanied it with a comprehensive press and public relations campaign, (Footnote: ⁶⁴) even establishing a so-called “truth squad”.

In December 2012, EY were asked to give their view as to whether the valuation of Autonomy in August 2011 was affected by the matters said to give rise to the impairment. EY prepared a memorandum dated December 2012, addressing:

EY recorded that they had reviewed HP’s presentations to the SEC and to the SFO, and Autonomy’s re-baselining review. They noted that, in determining Autonomy’s enterprise value in 2011, HP had adopted a Discounted Cash Flow (DCF) model, with input from Duff & Phelps (“D&P”).

EY stated their conclusion in the round to be that:

On the various matters suggested by the Claimants to give rise to the question whether the 2011 accounts should be re-drawn, EY concluded as follows:

That they did not believe that the stand-alone loss-making “pass-through” hardware sales would have had any material impact on valuation using HP’s model to fair value Autonomy, which was a DCF valuation model. Indeed, loss making hardware sales operated to reduce cash flows.

That neither reseller deals (referred to in the memorandum as “channel sales”) nor sales of upfront hosting licences had a significant impact on the cash flows of the business. They involved moving revenue between quarters, and because HP used a DCF model to fair value Autonomy EY did not believe the timing differences would materially impact the valuation.

As to concerns about reciprocal deals, if value received was legitimate, the arrangements did not impact cash flows either. The proposed adjustments identified by HP did not have a material impact on the original valuation model.

These conclusions were relied on by the Defendants as endorsing their case that (a) there was no fraud and (b) the Claimants’ case on loss was a confected one.

The Claimants rejected the Defendants’ reliance on EY’s memorandum as misplaced because:

The question at which it was aimed was whether HP, having recognised an impairment in November 2012, ought to restate its accounts in respect of the purchase of Autonomy in 2011. Ms Lesjak had explained that that question turned on whether the change in forecast underlying the impairment calculation resulted from correcting an error, or a change in estimate; and as HP told the SEC, it took the view that its revised forecast did not amount to correcting an error, because the original forecasts were not erroneous based on the facts that existed at the time. Rather, it considered the revised forecast a change in estimate based on new information;

The memorandum was not shared with HP at the time and its context is unclear because EY have taken the position that the Claimants were not entitled to disclosure of its working papers and the memorandum itself was made available to the parties in these proceedings only because it had been admitted into evidence in the US criminal proceedings. Its context was unknown, and it should not be read out of context.

The Defendants embraced EY’s view as “a far closer reflection of reality than the Claimants’ case. ” Their overall position as to the financial effect of the allegations, leaving aside their defence to the allegations themselves, was summarised as follows:

The announcement of the impairment write-down, and the reasons for it, served as a catalyst for proceedings by HP’s shareholders against HP and its directors in the US, including derivative claims as well as securities claims for their failures in relation to the Acquisition. Those proceedings were ultimately settled against the backdrop that HP was itself going to bring proceedings against Dr Lynch and Mr Hussain.

These proceedings, following criminal proceedings in Northern California against Mr Hussain which resulted in him being convicted of fraud, being sentenced to 5 years in prison, and having to pay circa $10 million in financial penalties, were issued on 30 March 2015, some two and a half years after HP had publicly accused the Defendants of fraud. They follow an astonishingly extensive trawl through Autonomy’s documents and transactions in an effort to make good those early accusations, for which they appear to have had precious little support at the time, and even to provide a basis for quantifying loss in exactly the same amount as HP had asserted (without any discernible basis) in its public announcement of those accusations.

Dr Lynch has been indicted and further to extradition proceedings commenced in September 2019 before the conclusion of evidence in this trial, the Home Secretary ordered his extradition on 28 January 2022. (On 9 February 2022, Dr Lynch applied for permission to appeal that decision.) Mr Chamberlain has also been indicted (on the same indictment as Dr Lynch), which Dr Lynch relied on to explain his absence (see paragraph [426(3)] below). A superseding indictment of 21 March 2019 showed that investigations had been extended to events relating to the founding and growth of Invoke and its investments, which the Defendants offered as the explanation for Mr Kanter and Ms Eagan declining to give evidence in these proceedings, for fear (according to the Defendants) of repercussions in potential US criminal proceedings.

The US criminal proceedings, actual and prospective, have cast long shadows over these proceedings. It is not for me to question the basis on which the US courts have been satisfied as to their jurisdiction, though it is to be noted that the only person likely to have suffered direct loss (Footnote: ⁶⁵) , wherever the claims are adjudicated, is HP; and HP agreed to this jurisdiction. But the fact is that the effect of the US proceedings in this, the natural forum for these claims, has been considerable:

Mr Hussain was in practice unable to attend proceedings brought against him in their natural forum and in his place of habitual residence alleging fraud and damages well in excess of all he probably has in financial terms.

Throughout the hearing and all through his mammoth cross-examination, Dr Lynch has had to bear the additional burden of indictments against him and, since September 2019, the persistent threat of extradition.

The Claimants noted that even before he was indicted, Mr Chamberlain did not provide any witness statement in these proceedings; but the fact is that he has long been under threat and has now been indicted by the US DoJ and charged by the FRC with acting dishonestly and/or recklessly as well as failing to act with competence and due care. Although the Claimants submitted that this had not been suggested to be and was not a good explanation for Dr Lynch not having called Mr Chamberlain, Dr Lynch did note in his written closing submissions that “it is to be expected that he would wish to remain silent pending those charges” : although I accept that and have not drawn inferences, without more, from his absence, he was a central character and I have felt it unavoidable and necessary to make findings in respect of his conduct notwithstanding his absence (as, in my assessment, he must always have appreciated was on the cards).

As mentioned above, Mr Kanter and Ms Eagan (who is based in the US) declined to appear before me, though both had provided witness statements. Dr Lynch was apparently informed by their lawyers on 19 and 21 August 2019 respectively of their decision not to give evidence. No explanation appears to have been provided by those individuals’ lawyers, and certainly none was elaborated to me. The Defendants submitted that neither of them (though both worked for DarkTrace) was within Dr Lynch’s control; but even if technically true, and even if it is a reasonable inference that the extension of the US criminal investigation may have made them more reluctant to give evidence at this trial, it seemed to me likely, in the light of the fact that they had provided witness statements and until August 2019 they were in his running order, that had Dr Lynch really wanted them to attend they would have been persuaded to do so. The Claimants also noted that Dr Lynch had caused the press to be informed that it was he who had decided not to call them. It is difficult and I consider unnecessary to decide why they were not called. I have not drawn general inferences from their absence; but I have necessarily in particular instances drawn conclusions without being able to weigh in the balance any contrary evidence. That is inevitable: both were central actors; and Mr Kanter especially could have provided useful evidence bearing on factual matters where I have in consequence had to make determinations without his assistance; and likewise, though less often, in the case of Ms Eagan.

Certain key witnesses whose evidence in the US criminal trial was admitted into these proceedings by hearsay notice had been given immunity in return for their evidence (namely, Messrs Joel Scott and David and Steven Truitt, and Ms Antonia Anderson as to whom see later). Mr Egan had entered into a deferred prosecution agreement (“DPA”) which bound him to give evidence admitting culpability in the terms of a statement of fact negotiated and agreed with the DoJ, and not to offer any contradictory evidence or arguments on pain of immediate prosecution without defence for breach. Inevitably this invites caution and at least prima facie wariness as to their reliability, especially if and when their documented conduct appeared to reveal a different explanation than they offered in their evidence.

The position of Mr Sullivan, and whether he was promised immunity, was unclear in this regard; but although he provided a signed witness statement and it appeared that he was to give oral evidence, the Claimants informed the Defendants at a late stage that he was unwilling to attend, and that the Claimants did not propose to invoke the procedure for compelling his attendance that they did embark upon in the case of Mr Egan. His evidence, and the transcripts of his evidence in the US criminal proceedings which dealt with many of the matters in issue here, but which he did not deal with in his witness statement, came in only as hearsay.

Evidence from those witnesses in the US criminal proceedings who also provided a witness statement was often over-lawyered and too rehearsed to be authentic. Further, it became apparent that a number of the Claimants’ witnesses had discussed their evidence with lawyers for the Claimants or the US prosecutors, often many times (Footnote: ⁶⁶) , to the extent that I often felt I was being treated to a received and then rehearsed version of events, rather than the witnesses’ actual recollection.

But for the earlier criminal trial in the US the Claimants would almost certainly have felt they had to call in these proceedings and to expose to cross-examination witnesses whose evidence in the US they instead relied on by its introduction under a hearsay notice. I have in mind (i) Mr Scott (ii) Mr Stephan (iii) Mr Steven Truitt (iv) Mr David Truitt (v) Mr Loomis (vi) Mr Channing (vii) Mr Johnson and (as indicated above) (viii) Mr Sullivan. Dr Lynch thus had no opportunity to cross-examine any of those witnesses. Throughout this trial I have had to be careful to remind myself that much of the evidence on a transcript was only hearsay evidence untested in these proceedings.

There were notable absences. The only oral evidence from anyone at Deloitte, was that of Mr Welham (who also gave evidence in the US criminal proceedings). He was not a partner of Deloitte at the relevant times. The audit partners, Mr Richard Knights (for 2009 and Q1 2010) and then Mr Nigel Mercer (for the 2010 yearly audit), did not give any evidence. Their attendance would greatly have assisted me; and indeed the Defendants maintain that it is to be inferred from the fact that the Claimants did not call either, that (in a different way) it would ultimately have helped them. The Defendants pointed out in that regard that a Settlement Agreement entered into with Deloitte in April 2016 gave the Claimants unusual control over Mr Knights as a potential witness: clause 6 gave them the right ( inter alia ) to require Mr Knights to prepare for and attend interviews, sign statements and attend voluntarily as a witness at this trial. It follows that the Claimants had the ability to procure his attendance. The only explanation offered for not calling him is that he was involved in disciplinary proceedings brought by the FRC. But the Defendants submitted that that was not a good explanation, and pointed out also that the defence served on his behalf (and that of Deloitte) in those proceedings supported aspects of Dr Lynch's (and Mr Hussain’s) position. I have had to bear that in mind also, though I do not consider any general inference such as the Defendants submitted should be made that their evidence would have supported their case is warranted. It depends obviously on the facts and circumstances out of which each issue arises.

There were gaps on the Defendants’ side in addition to those already noted. Mr Zanchini, who also works for Darktrace, provided a witness statement, but this was withdrawn when the Defendants decided not to call him. Dr Menell, a central character who was copied into many of the critical emails, provided no evidence at all. With regard to Dr Menell, Dr Lynch said in cross-examination “I guess you would call it a medical issue”. I invited Mr Rabinowitz to explore this; but he declined on the basis that he did not wish to embarrass Dr Menell. I have drawn no inferences; but again, his evidence could have assisted me, and again I have in consequence had to make determinations (including as to their conduct) without their assistance.

Mr Zanchini’s role was more limited. However, as will be seen, he was one of the individuals said by the Defendants to have spent all his technical time in Q1, Q2 and Q3 2009 on the development of a product called “SPE”. The issue became one of some importance. As no documents had been identified to suggest that Mr Zanchini had spent any of his time in the first three quarters of 2009 doing any such work on SPE, the Claimants invited the inference that at least one of Dr Lynch's reasons for declining to call him was that Mr Zanchini could not have given truthful evidence in support of Dr Lynch's case on the matter. I take that into account also when considering the issue later, as I do in respect of his direct involvement in one of the impugned VAR transactions (the “Vatican Library” transaction, “VT13”) where I have also had to make determinations (in the event, contrary to the Defendants) without the benefit of his evidence.

Lastly in this resume of problems with the witness evidence, there is always significant danger in rehearsed evidence, especially when expanded over time. Memory is never complete, and often it is an amalgam of real memory and the received, with confirmation bias accentuating those parts of it which have become important support for a given point of view. Further, the events in question took place nearly a decade before this trial. I have had to be careful throughout in my approach to the witness evidence, and the documents, as usual, provided a clearer and less tainted guide. Even in that context, however, a number of documents were labelled by the Claimants as “pre-textual”. I have sought to indicate those in the course of this judgment.

I turn next to various legal issues which arise in respect of the various claims.

Overview and background to the provisions of FSMA

Autonomy’s liability to Bidco is said to arise under s. 90A FSMA (“s. 90A”), and its successor Schedule 10A FSMA (“Sch 10A”). S. 90A, introduced into FSMA by the Companies Act 2006, was in force from 8 November 2006 to 30 September 2010. Sch 10A came into force on 1 October 2010, but s. 90A continues to apply to information first published before that date. (Footnote: ⁶⁷)

S. 90A and Sch 10A share the same broad scheme. They impose liability on the issuers of securities for misleading statements or omissions in certain publications, but only in circumstances where a person discharging managerial responsibilities at the issuer (a “PDMR”) knew that or was reckless as to whether the statement was untrue or misleading, or knew the omission to be a dishonest concealment of a material fact. The issuer is liable to pay compensation to anyone who has acquired (Footnote: ⁶⁸) securities in reliance on the information contained in the publication, for any losses suffered as a result of the untrue or misleading statement or omission, but only where the reliance was reasonable.

In the present case, the alleged liability of Autonomy under s. 90A/Sch 10A is used as a stepping-stone to a claim against Dr Lynch and Mr Hussain: it is something of a ‘dog leg claim’ in which, having accepted full liability to Bidco, Autonomy now seeks to recover in turn from the Defendants. There is no conceptual impediment to this. However, it is right to bear in mind that in interpreting the provisions and conditions of liability, the relevant question is whether the issuer itself should be liable, bearing in mind that in the usual course, the issuer’s liability will fall on the general body of its shareholders, rather than on the director or other individual primarily responsible for any misstatement.

FSMA was the first UK statute to provide a statutory scheme of liability to investors for misstatements made to the market by issuers, other than through prospectuses; and this case is believed to be the first s. 90A or Sch 10A case to come to trial in these Courts.

This statutory scheme for issuer liability for misstatements or deliberate omissions in published information has been some time in the making. English law had historically not provided a remedy for investors acquiring shares on the basis of inaccuracies in a company’s financial statements. Until the introduction of s. 90A in 2006, there was no statutory regime imposing civil liability for inaccurate statements in information disclosed by issuers to the market. This stands in contrast to the long-established statutory scheme of liability for misstatements contained in prospectuses, first set out in the Directors Liability Act 1890, and now contained in the widely drawn provisions of s. 90 FSMA. (Footnote: ⁶⁹)

The rationale for the different treatment of liability for misstatements in prospectuses and those in other disclosures – with a tougher regime for prospectuses – is easy to understand. “ [T]he prospectus is a ‘selling document’, produced by the issuer to promote its securities to investors, whereas periodic and ad hoc reports are expressions of routine reporting requirements which do not typically coincide with a selling effort on the part of the company ” . (Footnote: ⁷⁰) An untrue statement in a prospectus can lead to payments being made to the company on a false basis, but the same cannot be said of an untrue statement contained in, say, an annual report.

Nor did the common law provide an easy route to recovery for an investor who had acquired shares in reliance on a misstatement in a company’s periodic or ad hoc disclosures:

Claims in deceit did not readily give rise to liability for such misstatements, because of the requirement in Derry v Peek (1889) 14 App Cas 337 that the maker of the statement should have intended that the recipient of the statement rely on it: as noted in the Davies Review (Footnote: ⁷¹) , although that requirement might easily be satisfied in the case of a prospectus, which is a selling document, it is difficult to satisfy in the case of an annual or quarterly report, which is primarily a report to shareholders on the directors’ stewardship and not obviously intended to induce reliance by securities trading.

Claims in negligence were in general precluded by the House of Lords’ decision in Caparo v Dickman [1990] 2 AC 605 that liability for economic loss due to negligent misstatement was confined to cases where the statement or advice had been given to a known recipient for a specific purpose of which the maker was aware. Since statutory accounts are prepared for the purposes of assisting shareholders to exercise their governance rights, rather than enabling them to take investment decisions, an investor who acquired shares in reliance on a company’s published accounts would not normally have a cause of action.

Furthermore, the ‘safe harbour’ provisions of s. 463 Companies Act 2006 entirely preclude liability to third parties in respect of ‘narrative’ reports, subject only to s.90A/Sch 10A FSMA.

Accordingly, until the enactment of s. 90A, there was no scope for a claim of the present type. The explanatory note to the legislation by which s. 90A was introduced in 2006 stated that as at that time, no issuer had been found liable in damages under English law in respect of statements made in narrative reports or financial statements. (Footnote: ⁷²)

The ultimate catalyst for the introduction of a scheme of liability was the Transparency Directive (“TD”) (Footnote: ⁷³) in 2004. This required Member States to apply their national liability regimes to the disclosures required by that Directive and to apply them to “at least…the issuer or its administrative…bodies.” The TD’s insistence on enhanced disclosure obligations and the requirement for a responsibility statement gave rise to concerns that the English law’s restrictive approach to issuer liability would be disturbed, and that issuers (and directors and auditors) might be made liable for merely negligent errors contained in narrative reports or financial statements. (Footnote: ⁷⁴)

In light of these concerns, and in recognition of the historical tendency against liability, the regime for issuer liability was introduced in this jurisdiction piece-meal, starting with a stop-gap solution introduced into FSMA by the Companies Act 2006 in the original form of s.90A. This approach allowed for further enquiry and consideration whether any more extensive regime was appropriate.

Any such scheme inevitably involves a balance, especially between the desirability of encouraging proper disclosure and affording recourse to a defrauded investor in its absence, and the need to protect existing and longer term investors who, subject to any claim against relevant directors (who may not be good for the money), may indirectly bear the brunt of any award. The explanatory note referred to the uncertainty which the TD had created, and stated that the government was “anxious not to extend unnecessarily the scope of any duties which might be owed to investors or wider classes of third parties, in order to protect the interests of company members, employees and creditors”. (Footnote: ⁷⁵) In the debates on the introduction of this provision, a government minister referred to “the Government’s desire to avoid radical change to the law in this area, to respect the preferences of stakeholders and, especially, to ensure that company resources are not inappropriately diverted from shareholders, employees and creditors to the benefit of a much wider group of actual and potential investors” . (Footnote: ⁷⁶)

Amongst the matters for particular consideration were (a) whether negligence or (as in the USA) gross negligence should suffice or whether liability should continue to depend on proof of fraud; (b) what should count as published information; and (c) who apart from the issuer should be prospectively liable.

After the Davies Review, and a process of consultation and subsequent Final Report by Professor Davies (“the Final Report”), the original s. 90A provisions were extended (a) to issuers with securities admitted to trading on a greater variety of trading facilities (b) to relevant information disclosed by an issuer through a UK recognised information service (c) to permit sellers, as well as buyers, of securities to recover losses incurred through reliance on fraudulent misstatements or omissions and (d) to permit recovery for losses resulting from dishonest delay in disclosure. However, Professor Davies recommended that liability should continue to be based on fraud, and his recommendation was accepted. Further, no change was suggested or made to the limitation to PDMRs of the persons other than the issuer who could be made liable, and no specific provisions to determine the basis for the assessment of damages was introduced.

As emphasised by the Defendants, this history is not merely of antiquarian interest. It is relevant when considering the scope of the section, and in particular the matters left open by the draftsman such as the nature of the reliance that must be shown, and the measure of damages. I accept the Defendants’ general admonition that the Court should not interpret and apply the section in a way which exposes public companies and their shareholders to unreasonably wide liability.

S. 90A(3), in the form that applied until 30 September 2010, read as follows, so far as material:

The equivalent provision of Sch 10A is paragraph 3: