Quantum awarded in HP’s claim against former CEO and CFO of Autonomy

The long-awaited quantum judgment in Autonomy Corporation and others v Lynch and Hussain (2025 EWHC 1877 (Ch)) was handed down on 22 July 2025. It provides insight into the approach to quantifying claims under Section 90A FSMA.

HP claimed that the former Autonomy CEO (Dr Mike Lynch) and CFO (Sushovan Hussain) fraudulently inflated Autonomy's published accounts, causing an overvaluation in the price HP paid for the company in 2011.  After more than 11 years of litigation, the Court valued HP's loss at £740 million - significant, but considerably less than the approximately $5 billion HP originally claimed.  This was on the basis that HP would still have bought Autonomy, even if it knew its true accounting position, albeit at a lower price. 

Below, we distil the key practical takeaways for legal practitioners in such disputes.

Section 90A class actions 

Section 90A provides shareholders with a cause of action against UK-listed companies for losses suffered due to their reliance on misleading information that the company published.  Generally (though not here) such claims are brought by way of a class action with third-party litigation funding.  This is the first Section 90A case to reach a quantum judgment.  As such, this judgment is likely to be influential in other Section 90A cases.

We draw out more specific points of relevance to Section 90A class actions below. 

Private acquisitions of UK public companies

Instead of bringing a common law misrepresentation claim only against the former directors, HP constructed the lion's share of its claim as a Section 90A FSMA claim against Autonomy Plc (by then its own company), and then sought to attach that liability to Lynch and Hussain personally.  This construct has two obvious benefits over common law misrepresentation claims:

• It did not require HP, as purchaser, to prove that the Autonomy directors made a misrepresentation directly to it. Given the requirement to treat shareholders equally under UK disclosure rules, that kind of circumstance ought rarely to arise in public companies.  In recognition, Section 90A only requires investors to show that PDMRs¹ (essentially the statutory directors) knew that the company's Published Information² contained misleading information. While still challenging, it is a lower evidential threshold than for a direct misrepresentation claim.  

• It was not limited to the revised value of the 'contract' between HP and Lynch / Hussain, i.e. their combined 7.4% stake in Autonomy.  Instead, the revised value of Autonomy's entire issued share capital could, in theory, comprise HP's loss. This greatly increased the potential quantum of HP's claim.

This construct may be of interest to private equity firms considering their options where a fund has taken private a UK-listed company, only to discover, post-acquisition, that the company's finances were misleadingly presented by its former directors.  Where those directors are of high net worth, they present a potential target for recovering losses, as this claim demonstrates.

¹ Persons Discharging Managerial Responsibility
² Periodic reports and press releases

Assessing damages in Section 90A claims

The judge started with a reminder that Section 90A FSMA is closely modelled on tortious deceit claims; it too should put an injured party back in the same position as if the misleading information had not been included in the company's Published Information. 

This gives rise to the following approach to quantum:

1. what would have happened had the Published Information not contained the false or misleading statement, and included all matters required to be included? In particular, would the investor - armed with accurate information - have gone through with the transaction (at a lower price) or pulled out of the transaction altogether?  This distinction is sometimes referred to as a 'Transaction or No Transaction' scenario.

2. If the investor would have gone through with the transaction, what lower price would they have paid?

Constructing the Section 90A counterfactual

As to 1, the counterfactual is fairly simple for misstatements - the Published Information is corrected by not containing that statement.  That kind of binary counterfactual should in principle be fairly straightforward to model. For example, this can be done by an event study, as is common practice in US share-drop cases. The judge in this case even went so far as to suggest he would have been aided by such a study [102] to [106]. 

Arguably, however, it is harder to conceive of the appropriate corrective disclosure in omission cases, especially where the relevant information is not of a precise, quantitative nature.  Parties are still likely to disagree over the extent to which the company ought to have disclosed matters omitted from its Published Information, and the effect of that on both: (i) whether the counterfactual is a Transaction or No Transaction scenario; and if the former, (ii) what was the price impact. 

Adding to that complexity is the judge's decision in the 2022 liability judgment that a company can only be held liable for the specific misrepresentation that its PDMR is found to have known, even if there are others in the annual report.  This may result in a significant practical challenge for claimants at trial or even before.  Given the information asymmetry in these cases, it may be difficult for claimants to predict at the outset what the evidence will reveal a company's PDMRs to have known.  Facing that uncertainty, should claimants model loss based on all alleged misrepresentations being upheld, or should they anticipate what the court might find? The former could lead to significant wasted expenses, but the latter risks underselling the claimants' case.

Inconsistent counterfactuals

The liability judgment found, and HP accepted, that its Section 90A case proceeded on a 'Transaction basis', i.e. in possession of accurate information, HP would still have bought Autonomy but for a lower price.  Despite that, HP maintained a 'No Transaction' basis for its parallel common law misrepresentation / deceit claim in respect of Lynch and Hussain's personal shares. This is because in a 'No Transaction' counterfactual, investors can in principle claim for wider losses than price inflation, including for example the lost opportunity to buy something else. 

However, the Court objected as a matter of principle to HP relying on "diametrically opposite factual assertions" - 'Transaction' and 'No Transaction' - so as to increase its aggregate loss [639].  Instead, the Court held HP to a 'Transaction' basis for both its FSMA and misrepresentation claims.  This point may prove influential in class actions brought under Section 90A.

One can also see the inherent tension in trying to frame witness evidence for alternative 'Transaction' and 'No Transaction' cases, particularly in the more recent corporate governance-type Section 90A claims. This may tend towards picking one basis over the other.  Claimants will typically keep a close eye on the quantum impact of these alternative approaches to quantum when presenting their case.

Factual evidence for loss and split trials

The judge clarified that to calculate loss in a Section 90A FSMA claim "it is necessary as part of the inquiry to retrace the steps actually adopted by HP, and the assessments of value HP made, as best these can be identified, and to replicate these steps, but in the (hypothetical) context of and by reference to the published information as modified for the purposes of the FSMA" [37].  This highlights the real importance of reliance and loss particulars, which are increasingly becoming standard documents expected from claimants in Section 90A claims, together with evidence from witnesses of what they would have done in the counterfactual.

Originally, the case management directions in the HP/Autonomy case had not ordered a split trial.  Despite that, due to the complexity, the judge decided to split the liability and quantum judgments, and requested further quantum evidence and a 9-day quantum-specific trial between the two.  The HP/Autonomy case tends to suggest that a split trial, if ordered at the outset, would provide certainty as to the possible valuation methods and ultimately save costs for both parties. The debate continues about how much disclosure and evidence the claimants should be ordered to give on reliance issues ahead of the liability trial.

No 'disappointment discount'

The judge in the HP/Autonomy case stated that there should be no valuation 'disappointment factor' by reason of comparing the company you thought you had bought, to the company you in fact bought. The judge held:  

"First, it is easy to speculate that the market would have been greatly affected upon the revelation of the volume of hardware sales, and to assume a considerable downward pressure on Autonomy's share price accordingly. But any surprise or disappointment factor, or market perturbation, which would in all probability have adversely affected the share price upon the revelation of such an active programme of reselling hardware (and the result that Autonomy was not the pure software seller that it was thought to be), is not to be brought into account in the FSMA Counterfactual" [508(3)]

Avoiding a 'disappointment factor' may be reasonably straightforward in an accounting fraud case where you assume the market was given accurate information, e.g. the company's earnings were X as opposed to being misrepresented as Y.  However, it is harder to see how this would apply in corporate governance-type Section 90A cases.  On the face of it, any disappointment arising from the company's alleged involvement in a scandal will not be accounted for as an FSMA loss.  Judicial guidance on this matter will likely be needed, as it was not addressed in HP/Autonomy.   

Conclusion and open questions

The Autonomy quantum judgment is an important addition to the ever-growing judicial authority available on Section 90A cases. 

Nor is this the end of the saga. The Court left open the question of when HP's claim against Autonomy crystallised for the purpose of calculating liability in US dollars ([703]), along with other matters on interest and costs. These issues may have wider relevance. We anticipate that they will be addressed soon at a further hearing.