Reflective Loss: the Unprincipled Principle

Published November 2019

The reflective loss principle (‘RLP’) is designed to prevent a claimant from recovering damages for loss suffered because the company in which the claimant is invested has suffered loss.

The reflective loss principle (‘RLP’) is designed to prevent a claimant from recovering damages for loss suffered because the company in which the claimant is invested has suffered loss.

It was introduced by the English Court of Appeal (CoA) in the decision in Prudential Assurance Co. Ltd. v Newman Industries Ltd. and Others (No. 2) [1982] AC 204 (‘Prudential’), which held that the claimant shareholder could not recover a sum equal to the diminution in the market value of his shares or dividend because such a loss was merely a reflection of the loss suffered by the company.

Although the RLP has been applied in common law jurisdictions for almost forty years, in recent years it has become more significant and controversial. The increased scrutiny has arisen from an uncertainty about its ambit, which has been expanded by the courts, in particular in two recent decisions in the United Kingdom and the Cayman Islands. Both decisions are currently on appeal to the United Kingdom Supreme Court and the Judicial Committee of the Privy Council respectively.

The first is the decision of the CoA in Carlos Sevilleja Garcia v Marex Financial Limited [2018] EWCA Civ 1468 (‘Marex’), heard by the Supreme Court on 8 May 2019 and awaiting judgment. This case concerns the question of whether the RLP applies to claims by creditors as well as shareholders of a company and the breadth of the fraud exception to the RLP, where the wrongdoing of the defendant has disabled the company from being able to bring a claim. The CoA found that the fraud exception only applied in very limited circumstances.

The second is the decision of the Cayman Islands Cayman Islands Court of Appeal (CICA) in Primeo Fund (in Official Liquidation) (‘Primeo’) v (1) Bank of Bermuda (Cayman) Limited and (2) HSBC Securities Services (Luxembourg) SA (unreported, 13 June 2019). The CICA found that Primeo had valid claims but that its loss was reflective, notwithstanding that Primeo had suffered its losses before it acquired a shareholding in a company. The case raises a wide range of issues relating to the RLP.

As discussed below, the current uncertainty surrounding the limits of the RLP has come to the fore because of the draconian, and potentially very unjust, consequences of the RLP. Originally intended to prevent claimants from using a personal action to circumvent the rule in Foss v Harbottle (1843) 2 Hare 461 (‘Foss’) – the “proper plaintiff” in an action in respect of a wrong alleged to be done to a company is the company and not the shareholder – the scope of the RLP may now extend to any situation where there is a prospect of a claimant receiving a contribution to separate loss which it has suffered from a company which may also have an action, even if the company’s action is likely to fail, and the fraud exception may be so narrow that it has no practical effect. Given that the effect of the rule is to expropriate from the claimant its cause of action, and the courts have no discretion in relation to its application, obvious questions arise as to whether it operates in the interests of justice and is appropriate in its current extended form.

Expansion of the RLP since its origin

The rule in Foss operated to prevent shareholders from enforcing a cause of action belonging to the company.1 The basis for the rule was that a company is a separate legal entity and liable for its own contracts and torts. Any irregularity or wrong which occurred in the management of the company was capable of being waived or ratified by an ordinary resolution of the company in general meeting. This preserved the rights of the majority and would generally bind individual shareholders. The exception to the rule was where there was a fraud on the minority, such that the company was under wrongdoer control, which enabled a minority shareholder to bring a derivative claim on behalf of the company in respect of the company’s cause of action.

Approximately 140 years later, the courts began to embark on the exercise of applying the purpose of the rule in Foss to personal claims brought by shareholders to recover for their own loss. The first decision was Prudential in 1982. The CoA held that the shareholder’s personal claim was misconceived because, in seeking to recover a sum equal to the diminution in market value of his shares his loss was merely a reflection of the loss suffered by the company. A personal action by the shareholder for this loss would subvert the rule in Foss that the company has the cause of action in respect of breaches causing it to suffer damage and no such cause of action vests in the shareholder, who knows that his investment will follow the fortunes of the company. The RLP was therefore born in the ordinary situation where the shareholder’s loss was a reduction in the value of his shareholding. The scope was narrow, based upon company autonomy and preventing the “proper plaintiff” rule from being circumvented.

The RLP was then considered by the House of Lords in Johnson v Gore Wood [2002] 2 AC 1 (‘Johnson’)2. The company had pursued professional negligence proceedings against solicitors, which it settled for a payment. There was a shortfall in the recovery and Mr Johnson, the managing director and majority shareholder, pursued personal claims against the solicitors in respect of the same episode. An application was made to strike out these claims in light of the company’s claim. Lord Bingham set out the following three propositions, which are regarded as encapsulating the test for the RLP:

1. “Where a company suffers loss caused by a breach of duty owed to it, only the company may sue in respect of that loss. No action lies at the suit of a shareholder suing in that capacity and no other to make good a diminution in the value of the shareholder’s shareholding where that merely reflects the loss suffered by the company. A claim will not lie by a shareholder to make good a loss which would be made good if the company’s assets were replenished through action against the party responsible for the loss, even if the company, acting through its constitutional organs, has declined or failed to make good that loss…

2. Where a company suffers loss but has no cause of action to sue to recover that loss, the shareholder in the company may sue in respect of it (if the shareholder has a cause of action to do so), even though the loss is a diminution in the value of the shareholding…

3. Where a company suffers loss caused by a breach of duty to it, and a shareholder suffers a loss separate and distinct from that suffered by the company caused by breach of a duty independently owed to the shareholder, each may sue to recover the loss caused to it by breach of the duty owed to it but neither may recover loss caused to the other by breach of the duty owed to that other…”3

Lord Bingham explained the RLP by referring to the preservation of company autonomy, avoiding prejudice to the company’s creditors and preventing one party recovering for another’s loss. However, he envisaged some flexibility when applying the principle: “the court must be astute to ensure that the party who has in fact suffered loss is not arbitrarily denied fair compensation.”4 Lord Millett appeared to take a firmer approach, stating that the automatic bar to the shareholder’s claim was a “matter of principle and there is no discretion involved.” He also referred to a number of considerations justifying the RLP, explored further below.

Two years later, the CoA heard Giles v Rhind [2002] EWCA Civ 1428 (‘Giles’) and held that the lower court was wrong to strike out a claim for reflective loss because the wrong done to the company had made it impossible for the company to pursue its claim against the wrongdoer. The defendant, in breach of duty, had diverted the company’s most lucrative contract to another company in which he had an interest. The company commenced proceedings but soon went into administrative receivership and the defendant issued an application for security for costs. The company could not put up the security and had to discontinue the proceedings by consent order which precluded further action against the defendant.

Chadwick LJ decided that none of Lord Millett’s considerations in Johnson were engaged. The company had not settled its claim so concerns about going behind any settlement did not arise. The company also had no choice but to abandon its claim because it was unable to secure the defendant’s costs, so there was no break in the chain of causation between the defendant’s breach and the claimant’s loss. In his view, the House of Lords, when formulating the RLP propositions, had not had in mind nor addressed this kind of scenario and he expressly relied upon Lord Bingham’s warning against arbitrarily denying fair compensation to a party who has suffered loss.5 This has become known as the fraud exception to the RLP.

In Gardner v Parker [2004] 2 BCLC 554 (‘Gardner’), the CoA confirmed that the RLP extended to losses suffered in the capacity as a creditor or employee. Neuberger LJ identified two essential ingredients: i) that the losses were suffered in the claimant’s capacity as a shareholder/creditor/employee; and ii) that the damages would have been made good if the company had enforced its rights. He considered that the foundation and ultimate reason for the principle was the need to avoid double recovery.6 He also stated that the nature of the cause of action made no difference because the principle was not concerned with barring causes of action but with barring recovery of certain types of loss.5

The observation about the irrelevance of the cause of action is, however, ambiguous and begs the question what type of loss is barred. For example, where there are clear differences between the causes of action of the claimant and the company, this could potentially affect the analysis of whether the claimant’s loss arises from the depletion of the company’s assets and is therefore reflective. It is difficult to see how the RLP can properly be applied without considering the relevant causes of action giving rise to the loss incurred by the claimant and the company, in order to consider whether the claimant’s loss is merely reflective of the company’s loss. As noted below, the ambiguity in Neuberger LJ’s comments has given rise to some of the issues that have arisen in Primeo.

Policy justifications

The decisions since Prudential have expanded the RLP well beyond its company autonomy roots. This trend has been justified by reference to a number of policy objectives which the courts have suggested are engaged. In Marex, the CoA suggested that the authorities provided a fourfold justification for the RLP: i) double recovery; ii) causation; iii) conflicts of interest; and iv) company autonomy and prejudice to others.

However, these policy justifications are not compelling reasons for the general application of the RLP and only make sense in certain factual situations. Double recovery was not mentioned in Prudential but assumed particular significance in Gardner. It is based on the notion that it would be unsatisfactory for a defendant to pay out twice to both the claimant and the company. The authorities, however, do not explain why the risk of double recovery is of such significance where there is a corporate relationship in place that it requires a special rule. The potential for double recovery may arise in many contexts outside the scope of the RLP which are not subject to any bar and it can be dealt with satisfactorily without requiring the barring of the claimant’s claim. It is also not clear why the “risk” of double recovery is enough. In many cases the prospects of the company’s claim will not be clear at the time of striking out the claimant’s claim. Why should the risk of the defendant possibly having to pay out twice always outweigh the injustice of a claimant’s valid claim being expropriated? This risk could be addressed, as it is in other situations, by a requirement to give credit for other recoveries.

The causation point has been explained in the authorities as a break in the chain between the defendant’s conduct and the claimant’s loss, such that it can be said that the claimant’s loss arises not through the defendant’s conduct but rather through the decision of the company not to pursue a recovery. However, this is based upon the assumption that the company could have made a recovery but has failed to do so. This was not the case in Marex, where recovery by the company was impractical but not impossible, or in Primeo, where the company’s claim could not be said to be likely to succeed.

The conflicts of interest point has been justified as encouraging settlements between the company and the defendant. However, it is not clear that the RLP has much, if any, role to play in this context. A defendant is likely to analyse the risk posed by a claim, and accordingly whether or not to seek to settle it, by reference to the merits of the underlying claim and tactical considerations as to whether there are other ways in which to dispose of it. If having carried out this risk analysis a defendant perceives sufficient risk to wish to pursue settlement, then the defendant may adopt that course of action. The existence of another potential claim is unlikely to have much impact on the risk analysis carried out by the defendant. In fact, the RLP principle may make settlement less likely, by complicating the position.

Company autonomy was the sole foundation of the RLP in Prudential and in the narrow context of the rule in Foss that is easy to understand. However, company autonomy becomes less compelling as a justification for a broader RLP, and in particular in relation to the suggestion that a claimant may scoop the pool ahead of the company. Firstly, the law does not generally, outside the scope of formal insolvency proceedings, impose any kind of moratorium on proceedings. It is open to creditors, if they wish, to aggressively pursue repayment of their debts even if this has the effect of leaving others with nothing. But, secondly, in many cases, the defendant may well be able to pay both the claimant’s claim and the company’s claim in full, so no issue of scooping the pool will arise. And, if the defendant is in liquidation, both claims will be addressed though the claims procedure and issues of double recovery and scooping the pool will be properly managed by the liquidators. So, this justification might be said to provide a somewhat flimsy foundation for a rule which has the effect of expropriating property from the claimant.


The decision in Marex considered two key points:i) whether the reflective loss bar should apply to claims by unsecured creditors (who are not also shareholders); and ii) the ambit of the fraud exception in Giles.

Marex obtained a judgment in its favour against two companies. The companies could not pay Marex the amount due because Mr Sevilleja, the defendant, had stripped them of their assets after the release of the draft judgment. Marex therefore made a claim against the defendant in damages representing the judgment debt. The CoA found that a claim by an unsecured creditor for loss caused by the abstraction of money from the company was barred by the RLP because there was no logical distinction between a shareholder
and a creditor. Lewison LJ also observed that the authorities since Prudential, which had expanded the scope of the RLP, were currently binding on the CoA. It is therefore for the Supreme Court to determine whether the RLP should apply to unsecured creditors.

As for the fraud exception, the CoA adopted a narrow approach and found that it could only be invoked where it was legally impossible for the company to bring a claim. The unfunded insolvency of a company was not sufficient in itself to engage the exception because the situation could, in theory, be resolved by an injection of funds by a third party shareholder/creditor to enable the company’s liquidator to pursue the claim, or by taking an assignment of the company’s claim. On the facts, the only party that could have funded the insolvency of the company was Marex but this made no difference.

Whilst on one view it can be argued that Marex involved a traditional application of the RLP, the narrowing of the fraud exception and the overall outcome in the case, which is obviously unjust in allowing the fraudster to escape any liability, must be questionable. Given the overall merits of the case, the application of the RLP to bar entirely the claimant’s claim might be said to involve the tail wagging the dog. It will be interesting to see how the Supreme Court approaches this issue.


The other case is the decision in Primeo. Primeo was an investment fund which had invested with Madoff and which, following the collapse of Madoff, brought various claims in the Cayman Islands against its administrator (first defendant) and custodian (second defendant) for losses suffered in the fraud. The claims related to losses suffered by Primeo in the period up to May 2007 when it had invested directly into Madoff’s company, BLMIS, and the defendants had acted for it pursuant to contracts with Primeo.

The CICA found that Primeo had good claims against both defendants in relation to the losses it had suffered but that any recovery for those losses was barred by the RLP. This was the case, according to the CICA, because in May 2007 Primeo had restructured its investments so that it no longer invested directly in BLMIS but rather invested indirectly through another fund called Herald Fund SPC (‘Herald’). As long as Herald had a cause of action with a reasonable prospect of success (i.e. sufficiently strong to withstand an application to strike it out) which, if successful, would result in a recovery which went to reduce or extinguish Primeo’s loss, that was sufficient to invoke the RLP.

In this respect, the CICA rejected the argument that the RLP should only apply when the company’s claim satisfies the higher threshold of being likely to succeed on the balance of probabilities rather than the lower threshold of just having a real prospect of success. In the CICA’s view the lower threshold (which Herald’s claim satisfied) was sufficient.

However, in practice, the real prospect of success threshold can be relatively low. And it is not clear why a claimant which has an extremely meritorious claim with very high prospects of success should have that claim entirely barred simply because the company has a claim which just about manages to scrape over the real prospect of success threshold. If the purpose of the RLP is to prevent double recovery, then it seems arguable that the company’s claim should have to be likely to succeed.

This issue – the threshold merits test which the company’s claim has to satisfy in order for the RLP to be engaged – is a good example of the difficult issues which the extended application of the RLP gives rise to, which have barely been addressed in the authorities so far.

The other issue of principle which arises from the Primeo case concerns the fact that Primeo was not in fact a shareholder in the relevant company (Herald) at the time when it suffered the relevant losses and its own causes of action arose. It only became a shareholder subsequently. The CICA in effect extended the application of the RLP so that it applies not only where the claimant was a shareholder in the company at the time when its own cause of action arose, but also where the claimant was not a shareholder but subsequently acquired shares in a company which has its own cause of action which, if successful, might be said to reduce the claimant’s loss.


The cumulative effect of decisions on the RLP since Prudential and Johnson has been to very substantially expand the boundaries of the principle – so that it now applies where the claimant is a creditor as well as a shareholder in the relevant company, where the company’s claim cannot be said to be likely to be succeed but has sufficient merit to pass a real prospect of success test, and where the claimant was not in fact a shareholder in (or creditor of) the company at all at the time when its claim arose. The policy justifications for this expansionary approach seem questionable. It will be a matter of keen interest to see whether the Supreme Court in Marex and the Privy Council in Primeo seek to rein the RLP back in.

1See the classic definition of the “proper plaintiff” rule in the judgment of Jenkins LJ in Edwards v Halliwell [1950]
2All ER 1064.
2The other decisions in the intervening period are listed in Johnson at 35D.
3See Johnson at 35E– 36A.
4This phrase was expressly relied upon by Chadwick LJ in Giles to justify the fraud exception.
5See Giles at [66], [68] and [70].
6See Gardner at [49]and [57].

The content of the Digest is provided to you for information purposes only, and not for the purpose of providing legal advice. If you have a legal issue, you should consult a suitably-qualified lawyer. The content of the Digest represents the views of the authors, and may not represent the views of other Members of Chambers. Members of Chambers practice as individuals and are not in partnership with one another.
Read full article
Tom Smith KC
Peter Hayden Mourant
Peter Hayden Mourant
Jonathan Moffatt Mourant
Jonathan Moffatt Mourant
Featured in
Related Articles
South Square & RISA BVI Conference – 2024
South Square & RISA Cayman Conference – 2024
The Lawyer Awards Finalist 2024
Chancery Bar Association’s Charter for Fairness
South Square Digest
South Square Digest is our quarterly review of news, cases, judgments and articles covering all our practice areas.
Contact us
Scroll To Top