The ongoing regulatory investigations into the manipulation of LIBOR have prompted much speculation about the possibility of claimants bringing private law actions against banks based on the regulators' findings. In the first cases to come before the UK courts, the Court of Appeal has recently granted permission to appeal two High Court decisions as to whether LIBOR-based claims could be introduced into existing actions alleging mis-selling of interest rate hedging products.

In light of those decisions, this briefing considers the prospects of such LIBOR claims in the mis-selling context and identifies substantial obstacles claimants will face in seeking to 'piggyback' off the regulatory actions.

1. The High Court decisions

Both Graiseley Properties Ltd & Ors v Barclays Bank plc [2012] EWHC 3093 (Comm) and Deutsche Bank AG & Ors v Unitech Global Ltd & Anor [2013] EWHC 471 (Comm) concerned applications to the Court to amend the claimants' statements of case. The threshold for permission to amend is relatively low, with the party being required to show only that the proposed amendments are sufficiently arguable to have a real prospect of success at trial.

Graiseley Properties Ltd’s (“GPL”) claim for mis-selling in respect of an interest rate swap and loan originally included an allegation of innocent misrepresentation in relation to LIBOR manipulation. Following the regulatory findings last year against Barclays, Flaux J granted GPL permission to amend its particulars to include a plea of fraudulent misrepresentation. By contrast, Cooke J did not allow Unitech Global Ltd and Unitech Ltd (together, “Unitech”) to amend their defence and counterclaim - to a claim by Deutsche Bank for payment under a credit facility and interest rate swap – so as to include similar allegations in respect of LIBOR manipulation.

2. Tortious claims – misrepresentation / negligent mis-statement

Both GPL and Unitech focused on representations they allege were impliedly made by the respective banks as to the integrity of LIBOR. The implied representations were pleaded differently by GPL and Unitech, but in broad summary were that:

  1. LIBOR represented the interest rate defined by the British Bankers’ Association (“BBA”) (the body with responsibility for determining the rate);
  2. LIBOR was based on the panel banks’ good faith accurate estimates of the rates they could borrow from one another / the defendant bank had no reason to believe LIBOR represented anything other than the rate at which contributor banks could borrow funds;
  3. the bank in question had not undermined the integrity of LIBOR / made false submissions or intended to manipulate LIBOR, and had no intention of doing so; and
  4. the bank in question was not aware of any conduct which would undermine the integrity of LIBOR (the "LIBOR representations").

GPL and Unitech's primary case was for misrepresentation on the basis that the implied representations were untrue. Each also pleaded negligent mis-statement based on breach of a duty to take reasonable care that the representations were true.

As discussed below, the likely reason for this focus on claims based on misrepresentation and tort is that the remedies available and measure of damages are more attractive - in these particular circumstances - than those available in contract.

However, potential claimants face a number of formidable obstacles in establishing a bank's liability for misrepresentation or mis-statement in these circumstances.

Finding an implied representation

Absent an unusual fact pattern, claimants are unlikely to be able to point to any express representation by a bank as to the integrity of LIBOR. This was the case with both GPL and Unitech, who had to plead that the LIBOR representations should be implied from conduct and the surrounding circumstances Implied representations are not easily made out, with the legal test being that a court must consider what (if anything) a reasonable person would have inferred was being implicitly represented by the representor's words and conduct in their context.1

Cooke J and Flaux J took quite different approaches to this question and arrived at different answers. Flaux J referred to the US Department of Justice's (DoJ) finding that the traders and submitters at Barclays involved in the manipulation of LIBOR had been aware of the effect their actions could have on the bank's counterparties, such as the claimant. He also held that the same was arguably true of those individuals in Barclays’ senior management who had been involved. Flaux J concluded from this that, in the circumstances, there was an arguable basis for implying that Barclays had made the LIBOR representations. His reasoning in this regard is not entirely clear but he appears to have been primarily concerned with GPL's prospects of being able to establish that the representations were fraudulent, rather than that they had been made at all (see further below).

Cooke J, on the other hand, considered that the awareness of the representor (Deutsche Bank) as to the consequences of manipulating LIBOR in fact had no bearing on the test to be applied. By contrast, Cooke J focused – in line with the authorities – on whether a reasonable person in the position of the representee (Unitech) would have inferred a representation.

Cooke J had specific concerns about the width and uncertainty of the individual LIBOR representations - particularly the first two, which, being statements about the integrity of the overall system of LIBOR-setting, were not within any individual bank’s control. Unitech’s most significant problem though was that it was unable to point to any specific pre-contractual conduct or statements to underpin the alleged representations. As such, its case was simply that a bank entering into a transaction referenced to LIBOR when that bank was on the LIBOR setting panel in itself gave rise to the implied representations. Cooke J rejected that plea as unsustainable and having no prospect of success.

Why did the judges take these different approaches? As Cooke J noted, the Court in Graiseley had different material before it, most importantly specific documents and communications from which it was said the representations could be implied, as well as the FSA and DoJ findings against Barclays (Deutsche Bank remains in discussions with regulators). Furthermore, in Graiseley, GPL had already pleaded misrepresentation as to LIBOR and the Court was required to decide whether the particulars could be amended to allege fraud, for which the state of the bank's knowledge would clearly be relevant. This is perhaps why Flaux J focused on Barclays and its awareness of the potential impact of LIBOR fixing. Nonetheless, Cooke J's approach of concentrating instead on the swap counterparty, and what a person in its position would have inferred from the bank's statements and conduct, appears to be more consistent with the authorities on implied representations.


Banks' contractual documentation will generally contain some form of disclaimer relating to representations. Four such clauses were identified in Unitech:

  1. A disclaimer of responsibility in the swap term sheet for any representations as to information provided.
  2. An 'entire agreement' clause in the ISDA Master Agreement, in which the parties acknowledged they had not relied on any representations outside of those set out in the Master Agreement and the swap confirmation.
  3. Statements in the swap confirmation to the effect that Unitech had not relied on any communication as investment advice or a recommendation, and that they were capable of assessing and understanding on their own behalf the merits and risks of the transaction.
  4. A confirmation in the term sheet for the loan that the bank was not acting as an adviser or fiduciary.

As Cooke J found against Unitech on other grounds, he went no further than saying that such provisions "militate against [the LIBOR] representations". Contractual disclaimers of this sort are likely to be helpful to banks in two ways: (1) they may make it harder to establish that the bank in fact made any express or implied representation as to LIBOR; (2) they may set up a contractual estoppel such that – even if a representation as to LIBOR was made – the swap counterparty is precluded from pleading reliance on it.

Importantly, however, such disclaimers are not likely to be effective where fraudulent misrepresentation is found. Fraud is considered further below.

Remedies / damages

Cooke J’s judgment in Unitech articulated the key underlying dynamic – that remedies in misrepresentation or tort will usually be more attractive to claimants, as they may give rise to the transaction being set aside or a discharge from liability, whereas contractual damages are likely to be much more limited and difficult to establish factually.

This is because misrepresentation may entitle the injured party to rescind (set aside) the contract or to damages in lieu, which in the context of a hedging transaction could potentially amount to compensation for all sums paid to the bank under the swap and all break costs. Similarly, damages in tort seek to put the injured party in the position it would have been in had the relevant statements not been made, which position could be that the party would never have entered the transaction in the first place. Thus, Unitech did not seek to identify any specific loss suffered as a result of the LIBOR representations being made but, rather, alleged that they would not have agreed to the swap at all had the representations not been made. (Of course, claimants making such an assertion will need to satisfy the court of that as a matter of fact, on the balance of probabilities, and could expect to be strongly challenged in that regard.)

Damages for breach of contract, on the other hand, will be premised on the relevant statements having been true, and may therefore be restricted to the difference resulting from the application to the contract of the actual (manipulated) LIBOR and what would have been the “true” rate (see further below for the difficulties involved in establishing damages for breach of contract).

3. Fraud and senior management

Clearly, it will be advantageous to a claimant if it is able to establish fraud on the part of a defendant bank. Fraud is likely to vitiate the contractual disclaimers discussed above and may also be relevant to issues of limitation and damages.

In Unitech, fraud was not considered, as the judge found that no representations had been made in any case. The Court in Graiseley, however, held that a pleading of fraud had sufficient prospects of success to go forward to trial. For a representation to be fraudulent, the representor must know it is untrue or be reckless as to its truth. Flaux J relied in this regard on the regulatory findings that Barclays' senior management was responsible for or aware of LIBOR manipulation. Knowledge that the LIBOR representations were untrue could therefore be imputed to the bank.

In most such cases (as in Graiseley), the swap salesperson or relationship manager dealing with the client is unlikely to have known about the bank's manipulation of LIBOR at the time. The extent of senior management awareness or responsibility is therefore likely to be a key factor in the ability of a claimant to establish the requisite knowledge on the part of the bank. Where regulatory findings do not include senior management awareness of wrongdoing, those banks should be less vulnerable to allegations of fraud.

4. Contractual claims

Claimants may have better prospects of success in establishing contractual claims, for breach of either an implied term or collateral warranty as to the integrity of LIBOR. Flaux J dealt very briefly with the suggested terms in Graiseley, finding it was clearly arguable they were to be implied into the relevant contracts.

Cooke J on the other hand rejected Unitech's proposed warranty / implied term, as it was pleaded in the same terms as the misrepresentations and related to “past facts, present practice and present intention vis a vis LIBOR generally.” However, he indicated that, if differently pleaded, he could envisage a warranty / implied term which could meet the test – for example, a “future obligation not to jeopardise the proper process of LIBOR assessment by self-interested action.”

However, the fundamental obstacle to contractual claims in these circumstances will be to establish causation and loss. As Cooke J noted, the damages that might flow from a breach of the kind of implied promise he envisaged 'might be hard to establish..'. Contractual damages will put the claimant in the position it would have been had the warranty / term as to LIBOR’s integrity not been breached (that is, had LIBOR not been manipulated). Claimants will face particular difficulties establishing:

  • What the bank’s LIBOR submissions would have been but for the wrongdoing. On any given day, trader-submitter communications may or may not evidence a causal link between impropriety on the part of the bank and its LIBOR submissions (for the relevant currency and tenor). Further, as guidelines published by the BBA in 2009 state, “a bank’s LIBOR submissions are its own perception of where it could take funds” in the market and, as such, involve a degree of subjective assessment. It will therefore be an extremely challenging undertaking to re-create what the bank’s daily “true” LIBOR submissions would have been.
  • What the published LIBOR rate on any particular day would have been but for the individual bank’s manipulated submissions. LIBOR is calculated by way of a filtered average of the submissions of eight of the 16 contributor banks, with the four highest and four lowest submissions excluded. Given this method of determining LIBOR, the impact of any one bank’s false submissions on the rate is likely to have been extremely small (though as the submissions are publicly available, it would presumably be possible to calculate, as long as the individual bank’s “true” submissions can be established, as above).
  • What loss the claimant itself has suffered.
    • Some false submissions may have moved LIBOR in ways unfavourable to the claimant but some may have been favourable.
    • Furthermore, where a party has a linked loan and swap, any loss it incurred on one of those contracts due to an artificial LIBOR fixing may be (at least partially) offset by the related contract. Indeed, the purpose of many interest rate hedging products is of course to insulate the counterparty from the effects of movements in rates.
    • Further, even if (i) an impact on the published LIBOR rate and (ii) a resultant loss to the claimant can both be established for particular days in respect of which there is evidence of manipulation (such as trader/submitter emails), how are those individual daily losses to be extrapolated to the entire period over which the claimant alleges loss? Such an undertaking might require the application of relatively complex mathematical models. While the courts will not deprive a claimant of compensation for loss purely because the loss is difficult to quantify precisely, the 'piecemeal' nature of the evidence disclosed in the regulatory investigations could be expected to present claimants with an additional hurdle in establishing the extent of their loss.

Claimants therefore are likely to fact substantial difficulties establishing loss for breach of contract, and what loss they can show may ultimately be relatively limited in quantum.

5. Conclusion

It should be remembered that Graiseley and Unitech were both applications to amend statements of case and, as such, concerned only with the question of whether those amendments were sufficiently arguable to go to trial to be tested. Nevertheless, some of the difficulties that private law actions will face are apparent from the judgments. As Cooke J recognised, a collateral warranty or implied promise not to manipulate LIBOR is perhaps “a more logical plea to make than a series of implied representations by conduct”. However, the problem of establishing causation and substantial loss in respect of contractual claims is likely to force claimants to plead such implied representations. In doing so, they will face a significant evidential challenge, as in most cases it seems highly unlikely that they will be able to point to any specific focus by the parties at the time on the integrity of LIBOR.

The decision in Graiseley is arguably the more surprising of the two, particularly as Flaux J was prepared - on the basis of relatively limited specific materials - to permit a pleading not only of implied representations but also of fraud. However, as always, each case will turn on its facts and the regulatory findings particular to Barclays and the conduct of its senior management were clearly a central factor in the decision.

The Graiseley appeal is listed for between September 2013 and January 2014 and Unitech for between July and December 2013, and it is possible that the two will be heard together. We will continue to report any developments.