Clifford Chance

Almost 20 years on from the beginning of the conduct that led them to be prosecuted and 10 years after the first of them was convicted for conspiracy to defraud based on bench mark rate submissions, the Supreme Court has quashed the convictions of two former traders.

Summary

Tom Hayes and Carlo Palombo were tried and convicted of conspiracy to defraud for seeking to engineer LIBOR and EURIBOR submissions to favour their trading positions. The Court of Appeal twice upheld those convictions, including once by declining to take account of the effect of decisions in separate LIBOR related cases in the US.

The Supreme Court has now held that the convictions of Mr Hayes and Mr Palombo were unsafe. It found that the trial judges wrongly treated what was in truth a factual issue for the jury – namely whether the submissions reflected the submitter’s genuine opinion of the bank’s borrowing cost – as a question of law that the judges then decided.

Background

The essence of the allegations against each of Mr Hayes and Mr Palombo was that (in separate cases) they had dishonestly conspired with others to procure daily rate submissions that were false or misleading because, rather than reflecting a genuine borrowing rate assessment, the figures were chosen to benefit their trading books.

Both defendants were prosecuted for the common law offence of conspiracy to defraud. Mr Hayes was tried alone at Southwark Crown Court, convicted by the jury on eight counts on 3 August 2015 and initially sentenced to fourteen years' imprisonment (later reduced to eleven years on appeal). Mr Palombo was first tried with co-defendants in 2018 and later re-tried in 2019 (since the jury in his first trial was unable to reach a verdict). He was convicted in March 2019 of a single count of conspiracy to defraud and sentenced to four years’ imprisonment.

Initial appeals

Their original appeals to the Court of Appeal were dismissed (Mr Hayes' in December 2015 and Mr Palombo's in December 2020).

In his original appeal Mr Hayes argued that Cooke J, in effect, directed the jury that any LIBOR submission tainted by consideration of trading advantage could not, as a matter of law, be a “genuine or honest” assessment of the bank’s borrowing cost. That, it was said, removed from the jury the factual question whether the figures put forward truly represented the submitters’ opinions and thereby usurped the jury’s function. The Court of Appeal rejected the argument summarily, holding that the judge had done no more than articulate the legal obligation on a panel bank to tender a bona fide borrowing rate, and that any submission influenced by commercial interest necessarily fell outside that obligation.

Similarly, Mr Palombo argued that HH Judge Gledhill had misconstrued the EURIBOR Code of Conduct and that the jury should have been permitted to consider expert evidence showing that, under Belgian law, a panel bank could lawfully select any rate within a legitimate range even if motivated by trading advantage. The Court of Appeal also rejected that argument.

US proceedings

In the United States, the Department of Justice had brought a series of criminal actions springing from the same alleged manipulation of inter-bank reference rates. The most prominent of these was United States v Connolly & Black, tried in New York in 2018. Matthew Connolly and Gavin Black were convicted of conspiring to transmit false and deceptive USD-LIBOR submissions. Their convictions followed the Court's acceptance of prosecutors' “one-true-number” theory – namely that each panel bank was required to submit a single objectively correct borrowing rate. On 18 January 2022 the US Court of Appeals for the Second Circuit overturned those convictions, holding that the prosecution had failed to prove any false statement because the rates submitted remained within a range at which the bank could in fact borrow. That appellate ruling not only marked the effective collapse of the US prosecution campaign but also precipitated the dismissal of the outstanding US indictment against Mr Hayes, whose extradition had previously been sought on overlapping allegations concerning JPY-LIBOR.

Second appeals

Mr Hayes had already applied to the Criminal Cases Review Commission ("CCRC") in 2017, but in December 2021 its provisional view was that no miscarriage of justice had occurred. When Connolly & Black was decided, the CCRC invited supplementary submissions. In July 2023, it concluded that there was now a ‘real possibility’ that the Court of Appeal would regard the English jury directions as legally erroneous. On that basis it referred Mr Hayes’ 2015 convictions back to the Court of Appeal. For similar reasons, in October 2023 it referred the 2019 conviction of Mr Palombo, whose jury had received materially similar directions.

The Court of Appeal again dismissed both individuals' appeals.

The question for the Supreme Court

The Supreme Court granted permission to review a point of law of general public importance certified by the Court of Appeal. The certified question asked whether as a matter of law upon the proper construction of the LIBOR and EURIBOR definitions:

(a) If a LIBOR or EURIBOR submission is influenced by trading advantage, it is for that reason not a genuine or honest answer to the question posed by the definitions; and
(b) the submission must be an assessment of the single cheapest rate at which the panel bank, or a prime bank, respectively, could borrow at the time of submission, rather than a selection from within a range of borrowing rates.

The Supreme Court's judgment

The Supreme Court answered both elements of the certified question "no".

Mr Hayes

It held that Mr Hayes’ convictions were unsafe because the trial judge himself decided as a question of law the issue of whether the figures submitted to the LIBOR panel reflected the submitter’s genuine opinion of the bank’s borrowing cost. It held that this was a factual question that should have been left to the jury to determine. By telling the jury that any submission influenced by trading advantage was, as a matter of law, not a “genuine or honest” answer to the LIBOR question, the trial judge effectively deprived the jury of the opportunity to decide whether Mr Hayes had agreed to procure fraudulent misstatements, a task which was squarely theirs, rather than the judge's.

The Supreme Court drew a distinction between questions that engage a document’s legal effect and those that are concerned solely with what the words were taken to mean. It made clear that where a document (for example a statute, contract or code that actually creates or records legal rights and obligations) has to be construed so that the court can identify the legal relations it establishes, that task is for the judge. Construction in that sense involves determining the document’s legal consequence. The judge must then instruct the jury on the obligations, powers or duties that flow from it. By contrast, understanding the ordinary meaning of the language used, or deciding what the drafter or the parties intended a statement or question to convey, is a factual question for the jury.

Since the jury were directed that consideration of trading advantage automatically rendered a submission false, Mr Hayes’ defence - that he sought only figures within a legitimate range honestly held by the submitters - was never properly considered. The Supreme Court accepted Mr Hayes' argument that there would on any given day be a range of different borrowing rates that could be reasonably selected by the submitter. Those misdirections permeated the summing-up and the “route to verdict”. The Supreme Court held that accordingly the convictions had to be quashed.

Mr Palombo

In Mr Palombo’s case the Court reached the same result, although the context differed. Again, it accepted that there was likely to be a range of rates from which a submitter would have made a selection using their subjective judgment. It held that the EURIBOR Code of Conduct was a binding contract governed by Belgian law, and the trial judge was entitled to decide, as a matter of law, that the Code prohibited a submitter from taking commercial advantage into account. The trial judge's error in that case lay in instructing the jury that a submission breaching that contractual duty was, for that reason alone, a false or misleading representation capable of sustaining a charge of conspiracy to defraud.

The Supreme Court held that, as in Mr Hayes' case, that direction similarly prevented consideration by the jury of a factual question - whether the submitter’s figure misstated his or her genuine opinion. It held that this error was compounded by several other features of the trial judge's directions. Firstly, by telling the jury that taking account of trading advantage was against the law of England and Wales at the time of the agreement overstated the gravity of the alleged conduct. Secondly, the judge’s additional guidance on “deliberate disregard” failed to make clear that the prosecution had to prove knowledge of, and intentional departure from, the submitter’s true belief. The Supreme Court decided that, taken together, these misdirections meant the jury could have convicted without finding the essential ingredient of fraud, so Mr Palombo's conviction was also unsafe and had to be set aside.

Analysis – what is the effect of the decision?

The main significance of the Supreme Court's decision lies in the clear messages it conveys to judges about the limits of their decision-making remit in fraud cases. In this respect, it provides hugely important guidance on the dividing line between questions of law and construction, which are matters for a judge to determine, and questions of fact, which must be left to the jury.

The judgment also contains strong criticism of the drafting of the indictments in both cases. It traces the extensive efforts made by lawyers for both individuals to clarify the precise ways in which they committed the offence of conspiracy to defraud. It suggests that failures adequately to do so meant that the defence and the trial judge did not have sufficient particulars to know clearly and precisely the prosecution's case and contributed to creating the conditions in which the trial judges gave the defective directions which have rendered both convictions unsafe.

A significant period of time has passed since the underlying conduct upon which the prosecutions were based and the prosecutions themselves. Other, more simply formulated offences, have been added to the statute book during this intervening time (in particular under section 91 of the Financial Services Act 2012, which created a new criminal offence of making misleading statements in relation to benchmarks). The judgment also notes that offences under the Fraud Act 2006 could have provided an alternative basis for prosecution.

It is unlikely that the same issue will recur (not least because of the cessation of LIBOR and EURIBOR rates which were reference rates based on expert judgment or submissions and their replacement by Risk-Free Reference Rate alternatives based on actual transaction data and which involve minimal or no subjective elements). However, the offence of conspiracy to defraud remains available for use by prosecutors. The judgment provides a salutary warning about the need for clarity when using this offence and provides defence lawyers with an aid to challenges to vaguely characterised misconduct.

More immediately, the setting aside of the convictions of Mr Hayes and Mr Palombo has also called into question the safety of the convictions of seven others convicted of offences related to LIBOR and EURIBOR manipulation where similar directions were given to the juries. It is possible that further challenges may follow.