Benchmark-Rate Fraud: Commercial Influence Does Not Establish a False Submission as a Matter of Law

1. Introduction

This judgment quashed five convictions for conspiracy to defraud arising from the submission of interest-rate information used to calculate LIBOR and EURIBOR. Its central concern was the boundary between a judge’s directions on law and a jury’s responsibility to determine facts: commercial influence on a rate submission could support an inference that the submission did not represent a genuine opinion, but it could not legally compel that conclusion.

The decision principally applies, rather than creates, the rule established by the Supreme Court in R v Tom Hayes and R v Carlo Palombo [2025] UKSC 29; [2025] 2 Cr App R 21. Its significance lies in extending that reasoning to five further convictions, explaining how the same error spread through successive trials, and demonstrating the need to assess the actual directions and circumstances of each case.

Three appellants had been convicted in a LIBOR trial. The other two had been convicted in separate EURIBOR trials, the second being a retrial following jury disagreement. Their cases reached the Court of Appeal through references by the Criminal Cases Review Commission after earlier appellate proceedings had failed.

Although the Serious Fraud Office did not oppose the appeals, the court stressed that it retained an independent constitutional responsibility to decide whether the convictions were safe. The judgment determined five appeals only. The position of a further defendant who had pleaded guilty was reserved for a separate hearing.

2. Background and Procedural History

The benchmark-rate prosecutions

The prosecutions concerned submissions made in the processes for setting LIBOR and EURIBOR. The relevant definitions called for assessments concerning borrowing rates. The central question was whether submissions influenced by the commercial interests of a bank or trader nevertheless represented the submitter’s genuine opinion about the rate requested.

The trials proceeded on directions treating commercial influence as legally inconsistent with a genuine or honest response to the benchmark definition. That approach removed from the jury a factual question which the Supreme Court subsequently held belonged to it.

The five convictions

• The LIBOR proceedings: three appellants were convicted of conspiracy to defraud following a trial before His Honour Judge Leonard KC at Southwark Crown Court. Earlier challenges to their convictions had been unsuccessful, although one appellant obtained a sentence reduction.
• The first EURIBOR trial: one appellant was convicted in proceedings before His Honour Judge Gledhill KC. He was sentenced in his absence and was subsequently the subject of a European Arrest Warrant. The Court of Appeal of Paris refused his surrender.
• The EURIBOR retrial: another appellant was convicted after the first jury had been unable to agree. His retrial was conducted alongside the defendant whose conviction later came before the Supreme Court.

The convictions had also generated substantial financial orders, including confiscation and prosecution costs orders. The Court of Appeal recognised that quashing the convictions required further consideration of those consequences, but adjourned outstanding financial issues to the following day.

3. Summary of the Judgment

The court accepted the common ground of appeal: the prosecution approach and jury directions contained the same legal error identified by the Supreme Court, rendering all five convictions unsafe.

1. Commercial influence was evidence, not a legally conclusive answer. A jury could infer that a commercially motivated submission was not genuine, but the judge could not require that inference as a matter of law.
2. The error affected all three relevant trials. Although the directions were not identical, they shared the same fundamental defect.
3. Differences between the EURIBOR trials did not preserve the first conviction. Materially defective directions were common to both trials, notwithstanding differences in the prosecution’s submissions.
4. All five convictions were quashed. The prosecution had stated that there would be no retrials if that occurred.
5. The European Arrest Warrant was to be withdrawn. This concerned the appellant convicted in his absence.
6. Some matters remained unresolved. Outstanding financial consequences were adjourned; two additional grounds raised in one appeal were unnecessary to decide; and the guilty-plea case was reserved for separate consideration.

4. Precedents Cited and Their Influence

R v Tom Hayes and R v Carlo Palombo [2025] UKSC 29; [2025] 2 Cr App R 21

This was the controlling authority. The Court of Appeal considered that it determined the substance of the five appeals, while deliberately avoiding an extensive restatement of the Supreme Court’s reasoning.

Paragraph 8 of the Supreme Court judgment identified the essential error: a judge could not direct that any consideration of a bank’s or trader’s commercial interests necessarily made a submission nongenuine or dishonest. Commercial influence could justify an inference against the defendant, but whether to draw that inference remained for the jury.

Paragraphs 93 and 123 explained how the error arose in the LIBOR proceedings. The trial took a wrong turn when the judge treated the meaning of the benchmark document as supplying a legally decisive answer about the submission. The material question was how the defendant and other relevant individuals understood the document—a factual inquiry—not simply how the court construed it.

Paragraph 163 traced the same approach into EURIBOR prosecutions, where the benchmark documentation was accepted to constitute a binding contract. That contractual setting did not justify removing the relevant factual inquiry from the jury.

Finally, paragraphs 227–233 were important to the Court of Appeal’s examination of the differences between the two EURIBOR trials. The Supreme Court had considered not only the central misdirection but also additional defects in directions about the proper basis for submitting EURIBOR rates and particular prosecution submissions. The present court therefore examined whether the earlier trial shared enough of those defects to make its conviction unsafe. It concluded that it did.

R v H [2015] EWCA Crim 46

This interlocutory appeal formed an early stage in the development of the erroneous approach. It concerned a pre-trial ruling by Cooke J in the original LIBOR prosecution.

The Court of Appeal explained that Cooke J understood the appellate decision as affecting a further question which had not actually been before the appellate court: how the jury should be directed on the offence’s elements and the significance of the LIBOR Code of Conduct.

Its importance here was therefore historical and analytical. It illustrated how a decision on one issue came to be treated as resolving another, contributing to directions later reproduced in other trials.

The earlier appeal reported at [2015] EWCA Crim 1944; [2018] 1 Cr App R 10

This decision expressly approved the directions used in the original LIBOR trial. It was available before Judge Leonard directed the jury in the three appellants’ trial.

The present court regarded his adoption of those directions as unsurprising: they had received appellate endorsement. Nevertheless, their established use could not preserve the resulting convictions once the Supreme Court had identified the underlying error.

R v Farr and Others

In this intervening trial, Hamblen J modelled his directions closely on those given by Cooke J. All defendants were acquitted, so the directions did not receive appellate examination.

The case demonstrated the spread of the approach through successive prosecutions. The Court of Appeal did not treat those acquittals as an endorsement of the directions or as authority on their correctness.

R v Merchant and Mathew [2017] EWCA Crim 60

This decision formed part of the earlier appellate history of two LIBOR appellants. One appellant obtained permission to appeal against conviction and sentence; the conviction appeal failed, but the sentence was reduced. The other was refused permission to appeal against conviction.

Its relevance was procedural: it showed that earlier appellate consideration had not corrected the error subsequently identified by the Supreme Court. The Commission’s references provided the route for further consideration.

The appeal reported at [2018] EWCA Crim 420

The remaining LIBOR appellant had previously challenged the competence and integrity of a prosecution expert witness. Permission to appeal was granted, but the appeal was dismissed.

That challenge was distinct from the common ground now upheld. The judgment therefore distinguishes between an earlier unsuccessful evidential challenge and a later successful challenge to the legal basis on which the jury was directed.

5. Analysis of the Court’s Legal Reasoning

A. A possible inference must not become a compulsory conclusion

The core distinction was between evidence capable of supporting a conclusion and a legal rule requiring that conclusion.

A request for a submission favourable to a trading position may be powerful evidence that the submitted figure did not genuinely reflect the submitter’s assessment. But that possibility does not make the conclusion inevitable. The jury must evaluate the evidence, the submitter’s understanding and the circumstances.

The defective directions effectively treated the presence of commercial influence as resolving the issue. They therefore displaced the jury’s assessment rather than merely explaining the law it should apply.

This was not a ruling that commercially influenced submissions are necessarily lawful. Nor did it establish that a defendant’s assertion of genuine belief must be accepted. It preserved the jury’s responsibility to decide whether the relevant opinion was genuine and to evaluate the prosecution’s case under proper directions.

B. Construing a document is different from determining a defendant’s understanding

A court may decide what a document means as a matter of law. That does not necessarily determine how an individual understood it, whether a particular submission expressed that individual’s genuine opinion, or what the evidence establishes about participation in the alleged dishonest agreement.

The distinction was especially significant in the EURIBOR proceedings. A preparatory ruling had treated the documentation as a contract governed by Belgian law, and an interlocutory appeal had upheld that ruling.

The Court of Appeal did not hold that the contractual interpretation was wrong. It expressly recognised that it might have been correct. The error lay in using that interpretation to withdraw a separate factual issue from the jury.

The decision consequently does not rewrite the general law of dishonesty or make a defendant’s personal interpretation legally decisive. Its narrower point is that legal construction cannot substitute for the factual inquiries required in the particular prosecution.

C. Established directions remained open to correction

The judgment carefully traced how the error became embedded: an initial ruling was interpreted broadly, trial directions followed, those directions received appellate approval, and later judges adopted them.

This history explains the repetition without curing it. The court distinguished the understandable conduct of judges applying established authority from the separate question whether the resulting convictions remained safe after the Supreme Court’s correction.

D. The first EURIBOR trial required individual examination

The prosecution drew attention to differences between the first EURIBOR trial and the retrial considered by the Supreme Court. Some criticised prosecution submissions had been made only at the retrial. Moreover, the Supreme Court had questioned whether the central defect, considered alone, necessarily justified interference with the conviction.

The Court of Appeal therefore did not rely solely on a broad resemblance between the cases. It examined the accompanying directions. Both juries had received the same five-point explanation of the proper basis for submitting EURIBOR rates, and material aspects of that explanation had also attracted Supreme Court criticism.

Those shared deficiencies meant that the differences in advocacy did not alter the result. Read as a whole, the reasoning supports a contextual assessment of safety, not a universal rule that any reference to the defective approach automatically determines every related appeal.

E. The prosecution could not concede away the court’s function

The Serious Fraud Office’s decision not to oppose the appeals was important, but not conclusive. The court still required the relevant material, preparation and submissions to discharge its own responsibility.

This is a significant procedural point: agreement between the parties does not itself establish that a criminal conviction is unsafe. The appellate court must independently reach that conclusion.

6. Complex Concepts Simplified

Conspiracy to defraud

An offence concerned with an agreement to act dishonestly so as to prejudice another’s economic interests or rights. These appeals concerned whether the juries had been properly directed when deciding the allegations.

A genuine opinion

An assessment actually held by the person expressing it. A commercial motive may cast doubt on that assessment, but does not automatically prove that the person did not hold it.

Fact and law

Judges explain legal rules; juries generally determine disputed facts. Here, the error was treating a factual conclusion about a submission as already dictated by law.

An unsafe conviction

A conviction which cannot properly stand on appellate review. Quashing it for defective directions does not itself amount to a factual finding that the alleged conduct never occurred.

A Commission reference

A referral by the Criminal Cases Review Commission returning a conviction to the appeal court for consideration, including after earlier appeals have failed.

An interlocutory appeal

An appeal concerning an issue arising before the proceedings have finally concluded. Its effect must be understood by reference to the issue actually decided.

A confiscation order

A financial order directed at recovering the benefit of criminal conduct. This judgment recorded existing orders but left outstanding consequential financial issues for a further hearing.

7. Impact and Limits

For benchmark-rate prosecutions, the judgment confirms that commercial motivation cannot replace proof of the factual propositions underlying the alleged fraud. Directions must distinguish the evidence supporting an inference from a legal requirement to draw it.

For complex financial crime more broadly, the reasoning cautions against treating the interpretation of market documentation or contractual standards as a complete answer to questions about a defendant’s understanding and conduct. Its application outside benchmark cases will depend on the offence and the issues actually requiring proof.

For historical convictions, the decision demonstrates the importance of reviewing cases affected by a later authoritative correction. Earlier appellate failure does not necessarily prevent reconsideration through the Commission’s reference process. Nevertheless, the actual directions and surrounding circumstances still matter.

For consequential relief, the judgment separates quashing convictions from resolving every resulting financial issue. It did not finally determine the outstanding treatment of confiscation, costs or other financial consequences.

For guilty pleas, no conclusion should be inferred from this decision. The court expressly reserved the related guilty-plea case, indicating that its position required separate consideration.

8. Conclusion

The Court of Appeal quashed five convictions because the juries had been directed on an erroneous basis: commercial influence had been treated as legally determining whether benchmark submissions were genuine, rather than as evidence for the jury to assess.

The judgment’s broader significance is its insistence on maintaining the boundary between legal interpretation and factual adjudication. Neither established practice, prior appellate approval nor prosecutorial agreement could replace the court’s independent examination of whether the convictions were safe.