Merchant & Anor v R.

Permission to appeal following refusal by the Single Judge is sought by:

Jay Merchant, a 45 year old man and a former US dollar interest rate derivatives (“swaps”) trader at Barclays Bank plc (“the bank”), and

Jonathan Mathew, a 35 year old man who worked on the bank’s cash desk, buying and selling cash and making, with others, daily London Interbank Offered Rate (“LIBOR”) submissions to the British Bankers’ Association (“BBA”).

Mr Merchant and Mr Mathew, together with Alex Pabon, were convicted on 29 June 2016 in the Crown Court at Southwark, following a trial before HH Judge Leonard QC and a jury, of conspiracy to defraud in respect of fixing the LIBOR rate for the US dollar. Peter Johnson, who had been Mr Mathew’s manager on the cash desk and who also made LIBOR submissions on behalf of the bank, had earlier pleaded guilty to conspiracy to defraud. Mr Merchant was sentenced to 6½ years’ imprisonment, and Mr Mathew was sentenced to 4 years’ imprisonment.

The jury could not agree in relation to two co-defendants, who are awaiting a retrial. For that reason reporting restrictions were imposed prohibiting the reporting of the arguments on the appeal on 1 February 2017, and the issue about what can be reported will be revisited after this judgment has been delivered and representations have been made by the press and relevant parties.

Mr Merchant seeks permission to appeal against conviction and sentence. Mr Mathew seeks permission to appeal against conviction. Mr Pabon had sought permission to appeal against conviction but he has not renewed his application following refusal by the single judge.

For the detailed reasons set out below we have granted Mr Merchant permission to appeal against conviction on two grounds, but dismissed his appeal against conviction. We have granted Mr Merchant permission to appeal against sentence and allowed the appeal by reducing his sentence to a sentence of 5½ years’ imprisonment. We have refused Mr Mathew’s renewed application for permission to appeal against conviction. We should note that Mr Mathew sought permission to adopt the points against conviction made by Mr Merchant if we considered them to be well-founded. In circumstances where we have dismissed Mr Merchant’s appeal against conviction, it did not become necessary to consider this application.

In order to calculate the LIBOR setting, the BBA obtained daily submissions from 16 banks, known as panel banks, in answer to what was known as the LIBOR question. So far as is material the Libor question asked by the BBA was

Some trades by the bank and others would be made using the LIBOR setting. Trades were part based on predictions made by the traders of the future LIBOR setting, who would take a position against other traders. Depending on the position taken by the trader, a higher or lower LIBOR setting would increase or decrease the profits (or losses) made on the trade.

So far as is material the particulars of offence of conspiracy to defraud in the indictment alleged that:

It is necessary to set out a very brief summary of the prosecution and defence cases so that the issues on the appeal may be understood, though a summary cannot highlight all the points made on behalf of both the appellant and the prosecution.

The prosecution case was that the answer to the LIBOR question required a “genuine answer” from the submitter on behalf of each panel bank to the BBA. The prosecution asserted that a submitter would be guilty of the crime of defrauding another if they dishonestly prejudiced, or risked prejudicing, rights of others (namely those engaged in trades with the bank) by altering the LIBOR setting so that trades made on behalf of the bank would either make more money, or lose less money, knowing that they had no right to do so. The conspiracy to defraud was committed when an agreement was made to commit the crime of defrauding another. The prosecution alleged that both submitters and traders had agreed to procure or make submissions which were false and misleading because they were not genuine answers to the LIBOR question. The prosecution alleged that traders and submitters must have known that it would be unlawful and dishonest to manipulate the LIBOR setting to advantage the bank’s trades.

The evidence in relation to Mr Merchant included emails in which he had asked for the LIBOR figure to be set at a certain rate so that he would either gain more (or lose less) from his trades. For example in an email dated 27 September 2005 Mr Merchant said:

The evidence against Mr Mathew included his admissions in an interview on 30 November 2011 in which he confirmed that he would go through a process to form a judgment for the LIBOR but:

Mr Merchant’s case was that he was not involved in analysing the answer to the LIBOR question, and that he did not know that requesting the rate to be altered to advantage his trades was wrong. He was aware that many traders were making similar requests, and that he was only doing his job of making money for the bank. His case was that putting in an answer which had been adjusted to reflect the bank’s own trades was not a wrong answer if it was within the range of rates at which the bank had borrowed. Mr Merchant’s case was that he had not been dishonest, as part evidenced by the fact that he had sent his emails and communications about altering the LIBOR submissions openly on the bank’s systems which could be monitored. He accepted that counterparties had the right to have their positions calculated by an independent benchmark.

Mr Mathew’s case was that he did not know the exact relationship between Barclays and the BBA, but that he had been told by his manager, Mr Johnson, to make LIBOR submissions taking account of the traders’ requests, and he had not known that making LIBOR submissions in this way was wrong. Mr Mathew’s case was that he had not been dishonest in acting in accordance with what he had been taught. He had no reason to question Mr Johnson who was an “old school” teacher. He only realised what he was doing was wrong after his interview when he had answered the questions truthfully and deduced from Mr Johnson’s reaction that he had expected him to lie about it.

Before the commencement of the trial, the judge had given a Ruling on Indictment dated 1 March 2016 in which he decided to follow the approach taken by Cooke J in the first LIBOR trial which had involved Mr Hayes. The approach by Cooke J was upheld on a pre-trial appeal against the ruling by Cooke J by Davis LJ in R v H [2015] EWCA Crim 46. Mr Hayes was convicted by a jury and appealed against his conviction and sentence. The approach adopted by Cooke J was part implicitly and part expressly approved in the judgment of this Court in R v Hayes [2015] EWCA Crim 1944; [2016] 1 Cr App R (S) 63. The approach adopted by Cooke J was itself followed by Hamblen J in another LIBOR rate fixing trial.

In the Ruling on Indictment the judge stated, in paragraphs 43 and 44 of the ruling, that the words in paragraph 2(b) of the particulars of offence added little or anything to the words contained in paragraph 2 that the rate submitted was false and misleading. It would therefore not help the jury to direct them separately on that subparagraph.

The judge provided written directions to the jury at the beginning of the trial and further written directions (which we have been told were in the material respects the same) at the end of the trial which were consistent with his Ruling on Indictment. The jury were asked two questions. Question 1 asked the jury:

The judge also directed the jury (at paragraph 2.10 of the written directions).

The judge also gave directions on dishonesty. So far as is material question 2 asked the jury:

Under issues arising in relation to these questions on dishonesty the judge specifically told the jury to consider, among other matters, for each defendant:

In his summing up (at page 58 of the transcript dated 15 June 2016) the judge gave further directions in a series of propositions that had been used by Cooke J to explain matters to the jury in line with the legal definition of LIBOR (see paragraphs 34-37 of the decision in R v Hayes, as we set out at paragraph 31 below).

The grounds of Mr Merchant’s appeal against conviction are that the judge’s directions to the jury on the elements of the offence of conspiracy to defraud were wrong in law because the judge conflated issues of falsity in fact with the intention of the defendant when giving directions on the LIBOR question. It was submitted that the earlier judgment of this Court in R v H was wrongly decided. Other criticisms were also made of the summing up which it was said were not dependent on a finding that R v H was wrongly decided. These criticisms were that the judge: wrongly withdrew an issue of fact from the jury; did not give a proper direction on paragraph 2(b) of the particulars of offence set out in the indictment; gave a confusing direction to the jury; did not give a proper direction on what was meant by “agreement”; did not give a proper direction on the need to find a victim of the offence; and did not give a proper direction on the unlawful means required in the conspiracy. It is apparent that this appeal against conviction involved, as a principal though not exclusive ground, a comprehensive attack on the approach taken in other LIBOR trials, an approach which has been approved by this Court in R v H and R v Hayes.

In the light of the fact that it was Cooke J’s approach which had been approved and followed in subsequent trials we were provided, during the course of the hearing and thereafter, with rulings made by Cooke J dated 3 July 2014 and 5 December 2014. The parties made supplementary written submissions on the rulings made by Cooke J. The emphasis on the argument advanced on behalf of Mr Merchant was that Cooke J had not been concerned with the correct interpretation of the LIBOR question, but with the legal duty owed by the banks when answering the question. It was submitted that neither Cooke J nor this court nor at the subsequent trials had properly addressed the LIBOR question nor properly considered whether any duty was owed to the BBA when answering the question.

We should record that the rulings made by Cooke J were made in relation to the case before him. Some of the factual circumstances in that case were specific to that case; for example Cooke J recorded on 3 July 2014 that the Defendant had, in pre-trial interviews, made very extensive admissions. At paragraph 16 of the ruling dated 3 July 2014 Cooke J recorded that the issue for the jury was formulated by the defence as being “at the time of the contact with the submitters … was Mr Hayes acting dishonestly”. In a case where the existence of an agreement is in issue different directions and approaches may be required.

However, it is true to note that some issues, such as the meaning of the LIBOR question and the duty owed, are likely to be of general application in LIBOR trials arising out of the fixing of the LIBOR submissions. It is always necessary to remember that the evidence in each trial may be different, and directions appropriate for one trial might not be appropriate for another trial. It is the principled aim of every summing up to provide succinct, focussed directions on the issues raised in the specific trial. Directions which have been appropriate in one of a series of trials being heard separately for proper reasons of case management, even if those directions have been approved on appeal, might not be appropriate in a subsequent trial in the series in which there are different issues and evidence. As has been said on numerous occasions, crafting the directions for each case is essential.

In his pre-trial ruling dated 3 July 2014 Cooke J considered submissions about disclosure of returns from other banks and said:

In two further pre-trial rulings dated 5 December 2014 Cooke J considered whether loss to a victim could be shown. In the first of the rulings dated 5 December 2014 Cooke J said:

In the second of the rulings dated 5 December Cooke J expressly accepted the prosecution’s submission that:

As noted above, there was an appeal from these pre-trial rulings. In R v H Davis LJ recorded, at paragraph 42 that:

In paragraph 43 Davis LJ confirmed:

After Mr Hayes’ conviction he appealed against conviction and sentence; the judgment is reported as R v Hayes. Although the correctness of the decision in R v H was not raised in R v Hayes, this Court considered criticism of a specific direction given by Cooke J to the jury. This was Cooke J’s fifth proposition which was to the effect that if a submitter considered that there was a range of possible figures which could be submitted, if the submitter took account of the commercial interests of the bank, that would not be a genuine figure in accordance with the LIBOR definition, even if it would not have differed from the figure which would have been submitted. It is important to note that this Court said:

At the heart of the submissions made on behalf of Mr Merchant by Mr Jonathan Crow QC in seeking to persuade us that the approach in R v H was wrong, was the proposition that whether a statement is true or false is a question of fact which does not depend on the belief in which or the intention with which the statement is made. It was said that so long as the answer to the LIBOR question was within a range of permissible interest rates, the answer was not false just because the submitter had adjusted the rate to take account of requests made by traders, who were hoping for an advantage to their trading position.

The short answer to this submission is that whether belief or intention is relevant to the issue of whether a statement is true depends on what statement has been made. A false statement about a person’s belief or intention can be a false statement of fact. It has been long established that a statement about a person’s current belief or intention can be a statement of fact: see, for example, Edgington v Fitzmaurice (1885) 29 Ch D 459.

Mr Crow QC developed his submissions by reference to, among other matters: Property Alliance Group Ltd v Royal Bank of Scotland plc [2016] EWHC 3342 (Ch), a decision about implied representations being made by the Royal Bank of Scotland in relation to the LIBOR sterling rate; and a decision by the Regulatory Decisions Committee of the Financial Conduct Authority (“FCA”) dated 2 April 2015 relating to requests made by a trader to UBS’s LIBOR submitters. We do not consider that either decision, when properly analysed, assists Mr Merchant.

So far as Property Alliance Group is concerned, Asplin J was considering claims arising out of the sale of interest rate derivative products sold to the claimant by the Royal Bank of Scotland. One of the claims related to implied representations relating to LIBOR pleaded by the claimant against the bank. Reliance was placed by Mr Crow QC on the rejection by Asplin J of the implied LIBOR representations pleaded in that case. However Asplin J was clear that she was dealing with the implied LIBOR representations as pleaded in that case, and not what are standard implied terms about honesty in making representations. As Asplin J said at paragraph 405 of that case:

We consider, in agreement with Cooke J’s initial ruling and his second ruling on 5 December 2014, and the judgment of this court in R v H, that the person making the LIBOR submission was under an obligation to give their honest and genuine assessment. That the submitters will give their honest and genuine assessment is implied into the LIBOR submission; long established authority (some of which was referred to in paragraph 377-379 of Asplin J’s judgment) shows that, when an answer is given in such circumstances, it must be an honest or genuine assessment by the person making the answer.

It is clear from the transcript of the argument in R v H that the court fully considered both the meaning of the LIBOR question and the issue as to legal duty. The court clearly concluded that the operation of the LIBOR market and the answer to the LIBOR question entailed a legal duty to provide, when answering the question, an honest or genuine assessment. Indeed it is difficult to understand how the market which depended on the setting of a benchmark could have operated in any other way. As Davis LJ observed, unless when answering the question there was a legal duty to give an honest and genuine assessment the market could not operate. It followed that in making a LIBOR submission there was a legal duty to provide an honest and genuine assessment. That was the proposition which Cooke J accepted, and which was accepted without hesitation in both R v H and R v Hayes. It is hardly surprising in the circumstances that the matter was dealt with shortly by such an experienced commercial judge as Cooke J, and by the Court of Appeal in R v H.

We ourselves cannot see how a benchmark could have been set in any way other than through discharge of such an obligation when answering the question. Quite apart from the decisions in this court, it is important also to note that Hamblen J, another very experienced commercial judge, followed the same approach as Cooke J.

This is the answer to Mr Merchant’s point that the LIBOR question does not involve any question of intention or belief or honesty. We consider that it is better to use the term “genuine” when assisting the jury rather than “honest” in relation to the LIBOR submission. This is because the use of the word “honest” might be confused with the separate and distinct issue of “dishonesty” as an essential ingredient of the offence of conspiracy to defraud.

So far as the decision of the Regulatory Decisions Committee of the FCA dated 2 April 2015 is concerned, the panel considered a trader’s case who had made suggestions to the LIBOR submitters for UBS that:

We therefore reject the submission made on behalf of Mr Merchant to the effect that the LIBOR question requires only an answer of one of the rates at which the bank could borrow and no legal duty to the effect suggested was owed. In our judgement, the judges who considered this question in the earlier cases were correct and the bank was required to give a genuine assessment. They were right in concluding that this is so obvious that it was to be implied in the return. An answer which was not genuine assessment was a false answer. For these reasons R v H was rightly decided, founded as it was on well-established legal principles.

In these circumstances it was not necessary for the prosecution to prove that the actual submissions made by Barclays in answer to the LIBOR question were outside the permissible or acceptable range; what needed to be proved by the prosecution was that they were not genuine submissions. This is because a submission would be false, even if within the range, if it was either higher or lower than the bank believed a genuine answer would have yielded. As the evidence demonstrated, very small movements, within the permissible range, were capable of increasing profitability for the bank and reducing profits or increasing losses to the counter-parties.

There is therefore no merit in this ground of appeal.

We turn to the ground of appeal relating to the direction on paragraph 2(b) of the particulars of offence. This ground did not depend on showing that the decision in R v H was wrong, but it raised an issue as to the way the judge had crafted his directions for the circumstances of this trial by, in effect, taking the view that one of the particulars in the short particulars was otiose. The wording of the indictment appears to involve some surplus words; for example a submission which is “false and misleading” is likely to “deliberately disregard the proper basis for the submission”.

However the particulars of this offence had been drafted by asserting that to defraud, and commit a crime, the submission which was made or procured needed dishonestly to prejudice, or risk prejudicing, rights of others (namely those engaged in trades with the bank) by altering the LIBOR setting so that trades made on behalf of the bank would either make more money, or lose less money, deliberately disregarding the proper basis for the submission, effectively stating that it was known that there was no right to act in that way. We have no doubt that in each of the LIBOR cases detailed consideration would have been given to the importance of drafting the indictment carefully and crafting the particulars specifically (as has been pointed out by this court on many occasions: see for example R v K [2005] 1 Cr App R 25 at paragraph 36).

Cooke J dealt with the issue of deliberately disregarding the proper basis as a general issue for the jury under the issue of dishonesty. This may have been because, as noted above, the central issue before Cooke J had been identified as one of dishonesty. We accept that the issue can be addressed under dishonesty and because of the focus of the issues in that case, it was right to do so. However in the present case, properly analysed, deliberately disregarding the proper basis was, as drafted in the particulars of the offence of this indictment, part of the element of the particulars of this offence of defrauding.

The judge considered that the matter was best dealt with under the issue of “dishonesty”; he considered that the issue of dishonesty was capable of doing what he described as “the heavy-lifting” so far as the issues in this case were concerned. It therefore, in the judge’s view, encompassed the particulars within paragraph 2(b) without the need for a specific direction.

It is apparent from their verdict that the jury must have answered both the written questions 2(a) and (b) against Mr Merchant and decided that the agreement involved deliberately disregarding the proper basis for the LIBOR submission. They therefore must have been sure that Mr Merchant appreciated that what he agreed to do was dishonest by the standard of ordinary honest and reasonable persons. The jury had been specifically directed to consider in this respect the belief that Mr Merchant had that commercial influence was consistent with the way that the LIBOR question was being answered by the bank and other panel banks.

In these circumstances and on our careful view of the specific facts of this case, we are sure that the verdict against Mr Merchant is safe, notwithstanding the clear omission by the judge of a direction directed at paragraph 2(b) of the particulars of the indictment. This is because the jury must have been sure that Mr Merchant must have agreed to deliberately disregard the proper basis for the submission of rates, because otherwise he could not have been found to have been dishonest and found guilty by the jury.

After distribution of the judgment in draft to the parties, further submissions on this ground of appeal were made by Mr Merchant. However those submissions did not deal with the particular facts of Mr Merchant's case. The jury answered question 1 of the directions (see paragraph 17 above) against Mr Merchant and must have been sure that Mr Merchant agreed with another Barclays employee to procure a submission by the Bank which was not the bank's genuine perception of its borrowing rate. Mr Merchant himself accepted that counterparties to the Bank had the right to have their positions calculated by an independent benchmark (see paragraph 13 above). As noted in paragraph 48 above the jury had answered questions 2(a) and (b) on dishonesty against Mr Merchant having been asked to consider the belief that Mr Merchant had about the LIBOR question. The suggestion made in the further submissions that the jury would not have applied their mind to the written directions provided by the Judge is not sustainable. The joint effect of these answers is that the jury must have been sure that Mr Merchant deliberately disregarded the proper basis for submission of the rates. The safety of the conviction is a matter for the Court, and we are sure that the conviction is safe. The omission by the judge had in the particular circumstances of this case no material effect; it certainly does not in any way affect the safety of the conviction for the reasons we have given.

The other grounds of appeal against advanced by Mr Crow QC are also not dependent on showing that the decision in R v H was wrongly decided. The direction by the judge that:

We do not consider that the other grounds of appeal are arguable. The directions given by the judge were not confusing; the direction on agreement for the purposes of the conspiracy was conventional. There was no requirement to give directions about loss, victims and unlawful means because if the written questions 1 and 2 were answered against Mr Merchant, as they must have been, an unlawful conspiracy to defraud had been proved.

We therefore refuse leave to appeal on these other grounds.

Although we have granted leave to appeal on the two grounds we have considered at some length, we dismiss the appeal. We are satisfied that the conviction is safe.

The grounds of appeal relate to the admission of the conviction of Mr Johnson, a co-accused and Mr Mathew’s line manager at material times, and the judge’s directions in relation to that conviction. As we have mentioned Mr Johnson had been convicted on his own plea of guilty to the indictment.

It was common ground that Mr Johnson’s conviction following his plea was admissible pursuant to s.74 of the Police and Criminal Evidence Act 1984 (“PACE”). This was because the conviction was admissible to prove that there was a conspiracy to defraud to procure or make false LIBOR rates. However it was submitted on behalf of Mr Mathew and others at trial that the plea should not be admitted because it would have “such an adverse effect on the fairness of the proceedings” pursuant to the provisions of s.78 of PACE. Mr Mathew also submitted that a direction should have been given, pursuant to s.74(2) of PACE, that it could be proved that Mr Johnson was wrongly convicted. This second submission has been termed, by way of shorthand, as the “reverse burden” submission. Both submissions are renewed on appeal.

The trial judge had been invited by the prosecution to rule that Mr Johnson’s conviction was admissible in pre-trial hearings. By a ruling dated 1 May 2015 the judge recorded that it was common ground that the conviction was admissible pursuant to s.74, but deferred the determination of the s.78 issue until he had a fuller understanding of the issues at trial. This approach was in accordance with guidance from previous authorities, and it was a proper approach to take in this case.

The issue of Mr Johnson’s conviction was revisited during the trial. On 10 May 2016, at a time when the judge had a fuller understanding of the issues raised at the trial, he gave a ruling admitting the conviction.

The jury were directed about the plea of Mr Johnson on pages 13-14 of day 1 of the summing up on 15 June 2016. The jury were reminded of the plea and told that:

The judge reminded the jury that the plea did not decide the issue as to whether any other defendant was guilty, but it might assist them in deciding whether there was an agreement between Mr Johnson and one or more employees of Barclays to procure or make submissions of a LIBOR rate, which was not the bank’s genuine perception of its borrowing rate.

It was submitted that the conviction proved too much in the case of Mr Mathew, and should have been excluded. It was said that the conviction proved too much in this case because it not only proved the existence of a conspiracy to defraud in respect of fixing the LIBOR rate for the US dollar, but it also proved that Mr Mathew must have been a party to the conspiracy. It was said that this was because at material times Mr Mathew reported to Mr Johnson and was managed by Mr Johnson, and if Mr Johnson was conspiring with anyone, it must have been Mr Mathew.

We accept that admitting a conviction which proves too much is, in some circumstances, capable of having such an adverse effect on the fairness of proceedings such that it ought to be excluded, see R v Smith [2007] EWCA Crim 2105; (2007) 151 SJLB 1260 at page 5. That is because if a conviction proves too much, it may, depending on the circumstances of the case, have the effect of preventing a defendant from having the opportunity to challenge evidence against him in the process of a criminal trial with its procedures designed to ensure fairness.

However in our judgement this was not remotely such a case. The judge was entitled to admit the plea of Mr Johnson. In the context of this trial and Mr Mathew’s case, it did not prove too much. It assisted in the proof only of the existence of an agreement between Mr Johnson and other persons. As appears from the indictment this was charged as an open conspiracy meaning that named defendants were alleged to have conspired together and with other employees of Barclays Plc and its associated entities. This meant that Mr Johnson could have conspired with co-accused other than Mr Mathew, or other employees of Barclays plc and its associated entities.

We note that there was nothing in the plea to show that Mr Johnson was pleading guilty on the basis that he had conspired with Mr Mathew. Further it is apparent that the jury would not have interpreted the conviction of Mr Johnson to prove that Mr Johnson was pleading guilty on that basis, given the directions from the judge. The admission of this conviction did not prevent Mr Mathew from challenging the prosecution case against him in the normal way.

So far as is material section 74(2) of PACE provides:

It does not appear that any submission was advanced to the judge to the effect that he should give a reverse burden direction. We do not consider that surprising. That is because each defendant was able to advance their case that, although Mr Johnson was guilty, the relevant defendant had neither conspired with Mr Johnson nor had been dishonest. Even if a reverse burden direction should have been given, and we do not find that it should have been, the failure to give such a direction could not have caused any conviction to be unsafe. This is because it was apparent that Mr Johnson pleaded guilty in the light of overwhelming evidence against him. It was apparent that he had conspired with at least someone to put in LIBOR submissions which were not the bank’s genuine perceptions of its borrowing rate to advantage the trading position of an employee or employees of Barclays.

In our view none of these points was arguable. We refuse leave to appeal.

Mr Merchant renews his application for permission to appeal against sentence. It is submitted that the judge wrongly categorised the harm from the applicant’s activities, failed to have regard to mitigating factors, and ignored comparative sentences.

For the purposes of the “Definitive Guideline: Fraud, Bribery and Money Laundering Offences”, it was common ground that Mr Merchant’s culpability was high. So far as harm was concerned calculations submitted on behalf of Mr Merchant, which were not contradicted by other evidence, suggested that the direct loss to counter-parties of Barclays caused by Mr Merchant’s activities did not exceed $553,000 (just under £300,000 on the then applicable exchange rates). However the judge held that “the actual loss or risk of loss does not realistically identify the harm caused by the criminal conduct in this case … I have concluded that the risk of loss can be judged to run into millions of pounds”. This was, in the judge’s view, because whether profits were increased, or losses decreased, depended in part on the actions of other banks which was outside the control of Mr Merchant and the others.

The judge had heard the trial, and was best placed to make an assessment of the harm, and risk of harm. We agree that the effect of the conspiracy to defraud in this case was not limited to the effect on direct counter-parties with the bank. Altering LIBOR by returns which were not genuine affected anyone whose own trades or liabilities were governed by, in this case, a US$ LIBOR rate. Professor Anderson, who gave expert evidence to the jury, recorded a difference of 4 points over a year could benefit lending banks by US$500 million over a year. The submissions on harm made by Mr Merchant have overlooked the effect on the market as a whole, and the loss of confidence in the LIBOR setting.

The judge had proper regard to the mitigating factors, and there were many, in Mr Merchant’s case. However there is available to us further information about specific difficulties which have affected Mr Merchant and his family since he has been sentenced. This has included the birth of a child, albeit anticipated at the time of sentencing, and the illness suffered by Mrs Merchant after the birth of the child, as reported by a clinical psychologist. It is apparent that had this information been available to the judge these would have been reflected as further mitigating factors showing that the effect of imprisonment on Mr Merchant and his family is more severe than might have been expected. We consider that it is right to reflect these further factors and reduce the (otherwise appropriate) sentence imposed by the judge by a year.

To this extent the appeal against sentence succeeds. We will impose a sentence of 5½ years’ imprisonment in place of the 6½ years’ imprisonment imposed by the judge.