Rabobank LIBOR Traders: Why the Convictions Were Reversed

The Second Circuit reversed the Rabobank LIBOR convictions because the prosecution used evidence from a cooperating witness who had read the defendants’ compelled testimony. The court also dismissed the indictment against the bank’s former global head of liquidity and finance and its former primary U.S. dollar LIBOR submitter. For someone who knows a trading desk, the case connects familiar records with an unusually consequential question: what did the witness know from working there, and what did he learn from reading someone else’s account?

The Justice Department’s criminal case alleged that the former employees accommodated requests for benchmark submissions that favored trading positions. A jury convicted them on November 5, 2015, but the Second Circuit’s July 19, 2017 decision overturned those convictions. The reversal concerned how the government obtained and used its evidence.

How LIBOR Turned a Desk Request Into a Market Number

LIBOR was a daily reference rate intended to reflect what banks would pay to borrow from other banks. During the period covered by the criminal case, the British Bankers’ Association administered it using submissions from panels of banks, including Rabobank. The published rates fed into swaps, futures, mortgages and student loans, giving a borrowing-cost estimate consequences far beyond the submitting desk.

Each of the U.S. dollar and yen panels had 16 banks. Around 11 a.m. London time, submitters supplied estimates for 15 loan durations, known as tenors, ranging from overnight to one year. For each currency and tenor, the calculation discarded the highest four and lowest four submissions and averaged the remaining eight. A request for a particular tenor targeted one input into that calculation.

An interest-rate swap made the financial incentive concrete. One party paid a fixed rate and the other a floating rate tied to LIBOR, with the fixing date determining the floating payment. Banks supplying benchmark estimates also held contracts priced against the resulting benchmark. In the prosecution’s account, favorable movements benefited Rabobank and its traders at the expense of the businesses on the other side of those contracts.

The World Federation of Exchanges’ 2024 working paper, An Analysis of Market Manipulation Definitions Around the World, classifies benchmark rigging as manipulation through false or misleading submissions. It also explains why submitters might coordinate when any individual contribution has little influence. In this form of market manipulation, the submission serves the trading position that benefits from it. The published rate alone leaves the desk’s reason for wanting that number out of view.

The Requests Behind Rabobank’s LIBOR Submissions

The criminal prosecution centered on a manager and a submitter with different responsibilities. The global head of liquidity and finance supervised cash traders and had previously handled dollar submissions himself. Another cash trader took primary responsibility for dollar LIBOR in 2005. The yen submitter later became the cooperating witness whose evidence brought down the prosecution.

According to the appellate opinion’s account of the trial, derivatives traders requested higher or lower submissions to suit their positions. A London trader testified that he made requests aloud at the desks. Traders in other offices more often sent written messages, leaving exchanges that identified the currency, tenor and direction they wanted.

The opinion reproduces an October 31, 2006 message from a New York derivatives trader to the dollar submitter. After discussing market activity, the trader added:

The request meant a low two-month rate and a high three-month rate. The submitter replied that he would do it. We see the absurdity in the message itself: an estimate of the bank’s borrowing cost was being requested as a favor for another desk.

Another exchange connected a requested number with the bank’s submission. In August 2007, a derivatives trader asked for a high six-month dollar fixing of 5.42. The next day, the global head told another trader that he thought the requested rate was what the derivatives trader needed, and Rabobank submitted that number. The government presented the exchange at the criminal trial as evidence that requests had been accommodated.

The defendants disputed that they actually adjusted submissions to serve traders’ requests. Both sides agreed that the submitter’s job was to give an honest borrowing-cost estimate using market information, including brokers’ reports. The central dispute was whether requests became adjustments, and the yen submitter supplied crucial testimony on that point.

The opinion quotes the yen submitter’s criminal-trial testimony about how he accommodated requests himself. Brokers supplied a range of possible rates, and he said he would normally choose near the middle. When a trader wanted a higher rate, he moved toward the top of that range. The alleged manipulation could hide inside the estimate: the question was whose interests drove the choice.

Why Were the Rabobank LIBOR Convictions Reversed?

The convictions were reversed because the government failed to show that its cooperating witness’s evidence was independent of answers British regulators had required the defendants to give. The Second Circuit ruled that prosecutors could not use those compelled answers against them in an American criminal case, including through another witness whose account they had shaped.

The U.K. Financial Conduct Authority gave the yen submitter the other employees’ interview transcripts as evidence in its own enforcement action against him. He read and annotated them before becoming a cooperating witness for the Justice Department. The court compared his earlier account with what he later told American prosecutors and found material changes.

Those changes reached the desk’s alleged conduct. The witness learned about conversations he had not participated in from the transcripts, then discussed them with American prosecutors. He was also the only witness the government interviewed who said the global head directed submitters to account for trading positions. His general assurances that the transcripts had not influenced him could not resolve that evidence problem.

The same witness’s account had helped secure the charges, so the problem reached back beyond the trial. Prosecutors already held the written records but had not charged these defendants on that basis. The court could not conclude that the charges would have been brought without the coworker’s evidence, and it dismissed the indictment as well as reversing the convictions.

The Bank’s CFTC Resolution Was a Separate Action

The Commodity Futures Trading Commission ordered Rabobank to pay a $475 million civil penalty in its October 29, 2013 consent order. That civil enforcement action against the bank covered alleged manipulation involving dollar, yen and sterling LIBOR and Euribor. Its scope was broader than the criminal prosecution of the former global head and dollar submitter.

The CFTC alleged that Rabobank’s submitters favored trading positions and that the bank failed to provide adequate submission controls and supervision. The order described a conflict from assigning traders with benchmark-linked positions to submit rates, and alleged that seating them alongside other derivatives traders made requests easy to coordinate. Those are recognizable features of a desk’s organization, not just problems visible in a published rate. The FBI’s Washington Field Office investigated the criminal case following a referral from the CFTC’s Division of Enforcement.

What a Benchmark Insider’s Records Can Show

A benchmark insider may understand the connection between a trader’s position, a request for a particular tenor and the rate ultimately submitted. The Rabobank opinion preserves those connections through messages, submission figures and testimony. It also shows why the person’s own experience matters alongside the documents: who made a request, who heard it, and who decided what number to send.

We work with the trading logic as well as the records. A broker’s market indication, a desk message and a submission for the same fixing can explain different parts of a decision. Your knowledge of how the desk actually worked can give those records meaning, and our role is to develop that account with you and bring in CFTC benchmark-manipulation whistleblower lawyers who understand the transactions.

The CFTC whistleblower program provides an aggregate award range of 10% to 30% of qualifying monetary sanctions collected. A whistleblower case under the False Claims Act can pay 15% to 30% of what the government recovers. Those are different programs, and we work through which fits the conduct you know.

You Know Why the Desk Wanted That Rate

If you saw benchmark requests follow the trading book, tell our whistleblower experts what happened at your desk. We connect your market knowledge with the submissions, messages and positions, build your case alongside you, and find whistleblower lawyers experienced in CFTC benchmark cases. You bring the view from inside the business; we bring the expertise and commitment to pursue it with you.