The LIBOR Scandal: How Traders Rigged the World's Most Important Interest Rate (2008–2012)

A Number Built on Trust

For nearly three decades, the most important number in global finance was not calculated — it was polled. Shortly before 11 a.m. London time each business day, a handful of clerks at sixteen of the world's largest banks would glance at their screens, consult their traders, and send a single value to Thomson Reuters answering a deceptively simple question: at what rate, in reasonable market size, could the bank borrow unsecured funds from other banks, for each of ten currencies and fifteen maturities? The answers were sorted, the top and bottom quartiles discarded, and the remaining eight values averaged. Out popped the London Interbank Offered Rate.

LIBOR, in its finished form, had the crisp authority of a market price. It appeared on Bloomberg terminals and in loan documents. It set the coupon on adjustable-rate mortgages in Florida, the interest on student debt in Seoul, the funding cost for Australian corporates, and the settlement price of swaps contracts with notional values measured in hundreds of trillions. Industry estimates placed total LIBOR-referenced exposure at roughly $350 trillion by 2012 — a number that flattered nothing, since most of it was derivatives — but its reach into real-economy lending was itself staggering. Roughly half of all US adjustable-rate mortgages, and the majority of syndicated corporate loans globally, referenced some tenor of LIBOR.

What almost no one outside a small circle of regulators appreciated was how the number was produced. The submissions were not based on actual transactions. They were judgements. After 2007, they were judgements made in a market where interbank lending was seizing up and real trades at advertised rates had effectively stopped occurring. The benchmark that underpinned the plumbing of global finance was, in its final years, a daily survey of opinion about a market that had partially ceased to exist.

The Structure of a Vulnerability

LIBOR's origins were modest. Developed in the mid-1980s by the British Bankers' Association to service the nascent syndicated loan and interest rate swap markets, it solved a coordination problem. Banks and borrowers needed a common reference rate; polling the panel was an elegant compromise between market realism and operational simplicity. For twenty years it worked well enough to become ubiquitous.

The design, however, had two latent flaws that compounded as the stakes grew. First, the submitters were the same institutions whose derivative books were priced off the benchmark. A bank that had sold hundreds of billions in notional interest rate swaps had a direct financial interest in where LIBOR fixed on any given day. A move of a single basis point — one one-hundredth of a percentage point — could translate into hundreds of thousands of dollars of daily P&L on a large book, and far more on reset dates. The submitter and the beneficiary sat in the same building, often on the same floor.

Second, submissions were public. Each day's contributions were published with the contributing bank's name attached. During ordinary times this felt like a transparency feature. During the 2008 crisis it became a market signal. A bank submitting a high rate was advertising that it faced expensive funding — possibly because other banks were reluctant to lend to it. A bank in distress therefore had a strong reputational motive to submit a lower rate than its actual cost of funds.

Means, motive, opportunity. Each sat quietly inside the architecture of the benchmark.

Two Scandals in One

The investigations that unfolded between 2008 and 2015 ultimately uncovered two distinct, overlapping types of manipulation. Conflating them has muddied much of the public memory of the scandal, but the regulatory record is clear that they operated through different mechanisms and served different ends.

The first was trader-driven manipulation. Derivatives traders in yen, dollar, and Swiss franc desks across major banks routinely lobbied their own firms' rate submitters for a tick higher or lower on specific maturities, depending on their positions that day. Over time, networks of colluding traders at different banks coordinated their requests — a sort of distributed market-rigging in which each individual nudge was tiny but the cumulative effect on reported rates was consistent and profitable. Brokers in the interdealer market facilitated the collusion, passing messages between desks and, in some cases, fabricating "run-throughs" — quoted levels circulated to the panel to influence submissions.

The second was reputational, or so-called "lowballing." From August 2007 onward, and most acutely in the weeks around the collapse of Lehman Brothers, interbank funding markets seized. Banks that were still able to borrow paid steep premiums; banks that were not, could not borrow at any rate. In that environment, several panel members — Barclays most famously, but not only Barclays — submitted LIBOR rates that were below what their own treasury desks knew they would have to pay in the market. The logic was defensive: appearing stressed invited a run. Lowballing was a way of hiding distress from counterparties during the 2008 financial crisis.

Trader manipulation was about money. Lowballing was about survival. Both subverted a benchmark that half the world was using in good faith.

The Chatroom Record

What gave the LIBOR scandal its particular cultural charge was not the statistical anomalies that first attracted academic suspicion — although those were substantial — but the chatroom transcripts that regulators released with their enforcement orders. For the traders involved, Bloomberg and Reuters chat windows had replaced the telephone as the medium of professional life. Their messages, preserved by compliance systems, were written in the confidence that they were ephemeral. They were not.

A 2007 Barclays derivatives trader, asking his submitter for a low three-month fixing, sent: "Dude. I owe you big time! Come over one day after work and I'm opening a bottle of Bollinger" (FSA, 2012). A broker working with Tom Hayes referred to him, in a chat to another trader, as "the cartel." A UBS trader, discussing a successful attempt to move yen LIBOR, wrote that the submitter was "a top man." These fragments were released in dry regulatory prose, but they read like dialogue from a heist film.

The transcripts had a second, more technical consequence. They demonstrated that the manipulation was neither ambiguous nor occasional. It was a routine feature of how these desks operated, conducted with the casualness of professionals who had no reason to think the behaviour was problematic. That made it far harder for the banks involved to argue that a few bad apples were responsible, and it gave prosecutors the corroborating evidence needed to pursue individual criminal charges — a rare outcome in financial regulation, and one that had conspicuously not occurred after earlier scandals such as Enron or Madoff.

The Wall Street Journal and the Slow Awakening

Academic researchers began noticing strange patterns in LIBOR submissions during 2008. The dispersion between panel banks' submissions had collapsed to a degree inconsistent with the visible stress in credit default swap spreads and unsecured funding markets. In May 2008, The Wall Street Journal published an analysis suggesting that LIBOR submissions were running roughly 0.3 percentage points below what the banks' credit default swap levels implied their funding costs should have been (Mollenkamp and Whitehouse, 2008). The implication, politely worded, was that something was off.

The response from the British Bankers' Association, which administered LIBOR, was dismissive. Regulators in the US and UK began quietly asking questions, but the public focus in 2008 was on keeping the financial system from disintegrating, not on investigating the integrity of a benchmark that, for all its flaws, was at least continuing to print. The deeper investigations that would eventually produce the landmark enforcement actions did not accelerate until 2010 and 2011.

The Barclays Settlement

The scandal broke commercially on 27 June 2012, when Barclays announced a combined settlement with the US Commodity Futures Trading Commission, the US Department of Justice, and the UK Financial Services Authority totalling £290 million, or roughly $450 million at prevailing rates. The settlement documents quoted extensively from internal chats and emails. The bank admitted that both trader manipulation and lowballing had taken place across multiple years and multiple currencies. Barclays had also, by cooperating early, secured the smallest penalty among the major defendants — it was the first bank through the door, and it paid the lowest price for the distinction.

The political response in the United Kingdom was swift and unsparing. Barclays' chief executive Bob Diamond, an American banker who had come to embody the post-crisis hostility toward big bank pay and conduct, was forced to resign within a week. He appeared before the Treasury Select Committee of Parliament on 4 July 2012 in a session that made transatlantic news, parrying MPs with a mixture of contrition and defensiveness that played poorly. Chairman Marcus Agius also stepped down. Governor of the Bank of England Mervyn King publicly endorsed Diamond's departure — a rare intervention into the affairs of a commercial bank.

Total industry fines for LIBOR-related misconduct ultimately exceeded $9 billion, not counting private litigation settlements that continued to accumulate for years afterward.

Tom Hayes and the Criminal Prosecutions

Regulators could extract enormous sums from banks, but individual accountability was slower. The first criminal conviction came in August 2015, when Tom Hayes — a UBS and later Citigroup yen derivatives trader, a talented quant with a mild manner and an obsessive work ethic — was found guilty at Southwark Crown Court of eight counts of conspiracy to defraud. The jury's verdict was unanimous. The sentence was 14 years, later reduced to 11 on appeal.

Hayes's case was extraordinary for several reasons. He was the first person convicted anywhere in the world for rigging LIBOR. His defence did not dispute the facts — the chats were unambiguous — but argued that what he had done was standard practice, taught to him and encouraged by supervisors, and that he had believed it to be legitimate. The court rejected that defence, ruling that dishonesty by the standards of ordinary reasonable people was the correct test, regardless of industry norms.

In an unusual coda, Hayes's conviction was finally overturned by the UK Supreme Court in July 2025 on the grounds that the trial judge's directions to the jury had been incorrect — a legal, not a factual, reversal. Several of his co-conspirators had already completed their sentences. Whatever else the Hayes case established, it set a precedent for criminal liability in benchmark manipulation that banks and regulators have had to live with ever since.

The Wheatley Review and Regulatory Rebuild

In July 2012, the UK government commissioned a review of LIBOR under Martin Wheatley, the incoming chief executive of the Financial Services Authority's successor, the Financial Conduct Authority. The Wheatley Review, published in September, made a series of recommendations that were implemented rapidly: administration of LIBOR was transferred from the British Bankers' Association to an independent body (eventually ICE Benchmark Administration), submissions were required to be anchored in actual transactions wherever possible, the number of currencies and maturities published was reduced to eliminate those with the thinnest underlying markets, and manipulation of financial benchmarks was made a specific criminal offence under UK law (Wheatley, 2012).

The deeper problem, however, could not be solved within the existing framework. If the underlying unsecured interbank market had shrunk to the point where daily transactions were too few to support a benchmark, no amount of governance reform would restore integrity. The solution had to be a different benchmark altogether.

The Transition to SOFR, SONIA, and the End of LIBOR

In 2014 the Federal Reserve convened the Alternative Reference Rates Committee, a group of major market participants tasked with identifying a replacement for USD LIBOR. After extensive consultation, the ARRC selected the Secured Overnight Financing Rate — SOFR — calculated from the volume-weighted median of transactions in the Treasury repo market. The key distinction was that SOFR was transaction-based, derived from approximately $1 trillion of daily activity that could not be manipulated by a handful of submitters. The United Kingdom developed SONIA (Sterling Overnight Index Average) along similar lines. The euro area produced €STR. Switzerland adopted SARON. Japan adopted TONA.

The transition was enormous. Existing contracts referencing LIBOR had to be remediated, new contracts had to use the replacement rates, and the mathematical differences between term benchmarks and overnight compounding methodologies had to be bridged. Regulators set firm deadlines. In March 2021, the FCA announced that most LIBOR tenors would cease publication on 31 December 2021, with remaining USD tenors continuing on a limited, synthetic basis until 30 September 2024. On 30 June 2023, the main USD LIBOR settings ceased to be representative, and the architecture that had defined wholesale finance for four decades effectively ended.

Legacy: What the Scandal Revealed

The LIBOR scandal was not a story about a few bad traders. Traders such as Tom Hayes and his counterparts at other banks were, unquestionably, the immediate actors. But they operated within structures their employers had built and their regulators had tolerated. The benchmark had always depended on the honour of its submitters; the problem was that the submitters, over time, became people whose books were priced off the benchmark and whose employers sometimes needed it to lie on their behalf. Once the stakes grew large enough, the trust assumption collapsed.

The deeper significance of LIBOR runs in parallel to other great failures of financial self-regulation. The collapse of Barings in 1995 revealed how a single desk could overwhelm weak internal controls. Enron showed how accounting conventions could be corrupted. The 2008 crisis revealed how securitisation assumptions could be gamed. LIBOR revealed something distinct and arguably more damaging — that an entire industry's supposedly arm's-length benchmark could be quietly bent, day after day, by the same firms that used it as a reference. The fix was not better rules governing the old benchmark. The fix was to abolish the old benchmark and replace it with one that could not be judged into existence.

There is a narrow and a wide reading of what that replacement accomplished. The narrow reading: the plumbing now works more honestly than it did. SOFR cannot be rigged by a chatroom. The wide reading is harder. Every financial system rests somewhere on human judgement, and wherever judgement meets large positions, the LIBOR dynamic waits to re-emerge. The scandal's lesson is not that benchmarks have been solved. It is that the industry learned, belatedly and expensively, that when a number matters enough, it must be anchored in transactions rather than declarations — and that any corner of finance still running on a handshake is a corner where the handshake will eventually be monetised.

A benchmark built on trust died on 30 September 2024. Whatever replaces any other such benchmark will be watched more carefully, for a while, by people who remember what Bollinger used to be worth on a Barclays trading floor.