In the years before the scandal had a name, London’s money markets looked orderly from a distance and improvisational up close. LIBOR — the London Interbank Offered Rate — was supposed to capture the price at which major banks could borrow from one another, a daily snapshot built from submissions rather than completed trades. That design, according to regulators and later court records, left the benchmark vulnerable: it depended on judgment, convention, and an honor system that assumed traders and submitters would not treat a global reference rate as one more instrument to be gamed.
That vulnerability mattered because LIBOR was never a niche yardstick. It sat at the center of the plumbing of modern finance, feeding mortgages, corporate loans, floating-rate notes, and swaps across continents. By the time its weaknesses were exposed, products tied to the benchmark had been estimated at roughly $350 trillion. That was the scale of the prize, and the scale of the damage. A movement that looked trivial on a chart could produce real gains or losses when multiplied through derivatives books so large that a tiny adjustment could ripple into millions of dollars.
The environment in which LIBOR operated helped hide the problem. In the early 2000s, the market was opaque, the mechanics technical, and the people inside the system often knew more about the benchmark’s weaknesses than the public or many regulators did. The daily submission process, administered by the British Bankers’ Association until reforms came later, relied on banks providing estimates of their borrowing costs. Those estimates were not necessarily drawn from completed trades. They could be shaped by judgment, market conditions, and internal incentives. The result was a benchmark that looked objective but could, in practice, be influenced from within.
The first signs of abuse were easy to miss because they did not look like a grand conspiracy. There was no single switch being flipped. Instead, the structure itself invited pressure. Traders had incentives to move the number in directions that benefited their positions. Submitters, often inside the same institutions, were not insulated from trading desks. The separation between the people making the submissions and the people profiting from the rate was real on paper, but porous in practice, according to later enforcement actions. As regulators would later show, that porousness allowed requests for “high” or “low” submissions to circulate through chat rooms, emails, and calls.
Tom Hayes entered that world as a mathematically gifted trader, not a policy architect. Born in 1980 in the United Kingdom, he studied economics at Cambridge and moved into derivatives trading, eventually working on yen interest-rate products for major banks. The record that would later surround him portrayed a trader with a relentless focus on basis points, positioning, and P&L. In that culture, status came from speed, nerve, and results. The benchmark was not a sacred public utility; it was an input, and inputs could be optimized. Hayes was not alone in thinking this way, but he would become one of the most visible figures in the scandal because his trading sat directly on top of the benchmark’s weaknesses.
What made the setup so durable was its normality. In a dealing room, requests for a favorable submission could be folded into routine market chatter. A trader with exposure to yen derivatives could see a small movement in LIBOR translate into meaningful gains or losses. If a submitter could be nudged without immediate consequence, the benchmark became another tradable variable. The abuse began, in effect, as an operational shortcut. Once the shortcut worked, it acquired its own logic. The number no longer felt fixed; it felt negotiable.
At Barclays, one of the first major public flashpoints in the scandal, investigators later described a communications chain linking derivatives traders and submitters. Requests for advantageous rates were passed through chat rooms, emails, and calls. Deutsche Bank would later be pulled into the same web. But before the names became public, the system still functioned as a machine of plausible deniability. No single message had to carry the whole fraud. No single request had to look criminal. The conduct was only visible when the records were stacked together and read as a pattern.
That pattern mattered because the benchmark’s influence extended far beyond any one desk in London. A few thousandths of a percentage point, applied across enormous notional exposures, could alter the economics of trades and contracts throughout the financial system. The scale of the benchmark meant the abuse was not merely unethical in abstraction. It had the potential to shift money among banks, counterparties, and ultimately consumers and investors whose contracts referenced LIBOR without their knowledge. The hidden stakes were immense precisely because the number was trusted.
The mechanics of the scheme were not mysterious once investigators began to look. Requests moved between trader and submitter. Submissions were adjusted. Trading books improved. Because LIBOR was still widely seen as a neutral market fact rather than a manipulable input, the early benefit appeared in the ordinary language of profit and loss, not in the language of wrongdoing. The effect was incremental, but the accumulation was corrosive. Every successful nudge reinforced the idea that the system could be influenced, and every unchallenged adjustment lowered the barrier to the next one.
What could have exposed it sooner was the very thing the benchmark lacked: independent verification of the inputs against actual transactions. The market was structured around trust, and that trust was a weakness. Regulators did not have the same day-to-day visibility as the banks that generated the submissions. The BBA’s private governance offered limited external scrutiny. In a market built on discretion, there were few obvious alarms. The scheme survived not because it was invisible in the abstract, but because it was disguised as routine behavior inside institutions that had every reason not to ask too many questions.
By the time the practice matured, it had a rhythm. A trader would want a lower or higher submission depending on position. A submitter would accommodate, or at least consider accommodating, the request. The benchmark would move by an amount that looked insignificant. The trading book would benefit. And because the process was embedded in ordinary market operations, the first money flowing in did not arrive with sirens or headlines. It arrived as cleaner P&L, one trade at a time.
The scandal’s eventual unraveling would show how much had been hidden in plain sight. But in these early years, the decisive fact was simpler: a global reference rate built on trust had been placed in the hands of institutions that profited from influencing it. The question was no longer whether the system had a weakness. The question was how many people had learned to use it before anyone outside the dealing rooms understood what was happening.