Barings Bank and Nick Leeson

In 1995 Barings plc was Britain's oldest merchant bank, founded in 1762 and once banker to the British monarchy. On 26 February 1995 the Bank of England placed it into administration. No market crash or bad loan book sat behind that collapse. One trader in Singapore, Nick Leeson, spent more than two and a half years hiding losses on futures and options positions until they grew large enough to destroy the bank before anyone in London understood what was happening.

An arbitrage mandate becomes a hidden bet

Leeson's job, on paper, was arbitrage. Barings Futures (Singapore) traded Nikkei 225 index futures on the Singapore International Monetary Exchange (SIMEX) and, in principle, bought the contract on whichever of SIMEX or the Osaka Securities Exchange was cheaper while selling the other, pocketing a small, low-risk difference. That is a textbook hedged position, with every trade offset by an equal and opposite leg. Leeson built something else entirely. From July 1992 he used an internal "error account" numbered 88888, meant to net off minor booking mistakes, to hide losses from trades that were not hedged at all. He was general manager of the Singapore office and also controlled its back-office settlements function, the unit meant to check a trader's own numbers. That dual role is the whole story in miniature: nobody independent was checking what he reported.

Selling calm to fund the losses

By the end of 1993 the losses hidden in account 88888 already exceeded £20 million. By the end of 1994 they exceeded £200 million. From January 1994 Leeson began writing large volumes of Nikkei 225 options, mostly short straddles: selling a call and a put at the same strike price and collecting the combined premium up front. A short straddle pays off only if the market stays close to the strike until expiry. Leeson used the premium income to help cover the daily margin calls that SIMEX, correctly, kept issuing against his mounting futures losses. For a while the strategy worked, because the Nikkei was calm.

Kobe strikes both sides of the book at once

On 17 January 1995 the Great Hanshin earthquake struck Kobe in western Japan. The Nikkei 225 fell sharply in the days that followed, and this was bad news on both sides of Leeson's book at once: his long futures position needed the index to hold steady or rise, and his short straddles needed the index to stay near the strike. Instead it moved against both. Leeson doubled down, buying more futures in an attempt to prop the index up and recover the losses, a strategy that depended on a quick market turnaround. None came. By late February his positions were so large relative to the market that unwinding them without moving the price further against himself was close to impossible.

Collapse, sale, and sentence

Leeson left Singapore on 23 February. Three days later the Bank of England concluded Barings could not meet its obligations and placed the group into administration. The cumulative loss from the unauthorised trading stood at £827 million. ING Group, the Dutch bancassurer, bought Barings for a nominal £1 on 6 March 1995, taking on its liabilities and protecting depositors and staff, though not the shareholders. Leeson was arrested, extradited from Germany, and in December 1995 a Singapore court sentenced him to six and a half years in prison after he pleaded guilty to deceiving Barings' auditors and cheating SIMEX. He was released in 1999.

What the margin system could and could not see

Barings' capital at the time was around US$615 million. By the point of collapse, Leeson's actual long Nikkei futures position had a notional value of about US$7,000 million against a reported, supposedly hedged position of only US$2,809 million. He also held an unauthorised short position in Japanese Government Bond futures with a notional value of about US$19,650 million, and a short options book of nearly US$6,680 million notional that was not supposed to exist at all. None of this showed up as a problem to the exchange, because SIMEX's margining system worked exactly as it was designed to: it called cash from Barings every time the market moved against the position. That system had no way of knowing, and no reason to ask, whether the person answering those margin calls in Singapore was reporting the true position to his own head office in London. He was not, and by the time the truth surfaced, the size of the position itself had become the danger. By the end, Leeson's futures position represented roughly 49% of open interest in the March 1995 Nikkei contract and 85% of open interest in the March 1995 JGB contract. A position that size cannot be closed quietly, whatever anyone decides to do about it.

This case sits primarily on Lecture 1, introduction to derivatives, leverage and margin, with clear links to Lecture 2 (clearing and margin mechanics) and Lecture 11 (options and short volatility). Three ideas from the course show up directly in the numbers.

The first is leverage. Exchange-traded futures require only a fraction of contract value as margin. That is exactly what makes them capital-efficient for a genuine hedger or arbitrageur, and exactly what lets a directional bet reach a huge notional size relative to the capital actually behind it. The worked calculation below turns Barings' own reported capital and Leeson's actual position sizes into a leverage ratio, and the result is startling even by the standards of this course.

The second is short volatility. A short straddle, a matched short call and short put at the same strike, has a capped maximum gain (the premium collected) and, in principle, unbounded loss in either direction if the market moves far enough away from the strike. Leeson's straddle book earned steady premium income for a year precisely because the Nikkei was quiet. Then the Kobe earthquake arrived, an event entirely unconnected to his trading book, at the single worst possible moment: a long futures position that needed the market to hold up, and a short straddle book that needed the market to stay calm, both hurt by the same shock in the same direction.

The third, and the point worth holding onto from the whole case, is that the instrument was never the problem. Exchange-traded futures and options are standard, cleared, margined products, and SIMEX's margin system called cash correctly every single day the market moved against Leeson. What broke was governance. One person held both the trading mandate and the settlement function that was supposed to check his own reporting, so risk limits at Barings' head office had nothing real to act on. Derivatives redistribute risk rather than eliminate it, and in this case, the redistribution of that risk away from public view was the entire mechanism of the failure.