What makes a rogue trader?

Kweku Adoboli, a 31-year-old trader for the Swiss bank UBS, listened silently in the City of London magistrates court this September as his lawyer apologised for his actions. Adoboli, who was born in Ghana and educated in England, has been accused of fraud for allegedly hiding $2.3bn of rogue trading losses on UBS’s Delta One equity derivatives desk.

Adoboli “is sorry beyond words for what has happened here”, the barrister said. “He went to UBS and told them of what he had done and stands appalled at the scale of the consequences of his disastrous miscalculations.” After the loss was discovered, UBS executives privately called Adoboli “a terrorist” and Oswald Grübel, the bank’s former chief executive, told a Swiss newspaper, “If someone acts in a criminal way, there’s nothing you can do.” Except resign, which he shortly did.

Since February 1995, when Nick Leeson brought down Barings by hiding an £830m loss in his supposedly lucky secret “five eights” account in Singapore, we have become accustomed to rogue traders. Leeson’s fraud collapsed just after Joseph Jett, a trader at Kidder Peabody in New York, had hidden losses of $74m while pretending to make a profit of $264m trading bonds. Then came Toshihide Iguchi, who accumulated losses of $1.1bn in bond trading at Daiwa Bank in New York, confessing in September 1995. After that there was Yasuo Hamanaka, who was found in 1996 to have lost $2.6bn trying to corner the copper market in London for Sumitomo Corporation.

A second wave broke out in 2002 with the arrest of John Rusnak, a trader for Allied Irish Banks (AIB) in Baltimore, who lost $691m on bad currency bets that he hid with fictional trades. A group of four currency-options traders at the National Australia Bank (NAB) in Melbourne lost AU$360m in 2004. Finally, the biggest ever case emerged in 2008, when Société Générale found that its trader Jérôme Kerviel had hidden €4.9bn of losses on the bank’s Delta One desk.

Rogue trading, it is now clear, is not an aberration but integral to the banking system. Like the cycles of financial speculation and crashes that have occurred throughout history, rogue traders are always with us.

Yet rogue trading appears to make little rational sense. Most of those who are caught gain little from their frauds – certainly not enough to compensate for the strains of concealment they endure for years, and the punishment they face. Economists assume that people behave rationally – they will only gamble when the potential reward outweighs the risk of loss, or the gain is worth the effort.

Many people were baffled as to why Leeson went through the ordeal for little beyond the prospect of a higher bonus. People including Peter Baring, the bank’s chairman, speculated that he had an accomplice, or an account where profits were being hidden. It turned out to be untrue. The rogue’s enigma is that he has a vast amount to lose, and precious little to gain.

So why do traders go rogue? To know what goes on in the mind of a rogue trader – and that of every reckless gambler – it helps to be a bird-watcher.

. . .

Experiments on bees and junco birds show that stress affects risk-aversity – just as it does with rogue traders

The yellow-eyed junco is a type of sparrow found in Mexico and the southern US. Thirty years ago, three evolutionary biologists at the University of Arizona carried out a series of experiments with seven yellow-eyed juncos which had been caught in the south-east of the state. The experiments were designed to test the birds’ gambling instincts, and the results were intriguing.

One bird at a time was placed on a perch in an aviary 3.5 metres from two dishes covered with paper so that it could not spy the contents. The bird was trained to realise that if it flew to the first dish, it would always find two millet seeds to eat; if it flew to the second, it couldn’t be sure of what it would find. Half the time, there would be four seeds, and half the time there would be none.

The biologists were trying to discover how much risk the juncos would take in foraging for food – their main activity in the wild. In the first experiment, the birds were kept hungry for an hour and then allowed to choose dishes, with seeds being replenished every 30 seconds. This meant that they would get plenty to eat unless they were unlucky. The juncos responded by being risk averse: in 19 out of 25 cases, they chose the sure thing – the dish that contained two seeds – rather than take the risk of finding nothing.

Next, the birds were starved for four hours, and the seeds were replenished only every minute. That tipped the birds into what the scientists called “a negative net energy budget” – if they got only two seeds each time they made a choice, it would not provide enough food to keep them alive and enable them to reproduce. The birds responded by flying to the other dish instead.

Facing loss, they started to gamble.

The scientists, led by Thomas Caraco, now a biology professor at the State University of New York, concluded that “juncos in nature will generally avoid risk unless they face difficult energetic stress”. In other words, they will choose the safe option for feeding – the closest natural equivalent to financial traders making money – unless they are in danger, either through hunger or cold. They will then switch from “risk aversion” to “loss aversion” – gambling in search of a bigger pay-off.

. . .

NICK LEESON, 1995: brought down Barings

This phenomenon is not confined to sparrows. Experiments have found similar behaviour in mammals such as shrews and insects such as bumblebees. The difference is that bumblebees’ responses are triggered by collective stress – in one test, they preferred blue flowers with a constant amount of nectar to yellow flowers with a variable amount as long as the colony had plenty of honey in store. As soon as it ran short, its members chose the yellow flowers.

The choices make evolutionary sense for the species. “Facing the possibility of starvation, animals are willing to gamble on the ‘strike-it-rich’ policy of risk-prone foraging,” writes Barry Sinervo, a professor of ecology at the University of California. “Some foragers will have a string of bad luck and starve. Some will have a string of average luck and still starve. However, there will always be those lucky few that experience a string of good luck. It is those lucky few that survive and pass on genes to the next generation.”

Risk aversion is not only displayed by Junco phaeonotus and Bombus pennsylvanicus – the variety of bumblebee in the blue flower/yellow flower test. It is also embedded in Homo sapiens. This fact was first noted by the Swiss mathematician Daniel Bernoulli in a famous paper published by the St Petersburg Imperial Academy of Sciences in 1738 which was so far ahead of its time that it was largely forgotten until it was reprinted in 1954 in the economics journal Econometrica.

Bernoulli, reflecting on an 18th-century form of gambling by tossing coins, found that humans are governed by the same loss-aversion instincts as animals (although proof of the latter took another two centuries). He pointed out that a wealthy person will not be as gratified as a poor person by winning a lottery.

“There is no doubt that a gain of a thousand ducats is more significant to a pauper than to a rich man, even though both gain the same amount,” Bernoulli wrote. The richer a man is, the less of a risk he will take to gain a sum of money because he is happier with what he already has. Like bees, shrews and juncos, people who are wealthy (the human equivalent of animals with energy reserves) become risk-averse.

Bernoulli did not consider what would happen if the person faced with a choice of whether to gamble were not merely poor but had negative wealth – in other words, was in the same predicament as a starving sparrow. It turns out, however, that humans behave like other animals. When they have lost a significant amount of money, they become risk-seekers.

. . .

Daniel Kahneman and Amos Tversky, two Israeli psychologists, identified loss aversion in humans at about the same time that Caraco and his fellow biologists were observing yellow-eyed juncos. Their findings, published in 1979 in Econometrica, won Kahneman the 2002 Nobel award for economics and were seminal to the field of behavioural finance, the study of why Homo sapiens behaves differently from the way classical economics predicts.

In experiments at universities in Tel Aviv, Stockholm and Michigan, Kahneman and Tversky set groups of people a series of tests. In one, they asked them whether, given a choice of gaining 450 units of the local currency, they would settle for that or take a 50-50 gamble on winning 1,000. Mathematically, the average gain was 500 – more than the 450 – yet the subjects were overwhelmingly risk averse. They chose the safe gain rather than the risky bet.

The psychologists then reversed the experiment (and a series of related ones) by offering the subjects a choice either of settling for a definite loss of 450, or taking a 50-50 gamble on losing 1,000 or losing nothing. This time, they rejected the definite loss for a half-chance of escaping unscathed.

Like other animals, they started to gamble.

In his Nobel lecture, Kahneman argued that Bernoulli had failed to recognise the “reference point” in people’s financial choices. What matters most of all is not how much a gamble alters their wealth, but where they start – whether they are already satisfied, or have suffered loss. Their overriding trait is their inability to accept loss: they struggle instinctively – irrationally, in economic terms – to evade it.

“Utility [the value people place on something] cannot be divorced from emotion, and emotion is triggered by changes,” noted Kahneman, whose recent book Thinking, Fast and Slow has become a bestseller. “A theory of choice that completely ignores feelings such as the pain of losses and the regret of mistakes is… unrealistic.” He was describing the psychological make-up of the average human being, but the desperate scramble to avoid losses at all costs is a defining characteristic of the rogue trader.

KWEKU ADOBOLI, 2011: accused of fraud for allegedly hiding rogue trading losses at UBS

“I have never experienced worse feelings than when I had the wrong positions, couldn’t get out of them, and had to manage them so that the losses were minimised,” writes David Bullen, one of the traders who cost NAB A$360m in 2004, in his memoir Fake: My Life as a Rogue Trader. “It can become gut-wrenching and it is a regular occurrence in currency options trading.”

In practice, most people caught in a sequence of losing bets will sooner or later come to their senses. What Kahneman called errors of intuition “perhaps corresponding to evolutionary history” – the instinct to keep on gambling like a sparrow or a shrew – gets overridden by “the deliberate operations of reasoning”. But something in the rogue trader makes him carry on gambling when the normal response would be to stop.

“As the spring wore on, I traded harder and harder, risking more and more,” wrote Leeson in Rogue Trader of how he responded to having built up a loss of £6m in 1992. “I was well down, but increasingly sure that my doubling up and doubling up would pay off ... This is gambling at its simplest. If you double up, you halve the amount the market needs to turn for you to make your money back, but you double the risk.”

Rogue traders usually start by making small losses and hiding them, often for a year or two before their trading escalates. Once they start to gamble in an effort to cancel them out, they are drawn into an escalating pattern of risk-taking and doubling, which culminates in huge losses in the final weeks before they are caught. Kerviel took small gambles at Société Générale in 2005 and 2006 before building a €28bn stake in futures (on which he made a profit) during 2007 and then doubling up in a disastrous €49bn gamble; Leeson had lost £208m of Barings’ money by December 1994 but quintupled that loss in seven weeks through doubling; Rusnak had lost $300m for AIB by the end of 2000, and tried to avoid detection by selling $300m in options.

. . .

Rogue traders’ behaviour is financially stupid – traders are trained not to double their losses but to set risk limits and to retreat from risky positions when they face danger. Biologically, however, doubling is sound. Starving animals should gamble in search of a food windfall, even though each has only a small chance of success. Most will fail, but some will survive and reproduce.

Gambling even at bad odds pays because of the pattern of sexual reproduction – a few survivors produce a lot of descendants. “Living organisms, including humans, tended to propagate rapidly,” writes the anthropologist Azar Gat. “The small founder groups that arrived in the Pacific islands in the last two millennia rapidly filled up their new habitats, increasing in numbers to thousands and tens of thousands.”

It feels implausible that rogue traders are driven by Darwinian instinct, but humans have been found to use the emotional and instinctive regions of their brains when opting to gamble, instead of the prefrontal cortex that controls intellectual reasoning. The culprit is the amygdala, a part of the brain predating Homo sapiens and some great apes, which produces the “fight-or-flight” reflex.

Neuroscientists at University College, London who gave 20 people a series of risk-taking choices while they were in an fMRI scanner found that the amygdala was most active in risk-averse and loss-averse people – those who made decisions instinctively. “In evolutionary terms, this may confer a strong advantage,” they wrote. “In modern society ... where optimal decision-making often requires skills of abstraction and decontextualisation, such mechanisms may render human choices irrational.”

Thomas Brennan and Andrew Lo, two professors at Northwestern University and the Massachusetts Institute of Technology, believe that evolutionary instincts explain a lot of what appears to be “rogue” behaviour. “Faced with life-threatening circumstances,” they say, “even the most disciplined individual may not be able to engage in individually rational behaviour thanks to adaptive ‘hard-wired’ neural mechanisms that conferred survival benefits to the species.” The survival of the fittest, they add, underlies “modern adaptations such as boredom, thrill-seeking behaviour, rebellion, innovation, and most recently, financial market bubbles and crashes... From an evolutionary perspective, financial markets are neither efficient nor irrational – they are merely adaptive.”

The rogue trader may appear to be acting strangely, but he does what comes naturally.

John Gapper is the FT’s chief business commentator. This piece is adapted from his ‘How to be a Rogue Trader’, a short ebook published this week by Penguin Portfolio (£1.99/$3.99).

To comment on this article, please email magazineletters@ft.com