LONDON (Reuters) - When John Coates was on a winning streak during his days as a trader at Deutsche Bank and Goldman Sachs, the narcotic-like "high" he experienced was so powerful he was determined to find out more.
So after 13 years on trading floors on Wall Street he moved to the neuroscience labs of Rockefeller University in New York and of Britain's Cambridge University.
Here, the trader turned neuroscientist has been bent on uncovering the brain biology behind that high, what it did to him, and what it's probably doing to those he left behind.
What he's come up with, after several years reading up on animal studies and some interesting experiments with spit, is that risk taking is driven by a "winner effect" - a hormonal mechanism in which each competitive victory leads to more wins.
"The narcotic high was as powerful as anything I have ever felt," Coates said in an interview during a medical conference in London, describing the experience of making huge profits and big bonuses at some of the world's largest banks.
And as other experts in psychiatry and neuroscience at the conference agreed, the consequences of a winner effect gone out of control can lead some to become power-corrupted politicians, cruel military dictators and even surgeons who like to play god.
"You become euphoric, delusional, you have less need for sleep, you have racing thoughts, an expanded appetite for risk, and less stringent requirements in the risk and reward trade-off," said Coates. "Basically, you become a rogue trader."
Since publishing some initial scientific studies exploring these traits in traders, Coates says he has been contacted by researchers analyzing politicians, soldiers, and even sports people who believe his work can shed light on theirs.
"As our research progressed, it became clear we were doing a lot more that studying the biology of financial risk taking, we were studying the biology of (all) risk taking," he said.
"We only have one biology, and we take it with us into whatever world we're engaged in - whether it's the military, sports, politics or finance."
With evidence of extreme consequences of the winner effect - traders who turn rogue and bring down entire banks, political leaders corrupted by power who inflict cruelty on subordinates, or soldiers who become indiscriminate killing machines unchecked by the rules of conflict - Coates is looking deeper.
"When you see this transformation take place in people, they start carrying themselves like masters of the universe. And it's not a cognitive process. It isn't even about greed. It's more this feeling of consummate power, a feeling that you're dominating the world."
Coates says he was increasingly struck by the fact that "almost every blow-up north of a billion dollars - the sort of blow-up that shakes a bank to its foundation" - came down to the actions of a trader at the end of a winning streak.
"The winning streak seems to foster excessive risk-taking," he said.
Intrigued, and keen to bring his previous experience to his new role as a Cambridge research fellow in Neuroscience and Finance, Coates asked some of his former colleagues in London's City financial district to give him some time, and some spit.
Over eight consecutive business days, researchers took spit samples from 17 male traders, morning and afternoon, to measure levels of the hormone testosterone during daily trading.
The results were revealing. Daily testosterone was significantly higher on days when traders made more than their one-month daily average. And on mornings when they had high testosterone levels, their profits for the rest of the day were significantly larger than when testosterone levels were low.
These findings echo similar studies of animals in the wild, which also found a testosterone-driven winner effect among males who fight over territory or a mate, for example.
According to Coates, they also show without doubt that risk taking in humans is a physiological and not just a cognitive activity.
"Within economics, there's a belief that we wander around with this supercomputer in our heads that is unaffected by the body and has the ability to calculate returns, probabilities and the optimum allocation of capital," he said. "But of course the science doesn't support anything like that."
Coates' observations chimed with those of several other speakers at the conference, which gathered psychiatrists and neuroscientists to examine the phenomenon of hubris in public life - in other words what leads people in power to become corrupted and behave in arrogant and destructive ways.
Nassir Ghaemi, a professor of psychiatry at Tufts University Massachusetts, told the conference disorders like depression can often enhance political, economic and military leaders at times of crisis because depressives are more empathetic, more self critical and more realistic about the world around them.
Coates suggests what goes wrong in the case of rogue traders is that the hormonal mechanism behind the winner effect becomes pathological, fostering "irrational exuberance" and excessive risk taking.
He also said it is not enough for commentators and analysts to simply observe these activities, but argued that they should demand proper scientific studies which can provide robust answers to questions about what went wrong.
"What I'm describing is overlooked scientific data," he told the conference. "And what we're seeing in the corporate world is a desperate need for science conducted in the workplace. It's going to help us understand the sources of the instability, and how to control it."
Coates' hypothesis is that at a certain level of rising testosterone, effective risk taking gradually turns into a biological wave of excessively risky behavior. If that is the case, it should change the way traders are managed, he said.
"The trouble with the banks is that their risk management systems and compensation schemes have been amplifying these biological waves, when they should be leaning against them.
"What they should be doing with traders on a winning streak is not forever raising their risk limits, but holding their limits constant, or even telling them to close out their positions in the middle of a winning streak and take three weeks off until their biology resets."
Reporting by Kate Kelland, editing by Ralph Boulton