Fool's Gold: How the Bold Dream of a Small Tribe at J.P. Morgan Was Corrupted by Wall Street Greed and Unleashed a Catastrophe by Gillian Tett Read Free Book Online Page B

prepare. The terrible stock market crash of 1987 hammered that point home again when all the banks suffered losses. So, in 1989—two years after that crash—Weatherstone introduced a near-revolutionary practice into J.P. Morgan known as the “4–15 report.” At 4:15 p.m. each day, after the markets closed, a report was sent to him that quantified the level of risk the bank was running in all areas of its business. Initially the report was crude, limited to painting an approximate picture of all the bank’s trading businesses. Weatherstone decided he wanted more, and he asked a team of quantitative experts to develop a technique that could measure how much money the bank stood to lose each day if the markets turned sour. It was the first time that any bank had ever done that, with the notable exception of Bankers Trust.
    For several months the so-called quants played around with ideas until they coalesced around the concept of value at risk (VaR). They decided that the goal should be to work out how much money the bank could expect to lose, with a probability of 95 percent, on any given day. The 95 percent was an accommodation to the hard reality that there would always be some risk in the markets that the models wouldn’t be able to account for. Weatherstone and his quants reckoned there was little point in trying to run a business in a manner that would create obsessive worry about very worst-case scenarios. If a bank worried about those every day, it could barely afford to conduct business at all. What Weatherstone wanted to know was what levels of risk the bank was running in a bad-to-normal state of affairs, like a farmer who steels himself to expect periodic floods, snowstorms, and droughts but doesn’t worry about an asteroid impact that might bring on Armageddon. Hence the use of a 95 percent confidence level.
    The system essentially worked by taking data from the previous few years and working out how much money the bank risked losing at any one time if markets turned sour. The key simplifying assumption on which the models rested was that the future was likely to look like the recent past. Dennis Weatherstone was well aware that this assumption might not always be correct. He knew it was only a rough approximation to reality. Models were only tools and could not be used without human intelligence, as he sternly lectured the J.P. Morgan staff. They were not the only way of measuring potential future losses. “We were all taught to think that models are useful, but that they also have limits,” Peter Hancock later recalled. “It is an obvious point, but it is also something people so often forget.”
    Yet the beauty of VaR was that it allowed bankers to measure their risk with far more precision than ever before. A mind-boggling array of future dangers could be reduced to a single clear number, showing how much a bank stood to lose. The J.P. Morgan bankers assumed that would make it much easier for banks to control their risks, in relation to derivatives or anything else. That had a bigger implication: with tools such as VaR at their fingertips, banks now had both the incentive and the ability to navigate wisely in the derivatives world without the need for government interference. Or that, at least, was the argument that derivatives bankers liked to present.
     
    On July 21, 1993, the long-awaited G30 report was finally unveiled. It was a hefty, three-volume tome. Right from the start, Dennis Weatherstone had insisted that the study be a serious, highly detailed guide. If the report could show that the industry already had a credible internal code of conduct, there should be less need for bureaucrats to impose rules. “This should not be a study that gathers dust on a shelf,” Weatherstone declared. “I want to produce a guide by practitioners that has so much useful, practical advice that it will be referred to for years to come.”
    The G30 tome met that criterion in spades. The document started with