All the Devils Are Here [38]
By the early 1990s, Weatherstone had instituted something called the 4:15 Report. Every afternoon at 4:15—just fifteen minutes after the market had closed—all the top J.P. Morgan executives were sent a document that listed the firm’s overall VaR for that day, as well as the VaR number for all the various trading desks around the world. No longer were executives in New York blind to the risks being taken in London, or Tokyo, or anywhere else in the world.
Later, many Wall Street CEOs would view their daily VaR number as an expression of their firm’s worst-case scenario. But it was nothing of the sort. The most important information VaR conveyed was not the absolute number, but the trend over the course of weeks or months. Were the bank’s risks increasing or diminishing? Were problems arising on this desk or that one? And so on.
And then there was the tail risk issue. The fact that VaR told you how much your firm might lose 95 percent of the time didn’t say a thing about what might happen the other 5 percent of the time. Maybe you would lose a little more than the VaR number—no big deal. Or maybe you’d get caught in a black swan and lose billions. The fact that VaR had been created didn’t mean you could stop worrying about risk.
Weatherstone understood this completely. “I remember meeting with him,” says a former J.P. Morgan risk manager. “I would show him the VaR numbers and tell him that a certain currency trade had gotten 20 percent riskier. The currency guy would fight it. [Weatherstone] would listen to the arguments. He wouldn’t say a lot. Then he would make a decision about whether the currency desk had taken on too much risk. And it was based not just on VaR but on the deeper discussion that it sparked.” Which, for its creators at J.P. Morgan, was how VaR was supposed to work. Though it was an important data point, it was never meant to be the only data point.
Having created VaR, Guldimann then did something highly unusual: he gave it away. The theme of the bank’s 1993 client conference was risk. By then, other firms were grappling with the same set of issues that had led J.P. Morgan to create VaR. When Guldimann explained the bank’s new risk model at the conference, many of J.P. Morgan’s clients began clamoring to learn more about it. Some of them asked if they could purchase the underlying system.
Most banks would have likely declined; after all, VaR was J.P. Morgan’s intellectual property. But Weatherstone and Guldimann understood that if some banks took excessive risks they didn’t understand, it would be bad for everybody, J.P. Morgan included. It would be better, they believed, if everyone had access to the same risk tools. But since they also didn’t want to turn risk management into a side business, they decided to teach VaR to anyone who wanted to learn about it—free of charge. “Many wondered what the bank was trying to accomplish by giving away ‘proprietary’ methodologies and lots of data, but not selling any products or services,” Guldimann wrote years later. “It popularized a methodology . . . and it enhanced the reputation of J.P. Morgan.” By the late 1990s, VaR had become the de facto standard for risk models. Everybody used it. They had to.
Even as Guldimann was popularizing VaR, another group of J.P. Morgan quants, in a different corner of the firm, was creating a new kind of derivative: credit default swaps. The project grew out of the same impulse as VaR—the bank’s ongoing effort to better manage its own risks. In this case, though, the risk in question was perhaps the most basic in all of banking: the risk that a borrower might be unable to pay back his loan. Credit risk, in other words.
The best way to deal with the possibility of default of course, is to make good