To see into the future, economic and financial forecasters extrapolate from the past. So imagine their distress when, in 1974, Welsh-born economist Clive W.J. Granger showed that the statistical techniques forecasters were using to come up with patterns in historical data were simply wrong. "The reaction was chaos for a few years," recalls Princeton University economist Mark W. Watson. Forecasters, including those at the Federal Reserve, had to find new methods that would work.
Eventually it was Granger himself, along with colleague Robert F. Engle, who came up with the statistical techniques that are now used in areas as diverse as bank regulation, mortgage derivatives, and the pricing of stock options. It was for these techniques that Granger and Engle -- who had neighboring offices at the University of California at San Diego for 25 years -- were jointly awarded the Nobel prize in economics on Oct. 8. Granger, 69, a British citizen, retired from teaching earlier this year. Engle, 60, an American, moved from San Diego to New York University's Stern School of Business in 2000.
EASIER TO HANDLE
Granger and Engle's innovations go by such arcane names as "Granger causality" and "autoregressive conditional heteroskedasticity." Basically, though, they come up with improved ways of handling economic and financial variables such as interest rates and factory output.
Unlike tosses at a dartboard, these variables are somewhat predictable -- if an interest rate is 4% one day, it won't be 11% the next. And their volatility varies. In the financial markets, for example, periods of extreme fluctuations are followed by periods of calm. Older statistical tools couldn't cope with those features.
Granger's work has been a boon to macroeconomists, while Engle's has been crucial to modern finance. For instance, the modern version of the Black-Scholes formula for pricing options -- including stock options -- incorporates adjustments made possible by Engle's work, says Massachusetts Institute of Technology economist Andrew W. Lo. "Nobody," he says, "uses plain-vanilla Black-Scholes anymore."
Engle's work also underpins how financial firms calculate their "value at risk" -- namely, the most money they stand a chance of losing over, say, the next week. The Basel Committee on Banking Supervision is moving toward requiring all of the world's banks to employ value-at-risk models to determine how much capital they need to keep on hand as a cushion.
Granger and Engle helped build what might be described as the statistical plumbing of modern economics and finance. It may not be glamorous, but as the Nobel committee recognized, it's indispensable.
By Peter Coy in New York