Time-series pioneers share Nobel Prize for Economics

Recipients’ work on the analysis of time-series data has provided powerful mathematical tools for research in derivatives pricing, risk management and economics

The Nobel Committee said Engle’s award recognises his role in devising “methods for analysing economic time series with varying volatility”, while Granger created “methods of analysing economic time series with common trends”.

Up until the mid-to-late-1980s, many risk managers, traders and academics largely assumed volatility was constant over time. The jettisoning of this flawed assumption in the interim is in no small way due to Engle’s work in the early 1980s, when he developed the concept of autoregressive heteroskedasticity (Arch). This statistical technique ushered in a new era of more realistic volatility modelling.

Engle will speak about a new correlation estimator for credit risk evaluation at Risk’s Credit Risk Summit Europe 2003, which will be held in London on October 21 and October 22. He has a long relationship with Risk, with some of his earlier volatility research appearing in the magazine - see ‘Garch for Groups’, Risk August 1996, page 36; and ‘Grappling with Garch’, Risk September 1995, page 112. Both technical papers were co-authored with Joseph Mezrich.

Granger’s research has been particularly useful in macroeconomic analysis. Most significantly, he discovered certain combinations of non-stationary time series exhibit stationarity - that is, the series fluctuates around a specific value. This result allows robust statistical inferences to be made, which shed new light on the relations between consumption and wealth, and short- and long-term interest rates, for example.

Engle and Granger will receive a total award of SKr10 million ($1.3 million).