The Nobel prize for economics has been awarded to researchers whose mastery of mathematical methods could have consequences way beyond the 'dismal science'.

The prize was awarded to Robert Engle of New York University and Clive Granger of the University of California, San Diego, for their development of methods to analyse apparently random fluctuations in economic data.

The dynamics of economic statistics — including indices of national output, stock prices and exchange rates — pose a daunting challenge to scientific study. These indices tend to vary unpredictably over time, with periods of high volatility interspersed with calmer spells.

Engle, an American who trained as a physicist at Cornell University, devised statistical models to describe the plots of such indices over time — known as time series — using a concept that he called autoregressive conditional heteroskedasticity (ARCH). Models of market volatility based on the idea are now widely used by market analysts and bankers to evaluate risk.

Welsh-born Granger, meanwhile, developed statistical methods that help analysts to deal with time series that fluctuate around a moving baseline, by combining them in a way that removes their long-term drift, through a phenomenon that he called co-integration. This makes such data amenable to conventional statistical analysis, and helps to identify causal connections between data sets.

“If you are going to give a prize for economic time-series analysis, these are the two to give it to,” says economist Paul Ormerod of London-based company Volterra Consulting.

Besides economics, the winners' methods have found widespread application in the analysis of nonlinear systems in such diverse areas as climatology, physiology and physics. Granger's techniques have proved to be particularly versatile and have been used, for example, to assess human influences on climate change.