Chaos and the exchange rate

Authors:
  • DFDaniela Federici
    Daniela Federici
  • GGGiancarlo Gandolfo
    Giancarlo Gandolfo

  • 1 January 2001
Abstract
The interest of economists in chaos theory started in the 1980s. The first to draw the attention of economists to chaos theory was, in fact, Brock (1986), who examined the quarterly US real GNP data 1947-1985 using the Grassberger-Procaccia correlation dimension and Lyapunov exponents. Subsequent studies generally found absence of evidence for chaos in macroeconomic variables (GNP, monetary aggregates) while the study of financial variables such as stock-market returns and exchange rates gave mixed evidence. Studies aimed at detecting chaos in economic variables can be roughly classified into two categories. On the one hand, there are studies that simply examine the data and apply various tests, such as the studies mentioned. These tests have been originally developed in the physics literature and typically require several thousand observations. Apart from this data problem, such an approach is not very satisfactory from our point of view, which aims at finding the dynamic model underlying the data. On the other hand, structural models are built and analysed. This analysis can in principle be carried out in several ways: a) theoretically, namely showing that plausible economic assumptions give rise to dynamic structures having one of the mathematical forms known to give rise to chaotic motion; b) empirically, namely building a theoretical model and then b1) giving plausible values to the parameters, simulating the model, and testing the resulting data series for chaos; or b2) estimating the parameters econometrically, and then proceeding as in b1. Existing chaotic exchange rate models (De Grauwe and Versanten, 1990; Reszat, 1992; De Grauwe and Embrechts, 1992, 1993a,b; De Grauwe, Dewachter, Embrechts, 1993; Ellis, 1994; Szpiro, 1994; Da Silva, 1997) follow the structural approach: they are structural models built in discrete time (difference equations). From the theoretical point of view, these models show that with orthodox assumptions (PPP, interest parity, etc.)
All Related Versions