Nonlinear dynamics in foreign exchange rates
Academic Article
Overview
Identity
Additional Document Info
Other
View All
Overview
abstract
This paper investigates whether the behavior of real and nominal foreign exchange rates as well as interest rates are governed by nonlinear dynamics; it also explores whether observed deviations from parity conditions exhibit nonlinear dependence. Standard statistical tests for randomness, such as autocorrelation tests, have low power against a large class of deterministic, nonlinear processes. Discerning nonrandomness of innovations in exchange rates is important for a variety of reasons. For example, many models of international asset pricing assume exchange rates to follow a random walk. Furthermore, nonlinear patterns in deviations from various exchange rate parities have implications for the existence of a time-varying foreign exchange risk premium. With the use of the BDS statistic and a correlation dimension analysis, this paper's primary findings are that (1) foreign exchange markets have become increasingly complex and therefore less amenable to forecasting over time; (2) although forward exchange risk premia are statistically significant and display a deterministic structure, this structure is complex and therefore not easily discernible; and (3) innovations in real exchange rates are consistent with a Purchasing Power Parity equilibrium. 1999 Elsevier Science Inc.