Abstract
This chapter examines the nonlinear dynamics of exchange rate fluctuations using tools from chaos theory and nonlinear time-series analysis. The study analyzes four daily exchange rate series—euro/dollar, pound/dollar, dollar/yuan, and dollar/real—from 01/03/2010 to 09/30/2024, obtained from the Federal Reserve database. Daily returns were calculated using logarithmic differences of prices. To investigate the presence of deterministic chaos, the phase space of each series was reconstructed using the time-delay embedding approach. The time delay (τ) was determined through the average mutual information method, while the optimal embedding dimension (d) was estimated using the false nearest neighbors technique. Subsequently, the largest Lyapunov exponent was computed to evaluate the sensitivity of the systems to initial conditions. The empirical results indicate nonlinear behavior in exchange rate dynamics; however, no positive Lyapunov exponents were detected for the analyzed series. Therefore, the findings do not support the presence of deterministic chaos in these exchange rates.

This publication has 26 references indexed in Scilit: