Testing Long memory in exchange rates and its implications for the adaptive market hypothesis

Abstract

This paper focuses on testing the long memory in exchange rates series of 18 emerging and 10 developed foreign exchange markets against the USD, along with explaining the behavior of these markets by the implications of the adaptive market hypothesis. Our empirical approach is based on DFA methodology for computing Hurst exponent (a parameter for capturing long memory) along with the established bootstrapping procedure for improving the results obtained. We also analyzed the impact of the great financial crisis (2007–09) on the efficiency of these foreign exchange markets by dividing our sample period into three sub-periods; pre, during, and post-financial crisis. Finally, the rolling window approach is applied for capturing the time-varying dimension of market efficiency. All the aforementioned tests confirmed that the emerging and developed foreign exchange markets exhibit episodes of long memory (persistence long memory/anti-persistent long memory) and no long memory signaling their efficiency has been affected by the financial crisis due to the changes in the behavior of market participants, and they follow a time-varying market efficiency. Our results confirmed that the behavior of these markets is consistent with the implications of the Adaptive Market Hypothesis.