Forty Years, Thirty Currencies and 21,000 Trading Rules: A Large-Scale, Data-Snooping Robust Analysis of Technical Trading in the Foreign Exchange Market

Abstract

We carry out a large-scale investigation of technical trading rules in the foreign exchange market, using daily data over a maximum of forty years for thirty developed and emerging market currencies. Employing a stepwise test to safeguard against data-snooping bias and examining over 21,000 technical trading rules, we find evidence of substantial predictability in both developed and emerging markets, measured against a variety of returns and risk-adjusted performance metrics. We present time-series and cross-sectional variation in sub-periods and cultural and/or geographic groups, respectively, suggesting that temporarily not-fully-rational behavior and market immaturity lead to technical predictability and potential profitability.

Keywords: Data snooping, Foreign exchange, Technical analysis

JEL Classification: F31, C53, G15