Technical Analysis and Currency Trading: False Discoveries and Informative Covariates

Abstract

We develop a functional false discovery rate method that uses multiple informative covariates to examine the conditional performance of predictive models while controlling for data snooping. This method has superior power and is robust to data dependence, estimation errors and correlated covariates. We apply this new method to a large set of currency technical trading rules and construct a dynamic 30-currency portfolio that generates a Sharpe ratio around one for roughly 50 years. Technical trading profitability decreases with the computational power of traders and capital account openness, suggesting that such profitability is related to how fast foreign exchange traders can detect specific market patterns.

Keywords: Multiple testing, Multivariate functional false discovery rate, Informative covariates, Technical analysis, Foreign exchange markets.

JEL Classification: C12, C58, G11, G15