Long-Term Dependencies and the Profitability of Technical Analysis

Abstract

The profitability of nine trading rules and the Hurst exponent (H) were calculated on fifteen of the largest global equity markets in order to determine if technical trading rules are more profitable in markets characterized by long-term dependencies. Pearson correlation and regression analysis suggests that there is some association (rho = 26%) between the profitability of trading rules and the Hurst exponent. However, the fact that many of the trading rules were profitable in markets with low H suggests that the trading rules are able to generate useful information by processing past prices into a more informative trading signal than that provided by the random-walk model.

Keywords: Technical Analysis, Time-series Analysis, Hurst Exponent, Market Efficiency

JEL Classification: C4, G11, G14