A study of the predictive performance of the moving average trading rule as applied to NYSE, the Athens Stock Exchange and the Vienna Stock Exchange: sensitivity analysis and implications for weak-form market efficiency testing

Abstract

This work examines the variation of the simple Moving Average (MA) trading rule performance as a function of the MA length in New York Stock Exchange (NYSE), Athens Stock Exchange (ASE) and Vienna Stock Exchange (VSE) using daily data from May 1993 to April 2005. Results show that changes of the MA trading rule performance as a function of the length of the MA are in many cases random. Moreover, in the presence of a 0.5% fee per transaction, trading rule performance as a function of the MA length in several cases follow a random walk with a positive drift process, implying better performance for longer MAs. To an extent, due to the large variability of the trading rule performance observed in many cases, these results weaken previous conclusions regarding the predictive power of the rule where use was made of MAs with only specific lengths, as well as any conclusions regarding acceptance or rejection of the weak-form market efficiency hypothesis. Further, a preliminary qualitative analysis showed enhanced trading rule performance for very short MA lengths, a result which needs further investigation.