Year: 2004
Type: article
Abstract: Technical trading rules are tested to determine their economic and statistical significance as an investment technique using S&P 500 index Futures from 1982 to 2004 and three main subsets of rules: filter rules, trading range breaks and moving average rules.Using a cumulative net profit algorithm designed by Boswijk, Griffioen & Hommes in their 2001 study, we calculate the cumulative net profit of each rule at the end of the initial 12-year period and then compare to a similar test in the 10-year out of sample period.We find that the majority of rules were not consistently profitable within a single period or across both time periods.Individual rules are tested using a makeshift ranking system and it is again found that only a very small fraction of the rules were able to consistently classify as being in the upper rank(s).We then test for the ability to forecast daily Futures returns under the assumption that said returns are identically and independently distributed (IID) and not IID.In either case, only a very small portion of rules demonstrate significant return forecasting ability in the initial or final period.It is concluded that technical trading rules on the whole do not have significant return forecasting power and do not lead to profitable investment decisions which would suggest that technical analysis is not successful in the S&P 500 index Futures market.
Author: Martin Goulet
Language: English
Subfield: Finance
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W620804559
