A Quantitative Approach to Technical Analysis

The authors use a smoothing estimator based on neural networks. Smoothing estimators estimate nonlinear relationships (such as chart patterns) by averaging the data through a variety of sophisticated steps. The authors rigorously tested their method on four years of data for 20 stocks and on 20 years of data on 18 foreign currencies. They conclude that quantitative technical analysis can and should play a role in managing a portfolio. The authors' work in this area is a first step in raising technical analysis to an equal footing with other forms of financial analysis, and should change the way this practice is viewed forever.