Stock Trading Based on Principal Component Analysis and Clustering Analysis

Abstract

Predicting stock price becomes a hot problem in the realm of investment for investors, brokers and dealers. Stock selection in active project investment is a challenging and significant task. By using technical analysis, analysts make efforts to select stocks and set transaction rules. In this paper, 9 technical indicators sere used to select appropriate stocks. Principal component analysis and cluster analysis were applied to estimate the stock returns in the day of the training time by selecting the earnings in the top 30 stocks. Finally through calculating annual yield, maximum retracement, Sharpe ratio, ratio of information, quantitative evaluation of a strategy was analyzed. The result showed the validation of technical indicators in the analysis of stock trading. This paper provides theoretical support for investors in technical analysis. Investors could make a reasonable judgment on the price trend of the stock market.