Kelly Criterion under Stock Price Pattern Recognition Method

• Main Authors: Ya-ting Chang, 張雅婷
• Other Authors: Jen-Jsung Huang
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/38493864194241015661

碩士 === 國立中山大學 === 財務管理學系研究所 === 104 === Recently, previous research indicated that whether quantitative investment method can earn higher return in the stock market. This paper uses pattern recognition as predicting method, supposing history will happen again in the future. This paper compares the index pattern between current market index and whole market historical data. After comparing process, we can find the fitting pattern and use its next week data of return to calculate the probability of going up and down and the average return and loss. Moreover, this paper combine pattern recognition and Kelly criterion to back test. Although using Kelly leverage instead of fix leverage cannot increase annual return, it can decrease the risk of investment and increase Sharpe ratio. Finally, we apply this method to China and Taiwan stock market. Comparing two market, we find pattern recognition is more useful in China than in Taiwan. This research expects China’s high annual return will decrease in the future when China financial governance become more mutual as Taiwan market.