Using Genetic Algorithms to Search Optimal Technical Indicators

• Main Authors: Chih-Ho Hsu, 許智和
• Other Authors: Shung-Ming Tang
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/37556658335610829558

碩士 === 國立雲林科技大學 === 資訊管理系碩士班 === 91 === Use the technical indicators to predict stock index , that is the main concept of technical analysis. But the variety and inconsistency of technical indicators usually confuse the investor. Genetic algorithms own the character of parallel searching and adaptability can help investor to search the optimal trading rule. So this study try to use genetic algorithms to search the optimal trading rule, and divide into two periods. The searching period was from 1995 to 1999 and the testing period from 2000 to 2002. In the searching period, we find the six trading rules .The performance of the six trading rules is obvious better than buy-and-hold strategy. It prove the genetic algorithms can search the optimal trading rule. In the testing period, only one trading rule is worse than buy-and-hold strategy, the others are better than buy-and-hold strategy, so we can confirm the optimal trading rule have the ability of earning money. The technical analysis is worth understanding, and investor should not think of the common technical indicators as the best trading rules.