| Summary: | 碩士 === 長庚大學 === 企業管理研究所 === 95 === The issue discussed in this literature is that can investors get abnormal or excess return by observing price pattern of open and close trading period in spot market and futures market in Taiwan. Instead of identifying price patterns by human subjectivity, Self-Organizing Maps (SOM) is used to classify price patterns by its characteristic, which is more objective and appropriate in this issue. Moreover, the AIC rules in econometric field and the idea of hierarchical are also introduced to this article to improve some limitation while using SOM.
The empirical results shows that in open period, some price patterns are meaningful to following price performance in both market, but price patterns derived in close period of future market can’t appeal a good effect in forecasting following price performance. Besides, the patterns in spot market is dependent between open period and close period, but in future market, the patterns in different period are independent with each other. Furthermore, the patterns discovered in this study doesn’t like the patterns in traditional technical analysis, they don’t have much in common.
|
|---|
