An Application of the Dow Theory to the Taiwan Stock Index

Abstract

In this paper we propose a systematic approach to recognizing technical patterns and we apply this method to TAIEX index from 1990 to 2017 to evaluate the effectiveness of technical analysis Using OLS regression we find the two technical indicators head-and-shoulders top (HST) and head-and-shoulders bottom (HSB) are able to predict the accumulative returns after the patterns occur Using the Fama-and-French (1993) three-factor model as a style benchmark for the accumulative returns we find that the abnormal returns are statistically significant At last we back-test the trading strategies and find that the HST and HSB strategies outperform the sell-and-hold and buy-and-hold benchmarks respectively The strategy combining both HST and HSB earns about 8% annual returns and around 6% annual returns for time period of five trading days before and after transaction costs respectively Moreover we find that the strategies we proposed tend to generate more returns in bear markets than in bull markets

Date of Award	2018 Jun 26
Original language	English
Supervisor	Meng-Feng Yen (Supervisor)