| Summary: | 碩士 === 靜宜大學 === 企業管理研究所 === 90 === Weak form market efficiency implies that stock price has contained the related information in the past that investors cannot get excess return through the technical analysis. In this case, investors can only diversify the risk through a portfolio strategy. Owing to its importance, the weak form market efficiency has been a focus no matter in the academia or practice for years.
Price and quantity are two important variables in technical analysis. Ying(1966)、Copeland(1976)、Epps(1975) and Smirlock & Starks(1985) revealed that quantity of stock is positively related to the absolute value of price change. Some researches also emphasized the existence of linear relationship among price, quantity, and next day’s price change. However, the nonlinear relationships are rarely referred to. This paper assumes that the relationships among the variables are complicated more than just linear relationship, and tries to capture the nonlinear relationships by using a hybrid technique—neuro fuzzy. Moreover constructing a price-quantity based trading system to probe with Weak Form Efficient.
The objective of this paper is combining price-quantity technical index and neuro-fuzzy hybrid technique to construct a trading system for each Morgan stocks. Another we also compare the portfolio performance constructed by this proposed trading system with the performance of Markowitz model.
The empirical results show that the proposed model beats the market in return of year and sharpe ratio. It is also right in different market condition. When the trading cost increases, the return of neuro-fuzzy is eroded. Another proposed model beats the market in return of day and each of trading is profitable. The four kinds of excess return index are all positive. The return and sharpe ratio of portfolio is better than Markowitz, and all better than other indexes.
|
|---|
