The Research for Mean Reversion and Quasi-Arbitrage of Business Group Stocks

• Main Authors: Tsai-Ling Tseng, 曾彩玲
• Other Authors: Ai-Chi Hsu
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/38511476981310023904

碩士 === 國立雲林科技大學 === 財務金融系碩士班 === 95 === Taiwan’s business groups has significant influence on Taiwan’s economic activities. Investor also has certain appreciation of the importance of business groups. Thus, the variation of stock price between group members will be an important factor to investors. This research use stationary inspection method to find out non stationary series stocks in business group, and pairs of those have similar trend and long term relationship will become stationary series. Then, we use stationary series to do mean reversion research. Furthermore, base on the theory of mean reversion, when stock price overreact, it will revert to market price in long term. Therefore, we could reap profit by utilizing the relation of fluctuated price to do quasi-arbitrage; ARMA model is used for predicting the ratio between asset and stock price to analyze the investment return of individual stock. Finally, quasi-arbitrage are compared with ARMA model to find out which has the higher annual return ratio. This research shows that 18 asset pairs has mean reversion；The average annual return predicted by ARMA model is between –6.4% and 122.7%. After comparing the investment return among Min type of quasi-arbitrage, Max type OF quasi-arbitrage and ARMA prediction model, we find the Min type has the best return ,Max type is the second and ARMA model is the worst.