台灣股市之多空市場及資產定價因子之實證研究; An Empirical Research on Capital Asset Pricing Factor under Bull and Bear Trend in Taiwan Stock Market

NCU Institutional Repository > 管理學院 > 產業經濟研究所碩士在職專班 > 博碩士論文 > Item 987654321/25893

請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/25893

題名:	台灣股市之多空市場及資產定價因子之實證研究;An Empirical Research on Capital Asset Pricing Factor under Bull and Bear Trend in Taiwan Stock Market
作者:	何世宗;Shih-Zhong Ho
貢獻者:	產業經濟研究所碩士在職專班
關鍵詞:	多空;誤設比;一般動差法;Noise Ratio;Bull and Bear;GMM
日期:	2006-06-26
上傳時間:	2010-06-11 17:06:54 (UTC+8)
出版者:	國立中央大學圖書館
摘要:	長久以來，CAPM單因子模型的解釋能力就有許多爭論，對於系統風險貝他β的解釋能力有許多的研究，直到Rose（1976）提出套利定價理論（arbitrage pricing theory；APT）後，才發展出多因素模型，也發展出許多財務基本因子的相關研究，如 Fama and French（1992）的三因子模型。同時，也有研究發現，在不同的多空趨勢下，相關的變數表現會出現不穩定的狀態。本文主要分為三個部份，第一部份，參考Fama and Macbeth（1973）的實證模型，本研究採用月資料型態，檢視CAPM的單因子模型是否可以解釋台灣股票市場。第二部份，我們參考Fama and French（1992）幾個財務基本因素（公司規模、市價淨值比、益本比、負債比）建構出多因素模型，採用季資料型態，檢視各變數對股票預期報酬率的解釋能力。第三部份，我們參考Fabozzi and Francis（1977）的三種區分多頭空頭方法，調整為短、中、長期多空定義，比較個股日報酬率，以虛擬變數（dummy variable）方式，加入多因素模型中，觀察在不同多空定義下，各變數解釋能力的變化。本研究採用台灣證券交易所自民國75年10月至民國94年6月間，具有完整季財務資料的56家上市公司為研究對象，以GMM估計法進行實證研究，利用Durlauf and Hall（1989）所提的誤設比（noise ratio），比較模型的適用性。本研究結果得到以下結論： CAPM單因子模型無法完全解釋台灣股票市場，多因素模型較能解釋台灣股票市場，除了系統風險β外，尚有公司規模、市價淨值比、負債比可以解釋股票報酬率。但是在多頭與空頭的趨勢之下，系統風險β對股票報酬率並不具解釋能力。多因素模型中對報酬率具有解釋能力的幾個因子，在多頭與空頭的趨勢之下，只有市價淨值比與負的益本比對股票報酬率具有解釋能力，但負的益本比效應是短期的。 It’s been a long time that many disputations exist in explanation of CAPM single-factor model and there are many researches to explanation of system-risk beta. Multi-factor model and many relevant researches to basic financial factor were not developed until year 1976 that Rose raised the arbitrage pricing theory, such as three factor model of Fama and Frehch (1992). Meanwhile, there also some researches reveal that unstable situation appears on the displays of relevant variables under different bull and bear trends. There are three major parts in this research. The first part is to check if single-factor model of CAPM can give a full explanation for Taiwan stock market or not. And it is based on Fama and Macbeth’s (1973) empirical model by using monthly data. The second part is to check the explanatory ability for various variables in predicting the rate of return for stocks. The multi-factors model is learnt from several basic financial factors of Fama and French (1992) by using quarterly data, such as firm size, book-to-market ratio, EP ratio, leverage and so on. The third part, we use dummy variable as well as multi-factors model to observe the change for the explanation ability between various variables under different definition for the bull and bear. Under this part, we refer the three ways to discriminate the bull and the bear from Fabozzi and Francis (1977). And we adjust the bull and bear definitions into short, medium and long terms respectively and compare rate of daily return for specified stock. This research is subject to the completed quarterly financial statements from 56 firms listed on Taiwan Stock Exchange for the period from October 1986 to Jun 2005. This research has been done by using GMM estimate method for case proven. We also learnt and made the most use of the concept of noise ratio raised by Durlauf and Hall (1989). Finally, we have successfully compared the suitable for all models. The conclusion of this research is as following: Single-factor model of CAPM is unable to give a fully explanation to Taiwan stock market. Multi-factors model give better explanation than single-factor model, because not only system risk beta, but also firm size, book-to-market ratio, and leverage can explain the rate of return of specific stock. But when the bull and bear variables were jointed into multi-factors model, system risk beta could not have explanation ability to the rate of returns of specific stock, only book-to-market ratio and negative EP ratio can explain the rate of returns of specific stock, but the effect of the negative EP ratio is only for short-term.
顯示於類別:	[產業經濟研究所碩士在職專班 ] 博碩士論文

文件中的檔案:

檔案	描述	大小	格式	瀏覽次數
index.html		0Kb	HTML	1252	檢視/開啟

在NCUIR中所有的資料項目都受到原著作權保護.

社群 sharing

資料載入中.....