| Summary: | 碩士 === 淡江大學 === 財務金融學系碩士班 === 93 === After Sharpe (1964) and Lintner (1965) proposed the Capital Asset Pricing Model (CAPM), many scholars expressed differing views as to how well the model can reasonably and completely portray the rewards of risky asset. Among them, two theorists, Fama and French (1992, 1993, 1995, 1996), introduced the three-factor model, which helped extend the CAPM into a better, more complete model. This study will utilize the time-series regression analysis method, using the monthly return of the US Stock Market from July, 1934, to December, 2004. With the 10 Industry Portfolios and the 25 Portfolios as an example and the three-factor model as a basis, the Maximum Likelihood Estimates (MLE) and the Generalized Method of Moments (GMM) methods were used to estimate the value of the regression parameter. Through the perspective of the Sharpe-Lintner CAPM, one can find out whether using different methods of estimation for varying sample data will result in different empirical outcomes. Furthermore, the Joint Tests will be applied to the intercept to see if there are other influencing factors beside the three-factor model that might induce the abnormal return observed in the stock market.
The results are as follows:
1. Regardless of what categories are chosen, either the 10 Industry Portfolios or the 25 Portfolios, the conclusions resulting from applying either of the two methodologies, Maximum Likelihood Estimates (MLE) and the Generalized Method of Moments (GMM), are all in concordance. All of the three causal factors, the market factor, the size factor (SMB), and the book-to-market factor (HML), all seem to have evident effects on the returns of stock market.
2. In regards to the US Stock Market, it was discovered, using both the MLE and the GMM methods to calculate the three-factor model, that the significant levels are all below 0.01. This shows that the system risk is integral in the explanation of return of stocks. The above conclusion coincides with the linear relationship between the traditional CAPM system risk and the market return. In other words, the US Stock Market exhibits the high-risk, high return and low-risk, low return phenomenon.
3. Empirical results also show that, with the book-to-market factor controlled, investors will receive higher excess return if they invest in the stocks of small scale companies as compared to the stocks of large scale corporations. This proves the effectiveness of the model as well as supports the size effect theory proposed by Banz (1981) .
4. The empirical results also show that, with the size factor controlled, lower book-to-market ratios result in lower rates of return and higher book-to-market values result in higher rates of return, showing that there is a positive relation between excess return and the book-to-market factor. In other words, there exists the book-to-market effect in the US stock market.
5. The values of the size factor (SMB) and the book-to-market factor (HML) present a simply diminishing phenomenon, which proves that Fama and French’s three-factor model match the empirical results.
6. The value of the size factor (SMB) has a clearer and more direct linear relationship with the stock market returns than the book-to-market factor (HML), meaning that the size effect is more evident than the book-to-marke effect in the US Stock Market.
7. The results of the Joint Tests disapprove the unrealistic assumption of setting the intercept to zero. Thus, one can conclude that there may be other influencing factors beside the three-factor model that causes the market to have abnormal returns.
|
|---|
