| Summary: | 碩士 === 國立臺北大學 === 企業管理學系 === 89 === The importance of P-E ratios in the decision of portfolio selection has been highlighted again and again in Taiwan and many other countries. Many empirical studies demonstrate that there are P-E effects in the American and Japanese stock markets. In Taiwan, some empirical studies support that there are P-E effects, but some are not.
However, what determines P-E ratios is still a puzzle. The first purpose of the study reported here was to determine the P-E ratios and E-P ratios of Taiwan stock market that can be justified by macroeconomic fundamentals at any given time. When price to earnings, changeable rate of price to earnings, earnings to price or changeable rate of earnings to price was used as the dependent variable, several different kind of regression models were found to be significant. The second purpose is that which regression model is better than the others. The final main aim is to see whether some macroeconomic variables listed here have the phenomenon of time lag.
The summary of my empirical outcome is listed below:
1.For E-P as the dependent variable, using auto-regression model has higher R-Square of 0.8908.
2.In order to solve the problem of multicollinearity, factor analysis was used here. For many reasons, not all of macroeconomic variables were adopted. In the final, just leading index and price the two main factors were included.
3.Judged by AIC, the best number of time lag of leading index is two.
|
|---|
