| dc.description.abstract | The purpose of this study was to investigate the performance of the Arbitrage Pricing Theory (APT) in the Nairobi Securities Exchange (NSE) on a monthly basis, for the period January 2001 to December 2010.The study examined six pre-specified variables which are: Inflation Rate, Treasury Bill Rate, Exchange Rate, Price Earnings Ratio, NSE 20- Share Index and Dividend Yield. Some of the variables used were the same as those used by Chen, Roll and Roll for their study on the US stock market. In the study however, the researcher developed two more variables namely Price Earnings Ratio and Dividend Yield, which have a relation with the stock return. Using regression analysis and the analysis of variance, the researcher observed that including market based financial measures in the model add explanatory power to the APT model. The serial correlation problem was discussed using Durbin-Watson statistics. The test results confirmed that the portfolio constructed had no serial correlation. Our results show that the market portfolio reacts differently when the market based financial measures are introduced or dropped in the model. This can be seen through the variations in R – Squared, R –Predicted as well as the Durbin Watson statistic. The analysis appears to be the first empirical test of APT that includes market based financial measures in the model in a study of the NSE. The finding therefore indicates the importance of conducting more studies on APT by including the performance indicators of companies listed at the NSE. | en |
