САРМ MODIFICATIONS AND THEIR VERIFICATION

Abstract

This study aims to present the results from the testing of five CAPM modifications through two different approaches using data from the Bulgarian Stock Exchange. The first model uses a regression analysis to look for and measure the relation between historical stock returns and the beta-coefficient of companies – a separate regression is started for each of the beta-coefficient modifications, the determination coefficient of which can be compared to the determination coefficient of the regressions of the rest of modifications. The second approach is backtesting. In order to overcome the usual data incompleteness of the emerging markets an easily applicable approach is proposed for comparing the validity of different beta-coefficient modifications and hence of CAPM.