Authors: Eleni Thanou, George Polychronopoulos, Dimitris Kallivokas

Title: The CAPM revisited: is beta a useful tool in portfolio formulation? New evidence from the Athens Stock Exchange

Abstract

The purpose of this paper is to add to the extensive body of research on CAPM by applying the latest methodologies to individual stocks of the ASE and for a period with intense fluctuations.We follow an approach similar to that of Pettengill et al (1995) and  Theriou et al (2010). The conditional version of the CAPM, applied to portfolio formation acknowledges that in positive market excess return periods (up market), the relationship between beta and excess returns should be positive, while in negative market excess returns (down market) the opposite must be true.

Our sample, constructed from daily closing prices of ASE stocks between 2004-2014 aggregated in monthly averages, is divided into three successive sub periods. The first is used for the calculation of the betas of individual stocks and the formation of portfolios, based on the betas. The following period is used to estimate each portfolio's beta and then, the third period is used for the testing of hypothesis, using both the portfolios and individual stocks. First, the unconditional CAPM hypothesis is tested and as expected,  rejected, then we test the conditional hypothesis.  The results partially support the conditional relationship between beta and excess returns, supporting previous research.