CAPM (Capital Asset Pricing Model) with Stable Distribution

Abstract

In the classical finance theory, the CAPM models are developed using the Gaussian framework, that is, weassume the vector of returns can be modeled using the multivariate normal distribution. However, it is foundempirically that typically the financial data, especially the returns of assets, are leptokurtic (i.e., it is heavy tail andpeaked around the center). It has been shown in the literature that the stable distribution, where the normal is of aspecial case, becoming one of the popular model to model leptokurtic data. In this paper, we analyse the CAPMunder the assumption that the data follows the stable non-normal distribution with the index ofstability1 <α < 2 . We finally provide empirical application of the CAPM under the Gaussian and stable casesusing several returns data from Indonesian Stock Market.