The Efficiency Analysis of the CAPM Model in the Investment Decision Making.

<p>The paper is devoted to the problem of the CAPM practical application efficiency, analysis of quality of the model&rsquo;s risk and return estimates.</p><p>In particular the objective of the research is to test the model&rsquo;s assumption of investment return distribution normality. Since the quality of the coefficient Beta, which is based on the classic bilateral variance, depends on this assumption, then adequacy of risk and return measures also depends on the feasibility of a normal distribution in practice.</p><p>Research Methodology is based on graphical and statistical analysis of the distribution of return, the various tests of distribution normality, such as Jarque-Bera test, criterions of Cramer-von Mises, Watson, Anderson-Darling.</p><p>Return, the distribution of which is analyzed in this paper, in turn, represents net income of trading models on the basis of indicator &quot;moving average&quot; and daily quotations of financial indices S &amp; P 500, Nikkei 225 and the MICEX on 08.05.2001 08.05.2014. The choice of return precisely from the trading model, unlike its calculation in absolute terms (the difference of the index price of the current moment from the its earlier values, as many researchers on the topic suggest) is made primarily due to the purpose of this study, namely, the practical application of the CAPM model. Since, in our opinion, any rational investor possesses a certain strategy.</p><p>The analysis revealed that the return is not distributed according to the normal law. The paper draws conclusions about the bias obtained by CAPM risk values ​​(via beta coefficient), not appropriate in this case use of the classical bilateral dispersion, which presupposes the symmetry of the distribution. It is speculated that perhaps it makes sense to use unilateral variance and consider the deviation from the expected value of return in one direction (not two as it is done in the CAPM), because the distribution showed a significant negative asymmetry. However, the work does not include any specific conclusions about the effectiveness of unilateral dispersion, it is brought only as a hypothesis and required further consideration in a separate study.</p>