CAPM Test In Indonesian Stock Market Using Mean-Variance Optimal Portfolio As Market Return Proxy

Abstract

This study examines the validity of the market return proxy used in the standard capital asset pricing model (CAPM) test. CAPM describes the relationship between the risk and the expected return of assets and is commonly used to estimate the cost of capital and measure the performance of a managed portfolio. Unfortunately, the value-weighted Jakarta Composite Index (JCI) used in standard CAPM tests as a market proxy fails to satisfy the assumption of mean-variance efficiency. Using mean-variance portfolio optimization of Kompas 100, LQ45, and IDX30 index components to generate an optimal portfolio to be used as a market proxy in CAPM, this study shows that “optimal beta” has less error in the expected return prediction than “market beta”. Furthermore, the presence of bias in the valueweighted market index is also analyzed. The findings of this study imply that the mean-variance efficient portfolio from the portfolio optimization process should replace the value-weighted market index as a market proxy for CAPM’s beta estimation.

Keywords: CAPM, portfolio optimization, market proxy