A non-parametric test of the conditional CAPM for the Mexican economy

Abstract

Many models have been suggested to describe how investors manage risk and value risky cash flows. Among them, the most widely used is the Sharpe-Lintner-Black Capital Asset Pricing Model (CAPM).However, many anomalies and evidence against this version have been presented. To assume that the CAPM holds in a conditional sense is to assume that the betas and the market risk premium vary along time. We present a test of the conditional version of the CAPM for the Mexican economy, that uses a non-parametric methodology suggested by Wang that avoids the problem of functional misspecification of the betas of the assets.