摘要：In this paper, we consider the problem of estimating the systematic risk of stocks market by using a modeling formulation based on scale normal mixtures comparative calibration models. In this work, we emphasize the Student-t comparative calibration model, which is approached by considering the degrees of freedom parameter unknown. Inference is approached by using the EM algorithm and MCMC methodology. The results are applied to the stock returns of two Chilean companies.ons.
关键词：Bayesian inference Capital asset pricing model Maximum likelihood;estimation Scale normal mixture Structural comparative calibration model; Student-t distribution..
