Prediction of Systematic Risk: A Case from Turkey

Abstract This study compares Bayesian and time-varying models to adjust for the regression tendency of betas present in standard asset pricing applications. Beta adjustment techniques are applied to the Istanbul Stock Exchange (ISE) data. Empirical findings show that the Mean Square Error (MSE) is lowest among all models used in the study when log-linear or square-root linear Blume models are used and betas predicted according to Bayesian models, have lower MSE than unadjusted betas. Also, it is observed that the inefficiency part of the MSE changes most when various adjustment techniques are used. Arabic Abstract: