Heteroscedasticity and other misspecification issues in the market model

Abstract

In this paper the authors attempt to analyse the validity of the traditional market model in the complex environment of emerging and bullish markets where heteroscedasticity and other misspecification issues are detected through application of a wide range of appropriate empirical tests and tools. The issue which the authors investigate is generally related to excess volatility caused either by market over-reaction or by changes in the market regime accompanied by changes in the systematic risk (beta) or by short-period non-informational ‘noise’ accompanied by violations of the classical assumptions on the residuals of the market model. The results of this work highlight and reveal several sides of relatively unknown and empirically confused areas regarding market efficiency and appropriateness and adequacy of the market model, as well as certain unusual features of investor behaviour.