Dynamic Correlation Structure; Securities Risk and Return

Modelling dynamic nature of covariance of assets return almost always is a challenging area of finance. Econometrics models just pay attention to variance behavior longitudinally, however, core of numerous finance models need the analysis of the total covariance structure of returns.

Among the first models that analyze covariance behavior are multivariate GARCH models which were criticized for the need to estimate a large number of parameters. This paper is aimed to investigate the effect of stock return dynamic correlation structure on systematic risk, idiosyncratic risk and average stock return. To this end, a sample of 148 listed companies in Tehran Stock Exchange is examined during 2003 to 2014. GARCH framework is used for testing this claim.

According to the results, securities that were highly correlated with market wide risk factors in the past are likely to have low systematic risk, idiosyncratic risk and average return at present. It can be expected there is significant relationship between idiosyncrstic risk and correlation for lower turnover stock (information transparency proxy) although there is no relationship for smaller firms.