Modeling of Returns and Trading Volume by Regime Switching Copulas

Abstract

The structure of links between realized volatility and trading volume can be reflected by regime switching copulas. The estimation by means of copula based regime switching models delivered results concerning the interdependencies between realized return volatility and trading volume of selected companies listed in ATX. A copula in the first regime was chosen as an asymmetric copula with positive lower and upper tail dependencies. Conversely, Gaussian copula in the second regime is a symmetric copula and variables linked with it are tail independent. For all analyzed stocks the probability of being at the first regime appeared to be vitally greater than being at the second regime. This result suggest that there is considerable dependence between realized volatility and daily volume in extreme values. The results suggest that interdependencies between realized volatility and trading volume do not probably depend on the size but rather on the branch of a company.