Heterogenous market hypothesis evaluation using multipower variation volatility

High frequency trading activities is one of the common phenomena in nowadays financial markets. Enormous amounts of high frequency trading data are generated by huge numbers of market participants in every trading day. The availability of these information allow researchers to further examine the statistical properties of informationally efficient market hypothesis (EMH). Heterogeneous market hypothesis (HMH) is one of the important extensions of EMH literature. HMH introduced nonlinear trading behaviors of heterogeneous market participants instead of normality assumption under the EMH homogeneous market participants. In this study, we attempt to explore more high frequency volatility estimators in the HMH examination. These include the bipower, tripower and quadpower variation integrated volatility estimates using Heterogeneous AutoRegressive (HAR) models. The empirical findings show that these alternatives multipower variation (MPV) estimators provide better estimation and out-of-sample forecast evaluations as compared to the standard realized volatility. In other words, the usage of MPV estimators is able to better explain the HMH statistically. At last, a market risk determination is illustrated using value-at-risk approach