Advances in HFT on Russian Stock Market: Effects on Trading Activity and Prices Volatility

Market microstructure approach becomes more and more relevant due to the rapid changing nature of modern markets and advance in high-frequency trading (HFT), which consequences of spreading remain unclear. Theoretically, some productivity gains can be obtained, but in practice algorithmic traders increase the possibility of extreme market moves. Understanding the relationships between core market variables such as asset prices and volatility, trading activity in the context of particular market architecture is a key to analyzing market functioning and its efficiency. So in our paper we focus on modeling price volatility and market activity variables relationships over a long period of time in order to catch the effects of advances in HFT on Russian stock market behavior. We suppose that HFT extension may have a positive influence on the market quality and improve the efficiency under a proper regulation. We build a combination of the vector autoregression (VAR) with the EGARCH model, and fit it to data of 9 shares traded on MICEX for ten years by 3 month sub-samples. A set of simple ARMAX models is also estimated on overlapping 3 years sub-samples. Raw transaction data has been aggregated at two scales: 15-min and hour intervals. We obtained mixed results. ARMAX class of models demonstrates that in time key market indicators can be predicted by each other in less degree that evidences in favor of market microstructure theoretical assumptions and supports the idea of growing market efficiency under spreading of HFT. However, the VAR model does not allow providing any meaningful conclusions about the dynamics of relationships between market variables. So we call for more research in the area in order to comprehend decisively if effects of HFT are really present on Russian stock market and how exactly they are reflected at market indicators and overall market quality.