Title: Price endogeneity and volatility in the cryptocurrency market
Authors: Jan Sila - Univerzita Karlova (Czech Republic) [presenting]
Michael Mark - Ecole Polytechnique Federale Lausanne (Switzerland)
Abstract: The aim is to describe the endogenous dynamics of the emerging cryptocurrency markets by constructing the so-called reflexivity index. A univariate self-exciting Hawkes process with various kernels is fit to high-frequency Bitcoin transactions from BitMEX exchange. We then explore the predictive power of the reflexivity index for volatility forecasting. We conjure that this information will be able to improve forecasts made with the Heterogeneous Autoregressive model of Realized Volatility (HAR-RV model) as an external regressor. Thus, we create a novel link between market endogeneity and forecasting Realized volatility with an empirical application on cryptocurrency transaction dataset.