Identifying the underlying components of high-frequency data: Pure vs jump diffusion processes

Abstract

In this paper, we examine the finite sample properties of test statistics designed to identify distinct underlying components of high-frequency financial data, specifically the Brownian component and infinite vs. finite activity jumps. We conduct a comprehensive set of Monte Carlo simulations to evaluate the tests under various types of microstructure noise, price staleness, and different levels of jump activity. We apply these tests to a dataset comprising 100 individual S&P 500 constituents from diverse business sectors and the SPY (S&P 500 ETF) to empirically assess the relative magnitude of these components. Our findings strongly support the presence of both Brownian and jump components. Furthermore, we investigate the time-varying nature of rejection rates and we find that periods with more jumps days are usually associated with an increase in infinite jumps and a decrease in finite jumps. This suggests a dynamic interplay between jump components over time.