A Hawkes process model with a time-varying background rate is developed for analyzing the high-frequency financial data. In our model, the logarithm of the background rate is modeled by a linear model with variable-width basis functions, and the parameters are estimated by a Bayesian method. We find that the data are explained significantly better by our model as compared to the Hawkes model with a stationary background rate, which is commonly used in the field of quantitative finance. Our model can capture not only the slow time-variation, such as in the intraday seasonality, but also the rapid one, which follows a macroeconomic news announcement. We also demonstrate that the level of the endogeneity of markets, quantified by the branching ratio of the Hawkes process, is overestimated if the time-variation is not considered.
↧