Title: Variance estimation in the presence of self-excited jumps
Authors: Simon Kwok - University of Sydney (Australia) [presenting]
Abstract: Robust inference on the volatility process has become an important topic in high-frequency financial econometrics. There are methods for disentangling the variance component from jumps in jump diffusion models, but not all of them are robust to more realistic forms of jump dynamics. For example, bipower variation tends to overestimate the integrated variance when jumps are self-excited. The robustness of different estimators of integrated variance is studied and compared. It is found that the nonparametric threshold method delivers a more precise estimate than the multipower variation approach in the presence of autocorrelated jumps.