A0208
Title: Inference on volatility using first exit time
Authors: Wenhao Cui - Beihang University (China) [presenting]
Abstract: A valid uniform inference theory is established for the spot variance process based on the first exit time. By leveraging the concept of the first exit time, the proposed method achieves substantial efficiency gains by incorporating additional information from the observation times. Furthermore, the developed inference procedure is valid for both fixed-k and large-k scenarios across the entire spot variance process. The proposed uniform inferential theory can also accommodate the presence of discretization errors and market microstructure noise, significantly enhancing the empirical applicability of the method. An empirically calibrated simulation study demonstrates the practical reliability of methods in the presence of pricing errors. Through the simulation, it is found that the method delivers confidence bands with satisfactory coverage frequency while also exhibiting narrower widths. Additionally, an empirical application using high-frequency data illustrates the efficiency gains achieved by employing this approach.
