A1369
Title: On lead-lag estimation of non-synchronously observed point processes
Authors: Takaaki Shiotani - Graduate School of Mathematical Sciences, The University of Tokyo (Japan) [presenting]
Takaki Hayashi - Keio University (Japan)
Yuta Koike - University of Tokyo (Japan)
Abstract: A new theoretical framework for analyzing lead-lag relationships between high-frequency order arrivals is introduced. The problem is formulated in terms of the cross-pair correlation function of point processes, with the estimand defined as its peak location, providing a statistical interpretation of the seminal method. A more natural estimator based on kernel smoothing with an adaptive bandwidth selection procedure is proposed. The estimator achieves a near-optimal convergence rate for the peak estimation problem. The proposed method is validated through simulations and real-data analysis, demonstrating its empirical advantages.
