A1054
Title: Statistical analysis of a stochastic boundary model for high-frequency data from a limit order book
Authors: Markus Bibinger - University of Wurzburg (Germany) [presenting]
Abstract: Statistical methods to infer characteristics of a semimartingale efficient log-price process are presented. Observations are from a boundary model with one-sided microstructure noise for high-frequency prices of limit orders. The previous one-dimensional model is generalized to a multivariate model, and the focus is on the estimation of the covolatility matrix. Asymptotic results are established in a high-frequency regime. Convergence rates are shown to be faster than under standard market microstructure noise. They hinge on a tail index of the noise distribution. The estimation of this noise tail index and adaptive inference on the semimartingale is addressed.
