A0774
Title: Asymmetric information, information life span, and lead-lag effects on a multi-asset lagged adjustment price dynamics
Authors: Hai Dang NGO - Audencia Business School (France)
Iordanis Kalaitzoglou - Audencia Business School (France) [presenting]
Emilios GALARIOTIS - Audencia Business School (France)
Abstract: The classic martingale-plus-noise model is extended for high-frequency price formation dynamics to accommodate pricing errors and a lagged price adjustment mechanism. The generalized price formation dynamics are motivated by (i) asymmetric information is an inherent part of the market microstructure, triggering mispricing of true underlying values; (ii) residual information has a life span, decaying over time before fully incorporated into the price process. We identify the pricing of asymmetric information and residual information as separable sources driving the price formation process. The gradual discovery of the latter gives rise to temporal lead-lag effects, which, due to Epps effects, vanish in contemporaneous time of asynchronous trading. Econometric inference on the model is built on the bridge between microstructure models and Hawkes point process theory, and enables the estimation of price lagged adjustment to be computed at every point in time. This property allows for (i) statistical measures for the efficiency of the financial market; (ii) separable estimation of lead and lag correlations at different time scales; and (iii) an efficient estimator of locally integrated covariance over any time interval. Our empirical application to a set of selected DJIA stocks supports these theoretical postulations.
