JE Griffin and RCA Oomen
Covariance Measurement in the Presence of Non-Synchronous Trading and Market Microstructure Noise
Date: June 2006
Abstract: This paper studies the problem of covariance estimation when price observations are subject to nonsynchronicity and contaminated by i.i.d. microstructure noise. We derive closed form expressions for the bias and variance of three popular covariance estimators, namely realized covariance, realized covariance plus lead- and lag-adjustments, and the Hayashi and Yoshida estimator, and present a comprehensive investigation into their properties and relative efficiency. The key finding of this paper is that the ordering of covariance estimators in terms of efficiency depends crucially on the level of microstructure noise. In fact, for sufficiently high levels of noise, the standard realized covariance estimator (without any corrections for non-synchronous trading) can be most efficient. An empirical illustration using TAQ quote and transaction data confirms the validity of our methodology and points to some avenues for future research.