Integrated Covariance Estimation Using High-Frequency Data in the Presence of Noise
48 Pages Posted: 17 Jul 2006
Date Written: February 23, 2006
We analyze the effects of non-synchronicity and market microstructure noise on realized covariance type estimators. It is shown that non-synchronicity leads to severe biases, whenever synchronization methods that employ last-tick interpolation are used. We study a simple estimator which resolves that problem and is unbiased and consistent for the integrated covariance in the absence of noise. When noise is present, however, we show that this estimator is biased and suggest a simple bias correction procedure. Furthermore, a subsampling version of the estimator is proposed, which could improve its efficiency. Finally, a simulation experiment is carried out to illustrate the theoretical results.
Keywords: Integrated covariance, Epps effect, Non-synchronous trading, Market Microstructure
JEL Classification: C32, G00, G1
Suggested Citation: Suggested Citation