Covariance measurement in the presence of non-synchronous trading and market microstructure noise
This paper studies the problem of covariance estimation when prices are observed non-synchronously and contaminated by i.i.d. microstructure noise. We derive closed form expressions for the bias and variance of three popular covariance estimators, namely realised covariance, realised covariance plus lead and lag adjustments, and the Hayashi and Yoshida estimator, and present a comprehensive investigation into their properties and relative efficiency. Our main finding is that the ordering of the covariance estimators in terms of efficiency crucially depends on the level of microstructure noise, as well as the level of correlation. In fact, for sufficiently high levels of noise, the standard realised covariance estimator (without any corrections for non-synchronous trading) can be most efficient. We also propose a sparse sampling implementation of the Hayashi and Yoshida estimator, study the robustness of our findings using simulations with stochastic volatility and correlation, and highlight some important practical considerations.
Year of publication: |
2011
|
---|---|
Authors: | Griffin, Jim E. ; Oomen, Roel C.A. |
Published in: |
Journal of Econometrics. - Elsevier, ISSN 0304-4076. - Vol. 160.2011, 1, p. 58-68
|
Publisher: |
Elsevier |
Saved in:
Saved in favorites
Similar items by person
-
Covariance measurement in the presence of non-synchronous trading and market microstructure noise
Griffin, Jim E., (2011)
-
A blocking and regularization approach to high dimensional realized covariance estimation
Hautsch, Nikolaus, (2009)
-
Properties of Realized Variance Under Alternative Sampling Schemes
Oomen, Roel C.A., (2006)
- More ...