Estimating Unbiased and Precise Realized Covariances

35 Pages Posted: 9 Jun 2004

See all articles by Martin Martens

Martin Martens

Erasmus University Rotterdam (EUR); Robeco Asset Management

Date Written: June 11, 2004


Existing empirical work on realized covariances partitions the trading day in 5- or 30-minute intervals and then computes the sample covariance based on the resulting 5- or 30-minute returns. Such frequencies are heuristically or subjectively chosen to minimize the bias and avoid market microstructure effects. However, it is also important to consider the preciseness of the covariance estimator. This study compares different methods. Both when only non-trading matters (using midpoints of quotes) and when also bid-ask bounce matters (using transaction prices) the bias-adjusted sample covariance of returns (averaging over all possible subsamples) computed from interpolated prices around equidistant time marks performs well based on the root mean squared error. The optimal choice substantially lowers the RMSE compared to existing heuristic choices.

Keywords: Realized volatility, covariance, beta, subsampling

Suggested Citation

Martens, Martin P.E., Estimating Unbiased and Precise Realized Covariances (June 11, 2004). EFA 2004 Maastricht Meetings Paper No. 4299. Available at SSRN: or

Martin P.E. Martens (Contact Author)

Erasmus University Rotterdam (EUR) ( email )

P.O. Box 1738
3000 DR Rotterdam
+31 10 408 1253 (Phone)
+31 10 408 9162 (Fax)

Robeco Asset Management ( email )

Rotterdam, 3011 AG

Here is the Coronavirus
related research on SSRN

Paper statistics

Abstract Views
PlumX Metrics