cov_estim_evc_bp: Eigenvalue Clipping Covariance Estimation (Bouchaud-Potters)

In antshi/CovEstim: (High-dimensional) Covariance Estimation

Eigenvalue Clipping Covariance Estimation (Bouchaud-Potters)

Description

Computes the Eigenvalue Clipping (EVC) estimator of the covariance matrix with the Bouchaud-Potters (BP) technique.

Usage

cov_estim_evc_bp(data, cut_edge)

Arguments

data	an nxp data matrix
cut_edge	a double, indicating the proportion for the applied eigenvalue clipping.

Details

The eigenvalue clipping covariance matrix estimator is computed with the following formula:

\hat{\Sigma}=\Delta\hat{\Lambda}\Delta',

where \Delta is the matrix with the sample eigenvectors of the data matrix and \hat{\Lambda} is a diagonal matrix with the clipped sample eigenvalues. The clipping procedure follows \insertCitebouchaudpotters2009;textualcovestim. In particular, the user-defined cutting edge s gives the proportion for the applied eigenvalue clipping so that the (1-s)\times p largest eigenvalues are kept and the remaining s\times p eigenvalues are substituted by their average.

Value

a list with the following entries

• a pxp estimated covariance matrix.

• an estimation specific tuning parameter, here the user-supplied cut edge.

References

\insertAllCited

Examples

data(rets_m)
sigma_evc_bp <- cov_estim_evc_bp(rets_m, cut_edge = 0.3)[[1]]

antshi/CovEstim documentation built on June 10, 2025, 3:11 a.m.