Speaker
Zdzisław Burda
(AGH University of Science and Technology)
Description
The problem of estimating covariance matrix plays a fundamental role in portfolio selection. Recently a new estimator of large-dimensional covariance matrices has been proposed to reduce out-of-sample risk of large portfolios. The estimator is called non-linear shrinkage estimators. We derive an analytic formula for the non-linear shrinkage estimator of large dimensional covariance for correlated samples.
Primary author
Zdzisław Burda
(AGH University of Science and Technology)
