Ledoit-Wolf Shrinkage

A method that smooths a noisy covariance matrix toward a structured target, producing more reliable optimizer inputs.

Category: Methodology

What is Ledoit-Wolf Shrinkage?

Sample covariance matrices are noisy when the number of assets is large relative to history. Ledoit-Wolf shrinks the sample matrix toward a constant-correlation or identity target with a data-driven blend coefficient.

Related terms

Back to the GNG Research investing glossary