RESEARCH

Eigenfactor adjusted covariance matrices

This paper investigates the underlying sources for the biases of optimised portfolios, and identifies special portfolios, termed eigenfactors, that exhibit large systematic biases in the risk forecasts. It shows that the covariance matrix can be adjusted to remove these biases, and that removing eigenfactor biases essentially removes the optimised portfolio biases as well.

To access the article click here

 

Eigenfactor_Adjusted_Covariance_Matrices