Tuesday, December 8, 2015
Rethinking Performance Evaluation
By Campbell R. Harvey, Duke University – Fuqua School of Business; and Yan Liu, Texas A&M University, Department of Finance
The current approach to performance evaluation is to run equation-by-equation regressions to calculate alphas. With this approach, three issues arise: 1) the estimation does not take into account any cross-sectional information; 2) there is no allowance for parameter uncertainty; and 3) the estimated alphas do a poor job of predicting future alphas. In “Rethinking Performance Evaluation,” the authors depart from the existing literature by proposing a ‘random effects’ counterpart of the current performance evaluation model.
Download the full article here.
From the December 2015 issue of Barclay's Insider Report. Accredited investors can subscribe to the full newsletter for free.
The current approach to performance evaluation is to run equation-by-equation regressions to calculate alphas. With this approach, three issues arise: 1) the estimation does not take into account any cross-sectional information; 2) there is no allowance for parameter uncertainty; and 3) the estimated alphas do a poor job of predicting future alphas. In “Rethinking Performance Evaluation,” the authors depart from the existing literature by proposing a ‘random effects’ counterpart of the current performance evaluation model.
Download the full article here.
From the December 2015 issue of Barclay's Insider Report. Accredited investors can subscribe to the full newsletter for free.
Labels: Barclay Insider Report, Barclay Insider Report Guest Article, Bayesian, EM algorithm, Fixed effects, hedge funds, Multiple testing, Mutual funds, Performance evaluation, Random effects, Regularization
Copyright © 2010 by Barclay Hedge
Subscribe to Posts [Atom]