Sample Selection Bias, Return Moments, and the Performance of Optimal versus Naive Diversification
53 Pages Posted: 18 Aug 2016 Last revised: 10 Nov 2016
Date Written: September 18, 2016
Abstract
I examine the sample selection bias in portfolio horse race. Numerous studies propose mean-variance portfolio rules to outperform the naive 1/N portfolio rule. However, the outperformance is often justified by a small number of pre-selected datasets. Using a new performance test based on a large number of datasets, I compare thirteen "1/N outperformers" with the naive rule. Results show that not only a majority of the thirteen "1/N outperformers" on average underperform the 1/N rule significantly, but also none of these mean-variance rules significantly outperform the naive benchmark in more than 10% of the datasets. To further understand portfolio performance, this paper explores the theoretical relations between assets' return moments and the performance of optimal versus naive diversification. These relations not only imply strong performance predictability, but also can be exploited to deliver out-of-sample portfolio benefits.
Keywords: sample selection bias, mean-variance portfolio, performance test, predictability
JEL Classification: G11, G12
Suggested Citation: Suggested Citation