The Jury is Still Out On the Performance of Naive Diversification (1/N rule)

67 Pages Posted: 22 Jul 2020 Last revised: 5 Oct 2021

See all articles by Redouane Elkamhi

Redouane Elkamhi

University of Toronto - Rotman School of Management

Chanik Jo

CUHK Business School

Jacky S.H. Lee

Healthcare of Ontario Pension Plan Trust Fund

Marco Salerno

University of Toronto - Rotman School of Management

Date Written: June 30, 2020

Abstract

DeMiguel et. al. (2009b) made a compelling case that estimation error dwarfs diversification benefits resulting in naive diversification (1/N) dominating mean-variance portfolios. We illustrate the necessary and sufficient conditions for risk-based allocation rules to be optimal in a mean-variance framework. We show empirically that many common datasets satisfy such conditions, making these rules preferred to mean-variance in the presence of estimation error. Our out-of-sample tests show that these rules outperform both mean-variance and 1/N. Further, we show that clustering the data using machine learning enhances the diversification benefits of these rules by making the data closer to the required conditions for optimality.

Keywords: Portfolio Choice, Asset Allocation, Machine Learning, Clustering

JEL Classification: G11

Suggested Citation

Elkamhi, Redouane and Jo, Chanik and Lee, Jacky and Salerno, Marco, The Jury is Still Out On the Performance of Naive Diversification (1/N rule) (June 30, 2020). Rotman School of Management Working Paper No. 3638713, Available at SSRN: https://ssrn.com/abstract=3638713 or http://dx.doi.org/10.2139/ssrn.3638713

Redouane Elkamhi

University of Toronto - Rotman School of Management ( email )

105 St. George Street
Toronto, Ontario M5S 3E6 M5S1S4
Canada

Chanik Jo

CUHK Business School ( email )

Cheng Yu Tung Building
12 Chak Cheung Street
Shatin, N.T.
Hong Kong

HOME PAGE: http://sites.google.com/view/chanik-jo/

Jacky Lee

Healthcare of Ontario Pension Plan Trust Fund ( email )

1 York Street
Toronto, Ontario M5J0B6
Canada

Marco Salerno (Contact Author)

University of Toronto - Rotman School of Management ( email )

Toronto, Ontario M5S 3G8
Canada

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
184
Abstract Views
914
rank
208,590
PlumX Metrics