Title: The (ir)rationality of the 1/N heuristic

Authors: Guido Abate

Addresses: Department of Economics and Management, University of Brescia, C.da S. Chiara 50, 25122 Brescia (BS), Italy

Abstract: This paper provides an empirical evaluation of the rationality of the 1/N heuristic when applied to indexed investments linked to the most representative benchmarks of the US stock market. The 'rationality' of the investment is measured by the efficiency, both in the modern portfolio theory (MPT) and in capital asset pricing model (CAPM) frameworks, of the construction methodologies of each index. The empirical study employs four measures of efficiency: the small-sample Gibbons, Ross and Shanken test; the large-sample Wald test, implemented through a bootstrap simulation; the generalised method of moments (GMM) test, applied in a non-Gaussian framework through a block bootstrap simulation, and Kandel and Stambaugh's relative efficiency measure. Results provide strong evidence of the superior efficiency of equal weighting if compared to float- and capitalisation weighting. This also indicates that the 1/N heuristic can be regarded as rational behaviour for indexed investors.

Keywords: 1/N heuristic; diversification heuristic; indexed investments; equal weighting; portfolio efficiency measures; behavioural finance; rationality; modern portfolio theory; MPT; capital asset pricing model; CAPM; small sample; large sample; generalised method of moments; GMM; bootstrap simulation; relative efficiency; float weighting; capitalisation weighting.

DOI: 10.1504/IJBAF.2014.067623

International Journal of Behavioural Accounting and Finance, 2014 Vol.4 No.4, pp.305 - 324

Received: 12 Dec 2013
Accepted: 20 Oct 2014

Published online: 22 Feb 2015 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article