What Every Investor Should Know About Commodities, Part Ii: Multivariate Return Analysis

Alternative Investment Research Centre Working Paper No. 33

35 Pages Posted: 14 Jun 2006 Last revised: 20 Oct 2016

Harry M. Kat

Independent

Roel C. A. Oomen

Deutsche Bank AG (London); London School of Economics & Political Science (LSE) - Department of Statistics

Multiple version iconThere are 2 versions of this paper

Date Written: June 1, 2006

Abstract

In this paper we study the multivariate return properties of a large variety of commodity futures. We find that between commodity groupings (such as metals, energy, etc.) correlations are very low and mostly insignificant whereas within groups they tend to be much stronger. In addition, commodity futures are roughly uncorrelated with stocks and bonds. Still, correlations may vary somewhat over the different phases of the business cycle, suggesting that not all commodities make equally good diversifiers at all times. Copula-based tests do not indicate any deviant behaviour in the tails of the joint return distribution of commodity futures and stocks or bonds. Contrary to equities and bonds, we show that commodity futures returns are positively correlated with unexpected inflation (i.e. 25% on average with CPI inflation as opposed to -30% for equities and -50% for bonds). There are significant differences between the various commodities, however, with energy, metals, cattle, and sugar offering the best hedging potential. Altogether, assuming that the observed regularities will persist, our results confirm that a well-balanced commodity futures portfolio could offer a worthwhile diversification service to the typical traditional investment portfolio.

Keywords: Commodities, commodity futures, correlation, tail-dependence, SJC copula, inflation

JEL Classification: G11, E44, O13, Q19, Q49

Suggested Citation

Kat, Harry M. and Oomen, Roel C. A., What Every Investor Should Know About Commodities, Part Ii: Multivariate Return Analysis (June 1, 2006). Alternative Investment Research Centre Working Paper No. 33. Available at SSRN: https://ssrn.com/abstract=908609 or http://dx.doi.org/10.2139/ssrn.908609

Harry M. Kat

Independent

No Address Available

Roel C.A. Oomen (Contact Author)

Deutsche Bank AG (London) ( email )

Winchester House
1 Great Winchester Street
London, EC2N 2DB
United Kingdom

London School of Economics & Political Science (LSE) - Department of Statistics ( email )

Houghton Street
London, England WC2A 2AE
United Kingdom

Paper statistics

Downloads
3,992
Rank
1,610
Abstract Views
11,335