Short-term and long-term dependencies of the S&P 500 index and commodity prices

Abstract

We utilize wavelet coherency methodology with simulated confidence bounds to examine the short-term and long-term dependencies of the returns for S&P 500 and the S&P GSCI^® commodity index. Our results indicate no evidence of co-movement between S&P 500 total return and the S&P GSCI^® commodity index total return in the short term, thereby suggesting diversification gains for equity investors. Importantly, this finding encompasses the onset of the current financial crisis. However, long-term diversification benefits, particularly after the onset of the recent financial crisis, are limited. We find, moreover, no consistent evidence of co-movements between S&P 500 and 10 individual sub-indexes of the S&P GSCI^® commodity index. Of particular importance, we report weak co-movement of returns between S&P 500 and S&P GSCI^® Precious Metals total return and S&P 500 and S&P GSCI^® Softs at all frequencies, implying significant diversification gains both for short-term and long-term investors.

Keywords:

• Commodity markets
• Comovement
• Applied finance
• Correlation modelling

JEL Classification:

• C4
• E3
• E32
• F3
• F30
• G1
• G15

Acknowledgements

We would like to thank the anonymous reviewers and the editor of the journal for their insightful comments. The financial support of the Academy of Finland (project 132913) is also gratefully acknowledged.

Notes

¹Financial crisis interferes with the resource allocation process that moves funds to agents with the most productive investment opportunities in the financial system.

²See, for example, Dwyer and Tkac (2009), Bartram and Bodnar (2009), Claessens and Terrones (2010), Aloui et al. (2011) and Kenourgios and Padhi (2012).

³This paper deals with dependence in time of crisis. Other studies have concentrated on extreme co-movement in international financial markets. See Longin and Solnik (2001), Ang and Chen (2002), Patten (2004), Garcia and Tsafack (2009), Adel and Salma (2012) and references therein for a discussion of this literature.

⁴The value of commodities is best achieved as part of a diversified portfolio and not as a stand-alone investment. Historically, commodities have shown inferior returns and greater volatility when compared with other assets (Erb and Harvey 2006). However, when combined with conventional investment assets (e.g., stocks and bonds), they can enhance portfolio performance. See also, for instance, the article of Mark Huamani on investing in commodities and diversification: http://www.jpmorgan.com/tss/General/Investing_in_Commodities/1159337364479 (accessed 1 September 2011).

⁵ http://www.bloomberg.com/news/2011-01-07/s-p-gsci-says-commodity-assets-against-index-rose-as-high-as-100-billion.html (accessed 1 September 2011).

⁶See, for instance, http://www.ft.com/intl/cms/s/1/23fadee4-4803-11db-a42e-0000779e2340.html#axzz1YDqkOknx (an article by John Authers, Financial Times, September 22, 2006) (accessed 8 November 2010).

⁷Utilizing wavelet coherency methodology is not novel, however. The methodology has been applied to analyse financial time-series (see, e.g., Rua and Nunes (2009), Nikkinen et al. (2011), Graham and Nikkinen 2011) and Graham et al. (2012)).

⁸We acknowledge the existence of alternatives, e.g. public futures funds, commodity trading advisors, and private commodity pool operators, through which investors can invest in commodities (Edwards and Liew 1999). However, using the S&P GSCI^® is advantageous because it allows for a clean examination without the contamination of superior/inferior trading techniques of other active public futures funds. Our choice of index also avoids biases associated with databases that report futures returns. Akey (2005) provides an excellent review of available passive commodity indexes.

⁹See the following link for further details on the components, weights, and construction of the S&P GSCI^® index: http://www.standardandpoors.com/servlet/BlobServer?blobheadername3=MDT-Type&blobcol=urldata&blobtable=MungoBlobs&blobheadervalue2=inline%3B+filename%3DSP_GSCI_FAQ_Web.pdf&blobheadername2=Content-Disposition&blobheadervalue1=application%2Fpdf&blobkey=id&blobheadername1=content-type&blobwhere=1243748755917&blobheadervalue3=UTF-8.

¹⁰To reflect the performance of a total return investment in commodities, four separate but related indexes have been developed based on the S&P GSCI: (1) the S&P GSCI Spot Index, which is based on the price levels of the contracts included in the S&P GSCI; (2) the S&P GSCI Excess Return Index (S&P GSCI ER), which incorporates the returns of the S&P GSCI Spot Index as well as the discount or premium obtained by ‘rolling’ hypothetical positions in such contracts forward as they approach delivery; (3) the S&P GSCI Total Return Index (S&P GSCI TR), which incorporates the returns of the S&P GSCI ER and interest earned on hypothetical fully collateralized contract positions on the commodities included in the S&P GSCI; and (4) the S&P GSCI Futures Price Index (S&P GSFPI), which is intended to serve as a benchmark for the fair value of the futures contracts on the S&P GSCI traded on the Chicago Mercantile Exchange (CME).