Efficiency and unbiasedness of corn futures markets: new evidence across the financial crisis

Abstract

Recent years witnessed commodity prices increases which have fostered research works on their predictability and a renewed interest of practitioners and policy-makers. The objective of this article is to test the predictive ability of futures prices on the underlying spot prices by taking corn, which is one of the most important agricultural commodities in terms of trading volumes and for its role in the dietary regime of many countries. We consider the corn futures on the Chicago Board of Trade (CBOT) in the period May 1998 to December 2011 so as to extend previous studies on this market and to assess a possible effect of the financial crisis. Our results do not emphasize a role for the latter and, although we do not find evidence of efficiency and unbiasedness, the futures corn price turns out to be the best predictor of the spot price if compared with most used alternatives.

Keywords:

• futures prices
• corn futures
• efficiency
• unbiasedness

JEL Classification:

• C53
• G13
• Q14

Acknowledgements

We thank Barbara Cardelli for excellent research assistance, Andrea Cipollini for useful discussions and an anonymous referee for helpful comments.

Funding

The authors kindly acknowledge financial support from the Italian University Ministry.

Notes

¹  A parallel renewed interest on commodity futures pricing is also present in the literature (e.g. Realdon, 2012).

²  The CBOT merged with the Chicago Mercantile Exchange (CME) in 2007.

³  In fact the theory shows that S_T = F_T and commodity futures have a 15-day time span for delivery.

⁴  In all the analyses we take prices in logarithm (multiplied by 100). Preliminary to the cointegration test, we have verified by means of the ADF test that the two series are both integrated of order 1. Results are available upon request. For a discussion of stationarity of commodity prices, see Wang and Tomek (2007).

⁵  In order to choose for the optimal number of lag, we have implemented the Akaike Information Criterion (AIC) and obtained 2 as the optimal lag (results are available upon request). As for a discussion of the constant issue in the cointegration test, see Johansen (1991).

⁶  Preliminary to the cointegration analysis, we have performed stationarity analysis obtaining the same results (available upon request) as for the whole sample.