Informational Frictions and Commodity Markets
MICHAEL SOCKIN and WEI XIONG∗
ABSTRACT
This paper develops a model with a tractable log-linear equilibrium to analyze the
effects of informational frictions in commodity markets. By aggregating dispersed
information about the strength of the global economy among goods producers whose
production has complementarity, commodity prices serve as price signals to guide
producers’ production decisions and commodity demand. Our model highlights important
feedback effects of informational noise originating from supply shocks and
futures market trading on commodity demand and spot prices. Our analysis illustrates
the weakness common in empirical studies on commodity markets of assuming
that different types of shocks are publicly observable to market participants.
IN THE AFTERMATH OF THE DRAMATIC BOOM and bust cycle of commodity prices
in 2007 to 2008, there has been renewed interest among academics and policy
makers regarding the drivers of commodity price fluctuations, in particular,
whether fundamental demand and supply shocks are sufficient to explain the
observed price cycles and whether speculation in commodity futures markets
exacerbated these cycles are subjects of debate. In this debate, it is common for
academic and policy studies to treat different types of shocks (such as supply,
demand, and financial market shocks) as observable to market participants.1
In doing so, however, these studies ignore a key aspect of commodity markets,
namely, severe informational frictions faced by market participants. The
markets for major commodities, such as crude oil and copper, have become
globalized in recent decades, with supply and demand now stemming from
across the world. This globalization exposes market participants to heightened
informational frictions regarding the global supply, demand, and inventory of
these commodities.
The economics literature has developed an elegant theoretical framework to
analyze how trading in centralized asset markets both facilitates information
∗Sockin is with University of Texas at Austia and Xiong is with Princeton University and NBER.
We wish to thank Thierry Foucault; Lutz Kilian; Jennifer La’O; Matteo Maggiori; Joel Peress; Ken
Singleton; Kathy Yuan; and seminar participants at Asian Meeting of Econometric Society, Bank
of Canada, Chicago, Columbia, Emory, HEC-Paris, INSEAD, NBER Meeting on Economics of
Commodity Markets, Princeton, North America Meeting of Econometric Society, the 6th Annual
Conference of the Paul Woolley Centre of London School of Economics, and Western Finance
Association Meetings for helpful discussion and comments. We are especially grateful to Bruno
Biais, an Associate Editor, and three referees for numerous constructive comments and suggestions.
Xiong acknowledges financial support from Smith Richardson Foundation Grant #2011-8691.
1 See a recent review by Cheng and Xiong (2014).
DOI: 10.1111/jofi.12261
2063