activity are widely studied by academics
and articulated by policymakers (Bernanke, 2008), the
factors driving commodity price fluctuations have proved to be more difficult to uncover. However,
understanding the behavior of commodity prices is of key importance in designing economic policies
that limit their impact on real economic activity and inflation. Factors that have been identified to drive
the dynamics of spot commodity prices include convenience yields (Gibson & Schwartz, 1990;
Gospodinov & Ng, 2013; Schwartz & Smith, 2000), exchange rates (Chen, Rogoff & Rossi, 2010) and
interest rates (Frankel, 2008; Schwartz, 1997).
In this paper, we explore the effect of monetary policy uncertainty on individual commodity prices
as well as on commodity price indexes across time and in the cross-section. In view of the increasing
financialization of commodities (Cheng & Xiong, 2014; Gorton & Rouwenhorst, 2006; Tang & Xiong,
2012), we argue and provide empirical evidence that Monetary Policy Uncertainty (MPU) is an
important determinant of commodity risk premiums. Using data on the positions of traders for
individual commodities, we also uncover a novel empirical relationship between excessive speculative
activity, measured using Working (1960)'s Tindex, and MPU.
Despite the voluminous literature on the effect of monetary policy changes on commo dity prices
(Basistha & Kurov, 2015; Frankel, 2008) and other financial variables such as inte rest rates (Kuttner,
2001), exchange rates (Fatum & Scholnick, 2008) and stock retu rns (Bernanke & Kuttner, 2005), the
response of commodity prices to MPU has not been thoroughly investigate d in the literature. This is
unfortunate given the important role that uncertainty shocks play in deter mining real activity
(Bekaert & Hoerova, 2014; Bloom, 2009; Jurado, Ludvi gson & Ng, 2015), the equity risk premium
(Anderson, Ghysels & Juergens, 2009; Zhou, 2017) and bond risk premium s (Buraschi, Carnelli &
Changes in interest rates can, according to Frankel (2008), exert an impact on commodity prices
through the inventories and speculation channels as well as by changing the incentives for extracting
The tight link between the Federal funds target rate, the Fed's main policy instrument,
and various short-term interest rates, implies that monetary policy actions can also affect commodity
prices through the aforementioned channel. Given that our measure of MPU comprises information on
Despite the widely held view that commodity price increases pass-through to inflation, Gospodinov and Ng (2013) were
the first to provide evidence of a robust empirical relationship between commodity prices and inflation.
We should note at the outset that various definitions of economic uncertainty are employed in existing research. Bloom
(2009) uses the Chicago Board Option Exchange (CBOE)'s S&P 500 option-implied volatility index (VIX) to measure
economic uncertainty. Bekaert and Hoerova (2014) refine Bloom (2009)'s approach by decomposing the VIX into
uncertainty, proxied for using stock market volatility, and a variance risk premium. Jurado et al. (2015) define time-varying
economic uncertainty as the common component extracted from a large panel of unexpected changes in macroeconomic
variables. Anderson et al. (2009) measure uncertainty as the degree of disagreement among professional forecasters while
Buraschi et al. (2013) combine a Taylor rule estimate of the Federal funds rate with survey forecasts to measure MPU.
Baker, Bloom, and Davis (2016) devise an economic policy uncertainty index which comprises measures of uncertainty
relating to fiscal, monetary and regulatory actions. Given that our interest centers specifically on examining the effect of
MPU on commodity prices, we opt not to employ the existing measures of economic uncertainty. We discuss the reasons
we opt not to use the uncertainty measures discussed before as well as the advantages of our measure of MPU in greater
detail in section 2.3.
More specifically, Frankel (2008) argues that an increase in the interest rate increases firms’costs of carrying inventories
and entices speculators to reallocate their portfolios. Following an interest rate increase, speculators would increase their
holdings of Treasury bills, which become more attractive due to the higher interest rates, and decrease the share of
commodities (or commodity futures contracts) in their portfolios. In addition, an increase in interest rate would increase the
incentive to extract resources due to the increased opportunity cost of delaying extraction.
GOSPODINOV AND JAMALI