I’ve just completed a new research paper with University of Chicago Professor Cynthia Wu on the Effects of Index-Fund Investing on Commodity Futures Prices. Here was our motivation for writing the paper:
The last decade has seen a phenomenal increased participation by financial investors in commodity futures markets. A typical strategy is to take a long position in a near futures contract, and as the contract nears maturity, sell the position and assume a new long position in the next contract, with the goal being to create an artificial asset that tracks price changes in the underlying commodity. Barclays Capital estimated that exchange traded financial products following such strategies grew from negligible amounts in 2003 to a quarter trillion dollars by 2008 (Irwin and Sanders (2011)). Stoll and Whaley (2010) found that in recent years up to half of the open interest in outstanding agricultural commodity futures contracts was held by institutions characterized by the Commodity Futures Trading Commission (CFTC) as commodity index traders.
Hedge fund manager Michael Masters argued in testimony before the U.S. Senate in 2008 that purchases of commodity futures contracts by index funds must have influenced prices:
Index Speculator demand is distinctly different from Traditional Speculator demand; it
arises purely from portfolio allocation decisions. When an Institutional Investor decides
to allocate 2% to commodities futures, for example, they come to the market with a set
amount of money. They are not concerned with the price per unit; they will buy as many
futures contracts as they need, at whatever price is necessary, until all of their money
has been “put to work.” Their insensitivity to price multiplies their impact on commodity
In testimony the following year before the CFTC, Masters claimed:
Buying pressure from Index
Speculators overwhelmed selling pressure from producers and the result
was skyrocketing commodity prices.
In our new paper, Cynthia Wu and I review a simple model of how such market pressure might influence commodity futures prices. The basic idea is that the more futures contracts the funds want to hold, the more risk the counterparties who short the contract are exposed to. According to the model, the futures price must be bid high enough to compensate the short side for absorbing the risk. This compensation comes in the form of an expected profit to the short side of the futures contract. The empirical implication of the model is that if the long position of index funds is indeed exerting an effect on commodity futures prices, then the notional value of index-fund futures contract holdings should help predict the returns over time on the contracts themselves.
The framework thus offers an interpretation for a large previous literature that has looked for evidence of predictability of commodity prices based on various measures related to index-fund investing. Here is a brief summary of those previous studies provided by our paper:
Brunetti, Buyuksahin, and Harris (2011) used proprietary
CFTC data over 2005-2009 on daily positions of traders disaggregated into merchants,
manufacturers, floor brokers, swap dealers, and hedge funds. They found that changes in net
positions of any of the groups could not help to predict changes in the prices of futures contracts
for the three commodities they studied (crude oil, natural gas, and corn). Sanders and Irwin
(2011a) used the CFTC’s publicly available Disaggregated Commitment of Traders Report on
weekly net positions of swap dealers, and found these were no help in predicting returns on
14 different commodity futures contracts over 2006-2009. Sanders and Irwin (2011b) used
proprietary CFTC data to extend the public Supplemental Commitment of Traders (SCOT),
which categorizes certain participants as commodity index traders, back to 2004. They found
that changes in the positions of index traders did not help predict weekly returns for corn
or wheat but found some predictability for soybeans under some specifications. Stoll and Whaley (2010) used the public SCOT for 12 agricultural commodities over 2006-2009 and
found that changes in the long positions of commodity index traders predicted weekly returns
for cotton contracts but none of the other 11 commodities. Alquist and Gervais (2011) used
the public CFTC Commitment of Traders Report to measure net positions of commercial and
non-commercial traders, and found that changes in either category could not predict monthly
changes in oil prices or the futures-spot spread over 2003-2010, though there was statistically
significant predictability when the sample was extended back to 1993. Irwin and Sanders
(2012) used the CFTC’s Index Investment Data on quarterly positions in 19 commodities
held by commodity index funds. They found in a pooled regression that changes in these positions
did not predict futures returns over 2008-2011. They also separately analyzed whether
changes in futures positions of a particular oil- or gas-specific exchange traded fund could
predict daily returns on those contracts over 2006-2011, and again found no predictability.
Buyuksahin and Harris (2011) used proprietary CFTC data on daily positions broken down
by non-commercials, commercials, swap dealers, hedge funds, and floor broker-dealers. They
found the last category could help predict changes in oil futures prices one day ahead, but no
predictability for any of the other categories or other horizons. By contrast, Singleton (2011)
found that a variety of measures, including a 13-week change in index-fund holdings imputed
from the SCOT, could help predict weekly and monthly returns on crude oil futures contracts
over September 2006 to January 2010.
The model we investigated has a specific implication for how index-fund holdings could influence futures prices– the notional holdings of index-fund investors should help predict the returns for the counterparties to those contracts.
For 12 agricultural commodities, since 2006 the CFTC has been providing through its Supplemental Commitments of Traders Report weekly positions for traders it characterizes as “replicating a commodity index by establishing long futures positions in the component markets and then rolling those positions forward from future to future using a fixed methodology”. We looked at a variety of specifications in which the notional holdings associated with those positions might help predict returns on any of the futures contracts. Consistent with the findings of the other researchers summarized above who used different data sets and methods, we found essentially no helpful forecasting relations for any of the 12 agricultural commodities.
This was something of a disappointment for us, since we had what we thought was a neat model to capture the effect that Masters expected to see. But we found no basis in the data for assigning a value other than zero to any of the key parameters of that model– there just is no forecasting relation in the data between index fund positions in these 12 agricultural commodities and subsequent changes in their futures prices.
As noted above, the literature has reported some more mixed results with oil prices as opposed to agricultural prices. I will discuss what we found for those data in a follow-up post.