In the past decade, financial institutions have assumed an ever greater role in energy derivatives (or “paper”) markets. Numerous recent studies provide novel evidence of this “financialization” and analyze the extent to which it helps explain an important aspect of the distribution of energy returns. In this paper, we summarize their findings, identify some questions that remain unanswered, and discuss what data or theoretical breakthroughs could shed light on those issues.
Keywords: Financial Institutions, Energy Derivatives, Speculation, Financialization, Cross-Market Linkages
JEL Classification: G10, G12, G13 , G23
Suggested citation: Buyuksahin, Bahattin and Robe, Michel A., Does it Matter Who Trades Energy Derivatives? (March 1, 2012). FEEM (Fondazione Eni Enrico Mattei), Review of Environment, Energy and Economics (Re3), http://dx.doi.org/10.7711/feemre3.2012.03.001
See other useful links
In Spring 2011, commodity assets under management (CAUM) surpassed $400bn for the first time. Less than a year earlier, in July 2010, CAUM had broken the $300bn mark. Over the course of the prior decade, CAUM rose by $290bn – with investment inflows accounting for the bulk ($245bn) of the increase. This development partly reflects the growth of physically-backed financial assets, such as precious-metal exchange-traded products. Much of it, however, echoes the massive growth of commodity “paper” markets.
Intuitively, energy futures should account for a substantial part of this growth. Between July 2004 and July 2008, open interest in exchange-traded futures and futures options on crude oil, natural gas and corn (important for ethanol production in the United States) almost tripled. Although the number of open positions dropped substantially in the following year, open interest in those markets remains two to three times higher than just a decade ago – in part as a the result of greater market participation by index investors, hedge funds and other financial institutions.
What is behind these growth figures? Have they been accompanied by structural changes in energy derivatives markets, especially changes in activity patterns along the maturity curves or changes in the relative activity levels of different types of market participants? If so, have those changes in turn brought about material shifts in the distribution of energy prices – in their levels, their volatilities, or the strength of intra- and inter-market linkages?
Answers to these questions are of interest to both market participants and policy makers. The unprecedented roller-coaster of energy prices in the last five years, coupled with a greater need for some economic agents to hedge against energy price risk, have only added a measure of urgency to the search for theoretical and empirical answers.
A number of recent studies, including several by our co-authors and ourselves, utilize proprietary trader-level position data from the Commodity Futures Trading Commission (CFTC) to document changes in the U.S. energy futures markets and to analyze how these changes help explain the distribution of oil- and energy-market returns. In this paper, we summarize some of these findings. We also highlight some related papers, identify some questions that remain unanswered, and discuss what data or theoretical breakthroughs could shed light on those issues.
Changes in Oil Futures Market Participation and Intra-Market Linkages
The open interest in West Texas Intermediate (WTI) crude oil futures and futures options quintupled between 2000 and 2008. Büyüksahin, Haigh, Harris, Overdahl and Robe (BHHOR 2011) document that this growth involved a considerable lengthening in the market’s term structure and a structural change in the composition of trading activity. They identify two main trends from the WTI futures and options position data. They then use econometric analyses to establish a connection from those trends to a structural change in the relationship between crude oil futures prices across the term structure.
The first trend is a dramatic increase in liquidity for contracts with far-out delivery dates. Open interest at maturities greater than one year grew nearly twice as fast as open interest at shorter maturities. Several types of traders now carry much larger long-dated positions (one year or more) than they held in near-dated contracts (three months or less) in 2000. The second trend relates to the nature of market participants. Commodity swap dealers’ share of the WTI open interest grew markedly during the key years of 2002 and 2003, reflecting those traders’ over-the-counter (OTC) activities amid the start of a commodity-index investment boom. Meanwhile, in a development partly related to greatly expanded spread trading, the market share of financial traders more than doubled – from less than 20 percent of all open futures and futures-equivalent option positions in 2000 to more than 40 percent by 2010. During the same decade, “traditional commercial” traders’ share of the futures open interest halved.
The increased oil-futures market activity of index trading activity in near-dated contracts, and that of hedge funds and other financial traders at further-out maturities, has had a major impact on the relationship between crude oil futures prices at different maturities. Specifically, these two developments led to the emergence of a strong and stable relationship between the prices of short-dated and long-dated crude oil futures. After controlling for changes in physical-market fundamentals, BHHOR provide evidence that the growth in the market presence of those two (but not other) types of traders helps explain the emergence after 2003 of a cointegrating (i.e., stable long-term) relationship between the near-term, one-year-out and two-year-out crude oil futures. This new relationship has survived the market dislocations brought about by Lehman Brothers’ demise in September 2008.
This development is economically significant. A critical issue, for many energy market participants, is whether a long-term relationship exists between the prices of some contracts even though these contracts’ prices may diverge in the short run. Such long-run co-movements are precisely the ones that the cointegration analysis identifies. One implication is that long-term oil-price hedging strategies should be more effective when they are based on cointegrated futures contracts, i.e., when nearby and backdated contracts are not segmented. In the same vein, a lack of market integration across contract maturities could be deleterious to energy traders who might otherwise rely on its existence for price discovery or hedging purposes.
Cross-Market Trading and Cross-Market Linkages Linkages
A debate is ongoing as to whether the co-movements between returns on equity and energy (i.e., commodity) investments, and between the returns on different commodities, have strengthened amid the financialization of commodity-markets. The debate centers on two issues: whether there has been a secular increase in those correlations and, in the affirmative, the date and cause(s) of the structural break.
Büyüksahin, Haigh & Robe (2010) and Büyüksahin & Robe (2010, 2011) provide evidence that those correlations increased massively after Lehman’s collapse in 2008. In 1991-2008, in contrast, the strength of commodity-equity linkages fluctuated substantially around a mean close to zero.
In line with economic intuition (Gorton & Rouwenhorst, 2006), return data from 1991 to 2010 show that physical-market and macroeconomic conditions are important drivers of the observed commodity-equity correlation patterns. Still, after controlling for those fundamentals, statistical tests show that the dynamic conditional correlation between returns on energy-futures and equity indices increased amid greater activity by one type of financial traders – hedge funds.
In contrast, trader-level data show that the positions of other kinds of commodity-futures market participants (including, notably, commodity index traders) have little or no explanatory power. Instead, the explanatory power can be traced to a subset of hedge funds – those that trade across futures markets.
Commodity-equity correlations soared after the demise of Lehman Brothers in Fall 2008 and have remained unusually high since November 2008. Even before the 2008-2011 financial crisis, equity-commodity co-movements were positively related to financial market stress. Intuitively, hedge funds could be an important transmission channel of negative equity market shocks into the commodity space. In fact, the empirical evidence suggests that the impact of hedge fund activity is lower during periods of stress.
The sharp increase of energy-equity correlations since Fall 2008, to levels not seen in the previous two decades, holds the potential to be a momentous development. In particular, insofar as a large part of the commodity index surge can be attributed to the search by passive investors for assets uncorrelated with the stock market, this increase in correlations could have a dissuasive effect on further inflows. In turn, one can imagine scenarios in which reduced market participation by passive investors has an impact on intra-market linkages or on the slope of the term structure of oil futures prices (and, hence, on the rewards to oil storage).
For those and other reasons, it is important to understand why cross-market linkages have been so strong in the last two years. It could simply be that increases in investor risk-aversion have shortened their effective discounting horizons for equity-linked cash-flows, making equity prices relatively more susceptible to changes in current demand and supply conditions – in other words, making them behave more like commodities. Theoretical and empirical studies are needed to disentangle this from other (possibly trading-related) candidate explanations. Such empirical analyses would ideally rely on very long commodity data series.
Price Levels and Volatility: What We Know
A central question is why crude oil prices reached a record in 2007-2008. A number of academic studies find a role for physical-market fundamentals, including limits on crude oil spare production capacity, the sluggishness of production change decisions, possible producer decisions to limit production, or unexpected increases in the world demand for oil (e.g., Hamilton, 2009; Kilian and Murphy, 2011; Kilian and Hicks, 2011). An interesting issue is the extent to which environmental regulations may unwittingly have exacerbated some of those factors (Verlegger, 2009).
Notwithstanding those studies, and the reality that for each long investor there must be someone willing to go short, many oil-market observers continue to worry about the impact from the massive increase in oil speculation over the five years. This concern has started to find its way into some academic studies – see, e.g., Singleton (2011).
Several recent studies based on public (Irwin and Sanders, 2012) or on non-public CFTC data (Büyüksahin and Harris, 2011), however, suggest that the market activities of speculators (including index traders and hedge funds) do not Granger-cause commodity prices in general, and crude oil prices in particular. Indeed, univariate tests by Büyüksahin and Harris (2011) in the 2000-2009 period, together with another study using multivariate Granger-causality tests in a cross-section of commodity markets (Brunetti et al., 2011), conclude that hedge funds exert a calming influence on crude oil prices by lowering oil price volatility. In a similar vein, a recent paper by Boyd et al. (2010) finds evidence of hedge fund herding in the crude oil market, but document that this herding is countercyclical and conclude that “hedge fund traders serve to stabilize futures markets by buying when prices are dropping and selling when prices are rising.”
Conclusion and Further Research
The studies we have briefly summarized together establish that who trades helps explain several aspects of the distribution of prices and returns in energy-markets, though not the level of those prices and returns. We now use the question of what (beyond physical fundamentals) does drives energy-market returns, to illustrate the importance of devoting additional attention to data granularity and maturity issues.
An extensive theoretical and empirical literature asks whether returns in futures markets are related to commodity-specific risk (reflected in hedging pressures) or systematic risk (measured by some macroeconomic variables). The majority of the extant empirical studies find evidence that idiosyncratic risk is priced, with futures prices biased downward (upward) in commodity markets where hedgers’ net position is short (long).
Lacking access to more detailed position data, those studies estimate hedging pressures based on publicly-available CFTC information that aggregates hedgers’ net positions across all maturities – even though the risk premia they are trying to link to hedging pressure are defined using near-dated futures returns only. This inconsistency has become a more serious issue in the past decade, as the growth rates, compositions, and directions of traders’ net positions in back-dated contracts have all differed greatly from those in near-dated contracts. One must therefore revisit this empirical question based on (currently non-public) maturity-specific position data.
[Note 1] Much of this paper is informed by our joint research at the U.S. Commodity Futures Trading Commission (CFTC) and at the IEA. It incorporates feedback received at numerous seminars as well as from academics and key market participants in panels at the 33rd International Conference of the IAEE (Rio de Janeiro, Summer 2010), Stanford University’s Energy Modeling Forum (EMF, Fall 2010), the Haskayne School of Business and Bank of Canada “Bi-Annual Conference on Commodities, the Economy, and Money” (Calgary, June 2011) and the Energy Information Administration’s Workshop on Financial and Physical Oil Market Linkages (Washington, DC, August 2011). It also reflects many discussions with past and present colleagues at the CFTC, especially our co-authors James A. Overdahl, Jeffrey H. Harris, James T. Moser, David Reiffen and Celso Brunetti. The CFTC and the IEA, as a matter of policy, disclaim responsibility for any private publication or statement by any of their employees or consultants. The views expressed in this document are mine only and do not necessarily reflect the views of the CFTC, the IEA, the Commissioners, any government, or our colleagues at either institution. Errors or omissions, if any, are the authors’ sole responsibility.
[Note 2] While the composition of trading interests matters, fundamentals also matter. The study controls for fundamentals that might affect cointegration: spare crude oil production capacity and two proxies of world demand for commodities (an index of non-exchange-traded commodity prices, and the cost of shipping dry freight in bulk). The study also controls for an exogenous liquidity change (the advent of electronic trading); for the general cost of liquidity (the slope of the term structure of near-term crude oil prices, which captures the cost of maintaining an exposure to crude oil prices by means of a long position in the nearby contract); and, for cross-sectional differences in liquidity (to allow for the fact that most longer-term trading involves contracts maturing in June or December).
[Note 3] See also Stoll & Whaley (2010), Tang and Xiong (2010), and references therein.
Boyd, N.E., B. Büyüksahin, M.S. Haigh, and J.H. Harris (2010). “The Prevalence, Sources and Effects of Herding”. Working Paper, CFTC, July 2010.
Büyüksahin, B. and J.H. Harris (2011). “Do Speculators Drive Crude Oil Futures Prices?” Energy Journal, 32 (2), pp. 167-202. An early version is publicly available from the SSRN: http://ssrn.com/abstract=1435042
Büyüksahin, B. and M.A. Robe (2010). “Does It Matter Who Trades? Hedge Fund Activity and Commodity-Equity Linkages.” Working Paper, American University, November. Available online: http://ssrn.com/abstract=1707103
Büyüksahin, B., M.S. Haigh and M.A. Robe (2010). “Commodities and Equities: Ever a ‘Market of One’?” Journal of Alternative Investments, 12 (3), pp. 75-95.
Büyüksahin, B. and M.A. Robe (2011). “Does Paper Oil Matter?” Working Paper, American University, July. http://ssrn.com/abstract=1855264
Büyüksahin, B., M.S. Haigh, J.H. Harris, J.A. Overdahl and M.A. Robe (2011). “Fundamentals, Trading Activity and Derivative Pricing.” Paper presented at the 2009 Meeting of the European Finance Association. Updated, August 2011: http://ssrn.com/abstract=966692
Brunetti, C., B. Büyüksahin and J.H. Harris (2011). “Speculators, Prices and Market Volatility.” Working Paper. Available online: http://ssrn.com/abstract=1736737
Gorton, G. and K.G. Rouwenhorst (2006). “Facts and Fantasies about Commodity Futures”. Financial Analysts Journal, 62(2), pp. 47–68.
Hamilton, J.D. (2009). “Causes and Consequences of the Oil Shock of 2007-08”, Brookings Papers on Economic Activity, Spring, pp. 215-259.
Irwin, S.H. and D.R. Sanders (2012). “Testing the Masters Hypothesis in Commodity Futures Markets”, Energy Economics, 34 (1), pp. 256-269.
Kilian, L. and D. Murphy (2011). “The Role of Inventories and Speculative Trading in the Global Market for Crude Oil”. Working Paper, University of Michigan, May. http://www-personal.umich.edu/~lkilian/MS2010126R1.pdf
Kilian, L. and B. Hicks (2012). “Did Unexpectedly Strong Economic Growth Cause the Oil Price Shock of 2003-2008?” Forthcoming, Journal of Forecasting. Last draft available online at: http://www-personal.umich.edu/~lkilian/kilianhicks010711r1.pdf
Singleton, K.J. (2011). “Investor Flows and the 2008 Boom/Bust in Oil Prices.” Working Paper, Stanford University, 2011. http://ssrn.com/abstract=1793449
Stoll, H.R. and R.E. Whaley (2010). “Commodity Index Investing and Commodity Futures Prices.” Journal of Applied Finance, (1).
Tang, K. and W. Xiong (2010)• Index Investing and the Financialization of Commodities. NBER Working Paper No. 16325, September.
Verlegger, P.K., Jr. (2009). “Anatomy of the 10-Year Cycle in Crude Oil Prices”. Working Paper, University of Calgary, March.
International Workshop on "Financial Speculation in the Oil Market and the Determinants of the Oil Price"
Interviews with the Experts