John Maynard Keynes once remarked that “the ideas of economists […] are more powerful than commonly understood. Indeed, the world is ruled by little else” ( 2008:350). This view is reflected in the media, where interviews with economists frequently accompany the news (Weinstein 1992:73-4). However, the role of economic theory in real-world contexts is subject to substantive academic debate, even with regard to its central object of study: the market. By analyzing the relationship between economic theory and commodity futures markets (CFMs), I aim to contribute to the existing debate. I offer an explanation in how far aspects of economics are realized in practice, which role economic theory plays in their production and what distributive impacts result.
I start by introducing today’s dominant neoclassical theory of economics and its nineteenth century origins. Neoclassical theory assigns itself the role of an objective description of markets, which evolve naturally and tend towards a general equilibrium. In order to satisfy intricate mathematical models, it assumes omniscient, rational and optimizing agents. Yet this abstraction removes all social factors (Watson 2005:151) and deprives economics of the dimension which Fligstein calls ‘markets as politics’, where agency and power are employed by market actors to influence distribution (1996:657). The resulting discrepancy between economic assumptions and real-world contexts casts doubts on the explanatory power of neoclassical theory, leading some scholars to dismiss its findings completely (Hosseini 1990). In contrast, a look at CFMs shows that these closely reflect a number of neoclassical aspects. This seeming contradiction can be resolved using the theory of performativity, which is introduced in the second section. Accordingly, the fit of markets to economics is not the result of natural evolution but of conscious making. The corresponding role of economic theory is that of a ‘blueprint’, which serves to incorporate economic aspects during market formation without its assumptions having been met initially (Callon 1998b:23). Different approaches were developed within the performativity literature, enabling its use in comprehensive analyses. However, by focussing on the context rather than the action of successful performances, these approaches again downplay the ‘political’ factors of agency, power and distribution. I particularly emphasize these factors, because their marginal treatment obscures performativity’s potential as a theoretical alternative to neoclassical economics.
The insights from the first two sections are applied to CFMs in the third. These are an excellent example for this purpose, as they were created at a time when the foundations of neoclassical economics were already developed, yet no similar markets to emulate existed. Both circumstances support the idea of neoclassical theory as a ‘blueprint’, and I consequently illustrate historical steps that were designed to incorporate aspects of economics. At the same time, discourses were constructed to publicly legitimize the new markets and their agents. Throughout these performative processes, the political aspects of agency and power played a significant role in shaping outcomes and leading to a distribution that approximated neoclassical market equilibrium. However, the recent period of ‘financialization’ transformed CFMs, especially through their use for investment portfolio diversification. This provoked a loss of fit with neoclassical assumptions that can be classified as ‘counterperformativity’, and the ensuing formation of commodity price bubbles produced severe distributional consequences.
Based on the findings of this dissertation I conclude that aspects of economics were successfully realized on CFMs, yet neoclassical theory fails to explain how. Using performativity theory to substantiate this middle ground position, it can be shown that it was conscious market making that led to stable equilibrium. Subsequent dynamics disrupted this state, however, causing disequilibrating effects and negative distributional impacts. Despite political initiatives to reverse the current situation, whether CFMs will return to reflect economic theory in the near future remains to be seen.
From Neoclassical to Futures Markets
Economic theory from the mid-to-late nineteenth century, due to the concept of marginal utility often called ‘marginalist’ era, influenced today’s dominant ‘neoclassical’ market conception in fundamental ways (Niehans 1990:161). The core of this conception, described by Morgan as “formal treatments of rational, or optimizing, economic agents joined together in an abstractly conceived free-market, general equilibrium world” (2008:279), is illustrated in the first subsection. While neoclassical theory envisions itself as an objective description of ‘naturally’ emerging markets (Aspers 2011:149), its assumptions lack central ‘natural’ features. The noticeable suppression of the political aspects of agency, power and distribution is highlighted in the second subsection. While the epistemic culture of economics values a model’s isolation of crucial features more than its fit to reality (Friedman 1953:34,36), some non-economists argue that the shortcoming of central assumptions makes neoclassical theory an inadequate representation of concrete markets (Authers 2011). Yet conversely, looking at CFMs in subsection three shows how these reflect central theoretical aspects.
Neoclassical Market Theory
A cornerstone of economic theory since the classical era is the idea of equilibrating market forces, famously captured in Adam Smith’s ‘invisible hand’ metaphor ( 1904:29). Its assumption that the actions of economic actors with given preferences naturally lead to an ordered market remains part of orthodox thinking (Aspers 2011:150-1). Beginning in the 1830s, however, the focus of economic analysis changed decisively from producers to consumers and the utility the latter derive from goods (Mátyás 1985:15). While Smith already mentioned utility as a determinant of value, it is during the era of marginalist economics that it became the fundamental element (Deane 1978-1). This era laid the foundations of today’s neoclassical theory of the market, in which stable equilibriums and true prices are achieved through the rational decisions of utility-maximizing agents under perfect competition.
The combination of marginal supply and demand considerations enabled the modelling of a market in equilibrium. Based on this notion, Léon Walras (1834-1910) increased the number of markets to derive a general equilibrium theory. Accordingly, all markets move naturally towards a unique and stable state where all consumer demand is met by corresponding supply at the determined ‘true’ price (Geanakoplos 1992:120). As Arrow and Debreu later specified, the theory assumes a world in which everything is tradable (Bernstein 1992:216). Since no institution in the real world is able to fulfil the necessary coordination function, Walras resorted to the theoretical ‘tâtonnement’ process (Niehans 1990:213-4). It describes an auction that takes place before the actual goods exchange, in which demand and supply are adjusted until an equilibrium price for every market is determined. Walras thus assumes the ‘natural’ existence of markets (Watson 2005:144). The first theorem of welfare economics further states that any general equilibrium is Pareto optimal, i.e. an ideal distribution prevents any agent from being better off without another being worse of (Geanakoplos 1992:120). Since equilibrium is the natural state of the system, destabilizing shocks must originate externally. If these cause a deviation from equilibrium, consumers and producers adjust accordingly. Hence, the market possesses a built-in negative feedback mechanism that ensures the return to a steady state (Prasch 2008:45-6).
The ‘methodological individualism’ that informed classical economics was adopted by marginalist scholars (Deane 1978:14, 101). Supply and demand are thus aggregates derived from models of individual consumers and producers (Kuenne 1968:284-7). Their correct relationship can only be observed on markets under perfect competition. This concept was laid out by Antoine Augustine Cournot (1801-1877), who reasoned that an increase in the number of agents on the market would lead them to accept prices as a parameter for their consumption and output decisions (Niehans 1990:185-6). Besides an inability of market participants to consciously influence prices, perfect competition is also characterized by homogeneous products and negligible transaction costs (Van Der Hoek 2001:1204-5). In such a ‘perfect market’ goods forming identical substitutes are traded at the same ‘true price’, following William Stanley Jevons’ (1835-1882) ‘law of indifference’ (Rima 2009:263).
Neoliberal theory regards standardization as essential for order to emerge during market formation (Aspers 2011:119). It therefore assumes not only homogeneous products, but also homogeneous agents. Building on prior works, Jevons developed the concept of utility-maximization as the central motivation of economic actors (Backhouse 1985:70-2). This became an essential condition within general equilibrium theory to attain the greatest possible social welfare. In order to calculate optimal decisions, consumers must be able to judge the amount of utility a certain good provides (Begg et al. 2003:55-61, Kuenne 1968:387). Hence, they need to form rational expectations. In his seminal work, Muth defines rational expectations in orthodox economics as “informed predictions of future events, [which] are essentially the same as the predictions of the relevant economic theory” (1961:316). All information is thus incorporated into agents’ decisions, while irrelevant information is ignored. Only random external influences can temporarily disturb expectations (ibid:316-7).
The marginalist Carl Menger (1840-1921) highlighted the importance of information, knowledge and uncertainty over time for the formation of values and prices (Backhouse 1985:88). The fact that prices under perfect competition fully reflect all available information subsequently became known as market ‘efficiency’ (Fama 1970:383), and the efficient market hypothesis (EMH) turned into a fundamental assumption in (financial) economics (MacKenzie 2006:246). Fama distinguished three levels of market efficiency (1970:383): The weak form defines information in markets as historical prices, whereas the semi-strong and strong forms add public and proprietary information respectively. For the EMH to hold, the absence of transactions costs, the costless availability of information to all market participants and a shared understanding of the meaning of information are assumed (ibid:387). In this state of ‘perfect information’ all consumers and producers possess all relevant information to judge prices, but act as price-takers following Cournot (Tsaliki 2001:1205). Modigliani and Miller introduced arbitrage as central force behind market efficiency, because agents exploit riskless gains until prices converge to their ‘true’ values (1958:279-80). Fama thus suggests calling markets efficient when trading systems based on information fail to produce excess returns above the overall market rate (Bernstein 1992:137).
Politics and Neoclassical Market Theory
Neoclassical theory, as illustrated above, is an idealized abstraction from reality. The increasing use of calculus from the beginning of the marginalist era to solve optimization problems produced models simplified to enable their description in mathematical terms (Niehans 1990:178). This simplification led to the neglect of socio-political factors, which are either omitted from calculations as ‘ceteris paribus’ conditions or incorporated as random variables in individuals’ utility functions (Fine 1998:50,57, Hausman 1990:182). The resulting mechanistic and asocial world deprives its actors of the central political properties agency and power, and therefore their influence on economic distribution. Such a world cannot be found in empirical studies of real market environments, as these are ‘embedded’ in and dependent on ongoing social relations (Granovetter 1985:481-2). This makes it necessary to analyze ‘markets as politics’.
The concept of agency and the relationship with the structure in which agents find themselves is central to politics. Structure can be defined as the context within which economic, political and social effects occur, whereas agency refers to the actions of agents that try to consciously realize their intentions (Hay 2002:94). General equilibrium models, in the absence of heterogeneous social properties, assume standardized utility-maximizing agents (Watson 2007:65-6). As these agents are modelled at the individual level before being analyzed in macroeconomic aggregates, the resulting ‘methodological individualism’ (Arrow 1994:1) creates the impression that agency surpasses structure in determining outcomes. However, the fact that agents always follow the optimal adaption to their environment precludes any choice between alternative actions. Neoclassical economics thus suppresses agency in favour of structural properties of the economy (Hay 2002:103-4). Yet this theoretical supremacy of structure is not confirmed by observations of real-world markets. The diversity of agents and their interests create an environment in which simple optimal choices are largely impossible to determine, and different strategies are chosen under uncertain conditions (Fligstein 1996:659-60). On complex markets, like financial markets, the influences arising from agency severely complicate any attempts to build a general equilibrium model (Kuenne 1968:350-1).
Still, neoclassical theory adheres to the idea of market equilibrium, and attributes the corresponding welfare gains to free competition and agents’ negligible economic power (Kuenne 1968:387). Following the EMH and the assumption of rational expectations, all agents are equally informed and possess similar calculative capabilities. Yet even some economists refute this notion. Paul Samuelson argues that the theoretical unpredictability of prices, which is a property of efficient markets, is dependent on the proportion of skilled investors (Bernstein 1992:123,149). Thus assets are only priced correctly when a number of big investors know how to value them. This contradicts the assumption of strict price-taking behaviour, i.e. the absence of power. Clearly, the size and capability of actors on a market influence their ability to shape developments (Fligstein 1996:663). Nonetheless, theory maintains that a general equilibrium is also the point where no agent is able to influence the market in a way that increases his utility, as prices reflect the cost of production and market participants make neither profits nor losses (Backhouse 1985:79, Geanakoplos 1992:120).
This implies that agents refrain from attempts to influence distribution, as no individual gains are achievable. Walras went even further by conceptualizing ‘justice in exchange’ (Watson 2005:151). The idea to hold all existing distributions constant thus makes his general equilibrium models distribution-neutral (Niehans 1990:211-2). Again, this assumption does not correspond to empirical observations. Especially on markets where speculation forms an integral part, agents are exploiting power asymmetries in order to increase their profit. This is acknowledged by the imposition of speculative position limits on CFMs “to reduce the potential threat of market manipulation” (§7d.5, Commodity Exchange Act 2010).
The elimination of socio-political factors in the world of neoclassical markets reduces its correspondence with reality. Some authors infer from this ‘unrealistic’ point of departure that real-world markets with neoclassical properties do not exist (Hosseini 1990). This conclusion goes too far, however, as the introduction of CFMs in the next subsection shows that they closely reflect a number of theoretical aspects.
Commodity Futures Markets
CFMs organize the trading of derivatives, which according to the International Accounting Standards Board (IASB) possess three main characteristics (IASB 2009): their value is based on the price of an underlying commodity, the initial net investment is smaller than the value of the underlying and they are settled at a future date. Commodity futures contracts are thus non-optional commitments to buy or sell a given quantity of an underlying commodity, e.g. wheat or oil, by an agreed future date at an agreed price. They are standardized regarding the quality of the commodity, contract size and delivery date. The main difference to forward contracts is that there is generally no intention to exchange the physical commodity. Instead, contracts are cancelled out with opposite contracts and differences settled in monetary terms. The possibility of physical delivery solely ensures the tie between futures price and the price of the underlying commodity (Valdez and Molyneux 2010:402-4). However, even the requirement of deliverability became obsolete with the introduction of financial futures in 1983 (Goss 2000:2-3). Commodity futures contracts are traded on exchanges or as over-the-counter (OTC) products. A commodity exchange generally provides the facilities for spot, futures and options trade. Its clearing house is the nominal counterparty to all trades, and traders are required to deposit a margin of 10-20 percent of contract value. They are allowed to withdraw a part of their margin if changes in prices are favourable, but have to top it up in adverse circumstances to prevent default risk (Radetzki 2008:92). In 1974, the US Congress created the Commodity Futures Trading Commission (CFTC) as an independent agency to regulate commodity exchanges.
CFMs provide two central economic functions: an authoritative mechanism for long-term price discovery and an opportunity to transfer risks (Radetzki 2008:91-2). Before futures trading, price setting occurred directly on physical commodity markets. At these ‘spot’ markets, contracts are agreed for the immediate exchange of goods. But as a commodity’s supply and demand schedules are vague and unstable, price developments can only be determined in the short-term and are generally highly volatile (Radetzki 2008:59-65, Black 1986:536). Fluctuating prices mean producers and consumers face a substantial risk of adverse price changes on the spot market (Atkin 1989:2-3). At this point, futures markets support commodity market stability through the determination of long-term prices. CFMs are characterized as the economically optimal institution to minimize “the transaction costs of exchange of different private information sets concerning the future” (Francis 2000:15). Since futures contracts normally do not lead to physical delivery, there is no theoretical limit to the amount of traders and volume of contracts traded on futures exchanges (Atkin 1989:3). Hence, the same applies to the amount of information introduced by traders. Futures markets are technologically and physically optimized to incorporate the diversity of information instantaneously into prices, thereby harmonizing expectations and predicting the commodity price in the future (Francis 2000:32-3). The futures price is then used by producers and consumers in their supply and demand planning (Hoffman 2000:2). While the futures price at a certain date normally deviates from the spot price, it is expected to converge with it upon reaching maturity. Depending on market conditions and expectations, the futures price may exceed the spot price (‘contango’) or lie below it (‘backwardation’) (Radetzki 2008:102). The trading of futures in isolation is generally depicted as a zero-sum game, as the gains of one party in a trade are exactly offset by losses by the opposing party (Francis 2000:15). However, based on the ideas of John Maynard Keynes, a ‘normal backwardation’ is assumed (1930:143-4). This means that futures contracts are usually priced below the expected future spot price, because they contain a ‘risk premium’. While there has been substantial evidence of normal backwardation, e.g. on wheat markets (Miffre 2000:819), the topic remains debated among economists.
The second function of CFMs, risk transfer, is possible through the interplay of its market participants, the two basic types being commercial and non-commercial (CFTC 2009). Agents dealing commercially with the physical commodity, e.g. producers or consumers, use the futures markets to manage or hedge risks associated with those activities. An owner or producer of a physical commodity can reduce the risk of a price decline by making a ‘short hedge’, i.e. by selling futures for the date he plans the disposal of his physical holding. If the price declines, the loss on the physical market is offset by gains on the futures market, where he locked in the higher price. Conversely, a consumer of a physical commodity can use a ‘long hedge’ (Radetzki 2008:98-9). However, a full hedge precludes profits from price changes, as it simultaneously eliminates losses and gains (Kuenne 1968:349). It is further only effective if spot and futures prices move in tandem, since anomalies in market behaviour can prevent effective risk reduction (Chorafas 2008:79).
The second type, non-commercial agents, mainly encompasses those trading on futures markets to profit from a move in commodity prices. This speculation arises when there is uncertainty, and speculators assume the price risk of hedgers by forming the counterparty to the contract (without owning an offsetting physical position) (Radetzki 2008:100). Assuming risk is an essential part of speculation, as the speculator “can gain expected return by taking on variance” (Markowitz 1952:79). CFMs allow the use of substantial leverage, as margin payments at the clearing house are far below the total contract values (Radetzki 2008:101). Investors may be regarded as a subgroup of speculators. They generally establish long-term positions in CFMs to gain a fairly safe rate of return or diversify their portfolios (Radetzki 2008:103). This is mainly done by rolling over futures contracts, i.e. selling them shortly before expiry and purchasing new contracts. A profit is generated if the market is in backwardation, as the price of futures contracts rises over time (Miffre 2000:804). The structure of concrete investments can take many different forms, participating either directly in commodity futures or using these indirectly to hedge against risks on swap markets (UNCTAD 2009:25).
The description of CFMs enables a comparison of these real-world markets with neoclassical theory. The trading of standardized contracts by potentially unlimited sellers and buyers corresponds well to the assumption of perfect competition. The rapid processing of information at low transaction costs further supports market ‘efficiency’. A CFM can also be regarded as a possible realization of the ‘tâtonnement’ process for a single market (Bliss 1972:98). As most contracts are not settled by delivery, the price determination function resembles the determination of equilibrium prices on an auction before actual exchange takes place. These apparent incorporations of economic aspects on CFMs are backed by several empirical studies (Cox 1976, Telser and Higinbotham 1977).
This section illustrated how neoclassical economic theory is based on the idea of informed and rational agents establishing equilibrium prices on markets. But the abstractions made from observable markets, especially the marginal treatment of the central political factors agency, power and distribution, render the assumptions underlying the theory ‘unrealistic’. While this may cast doubts on its applicability, a look at CFMs shows that they correspond remarkably well to some neoclassical aspects. In order to analyze this seeming paradox, a theoretical approach is needed that acknowledges the influence of economics without being dependent on its assumptions.
 The definition incorporates minor changes as proposed in a draft to the prior definition in International Accounting Standard (IAS) 39.
 For example, wheat futures at the Chicago Board of Trade (CBOT) are either quality class 1 or 2 of the soft red winter variety, encompass 5,000 bushel and are delivered in March, May, July, September or December (CBOT 2011).