Making Things Deliverable: The Origins of Index-Based Derivatives

Introduction

Index-based derivatives: financial contracts that use financial market indices as their underlying ‘assets’ are currently amongst the most commonly traded financial contracts (BIS, 2006). ¹ Furthermore, the introduction of index-based derivatives is considered by many as the single most significant development in contemporary financial markets (Chance, 1995; Arditti, 1996; Kolb, 1997a, 1997b). The use of index-based futures has become a standard practice in the financial world and today banks, pension funds, insurance companies and governments hold portfolios that include index-based derivatives. In fact, index-based contracts have become such an indispensable feature of the global financial system that it would be safe to say that there are many millions in the West who own, either directly or indirectly (even unknowingly), index-based derivatives.

In spite of their ubiquity in the financial world, index-based derivatives represent a fundamental ambiguity. Derivative contracts, as their name implies, derive their prices from the prices of a variety of assets, such as agricultural commodities, precious metals, currencies and many others. The contract themselves, (eg, futures contracts), state the terms of future transaction: price to be paid and time for the delivery of assets. In contrast with physical and deliverable assets, market indices are the products of mathematical procedures applied to market data. Hence, index-based derivatives are contracts written for ‘strange assets’, assets that do not have straightforward physical characteristics, and therefore cannot be delivered, upon buying and selling, in a similar manner to physical assets. This fact raises an interesting historical and sociological question: how did it happen that these abstract mathematical entities became the basis for the most popular financial contract of our time? More specifically, how did the non-physical and non-deliverable nature of market indices change in such a way that allowed indices to serve as a basis for the popular derivative contract we know today? This is the question that this chapter discusses.

Economic sociology has paid little attention so far to the evolution of index-based contracts. Several sociologists refer to the role that index-based contracts play constructing hedging positions and to the way in which they are used in arbitrage trading (for example, Beunza and Stark, 2004; MacKenzie, 2006). Yet, there exists no published sociological work that traces the social history of this market and analyses the evolution of index-based derivatives and the network of institution in which this processes unfolded. This chapter is hoping to be a first step in recognising the importance of index-based financial instruments as a topic for sociological investigation.

How are we to conceptualize index-based derivative contracts sociologically? The empirical material in this chapter suggests that the question of the qualification of index-based derivatives – that is, the process through which the derivatives gained their particular qualities – may provide a useful entry point. The qualification process can be described as an interactive (network-like) process that involves not only various actors located in different organizations and institutions (the staff of commodity exchanges, commodity traders and financial regulators), but also the products themselves.

The notion of product qualification is related to the work done by Geoffrey Bowker and Susan Leigh Star (Bowker and Leigh Star, 1999). Bowker and Leigh Star analyse classificatory systems and trace the ways in which they were embedded into bureaucratic infrastructures and become part of the taken-for-granted reality of organizational structures. Using detailed case studies, Bowker and Leigh Star outline several general historical heuristics along the lines of which classificatory systems come about and how their legitimacy is being proposed and contested. This process of institutionalisation is a vital part of the organizational implication of the qualification process. That is, it describes how the intra- and inter-organizational spaces internalise categories-making rules and practice. Complementarily, this chapter focuses on qualification practices and their performance and less on the resulting structures of such actions. In fact, by focusing on the action-related dimensions of product qualification, this chapter tries to strengthen one of Bowker and Star's more provocative conclusions, the futility of separating between actors and the structure in which they operate. In addition, by regarding qualification as a dynamic process, this chapter refers to a specific theory of action: the Actor-Network theory (ANT), and more specifically, to the work about qualification done by French sociologists Michel Callon, Cecile Méadel, and Vololona Rabeharisoa (2002). As we will see in the discussion below, ANT and the specific form of qualification proposed by Callon capture the dynamic nature of the process. That is, according to Callon qualities of a product are not simply assigned to it, but instead are the outcomes of multiple interactions, including, as we see in the empirical case, the products themselves.

In order to analyse this process, involving heterogeneous entities, we need first to assume that it does not operate in a linear fashion, but rather in network-like one. That is, the various actors that shape the products do not operate one after the other, but rather interact with each other, and products are shaped through these interactions. A second assumption that stems from the interactive and networked perspective of product qualification is that the various actors may promote different and even conflicting ideas about the desired shape and function of the product. As a result, the qualification process should be regarded less as a centrally orchestrated effort, and more as an emergent, and potentially a competitive one. In this respect, it can be expected that any set of qualities attached to a product through the interaction of actors may be challenged by alternative sets of qualities that other actors, or coalitions of actors, propose. Lastly, the networked nature of the qualification process places the evolution of markets themselves in a new light. In particular, the relations between the products and the markets in which they are traded may need to be re-examined.

If qualification is a process in which take part different actors whose conflicting worldviews clash, can we assume that the market itself does not change as a result of such process? In other words, should not we consider qualification not only as a process that takes place in the market and affects products, but also as a process of qualification of the market itself? This perspective departs from Granovetter's influential approach, which insists upon the structural embeddedness of markets (Granovetter, 1985) and tends to bracket off the content of economic goods. In certain aspects it echoes Zelizer's notion of multiple markets and her seminal work about the creation of ‘types of money’ that can be seen as a description of a multifaceted process of product qualification (Zelizer, 1989). In fact, this chapter expands Zelizer's work by focussing on the controversial dimension of the (networked) process of qualification.

As mentioned before, the notion of qualification developed in this chapter is related to a concept presented in an influential paper by Callon, Méadel, and Rabeharisoa (2002) as well as by Callon and Muniesa (2005). The motivation behind the analytical effort in this chapter and the one behind Callon's paper are similar: an effort to understand where products come from. These similarities, however, are limited to the contours of the theoretical approaches, while the contents are significantly different. Callon suggests that qualification operates through the continual creation and recreation of relations between the evolving products and existing products. The temporary outcomes of these comparative exercises are performed each time the qualified product is bought and sold. This schematic framework is a very powerful conceptual tool, because it allows us to treat products as a chain composed of connected qualification attempts. Yet, the qualification of financial products includes aspects that call for broadening and reconfiguration of existing concepts in economic sociology. The relationship between the products and their attached practices are different in financial markets, the markets described in this chapter, from those that exist in most other markets that empirical works in economic sociology describe. In the cases of products such as wine bottles or cars, for example, there is a visible distinction between the practices of trading, in which the qualities of the products take part in constructing the prices of the products, and between the practices through which the qualities are established and tested. In the case of financial products, like index-based derivatives, no such distinction exists; a crucial part in the qualification of financial products is performed on the trading floor. Namely, the way financial products ‘behave’ in the market is not only a trial of their qualities, but the market is also one of the arenas where these qualities are formed and attached to the product. As the empirical material in this chapter shows, the practices through which the qualities of financial products were performed had a dramatic impact on the evolution of the markets.

The construction of deliverability: early history of commodities futures

Derivatives are financial contracts whose market price, as their name implies, is derived from the price of another asset that underlines the derivative. Agricultural commodities, shares and stock indices are just some of the most common entities for which derivative are designed. The central claim that this chapter makes is that to understand the social dimension of derivatives in general, and index-based ones in particular, it is necessary to trace the development of the links between the underlying entities and the derivative contract. It is vital to understand the nature of the qualities of the underlying assets and how these qualities are embedded into institutionalized market mechanisms, practices and conventions. This dynamic historical process through which the qualities of the underlying assets are linked to the derivative contract – the qualification of the contract – is central to the analysis of index-based contracts. To analyse the complex historical process of the qualification of index-based derivatives we need to trace the evolution and the influence of two key factors in the markets where these financial contracts were first traded – the commodities futures markets: (1) the deliverability of underlying assets and (2) the nexus of connections between the futures exchanges and the regulatory establishment.

The first American market to trade derivative contracts, the Chicago Board of Trade (CBOT), started trading in 1848. The commodities-based contracts (known as ‘forwards’) traded in the CBOT specified the terms of mutual obligations of the buyer and the seller to, respectively, deliver and pay for a specified amount of product (of a certain quality) on a set date. For example, a typical contract might include the obligation to deliver 20 tons of potatoes of a given variety and of a given quality at a given date in return for a set amount. The terms of each contract had to be negotiated by the buyer and seller: the date in which delivery of the agricultural product was to take place (the contracts’ ‘expiration date’), the exact nature and quality of the product and the price to be paid for the product on delivery. Hence, since forward contracts were designed for the needs of the two specific parties to the future transaction and they had little use outside the particular setting.

Three years after its inception, in 1851, the CBOT took a revolutionary step and standardized the bilateral forwards. The standardized forward contracts, which came to be known as ‘futures’, were templates that included the terms of the contract and all that was left for the parties to negotiate was the contract's price. Any futures contract bearing the same expiration date and underlying asset became interchangeable with any similar contract, regardless of the identity of the trader who bought it initially. By standardizing the contracts, the time-consuming stage of negotiations over the details of the contracts was eliminated and the trading process became faster and more efficient. There was a more far-reaching implication to contract standardization, however: the standardization made the contracts themselves tradable, thus relieving the traders from the necessity of owning the underlying agricultural products on which the contracts were based. In other words, the traders could buy and sell the contracts and, when the expiration date was approaching, offset their obligations by acquiring contracts that were the ‘other side’ of their transactions. For example, a trader holding a contract requiring her to buy 500 bushels of corn at the end of the month would offset her obligations by buying a contract requiring her to sell the same amount and the same type of corn at that date.

After the standardization of the contracts the CBOT and futures trading witnessed a long period of growth. The popularity of futures trading gave rise to bets that were made on the prices that the market would quote. In the late 1880s, that practice of betting on futures’ prices gained popularity and by the end of the century, there were numerous shops that sold contracts that were based on CBOT-traded prices of agricultural futures (Fabian, 1990). CBOT regarded those establishments, commonly known as ‘bucket shops’, as illegitimate competition. First, the bucket shops were piggybacking on information that originated from the CBOT (where the prices were determined) and offered no compensation for that service. Second, by betting on futures’ prices the bucket shops were taking potential customers from the CBOT and thus were denying commissions to CBOT members. During the last years of the 1890s and the early 1900s, CBOT went on a fierce legal battle against the bucket shops, at the end of which, after several landmark court cases (Ferris, 1988) the operation of the latter was declared illegal and was effectively terminated.

The main argument that the CBOT used in its legal struggle against the bucket shops was that contracts that did not include a specific obligation for the delivery of the underlying assets (i.e., were settable only through the payment of cash), could not count as legitimate commercial contracts but were essentially gambling. The carefully constructed CBOT's argument struck a chord with broader anti-gambling emotions in the American society and following this initial success, the association between cash-settlement and gambling did not remain limited to institutions like the bucket shops, but was extended, at least implicitly, to all other possible cash-settled contracts (Fabian, 1990). A lobbying effort by the CBOT also brought about a change in Illinois’ gambling laws, forbidding the trading of cash-settled contracts and allowing only contracts that included the option to deliver the goods. The changes in Illinois laws were followed by a number of other Midwestern states (Cronon, 1991). Moreover, the notion about the similarity between cash-settlement and illegal betting, in its wider form, became a common argument against the trading of financial contracts and even against the immoral nature of financial markets in general. For example, bucket shops were mentioned in the Congressional debates in the early 1930s that led to the establishment of the SEC (Shapiro, 1984).

Yet, although assets’ deliverability was a milestone in the establishment of CBOT as an economic and political institution, the actual performance of assets’ delivery has all but disappeared from CBOT. The enormous success of standardized futures eroded the importance of deliverability. Over the decades since the introduction of standardized futures in the mid 19th century, the ratio of traders who actually took part in a delivery of assets on expiration of the futures contracts dwindled constantly. By the 1950s the vast majority of futures contracts were not settled in delivery: most estimates were that only in 3 to 5 per cent of transactions did products actually change hands (Clark, 1978; Markham, 1987). The rest of the transactions were settled by offsetting the obligations in the contracts through buying/selling ‘opposite’ contracts. This yawning gap between the volumes of futures trading and actual deliveries of the assets on the basis of which the contracts were written was not merely an evolving feature of agricultural futures exchanges like the CBOT. In effect, the discrepancy between trading futures and delivery became one of the main growth engines for futures market. Nominally, each futures contract was backed up by the underlying agricultural product in the amounts and qualities specified in the contract. Futures trading grew about a 100-fold in the first century of CBOT (Tamarkin, 1993). In practice, therefore the volume of futures contracts traded made it impossible for delivery of assets to taken place for more than a fraction of the futures contracts. It would be safe to say that such phenomenal growth would not have been possible had all the transactions ended up in delivery of the agricultural products.

This short analysis of the historical path of futures markets shows that the deliverability of the underlying asset was crucial for the legitimacy of futures, and indeed was embedded in the coded norms of gambling laws. While the concept of deliverability became crucial to futures markets, however, the actual practice in the markets rendered deliverability and indeed the sheer physicality of assets irrelevant. Moreover, it was crucial for the growth and prosperity of futures exchanges that deliverability would be possible in principle, but that it would not be performed in practice in all but a tiny minority of the transactions.

Regulators and exchanges in the network of qualification

By the 1960s, the tension between deliverability-in-principle and market practices has become an inherent part of the futures market. Yet, in the following few years that reality was about to change. In 1969–1971 the agricultural commodities markets witnessed a period of low trading volume (Yamey, 1985). That period coincided with the gradual demise of the currencies’ gold standard; a process that allowed currency exchange rates to float with less regulatory intervention and turned currencies into a much more risky asset than they had been previously. The futures exchanges, whose members were struggling due to the slow grains market, saw the more volatile currencies markets as a promising business opportunity and in spring 1971, Leo Melamed of the Chicago Mercantile Exchange (CME), the CBOT's archrival Chicago futures exchange, began to promote a plan for trading futures contracts based on foreign currencies (Melamed, 1988). Melamed's initiative, the International Monetary Market (IMM), an exchange that traded futures based on currencies, commenced trading in April 1972. The volatile currency markets contributed to high volumes in currency futures and following this initial success other agricultural futures exchanges developed similar contracts. Within months, the commodities regulator, the Commodity Exchange Authority (CEA) in the Department of Agriculture, received notices from several other American commodities exchanges about their intentions to apply for permissions to trade futures on non-agricultural products (R* interview).

Currency futures broke the virtually exclusive association that futures had had for over a century with agricultural products. In addition, non-agricultural futures signalled that the futures exchanges were aiming to expand their potential customers’ base by attracting investors from the general business community and in particular from the financial sector. These trends created a challenge for the regulatory process related to approving new contracts. Non-agricultural futures were an unknown territory for the Department of Agriculture and following the success of futures on currencies concerns were raised among the futures exchanges that the regulatory approval of contracts might be slowed considerably due to lack of knowledge. The severity of the regulatory challenge was manifested in the form requesting information about proposed contract that was sent to the CME by the Commodity Exchange Authority when permission to trade currency futures was sought. That form asked, among other things, about the size, location and condition of the warehouses in which the underlying commodities would be stored (Melamed, 1988).

These growing concerns about the suitability of the existing regulatory regime led CBOT in early 1973 to initiate an intensive lobbying offensive in Washington intending to persuade the American Congress to change the commodities regulatory structure. Concerns about the CEA's suitability to regulate the evolving futures world were not the only reason for the lobbying campaign. Since futures contracts were no longer associated exclusively with agricultural products there were now potential candidates to be transferred to another regulatory authority, one more suitable for the regulation of financial products. The possibility that the CBOT was particularly concerned about was that the American Congress would define futures on financial products as securities and consequently transfer the regulation of the contracts to the Securities and Exchange Commission (SEC).

The CBOT's concerns were not without basis. The same weak trading period in grain markets in the late 1960s that motivated the CME to develop the currency futures also drove the CBOT to fund research into the possibility of trading options on stocks. ² Options, being underlined by stocks, were regarded as securities and thus the SEC was given jurisdiction over that contract and the CBOT's proposal to trade stock options in an organized exchange underwent a long and exhaustive approval process by the SEC. The regulatory process took more than three years and acquired considerable effort by a team of lawyers and the exchange staff. Following those events, when the commodities exchanges’ lobbying effort succeeded in bringing about a Congressional hearings about the future of futures regulation, it promoted one message above all others: financial futures should not be regulated by the SEC. R*, at that period a CBOT lawyer who was heading the lobbying effort in Washington:

The main argument of the lobbying team was that the CME's currency futures opened the floodgates to a new torrent of financial futures and soon futures based on a variety of financial products would be proposed. That situation, the argument continued, may lead to an ‘administrative hell’ in which an exchange that would offer, say, futures on Treasury bills, crude oil and orange juice would have to go through separate approval procedures with the Treasury Department, with the Department of Energy and with the Department of Agriculture. Moreover, whenever a change to the contracts would be necessary, each of the agencies that regulated the underlying assets would have to be notified. Hence, the argument concluded, in a world where futures were no longer limited to agricultural products, regulation according to the underlying products was no longer feasible. Instead, the exchanges’ lobbying mission suggested that a new agency would be created that would have exclusive regulative authority over all futures contracts. In other words, a new principle of regulatory taxonomy was presented: regulation based on the type of contract instead of one based on the type of the underlying commodity.

The lobbying efforts were successful and in May 1974 the American Congress amended the Commodity Exchange Act (CEA) of 1936 and facilitated the creation of a new regulatory agency – the Commodity Futures Trading Commission (CFTC). The meaning of this new legal structure was that the CFTC was given exclusive rights over the regulatory approval of futures contracts. In other words, since futures could not be traded without such an approval the CFTC became an obligatory point of passage in the topology of the qualification network. That is, in the regulatory reality that was put in place by the 1974 Act, the CFTC has become an indispensable part of any path connecting a potential underlying asset, an exchange and a tradable futures contract.

The broad regulatory definition of futures created a conceptual blur between non-agricultural futures and between securities, and created uncertainty regarding the regulatory domains of financial markets. For example, being the exclusive regulator of futures the CFTC had jurisdiction over futures written on any asset, including, potentially, securities. The SEC, however, already had exclusive jurisdiction over securities. This may lead to the following dilemma: if futures on securities were to be proposed, which of the two, CFTC or SEC, would regulate them?

Such a qualification challenge was not merely hypothetical. In October 1975, the CFTC approved an application by the CBOT to trade futures on a financial product – mortgage-backed certificates known as GNMA's (US GAO, 2000: 5). ³ At the same time, an application by the CME was pending to trade futures on Treasury Bills (Johnson, 1976), and several of the other 12 American commodity exchanges also had applications for the trading of futures on Treasury bills at various stages of completion. It was the common opinion among the SEC's staff that futures on GNMA's would erode the distinction between securities and commodities. This was a trend that the SEC observed with much concern as it threatened its regulatory territory. This potential conflict between the interests of the two regulators would have had serious implications for the qualification of financial futures. Futures could not be traded without a regulatory approval and such an approval was not likely to be given while the regulatory identity of the contracts was disputed.

The challenge to the qualification of financial futures was embedded in a broader, deeply-rooted rivalry between the two regulators. It is true that the immediate issues that troubled the SEC's staff were related to the threatening possibility that the SEC's jurisdictional turf would be limited as a result of the new broad definition, but the concerns were also underpinned by a more general perception about the nature of commodities markets and their regulation. A senior staff member of the SEC's division of market regulation in the mid 70s described the SEC's staff attitude to the CFTC:

The views, common among the SEC's staff, that the SEC was a better regulator than the CFTC and that its staff was more professional than that of the new regulator should not be dismissed merely as a sign of inter-regulatory rivalry. These views belong to a broader perception that contributed to the shaping of qualification struggle between the two regulators. G*, who was the chief economist of the SEC when the CFTC was established, described the common view about commodities futures at the SEC:

This quote encapsulates both the nature of qualification and the contours of the qualification conflict that took place in the inter-regulatory sphere. As the history of markets for agricultural commodities indicates, assets gain their relative positions in the market and their qualities through the market practice.

Hence, if commodity trading is comparable to gambling, and the actual practice of trading is compared to pulling the handle of a slot machine, then the massage is clear: futures cannot be used for conducting sound, calculable investment and trading them is equal to the entirely luck-determined practice of gambling. This is not the entire message, however. At the time, the SEC and the CFTC were engaged in a struggle over the definition of futures contracts and the struggle was, in many respects, a zero-sum game. Any regulatory ‘territory’ lost by one regulator would have most likely be given to the other one. In this light, the quote above, and the common opinion about commodities should be regarded as an implicit opinion about the SEC, as much as it was a direct opinion about the CFTC. That is, if commodities trading is equal to gambling, then securities trading, the other activity in this dichotomy should be seen as legitimate investment.

Such views underlined the conflict about the qualification of futures. Yet, for the SEC, having a decisive impact on product qualification was not the ultimate goal. Qualification was seen as a step towards the more important goal, to distinguish its regulatory domain from that of the CFTC. The SEC, established in the early 30s, was the more well-established of the two regulators and had better chances to recruit influential supporters and to impose its definition. Given that the CFTC had exclusive statutory rights over regulation of all futures, however, and since the definition of commodities included financial products, the separation between the two regulatory fields of securities and commodities no longer existed. In this situation, the SEC's staff knew that in order to avoid the threat of having its regulatory domain taken over by the CFTC, the boundaries between the regulatory areas would need to be reconstructed.

This insight motivated the staff of the SEC to promote a message according to which there is a need to maintain the distinction between securities, the exclusive domain of the SEC, and between the types of assets that could be served as underlying for futures. Internal discussions in the SEC took place in the months after the American Congress passed the amendment to the Commodity Exchange Act and in December 10, 1975, Roderick Hills, the chairman of the SEC sent a letter to the chairman of CFTC suggesting that:

The two regulators were not the only actors involved in the political struggle. Due to the dense network of ties between the organizational actors, the debate over the shape of qualification had important implications for the exchanges. The futures exchanges, which were regulated by the CFTC, wished to expand their catalogue of contracts, and not to transfer to the stricter regulatory regime of the SEC. CBOT, the leading futures market of the time, directed an additional lobbying effort aiming to persuade the American Congress to incorporate into the Futures Act a wide as possible a definition of underlying assets, so as to include as many potential financial assets under the jurisdiction of the CFTC. R*, one of the leading commodities lawyers at the time coordinated CBOT's lobbying effort. R*:

The regulatory compromise solved the discursive aspect of the qualification process but the disputes over market practices did not end. The following six years, between 1974 and 1980, witnessed a long chain of ‘border incidents’ between the SEC and CFTC that flared around the regulatory approval processes of futures contracts based on financial assets, contracts that could be interpreted as being securities. In several cases, securities exchanges sued commodities exchanges for trading financial futures contracts, claiming that the futures contracts were actually securities in disguise and that the futures exchanges were illegally expanding their trading territory at the securities exchanges’ expense. As a general rule, the SEC and the CFTC provided advice and support to ‘their’ exchanges that were involved in cases, but mostly remained out of the courtrooms themselves. One exception was a court case related to the GNMA-based contracts mentioned earlier. CBOT traded GNMA futures since 1975 with considerable success. In 1981, there were approximately 2,293,000 sales of the contract, each representing $100,000 in unpaid mortgage principal (Board of Trade of City of Chicago v. SEC. 1982: 25,719). In early 1981, the Chicago Board Options Exchange (CBOE) submitted an application to its regulator, the SEC, to trade options on GNMA's. The CBOT, fearing that options on GNMA's would compete with its lucrative futures contract, sent a complain to the SEC and when, later on, the SEC approved CBOE's option contract, the CBOT filed a petition, objecting the approval, at a Federal court of appeals. The GNMA's case brought the two regulatory agencies into a direct confrontation in court. The court case resulted in a call by one of the judges for commencement of negotiations between the two parties:

The GNMA's case exposed the full extent of the regulatory struggle to the public and forced the SEC and the CFTC to start negotiations over the regulatory qualification of futures contracts.

As we shall see, the talks between the SEC and the CFTC that followed the court case discussed above eventually decoupled the exclusive link between derivative contracts and deliverable assets. Before we return to the historical narrative, however, a brief explanation about index-based contracts may be necessary. Stock indices are mathematical averages of the market prices of set groups of stocks at a given time. For example, a list of 500 stocks complied by Standard and Poor (S&P) is used as a basis for the S&P 500 index. On their own, indices are little more than mathematical representations of markets’ price levels. In contrast, when incorporated into financial contracts, like futures or options, stock indices can serve as a useful market tool. Index-based futures contracts require their owners (buyer) and its seller to pay or receive an amount of money proportional to the difference between the index level at the market on expiry and the index level stated in the contract. These contracts allow market participants to protect their holding against sudden drops in prices. For example, a contract that would grant its owner, say, $25 for each index point below a certain value at a certain date could serve as a safety net for investors. Similarly, a contract that would pay its owner $25 for each index point above a certain value would make a good device for profit-seeking traders who hope to gain from increasing prices. Index-based contracts would also have a significant advantage over contracts written on the basis of individual assets. Index-based contracts, being based on a relatively general ‘prices generator’ – the index – were less specific. That meant that as long as a person's portfolio is correlated with the price movements of the whole market, a derivative contract based on a market-wide index could be used to protect against losses to that person's portfolio.

These attributes explain why index-based contracts were attractive for investors. In addition, the nature of the assets underlying index-based contracts – indices – made them attractive from an organizational perspective. The SEC approved only a limited number of stocks to be used as bases for options. Consequently, as option trading became more popular, competition among exchanges for available stocks increased and so did the motivation of futures exchanges (regulated by the CFTC) for the approval of index-based futures. ⁴ Unlike stocks, indices represented a virtually infinite source of underlying assets and a promising source of growth for the exchanges. Index-based contracts also carried with them the promise of a light organizational burden as they allowed the exchanges to write one type of contract per market-wide index, instead of writing contracts on many individual stocks, thus their maintenance required less organizational resources than single stock contracts.

As promising as indices were, the deliverability problem stood between the exchanges and the realisation of index-based contracts. Index-based contracts could not guarantee the delivery of goods – such deliverable goods simply did not exist! Because no exchange of goods and funds was possible, index-based contracts could only be settled through the transfer of cash (cash settlement) and this was illegal in most American states. Indeed, as discussed earlier in the chapter, the evolution of commodities markets and the establishment of the notion of deliverability were contingent on each other.

Deliverability and Witching Hours: the contracts in the qualification network

The contrasting arguments of the SEC and the CFTC that evolved in the period leading to the GNMA case framed the range of qualification options in such a way that left the two regulators little choice other than to cooperate. Futures were put under exclusive jurisdiction of the CFTC, yet the products on which the indices were based – the stocks – were regulated by the SEC. This situation led the two regulators to the realisation that in spite of the rivalry between the two agencies the cash-settlement issue should be solved co-operatively. Even the more militant among the SEC staff realised that cooperation was necessary. For example, H*, a senior staff member of the SEC who was involved in the discussions, described SEC's the situation:

In the Spring of 1981, while the GNMA case was still discussed in court, John Shad and Philip Johnson were appointed as the chairs of, respectively, the SEC and the CFTC and according to Johnson, even before he took up office in Washington he contacted Shad and they both agreed to meet and discuss the overlapping regulatory areas of the two agencies (R* interview). ⁵

Although it was essential to solve the deliverability issue, tackling it proved difficult. Delivering the underlying was the practice through which the distinction between legitimate financial contracts and illegitimate, illegal, gambling was performed. As such, deliverability performed a crucial part of the contract qualities – their legality. From this legal and social perspective, doing away with the contractual obligation to have a delivery and replacing it with cash settlement would have amounted to obliterating the distinction between financial markets and casinos. Hence, the dilemma that the Shad and Johnson faced was: how could the heads of the two most important financial regulatory agencies in the US decide suddenly that cash-settlement was different from gambling, after their agencies have been condemning the practice since they were formed?

Shad and Johnson considered an approach that would circumvent the problems rather than tackle them directly. They simulated a scenario in which index-based contracts would include an obligation to deliver the assets. For example, if sellers of index-based futures would choose to exercise their contracts and deliver the underlying assets they would have to buy the stocks that composed the index that underlined the contract. Considering that indices are composed of any number of stocks ranging from 30 (Dow Jones) to a few hundreds (Standard and Poor's 100 or 500), and also that many series of futures would expire at the same date, deliveries of the underling assets would certainly result in sudden demands for stocks, leading to sharp surges in prices. Following such a scenario, Shad and Johnson understood that even if a fraction of index-based contracts would be settled by delivery, then the consequential transactions might still cause extreme volatility in the securities markets; a situation that both parties did not want to induce:

The ‘witching hours’ R* referred to were the last trading hours before the expiration of stock options. These contracts were written typically for periods of one month, three months, or six months and usually expired at the end of trading at the third Friday of the month in which the contract expires. Due to these standard time intervals, four times a year (on March, June, September and December) there occurred mutual expiration of the contracts. On those dates, at the hours before expiration, participants in stock markets witnessed huge price movements that seemed to be completely unrelated to the known information about the stocks. Such waves of sale and buy orders, frequently amounting to tens of millions of dollars, did not only leave the traders bewildered, but also caused significant losses to many veteran traders (Stein, 1986). These periods, immediately prior to the expiry of the contracts, came to be known as witching hours in the Wall Street parlance. The reason for this moniker was that market participants realized that during these near-expiration periods markets tended to display behaviour that was dramatically different from what was normally expected. It was gradually understood that usual patterns of trading did not hold true on those Fridays, and the markets behaved as if they were put under a mysterious spell.

Shad and Johnson's simulation exercise brought to the fore the role that deliverability played in the qualification and the realization of index-based contract. Shad and Johnson showed that the environments of deliverable assets and the non-deliverable assets were incompatible. According to the deliverable assets worldview, the absence of a delivery clause from financial contracts meant that those contracts were no different from betting. In contrast, in a market where index-based contracts are traded, an obligatory delivery would be equal to inviting a market crash. Therefore, what was a condition of the legal existence of trading in the deliverable assets world became unbearably dangerous in the world of non-deliverable index-based contracts.

This conceptual reconfiguration of the meaning of cash-settlement was a step in the creation of a new ‘language game’ in its Wittgensteinian sense (Lyotard, 1984). The two regulators created a constitutive language act (Barnes et al., 1996) and by so doing, by assigning a new discourse to the situation they resolved a century old concept. Namely, by connecting the trading practice of delivering underlying assets with the new index-based contracts, obligatory delivery was denounced as irrelevant and even dangerous. It has to be noted that although crucial for qualification, the interaction between the regulator was not the only site where index-based futures were born. The new environments in which financial futures and options were traded and the products themselves, the financial derivatives, gave the regulators with the essential vocabulary. Shad and Johnson were able to use the notion of ‘witching hours’ as a discursive tool in their discussions because such phenomenon had existed in organized options markets for several years before they met. Actors who traded financial futures and options created a new market nexus of practices and norms, constituting a new lingual and communicative medium, which was later recruited by the regulators. Together, this network of connections among traders, exchanges and regulators qualified index-based derivatives and turned the non-deliverable into tradable assets.

Discussion

The ability to design derivatives on the basis of market indices is arguably one of the reasons behind the explosive growth of these markets in the last decades and is an integral part of contemporary financial markets. Unlike other entities that are tradable assets in their own rights, however, indices are merely the products of mathematical procedures. Therefore, a crucial element in the qualification of index-based derivatives was the construction of the indices as legitimate underlying assets. As we saw, the qualification of index-based derivatives depended on a concentrated effort by a heterogeneous forum of agents (exchanges and regulators) that together transformed the cultural, political and practical aspects of commodities trading into qualities that were assigned to the new financial contract. This analysis serves as a basis for a more general discussion regarding financial markets. In particular, it raises a number of interesting questions. This concluding section will address two of them. First, the case of index-based derivatives illuminates the complex nature of the network of heterogeneous actors that constitutes contemporary financial markets, and in particular the connections between financial regulators and exchanges. Second, the dynamic process through which index-based derivatives were qualified raises the more general issue of the nature of agency in evolving market networks.

The qualification of index-based derivatives reveals the dense nexus of connections between regulators and exchanges. If a commonly accepted worldview were to be applied to this case then the various actors involved would probably be classified as belonging to one of two archetypical groups: regulators and regulated. A good example for the application of such a dichotomy to financial markets is the characterisation of financial economist Merton Miller (1986) of the constitutive powers of financial regulators. Miller suggests that many of the sophisticated financial derivative products existing today were developed because financial entrepreneurs were trying to break away from regulation. ⁶ According to Miller, new and innovative financial products did not fall under the existing regulatory definitions and thus allowed their users to be free from regulatory constrains such as reporting, or compliance with strict risk-mitigation practices. The ‘action-reaction’ hypothesis makes an implicit assumption about the nature of the financial entrepreneurship process. According to this assumption the regulators and the entrepreneurs are locked in an endless symbolic tennis game: the financial entrepreneurs launch a new type of product, which challenges the abilities of the existing regulatory regime, and the regulators react by changing the regulations.

As index-based derivatives show us, such sequential, bilateral model is inaccurate. Instead, we saw how regulators and exchanges form and dissolve coalitions that cross the boundaries between regulators and regulated. Ian Ayers and John Braithwaite offer an alternative to the mutually excluding division between regulation and deregulation, a dichotomy they regard as arbitrary and contrived (Ayers and Braithwaite, 1992). In their analysis, Ayers and Braithwaite predict that in complex fields, such as financial markets, the relations between regulators and regulated would tend to shift from a pattern of command and control, in which the regulators dictate to regulated the regulatory goal and the means through they are to be achieved, to an interactive pattern they refer to as ‘enforced self-regulation’. In the second phase, the regulatory goals are still chosen by the regulator, but the ways in which they are attained are dependent on the expertise of the regulated. The ‘enforced self-regulation’ scheme assumes implicitly that there is a separation between the ‘what’ element of regulation, the value-based, normative demands that underline the regulatory practice, and the ‘how’ element, the means through which these demands are tackled. Andrew Leyshon and Nigel Thrift (1996) follow a different disciplinary path but offer similar conclusions: they suggest that there is a discursive plurality in the interfaces between regulators and corporations, which brings about frequent changes in the content and boundaries of the economic system. If we add this insight to the hypothesis about enforced self-regulation, we see that complex regulatory fields, like the one that evolved around derivatives markets, call for very different analytic perspectives from the ones that divide the institutional agents to regulators and regulated. Instead, as the historical description implies, in an environment where it is necessary for the institutional agents to co-operate in order to influence the shape of the regulatory action, the nature of the ties among the various actors is as important as their motivations.

This theoretical perspective, which assumes that there is a dynamic, multi-focal regulatory environment, sets the stage for a concept of a more decentred regulation, such as the one offered by Julia Black (2001, 2002). According to Black, regulation is not a process that the state or its agents activate, but it is rather an outcome, or multiple outcomes, of interactions among actors. This approach differs radically from command and control approaches not only because it distributes the regulatory action among the agents but also because it also detaches the responsibility for the regulatory process from a single agent, or a group of agents, and transfers it to the relations among the different actors.

In other words, the question ‘who regulates?’ is replaced by the question ‘how is regulation performed?’ This question, which according to previous theoretical approaches to regulation was seen as a technical derivative of the worldview of the regulator, has moved to the fore. Since no single agent performs the regulation, but instead it is seen as an emerging organizational, political and, more recently, technological phenomenon, it cannot be reduced to a string of pre-determined procedures. Instead, the network of connections through which regulatory activity takes places should be regarded as the organizational infrastructure where rules, practices and procedures evolve and take shape. As the case of index-based derivates shows us, such a network includes the various interfaces between the actors, as well as the material and technological artefacts that they use.

The qualification process also raises questions about the nature of agency in financial markets. If we use the historical narrative in this chapter as a possible answer for the question: ‘who created index-based contract?’ we would find that the answer is far from straightforward. The regulators did not create the market for index-based contracts on their own because the options traders were the ones who developed the new conceptual meaning of non-delivery contracts. In particular, it was the notion of ‘Witching Hours’ that motivated Shad and Johnson to relinquish the demand for delivery. Similarly, it cannot be argued that the exchanges were the ones responsible for the creation of index-based contracts because, as the data show, critical parts in the qualification process took place within organizational settings in which the exchanges had relatively little influence.

A possible answer to the question of who created the first index-based derivatives can be that the network of connections that makes up the market is responsible for their creation. In other words, qualification provides us with an explanation why we should regard markets as a case of networked, distributed agency. This concept is taken from Marvin Minsky (Minsky, 1986). Minsky argued that intelligent action should be conceptualized as a large system of agencies that can be assembled together in various configurations. Edwin Hutchins (Hutchins, 1995) followed Minsky and expanded the concept to systems that include both humans and material objects. Hutchins showed that in a complex techno-social network the attribution of exclusive decision-making capacity to one actor would not be accurate. In such networks (Hutchins analyses aircrafts and boats), no single actor is the ‘commander’ while the rest are ‘subordinates’. Instead, whole network of humans and machines makes the decisions and performs the practices. In accordance, it can be said that the creation of markets for financial derivatives, a process that included a string of interpretations and decisions, could not be reduced to a simple ‘action-reaction’ narrative between the regulators and the exchanges. Indeed, the data shows us that each of the agents had a set of goals that was distinctly different from those of the rest. Instead, the connections between the differential actors were responsible for the transformation of index-based contracts from general concepts to tradable products.

Interviews

All interviews were conducted by the author, recorded and transcribed in full.

All interviewees’ names and details were kept anonymous.

R* – Chicago, February 2000

M* – Washington, DC., March 2001

G* – Washington, DC., April 2001

H* – Washington, DC., April 2001