Those who have read SWP regularly will have deduced that I have a historical bent. Although historians and economists (and other social scientists) tend to think and analyze problems in very different ways, I think that history is quite often a useful source of material on which to test theories and conjectures, and a source of inspiration and ideas useful in the generation of new theories and models. It also frequently proves quite useful in debunking assertions about the necessity or sufficiency of a certain factor causing some outcome.
Today’s historical excursion examines a couple of episodes in the past relating to clearing. Specifically, two occasions in which important exchanges, the London Metal Exchange and the Chicago Board of Trade, adopted central clearing under government pressure only after a long period of resistance. In each case, the memberships and/or leaderships of these exchanges steadfastly refused to introduce clearing until forced to do so by regulators. In each of these cases, the regulators’ rationales were tangential, at best, to the actual economic purpose and effects of a clearinghouse. It seems that as various government entities are pushing a clearing solution for the CDS market that we are in the process of repeating history.
A central counterparty/clearinghouse was first adopted in the US in 1891 in Minneapolis. From time to time in the following years, some members of the Chicago Board of Trade proposed that the exchange adopt the “Minneapolis Model.” Each time, either the membership or the board of directors rejected these proposals.
All of the merits of central clearing were pretty well understood. In particular, CBT members understood that clearing would reduce the amount of money that would be tied up in margins–although some considered this a bug, not a feature.
The standoff between the advocates and opponents of clearing continued literally for decades. The impetus for the move that actually resulted in the creation of the Board of Trade Clearing Corporation originated in the passage of the Grain Futures Act of 1922. This Act gave the Secretary of Agriculture considerable power over grain exchanges. The Secretary could put an exchange out of business by revoking its designation as a contract market.
In 1925, William Jardine was Calvin Coolidge’s Agriculture Secretary. He was convinced that large speculators were manipulating the grain markets, to the detriment of farmers, and wanted to do something to stop it. In another foreshadowing of current events, Jardine was convinced that identifying who held large positions would facilitate deterrence. He further thought that a clearinghouse would achieve this objective because it would require the registration of all positions in order to perform its work.
Of course, it would have been possible to register positions without creating a central counterparty. Indeed, such a change to clearing procedures had been proposed, but not adopted, in 1911.
The directors of the CBT continued to balk at creating a clearinghouse, until two influential members of the exchange, L.F. Gates and J.C. Murray, who served on the exchange’s Grain Exchange Legislative Committee, wrote a letter warning of dire consequences if the exchange failed to act. Gates and Murray said that failure to implement the clearinghouse plan could lead to a revocation of contract market status. Stung into action by their warning, the directors approved the creation of a clearinghouse (and a Business Conduct Committee to police the exchange’s members.)
Jim Moser argues that Jardine’s pressure was not what caused the CBT to adopt the central counterparty. He notes that the exchange could have satisfied Jardine’s demands for transparency by registering every trade, but not providing a central guarantee. He concludes that the directors’ reluctance to adopt the CCP was based on privacy concerns that Jardine alleviated. From this he infers that the directors’ objection to the clearinghouse was not the risk sharing aspect of its operation, but the possibility that its operation could result in disclosure of position information.
Perhaps. But it seems that if there was a strong demand for risk sharing via a CCP, the exchange could have found away to address the privacy concerns. Thus, another, and in my view more straightforward, interpretation of events is that there was not a uniform desire for a CCP, and there was strong opposition to the concept; and that Jardine’s threat to shut down the exchange compelled the directors to pacify him by adopting a method that he had endorsed–the Minneapolis Method.
That’s one of the problems with history–it seldom permits definitive answers as to cause and effect. But regardless of the interpretation, the extreme reluctance of the Board to adopt a CCP suggests that the concept is not nearly as beneficial as its advocates suggest. To be sure, there were legal and privacy objections, but (a) where there’s a will, there’s a way, and (b) it is often the case that explicit objections camouflage the true reasons for opposition. And to me, the most straightforward interpretation of the facts is that the CBT adopted a central counterparty not because of its inherent economic benefits, but only as result of government pressure motivated by considerations completely unrelated to the costs and benefits of sharing default risks. The failure of the CBT to adopt a central counterparty voluntarily is consistent with the view that this method of allocating default risk has costs that exceed the benefits.
The LME was another exchange that adopted central clearing only when its very existence was threatened by government action. The LME had suffered a major crisis in 1985-1986 with the collapse of the International Tin Council’s price support program. Tin prices collapsed, contracts were breached, and many LME members suffered huge losses, with some closing their doors. Prior to the Tin Crisis, the LME had been strictly a principals’ market, with all dealings done on a bilateral basis just as is the case in OTC markets today. The LME clearinghouse was merely a mechanism for netting payments and deliveries, similar to a bank clearinghouse. It was not a CCP. Indeed, the LME had consciously resisted creation of a CCP for years after this practice had become standard in both American and British commodity markets.
Remarkably, the LME’s resistence to the idea continued after the Tin Crisis. Thus, even a major series of dealer defaults was not sufficient to convince the membership that a central counterparty was a superior method of sharing default risks. The LME adopted a CCP only after the newly formed U.K. Securities and Investment Board (SIB) made it clear that it would deny the LME its license to operate unless it introduced an independent CCP.
Most of the benefits attributed to CCPs are private benefits that accrue to their members–lower collateral, more efficient sharing of default risk, prioritization of derivatives counterparties over other creditors. If this was all there was to the story, derivatives markets participants should adopt a CCP without fail. But, as these historical episodes demonstrate, they don’t. Indeed, they demonstrate that the resistance to the idea can be so strong that only intense government pressure can overcome it. Thus, there must be more to the story. There must be some costs sufficiently large to overwhelm the private benefits commonly attributed to a CCP. These costs, unfortunately, are seldom–if ever–the subject of serious analysis. They should be. A reasoned judgment about the efficiency of a CCP must be predicated on a judicious weighing of both costs and benefits. Presumably 1920s grain traders and 1980s metals traders whose livelihoods depended on the efficient operation of their markets made such a judicious appraisal, and decided that the costs outweighed the benefits. That should give advocates of CCPs pause, and spark a more thoughtful consideration of their full effects.
Throughout the period of explosive growth in OTC derivatives, market participants have arrived at a similar judgment. The concept of clearing is well known. The subject has been mooted–and rejected (for the most part). That judgment should not be overridden without excellent justification.
Heretofore, that justification has been totally lacking, but government pressure to adopt a CCP for credit derivatives, and perhaps eventually other derivatives, is unrelenting. There are repeated invocations of “systemic risk” and the bandying about of scary (and often mixed) metaphors–“weapons of mass destruction,” “time bombs,” etc. But the concept of systemic risk is never precisely defined, nor do the advocates of clearing of OTC derivatives present any serious argument as to how a CCP would reduce that which is scary (but undefined.)
Typically, the argument stops at “a CCP would net which would reduce replacement costs in the event of a default, thereby lowering the costs that other dealers pay when one dealer defaults.” That argument is incomplete, not generally true, and arguably false. It fails to recognize that netting just redistributes default costs from derivatives counterparties to other creditors, some of whom may be systemically important; it does not consider that the formation of a CCP may increase the sizes of positions that traders hold, thereby increasing the potential for default losses; it ignores the effects of a change in risk allocation on asymmetric information and the incentives to monitor counterparty risk; it overlooks the fact that a CCP, by effectively guaranteeing customer positions, requires dealers that are members of a clearinghouse to bear the costs of a defaulter’s entire net position (including trades done with non-member customers), rather than merely their direct trades with the defaulting member.
If systemic risks are an increasing function of the default losses that dealers incur, it is a straightforward exercise to create examples in which such losses (and hence systemic risks) are higher, rather than lower, with a CCP. This is true even if one ignores the effects of asymmetric information. The monomoniacal focus on netting obscures the myriad effects of the mutualization of default risk, and leads to a wild overestimate of the efficacy of clearing in reducing systemic risks.
The government forcing grain traders or metals traders to adopt an inefficient method of allocating default risks is bad enough, but the social costs of these actions were likely bearable because these markets were small in the scheme of things. Forcing the adoption of clearing in the massive OTC financial derivatives markets is another thing altogether. Here the costs of a mistake could be huge. The incomplete and faulty analysis that is being used to justify the creation of a CDS CCP greatly increases the odds that its adoption would in fact be such a mistake.
The seeming compulsion to “do something” in the face of a burgeoning credit crisis threatens to hasten the implementation of an ill-considered plan with major systemic consequences. Let’s hope that that compulsion is replaced by contemplation and measured action before it is too late.