More on OTC derivatives
I have several comments by email on my blog post yesterday on OTC derivatives. This post responds to some of them and adds some more material on the subject.
One of the papers that I did not refer to in yesterday’s post was a paper by Riva and White on the evolution of the clearing house model for account period settlement at the Paris Bourse during the nineteenth century. Streetwise Professor linked to a conference paper version of this in his post while Ajay Shah pointed me to an NBER version of the same paper.
The account period settlement at the Paris Bourse was similar to the ‘badla’ system that prevailed in India until the beginning of this century. All trades during a month were settled at the end of the month so that the stock market at the beginning of the month was actually a one month forward market. At the end of the month, the settlement could be postponed for another month on paying the price difference and a market determined backwardation or contango charge. In fact, this system for trading individual stocks continued even after the introduction of stock index futures (CAC 40) in the Paris market. Crouhy and Galai, “The settlement day effect in the French Bourse,” (Journal of Financial Services Research, 1992) provide a good description of this market and explore the working of the cost of carry model in this market.
Coming back to Riva and White, they document the emergence of a clearing house model in which the Paris Bourse guaranteed all trades on the exchange. The bourse not only guaranteed settlement of trades between two brokers but also repaid the losses suffered by the defaulting broker’s clients (except during a period from 1882 to 1895). This clearing house guarantee was supported by a capital requirement for all brokers and by a guarantee fund, but not by any margins. The entire process seems to have been driven by the government and the central bank.
By contrast, the US futures exchanges introduced initial and variation margins well before they introduced the clearinghouse in 1883 as documented in the Kroszner paper that I mentioned in my post yesterday. The existence of margins eliminates most of the moral hazard problems that plagued the clearing house in Paris and required state intervention of some form or the other. The US thus saw a private ordering emerging without any involvement of the state.
India ran the ‘badla’ system in the nineteenth and through most of the twentieth century without any margins and without any clearinghouse guarantee. While France solved the risk problem in its usual dirigiste style and the US solved it using private ordering, India seems to be a case of state failure and market failure until the last decade of the twentieth century. In fact, the problem was solved only a few years before the abolition of ‘badla’ itself.
Now I turn to some other papers relevant to the regulation of OTC derivatives.
Viral Acharya and several coauthors have written extensively on the regulation of OTC derivatives, and I will mention two. Acharya and Engle have a nice paper that explains the key issues in the context of the proposed US legislation.
Acharya and Binsin have a conference paper explaining how an exchange is able to price counterparty risk better than the OTC market because it is able to see the entire portfolio of the counterparty. Streetwise Professor criticizes this on the ground that exchanges charge the same price to everybody and do not discriminate. I think this criticism is incorrect – exchanges do not discriminate in the sense that they apply the same risk model to everybody, but the standard SPAN type model is a portfolio model where the margin is not on an individual position but on a portfolio. The incremental margin requirement for any position thus depends on what else is there in the portfolio. Thus the risk is priced differentially.
The Acharya and Binsin paper must be read in conjunction with a paper by Duffie and Zhu who show that the efficiency gain from central clearing is best realized when there is a single clearing house for all derivatives and the gains may disappear if there are separate clearing houses for different products and even more so when there are competing clearing houses for the same product. This in my view is only an efficiency issue and does not detract from the reduction of systemic risk from the use of central clearing.
For those interested in data about the magnitudes involved in these markets in terms of risk and collateral requirements, a good source is an IMF Working Paper on “Counterparty Risk, Impact on Collateral Flows, and Role for Central Counterparties.” For more detailed information about the CDS market, there is an ECB paper on “Credit default swaps and counterparty risk”
Posted at 6:10 pm IST on Wed, 13 Jan 2010 permanent link
Comments