Friday, June 18, 2010

Making OTC derivatives less OTC (BIS's Stephen Cecchetti)

The Squam Lake report 2 provides recommendations to, among other things, (i) protect against a systemic failure arising from a failure in the credit default swap (CDS) market, (ii) improve transparency in the CDS market and (iii) reduce the risk of runs on prime brokers and dealers. The recommendations focus on the use of central counterparties (CCPs), derivatives trade reporting, stricter regulation of liquidity requirements for dealers, and segregation of customer assets. Additional steps that would help further the goals of the report are: move end user derivatives trades onto CCPs; adopt standardised exchange traded derivatives for all risk types covered by OTC derivatives and higher regulatory capital requirements for non-standardised contracts; and establish safety-related registration of all financial products.

I would like to thank the Squam Lake Group for inviting me to participate in this important and timely conference. It is a pleasure to have the opportunity to discuss ongoing efforts to reform the regulation of the global financial system with all of you.

Let me start with a very short story about Bombardier, a Montreal-based transportation-equipment giant, which, among other things, used to make and sell snowmobiles. 3

In the winter of 1998, Bombardier came up with an innovative way to improve snowmobile sales. It offered buyers a $1,000 rebate should snowfall in 44 cities total less than a pre-set amount. In short, it offered a "snow-guarantee". 4 Sales increased by 38%.

Not wanting to retain the risk in this guarantee, Bombardier bought insurance. It hedged the snow-guarantee with an over-the-counter (OTC) weather derivative - an option based on a snowfall index - under which Bombardier would be paid the same amount as it would have to pay its customers in the event of low snowfall. In the end, it snowed enough that Bombardier did not have to pay their snowmobile customers, and its insurance contract did not have to pay off. Since the increase in sales generated by the snow-guarantee more than covered the cost of the derivative, it looked like a winning strategy all around. The transfer of risk through the derivative was welfare improving.

But, as we know, derivatives are not always so benign in their economic impact. We can find numerous instances in which they contributed to finanical instability. In the recent crisis, credit default swaps (CDS) were vilified - especially those written by the now infamous financial products division of AIG.

With that as a very brief introduction, I will now turn to the Squam Lake Group's report, in particular Chapters 9 and 10, which cover CDS, clearing houses, exchanges, dealers, prime brokers and runs.

I will then turn to some recommendations I would like to have seen in the report - all of which are only a small step beyond the ones that are there.

The recommendations in the Squam Lake report

Chapter 9: Credit default swaps, clearing houses and exchanges

Chapter 9 makes recommendations to strengthen the infrastructure of OTC derivatives markets. The proposals are aimed at two goals.

The first goal is to lower the risk of a systemic failure arising from a counterparty failure in the CDS market. Imagine a circle of people in which everyone sells to the person on the left the same OTC derivative. Each person is perfectly hedged because each has bought and sold the same security. But if just one person goes broke, the circle is broken. Likewise, in an OTC market, the failure of one firm can create a chain of failures ending in a complete collapse of the system.

The report recommends that financial firms be encouraged to use clearing houses, which I shall also refer to here as central counterparties (CCPs). The encouragement would come in part by requiring additional capital for contracts not cleared through a recognised CCP. If, in my example, the trading had been through a central clearing house, netting would have eliminated the systemic risk.

Importantly, the report notes that CCPs concentrate risks and so should be "well designed". That is, they should be required to have strong operational controls, appropriate collateral requirements and sufficient capital.

The second goal is to increase transparency in the CDS market. Doing so would improve the ability of market participants and regulators to identify "potential trouble spots". Transparency here is about information collection and dissemination.

To increase transparency, the group would target the index and single-name CDS contracts that are relatively liquid and standardised. In particular, the group suggests introducing trade-reporting similar to that in the TRACE system, which provides post-trade price transparency for US corporate bonds.

Underlying the transparency recommendation is the well-known fact that information asymmetries are often the fuel for financial panics. During the last quarter of 2008 as well as more recently, we saw contagion due to uncertainty over counterparty exposures - that is, not knowing who will bear losses should they occur. It follows that transparency is critical if we are to avoid panics driven by uncertainty.

Chapter 10: Dealers, prime brokers and runs

Chapter 10 focuses on reducing the risk of runs on prime brokers and dealers. To reduce the risk, the group recommends imposing liquidity requirements on systemically important banks and broker-dealers. And it would exclude from regulatory liquidity any short-term financing based on assets from counterparties or customers. To head-off attempts by prime brokers to avoid the proposed rules on segregation of customer assets, the report recommends that regulation on that point in major financial centres be at least as tight as it is in the United States.

Implementation: how far along are we?

It should be a source of some satisfaction to the Squam Lake Group that a number of its recommendations have already been taken to heart by policymakers, regulators and, to some extent, market participants. 5 That said, complete adoption is still some distance away. Here is a status report on four of the more important items.

First, we now have trade depositories for CDS and interest rate derivatives. These depositories feature electronic databases of open OTC positions and publication of aggregate numbers on volumes and market activity. But very little information on exposures is publicly available. 6 And while there is some pre-trade price transparency, there no post-trade price transparency at all.

Second, when it comes to the actual use of CCPs, the interest rate swap market is the only OTC derivatives market in which market participants and financial institutions rely on central clearing in a systematic way. The London-based CCP called SwapClear covers roughly 45% of the total interest swap market. The use of clearing houses for other OTC derivatives contracts, however, ranges from very limited to non-existent.

Third, under the Basel III process, regulators have proposed more stringent liquidity requirements, but they have yet to be adopted.

Fourth, the need to segregate broker and client assets addresses not only the need to prevent rehypothecation (ie the pledging of securities in customer margin accounts as collateral for a brokerage's loan) but also the need to prevent co-mingling (ie using the same account for) broker and client assets. Regulations to address the co-mingling problem have been in place for some time, and enforcement has recently become more vigilant. 7 Moreover, anecdotal evidence suggests that hedge funds and other clients have started to insist that their assets not be co-mingled with prime broker assets. I should also note that, with respect to rehypothecation, private contracting practices are beginning to rule it out in response to the Lehman Brothers bankruptcy.

Recommendations I would like to have seen in the report

I agree with most of the report's recommendations, but I believe they do not go far enough. Here are three more proposals that I wish had been there.

1. Corporate derivatives users should be required to rely on central clearing houses

If we are to fully reap the benefits of having CCPs, central clearing houses will have to cover large swaths of the derivatives market, both in terms of counterparties and volume.

Current draft legislation in the United States and the European Union requires important financial institutions to trade through CCPs, as is recommended by the group. But end users could be exempt. The argument for the exemption is that, if end users have to use central clearing, they will have to post more collateral, which would drastically increase their cost of using derivatives for risk management. The result would be too little hedging.

There may be something to this. But the argument implicitly assumes that end users (such as Bombardier in the example I gave at the beginning) are not being charged for the credit risks that their counterparty takes on by not asking for collateral.

I find it hard to believe that banks do not charge their clients for the services they provide. Just as retail clients pay banks for their free checking accounts in one way or another, I have no doubt that end users are already paying for the services they receive. In fact, bankers are known to derisively refer to these services as "fee checking accounts" because of the hidden nature of the charges. What is true in checking is surely true for derivatives. That is, banks are surely compensated via prices, higher bid-ask spreads or higher costs for other services provided by the bank. Because of the opacity of the OTC market, it is very difficult for end users to know what they are actually paying.

In my view, end users wrongly perceive central clearing houses as being more expensive than the current solution simply because (i) CCPs allocate costs directly to the services provided and (ii) CCP costs are transparent.

2. Market participants should be encouraged to create standardised exchange traded derivatives for all risk types currently covered by OTC derivatives. Regulated financial institutions should have higher capital requirements for non-standardised contracts.

Current reforms focus on ensuring central clearing for "standardised" contracts. But to define the contracts that they think should be centrally cleared, the financial industry is using the phrase "eligible " instead of "standardised". This subtle rephrasing must not become a loophole that allows them to retain the status quo ante.

The argument typically advanced in favour of non-standardised OTC derivatives is that markets need tailored hedging tools. I would argue, however, that one can design standardised contracts for nearly all risk types and that standardised contracts are very good hedging tools.

The choice between a tailored hedge and a standardised hedge boils down to a choice between:

    1. a perfect hedge with a high bid-ask spread and counterparty risk, or

    2. an imperfect hedge with a low bid-ask spread and basis risk.

Let me give two examples.

First, weather derivatives. As many of you probably know, there are active exchange-based markets for US weather derivatives. The risks covered include hurricanes, snowfall, extreme temperature and frost. 8 Some of the contracts look remarkably like the one used by Bombardier to insure the snow guarantee it gave in 1998 to the buyers of its snowmobiles.

Clearly, standardised, exchange-traded weather derivatives are unlikely to be a perfect hedge. But a decade ago, they were very much an OTC product. In fact, Bombardier was a pioneer. But who sold the snowfall-option that Bombardier bought? Would it surprise you if I told you the seller was Enron? A perfect hedge using a bespoke OTC derivative - but alas, with counterparty risk.

The second example is exchange-traded futures and options. Very large and liquid derivatives markets exist for US and German government bonds. Contracts have been successfully standardised to provide good but not perfect hedges for these two most liquid government bond markets.

A few observations on the futures and options contracts for US and German government bonds:

1. None of them corresponds to an existing outstanding bond. Rather, each is designed as a contract on a non-existent or synthetic bond.

2. The delivery schedule matches the open interest with the deliverable quantity for bonds that do exist. The conversion factors are calculated as a fraction of the synthetic bond, taking into account the relative riskiness of the delivered bond.

I am hard pressed to understand why the same approach cannot be used to standardise interest rate, foreign exchange and credit derivatives. It would be well worth the effort simply to explore whether one could persuade market participants to seriously attempt such standardisation.

3. Consideration should be given to the introduction of product registration for financial contracts.

I can't help wonder if one should consider a scheme for financial contracts akin to drug regulation. The idea is to balance the need for innovation in financial instruments with the need to limit the capacity of any individual security to weaken the whole system.

The solution is some form of product registration that would constrain the use of instruments according to their degree of safety. The safest securities would be available to everyone, much like non-prescription medicines; next would be financial instruments available only to those with a licence, like prescription drugs; then would come securities available only in limited amounts to qualified professionals and institutions, like drugs in experimental trials; and securities at the lowest level of safety would be deemed illegal. An instrument could move to a higher category of safety only after successful tests analogous to clinical trials. Testing would combine real-world issuance in limited quantities with simulations of how the new instrument would behave under stress. Such a certification system creates transparency. As in the case of pharmaceutical manufacturers, so it should be for manufacturers of financial products: there must be a mechanism to hold them accountable for the quality of what they sell. That mechanism will increase the responsibility of financial institutions for assessing the risk of their products.

Concluding remark

I would like to congratulate the Squam Lake Group on an excellent and thoughtful report on how to fix the financial system. I remain hopeful that many if not most of the recommendations in the report will be implemented and that they will help improve the resilience of the financial system.

Thank you for your attention.


1 I thank Jacob Gyntelberg for his contributions to this presentation. The views expressed here are those of the author and do not necessarily reflect those of the BIS.

2 K French, M Baily, J Campbell, J Cochrane, D Diamond, D Duffie, A Kashyap, F Mishkin, R Rajan, D Scharfstein, R Shiller, H S Shin, M Slaughter, J Stein and R Stulz, The Squam Lake Report: Fixing the Financial System, Princeton University Press, 2010.

3 The Bombardier Recreational Products division was sold in 2003 and is today an independent company owned by, among others, members of the Bombardier family, Bain Capital and Caisse de dépôt et placement du Québec.

4 The rebate would be paid if the snowfall was less than half of what it had averaged in the past three years.

5 In May 2010, the Committee on Payment and Settlement Systems (CPSS) and the Technical Committee of the International Organisation of Securities Commissions (IOSCO) jointly released two companion reports that provide guidelines for the establishment of CCPs as well as trade repositories for OTC derivatives: "Guidance on the application of the 2004 CPSS-IOSCO Recommendations for Central Counterparties to OTC derivatives CCPs: Consultative report", CPSS Publications, no 89, www.bis.org/publ/cpss89.htm; and "Considerations for trade repositories in OTC derivatives markets: Consultative report", CPSS Publications, no 90, www.bis.org/publ/cpss90.htm.

6 The Depository Trust & Clearing Corporation operates the trade repository and data warehouse for CDS, while TriOptima operates a trade repository for interest rate derivatives. A trade repository for equity derivatives is expected to become operational in July 2010. In addition, aggregate data based on surveys are published by the International Swaps and Derivatives Association (ISDA); and by the Bank for International Settlements in collaboration with a number of central banks.

7 In early June 2010, the United Kingdom's Financial Services Authority (FSA) fined JPMorgan a record £33 million because they co-mingled accounts. Later, the FSA fined divisions of Astaire Group and Close Brothers also for failing to segregate clients' money from their own.

8 www.cmegroup.com/trading/weather/.

No comments: