U.S. Securities & Exchange Commission
SEC Seal
Home | Previous Page
U.S. Securities and Exchange Commission

Testimony Concerning Credit Default Swaps

by Erik Sirri
Director, Division of Trading and Markets
U.S. Securities and Exchange Commission

Before the House Committee on Agriculture

October 15, 2008

Chairman Peterson, Ranking Member Goodlatte, and Members of the House Committee on Agriculture:

I am pleased to have the opportunity today to testify regarding the credit default swap (CDS) market. The over-the-counter (OTC) market for CDSs has drawn together some of the world's important financial institutions into a complex web. These institutions have diverse roles in the market for CDSs, including as market makers, hedgers, and speculators who take proprietary positions in the credit risk of the underlying entity. The CDS market has experienced explosive growth in recent years. As of the end of the first half of 2008, the total notional value of CDSs is estimated to be approximately $55 trillion, according to the International Swaps and Derivatives Association (ISDA), doubling its size in only two years. AIG alone is reported to have sold over $440 billion of CDS protection on a notional basis. It is important, however, to keep in mind that notional value is not a precise measure of the total risk exposure.

The SEC has a great interest in the CDS market because of its impact on the debt and cash equity securities markets and the Commission's responsibility to maintain fair, orderly, and efficient securities markets. These markets are directly affected by CDSs due to the interrelationship between the CDS market and the claims that compose the capital structure of the underlying issuers on which the protection is written. In addition, we have seen CDS spreads move in tandem with falling stock prices, a correlation that suggests that activities in the OTC CDS market may in fact be spilling over into the cash securities markets.

The Commission's current authority with respect to OTC CDSs, which are generally "security-based swap agreements" under the CFMA, is limited to enforcing antifraud prohibitions under the federal securities laws, including prohibitions against insider trading. The SEC, however, is statutorily prohibited under current law from promulgating any rules regarding CDS trading in the over-the-counter market. Thus, the tools necessary to oversee this market effectively and efficiently do not exist.

SEC staff are actively participating with other financial supervisors and industry members in efforts to establish one or more central counterparties, or CCPs, for credit default swaps. Improving market infrastructure and the ability to monitor the CDS market, for example by establishing a CCP, would be an important first step in reducing systemic and operational risks in the market. The Commission staff fully supports these efforts.

In addition, when Chairman Cox spoke before the Senate Committee on Banking, Housing, and Urban Affairs three weeks ago, he called the lack of regulation of the CDS market a "cause for great concern." The CDS market's considerable size and importance to the financial system, particularly during periods of significant market turbulence, compel greater oversight. Recent credit market events, notably the default by Lehman Brothers and the intervention by the Treasury with respect to Fannie Mae and Freddie Mac, have required an ad hoc response by market participants, generally under the auspices of industry groups such as ISDA. In all three cases, the industry had to orchestrate an auction to permit cash settlement of CDSs intended to be settled through physical delivery of bonds as a means to reduce operational frictions. In fact, the industry had to meet under the auspices of ISDA to even determine with certainty that the Treasury actions with respect to Fannie Mae and Freddie Mac were an event of default for purposes of credit default swaps written on the debt securities of those two reference entities. While ad hoc approaches have worked remarkably well to date, Chairman Cox and others have questioned whether the size and importance of the market make more oversight, including a more developed infrastructure, prudent.


As you know, CDSs, like other credit derivatives, are a type of financial contract whose value is based on underlying debt obligations. By their very nature, CDSs transfer risk rather than directly raise capital in the way a bond or stock does. However, the transference of risk can indirectly aid in raising capital. A CDS can be tied to the performance of the debt obligations of a single entity or security, or — with more complex CDSs — an index of several such entities or securities. In a CDS, as in an insurance contract, the CDS "buyer" is buying protection and the CDS "seller" is selling protection against a default or other credit event with respect to the underlying debt obligations. The buyer pays the seller a premium for this protection, and the seller only pays the buyer if there is a default or other credit event that triggers the CDS contract. The premium — cost of protection for the buyer — increases as the risk associated with the underlying obligation increases. In other words, as the creditworthiness of the underlying entity goes down, the cost of protection goes up.

CDSs are executed bilaterally with derivatives dealers in the OTC market, which means that they are privately negotiated between two sophisticated, institutional parties. They are not traded on an exchange and there is no required recordkeeping of who traded, how much and when. The dealers include more than a dozen large, globally active banks. London and New York are the centers of CDS trading. In addition to the dealers, active participants in the CDS market include hedge funds and registered investment companies, as well as insurance companies, among others.

Although CDSs are frequently described as insurance (buying protection against the risk of default), they, in fact, also are used by investors for purposes other than hedging. Institutions can and do buy and sell CDS protection without any ownership in the entity or obligations underlying the CDS. In this way, CDSs can be used to create synthetic long (or short) positions in the referenced entity. Because a CDS transfers the risk of default on debt obligations from the buyer to the seller, a CDS buyer is analogous to being "short" the bond underlying the CDS. Whereas a person who owns a bond profits when its issuer is in a position to repay the bond, a CDS buyer profits when, among other things, the bond goes into default. Conversely, a CDS seller can be said to be taking a "long position" on the underlying credit. In other words CDSs may be used to replace cash bonds in establishing trading positions in a credit.

Indeed, for a typical corporate debt issuer, the notional amount of activity in OTC derivatives tied to its debt or credit can be substantially larger than the outstanding balance (principal amount) or trading in the issuer's actual debt securities. CDSs, therefore, can be used to manage the risk of a portfolio of assets or to mitigate a firm's exposure to an entire financial institution. Writers of CDSs can develop concentrated exposures to particular credits, which if large enough, could raise serious systemic issues for the global financial system.

Establishing a Central Counterparty for the CDS Market

Although the clearance and settlement of CDSs are not currently regulated, the SEC has regulated the clearance and settlement of securities, including derivatives on securities, since the Securities Acts Amendments of 1975. The SEC has registered approximately 20 clearing agencies under the Exchange Act, and SEC staff have performed many compliance inspections and program reviews. During the more than 30 years the SEC has regulated clearing agencies, the SEC has continued to develop expertise in this area, and no registered clearing agency under the securities laws has failed to perform its obligations or contributed to the failure of another institution through poor performance.

As noted above, there are important relationships between the securities markets and the market for CDSs. Accordingly, the SEC is participating in discussions with the Federal Reserve Board (Fed), the Federal Reserve Bank of New York, the Commodity Futures Trading Commission (CFTC), and industry participants to create a central counterparty (CCP) for credit default swaps. Last week, senior SEC staff attended meetings with other regulators, hosted by the Federal Reserve Bank of New York, at which industry members discussed their proposed CCPs. There are currently four potential CDS central counterparties: Eurex, NYSE Euronext, CME Group/Citadel, and Intercontinental Exchange/The Clearing Corporation. The SEC staff will continue to work in close cooperation with the Fed, the Federal Reserve Bank of New York, and the CFTC to facilitate the creation of at least one CCP.

As addressed in the testimony of my colleague, Dr. James Overdahl, before the Senate Subcommittee on Securities, Insurance, and Investment on July 9 of this year, a CCP could be an important step in reducing the counterparty risks inherent in the CDS market, and thereby help to mitigate the potential systemic impacts. As I noted earlier, CDS are bilateral contracts between market participants. As is the case with all contracts, each party to the transaction needs to be concerned about the willingness and capacity of the party on the other side to perform its obligations.

To illustrate how CDSs work, suppose that Dealer X sells protection on ABC to Dealer Y. Dealer Y needs to be concerned about Dealer X's ability and willingness to perform in the event of a default or other credit event by ABC. While the risk being transferred from Dealer Y to Dealer X relates to the credit quality of ABC, Dealer Y, while shedding risk related to ABC, is taking on counterparty risk to Dealer X. Market participants manage this counterparty risk using a variety of tools, including marking positions to market and posting collateral, as well as documentation that provides for other mitigants.

A central counterparty could further reduce systemic risk by novating trades to the CCP, meaning that Dealers X and Y no longer are exposed to each others' credit risk. In addition, the CCP could reduce the risk of collateral flows by netting positions in similar instruments, and by netting all gains and losses across different instruments. So, instead of Dealer Y having a large volume of trades, some offsetting, with many counterparties, Dealer Y could have a single net position in ABC with the CCP. Likewise, Dealer X could have a single net position in each underlying credit, perhaps related to a large volume of individual trades, with the CCP. By replacing the current "web" of CDS exposures with a "hub and spokes" architecture, a CCP could vastly simplify containing the failure of a major market participant.

Moreover, a CCP could further reduce risk through uniform margining and other robust risk controls over its exposures to its participants, including specific controls on market-wide concentrations that cannot be implemented effectively when counterparty risk management is decentralized. A CCP also could aid in preventing the failure of a single market participant from destabilizing other market participants and, ultimately, the broader financial system.

A CCP also could help ensure that eligible trades are cleared and settled in a timely manner, thereby reducing the operational risks associated with significant volumes of unconfirmed and failed trades. It may also help to reduce the negative effects of misinformation and rumors that can occur during high volume periods, for example when one market participant is rumored to "not be taking the name" or not trading with another market participant because of concerns about its financial condition and taking on incremental credit risk exposure to the counterparty. Finally, a CCP could be a source of records regarding CDS transactions, including, for each day, by underlying reference entity, the identity of each party that engaged in one or more CDS transactions. Of course, to the extent that participation in a CCP is voluntary, its value as a device to prevent and detect manipulation and other fraud and abuse in the CDS market may be greatly limited.

There is no guarantee, however, that efforts to establish CCPs or other mechanisms would achieve success, or that OTC CDS market participants would avail themselves of these services. Even if a dealer does participate in the CCP, trades the dealer elects to do away from the CCP would escape its risk management oversight. Accordingly, one should not view a CCP as a panacea for concerns about the management of exposures related to credit derivatives. Even with a CCP, preventing a systemic risk buildup would require dealers and other market participants to manage their remaining bilateral exposures effectively, and the dealers' management of their bilateral exposures would require ongoing supervisory oversight. Nonetheless, developing a CCP for clearing CDSs would be an important step in accomplishing this goal.

Exchange Trading of CDSs

It is not uncommon for derivative contracts that are initially developed in the OTC market to become exchange-traded as the market for the product matures. While the contracts traded in the OTC market are subject to individual bilateral negotiation, an exchange efficiently creates a market for a standardized form of the contract that is not subject to individual negotiation (other than price and quantity). These standardized exchange-traded contracts typically coexist with the more varied and negotiated OTC contracts. In this regard, we note that last year the Commission approved a proposal by the Chicago Board Options Exchange to list and trade Credit Default Options ("CDOs") and Credit Default Basket Options. The CDOs are modeled after CDSs and structured as binary call options that settle in cash based on confirmation of one or more specified adverse credit developments (such as payment default) involving obligation(s) referenced in the CDO, such as a debt security.

Some of the prospective central counterparties for CDSs also propose offering some type of trading facility. Exchange trading of credit derivatives could add both pre- and post-trade transparency to the market that would enhance efficient pricing of credit derivatives. Exchange trading also could reduce liquidity risk by providing a centralized market that allows participants to efficiently initiate and close out positions at the best available prices.

Primary Regulatory Concerns

CDSs serve important purposes as a tool that can be employed to closely calibrate risk exposure to a credit or a sector. CDSs can be especially useful for the business model of some financial institutions that results in the institution making heavily directional bets, and others — such as dealer banks — that take both long and short positions through their market-making and proprietary trading activities. Through CDSs, market participants can shift credit risk from one party to another, and thus the CDS market may be an important element to a particular firm's willingness to participate in an issuer's securities offering.

CDSs also raise a number of regulatory concerns, including the risks they pose systemically to financial stability and the risk of manipulation.

With regard to financial stability, the OTC CDS market, together with other derivative products, has drawn together the world's major financial institutions and others into a deeply interconnected network. Their activities in the CDS market generate significant market, credit, and operational risk that extend beyond the willing counterparties to the CDS transaction. As I described earlier, the buying and selling of default protection through CDSs creates short and long exposures — market risk — to the index, debt security, or other obligations referenced in the CDS contract. At the same time, the buying and selling of default protection creates credit risk exposure to counterparties. The default of one major player therefore impacts not only the financial health but also the market and operational risks experienced by financial market participants distant to these transactions.

In addition, like all financial instruments, there is the risk that CDSs are used for manipulative purposes, and there is a risk of fraud in the CDS market, in part because trade reporting and disclosure to the SEC are limited. Further, very small trades in a relatively thin market can be used to "paint the tape" and suggest that a credit is viewed by the market as weak. The focus by current data providers in CDS is on the spreads at which trades are concluded, rather than the volume transacted at that price.

One way to guard against misinformation and fraud is to create a mandatory system of recordkeeping and reporting of all CDS trades to the SEC. The information that would result from such a system would not only reduce the potential for abuse of the market, but would aid the SEC in detection of fraud in the market as quickly and efficiently as possible. Given the interdependency of financial institutions and financial products, it is crucial that we have a mechanism for promptly obtaining CDS trading information — who traded, how much and when — that is complete and accurate.

OTC market participants generally structure their activities in CDSs to comply with the CFMA's "swap exclusion" from the Securities Act and the Exchange Act. These CDSs are "security-based swap agreements" under the CFMA, which means that the SEC currently has authority to enforce antifraud prohibitions under the federal securities laws, including prohibitions against insider trading. If CDSs were standardized as a result of centralized clearing or exchange trading or other changes in the market, and no longer individually negotiated, the "swap exclusion" from the securities laws under the CFMA would be unavailable.

Notwithstanding the lack of statutory authority, the SEC is doing what it can under its existing statutory authority to address concerns regarding this market. Most recently, the Commission announced a sweeping expansion of its ongoing investigation into possible market manipulation involving certain financial institutions. The expanded investigation will require hedge fund managers and other persons with positions in CDSs to disclose those positions to the Commission and provide certain other information under oath. This expanded investigation should help to reveal the extent to which the risks I have identified played a role in recent events. Depending on its results, this investigation may lead to more specific policy recommendations.

However, investigations of over-the-counter CDS transactions have been far more difficult and time-consuming than those involving cash equities and options. Although the SEC clearly has antifraud jurisdiction over the CDS market, the SEC faces a much more difficult task in investigating and taking effective action against fraud and manipulation in the CDS market as compared to other markets. Because of the lack of uniform recordkeeping and reporting to the SEC, the information on CDS transactions gathered from market participants has been incomplete and inconsistent.

Recent private sector efforts may help to alleviate some of these concerns. For example, Deriv/SERV, an unregulated subsidiary of DTCC, provides automated matching and confirmation services for over-the-counter derivatives trades, including CDSs. Deriv/SERV's customers include dealers and buy-side firms from more than 30 countries. According to Deriv/SERV, more than 80% of credit derivatives traded globally are now confirmed through Deriv/SERV, up from 15% in 2004. Its customer base includes 25 global dealers and more than 1,100 buy-side firms in 31 countries. While programs like DerivSERV may aid the Commission's efforts, from an enforcement perspective, such voluntary programs would not be expected to take the place of mandatory recordkeeping and reporting requirements to the SEC.

In the future, Deriv/SERV and similar services may be a source of reliable information about most CDS transactions. However, participation in Deriv/SERV is elective at present, and the platform does not support some of the most complex credit derivatives products. Consequently, not all persons that engage in CDS transactions are members of Deriv/SERV or similar platforms. Greater information on CDS trades, maintained in consistent form, would be useful to financial supervisors. In addition to better recordkeeping by market participants, ready information on trades and positions of dealers also would aid the SEC in its enforcement of anti-fraud and anti-manipulation rules. Finally, because Deriv/SERV is unregulated, the SEC has no authority to view the information stored in this facility for supervision of risk associated with the OTC CDS market.

In crafting any regulatory solution, it is important to keep in mind the significant role CDS trading plays in today's financial markets, as well as the truly global nature of the CDS market. Further, the varied nature of market participants in CDSs and the breadth of this market underscore the importance of cooperation among U.S. financial supervisors at the federal and state level, as well as supervisors internationally.

Thank you for this opportunity to discuss these important issues. I am happy to take your questions.


Modified: 10/15/2008