The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
Credit Derivatives: Where's the Risk? Paula Tkac Economic Review, Vol. 92, No. 4, 2007 Tkac is an economist and associate policy adviser in the Atlanta Fed's research department. The following text is adapted slightly from an interview with Tkac recorded shortly after the Atlanta Fed's 2007 Financial Markets Conference, "Credit Derivatives: Where's the Risk?" held May 14–16. The interview is part of the Atlanta Fed's Research Insights podcast series. To listen to this podcast or to subscribe to any of the Atlanta Fed podcast series, visit www.frbatlanta.org and click "Podcasts" on the home page. Moderator: Welcome to Research Insights, a podcast from the Federal Reserve Bank of Atlanta. Our topic today is credit derivatives. . . . Our first question is, What are credit derivatives? Tkac: . . . Credit derivatives are a relatively recent financial innovation that effectively shifts credit risk, or the risk of default, from one party to another. The vast majority of credit derivatives are actually known as credit default swaps, and I think it might help if I give you a simple example. [S]uppose you want a bond, issued by a major corporation, such as GM or IBM. You're exposed to credit risk in the sense that if the corporation defaults on its bond payment, your bond decreases in value or becomes worthless. A credit default swap allows you to transfer this credit risk to another market participant for a price. Essentially, you pay a yearly fee to your counterparty, and they agree to pay you the par value of the debt in the case of default. Credit default swaps can also be written on sovereign debt of a nation, baskets of corporate bonds, indexes of debt-issuing corporations, and even tranches of these indexes. The market has grown rapidly in the past decade or so and has gotten increasingly complex as we've moved ahead. Moderator: . . . Why do credit derivatives matter to financial markets? Tkac: . . . I think the first reason is, as I mentioned, the large growth in the market. As of 1997, there was $180 billion in credit default swaps traded. That ballooned to over $20 trillion in 2006, so the sheer magnitude of the trade in these instruments reflects the value of this innovation to market participants but also causes us to want to look at them a little more closely. I should mention [that] the major participants are banks, hedge funds, and insurance companies. Conceptually, any innovation like this that unbundles risk allows these risks to be priced more efficiently. This [efficiency] allows participants to hedge and allows capital to flow more freely to its highest-valued use [and thus] has great benefits not only for financial markets but for our economy more generally. However, there are some concerns or potential for what we call "systemic risk"—the risk of spillovers of any kind of adverse event in this market to the larger financial market. The reasons that there are some concerns are, one, that these contracts are largely bilateral. They're under the radar. They're agreed to by two individual counterparties, not regulated through any exchanges. They involve a large notional amount. That means the value of the credit default swaps being traded out there is some fourteen to eighteen times the value of the underlying bonds on which these instruments are written. In addition, the participation of hedge funds causes some concern because, again, hedge funds largely operate under the regulatory radar, and so we're never certain from a policy perspective exactly what risks are being taken and whether they're being effectively monitored. Finally, this market has largely been untested. We've been through a very positive credit environment recently, and we don't really know what will happen if there is a major credit event on a large scale and how that might ripple through the financial system Moderator: . . . [Y]ou just got back from the Atlanta Fed Financial Markets Conference. Fed Chairman Ben Bernanke spoke there. There were other participants including policymakers, academics, and market participants. What were some of the main points that were raised at this year's conference? Tkac: [W]e spent a lot of time talking about both the risk management and, at individual firm level, the important things that market participants need to do to manage and evaluate their risk when trading these instruments. But [we also discussed] the risks at the systemic or macro level in terms of understanding the integrity and stability of the financial system and how credit derivatives and credit default swaps play into those concerns. [T]wo points that I would like to mention as probably most important coming out of the conference [are], first, [that] we heard from many practitioners and policymakers about the role of financial market participants themselves in monitoring and evaluating and managing their own risk—not only the risks of the instruments that they're holding, but importantly, for systemic reasons, the risk of their counterparties and their potential distress. So, when you trade with another . . . entity [that] has distress or losses and can shut down and can't meet [its] obligations, that may spill over onto your concerns, and that's where we worry about this domino effect being more pervasive. So . . . what we heard was that each individual market participant has a great incentive to evaluate and monitor and manage [his] own risk and exposure, and this is a powerful market-centered force to help us in constraining the potential for systemic risk. And indeed the industry itself has taken steps to improve back-office clearing and settlement, to move to auction and cash-based settlement, away from the physical delivery of bonds. All of these contribute to the integrity and stability of the market for credit derivatives. Second, I'd like to reference Chairman Bernanke's remarks. He spoke more generally and brought up, I think, some very important points about regulating financial innovations such as credit derivatives. He noted that consistency and a principlesbased paradigm for the regulation of innovations is important for achieving three public policy objectives: financial stability, investor protection, and market integrity. This [approach] is in contrast to a more rules-based approach or an ad hoc approach, which might regulate each instrument as it becomes available. As we all know, regulating in a rule-based fashion only encourages market participants to innovate and behave in response to those rules in ways [that are] sometimes . . . adverse to the initial policy objective. So Chairman Bernanke's remarks, I think, provide a good roadmap going forward as the market continues to innovate. F E D E R A L R E S E R V E B A N K O F AT L A N TA Credit Derivatives: An Overview DAVID MENGLE The author is the head of research at the International Swaps and Derivatives Association. This paper was presented at the Atlanta Fed’s 2007 Financial Markets Conference, “Credit Derivatives: Where’s the Risk?” held May 14–16. derivative is a bilateral agreement that shifts risk from one party to another; its value is derived from the value of an underlying price, rate, index, or financial instrument. A credit derivative is an agreement designed explicitly to shift credit risk between the parties; its value is derived from the credit performance of one or more corporations, sovereign entities, or debt obligations. Credit derivatives arose in response to demand by financial institutions, mainly banks, for a means of hedging and diversifying credit risks similar to those already used for interest rate and currency risks. But credit derivatives also have grown in response to demands for low-cost means of taking on credit exposure. The result has been that credit has gradually changed from an illiquid risk that was not considered suitable for trading to a risk that can be traded much the same as others. This paper begins with a description of credit default swaps, total return swaps, and asset swaps and then focuses on the mechanics and risks of credit default swaps. The paper then describes the market for credit default swaps and how it evolved and provides an overview of pricing and the risk-management role of the dealer. Next, the discussion considers the costs and benefits of credit derivatives and outlines some recent policy issues. The conclusion considers the possible future direction of the market. A How Credit Derivatives Work The vast majority of credit derivatives take the form of the credit default swap (CDS), which is a contractual agreement to transfer the default risk of one or more reference entities from one party to the other (Figure 1). One party, the protection buyer, pays a periodic fee to the other party, the protection seller, during the term of the CDS. If the reference entity defaults or declares bankruptcy or another credit event occurs, the protection seller is obligated to compensate the protection buyer for the loss by means of a specified settlement procedure. The protection buyer is entitled to protection on a specified face value, referred to in this paper as the ECONOMIC REVIEW Fourth Quarter 2007 1 F E D E R A L R E S E R V E B A N K O F AT L A N TA Figure 1 Credit Default Swap XX basis points per annum Protection buyer Protection seller Default payment Reference entity notional amount, of reference entity debt. The reference entity is not a party to the contract, and the buyer or seller need not obtain the reference entity’s consent to enter into a CDS. Risks associated with credit default swaps. In contrast to interest rate swaps but similar to options, the risks assumed in a credit default swap by the protection buyer and protection seller are not symmetrical. The protection buyer effectively takes on a short position in the credit risk of the reference entity, which thereby relieves the buyer of exposure to default.1 By giving up reference entity credit risk, the buyer effectively gives up the opportunity to profit from exposure to the reference entity. In return, the buyer takes on (1) counterparty default exposure to simultaneous default by the reference entity and the protection seller (“double default”) and (2) counterparty replacement risk of default by the protection seller only. In addition, the protection buyer takes on basis risk to the extent that the reference entity specified in the CDS does not precisely match the hedged asset. A bank hedging a loan, for example, might buy protection on a bond issued by the borrower instead of negotiating a more customized, and potentially less liquid, CDS linked directly to the loan. Another example would be a bank using a CDS with a five-year maturity to hedge a loan with four years to maturity. Again, the reason for doing so is liquidity, although as CDS markets expand the concentration of liquidity in specific maturities should lessen. The protection seller, in contrast, takes on a long position in the credit risk of the reference entity, which is essentially the same as the default risk taken on when lending directly to the reference entity. The main difference between the two is the need to fund a loan but not a sale of protection. The protection seller also takes on counterparty risk because the seller will lose expected premium income if the buyer defaults. One exception to the above risk allocation is the funded CDS (also called a creditlinked note), in which the protection seller lends the notional amount to the protection buyer in order to secure performance in the event of default. In a funded CDS the protection buyer is relieved of counterparty exposure to the protection seller, but the seller now has exposure to the buyer along with exposure to the reference entity. In order to reduce the seller’s exposure to the buyer, the parties sometimes establish 2 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA a bankruptcy-remote entity, known as a special-purpose vehicle, that stands between the two parties and is independent of default by the protection buyer. CDS mechanics. The reference entity is the party on which protection is written. For the simplest (single-name) form of CDS, the reference entity is an individual corporation or government. If a corporate reference entity is taken over by another, the protection typically shifts to the acquiring entity. If a reference entity de-merges or spins off a subsidiary, CDS market participants have developed a set of criteria, known as successor provisions, for determining the new reference entities. A CDS with two or more—usually between three and ten—reference entities is known as a basket CDS. In the most common form of basket CDS, the first-to-default CDS, the protection seller compensates the buyer for losses associated with the first entity in the basket to default, after which the swap terminates and provides no further protection. CDS referencing more than ten entities are sometimes referred to as portfolio products. Such products are generally used in connection with synthetic securitizations, in which a CDS transfers credit risk of loans or bonds to collateralized debt obligation (CDO) note holders in lieu of a true sale of the assets as in a cash securitization (Choudhry 2004). A major source of credit derivatives growth since 2004 has been the index CDS, in which the reference entity is an index of as many as 125 corporate entities. An index CDS offers protection on all entities in the index, and each entity has an equal share of the notional amount. The two main indices are the CDX index, consisting of 125 North American investment-grade firms, and the iTraxx index, consisting of 125 euro-based firms, mainly investment grade. In addition, indices exist for North American subinvestment-grade firms, for European firms that have been downgraded from investment grade, and for regions such as Japan and Asia excluding Japan. If a firm included in the index defaults, the protection buyer is compensated for the loss and then the CDS notional amount is reduced by the defaulting firm’s pro rata share. In addition to CDS on indices, market participants can buy or sell protection on tranches of indices— that is, on a specific level of losses on an agreed notional amount of an underlying index. For example, an investor can sell protection on the 3–7 percent tranche of the CDX Investment Grade Index with a notional amount of $100 million, which means the investor could be required to compensate a protection buyer for losses on the index in excess of $3 million but not beyond $7 million, for a maximum of $4 million. Recent innovations in CDS have extended protection to reference obligations instead of entities. CDS on asset-backed securities (ABS), for example, provide protection against credit events on securitized assets, usually securitized home equity lines of credit. In addition, CDS can specify CDO notes as reference obligations. Finally, loan CDS can reference leveraged loans to a specific entity. With regard to credit events, the confirmation of a CDS deal specifies a standard set of events that must occur before the protection seller compensates the buyer for losses; the parties to the deal decide which of those events to include and which to exclude. Which events are chosen varies according to the type of reference entity. First, the most commonly included credit event is failure to pay. Second, bankruptcy is a credit event for corporate reference entities but not for sovereign entities. Third, restructuring, which refers to actions such as coupon reduction or maturity extension undertaken in lieu of default, is generally included as a credit event for corporate entities. Restructuring is sometimes referred to as a “soft” credit event because, 1. Credit traders in fact refer to bought protection as a short position in the reference entity and to sold protection as a long position. ECONOMIC REVIEW Fourth Quarter 2007 3 F E D E R A L R E S E R V E B A N K O F AT L A N TA in contrast to failure to pay or bankruptcy, it is not always clear what constitutes a restructuring that should trigger compensation. Fourth, repudiation or moratorium provides for compensation after specified actions of a government reference entity and is generally relevant only to emerging market reference entities. Finally, obligation acceleration and obligation default, which refer to technical defaults such as violation of a bond covenant, are rarely included. The third feature of a CDS, the settlement method, refers to the means by which the protection seller compensates the buyer in the event of default. The two types of settlement are physical settlement and cash settlement. If a credit event triggers a In contrast to interest rate swaps but simiCDS with physical settlement, the proteclar to options, the risks assumed in a credit tion buyer delivers to the protection seller default swap by the protection buyer and the defaulted debt of the reference entity with a face value equal to the notional protection seller are not symmetrical. amount specified in the CDS. In return, the protection seller pays the par value—that is, the face amount—of the debt. If the event occurs in a CDS with cash settlement, an auction of the defaulted bonds takes place to determine the postdefault market value. Once this value is determined, the protection seller pays the buyer the difference between the par value, which is equal to the CDS notional amount, and the postdefault market value. Physical settlement was the standard settlement method for most CDS until 2005 but is being replaced by cash settlement for reasons that will be discussed in a later section. The last major feature of a credit default swap is the premium, commonly known as the CDS spread; this feature will be discussed in more detail in a later section. The spread is essentially the internal rate of return that equates the expected premium flows over the life of the swap to the expected loss if a default occurs at various dates. The buyer and seller agree on the spread on the trade date, and the spread remains constant for the life of the CDS; the only exception is a constant maturity CDS, in which the credit spread is reset periodically to the current market level. The CDS spread is quoted as an annual premium, such as 1 percent or 100 basis points per annum, but is actually paid in quarterly installments during the year. Transaction mechanics. In the early stages of a trading relationship, the contracting parties conduct credit analyses of each other and negotiate the terms of the agreement under which future transactions will take place. For over-the-counter (OTC) derivatives, including credit derivatives, the most commonly used agreement is the International Swaps and Derivatives Association (ISDA) Master Agreement. The agreement includes terms that the parties wish to include in all future transactions—for example, governing law, covenants, and so on. Once the parties execute the agreement, it serves as the contract under which all future OTC derivative deals take place. Each deal is evidenced by a confirmation, which contains the terms of the individual transaction such as reference entity, maturity, premium, notional amount, credit events, settlement method, and other transaction-specific terms. The terms of the confirmation in turn draw from the ISDA definitions pertaining to the product; for CDS, the relevant definitions are the 2003 ISDA Credit Derivatives Definitions. Execution of a deal involves negotiating the deal terms, which as mentioned above are listed in the confirmation. The generation of the confirmation is of particular importance because both parties must agree on the same terms; if they do not specify precisely the identity of the reference entity, for example, a protection buyer could claim that the entity defaulted, but the payer could refuse payment because the entity described in the confirmation is not identical to the one that defaulted. In most trans- 4 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Figure 2 Total Return Swap LIBOR + X basis points Total return receiver Funding cost (< – LIBOR) Total return payer TR of reference obligation Reference obligation TR of reference obligation Total return swap Total return (TR) = interest + fees ± (appreciation/depreciation) – default losses actions, market participants will choose from a standard menu of contract terms that have been developed collectively by ISDA member firms. As in all OTC derivatives, however, the parties are free to negotiate terms that differ from market standards. Following the execution of the trade, the parties will monitor for occurrence of credit events. In addition, the parties will also have to amend trades to account for succession events in which the reference entity changes form as mentioned previously. Finally, if a credit event occurs, the parties settle the CDS obligations according to procedures set forth in the ISDA documentation. Other credit derivatives. The credit default swap in various forms accounts for the vast majority of credit derivatives activity. Three related products deserve mention, however. First, a total return swap transfers the total economic performance of a reference obligation from one party (total return payer) to the other (total return receiver). In contrast to a credit default swap, the total return swap transfers market risk along with credit risk. As a result, a credit event is not necessary for payment to occur between the parties. A total return swap works as follows (Figure 2). The total return payer normally owns the reference obligation and agrees to pay the total return on the reference obligation to the receiver. The total return is generally equal to interest plus fees plus the appreciation or depreciation of the reference obligation. The total return receiver, for its part, will pay a money market rate, usually LIBOR (London Interbank Offered Rate), plus a negotiated spread, which is generally independent of the reference obligation performance. The spread is generally bounded by funding costs: The upper bound is the receiver’s cost of funding, and the lower bound is the payer’s cost of funding the reference obligation. If a credit event or a major decline in market value occurs, the total return will become negative, so the receiver will end up compensating the payer. The end result of a total return swap is that the total return payer is relieved of economic exposure to the reference obligation but has taken on counterparty exposure to the total return receiver. The most common total return receivers are hedge funds seeking exposure to the reference obligation on terms more favorable ECONOMIC REVIEW Fourth Quarter 2007 5 F E D E R A L R E S E R V E B A N K O F AT L A N TA Figure 3 Asset Swap LIBOR Money market 6.30% Investor LIBOR + 0.85% Corporate note (5-year) 6.30% Dealer Assume that 5-year U.S. dollar interest rate swap rate = 5.45% Par bond coupon = 6.30% ⇒ Asset swap spread = 0.85% than by funding a direct purchase of the obligation; this tactic is sometimes known as “renting the balance sheet” of the total return payer, which is normally a wellcapitalized institution such as a bank. In addition to total return swaps, asset swaps are sometimes classified as credit derivatives although they are in fact interest rate derivatives. Whatever their classification, they are relevant to credit derivatives because they are related by arbitrage to credit default swaps. An asset swap combines a fixed-rate bond or note with an interest rate swap (Figure 3). The party that owns the bond pays the coupon into an interest rate swap with a similar maturity to the bond. Because the bond coupon is typically larger than the current swap rate for that maturity, the LIBOR leg of the floating rate swap is increased by a spread equal to the difference between the underlying bond coupon rate and the interest rate swap rate prevailing on the trade date. Because the interest rate swap effectively strips out the interest rate risk of the bond, the bondholder is left mainly with the credit risk of the bond (along with some counterparty credit risk on the swap). The asset swap spread compensates the bondholder for the credit risk; for this reason, the asset swap spread should be related by arbitrage to the credit default swap spread. This relationship will be discussed in more detail in the section on pricing. One last type of credit derivative is the credit spread option, which gives the buyer the right but not the obligation to pay or receive a specified credit spread for a given period. Such products were never more than 5 percent of notional amounts outstanding and are now about 1 percent (British Bankers Association [BBA] 2006), so they are of mainly historical interest. Credit spread options appear to have given way to swaptions on CDS, which give the buyer the right but not the obligation to buy (put swaption) or sell (call swaption) CDS protection. In the remainder of this paper, credit derivatives and credit default swaps will mean the same thing unless otherwise specified. The Market for Credit Derivatives According to the BBA (2006), the notional amount outstanding of credit derivatives has grown from $180 billion in 1997 to over $20 trillion in 2006 (Figure 4). Other sur- 6 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Figure 4 Growth of Credit Derivatives 50,000 45,000 BBA ISDA 40,000 $U.S. billion 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Notional amounts outstanding Sources: BBA (2006); ISDA Market Surveys, 2001–07 veys report higher numbers. ISDA, for example, began collecting CDS notional amounts in 2001 and reports growth from $632 billion in 2001 to over $45 trillion by midyear 2007; annual growth has exceeded 100 percent from 2004 through 2006 but slowed to 75 percent by mid-2007. And the Bank for International Settlements, which began collecting comprehensive statistics in 2004, reports growth of notional amount from $6.4 trillion at the end of 2004 to almost $43 trillion as of June 2007 (BIS 2007). Average notional amounts for individual deals range from $10 million to $20 million for North American investment-grade credits and are about €10 million for European investment-grade credits; sub-investment-grade credits have notionals that average about half the amounts for investment grade (JPMorgan Chase 2006). The most liquid maturities center on five years, but liquidity is increasing for shorter maturities and for longer maturities out to ten years (BBA 2006). Table 1 shows the credit derivative breakdown by product type. According to the BBA, CDS on indices have recently passed CDS on single names as the dominant product type (BBA 2006). Single-name CDS, which were 38 percent of notional amount outstanding in 1999, grew to as high as 51 percent in 2004 and are 33 percent as of 2006. CDS linked to indices and to tranches of indices have grown from virtually nothing in 2003 to 38 percent of outstandings. Finally, CDS referencing portfolios of names in synthetic securitization transactions have declined slightly from 18 percent in 2000 to just over 16 percent in 2006. The “others” category includes total return swaps and asset swaps, which are now less than 6 percent of outstandings; in 2000, in contrast, total return swaps were 11 percent of outstanding amounts, and asset swaps were 12 percent (BBA 2002). Tables 2 and 3 show the breakdown of market participants by type. Banks and securities firms were dominant in 2000, at 81 percent of protection buyers and 63 percent of protection sellers. By 2006, they had declined in importance to 59 percent of buyers and 44 percent of sellers. Recent data distinguish between banks’ trading ECONOMIC REVIEW Fourth Quarter 2007 7 F E D E R A L R E S E R V E B A N K O F AT L A N TA Table 1 Credit Derivative Product Mix 2000 2002 2004 2006 38 6 — — — — 10 5 — — 41 45 6 — — — — 8 5 — — 36 51 4 9 2 6 10 6 2 1 1 8 33 2 30 8 4 13 3 1 0 1 6 Single-name credit default swaps Basket products Full index trades Tranched index trades Synthetic CDOs—fully funded Synthetic CDOs—partially funded Credit-linked notes (funded CDS) Credit spread options Equity-linked credit products Swaptions Others Source: BBA (2006) activities and credit portfolio management activities: Trading activities are roughly balanced between buying and selling protection while credit portfolio managers appear more likely to hedge by buying protection than to seek diversification through selling protection. Insurance companies tend to be active as sellers of protection; they were 23 percent of sellers in 2000 but dropped to 17 percent by 2006. The most significant change has been in the importance of hedge funds, which tend to function as both buyers and sellers: In 2000, hedge funds were 3 percent of buyers and 5 percent of sellers but by 2006 had grown to 28 percent of buyers and 32 percent of sellers. Table 4 shows the most common CDS counterparties—essentially, the most active dealers in the market—from 2003 through 2006. Table 5 shows the most common reference entities for single-name CDS, both by deal count and by underlying notional amount, as of year-end 2006 (Fitch Ratings 2007). Evolution of the market. Smithson (2003) identified three stages in the evolution of credit derivatives activity. The first, “defensive” stage, during the late 1980s and early 1990s, was characterized by ad hoc attempts by banks to lay off some of their credit exposures. In addition, products such as securitized asset swaps bore some resemblance to credit default swaps in that they paid investors a credit spread while providing for delivery of the underlying asset to the investor in the event of a default (Cilia 1996). Stage two, which began about 1991 and lasted through the mid-to-late 1990s, saw the emergence of an intermediated market, in which dealers applied derivatives technology to the transfer of credit risk while investors entered the markets to seek exposure to credit risk (Spinner 1997). Examples of dealer applications of derivative technology include two transactions by Bankers Trust (Das 2006, 269–70). The first involved a total return swap with another bank client seeking to free up credit lines with a major client. The swap enabled the bank to pass its credit risk to Bankers Trust, which in turn hedged its risk by selling the client’s bonds short. The second transaction involved a funded first-to-default CDS on several Japanese client banks, against which Bankers Trust had substantial credit exposure in the form of in-themoney options. Although defensive in nature from Bankers Trust’s viewpoint, the 8 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Table 2 Buyers of Protection by Institution Type Type of institution 2000 2002 2004 2006 67 — — 7 59 39 20 6 2 2 2 28 2 2 2 1 Banks (including securities firms) Banks—trading activities Banks—loan portfolio Insurers Monoline insurers 81 — — 7 — 73 — — 6 3* Reinsurers Other insurance companies Hedge funds Pension funds Mutual funds Corporates Other — — 3 1 1 6 1 3 12 1 2 4 2 2 3 2 16 3 3 3 1 2002 2004 2006 54 — — 20 10 7 3 15 4 4 2 1 44 35 9 17 8 4 5 32 4 3 1 1 *Monoline insurers and reinsurers combined Source: BBA (2006) Table 3 Sellers of Protection by Institution Type Type of institution Banks (including securities firms) Banks—trading activities Banks—loan portfolio Insurers Monoline insurers Reinsurers Other insurance companies Hedge funds Pension funds Mutual funds Corporates Other 2000 63 — — 23 — — — 5 3 2 3 1 55 — — 33 21* 12 5 2 3 2 0 *Monoline insurers and reinsurers combined Source: BBA (2006) transaction appealed to investors seeking yield enhancement by buying the creditlinked notes issued by Bankers Trust. Another innovation during this phase was the synthetic securitization structure. Synthetic securitization represented the extension of credit derivatives to structured finance, that is, to the combining of derivatives with cash instruments or with other derivatives to attain a desired exposure. The first synthetic securitization transactions included the Glacier transaction, developed by SBC Warburg (now UBS), and the Bistro transaction, developed by J.P. Morgan (now JPMorgan Chase). Glacier was a funded structure, in which SBC transferred to investors the entire credit risk of approximately $1.75 billion of loans by means of credit-linked notes. Bistro, in contrast, was ECONOMIC REVIEW Fourth Quarter 2007 9 F E D E R A L R E S E R V E B A N K O F AT L A N TA Table 4 Twenty Largest CDS Counterparties, 2003–06 2003 2004 2005 2006 JPMorgan Chase Deutsche Bank Morgan Stanley Morgan Stanley Deutsche Bank Morgan Stanley Deutsche Bank Deutsche Bank Goldman Sachs Goldman Sachs Goldman Sachs Goldman Sachs Morgan Stanley JPMorgan Chase JPMorgan Chase JPMorgan Chase Merrill Lynch Merrill Lynch UBS Barclays CSFB CSFB Lehman Brothers UBS UBS Lehman Brothers Barclays Lehman Brothers Lehman Brothers Merrill Lynch Citigroup Credit Suisse Citigroup Citigroup CSFB Merrill Lynch Bear Stearns Bear Stearns BNP Paribas BNP Paribas Commerzbank Barclays Merrill Lynch ABN Amro BNP Paribas BNP Paribas Bear Stearns Bear Stearns Bank of America Bank of America Bank of America Citigroup Dresdner Dresdner Dresdner Société Générale ABN Amro HSBC ABN Amro HSBC Société Générale Commerzbank HSBC Dresdner AIG Royal Bank of Scotland Société Générale Bank of America Barclays Société Générale Calyon Royal Bank of Scotland Toronto Dominion ABN Amro Royal Bank of Scotland Calyon Calyon Toronto Dominion AIG CIBC Source: Fitch Ratings (various years) a partially funded structure, in which Morgan transferred to investors approximately 10 percent of the credit risk by means of a credit default swap while retaining any loss beyond that in the form of a “super-senior” tranche (Choudhry 2004). Although the transactions appear defensive from UBS and Morgan’s point of view, they also appealed to investors seeking exposure to credit risk. Investors benefited from the above second-stage innovations in at least two ways. First, investors could attain exposure to loans, which had previously been out of reach becaue of the lack of a credit processing infrastructure among buy-side firms. Second, investors could attain exposure to credit risk without having to accept exposure to interest rate risk as well; asset swaps were an early means of attaining such exposure. The third stage saw the maturing of credit derivatives from a new product into one resembling other forms of derivatives. Single-name credit default swaps emerged during this period as the “vanilla,” or generic, credit derivatives product, while structured finance groups combined credit derivatives into “arbitrage” CDO packages geared to investor demands. Major financial regulators issued guidance for the regulatory capital treatment of credit derivatives, which served to clarify the constraints under which the emerging market would operate. Further, ISDA in 1999 issued a set of standard credit derivatives definitions for use in connection with the ISDA Master Agreement. Finally, dealers began warehousing risks and running hedged and diversified portfolios of credit derivatives. 10 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Table 5 Top Reference Entities, Gross Protection Bought and Sold, Year-End 2006 Largest by count Largest by volume Protection sold Protection bought Protection sold Protection bought 1 General Motors/GMAC General Motors/GMAC General Motors/GMAC General Motors/GMAC 2 DaimlerChrysler DaimlerChrysler Brazil Brazil 3 Telecom Italia Ford Motor Corp./ Ford Motor Credit DaimlerChrysler DaimlerChrysler 4 Italy France Telecom Ford Motor Corp./ Ford Motor Credit France Telecom 5 Deutsche Telekom Telecom Italia Turkey Turkey 6 Ford Motor Corp./ Ford Motor Credit Telefonica Telecom Italia Ford Motor Corp./ Ford Motor Credit 7 Brazil Brazil Russia Telecom Italia 8 Telefonica Deutsche Telekom France Telecom Deutsche Telecom 9 France Telecom Italy Deutsche Telecom Russia 10 Russia Volkswagen Telefonica Telefonica 11 BT Group Russia United Mexican States AT&T 12 Fannie Mae Time Warner BT Group BT Group 13 General Electric/GECC Turkey Italy AIG 14 Spain Argentina AT&T Volkswagen 15 Turkey BT Group General Electric/GECC General Electric/GECC 16 Portugal General Electric/GECC AIG Gazprom 17 United Mexican States Altria Group Fannie Mae Banco Santander Central Hispano 18 France Bombardier Altria Group Safeway 19 Germany Merrill Lynch KPN United Mexican States 20 Altria Group Philippines Vodafone Altria Group 21 Deutsche Bank United Mexican States Portugal Telecom Argentina 22 Merrill Lynch AIG VNU KPN 23 Gazprom Bayer Safeway Venezuela 24 Time Warner Citigroup Gazprom AXA 25 Volkswagen Clear Channel Venezuela Supervalu Source: Fitch Ratings (2007) During this stage, the market encountered a series of challenges that led to calls for further refinement to the documentation. One such problem was restructuring. The 1999 definitions included debt restructuring—that is, actions such as lowering coupon or extending maturity—as a credit event triggering payment under a CDS. The definition was put to the test with the restructuring in 2000 of loans to Conseco. Banks agreed to extend the maturity of Conseco’s senior secured loans in return for higher coupon and collateral; protection was thereby triggered on about $2 billion of CDS. Protection buyers then took advantage of an embedded “cheapest to deliver” option in CDS by delivering long-dated senior unsecured bonds, which were deeply ECONOMIC REVIEW Fourth Quarter 2007 11 F E D E R A L R E S E R V E B A N K O F AT L A N TA discounted—worth about 40 cents on the dollar—relative to the restructured loans, which were worth over 90 cents on the dollar. Protection sellers ended up absorbing losses that were greater than those incurred by protection buyers, which led many protection sellers to question the workability of including restructuring. The problem was complicated further by the insistence by some regulators that CDS cover restructuring for a CDS hedge to qualify for capital relief. The result, arrived at through ISDA committee efforts, was a set of modWith the new ISDA credit derivatives defiifications to the definition of restructuring that placed some limits on deliverable bond nitions in place, dealers in 2003 began to maturity and therefore on the cheapesttrade according to certain standardized to-deliver option. practices that went beyond those adopted Another problem involved apportioning credit protection when a reference for other OTC derivatives. entity de-merges or spins off part of its activities into new entities. The problem arose in the United Kingdom in 2000, when National Power de-merged into two companies; one company inherited 56 percent of National Power’s obligations, and the other held the rest. The problem was to determine the new reference entity for CDS referencing National Power, but the 1999 definitions did not provide sufficient guidance to assure the market that courts would agree on the outcome. The result was to develop a set of detailed “successor” provisions, which provided quantitative thresholds for such cases. Yet another problem was debt moratoriums or repudiations in emerging markets. During the Argentine debt crisis of 2002, there were several changes of government, involving a succession of officials who made threats regarding debt repudiation. The problem arose that, under the 1999 definitions, it was possible to declare a repudiation credit event following a statement by a government official even if in the end the government did not fail to pay its obligations. To reduce the risk of declaring a credit event prematurely, ISDA developed more stringent criteria for such an event. The foregoing problems led ISDA to issue in 2003 a new set of credit derivatives definitions. At this point, one can add a fourth stage to those cataloged by Smithson, namely, the development of a liquid market. With the new ISDA credit derivatives definitions in place, dealers in 2003 began to trade according to certain standardized practices—standard settlement dates, for example—that went beyond those adopted for other OTC derivatives. Further, index trading began on a large scale in 2004 and grew rapidly. The wide acceptance of index trading at that time was in part the result of the merger of the iBoxx and Trac-x credit indices into iTraxx for Europe and CDX for North America. The mergers provided market participants with a single index, subject to transparent rules and a high degree of standardization, for each major market. At the same time, dealers took deliberate measures to promote liquidity in index trading; such measures included developing master confirmations, committing to quote tight bid-offer spreads, and allowing investors to trade out of an old index and “roll” into the new one at mid-market spreads. Index trading was more appealing to investors than single-name trading because indices provide diversified exposure instead of concentrating it on one name. The results went well beyond expectations: According to one survey, index product growth was 900 percent in 2005 (Fitch Ratings 2006). This last stage saw the entry of hedge funds on a large scale as both buyers and sellers (Tables 2 and 3). Hedge funds use credit derivatives in a variety of ways. First, hedge funds use credit derivatives in their convertible bond arbitrage activities to strip out unwanted credit risk. Second, hedge funds can buy and sell protection to profit 12 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA from perceived mispricing. Finally, hedge funds engage in basis trading between credit default swaps and assets swaps on cash bonds. All these activities serve to increase liquidity, price discovery, and efficiency in the market. Pricing, Valuation, and Risk Management CDS pricing and valuation. The premium for a credit default swap is commonly known as a CDS spread and is quoted as an annual percentage in basis points of the notional amount. Although quoted as an annual percentage, protection buyers actually pay the spread on a quarterly basis, that is, in four installments per year. Further, in contrast to other OTC derivatives, CDS have standard payment dates, namely, March 20, June 20, September 20, and December 20; these standard payment dates also serve as standard maturity dates. CDS transacted prior to a standard payment date are subject to a “stub” period up to the first standard payment date and follow the standard schedule afterwards. A CDS with a five-year maturity agreed on May 1, 2007, for example, would become effective on May 2 with the accrued premium due on June 20; subsequent payments would occur on regular dates until maturity on June 20, 2012. If the spread for a distressed credit is sufficiently high, the CDS will trade “up front”—that is, the buyer will pay the present value of the excess of the premium over 500 basis points at the beginning of the trade and pay 500 basis points per annum for the life of the swap (Taksler 2007, 44). There are two basic ways of determining a CDS spread, namely, from asset swap spreads and from calculation of expected CDS cash flows. Assets swaps, which were described in a preceding section and shown in Figure 3, are related to credit default swaps because both products serve to unbundle credit risk. In an asset swap, an investor purchases a reference entity’s cash bond—preferably priced at par—and pays the bond coupon into an interest rate swap of the same maturity. As mentioned previously, the swap counterparty adjusts the LIBOR leg of the swap for the difference between the bond coupon rate and the par swap rate for the same maturity; the difference is known as the asset swap spread and compensates the investor for the default risk on the bond. The asset swap spread performs essentially the same function as a CDS spread, so the two should be related by arbitrage. If CDS spreads are low relative to asset swap spreads, for example, a dealer or investor could buy an asset-swapped bond and offset it by buying protection (equivalent to selling the reference entity short) and locking in a profit. Such arbitrage should lead to convergence between CDS spreads and asset swap spreads (narrowing of the basis). Arbitrage in the other direction is not as straightforward, however: If CDS spreads are higher than asset swap spreads, arbitrage requires selling protection (long the credit) and selling the bond short. Shorting a bond is often not feasible, however, and will depend crucially on the liquidity of the underlying bond market. The possibility of arbitrage between CDS and asset swaps will nonetheless tend to reduce the basis between the two rates. But other factors are also at work to keep the rates from converging (Choudhry 2006). Supply and demand factors might affect the price of CDS relative to bonds in several ways. Structured finance activity, for example, might lead to sales of protection to fund CDO notes, thereby driving down CDS spreads relative to bonds. Similarly, investors with high funding costs might prefer to take on credit risk by selling protection rather than by purchasing bonds financed by borrowing, again driving spreads down relative to bond yields. And in the other direction, a bond trading below par will tend to push CDS spreads higher relative to bond yields. The reason is that a protection seller pays out the difference ECONOMIC REVIEW Fourth Quarter 2007 13 F E D E R A L R E S E R V E B A N K O F AT L A N TA between par value and postdefault price, while an investor who bought the bond below par has lost only the difference between the below-par purchase price and the postdefault price. The result of the above factors is that, even if asset swap spreads will not in most cases converge to CDS spreads, they are a reasonable starting point for calculating the spread. As an alternative to relying on asset swap spreads, CDS pricing models seek to calculate CDS spreads by calculating expected cash flows. In such models, the CDS spread is an internal rate of return that equates present value of expected premium payments to present value of expected loss payments; that is PV(expected spread payments) = PV(expected default losses), where n Σ DCF PV(expected spread payments) = Spread × i × PSi × PVi (N) and n i=1 Σ [(PSi–1 – PSi) × PVi(N)] PV(expected default losses) = LGD × i=1 using the following notation: Spread = fixed CDS spread; DCFi = day count fraction relevant to period i; PSi = survival probability, that is, probability of no default from inception to period i; PVi(.) = present value operator for period i; N = notional amount of CDS protection; LGD = loss rate given default, equal to (1 – recovery rate), assumed fixed; and PSi–1 – PSi = PDi = probability of default in period i. Solving the model involves calculating the spread that equates net present value to zero—that is, an internal rate of return—under an assumed LGD. The survival and default probabilities come from outside the model; alternatively, market spreads can be used to calculate implied probabilities of default under an assumed LGD. For simplicity, the above equations ignore accrued spread, which in the event of default would be payable by the buyer to the seller for the fraction of the period from the last premium payment date to the default date. After inception, the value of the CDS will depend mostly on changes in credit quality as reflected in current credit spreads. Given that the CDS spread for a transaction remains fixed, the mark-to-market value of the CDS will be equal to the present value of the spread differences over the life of the CDS, taking account of survival probabilities and, again, ignoring the accrued premium. Letting Spread0 equal the CDS spread fixed at inception and Spreadi equal the current market spread, markto-market value from the buyer’s point of view is n Σ DCFi × PSi × PVi(N). MTMi = (Spreadi – Spread0) × i=1 If the buyer were to unwind at this point (to be discussed in the section on novations), the above equation represents the amount payable to the buyer.2 Just as difficulties exist using asset swap spreads, problems are associated with models such as that described above. A major difficulty is that the model requires that one assume an LGD and furnish exogenous probabilities of default to calculate an implied CDS spread. Alternatively, one could use the current market CDS spread to calculate an implied probability of default, but doing so would still require assuming an LGD. Assumptions regarding recoveries therefore are important to CDS pricing and valuation. In practice, market participants can model the effect of alternative 14 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA LGD assumptions and can set aside reserves against differences in assumptions (JPMorgan Chase 2006, 92–93; Chaplin 2005, 105). Risk management. The purpose of a derivatives dealer, along with making a twoway market, is to earn profits by managing the risk of a portfolio of derivatives. For credit derivatives, the risk management function is similar in form to that for other types of derivatives. When a dealer takes on risk from a client, the dealer typically hedges the risk but might elect to leave some of the risk uncovered. The willingness of dealers to leave some risks uncovered—that is, to speculate—adds liquidity to the market but requires close management. Typically, however, dealers hedge their risks in some manner. Most simply, a dealer might offset the risk of a new deal against that of other clients. Further, the dealer might hedge the risk of a deal in the underlying market, that is, the cash bond market; for that reason, credit derivatives and corporate bond risks are often managed together. Finally, a dealer might choose to offset risks by means of offsetting transactions with other dealers; this is a primary function of the interdealer market. In the early stages of credit derivatives as described above, risk management essentially consisted of laying off one’s own risks. As the market developed into a twosided market, dealers assumed the role of intermediaries between buyers and sellers, taking on the basis risk between the two. In these early stages of development, dealers tended to hedge new transactions with offsetting cash market positions to the extent feasible or else with offsetting transactions. As the market has developed, additional hedging alternatives have become available. The growth of index CDS, for example, has increased the flexibility of dealers in their risk-management activities. After the advent of widely traded index CDS, market liquidity increased significantly, and new market participants entered both as buyers and sellers. In such an environment, the business has evolved into a “flow” business: that is, traders tend to remain “flat” by buying and selling continually instead of by taking large open long or short positions. But trading desks also can use index CDS to hedge their single-name CDS. For example, on any given day spreads tend to move in the same direction, so index swaps might be a reasonable hedge of a diversified single-name CDS portfolio; as with other hedges, the dealer would assume and manage the basis risk between the two.3 Along with the market risks of their deal portfolios and the accompanying basis risks, dealers manage a host of other risks. First, counterparty credit risk is a major component of all OTC derivatives activity. Counterparty risk management begins with ISDA or other relevant transaction documentation, followed by measurement of both current exposure and potential losses if default were to occur in the future and finally by collateralization of net exposures. Second, dealers manage such risks as time decay, in which deals lose value as they approach maturity. Finally, dealers manage model risks, which are associated with the simplifying assumptions as well as unidentified errors in pricing models; to anticipate the possibility of model deficiencies, dealers calculate potential losses from modeling errors and set aside reserves to cover them (Chaplin 2005, 100–106). One unique aspect of credit derivatives activity compared with other OTC derivatives is liquidity management. Credit derivatives are characterized by a higher degree of standardization than are other forms of OTC derivatives, although the standard 2. For a more detailed practitioner-oriented discussion of CDS pricing, see Chaplin (2005). 3. This type of hedging is roughly equivalent to hedging interest rate swaps with Treasury bonds or an equity portfolio with S&P 500 futures. ECONOMIC REVIEW Fourth Quarter 2007 15 F E D E R A L R E S E R V E B A N K O F AT L A N TA terms are not mandatory as in exchange-traded futures. As noted above, CDS involve standard payment and maturity dates. Further, each type of reference entity involves a standard set of credit events and other terms. Standard terms facilitate trading by simplifying negotiation and tasks such as unwinds; they also make it easier for market participants to engage in arbitrage between CDS indices and underlying names. Credit derivatives participants have adopted a higher degree of standardization because credit risk is different from other underlying risks. Unlike interest rate swaps, in which the various risks of a customized transaction can be isolated by traders and offset in liquid underlying money and currency markets, credit default swaps involve “lumpy” credit risks that do not lend themselves to decomposition. Standardization is therefore a substitute for decomposition. Yet despite the higher degree of standardization, CDS retain their essential nature as OTC rather than standardized transactions: Parties to CDS deals remain free to diverge from the market standard and to customize transactions to whatever extent they agree. Benefits and Costs of Credit Derivatives Benefits. Credit derivatives emerged in response to two long-standing problems in banking. First, lending operated under the handicap that hedging credit risk was seldom, if ever, feasible. In financial terms, the problem was that taking a short position in credit was not generally feasible. Although selling a corporate bond short is theoretically possible, many borrowers do not have liquid debt outstanding, so borrowing for a short sale is often not feasible. As a result, if a credit deteriorates, a lender can do little to protect itself prior to default other than taking collateral, which might not be effective in many distressed cases, or by selling the loan, which normally requires the consent of the borrower. A second problem was diversification of credit risk. Financial economists have long noted the benefits of applying a portfolio approach to loans by means of diversification (Flannery 1985), but practical considerations made diversification difficult to achieve. Relationship considerations, for example, posed an obstacle to diversifying by deliberately reducing exposure to major clients. Further, the statistical properties of credit risk—that is, non-normality of loss distributions and the resulting effect of specification errors in determining losses in the tail of the distributions— suggest that a truly diversified loan portfolio requires a significantly larger number of credits than would an equity or bond portfolio (Smithson 2003, 34–38). Given such obstacles, the only practical way to diversify a lending business was to grow to a large size by means of acquiring other banks. Buying protection by means of credit derivatives provides solutions to both of the foregoing problems. By allowing banks to take a short credit position, credit derivatives enable banks to hedge their exposure to credit losses. A major benefit is that, in contrast to loan sales, CDS do not require consent of the reference entity. As a result, lenders also have a solution to the second problem, diversification. By hedging selectively, a bank can reduce its exposure to certain entities, thereby attaining its diversification objective without jeopardizing the client relationship. Another benefit to the protection buyer is the ability to act on a negative credit view. If an investor believes that the market is overly optimistic about a reference entity’s prospects, for example, the investor can buy protection now in anticipation of deterioration. If the investor’s view turns out to be correct, the investor can unwind the transaction at a profit by selling protection on the entity. Such speculative activity has the beneficial effect of adding liquidity to the market and of increasing the quality of price discovery (International Monetary Fund [IMF] 2006, chap. 2). 16 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Market participants can also use CDS to engage in arbitrage between markets. In convertible bond arbitrage, for example, an investor buys a convertible bond in which the embedded equity option is underpriced, uses an asset swap to hedge out the interest rate risk, and then buys credit protection to hedge out the credit risk. The investor is then left with a pure equity exposure, which is the object of the arbitrage. With regard to sellers of protection, credit derivatives enable market participants to attain exposure in the form of a long credit position. A financial institution seeking to diversify its credit exposure might, for Unlike interest rate swaps, in which risks example, sell CDS protection as an alternacan be isolated and offset in liquid undertive to making loans or buying bonds. This alternative is especially helpful to institulying money and currency markets, tions that seek credit exposure but lack the credit default swaps involve “lumpy” legal infrastructure for lending; it is also credit risks that do not lend themselves helpful to banks seeking to diversify their loan portfolios but lacking direct relationto decomposition. ships with desired credits. Further, selling protection allows an investor with a high cost of funding to take on credit exposure without incurring the cost of funding. It is important in such cases that the investor realizes that the exposure to losses is the same as if it were lending directly. The ability to sell protection also allows market participants to act on a view that a reference entity’s credit quality will improve. In this case, the investor would sell protection now in the hope of unwinding it later by purchasing it at a lower price. As mentioned above, such activity adds liquidity to the market and increases the quality of price discovery. Another benefit of credit derivatives is that they add transparency to credit markets (Kroszner 2007). Prior to the existence of credit derivatives, determining a price for credit risk was difficult, and no accepted benchmark existed for credit risk. As credit derivatives become more liquid and cover a wider range of entities, however, lenders and investors will be able to compare pricing of cash instruments such as bonds and loans with credit derivatives. Further, investors will be able to engage in relative value trades between markets, which will lead to further improvements in efficiency and price discovery. At a higher level, economic stability stands to benefit from the ability to transfer credit risk by buying and selling protection. As with other derivatives, the cost of risk transfer is reduced, so risk is dispersed more widely into deeper markets. The result is that economic shocks should have less effect than was the case prior to the existence of derivatives. Several objections can be made to such an argument, however, and these will be considered in the next section. Costs. It is often argued that the flip side of wider and deeper risk transfer is that, instead of exerting a stabilizing influence on markets, it is potentially destabilizing because it transfers risk from participants that specialize in credit risk (that is, banks) to participants with less experience in managing credit risk—for example, insurers and hedge funds (“Risky business” 2005, for example). In addition, there is the danger that anything used to disperse risk can also be used by investors seeking yield enhancement to concentrate risk. Finally, these new institutions generally fall outside the regulatory reach of agencies that oversee various aspects of the credit markets. Such arguments have weaknesses, however. While it is true that banks traditionally specialize in managing credit risk, for example, it is also true that traditional lending has tended to concentrate credit exposures in a narrow class of institutions, namely, commercial banks. Further, one could argue that nonbank institutions might ECONOMIC REVIEW Fourth Quarter 2007 17 F E D E R A L R E S E R V E B A N K O F AT L A N TA in many cases have liability structures that are more suitable than those of banks for bearing credit risks (IMF 2006, chap. 2). But even if one were to accept the questionable argument that nonbank investors are inevitably less skilled than banks at managing credit risk, it would also be the case that credit losses would have less effect on any one institution than was the case when credit was limited mostly to banks. Finally, the argument that credit derivatives increase overall risks by transferring credit risk outside strictly regulated institutions makes an implicit assumption that government regulation automatically leads to more prudent risk-taking. But this argument ignores the potential moral hazard associated with such an assumption. Indeed, because less regulated institutions are less likely to be protected by an official safety net, such institutions are likely to have substantial incentives to identify, measure, and manage credit exposures (Kroszner 2007). Another commonly cited cost of credit derivatives is that they reduce incentives for lenders to analyze and monitor credit quality because they now have the ability to off-load credit risk (Jackson 2007, for example). The result is a decrease in overall credit quality. Again, there are weaknesses to such arguments, mainly that hedging is not costless. As is true with other risks, when one hedges away a risk, one also hedges away the opportunity to profit. A possible exception to this rule would be systematic underpricing of CDS protection relative to loan risk, for which no evidence exists. Another possible exception is a “lemons” argument that lenders use collateralized debt obligations to off-load risks to protection sellers, although one would expect that such a practice, if widespread, would induce CDO note buyers to build expectations of higher losses into the price of the credit protection they provide. Yet another exception would be lenders’ possessing inside information about credit quality, on which they could act by buying underpriced protection. This issue has already received extensive attention by the financial industry (Joint Market Practices Forum 2003), however, and one would not expect such activity to be a systematic feature of credit derivatives markets. A corollary to the argument that lenders with access to credit protection are indifferent to risk is that credit derivatives, as do other forms of risk transfer, inevitably involve a moral hazard effect that leads to higher risk overall (Plender 2006). In other words, risk reduction at the individual entity level can mean higher risk at the system level. Such an argument has an element of plausibility in that, when market participants are able to hedge certain risks, they are able to increase the amount of risks they take overall. But even if firms do take on more risk than before, one could argue that, as long as firms do not take on excessive amounts of risk, the system is in fact safer because the individual institutions that hedge are less vulnerable to market shocks. Recent Credit Derivatives Policy Issues Novations and operational backlogs. The entry of hedge funds and other investors into credit derivatives has been an important factor in the development of CDS market liquidity and efficiency. Such investors enter primarily to take positions. As described above, an investor who believes protection on a reference entity is underpriced can buy protection in anticipation that spreads will widen; if the view turns out to be correct, the fund reverses the transaction at a profit. Similarly, a fund that believes that a distressed credit’s prospects will improve could sell protection in the hope of unwinding at a profit if the improvement occurs. In order to understand the novations problem, one must understand how OTC derivatives trade. OTC derivatives do not trade in the same way as securities, that is, 18 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA by means of transfer of ownership. Instead, they trade “synthetically” by three different means, each of which involves payment by one party to the other of a transaction’s mark-to-market value. First, the parties can agree to a termination (or tear-up), under which they agree to extinguish the original obligation following payment. Second, one party can enter into an offsetting transaction, which leaves the original transaction in place but effectively cancels out its economic effect. Finally, a party can enter into a novation, also known as an assignment, under which the party (transferor) transfers its rights and obligations under the transaction to a third party (transferee) in exchange for a payment. Following the novation, the parties to the transaction are the transferee and the remaining party. The ISDA Master Agreement requires a transferor to obtain prior written consent from the remaining party before a novation takes place. Until relatively recently, novations were relatively infrequent; the usual method of exiting a transaction was an offsetting transaction. But as hedge funds have become more active in CDS, novations have become increasingly common. Investors, and especially hedge funds, tend to prefer to unwind through novation rather than through offset because they are reluctant to incur credit exposure and the resulting need to post collateral on the offsetting swap.4 And they generally prefer novation to termination because termination limits unwind possibilities to the original counterparty and can provide insights into trading strategies to the counterparty. As a result, novations increased, especially as index trading grew; one estimate placed novations at 40 percent of trade volume as of 2005 (CRMPG II 2005). Novations became a problem because of participants’ failure to follow established procedure. First, a hedge fund wishing to step out of a transaction via novation might not obtain prior consent from the remaining party. Second, the transferee might not verify that the transferor had obtained clearance. Finally, the remaining party, which might not have been aware of the novation until the first payment date following, might later back-date its books to the novation date and simply change the counterparty name. The finger pointing went further: When dealers complained that investors had failed to obtain consent, investors countered that remaining parties had given consent but had failed to transmit the necessary information to the back office in a timely manner. Although novations in such cases did not typically lead to significant adverse credit exposures for dealers—transferees are virtually always dealers and therefore better capitalized than the hedge fund transferors—they did present substantial operational problems in the form of confirmation backlogs. The industry was aware of the problem. ISDA addressed the issue in 2004 by developing novation definitions, a standard novation confirmation, and a best-practices statement. Regulators were also aware of the problem and in some cases expressed concern publicly (Evans 2005). A solution did not ensue, however, because of competitive pressures and the lack of incentive to act alone. On the one hand, dealers were aware of the problem and would benefit if all parties to novations followed established procedures. But on the other hand, refusing to agree to novations if procedures were not followed would lead to losing potentially profitable business to those dealers that did not insist on proper procedures. The industry consequently found itself in a “prisoners’ dilemma” situation in which each party would benefit from adhering to proper procedures but had no means of knowing whether other parties would do so as well. The result was no change, and confirmation backlogs increased. 4. Hedge funds generally post up-front collateral (known as the independent amount) with dealer counterparties regardless of the direction of exposure. They then post additional collateral to cover subsequent variations in exposure. ECONOMIC REVIEW Fourth Quarter 2007 19 F E D E R A L R E S E R V E B A N K O F AT L A N TA During August 2005, however, Federal Reserve Bank of New York President Timothy Geithner invited fourteen major credit derivative dealers to a meeting to discuss CDS operations issues, with particular attention to confirmation backlogs. At the meeting, which occurred the following month, the dealers agreed to reduce backlogs and to report their progress periodically. The effort to reduce backlogs led to increased efforts within ISDA to complete a solution to the novations issue. The solution, known as the ISDA Novation Protocol, was announced just before the New York Fed meeting in September (Raisler and Teigland-Hunt 2006). The protocol entailed extensive negotiation between dealers, The entry of hedge funds and other hedge funds, and other participants and investors into credit derivatives has been specified a set of explicit duties for the parties to a novation. Under the protocol, an important factor in the development parties wishing to act as transferees are of CDS market liquidity and efficiency. required to obtain prior consent but are now able to do so electronically. If the remaining party provides consent prior to 6 PM New York time, the novation is complete; the remaining party can respond by email. If the remaining party does not provide consent prior to 6 PM, the transferor and transferee enter into an offsetting transaction that obtains a similar economic result to the novation. Market participants were given a deadline to sign on to the ISDA Novation Protocol; dealers agreed not to transact novations with parties that did not agree. In order to provide assurance that remaining parties would respond promptly to novation requests, dealers committed to specific standards for responding by the deadlines in the protocol. The result has been considered a success: 2,000 parties signed on to the original Novation Protocol, and almost 190 entities have signed on to a version designed for new participants. Initial assessments of the protocol have been favorable. These assessments have corresponded to reports that the industry has made considerable progress in reducing confirmation backlogs and increasing overall operational efficiency. According to the New York Fed, by September 2006 the fourteen largest dealers had reduced the number of all confirmations outstanding by 70 percent and of confirmations outstanding past thirty days by 85 percent. Further, the dealers had doubled the share of trades confirmed electronically to 80 percent of total trade volume (Federal Reserve Bank of New York 2006). The case of novations demonstrates that collective action problems can threaten the feasibility of private sector efforts but that thoughtful regulatory action can facilitate a solution. Although all parties had an interest in a solution, none believed the other side was willing to take the necessary steps. Further, competitive considerations made dealers reluctant to exert pressure on one of their most active client groups. The regulatory intervention provided sufficient cover for dealers to insist on adherence by their clients. In this case, a relatively light touch by a regulator was sufficient to bring about a solution. CDS settlement following credit events. As mentioned earlier, credit derivative index trades have been the major factor in recent growth in credit derivatives. The result of this growth was a new challenge, namely, that the amount of credit protection outstanding is far greater than the supply of underlying debt that could be delivered if a credit event were to occur. The problem manifested itself in a series of corporate bankruptcies in the North American auto parts companies (Collins & Aikman, Delphi, Dana, and Dura), airlines (Delta and Northwest), and power companies 20 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA (Calpine). Because of the expanded interest in credit derivatives caused by the introduction of indices, the amount of credit derivatives outstanding was in some cases reported to be as much as ten times the amount of bonds actually available to settle trades when a credit event occurs. This imbalance called into question the ability of the industry to achieve traditional physical settlement in an orderly manner, which led to calls by industry participants to substitute cash settlement for physical settlement for index trades. The problem was that existing CDS contracts called for physical settlement after credit events. In addition, current documentation provides that the cash settlement will be determined by dealer poll by each dealer, which was not considered feasible given the large number of market participants that would be trying to buy deliverable debt at the same time. With regard to the first problem, counterparties are free to substitute cash for physical settlement if they so agree, but doing so on a large scale could require bilateral negotiations between each pair of counterparties. Further, developing an alternative to the dealer poll required a collective industry solution. The solution proposed by ISDA was that firms move to cash settlement by means of a protocol that allows market participants to amend their contracts on a multilateral basis rather than engaging in bilateral negotiations. In essence, market participants can agree to industrywide and standardized amendments to their contracts. Parties that agree to be bound by the protocol’s terms effectively amend their credit derivative contracts without negotiating directly with other firms. In addition, the protocol provided an alternative means by which a cash settlement price could be determined. The protocol in this case allowed market participants to shift from physical settlement to cash settlement using a price generated in an auction for the defaulted bonds. In the first protocol, held in May 2005 for auto parts supplier Collins & Aikman, there was a single deliverable obligation. The next protocol addressed the bankruptcy filings of U.S. airlines Delta and Northwest, which offered multiple deliverable obligations. Delphi, another auto parts company, was a particularly challenging one as the volume of trades on the name was high: Merrill Lynch and Fitch Ratings estimated that there was $28 billion of exposure on this name but only $2.2 billion par value of bonds available and $3 billion in loans outstanding (Fitch Ratings 2005). Experience with subsequent credit events has led to what appears to be a longterm solution. First, although cash settlement will be the standard, institutions will have the option to settle physically with their dealers if they so choose. Second, the cash settlement protocol will, unlike the early versions, apply to both index and single-name CDS as well as other products such as swaptions. Third, the new system, following further experience, will be incorporated into the next set of credit derivatives definitions. Remaining Issues In considering the future of credit derivatives, two subjects come to mind. The first is the potential for further innovation and growth of credit derivatives. The second is the possibility that credit derivatives might evolve into a standardized, exchangetraded product. Growth and innovation could occur along several dimensions. One dimension is type of contract. There have been some variations on the CDS such as credit swaptions and constant-maturity credit default swaps, but the CDS has proved to be an adaptable product and is unlikely to be displaced. To the extent that product innovation occurs, it is likely to take the form of structured finance applications tailored ECONOMIC REVIEW Fourth Quarter 2007 21 F E D E R A L R E S E R V E B A N K O F AT L A N TA to investor demands for tailored exposures. An example of such a product is the single tranche CDO, which provides investors with exposure to credit risk through a customized CDS portfolio. Another dimension is type of risk. Until recently, CDS were written on entities or groups of entities. But recent innovations have extended CDS protection to obligations instead of industries. CDS on asset-backed securities, for example, have enabled investors to access securitized risks without having to make a direct investment. As a result, supply constraints are less of a factor. Other examples include CDS on leveraged loans and on preferred stock, which again reference financial instruments of a Growth and innovation of credit derivaparticular type. tives could occur along several dimensions: Yet another dimension is new market participants. The major new entrant has type of contract, type of risk, and new been hedge funds. Whether there are other market participants. significant entrants waiting in the wings is not clear, however. Tables 2 and 3 suggest that mutual funds and pension funds have shown some growth, but prudential restrictions on allowable risks might continue to limit the role of such buy-side firms. Second, retail investors might begin to participate, but the history of over-thecounter derivatives, which have remained overwhelmingly wholesale in nature, suggests that retail investors are unlikely to be a major factor. If retail investors do show an appetite for credit investing, they will likely participate through banks and securities firms that serve as intermediaries. A third possibility is that regional banks will increasingly participate in the market, possibly by selling protection as a means of diversifying their portfolios. As Tables 2 and 3 show, however, bank portfolio managers are significantly more likely to use credit derivatives to shed credit risk than to take on credit risk. A final possibility is that nonfinancial corporations might enter the market as they have for interest rate, currency, and commodity derivatives, but so far they have not done so in any significant way (Smithson and Mengle 2006). In the early days of credit derivatives, corporations were considered a potential source of business because of credit risks embedded in corporate balance sheets as receivables and supplier relationships. Further, corporate credit exposures could be a natural hedge for dealers’ exposures to investors and would help dealers balance their CDS portfolios at low cost. Corporate activity did not materialize, however, largely because of basis risks: The nature of corporate exposures is difficult to match with a specific amount of protection and possibly on a specific reference entity. With regard to the evolution of credit default swaps into standardized, exchangetraded futures contracts, several projects are under way as of this writing. First, futures on the iTraxx Europe credit index began trading on Eurex on March 27, 2007. The product is based on a five-year index with a fixed income; a new contract will be issued every six months. The contract will be cash settled, with the payout based on the ISDA CDS settlement auction. Second, the Chicago Mercantile Exchange is developing a futures contract on single-name credit default swaps. Contracts will be written on three reference entities chosen by the exchange. If a credit event occurs, the contracts will have a fixed recovery rate. The CME is also developing an index contract. Third, the Chicago Board Options Exchange is developing a credit default option, which involves an up-front payment and a binary payout of $100,000 per contract if a credit event occurs. Finally, the Chicago Board of Trade and Euronext.Liffe are each planning futures contracts based on a credit index, but details are not yet available. 22 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA The arguments for exchange-traded credit derivatives products are similar to those for other types of derivatives. First, exchanges could provide enhanced liquidity and price discovery by means of standardization and centralized trading. Second, by making the exchange clearinghouse the counterparty to each trade and by imposing universal margin requirements, credit futures could provide a means of reducing counterparty credit risk to users. Finally, credit derivatives might in some cases provide a means for dealers to hedge their exposures as they do for interest rate, commodity, and equity derivatives. Whether CDS trading migrates to exchanges will depend largely on the degree of substitutability between over-the-counter CDS and the new credit futures products. At the same time, whether CDS and credit futures will share the same complementarities that they have in other markets is not clear. With regard to substitutability, OTC derivatives and futures compete to some degree as substitutes but, on closer examination, tend to appeal to different groups of users. If investors perceive the CDS market as being insufficiently liquid, or if counterparty risk is a major consideration, then some volume might move to the exchanges. But index CDS appear to have attained a high degree of liquidity already, so whether they will be motivated to abandon dealers for exchanges is not yet clear. With regard to complementarity, it is not apparent that futures would provide the same hedging and price discovery function that they do for other over-the-counter derivatives. Given the degree of standardization of CDS, dealers are apparently able to trade balanced books without significant residual risks that need to be laid off on exchanges. Further, dealers might not find price discovery information for a small number of selected reference entities to be particularly useful. Still, if credit futures attract significant liquidity, dealers might seek to incorporate the price information into their risk management activities. It is possible, however, that credit derivatives have already evolved into a mature product and that future growth will resemble that of interest rate and other derivatives. That is, products will become increasingly commoditized but will also become known to a wider range of users. The past ten years have seen credit evolving from a largely illiquid product into an increasingly tradable product in which risks are managed in the same way as other market risks. Perhaps the next ten years will see the spread of this new credit risk management technology more deeply and widely into the financial system. ECONOMIC REVIEW Fourth Quarter 2007 23 F E D E R A L R E S E R V E B A N K O F AT L A N TA REFERENCES Bank for International Settlements (BIS). 2007. Triennial and semiannual surveys on positions in global over-the-counter (OTC) derivatives markets at end-June 2007, November. Flannery, Mark. 1985. A portfolio view of loan selection and pricing. In Handbook for banking strategy, edited by Robert A. Eisenbeis and Richard C. Aspinwall. New York: John Wiley and Sons. British Bankers Association (BBA). 2002. BBA credit derivatives report 2001/2002. London: BBA Enterprises Ltd. International Monetary Fund (IMF). 2006. Global financial stability report, April. ———. 2006. BBA credit derivatives report 2006. London: BBA Enterprises Ltd. Chaplin, Geoff. 2005. Credit derivatives: Risk management, trading and investing. West Sussex, U.K.: Wiley. Choudhry, Moorad. 2004. Structured credit products: Credit derivatives and synthetic securitisation. Singapore: Wiley. ———. 2006. The credit default swap basis. New York: Bloomberg Press. Cilia, Joseph. 1996. Product summary: Asset swaps. Financial Markets Unit, Supervision and Regulation, Federal Reserve Bank of Chicago, August. Counterparty Risk Management Policy Group (CRMPG) II. 2005. Toward greater financial stability: A private sector perspective, July. Das, Satyajit. 2006. Traders, guns, and money. Harlow, U.K.: Pearson. Evans, Gay Huey. 2005. Operations and risk management in credit derivatives markets. CEO Letter, Financial Services Authority, February. Federal Reserve Bank of New York. 2006. Statement regarding progress in credit derivatives markets, September 27. Fitch Ratings. 2005. Delphi, credit derivatives, and bond trading behavior after a bankruptcy filing, November 28. ———. 2006. Global credit derivatives survey: Indices dominate growth as banks’ risk position shifts, September 21. ———. 2007. CDx Survey, July 16. 24 ECONOMIC REVIEW Fourth Quarter 2007 Jackson, Tony. 2007. Derivative risk threatens private equity. Financial Times, February 26. Joint Market Practices Forum. 2003. Statement of principles and recommendations regarding the handling of material nonpublic information by credit market participants. International Association of Credit Portfolio Managers, International Swaps and Derivatives Association, Loan Sales and Trading Association, and the Bond Market Association. JPMorgan Chase. 2006. Credit derivatives handbook. Corporate Quantitative Research, December. Kroszner, Randall. 2007. Recent innovations in credit markets. Credit Markets Symposium, Federal Reserve Bank of Richmond, March 22. Plender, John. 2006. The credit business is more perilous than ever. Financial Times, October 13. Raisler, Kenneth, and Lauren Teigland-Hunt. 2006. How ISDA took on the confirmations backlog. International Financial Law Review, February. Risky business. 2005. Economist, August 18. Smithson, Charles. 2003. Credit portfolio management. Hoboken, N.J.: Wiley. Smithson, Charles, and David Mengle. 2006. The promise of credit derivatives in nonfinancial corporations (and why it’s failed to materialize). Journal of Applied Corporate Finance 18, no. 4:54–60. Spinner, Karen. 1997. Building the credit derivatives infrastructure. Derivatives Strategy, Credit Derivatives Supplement, June. Taksler, Glen. 2007. Credit default swap primer, 2d ed. Debt research, Bank of America, January 5. F E D E R A L R E S E R V E B A N K O F AT L A N TA Credit Derivatives and Risk Management MICHAEL S. GIBSON Gibson is a deputy associate director in the Division of Research and Statistics at the Board of Governors of the Federal Reserve System. He thanks Patrick Fleming and Ankur Rughani for excellent research assistance and Michael Gordy, Pat Parkinson, Matt Pritsker, and Pat White for comments on an earlier draft. This paper was presented at the Atlanta Fed’s 2007 Financial Markets Conference, “Credit Derivatives: Where’s the Risk?” held May 14–16. he growth of credit derivatives suggests that market participants find them useful for risk management. Figure 1 shows the growth trajectory for credit derivatives from two surveys of derivatives dealers: the International Swaps and Derivatives Association (ISDA) Market Survey, which goes back to 2001, and the Bank for International Settlements (BIS) semiannual derivatives statistics (2007), which go back to 2004. The BIS survey is more accurate because it adjusts for double counting of interdealer trades, but both surveys show a similar pattern of rapid growth. Notional amounts of credit derivatives outstanding have roughly doubled each year for the past five years. Credit derivatives have been used by a wide variety of market participants. No single data source provides definitive information on the activity of different types of market participants, but combining several available data sources provides a relatively clear picture. I will refer to three data sources: the BIS semiannual derivative statistics (2007), a report on credit risk transfer by the Joint Forum (BIS 2005a), and the surveys by Fitch Ratings (2006). All three data sources measure activity in the credit derivatives market with notional amounts. Notional amounts are often not a good measure of the credit risk that is actually transferred in a particular transaction. However, notional amounts are relatively easy data to collect, and for this reason they are the most common data reported. I will present the notional amount data while keeping their limitations in mind. The most comprehensive data source is the BIS semiannual derivative statistics (BIS 2007). About fifty-five dealers contribute to this survey, which breaks out credit derivative notional amounts by the type of counterparty. Table 1 shows these data for December 2006. The largest category is reporting dealers, reflecting the interdealer nature of the market. In any dealer market, dealers rely on interdealer trading to adjust their risk profiles in response to trading flows from end users. According to dealers, T ECONOMIC REVIEW Fourth Quarter 2007 25 F E D E R A L R E S E R V E B A N K O F AT L A N TA Figure 1 Notional Amounts of Credit Derivatives Outstanding 40,000 35,000 30,000 ISDA Market Survey $U.S. billion 25,000 20,000 15,000 10,000 BIS semiannual derivative statistics 5,000 0 2001 2002 2003 2004 2005 2006 Source: ISDA, BIS only 5 to 10 percent of their notional amount of derivatives represents hedges of their own credit exposures; the balance reflects interdealer trading and accommodation of customer demands (BIS 2005a, 16). Banks and security firms that are not reporting dealers make up one-fifth of the total. Some of this percentage captures nondealer banks investing on their own account in credit derivatives. Some likely captures banks acting as fiduciaries for private banking or high-net-worth investors. The category of “other financial institutions” includes hedge funds, pension funds, and special-purpose vehicles and makes up another fifth of the total. Many structured credit products, including collateralized debt obligations (CDOs), make use of special-purpose vehicles. Hedge funds are active traders but tend to maintain their positions for a short amount of time; their share of trading volume would likely be larger than their share of notional amounts outstanding. This category is the fastest-growing among the nonreporting dealer categories. Insurance firms account for a small portion of outstanding notional amounts, around 1 percent, but are notable for their one-sided participation as net sellers of credit protection to dealers. Of course, exactly how much risk transfer those data represent is unclear, given that notional amounts cannot be equated with risk. The Joint Forum report, which relied on surveys by Fitch Ratings and Standard and Poor’s, also noted that insurance and financial guaranty firms were net sellers of credit protection, along with European banks (BIS 2005a). Banks overall used credit derivatives to shed credit risk. At the banks that took on credit risk with credit derivatives, exposures taken on with credit derivatives were only 2–6 percent of exposures from traditional lending. Large banks tended to be net buyers of credit protection. Fitch Ratings has repeated its survey annually. The most recent survey, done in 2006, confirmed the broad outlines of the patterns discussed above (Fitch Ratings 2006). Insurance and financial guaranty firms remain net sellers of credit protection, mainly through portfolio credit derivatives, a category that includes synthetic CDOs, credit default swap indexes, and credit index tranches. While banks as a group remain 26 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Table 1 BIS Semiannual Derivatives Statistics, December 2006 (Notional Amounts, $Billions) Type of counterparty Dealer bought protection from counterparty Dealer sold protection to counterparty Total (adjusted for double counting) Reporting dealers Nonreporting banks and security firms Other financial institutions Nonfinancial institutions Insurance and financial guaranty firms 16,044 2,928 2,826 561 211 16,165 2,758 2,824 530 95 16,104 5,686 5,650 1,091 306 Total 22,571 22,372 28,838 Source: BIS (2007, table 4) net buyers of credit protection according to Fitch’s data, Fitch noted a shift in its most recent survey. Some individual banks, as well as German banks as a group, shifted to net sellers of credit protection via derivatives. One possibility is that these banks are relying more on securitization, rather than derivatives, to shed credit risk. Fitch also noted that firms responding to its survey reported having sold $377 billion of notional amount of credit protection more than they bought. Firms not reporting to the survey, including hedge funds, asset managers, and pension funds, must be shedding this $377 billion of notional credit risk. Market Participants Using Credit Derivatives for Risk Management Clearly, market participants are finding credit derivatives to be useful tools for risk management to support such rapid growth. To dig deeper into the usefulness of credit derivatives for risk management, I discuss how they are used by three types of market participants: commercial banks, investment banks, and investors. Commercial banks. Commercial banks use credit derivatives to tailor their credit risk exposure. Broadly speaking, they shed credit risk via credit derivatives. Banks have used credit derivatives and other means of credit risk transfer, such as securitizations, to shed risk in several areas of their credit portfolio, including large corporate loans, loans to smaller companies, and counterparty credit risk on over-the-counter (OTC) derivatives. Banks use single-name credit default swaps (CDS) to shed the credit risk of issuers to whom they have a large exposure. Banks can transfer the credit risk of a portfolio of exposures to investors via securitization transactions, such as collateralized loan obligations (CLOs). The Joint Forum (BIS 2005a), reporting on interviews held in 2004 with about sixty market participants, found that the largest commercial banks had shed a material, but small, amount of credit risk via credit derivatives, mainly to their large, investment-grade corporate customers. The Joint Forum also reported that a number of commercial banks had scaled back their credit hedging activity. However, these conclusions may no longer hold. The amount of credit risk shed by banks may be rising, and hedging has spread to categories of credit risk beyond investment-grade corporate loans. A number of banks, mainly European, have undertaken large hedging transactions in the past couple of years. Table 2 reports several recent hedging transactions by large banks. These transactions are larger and more numerous than what had been reported at the time of the Joint Forum survey. In total, these transactions represent the equivalent of $88 billion notional amount of ECONOMIC REVIEW Fourth Quarter 2007 27 ECONOMIC REVIEW Fourth Quarter 2007 Date Bank Name of deal Cash or synthetic? Collateral Amount June 2005 ABN Amro Amstel 2005 Synthetic Corporate loans EUR 10 bil December 2005 ABN Amro Smile 2005 Synthetic Dutch SME loans EUR 6.75 bil November 2006 ABN Amro Amstel 2006 Synthetic Corporate loans EUR 10 bil December 2006 ABN Amro Amstel SCO Synthetic Counterparty exposures on derivatives EUR 7 bil February 2007 ABN Amro Smile Securitization 2007 Cash Dutch SME loans EUR 4.9 bil December 2005 Barclays Gracechurch Corporate Loans Series 2005-1 Synthetic UK midsize corporates GBP 5 bil January 2007 Barclays Gracechurch Corporate Loans 20071 Synthetic UK SME loans GBP 3.5 bil February 2007 Credit Suisse Clock Finance Synthetic Swiss SME loans CHF 4.8 bil July 2005 Deutsche Bank GATE SME CLO Synthetic SME loans EUR 1.5 bil June 2006/ February 2007 Deutsche Bank Craft EM CLO Synthetic Emerging market loans, bonds, and counterparty exposures USD 500m/1 bil February 2007 HSBC Trinkaus HEAT 3 Cash SME loans EUR 314 mil November 2005 HSBC Metrix Funding Cash Corporate loans GBP 2 bil November 2006 HSBC Metrix Securities Synthetic Corporate loans GBP 2 bil November 2006 Mizuho n/a Synthetic Non-Japanese large corporate loans JPY 560 bil October 2006 SocGen Atlas III Synthetic Corporate loans EUR 2.8 bil November 2006 UBS n/a Bilateral swap High-yield corporate loans USD 600 mil Note: “SME” stands for small and medium-sized enterprises. Source: Company and rating agency reports and financial press F E D E R A L R E S E R V E B A N K O F AT L A N TA 28 Table 2 Recent Hedging Transactions by Large Banks F E D E R A L R E S E R V E B A N K O F AT L A N TA Table 3 Hedging Done with Credit Default Swaps Date Bank Year-end 2006 Year-end 2006 Year-end 2006 2006 Q1 Bank of America Citigroup JPMorgan Chase Société Générale Credit exposure before hedging (billions) Amount of hedging reported (billions) Exposure hedged (percent) USD 618 USD 633 USD 631 EUR 60 USD 8 USD 93 USD 51 EUR 15 1 15 8 25 Source: For U.S. banks, 2006 annual reports; for Société Générale, “Safety first,” Risk, August 2006 credit risk shed by eight large international banks over 2005–07.1 In many of these transactions, and in contrast to similar transactions in the late 1990s, the issuing bank sold off the first-loss equity tranche of the credit risk. The categories of credit risk shed include not only loans to large corporates but also loans to small and medium-sized enterprises, loans to emerging markets, and counterparty exposure on derivatives. Most transactions listed in the table are synthetic, using credit derivatives to transfer risk off the balance sheet. Table 3 shows reported amounts of CDS hedging by the three largest U.S. commercial bank holding companies, as reported in their 2006 annual reports, and by one European bank, as reported in the financial press. (Only U.S. banking organizations appear to disclose CDS hedging in their annual reports.) For this admittedly small sample, the average percentage of credit risk hedged appears to be larger than that reported by the Joint Forum. Several reasons could explain why commercial banks appear to be hedging more of their credit risk than they were in 2004. First, credit spreads are at low levels, reducing the cost of hedging. Second, accounting changes in Europe have made it possible for banks to carry loans at fair value, reducing the conflict that was perceived between the accounting treatment of credit derivatives and their use in risk management (BIS 2005a, 11). Third, the Basel 2 capital accord will align regulatory capital charges more closely with actual credit risks and will allow greater recognition of hedging. Investment banks. An investment bank can use credit derivatives to manage the risk it incurs when underwriting securities. An underwriter assumes credit risk for the short time period between taking the risk onto its own books and selling it into the market. By virtue of the growth of credit derivatives, the underwriter might then be able to hedge some of that credit risk more easily. Nonagency residential mortgage-backed securities (RMBS) have been a rapidly growing market for securities underwriting in recent years. In 2006, $574 billion of securities were underwritten and issued in this segment, up from less than $100 billion in 2000 (SIFMA 2007b). The increase in issuance volume would naturally lead to a rise in credit risk borne by underwriters because underwriters must warehouse residential mortgage loans on their books during the time it takes to assemble a pool large enough to launch a securitization. Underwriters must find a way to cope with the potential increase in credit risk, which, given the numbers cited above, might be 1. According to global CDO market issuance data reported by the Securities Industry and Financial Markets Association (SIFMA) (2007a), $107 billion of balance-sheet CDOs were issued in 2005–06. ECONOMIC REVIEW Fourth Quarter 2007 29 F E D E R A L R E S E R V E B A N K O F AT L A N TA so large as to discourage them, at the margin, from taking on additional underwriting business. One way for underwriters to cope with such a potential increase in credit risk is to hedge more of it. New credit derivative instruments appear to be useful to underwriters who want to hedge the risk of a residential mortgage loan warehouse. Beginning in mid-2004, dealers began to trade credit default swaps on asset-backed securities (referred to as Banks have used credit derivatives and ABS CDS). By year-end 2005, Fitch Ratings other means of credit risk transfer, such as (2006, 2) put the size of the ABS CDS securitizations, to shed risk in several areas market at $495 billion in notional amount outstanding and growing rapidly.2 An underof their credit portfolio. writer can use an ABS CDS to buy credit protection on an RMBS with similar characteristics to the loans in its warehouse. The performance of the ABS CDS should roughly offset the performance of the warehouse loans. As is typical of successful and liquid new markets, there appears to be a healthy balance of supply and demand of credit risk in the ABS CDS market. In addition to underwriters seeking to hedge warehouse loans, asset managers with a negative view on the housing sector are also natural buyers of credit protection on RMBS. Investors seeking exposure to the RMBS market, including CDOs, are natural sellers of credit protection. ABS CDS have proved to be relatively liquid compared to the markets for individual RMBS. Using ABS CDS on large, recently issued RMBS, dealers created an index called the ABX.HE, which can be used to trade the credit risk of a pool of twenty RMBS deals. Of course, if sellers of credit protection become scarce because of a weaker-than-expected housing market, the ABS CDS and ABX.HE markets could see much of their recent liquidity dry up, and underwriters would lose a useful tool for credit risk management. Investors. Investors are the third group that uses credit derivatives for risk management. An investor can use credit derivatives to align its credit risk exposure with its desired credit risk profile. Credit derivatives can be more flexible and less expensive than transacting in cash securities. The surveys cited earlier show that insurers are an important class of investors that use credit derivatives. However, other fixedincome asset managers, including hedge funds, also participate in the market. Most observers agree that of the three groups, demand from investors is most responsible for spurring the growth of the credit derivatives market. “Investors” are a heterogeneous group, and they participate in the credit derivatives market in different ways. Some are “buy and hold” investors that seek to earn a return from a broad exposure to issuers of fixed-income securities, and others are “active traders” that seek to earn a return by predicting short-term price movements better than other market participants. The advantages of credit derivatives as a riskmanagement tool are different for the two groups. Traditionally, insurance companies and pension funds have been thought of as buy-and-hold investors, and hedge funds have been thought of as active traders. But the distinctions between different types of asset managers are becoming increasingly blurred. A given asset manager may place some assets in a buy-and-hold index strategy, some with an in-house team of active traders, and the remainder with external managers who could pursue either type of strategy. Buy and hold. Suppose a buy-and-hold investor develops a negative view about a particular sector—for example, telecom. Consider an investor that does not use credit derivatives. It can only rebalance its portfolio away from telecom issuers by 30 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Table 4 Alternative Scenarios Faced by an “Active Trader” Scenario 1. XYZ defaults 2. XYZ credit spread falls by 10 basis points at all maturities 3. XYZ credit spread increases by 10 basis points at all maturities Change in market value Sell $10MM Buy $10MM at ten years at five years Deliver bonds +$76,347 –$76,347 Receive bonds –$43,837 +$43,837 Net Zero +$32,510 –$32,510 Source: Bloomberg page CDSW, using a BBB spread curve as of April 16, 2007 selling some of the telecom bonds it holds. Secondary markets for corporate bonds are notably illiquid for seasoned issues, so the transaction cost of selling seasoned telecom bonds will be high. The investor’s best option for replacing the telecom bonds will be newly issued bonds, which have low transaction costs because they are relatively liquid. However, the particular bonds that happen to be issued at a given time may be influenced by the particular trends in the market at that time, such as which sectors are undergoing leveraged buyouts, and may not be exactly the replacements that the investor is looking for. Now consider an investor that does use credit derivatives. This investor can shift its exposure away from telecom issuers by buying credit protection on telecom issuers using credit default swaps. The bid-ask spread is generally lower for credit default swaps than for corporate bonds, and the difference is larger when the bonds are seasoned. To replace the telecom exposures, this investor can sell credit protection on other, nontelecom issuers or simply sell credit protection on a credit default swap index. Active trader. Now consider an investor that is an active trader with a view that, over the next three months, Issuer XYZ’s credit risk standing will improve and its credit spreads will tighten. One obvious trade based on such a view is to buy one of Issuer XYZ’s bonds or sell credit protection on Issuer XYZ with a single-name credit default swap. However, buying a bond or selling credit protection exposes the investor to the risk that Issuer XYZ defaults, a risk the investor may not want to take. As mentioned above, one benefit of credit derivatives is that an investor can use them to take a customized exposure to particular components of credit risk, such as spread risk, default risk, recovery risk, or correlation risk. In this example, the investor wants to be exposed to the spread risk of Issuer XYZ but not default risk. To achieve this goal, suppose that the investor sells $10 million notional amount of credit protection on Issuer XYZ with a ten-year maturity and buys $10 million notional amount of credit protection on Issuer XYZ with a five-year maturity. These two positions have the same $10 million exposure to default risk, but the longer maturity position has a greater sensitivity to credit spreads (higher credit duration). Table 4 shows what happens in three different scenarios. Scenario 1 shows what happens if Issuer XYZ defaults. The investor will receive $10 million face value of Issuer XYZ’s bonds on the five-year CDS and will deliver $10 million face value of bonds on 2. While the asset-backed securities market includes securities backed by a range of collateral, including credit card loans and auto loans, nearly all ABS CDS contracts reference RMBS or commercial mortgage-backed securities. ECONOMIC REVIEW Fourth Quarter 2007 31 F E D E R A L R E S E R V E B A N K O F AT L A N TA the ten-year CDS. Clearly, such a trade is hedged against the default of Issuer XYZ within the next five years. Scenarios 2 and 3 show what happens Tranche Five-year Ten-year when Issuer XYZ’s credit spread curve narrows or widens at all maturities in a paral0–3 percent 500 + 9.98% 500 + 40.85% lel shift. As expected, in scenario 2, the up-front up-front issuer gains on net when the credit spread 3–6 percent 46 334 narrows, and the opposite occurs in sce6–9 percent 13 88 nario 3 when the credit spread widens. Of 9–12 percent 6 39 course, credit spread curves do not always 12–22 percent 2 13 shift in parallel, and an additional risk of Index 23 43 this trade (not shown in the table) is that Source: www.creditfixings.com the credit spread curve steepens. Without credit derivatives, such a trade would only be possible if Issuer XYZ hapTable 6 pened to have bonds outstanding with Deltas on iTraxx Europe Tranches five-year and ten-year maturities and if the on March 1, 2007 investor could borrow a bond to establish a Tranche Five-year Ten-year short position. While the stars may align on occasion for both of these conditions to be 0–3 percent 27.5 13 satisfied, a liquid credit derivatives market 3–6 percent 4 11 clearly offers more possibilities for cus6–9 percent 1.25 4.25 tomizing risk exposures along these lines. 9–12 percent 0.6 2.1 Credit index tranches. Credit index 12–22 percent 0.2 0.8 tranches are another example of how credit Index 1 1 derivatives can produce different riskSource: www.creditfixings.com return trade-offs. A credit index such as CDX (in North America) or iTraxx (in Europe) is a liquid product that provides exposure to a broad segment of the credit derivatives market. Credit index tranches take the risk of a credit index and divide it into pieces with different seniority. Because these tranches on credit indexes are standardized, they are relatively liquid compared to other tranched credit products, which are usually customized on a one-off basis. Table 5 shows the tranches for the iTraxx Europe index, along with the spreads on each tranche at the five- and ten-year maturities as of March 1, 2007. The spread represents the cost paid by a buyer of credit protection. Understanding the relative risk of credit index tranches is difficult but is obviously important for investors who are choosing the risk and return of their investment portfolio. A common way that market participants compare the risk of different tranches is to use a model to compute the relative size of the position in the underlying index that would have the same sensitivity to a small movement in the index credit spread as the tranche. This measure is called “delta,” and, by construction, the delta of a position in the index equals one. Delta can be seen as a measure of the tranche’s leverage. Table 6 shows the deltas of the iTraxx tranches. The deltas themselves purport to measure the risk of a tranche relative to a position in the index. One can also use deltas to compare the risk of different tranches. For example, at the fiveyear maturity, the 3–6 percent tranche is twenty times riskier than the 12–22 percent tranche. Table 5 Spreads on iTraxx Europe Tranches on March 1, 2007 (Basis Points per Annum) 32 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Table 7 Market Spreads and Expected Weighted Average Spreads on iTraxx Europe Tranches on March 1, 2007 (Basis Points per Annum) Five-year Tranche 3–6 percent 6–9 percent 9–12 percent 12–22 percent Index Ten-year Market Expected Market Expected 46 13 6 2 23 45.5 12.8 5.8 2 22.9 334 88 39 13 43 289 84 38 12.9 42.2 Source: Author's calculations However, delta only measures one dimension of a tranche’s risk: exposure to credit spread risk. Other dimensions of risk, such as default risk, may give a different sense of the relative risk of different tranches. In the field of derivatives pricing, the risk and return of different positions that are exposed to the same single underlying risk factor must satisfy the formula expected excess return = a constant. risk This idea underpins the Black-Scholes formula and many other asset-pricing models. I will apply this idea to get a back-of-the-envelope sense of the relative default risk of iTraxx tranches. Defaults among names in the iTraxx Europe index correspond to the single underlying risk factor of the theory. The 0–3 percent tranche is quite sensitive to the timing of defaults, in addition to the number of defaults, so it does not fit the assumption of exposure to a single risk factor, and I therefore omit it from this analysis. I compute the expected excess return on each tranche as the expected weighted average spread less the expected default loss. Each of these is measured in basis points per annum. To compute the expected weighted average spread per annum, I start with the market spreads from Table 5. The spread actually received is less than the market spread because the spread is paid on the remaining balance outstanding on the tranche, which can be reduced when defaults hit the tranche. Using the singlefactor Gaussian copula model of Gibson (2004), I use single-name CDS spreads on the 125 names that make up the iTraxx Europe index to compute the probability distribution of defaults on the index. This calculation assumes a flat correlation of 15 percent across all names. Using this market-implied probability distribution of defaults, Table 7 shows the expected weighted average spread on each tranche from Table 5, along with the market spread. I compute the expected default loss using the same model of Gibson (2004) and the same flat correlation of 15 percent, but I use historical default probabilities for each credit in the index based on its credit rating and assume a constant recovery rate on defaulted issuers of 40 percent.3 The expected default losses for each tranche are shown in Table 8. 3. Historical default probabilities are taken from Moody’s Investors Service (2007, exhibit 26). ECONOMIC REVIEW Fourth Quarter 2007 33 F E D E R A L R E S E R V E B A N K O F AT L A N TA Subtracting the expected default loss in Table 8 from the expected weighted average spread in Table 7 gives the expected excess return for each tranche. This calcuTranche Five-year Ten-year lation is the concept in the numerator of the above formula. Using the formula, 3–6 percent 29 73 the ratio of expected excess return for 6–9 percent 3.2 14 two tranches of the same maturity 9–12 percent 0.5 3 should equal the ratio of risk for the two 12–22 percent 0.03 0.3 tranches. Dividing the expected excess Index 15 16 return of each tranche by the expected Source: Author's calculations excess return of the index gives a relative ranking of the risk of the tranches, similar to the concept of delta introduced above. Table 9 Table 9 shows the result of my backRelative Risk of iTraxx Europe Tranches on of-the-envelope calculation to compare March 1, 2007 (Risk of Underlying Index = 1) the sensitivity to default risk of the iTraxx Tranche Five-year Ten-year tranches. Perhaps it is comforting that when one compares Table 9 with the deltas 3–6 percent 2.0 8.1 in Table 6, the two independent risk mea6–9 percent 1.2 2.6 sures agree on which tranches are more or 9–12 percent 0.7 1.3 less risky than the index (that is, which 12–22 percent 0.3 0.5 have a relative risk greater or less than Index 1 1 one). However, the actual numbers in the two tables differ by quite a bit in some Source: Author's calculations cases. For example, according to Table 6, the 3–6 percent five-year tranche is four times riskier than the index, while according to Table 9, it is only twice as risky. The conclusion from this discussion of the risk of credit index tranches is that, while they broaden the range of risk-return choices that investors have available in the credit markets, they also pose a challenge to understand the various dimensions of risk they are exposed to. I now turn to a more general discussion of the riskmanagement challenges posed by credit derivatives. Table 8 Expected Default Loss on iTraxx Europe Tranches on March 1, 2007 (Basis Points per Annum) Credit Derivatives and Risk-Management Challenges The first half of this paper has shown how commercial banks, investment banks, and investors use credit derivatives for managing credit risk. However, credit derivatives pose risk-management challenges of their own. In the second half of the paper, I discuss five of these challenges. Credit risk. One fundamental reality of credit derivatives is that they do not eliminate credit risk. They merely shift it around. As a result, when the credit cycle turns and default rates rise, someone, somewhere, will lose money. Consider Figure 2, which shows global speculative grade default rates since 1980. Clearly, no one should be surprised if, when the credit cycle turns, the speculative grade default rate hits 10 percent, which is what it hit in 1990–91 and in 2001. Although credit derivatives cannot eliminate losses from credit risk, they can transform credit risk in intricate ways that may not be easy to understand. This issue does not arise with single-name credit default swaps, where the exposure is nearly identical to that of a corporate bond, or with credit default swap indexes, where the exposure is nearly identical to that of a portfolio of corporate bonds. But where complex 34 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Figure 2 Speculative Grade Default Rate 12 10 Percent 8 6 4 2 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Source: Moody’s Investors Service (2007, exhibit 21) credit derivatives such as CDO tranches are concerned, a legitimate risk-management issue may develop. Do market participants understand their exposures to credit risk that they have taken on with complex credit derivatives? Given the breadth of market participants who are active in the credit derivative market, there is no definitive way to answer this question. However, we can point to evidence from the last credit cycle that some market participants did not fully understand the exposures they had from their participation in the credit derivatives market. In 2001, American Express “lost hundreds of millions of dollars on investments in collateralized debt obligations.”4 The chief executive officer of American Express was quoted as saying he “did not comprehend the risk” of its CDO holdings. The U.K. bank Abbey National was reported to have suffered “disastrous losses in its high-yield portfolio, including CDOs”5 and, as a result, to have liquidated its wholesale credit portfolio, including selling off $8 billion of CDO tranches in 2003 (Lucas, Goodman, and Fabozzi 2006, 383). In both cases, the banks were reported to have retained first-loss tranches of CDOs they had underwritten. And first-loss tranches naturally contain a great deal of credit risk. CDO investors, too, naturally suffered losses in the 2001–02 credit cycle. But since many investors do not account for their investments on a mark-to-market basis or may not make detailed accounting statements public, less information was made available in the public domain about these losses. According to reports in the financial media, dealers commonly restructured CDO tranches that were exposed to troubled issuers. Credit risk would have manifested itself in a decline in the mark-to-market value of such a tranche. The dealer could replace a troubled issuer in the CDO’s reference portfolio with a less risky issuer. The mark-to-market loss would be “paid for” 4. See Financial Times, “Out of depth in the collateralized debt pool,” July 22, 2001. 5. See Euromoney, “Market dislocation boosts CDO trading,” April 2003. 35 F E D E R A L R E S E R V E B A N K O F AT L A N TA by lowering the coupon that the investor would receive on the CDO tranche for the remaining life of the deal.6 This brief review of the experience in the last credit cycle of 2001–02 reinforces the point that credit derivatives do not eliminate losses from credit risk. These lessons that I have reviewed here are certainly no secret to participants in the credit markets, many of whom had first-hand experience of living through that credit cycle. Given the rapid growth of the credit derivatives market, it may be fortunate that one of the most widely used complex credit derivative structures, the CDO tranche, is a mature product has already been through Although credit derivatives cannot eliminate a stressful credit cycle. This experience should contribute to financial stability durlosses from credit risk, they can transform ing the next credit cycle, whenever that credit risk in intricate ways that may not be may come to pass. Of course, new flavors of CDOs will easy to understand. always present new challenges. One relatively new product is a CDO using assetbacked securities for collateral instead of corporate debt. In 2006, 60 percent of CDO issuance used asset-backed securities as collateral (SIFMA 2007a). These CDOs transfer the credit risk of asset-backed securities, primarily RMBS. Given the slowing growth of house prices in recent months, credit risk in the RMBS sector is likely to be increasing. Counterparty risk. Counterparty risk is the risk that the counterparty to a credit derivative contract will default and not pay what is owed under the contract. For credit derivatives, as with other OTC derivatives, counterparty risk is an important risk that needs to be managed. Given the growing role of hedge funds in the credit derivatives market, counterparty risk is becoming even more prominent, since hedge funds generally are among a dealer’s riskier counterparties. In many cases, dealers use collateral to reduce counterparty risk. According to the 2006 ISDA Margin Survey, 63 percent of all counterparty risk exposure on credit derivatives is currently collateralized by large dealers. For hedge fund counterparties, a larger share is likely to be covered by collateral since dealers nearly universally require hedge fund counterparties to post collateral to cover current credit exposures. However, despite the widespread use of collateral and margin, some important riskmanagement challenges are associated with counterparty risk on credit derivatives. One challenge is simply measuring the exposures on complex credit derivatives. One of the key measures of counterparty risk is potential future exposure. Potential future exposure takes into account the possible future moves in credit spreads or future defaults that could create a larger credit exposure if the market moves in the dealer’s favor. This potentially larger credit exposure is something that is already present in the current derivative contract and therefore should be measured like any other credit exposure. Market participants are aware of not only the need to measure potential future exposure on complex credit derivatives but also the difficulties in measuring it. As one article by a practitioner puts it, “unfortunately, models that can estimate [counterparty risk exposure] exactly are hard to build and calibrate” (Pugachevsky 2006, 372). That article describes a technique to approximately measure counterparty risk exposure on synthetic CDO tranches, defining counterparty risk exposure as the amount that would be expected to be lost if the counterparty defaults in the future. In a stylized example of CDO tranches with a notional amount of $5 million, the article estimates the counterparty risk exposure to be around $50,000, or 1 percent of notional. Certainly that seems like a material amount of counterparty risk. 36 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA According to an estimate from www.creditflux.com, there were $450 billion of synthetic CDO tranches and $1.7 trillion of credit index tranches traded in 2006. One percent of this roughly $2.2 trillion in notional amount would total $22 billion in counterparty risk exposure, the amount that dealers would collectively expect to lose if all their CDO counterparties simultaneously defaulted. If two-thirds of that is collateralized, dealers in aggregate would have roughly $7 billion in uncollateralized counterparty risk exposure currently in their portfolios, before accounting for hedging. These figures are certainly only a very rough approximation of the order of magnitude of the counterparty risk created by complex credit derivatives. In particular, the actual loss from counterparty default could well be larger than the expected loss. And, of course, any counterparty credit exposure amount should be compared with a dealer’s capital that is available to absorb potential losses.7 All told, it appears that counterparty risk should be a material concern of participants in the credit derivatives market. Model risk. Complex credit derivatives require complex models for valuation and hedging. While a few complex credit derivatives, such as credit index tranches, are traded in liquid markets with some price transparency, most are not. Products without a liquid market are referred to as “mark-to-model.” The risk of loss due to a flawed model is known as model risk.8 Model risk materialized in the market for tranched credit derivatives in May 2005.9 Following the downgrade of General Motors to below-investment-grade status, the market prices of some credit index tranches moved in ways that would be considered as either extremely implausible or impossible, according to the way certain models were being used for valuation and risk management at that time. For example, in the first week of May 2005, the credit spread on the CDX.NA.IG index widened, signaling higher credit risk, but the spread on the 3–7 percent mezzanine tranche tightened, signaling lower credit risk. Market commentary attributed this to an imbalance of market liquidity in the mezzanine tranche market. In some cases the models themselves were not the problem, but models were being used in a way that gave false confidence about the effectiveness of hedging strategies. In fairness to those who build models for a living, it has to be said that the flaws that were revealed by the May 2005 episode were not a surprise to many model builders. Even before May 2005, modelers were documenting the flaws of the standard 6. One example of such a report appeared in Risk magazine in September 2004: “By autumn 2002, all the talk in the structured credit community was about restructuring. Investors were increasingly looking for resilience in the ratings of their holdings. In its public CDO ratings, Moody’s began explicitly mentioning the removal of WorldCom from portfolios as a reason for ratings upgrades. “But such new-found resilience comes at a price. Although some claim it is a cheaper option than selling an entire CDO tranche, restructuring is by definition an expensive process, involving the unwinding of distressed default swap positions and replacing them with stronger credits at tighter spreads. And that process is likely to mean wider bid/offer spreads for investors” (Dunbar 2004, 32). 7. For comparison, the three largest U.S. bank holding companies each had over $75 billion in tier 1 capital in 2006. 8. Rebonato (2001) provides the following more complete definition of model risk: “Model risk is the risk of occurrence at a given point in time (today or in the future) of a significant difference between the mark-to-model value of a complex and/or illiquid instrument held on or off the balance sheet of a financial institution and the price at which the same instrument is revealed to have traded in the market—by brokers’ quotes or reliable intelligence of third-party market transactions—after the appropriate provisions have been taken into account.” 9. The International Monetary Fund (2005, 21–23) describes the episode. ECONOMIC REVIEW Fourth Quarter 2007 37 F E D E R A L R E S E R V E B A N K O F AT L A N TA model used for tranched credit products, the Gaussian copula model. As one paper published in 2004 noted, “despite the popularity of the Gaussian copula model, there are clear and valid questions over its theoretical foundations” (Gregory and Laurent 2004). Of course, any model is only an approximation of reality, and model improvement must be a continuous process for products as new as tranched credit derivatives. In the two years since the May 2005 episode, research into alternatives to the Gaussian copula model has exploded. While eventually this research is likely to lead to better models and a reduced level of model risk for complex credit derivatives, there could be a long wait until that occurs. For the foreseeable future, those who trade comMarket participants are aware of not only plex credit derivatives will need to pay the need to measure potential future expocareful attention to measuring and managing their exposure to model risk. sure on complex credit derivatives but also Rating agency risk. Rating agencies the difficulties in measuring it. play an important role in the credit derivatives market. As noted in a recent central bank research report, the structured finance market, including the credit derivatives market, relies heavily on ratings (BIS 2005b). Given the complex nature of many credit derivatives, many investors rely on rating agencies to assess the credit risk of a particular transaction. However, according to that report, large institutional investors do not rely solely on ratings for making investment decisions. The debate over the role of rating agencies in the market for complex credit derivatives has two sides. One side argues that rating agencies are fully transparent in the methodologies they use to rate synthetic CDOs. They publish detailed criteria reports that are available to the general public without charge, and in some cases they allow their models to be freely downloaded. They implicitly acknowledge that their ratings of structured finance transactions are fundamentally different than their ratings of corporate debt, for example, by compiling and publishing separate default and migration statistics for the two groups, rather than pooling them into a single group. This approach should discourage investors from treating an AAA rating on a structured credit derivative exactly like an AAA rating on a corporate bond. The other side of the debate argues that the one-dimensional nature of traditional credit ratings makes them insufficient for comparing the risk of corporate debt and structured credit derivatives and that using the same rating scale for the two is misleading. While the expected loss or probability of default of a BBB-rated corporate bond and a BBB-rated synthetic CDO tranche may be the same, their risk differs materially in other important dimensions. For example, synthetic CDO tranches are much more sensitive to the credit cycle, or to business cycle risk, than a portfolio of similarly rated corporate debt (Gibson 2004). Calculations reported in Gibson (2004) show the expected loss measured as a percent of notional amount on a stylized mezzanine CDO tranche in a recession could be eight times larger than on a portfolio of corporate bonds. The tranches of the iTraxx index, discussed in detail in the first half of this paper, can also be used to illustrate some of the points about rating agency risk. A credit rating provides a third way to look at the relative risk of the various tranches. Table 10 reproduces the two relative risk measures introduced above along with the ratings associated with each ten-year maturity tranche from one rating agency. In several ways, the ratings give a quite different message about the relative risk of various tranches. According to the rating, the 6–9 percent tranche is less risky, rated A, than the index itself, rated A–/BBB+. But the other two risk measures consider that tranche 38 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA to be at least twice as risky as the index. Table 10 Various Relative Risk Measures for iTraxx Europe The tranche’s spread of 88 basis points is Tranches at Ten-Year Maturity on March 1, 2007 twice as high as the index’s spread of 43 basis points. Comparing the ratings of the Spread risk Default risk 6–9 percent and 9–12 percent tranches Tranche (Table 6) (Table 9) Rating shows a large difference between an AAA rating and an A rating. In terms of histori3–6 percent 11 8.1 B+ cal default probability for corporate bonds, 6–9 percent 4.25 2.6 A an A rating has roughly ten times higher 9–12 percent 2.1 1.3 AAA default probability than AAA over a ten12–22 percent 0.8 0.5 AAA Index 1 1 A–/BBB+ year horizon. Yet the other two risk measures consider the 6–9 percent tranche to Source: Author's calculations; ratings from Fitch Ratings as of be roughly twice as risky as the 9–12 perSeptember 20, 2006 cent tranche. And it seems particularly odd that the 9–12 and 12–22 percent tranches can be rated the same when no scenario exists in which the 12–22 percent tranches takes a single dollar of loss without the 9–12 percent tranche losing its entire principal amount. Without taking a stand on which risk measure is better or worse, it seems clear that relying on a rating to tell the entire story of the risk on a tranched credit derivative product is a bad idea. Settlement risk. When an issuer defaults, credit derivatives that reference the issuer’s debt must be settled. Traditionally, settlement in the CDS market was based on physical delivery by the protection buyer of the referenced issuer’s debt securities in exchange for par. Physical settlement is the natural settlement mechanism when a CDS is used to hedge the credit risk of owning a bond. Cash settlement is less desirable in that situation because the value of owning the bond of the defaulted issuer may diverge from the cash settlement price on a CDS, reducing the effectiveness of the hedge. As the credit derivative market has grown, it has become common for the notional amount of CDS outstanding referencing a particular issuer to be larger than the face value of the issuer’s bonds outstanding. In October 2005, Delphi Corporation defaulted with $2 billion of deliverable bonds and approximately $28 billion of credit derivatives outstanding. Because settlement must occur within a fixed time period after a default, a single bond can be used (and reused) for settlement of CDS only so many times. The potential exists for an artificial scarcity of the bonds of defaulted issuers that are needed for CDS settlement, driving up the price of the bonds. In the worst case, if the protection buyer cannot obtain the bonds it needs to settle its contracts by the deadline, the contract expires worthless. This situation has the potential to affect the price of CDS in advance of a default, making CDS less useful as hedges and distorting the price signals that the CDS provides to the market. Since the growth of the credit derivatives market shows no signs of slowing down, settlement risk is likely to continue to increase as long as physical settlement is the standard in CDS contracts. Market participants are certainly aware of the issue and are working on a solution. In the wake of the Delphi default, dealers rushed to organize a cash settlement auction in which more than 570 counterparties participated. Although all participants in the credit derivatives market have a broad interest in seeing the market function well, their interests may diverge in a settlement situation when some are protection buyers, some are protection sellers, some would probably prefer physical settlement, and some would prefer cash settlement. Getting marketwide agreement on an auction mechanism may not be easy, especially when the ECONOMIC REVIEW Fourth Quarter 2007 39 F E D E R A L R E S E R V E B A N K O F AT L A N TA agreement is made after the default occurs. Moreover, the example of the European auctions of mobile-phone licenses reinforced the basic fact that differences in auction design can lead to vast differences in outcomes (Klemperer 2002). The auction mechanism that was used for the Delphi auction in November 2005 has been tweaked since then to discourage gaming and to encourage broader participation. In the most recent large default in the CDS market, Dura Automotive Systems in late 2006, the most recent auction mechanism was tested and seemed to work well. However, each auction is an ad hoc process that must be quickly agreed to following a default. Settlement risk will still be high until the auction settlement mechanism is incorporated into standard CDS documentation and is tested in actual defaults, including some in less benign market environments. Conclusion I have documented the striking growth of credit derivatives, from nearly nothing a decade ago to tens of trillions of dollars in notional amounts outstanding at the end of last year. Driving this growth, market participants—including commercial banks, investment banks, and investors—appear to find a variety of credit derivative products to be useful for their own risk-management purposes. I discussed a number of the ways that credit derivatives can be useful for risk management. At the same time, credit derivatives are posing some significant risk-management challenges. Many of these challenges reflect the immaturity of the credit derivatives market. For the credit derivatives market to develop and mature, market participants must address these risk management challenges. 40 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA REFERENCES Bank for International Settlements (BIS). 2005a. Credit risk transfer. The Joint Forum, March. <www.bis.org/publ/joint13.pdf> (November 2, 2007). ———. 2006. ISDA margin survey 2006. <www.isda.org/c_and_a/pdf/ISDA-Margin-Survey2006.pdf> (November 27, 2007). ———. 2005b. The role of ratings in structured finance: Issues and implications. Committee on the Global Financial System Publication No. 23, January. <www.bis.org/publ/cgfs23.pdf> (November 2, 2007). Klemperer, Paul. 2002. How (not) to run auctions: The European 3G telecom auctions. European Economic Review 46, nos. 4–5:829–45. ———. 2007. OTC derivatives market activity in the second half of 2006. Monetary and Economic Department, May. <www.bis.org/publ/otc_hy0705.pdf> (November 2, 2007). Dunbar, Nicholas. 2004. Seduced by CDOs. Risk (September): 38–44. Fitch Ratings. 2006. Global credit derivatives survey: Indices dominate growth as banks’ risk position shifts, September 21. Gibson, Michael S. 2004. Understanding the risk of synthetic CDOs. Board of Governors of the Federal Reserve System, Finance and Economics Discussion Series 2004-36, July. <www.federalreserve.gov/pubs/ feds/2004/200436/200436pap.pdf> (November 2, 2007). Gregory, Jon, and Jean-Paul Laurent. 2004. In the core of correlation. Risk (October): 87–91. International Monetary Fund. 2005. Global Financial Stability Report: Market Developments and Issues. September. <www.imf.org/External/Pubs/FT/GFSR/ 2005/02/index.htm> (November 2, 2007). Lucas, Douglas J., Laurie S. Goodman, and Frank J. Fabozzi. 2006. Collateralized debt obligations: Structures and analysis, 2d ed. Hoboken, N.J.: John Wiley & Sons. Moody’s Investors Service. 2007. Corporate default and recovery rates, 1920–2006. Pugachevsky, Dmitry. 2006. Pricing counterparty risk in unfunded synthetic CDO tranches. In Counterparty Credit Risk Modeling, edited by Michael Pykhtin. London: Risk Books. Rebonato, Riccardo. 2001. Managing model risk. In Mastering risk—volume 2: Applications, edited by Carol Alexander. London: Financial Times/Prentice Hall. Securities Industry and Financial Markets Association (SIFMA). 2007a. Global CDO Market Issuance Data. <www.sifma.org/research/pdf/SIFMA_CDOIssuanceData 2007q2.pdf> (November 2, 2007). ———. 2007b. Research Quarterly, February. <www.sifma.org/research/pdf/Research_Quarterly_ 0207.pdf> (November 2, 2007). International Swaps and Derivatives Association (ISDA). Various years. ISDA Market Survey historical data. <www.isda.org/statistics/historical.html> (November 27, 2007). ECONOMIC REVIEW Fourth Quarter 2007 41 F E D E R A L R E S E R V E B A N K O F AT L A N TA Credit Derivatives, Macro Risks, and Systemic Risks TIM WEITHERS The author is associate director of the Master of Science Program in Financial Mathematics at the University of Chicago. The author thanks Jerry Dwyer for the invitation to write the paper, conference discussants Dick Berner and Nigel Jenkinson and moderator Charlie Plosser for valuable insights, conference coordinator Jess Palazzolo and the staff at the Atlanta Fed for making the presentation a genuine pleasure, and Lynn Foley for editorial assistance. He thanks Bill Sullivan, Sanjeev Karkhanis, and Joe Bonin at UBS; Mark Hurley at JP Morgan; and William Y. Chan at Credit Suisse for comments on earlier drafts. He acknowledges the support of Niels Nygaard, director of the Program in Financial Mathematics at the University of Chicago; Regenstein Library at the University of Chicago; and the Social Science Library at Yale University. This paper was presented at the Atlanta Fed’s 2007 Financial Markets Conference, “Credit Derivatives: Where’s the Risk?” held May 14–16. n the early to mid-1990s, derivatives received a great deal of negative publicity in the popular media. Several unfortunate incidents ultimately led Gastineau and Kritzman (1996), in the revised edition of their Dictionary of Financial Risk Management, to define a derivative as, “in the financial press, anything that loses money.” The proximate causes of these derivatives disasters were a variety of factors: Metalgesellschaft experienced a cash flow mismatch between long-term over-thecounter (OTC) forward contracts and marked-to-market short-term exchange-traded futures; Gibson Greeting was encouraged to enter into complex, and probably inappropriate, financial transactions that it apparently didn’t fully understand; Procter & Gamble and Robert Citron of Orange County assumed significant investment risk, exacerbated by a “surprise” interest rate hike; Barings Bank employed a rogue trader who was able to engage in fraud because of the lack of institutional risk control; and, of course, just about everything went wrong at Long-Term Capital Management (LTCM). Many of these incidents were highlighted prominently soon thereafter in books with titles such as Derivatives: The Wild Beast of Finance (Steinherr 1998). At least one market participant (an investment bank) felt that the label “derivatives” was so detrimental that it renamed its offerings “risk management products.” Many remain skeptical of the value that derivatives can provide; one hedge fund manager, speaking to a group of summer MBA interns at an investment bank in New York a couple of years ago, when asked if he used options as part of his investment strategy, replied, “I don’t go to that crack house.” The (interest rate) swap market has been around for only about twenty-five years, yet it is one of the largest and, arguably, one of the most important and successful financial markets in the world. Credit derivatives are much newer, having been first publicly introduced by the International Swaps and Derivatives Association (ISDA) in 1992 but not broadly traded until after the standardization of the documentation in 1999.1 I ECONOMIC REVIEW Fourth Quarter 2007 43 F E D E R A L R E S E R V E B A N K O F AT L A N TA What about “credit”? The origin of the word, as our classics scholars know, comes from the Latin—proximately from creditum (meaning “loan”) and ultimately stemming from credere (to entrust) and credo (I believe), which, for our purposes, is what every bank or lender does (in terms of expecting to be paid back with interest) and, more generally, what every counterparty expects (in terms of performance) when entering into an OTC derivative contract. There is nothing new about lending and borrowing, though Grant (1992) has chronicled the alleged long-term relaxation, and accompanying deterioration, of credit practices in the United States. Chacko et al. (2006) go one step further—identifying credit risk as a disease: “It makes you uneasy, queasy, almost to the point of nausea. Well, we are here to inform you that you have just been infected with the Credit Risk virus. And you won’t be cured until the money is safely returned. In the modern world, this is a virus as ordinary as the common cold” (3). Ryan and Risk (2006) refer to the “predicament” relating to credit derivatives as “akin to battling a rare disease” (at least rare thus far). Others have used expressions like “contagious” and “cancerous growths” in their descriptions of these instruments. What happens when you combine wild beasts with some ubiquitous, virulent pathogen? Avian flu? No, credit derivatives! Who wouldn’t be scared? When the topic for this session was first proposed, the distinction between macro risks and systemic risks initially struck me as quite different. I would like to address the first (on which, I believe, there has been a fair amount of both academic and practitioner research and to which I will dedicate only something of an overview) and then transition to the second set of risks (which, I believe, constitutes the actual issues relevant for the policy discussions to be subsequently addressed here). Credit Derivatives and Macro Risks When one thinks of macro risks, what come to mind are exposures to changes in those aggregate or fundamental economic factors that could affect the economy as a whole in general or the financial markets and the banking sector in particular. Before considering the macro risks that might affect the credit markets, a distinction should be drawn—one that I heard made by a credit derivative market maker a few years ago. He pointed out that, while trading credit derivatives is surely trading credit, there is a difference between trading the market’s perception of credit (as realized in corporate and some sovereign bond spreads) and trading “real” credit. By real credit he meant trading instruments that are triggered not by the possible likelihood of bankruptcy; not by changes in default probabilities, recovery rates, or credit ratings or by changes in those ratings; and not by any other circumstances that may influence the market price of credit risk in any particular name but by the actual act of filing for bankruptcy, by missing payments on borrowed money, by debt repudiation or moratorium, or by restructuring under financial duress—in other words, trading instruments that kick in when one comes to not believe in some institution’s ability and willingness to repay debt. Of course, one would like to think that there is a fairly close correlation between these two types of credits and that the marketplace would respond by providing financial capital to what is perceived to be a potentially rewarding arbitrage strategy between the two (capital structure arbitrage having been one of the faster growing of the hedge fund strategies out there). But the distinction between real credit and perceived credit is not trivial, as most commonly seen reflected in the presence (or absence) of total returns swaps for corporate securities in (from) the catalogue of credit derivatives.2 What’s in a name? Insurance or derivative? One of the fundamental reasons for the success (or, at least, the popularity) of credit derivatives is their ability to sep- 44 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA arate the hedging or acquisition of credit risk from the traditional vehicles that have allowed a position in credit (that is, bonds and loans). Credit derivatives are often likened to “financial credit insurance” (and, indeed, they have been referred to in that manner and certainly can be utilized in that way),3 even if the NAIC (National Association of Insurance Commissioners) constantly reminds derivative salespeople (and their compliance departments) that they cannot market derivatives as insurance, which is a unique product, separate from financial contracts: swaps, forwards, futures, and options.4 Obviously, investment banks that have lending relationships with corporates and sovereigns welcome the ability to lay off credit risk without the consent, or even the knowledge, of their counterparties. This lending goes to the very heart of relationship banking. Moreover, thanks to credit derivatives, these banks have embraced the relaxation of capital requirements previously imposed on the traditional lending businesses. Consideration of macro risks for credit derivatives raises three issues. The first is whether the ability to lay off credit risk has influenced the activities associated with bank lending or capital market issuance practices. The second is whether 1. See Skinner (2005). Also see Neal and Rolph (1999), who wrote, “Estimates from industry sources suggest the credit derivatives market has grown from virtually nothing in 1993” (3). A very entertaining article (Tett 2006a) gives some insights into the development of the credit derivative market. 2. Nelken (1999) notes, “There is considerable uncertainty in the market about when an instrument is a credit derivative and when it is not. One definition of a CD [credit derivative] is any contract whose economic performance is primarily linked to the credit performance of the underlying asset. This definition would technically rule out TR [total return] swaps, because their performance is only partially linked to the credit quality of the underlying and is mostly linked to the market risk of the underlying” (173). 3. Skinner (2005) says, “Credit default swaps . . . are actually default insurance” (280). Nelken (1999) notes that “a credit derivative works very much like an insurance policy. . . . The credit swap market is very similar to the insurance and reinsurance markets” (5). Goodman (2001) argues that “credit default swaps are really quite simple—they are conceptually similar to insurance policies” (144). And Anson (1999) states, “This type of swap may be properly classified as credit insurance, and the swap premium paid by the investor may be classified as an insurance premium. The dealer has literally ‘insured’ the investor against any credit losses on the referenced asset” (44). 4. In a March 16, 2007, e-mail message to me from Matti Peltonen, Chief Risk Management Specialist, New York State Insurance Department, Peltonen cites a letter, dated April 30, 2002, written by James Everett, Capital Markets Counsel, New York State Insurance Department, providing the department’s legal interpretation in response to an inquiry asking whether credit default swaps constitute insurance. Peltonen notes in his e-mail: “The New York Insurance Department (NYID) consistently finds that derivative contracts are not insurance contracts as long as the payments due under the contracts are not dependent on proving an actual loss. For example, in considering catastrophe options (cat options) that provide for payment in the event of a specified natural disaster (such as a hurricane or major storm), the NYID stated that cat options were not insurance contracts. A cat option purchaser did not need to be injured by the event or prove it had suffered a loss from the event. In reaching this conclusion, the NYID distinguished between a ‘derivative product,’ which transfers risk without regard to an actual loss, and ‘insurance,’ which only transfers the risk of a purchaser’s actual loss.” This distinction is not to be taken lightly. Risk Transfer (May 26, 2004) informs us: “If a derivative contract were found to be an insurance policy, the derivative could only be sold by a licensed insurance broker. Thus a derivative counterparty that is not so licensed—one ultimately found to have been selling an insurance policy—would be acting unlawfully. In California, this would be a misdemeanour. In Connecticut, fines, imprisonment, or both can be imposed for acting ‘as an insurance producer’ without a license. Under Delaware law, a Delaware corporation can lose its ‘charter’ to do business if it acts ‘as an insurer’ without a ‘certificate of authority’ to conduct an insurance business.” ECONOMIC REVIEW Fourth Quarter 2007 45 F E D E R A L R E S E R V E B A N K O F AT L A N TA macroeconomic factors might act as catalysts in initiating widespread credit crises and their associated implications for credit derivative markets. The third is whether the greater dispersion of credit risk in the economy among a broader class of firms, investors, and institutions is a positive and stabilizing development. Credit derivatives and lending behavior: Moral hazard? The first question asks whether lending practices have changed in light of the new credit risk management products. This question addresses the ability to lend, the willingness to lend, and possibly the degree of thoroughness contained in the process of due diligence that has typically attended most bank lending activities. We tend to use the expression “moral hazard” technically to refer to a situation in which an additional or heightened risk arises because of the presence of a contract or mitigating arrangement, which subsequently causes one of the naturally risk-averse parties involved to relax its behavior with respect to its efforts to avoid a negative underlying outcome. The prototypical example of a market instance of this phenomenon is, not surprisingly, insurance; for example, a homeowner who possesses fire insurance may reduce her actions and expenditures to keep her domicile free from circumstances that might cause inadvertent combustion. Gladwell summarized this problem nicely: Insurance can have the paradoxical effect of producing risky and wasteful behavior. Economists spend a great deal of time thinking about such moral hazard for good reason. Insurance is an attempt to make human life safer and more secure. But, if those efforts can backfire and produce riskier behavior, providing insurance becomes a much more complicated and problematic endeavor. (2005, 2) Have banks really become less cautious in their lending behavior? A number of factors make this question more of a discussion point than a well-posed question in search of a definitive answer: Recent advances in banking deregulation, the Basel Accords, modernization of financial markets, the evolving role of financial institutions, consolidation in the banking (especially the investment banking) industry, heightened competition, collapsing spreads, innovative products, and new technology all add noise to the question at hand. That said, Nout Wellink, President of Netherlands Bank and Chairman of the Basel Committee on Banking Supervision, properly pointed out in February 2007: “The role of banks as the ultimate holders of credit assets has become less important. . . . We are therefore witnessing a fundamental change in the business of banking from buy and hold strategies to so-called ‘originate-to-distribute’ models” (2007). There have been claims that the current state of credit markets has been altered by the existence and infusion of credit derivatives. More specifically, it has been posited that traditional lenders have become less concerned with the accurate credit quality assessment of their borrowers because the lenders, through the use of credit derivatives, will no longer be the ones “holding the bag” when the ultimate creditors “cease to believe.” Plender tells us, If the real worry is systemic risk, a more fundamental threat comes from the change in the structure of the banking industry whereby credit risk is packaged into tradeable IOUs or hedged via credit derivatives and shunted off bank balance sheets. Yet . . . moral hazard . . . complete with the marked decline in risk premiums and in lending standards, is the story of credit markets this decade. The mechanics of moral hazard in the exponentially growing newer financial markets entail the destruction of the old relationship between banker and borrower. This is 46 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA because banks no longer retain the credit risk in much of their lending. They originate and distribute; and where the intention is to distribute, the lender is inevitably less bothered about loan quality. (2006) With the recent events in the subprime lending market (which, I believe, have little to do with credit derivatives), one could argue that this situation may have resulted simply from a turn in the credit cycle and housing market and might attach no greater significance than that. The ability to minimize financial fluctuations and lessen price volatility is typically not included among the benefits associated with free markets. Was the unprecedented level of subprime lending a result of a change in the market’s appetite for credit risk, a reflection of the influx of ready, new investors into this area, or simply an error on the part of those who assessed the risks in this case? Those who sing the praises of free markets usually assert that, while markets are not always correct and can frequently be “wrong,” they are generally not stupid. There may be a more subtle dynamic at work in this context. Whalen reports, In the age of derivatives-enabled structured finance, the term “private equity” has become passé. Nearly every financial buyer deal we see coming to market involves a large degree of debt finance, regardless of the type of sponsor. Looking at the staggering numbers for public and private bond issuance in 2006, measured in the trillions of dollars, it seems clear to us, at least, that OTC derivatives and kindred structures like collateralized debt obligations [CDO] are driving a process whereby assets are being packaged and sold at prices that understate the true economic risk. (2007) One last thought: Knowing that insurance is available is quite different from having a policy “in hand”; it is not wise to wait until flames are coming from the roof to seek an insurance quote. The issue of liquidity will be explored later. Debt: The big picture. Currently, the United States is seriously in debt. On an aggregate level, U.S. households “owe,” on average, 122 percent of their net income. National debt is ready to top $9 trillion (and this amount does not include future Social Security and future Medicare liabilities). Corporate debt is at an all-time high; business-sector and financial-sector debt exceeds $23 trillion. Moreover, the United States is relying on significant amounts of foreign funding. By the end of the third quarter of 2006, the United States had borrowed in excess of $860 billion (around 6.5 percent of gross domestic product [GDP]) from abroad to finance its expenditures, and BusinessWeek predicts more than 6 percent GDP growth in 2007. Overall, the debt of the United States was estimated (at the end of 2006) at $48.4 trillion. The question that begs answering is whether any changes have occurred in the banking system, lending markets, regulatory framework, or institutional landscape to warrant this explosion in credit risk. The presence of credit derivatives is probably more a reflection of an attempt to manage credit risk than a manifestation of the spread of this credit disease. Macro risks and contagion. The second macro issue is the extent of the potential impact of changes in the credit cycle and the ability of the system and the market participants to handle such changes. By analogy, if the government were to put something into the water that drove the death rate to zero, the life insurance business, one would think, would become extremely stable and relatively uninteresting; no one, or at least no one who knew what was in the water, would subsequently refer to this industry as risky. Insurance companies would collect the premiums and never ECONOMIC REVIEW Fourth Quarter 2007 47 F E D E R A L R E S E R V E B A N K O F AT L A N TA need to make a payout. If there were no bankruptcies, defaults, repudiations, or need for restructuring, credit markets (and credit derivative contracts in particular) would be dull and uninteresting. In the end, it will be credit events that test these products, contracts, markets, and institutions. If credit derivatives are triggered by credit events, then, on a macro scale, we might want to consider what tends to influence the incidence of these events. Neal and Rolph tell us, Credit risk is influenced by both business cycles and firm-specific events. Credit risk typically declines during economic expansions because strong earnings keep overall default rates low. Credit risk increases during economic contractions because earnings deteriorate, making it more difficult to repay loans or make bond payments. Firm-specific credit risk is unrelated to business cycles. (1999, 5) Credit derivative modeling will be looked at in more detail later, but some credit models have incorporated aggregate economic variables as potential explanatory drivers of credit conditions. For example, Das (2005) identifies the model developed by McKinsey and Company (under Tom Wilson) as one in which macro variables play a primary role: “The model focuses on the risk of a credit portfolio explicitly linking credit default and credit migration behaviour to the macro-economic factors that are major drivers of the credit quality of the portfolio” (590). Although one might think that inclusion of these macro variables could enhance/improve credit analysis, Das informs us that “in practice, the increasingly favoured models are reduced form models” (590). There is no shortage of academic or practitioner research attempting to identify and evaluate those discernable variables that influence the number and severity of bankruptcies, defaults, and so on. While it is intuitive that economic downturns would generally coincide with the incidence of credit events, we can ask what macroeconomic factors in particular are the most significant in that context. Ed Altman (a professor of finance at New York University’s Stern School of Business and one of the foremost authorities on credit, bankruptcy, and defaults) and other academic researchers have incorporated various macro factors into their credit models and analyses and have attempted to evaluate the importance of those variables. These factors have included the level of interest rates, leverage, inflation, unemployment, aggregate measures of indebtedness, nominal and real GDP growth rates, changes in those growth rates, savings rates, liquidity premiums, the ratio of highyield debt to total debt outstanding, returns (and changes in returns) of aggregate equity indices, and, in a few cases (see Frye 2000 and Gordy 2000), a single systematic factor referred to as “the state of the economy.” The inclusion of these factors is intended to capture the drivers of the probability of default and/or the recovery rate (or, conversely, loss given default) in the event of bankruptcy/default. In some instances, these variables are examined in conjunction with a number of other firmspecific factors such as industry or sector or geography as well as more traditional credit indicators like the degree of corporate leverage, the ratio of free operating cash flow to total debt, and EBIT or EBITDA (earnings before interest, taxes, depreciation, and amortization) interest coverage multiples. Interestingly, in examining the empirical importance of macroeconomic variables that have been recognized as statistically significant in the work of others, Altman et al. (2003) find that these variables add little in terms of explanatory power or incremental statistical significance.5 48 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA The open state of this research is reflected in the current work of Professor John Binder (2006) of the University of Illinois–Chicago, who has recently found a counterintuitive positive empirical relationship between probability of default and recovery rates. Furthermore, in a recent telephone conversation I had with a senior risk manager at a large, high-profile hedge fund, the manager articulated an unsolicited belief in support of the notion that default probabilities and recovery rates (on bank debt, at any rate) should be expected to exhibit a positive relationship. To summarize: If the level of interest rates, the state of the credit cycle, the dummy variable acting as a proxy for boom or recession, or any of the macro variables included in these credit studies proved likely to announce or even trigger widespread defaults, then we might consider these macro risks Macro risks are exposures to changes in as a potential source of systemic risk. Perhaps less ambitiously, consider the those aggregate or fundamental economic heretofore generally accepted negative factors that could affect the economy in relationship between probabilities of default general or financial markets and the bankand recovery rates. If the deterioration of the economy serves as the single driving ing sector in particular. factor (raising default probabilities and reducing recovery rates), then this deterioration could potentially, on an economywide basis, trigger credit derivatives and simultaneously generate systemic risk in the banking and financial sectors. The lack of unambiguous significance in the literature of aggregate macro phenomena on credit (and credit events in particular), viewed in conjunction with standard firm-specific characteristics, tends to mitigate our immediate and urgent concern with macroeconomic risks per se as a source of systemic risk via the conduit of the credit derivatives markets. But the likelihood of a macro event as a catalyst for triggering credit derivatives certainly remains a possibility. One final aspect of the macro relationship to credit involves what Lucas and others have referred to as “policy rules” (and the associated critique of attempting to estimate relationships econometrically when the behaviors of market participants change with changes in policy regime). While this consideration may serve to challenge the weak statistical significance in the empirical studies of macro variables and credit events, what it really introduces is the notion that Federal Reserve and governmental policies (in particular, monetary and credit policies) themselves respond to the myriad economic data and financial considerations discussed at each Federal Open Market Committee meeting. While that relationship may be obvious enough, it raises the issue of whether policy action itself may trigger a series of credit events. After all, if it were not for the unexpected tightening of interest rates in 1994, the first interest rate hike since 1989, there never would have been a Procter & Gamble derivative fiasco or an Orange County bankruptcy. Before leaving this section, I would like to quote Ed Altman’s conjecture that we may be navigating in a new and heretofore unexplored world of credit. One fact that he pointed out is that the U.S. high-yield market had less then $10 billion notional outstanding in 1978, whereas currently there is over $1 trillion outstanding—exceptional growth by any standard. Furthermore, 5. Altman et al. (2003) note: “Macro variables are added in columns 7–10; we are somewhat surprised by the low contributions of these variables since there are several models that have been constructed that utilize macro-variables, apparently significantly, in explaining annual default rates” (16). They also observe: “Macro variables—as before—tend to have no evident effect on BDR (the weighted average default rate on bonds in the high yield bond market)” (19). ECONOMIC REVIEW Fourth Quarter 2007 49 F E D E R A L R E S E R V E B A N K O F AT L A N TA the (junk) market is not dominated by fallen angels, despite GM and Ford’s inclusion in 2005, but by newly issued non-investment grade securities. . . . In addition, the U.S. has seen a substantial rise in the size of the syndicated loan market. Syndicated lending has risen more than 60% in the last three years and rose to total outstandings of $1.5 trillion in 2005. The growth in this sector has been paced by more risky leveraged loans. Leveraged loans . . . are now estimated to be about $500 billion, or about one-third of the syndicated loan market in the U.S. These higher risk and return loans are increasingly being financed by non-bank institutions, such as CLO (collateralized loan obligation) hedge funds. While large banks typically arrange these highly leveraged syndicated loans, in recent years more than three-quarters of the funds have been provided by non-bank institutions. . . . As is readily apparent from examining the history of high-yield bonds, however, markets are dynamic and constantly shifting. And there are times when even the most carefully constructed and tested forecasting models can be off the mark. The last few years has been one such period. Given the unique environment in the credit markets during the last several years, which has been fueled by massive liquidity and the advent of new participants like hedge funds, it is worth asking whether historically based estimates of default probabilities and recovery rates are still relevant. (Altman 2006, 2-6) The next section provides more discussion on hedge funds and who is taking on this mushrooming credit risk. Concern with Credit Derivatives from Market Professionals Concern has been articulated from many quarters about the rapidly expanding market in credit derivatives. With nearly $35 trillion notional outstanding and annual growth rates that have ranged between 40 percent and 160 percent, credit derivatives easily qualify as one of the most quickly developing product areas within the capital markets. The explosive growth in credit derivatives in recent years (in terms of face amount outstanding, trading volume, and the sheer variety of products available) has raised questions about many facets of this phenomenon. Like any new market, credit derivatives have experienced some growing pains (and I will mention a few of the problems that have arisen), but most of the anxiety that has been voiced centers on three aspects of this market: 1. the sheer size of the notional outstanding (and, more importantly, the fact that the face amounts being traded in many names—independent of the added volume via credit indices—are integer multiples of the current notionals outstanding in that name’s debt [bonds and loans]); 2. the increasing involvement of the hedge fund community in this market; and 3. the operational backlogs and issues surrounding confirmations, clearing, and settlement. Credit derivative notional versus underlying outstanding debt. Currently, in the auto industry alone, primarily at General Motors and Ford, the notional outstanding in credit default swaps (alone) is estimated to be fourteen to eighteen times higher than the underlying bonds, notes, and loans. Gillian Tett, the capital markets editor at the Financial Times, tells us that, in the overall market, “the total size of the CDS [credit default swap] universe is now believed to be 10 times bigger than the total pool of underlying cash bonds” (2006b). 50 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA What would happen if “something went wrong”? What do we even mean by “wrong”? For some in this market, a credit event could be interpreted as “something gone right.” Nevertheless, concern about the size of this market should not be underestimated. Gerry Curtis, a distinguished investment adviser based in Boston, recently noted the following: Possibly a bigger source of risk (than the issuance of low quality debt securities) is the sale of credit default swaps to buyers who do not own the bonds that are insured. There are many bond issues outstanding in which the amount of credit default swaps is substantially greater than the amount of bonds outstanding. If the issuer defaults on the bonds, the loss to the seller of the credit default swap is many times greater than the premium received by the seller. The favorable default rate on junk bonds in 2005 and 2006 has enhanced the willingness of buyers to purchase “junk” bonds and/or sell credit default swaps. (2007) This sentiment has been echoed by many other traditional institutional asset managers who are wary of what credit derivatives might do to their portfolios and markets. Part of the concern seems to stem from a general belief that credit default swaps, which originally played a useful role in hedging default risk associated with debt issuance, have been carried to excess and are now vehicles for speculation and counterproductive. There is absolutely no scarcity of negative sentiment regarding credit derivatives, as illustrated by article titles such as “Somebody Turn On the Lights” (Mayer 1999), “Credit Derivatives Trigger Near System Meltdown” (Dodd 2005), and “[Credit] Derivatives Will Collapse the World’s Financial System” (Jeffolie 2006). By some measures, Lyndon LaRouche’s admonition that “the amount of indebtedness outstanding is greater than could ever be repaid, so the system is hopelessly bankrupt” (Gallagher 2007) appears discerning and contemplative by comparison (and is only slightly less disturbing than Gallagher identifying LaRouche as a “leading economist”). On the topic of this rhetoric, I agree with Partnoy and Skeel (2006a), who write, “Unfortunately, opinion on the credit derivatives issue is polarized between alarmists who oppose financial innovation and supporters who naïvely embrace it.” Let’s examine what has gone wrong and could go wrong as a result of the volume mismatch. Historically, the credit default swap market has been primarily a physically settled market. By that, I mean that upon exercise (following the declaration of a credit event), the buyer of credit protection would deliver acceptable debt (as in a previously agreedupon range of bonds and/or loans) or deliverables in exchange for the face value of that debt (with little in the way of variations from par). For securities that are already distressed, the typical quote would involve points up front in addition to the periodic credit default swap premium (where that premium is quoted in terms of basis points per annum for any fixed horizon—five years being the most common as well as the de facto default tenor—and typically paid quarterly). The point is that, if there are multiples of the underlying debt being traded in the credit default swap market, then it would seem obvious that physical settlement could be problematic. There is at least the potential for a bottleneck. When Argentina defaulted in January 2002, the major broker-dealers got together to net all the trades before the securities ultimately traded hands; as in the past, an orderly capital markets settlement occurred following this credit event (as opposed to the protracted cross-border legal proceedings that have accompanied sovereign defaults). For the most part, the primary credit derivative dealers are the large global investment banks; ECONOMIC REVIEW Fourth Quarter 2007 51 F E D E R A L R E S E R V E B A N K O F AT L A N TA in most of the recent industry polls (BBA [British Bankers Association], ISDA, Fitch, Risk), banks account for around 55 percent of credit derivative buying and 40 percent of credit derivative selling. These figures are not surprising because banks often act as the market makers and intermediate counterparties in this product area. That said, in the case of Argentina, it took an unusual proactive measure on the part of the banks and dealers to ensure a smooth settlement; without this coordinated action, there could have been a problem. Why not move to cash settlement? Would that not eliminate the possible squeeze scenario? There have actually been instances in which, upon the occurrence of credit events, the outstanding debt has traded up (as it needed to be acquired to be subsequently delivered). For example, when Delphi entered bankruptcy, its debt, which had been trading around 57 cents on the dollar, traded up, peaking at 71 cents before ultimately falling back to around 60 cents. Many of the academic texts suggest that cash settlement of credit derivatives has not only been possible but common. These books are wrong (or at least “more wrong” in the case of the United States, as opposed to Europe, where cash settlement is more common). Even credit index documentation (and it is in the indices that cash settlement makes the most sense), when last I looked, indicated physical settlement on the term sheets. The reason for the staunch resistance to cash settlement (where the payout would be based on the difference between face value and market value) hinges on the process for the determination of what market value really is. In the past, with other products, recourse to polling a number of other market makers and broker-dealers and then possibly averaging the quotes (midmarket or otherwise) would serve to determine the unwind cash flow. So what had been the objection to cash settlement for credit derivatives (and it had been a large one)? One Morgan, for example, may feel singularly uncomfortable if another Morgan (which may have positions in that name’s debt and/or the credit derivatives themselves) is a significant contributing factor in the broker poll. Physical settlement, then, avoids valuation disagreements and the need for market polls. This overarching concern with the notional imbalances has led to concerns along these lines: With more credit derivatives being traded than bonds available, a default by GM could spark panic buying of the company’s bonds, driving up prices. The contracts would be worthless if prices rose to 100 cents on the dollar because investors would have to pay the same amount for the bonds as they received in payouts. ‘The current method has the potential to significantly distort the economics of the trade,’ says James Batterman, an analyst at Fitch Ratings in New York. ‘There are no limits on the amount of derivatives exposure vis-à-vis deliverables.’ (Hamish Risk 2006) To be blunt, I have to question Risk’s use of the word “worthless” (or at least ask for clarification “to whom?”); I would replace his use of “investors” with “speculators”; and as for the use of the expression “the economics of the trade,” I think the economics speak for themselves. Another concern after a bankruptcy or default, not unrelated to the necessity of a broker poll or some other process for the determination of market value, is the likely loss of liquidity in the securities of the affected debtor. Thin markets tend to make people uncomfortable about taking, for example, the last traded price as a market consensus, and, following credit events, even if the debt continues to trade, it is often 52 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA accompanied by spotty markets. Some have argued that the downgrade of the autos (which was not a credit event in and of itself) was not such a tremendous shock but that the significant market impact resulted from the institutional response—as bond funds, whose prospectuses require they hold investment-grade paper, scrambled to dump Ford and GM and sought other investment opportunities. The credit derivative market has responded to the credit-derivative-notional-versusunderlying-debt mismatch and the issues related to polling by developing a process that seems to meet the needs of market participants: an auction. Going forward, with credit events, the broker-dealers (supported by Creditex and Mark-It Partners and in line with ISDA protocol) will participate in an actual auction (not just a polling) through which the investment banks will provide inside markets, market orders, limit orders, and automated electronic trades and arrive at a final settlement price. If one Morgan thought the other Morgan was too low on his valuation for the defaulted debt, the first Morgan could express his belief by buying it in that market (voting with his dollars as it were), independent of credit derivative positions. This process has already been successfully implemented for Calpine, Collins & Aikman, Dana, Delphi, Delta, Dura, and Northwest Airlines over the last couple of years, has been supported and embraced by the dealer community (contributing members include ABN Amro, Bank of America, Barclays, Bear Stearns, BNP Paribas, Citigroup, Commerzbank, Credit Suisse, Deutsche Bank, Dresdner, Goldman Sachs, HSBC, JP Morgan, Lehman, Merrill Lynch, Morgan Stanley, Royal Bank of Scotland, Société Générale, and UBS), and has recently (February 2007) been extended to electronic tradable tranche fixings for credit indices (see Markit 2007). This auction process now allows for the cash settlement of credit default swaps following a credit event (making the derivative/underlying debt imbalance something of a nonissue as well as making the invariably uncomfortable polling unnecessary) and should help allay fears about the sheer number and notional magnitude of these derivatives being traded. Many market professionals remain largely unfamiliar with the specifics of these contracts: However, for a CDS (credit default swap) contract to be valid, it needs to be backed up by some tangible bonds in the marketplace (even if far smaller in size). Usually that is not a problem, since few companies are debt free. But if corporate events occur which prompt a company to withdraw its bonds—such as a merger— this can suddenly make CDS contracts worthless. . . . For the CDS market is now so monstrously large that the behaviour of the derivatives is exerting an increasingly large impact on the cash market. The tail, as they say, is wagging the dog. (Tett 2006b) There are actually well-defined protocols for such corporate activities as mergers, acquisitions, spin-offs, and other corporate actions called succession events (which I will not go into here). I will offer one last thought on underlying mismatches before leaving this topic (as it is one of the main sources of concern regarding credit derivatives). There are a number of (very successful and important) derivative contracts that cover underlyings that themselves are relatively small, illiquid, not traded, or even nonexistent as a stand-alone asset. Dozens of instances come to mind. The Treasury bond futures contracts are on a notional 6 percent (semiannual) coupon twenty-year U.S. Treasury bond; there is no such thing (and even if, by chance, there were today, there ECONOMIC REVIEW Fourth Quarter 2007 53 F E D E R A L R E S E R V E B A N K O F AT L A N TA wouldn’t be tomorrow). What’s made this contract particularly interesting is (1) the fact that it has been, and continues to be, physically settled (giving rise to lists of eligible-for-delivery securities, conversion factors, cheapest-to-deliver instruments, embedded options, etc.) and (2) the fact that the U.S. Treasury stopped issuing bonds for a time. While the futures contracts never stopped trading (though deliverables always did remain) and while a large portion of the volume of trade has shifted to the ten-year Treasury note futures contract, there is no reason why bond futures, The reason for the staunch resistance to in principle and in practice, could not cash settlement of credit derivatives hinges trade even if there were no deliverables. on the process for the determination of CMTs (constant-maturity Treasuries) also qualify by this criteria. Eurodollar futures, what market value really is. the most actively traded futures contract in the world, are cash-settled three-month LIBOR futures (and they have their own quirks), but they are nominally on ninety-day deposits (which the Chicago Mercantile Exchange will never make or take). The S&P 500 derivatives complex (futures and options on the futures at the Merc and options on the Chicago Board Options Exchange [CBOE]) pay off based on where the underlying stocks close; we once claimed that there was no S&P 500 cash product, but exchange-traded funds (SPDRs or ticker “SPY”) have mitigated that assertion. VIX derivatives traded on the CBOE are contracts that have payoffs based on the implied volatility as determined by several option quotes. OTC variance swaps also have payoffs based on actual volatility (in this case, usually the non-detrended historical variance of returns). There is no variance (per se as an asset) that trades, but no one worries about the settlement of these contracts. Nondeliverable forwards (NDFs) on Chinese yuan or renminbi have paid off without involving the underlying currency, and the foreign exchange (FX) market, the largest market of them all, generally trades on an order of magnitude forty times larger than the volume associated with the entire global value of international trade; if excessive volume or speculation were reasons to terminate trading in a product, FX would be the first to go. Of course, with every derivative (be it a future, forward, swap, or option), for every seller, there’s a buyer, and for every buyer, there’s a seller. While I am decidedly not of the opinion that derivatives are zero-sum instruments, I understand the statement that “risk is neither created nor destroyed, just repackaged and redistributed.” Given the propagation of derivatives in general and the growth of credit derivatives in particular (and recognizing that many of these OTC trades are leveraged), there are those who think their existence adds risk to the marketplace. Risk is a two-edged sword. Whether one gets long a credit name by buying its corporate bonds or selling credit protection via a credit default swap, the major difference is funding (and therefore leverage). If this fact sounds odd, consider that, far and away, the most common equity derivative strategy is selling puts—synthetically; this overlay strategy, which involves buying (or owning) the underlying stock and writing (or selling) calls against that long stock position, is most often referred to as a buy-write or covered-call or covered-write (or over-write). Many consider this strategy to be a low-risk investment play. Most would consider naked put selling, though, to be extremely risky. The primary difference between these two strategies is basically funding. So why would someone prefer one strategy over the other? That’s a good question. By the way, in 2003 the size of the OTC credit derivative market topped the size of the entire OTC equity derivative market (Banks, Glantz, and Siegel 2007, table 1.2), and this ratio now stands at around five. 54 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Hedge funds and credit derivatives. Although hedge funds have been involved in some of the larger derivative disasters (I once heard someone on a trading floor say, “Long Term Capital Management,” to which someone else interjected, “They were neither. They didn’t last long and apparently didn’t manage their capital very well either.”), many hedge funds understand the risks of derivatives (and credit derivatives in particular) well, use them responsibly and effectively, and provide support and depth to a market dealing in risks that were once concentrated in the banking industry. Independent of the ongoing trend that continues to see flows into hedge funds, they command under 3 percent of global investable wealth (around $1.25 trillion). Although any statement that begins with the words “every hedge fund” is likely false (given the range of strategies employed by the myriad hedge funds out there today), most do indeed “hedge.” The most common hedge fund strategies continue to be equity long-short. This approach might involve, for example, going long General Motors stock and short Ford stock. While there are many ways to get market neutral, the main idea is that if the market goes up or down, you’re okay if you’re simultaneously long and short; if the auto sector goes up or down, you’re covered (because you’re long and short). This strategy bases its returns on the specific overperformance/ underperformance in the chosen pair of securities. Variants of this strategy typically do not involve very high leverage (either using borrowing to magnify one’s positions or using derivatives to command greater positions than the cash market would provide). Typically, greater leverage is employed in risk arbitrage (that is, merger or takeover strategies) and in convertible bond arbitrage (buying convertible corporate debt, hedging the equity risk by shorting the corporation’s stock, and turning the exposure into a volatility trade). The one strategy that usually involves larger degrees of leverage is fixed income arbitrage; LTCM (which was, after all, primarily a fixed income hedge fund) told its investors that it intended to lever its positions twenty to twentyfive times (that is, for every $1 they received, they were going to take on $20 to $25 of risk). That said, it’s been argued that one of the most problematic aspects of the LTCM debacle is the ease with which the firm was able to lever its positions and access financial resources from the major banks. In that regard, I think the banks have learned their lesson. Nevertheless, Alex Ineichen (2001), a world-class authority on hedge funds, has argued that, “many of LTCM’s strategies would have worked if they could have held onto their assets for some months longer” (7). Many hedge funds use credit derivatives to lay off risk. Consider one of those convertible arbitrage funds (buying convertible bonds and selling stock). If the funds want to strip out the credit risk of these bonds (which they own), they could pay so many basis points per annum to know that, worst case, they have the right to sell this debt for its face value. On the other hand, some hedge funds are engaged in more sophisticated strategies (for example, buying five-year credit protection on Ford and selling five-year credit protection on General Motors—with no intention of holding this trade for five years). Unlike buying straight corporate debt and attempting to short another corporate bond (thus tying up financial capital), doing two credit default swaps may give the hedge fund exactly the exposure it would like (with only a net capital charge or net margining on the part of its counterparty/counterparties). Chilcote (2006) reports that “hedge funds lost hundreds of millions of dollars, owing to their exposure to derivative contracts and the downgrading of General Motors’ and Ford’s debt in May” (1). One need only hear this assertion to raise the obvious question, If the hedge funds lost, then who won? Chilcote goes on to characterize “hedge funds . . . that specialize in credit-default swaps” as “secretive.” Louis Moore Bacon is one of the grand old men of the hedge ECONOMIC REVIEW Fourth Quarter 2007 55 F E D E R A L R E S E R V E B A N K O F AT L A N TA fund industry (and credited with an extremely impressive track record at Moore Capital). Bacon, at a Hedge Fund Symposium in London in 2000, identified what he called the five warning signs for hedge funds: (1) size (getting too big and exhausting the available investment opportunities within one’s area of expertise—and beyond some point morphing from being one of the hunter-gatherers to “becoming the game”), (2) leverage (taking on too much risk), (3) transparency (in tremendously understated fashion, Ineichen [2001] tells us, “Full transparency of current positions is commercially unwise.”), (4) funding (asset and liability mismatches), and (5) hubris (what Lowenstein [2000, 89] has identified as potentially the most dangerous “Greek” of all). Perhaps the greatest detriment to hedge funds today is their association with LTCM (where all five of the above factors came into play in a significant and negative way). At any rate, many hedge funds are understandably reluctant to disclose their positions. Not only is this their stock in trade (that is, their security selection process, hedging techniques, valuation models, portfolio construction methods), but hedge funds know that a market participant with deeper pockets could trade against them. This scenario is not just the creation of the paranoia of a few hedge fund managers; it is probably far more likely to occur than one would think. Take the case of Amaranth Advisors LLC (a large hedge fund that was based in Greenwich, Connecticut). Amaranth apparently got into trouble in the fall of 2006 with losing positions in energy derivatives, though it did utilize what it referred to as a multistrategy approach and traded convertible bonds as well as other instruments. Amaranth’s typical leverage ranged between 6 and 8. The Wall Street Journal reported the following (after Amaranth’s $6 billion loss): Hedge funds are among Wall Street’s biggest customers, and the Street gives them red carpet treatment as the fees roll in. But the Amaranth case shows how Wall Street dealt with a fund after it had traded its way into a deep hole. Information the fund revealed about its holdings as it grasped for a lifeline let other commoditymarket players, Wall Street firms included, exploit its positions. As they drove prices relentlessly against Amaranth, its losses swelled, and instead of facing a big but possibly survivable setback, it collapsed. (Davis, Zuckerman, and Sender 2007; also see Stoyeck 2007) There were disturbingly similar allegations in the case of LTCM. If someone were to claim that hedge funds constitute a major source of systemic risk, the natural place to start looking for it would be with the investment banks. None of the investment banks or securities houses, to my knowledge, have complained about the fact that around half of all trades on the New York Stock Exchange are done by hedge funds. Furthermore, don’t hold your breath—many of the larger investment banks are generating 15 percent, 20 percent, 25 percent or more of their revenues from hedge funds. This revenue is not surprising because many hedge funds trade very actively and opportunistically. In principle, the investment banks, as prime brokers, clearing agents, and flow trading counterparties, should be in an excellent position to properly assess a hedge fund’s credit risk and charge/margin for market exigencies, but there is at least the potential for a perceived conflict of interest. Moreover, Thomas F. Huertas, director of the Financial Services Authority (FSA) in London, has shared his further concerns, as far as margining goes, with Risk magazine regarding the issues of rehypothecation, cross-margining, and the geographic and legal access to capital. 56 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Blaming hedge funds in general for market disasters is like blaming well-fed vultures for dead animals; while the two are often seen together, it doesn’t mean that one is the cause of the other. Contrary to what seems to be the norm, Alan Greenspan (2005) has praised the ability of hedge funds to make the financial markets more efficient, to bring some contrarian balance in times of overly enthusiastic exuberance, and to provide needed liquidity to markets, especially in turbulent market scenarios. In situations in which hedge funds have gotten into trouble, we should ultimately look for the real source of the problem (which may have been nothing more sinister than a bad investment or a strategy gone awry). Although it was felt at the time that LTCM required a Fed-orchestrated bailout for the good of the financial system as a whole, subsequent hedge funds have gone away with little in the way of concern that the banking system or financial markets (national or global) might be at risk or in peril. Furthermore, LTCM was atypical (particularly at the time) in its size; it was far and away the largest hedge fund at that time (based on assets under management). It’s seldom pointed out that LTCM returned financial capital to investors as investment/ trading opportunities in the market waned. With regard to the exceptional events surrounding LTCM, I’d like to quote one authority: The primary mechanism for regulating excessive leverage and other aspects of risk-taking in a market economy is the discipline provided by creditors, counterparties, and investors. In the LTCM episode, unfortunately, market discipline broke down. LTCM received generous terms from the banks and broker-dealers that provided credit and served as counterparties, even though LTCM took exceptional risks. Investors, perhaps awed by the reputations of LTCM’s principals, did not ask sufficiently tough questions about the risks that were being taken to generate the high returns. Together with the admittedly extraordinary market conditions of August 1998, these risk-management lapses were an important source of the LTCM crisis. (Bernanke 2006) (One can only wonder whether LTCM would be around today if they had utilized credit derivatives as part of their arbitrage strategy.) The demise of Amaranth is an excellent counterpoint. There was no furor in the financial press (at least, not until the role of the investment banks started to become better understood); there was no talk of a government-sponsored bail-out; and the possibility of collateral damage or systemic risk never even seemed to have been mentioned. Moreover, no one blamed derivatives for this implosion. Amaranth was a hedge fund (but at least part of its portfolio was assumed by another large hedge fund). There were certainly losses, but no former employees appeared on television lamenting the loss of their retirement savings. Maybe we’re getting it right. Or, at any rate, bashing hedge funds just because they are hedge funds seems to be losing popularity. Having worked at UBS, I believe I have some insight into the investment banks’ point of view. Alarm bells would surely be tolling if a bank knew that every hedge fund had on exactly the same trade(s); this sort of concentration of risk (gone wrong) may have repercussions for hedge funds’ banking counterparties—even if the klumpenrisk (the individual net exposure of the broker-dealer to a particular entity) is nominally managed to be small. In essence, if every hedge fund were doing the same thing, although booked as separate institutional relationships, it would be nothing more than a multiple-counterparty LTCM scenario. The job of credit risk control for hedge funds has to be one of the more challenging roles at an investment bank today. ECONOMIC REVIEW Fourth Quarter 2007 57 F E D E R A L R E S E R V E B A N K O F AT L A N TA This systemic danger (of hedge funds taking on similar positions) has not gone unnoticed. The European Central Bank (ECB) has warned, The increasingly similar positioning of individual hedge funds within broad hedge fund investment strategies is another major risk for financial stability which warrants close monitoring despite the essential lack of any possible remedies. This risk is further magnified by evidence that broad hedge fund investment strategies have also become increasingly correlated, thereby further increasing the potential adverse effects of disorderly exits from crowded trades. (2006, 142)6 The influx of financial capital into hedge funds, in conjunction with the concentration of trading strategies in this universe, probably explains the recent less-than-stellar industry performance. To say that hedge funds have been under tremendous, continual, ongoing scrutiny would be an understatement. The question is whether (and how) regulatory intermediation would help. Greenspan spoke at an IMF Conference in Beijing in June 2005 on hedge funds; Risk magazine reported as follows: Greenspan said (beyond his belief that some market participants were taking on “risks for which their compensation is inadequate,” that the hedge fund industry had expanded too quickly and should temporarily shrink, and that CDO returns were destined to be disappointing in the near term) he was not particularly concerned that this may have a negative impact on financial stability, as long as banks and other lenders are managing their credit risks effectively. (2005, 10) In other words, for those who qualify as eligible investors in hedge funds, laissez faire, and as for the investment banks that are the ultimate risk watchdogs, watch your credit risk! As Juvenal asks, though, “Quis custodiet ipsos custodes?” (Who will guard the guardians?) I’d like to make one last point about hedge funds and credit derivatives. Philippe Jorion (2005), a recognized authority on risk management (both market risk and credit risk), reports an interesting (and possibly surprising) fact about the use of credit derivatives by hedge finds (based on a 2003 BBA survey): “Hedge funds and securities firms . . . are fairly balanced, each with about 16% of protection buyers and sellers” (546). This statement makes you wonder where the credit risk is going, then, doesn’t it? Operational risks. When I first entered the financial world, it was with a proprietary option trading firm based in Chicago known as O’Connor and Associates. At the time, much of its trading took place on exchange floors (in Chicago and around the world). O’Connor was recognized as being among the best at what it did (and what it claimed to understand, better than anyone else, was risk management). For an O’Connor trader, there was one ultimate cardinal sin—not knowing your position. It is this unpardonable offense, for the world of credit derivatives generally, that has led to well-warranted criticism and ill-informed hysteria. Greenspan, over the years, has been among the staunchest defenders of derivatives, claiming that they reallocate risk into the hands of those who are best capable to take on, to warehouse, and to dynamically manage those risks. In 1999 Greenspan said, “By far the most significant event in finance during the past decade has been the extraordinary development and expansion of financial derivatives. . . . These instruments enhance the ability to differentiate risk and allocate it to those investors 58 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA most able and willing to take it . . . a process that has undoubtedly improved national productivity growth and standards of living” (1999). There are also those at the other end of the spectrum. Indeed, there have been some interesting articles comparing and contrasting the thoughts and beliefs of Alan Greenspan and Warren Buffett on this topic since both have been outspoken on the uses and value of these instruments (see Weinberg 2003). For those who have not been following these discussions, Buffett has labeled derivatives “financial weapons of mass destruction.” It’s been said that much of what Buffett has claimed is disingenuous because he has used derivatives himself, but he does make an important point to be revisited later. It is interesting, then, to hear of not only a criticism of derivatives from Greenspan but to hear of a Federal Reserve intervention (back in September 15, 2005) ordering a group of credit derivatives dealers “to get their act together” on the heels of a revelation that significant unprocessed credit derivative trades were outstanding—the cardinal sin. How can one manage risk if one doesn’t know what the risks are? And how can one know what the risks are if one doesn’t know what one’s positions are? Timothy Geithner, president of the Federal Reserve Bank of New York, last year touched on (and reiterated) this potential problem: These concerns . . . suggest the need for greater caution by financial institutions in several important areas. . . . It is very important that the major dealers make the investments necessary to improve the operational infrastructure that underpins the credit derivatives and broader OTC derivatives markets. Operational risk and infrastructure failures have played a prominent role in past financial crises, and the infrastructure weaknesses that have characterized the credit derivatives markets since their inception are an ongoing source of concern. (2006, 3) Since the September 15, 2005, castigations (which reflected concerns originally articulated in June 1999 on the heels of the LTCM disaster),7 the industry has worked diligently to reduce those trade backlogs and expedite the processing, confirmation, and settlement of credit derivatives. Originally, the fourteen banks agreed, among other things, to cut the number of unsigned trades by 70 percent before July 2006. Not only was that goal exceeded,8 but in 2005 the larger credit derivative traders reduced the average confirmation lag from twenty-three days to sixteen days. The FSA (the U.K. financial regulatory authority) in their Financial Risk Outlook 2006 wrote, Credit derivatives provide a valuable mechanism through which financial market participants can manage their credit risk, bringing together those who wish to reduce credit exposures with those who are prepared to increase them. The market 6. This sentiment was also contained in the International Monetary Fund’s annual report (2005), which suggested that there might be a meltdown in the credit derivatives market if all the investors were to “run for the exit at the same time.” 7. These concerns were outlined in a document known as “Improving Counterparty Risk Management Practice,” put out by the Counterparty Risk Management Policy Group under the direction of Gerald Corrigan and Stephen G. Thieke and then updated and reissued, again by Corrigan, in July 2005 with the title “Toward Greater Financial Stability—A Private Sector Perspective,” addressing current topics of concern. 8. “Credit derivative dealers have reduced a backlog in processing trades by more than 80%, more than their target, an industry trade association said” (Credit derivative banks 2006). ECONOMIC REVIEW Fourth Quarter 2007 59 F E D E R A L R E S E R V E B A N K O F AT L A N TA has continued to grow at a rapid pace and firms such as hedge funds have become increasingly important, as both buyers and sellers of these instruments. Operational and legal risks may arise if the market is unable to keep up with this growth. Without confirmation that a trade has taken place, parties to the transaction are exposed to legal and financial uncertainty. If a credit event occurs while a credit-derivative transaction remains unconfirmed, doubt as to its legal validity and contractual responsibilities could prevent the transaction from being executed. This uncertainty could create liquidity problems and act as an accelerant in a financial crisis. (2006, 15) Similarly, Platt (2006) tells us, “The rapid growth in global credit derivatives is putting stress on settlement systems and operational controls, despite significant progress in clearing a big backlog of unconfirmed trades.” Although improvements are reassuring, this concern over the recent rapid growth points to the possibility of a catalyst for a systemwide breakdown. It is neither the instruments themselves nor the fact that hedge funds are increasingly involved in credit derivatives that constitutes the greatest concern. Operational risk is certainly a well-founded consideration on its own. Some initiatives have been proposed that could act to mitigate some of the operational risks. For one thing, the European exchange Eurex started trading futures on the iTraxx index at the end of March 2007. These futures behave like the credit default swaps that trade over the counter but with the exchange counterparty support (reducing counterparty risk), with much more transparent pricing, and with the associated daily mark-to-market margining. The Chicago Mercantile Exchange has also reported its intention to list credit event futures contracts (originally targeted for first quarter 2007, revised to a June 17, 2007, start date) and, as usual, the CBOE is close behind. As is not uncommon, given the rivalry between the OTC market and the exchanges, the banks initially declined to participate in trading these exchange-listed contracts, one head of credit trading in London calling the contracts flawed. So What Are the Risks? Early problems. Mention was made of glitches in the development of credit derivative products and markets. There are some classic errors that were made early on in this market’s history (many of which have achieved almost folklore status). In one instance, namely Anderson, credit protection was sold and bought on a company that turned out to be a parent/holding corporation that did not have any outstanding debt. In essence, there were no deliverables. This sort of slip has been addressed, among other safeguards, by the creation of the REDs (Reference Entity Database Service), which is intended to eliminate any ambiguity regarding the precise legal names of counterparties (that may have similarly labeled affiliates or possibly unrelated but close-sounding names) and which links a particular name to a specific debt issue. For the purposes of, say, a credit default swap, this service ensures that the credit being traded is properly identified. For the record, what one is buying/selling protection on, in terms of the institution and the level of debt (for example, senior unsecured) referenced, may differ from the deliverables in the case of a credit event—obviously an issue, but hardly a source of systemic risk. A facet of this market that is often not discussed is the bilateral nature of a credit default swap. Reference has been made to these contracts as options—specifically put options on debt. A credit default swap (CDS), though, is a swap (not an option) 60 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA and can be triggered by either the buyer of protection (which is what most of us think of) or the seller of protection. One might ask why a protection seller would trigger a CDS when that action would result in his/her receiving defaulted debt that is trading, say, 70 cents on the dollar. The answer is that there are a number of reasons why a protection seller may wish to do this. Perhaps the most important (particularly in the early days before the standardization of the documentation, or “standard docs”) involved the need for a market maker who had, say, sold protection for 45 basis points and then purchased it, from another counterparty, for 42 basis points (if this trade is not assigned) to ensure that there was not a substantive difference between the debt that was being received and the debt that was required for delivery (which, at some point, involves a notice of physical settlement—a letter indicating the specifics of the debt that is going to be proffered). If the market were to continue with physical settlement, this procedure could pose a systematic concern, but, in light of the new auction process (and accompanying cash settlement), is not a cause for a systemic breakdown. Conseco. Because of its landmark nature, let me very briefly review the Conseco case. In September 2000, Conseco (an insurer, lender, and financial services company based in Indiana) found itself in financial difficulty. It had acquired a home lender known as Green Tree Financial, which made mobile home loans; unlike most home equity loans, where the value of the home tends to rise, this often is not the case with mobile homes. Conseco, therefore, experienced an urgent need for financing and, through its bankers (Bank of America and Chase Manhattan), was able to renegotiate its debt. Officially, this renegotiation constituted a credit event (under the category of restructuring). It was alleged that at least one of the banks, having bought credit protection on Conseco, subsequently triggered credit default swaps. Moreover, since there was not a bankruptcy (in which the majority of debt might have traded pari passu) because Conseco was still a viable business, there remained a credit term structure. Longer-dated Conseco debt was trading 68 cents on the dollar (whereas the extended fifteen-month loan was trading 92 cents on the dollar). Those who had bought protection and chose to exercise obviously delivered the longer-dated debt. The subsequent clarification of exactly what constitutes restructuring, who may “call” the credit event, and what may be delivered (leading to the definition of modified restructuring, or “Mod R,” which requires the deliverable to be like the protection traded) reflects some of the growing pains associated with the credit derivative marketplace. Credit risk models. It has long been known that credit risk is not an easy nut to crack (read, “concept to model”). Anyone who has ever traded options, for example, knows that the Black-Scholes formula does not precisely fit the real world, but for European-style options on non-dividend paying stocks, it works fairly well. There are a great many quotes along the following lines: “Models are to be used, but not to be believed,” “All models are wrong” and “I’ll take a good trader over a great model any day.” Nevertheless, models have their uses. Why is modeling credit risk so difficult? As far back as 1999 (the year of the publication of the first comprehensive, standardized ISDA credit derivative documentation), Phelan and Alexander concisely summarized what they perceived to be the primary impediments to developing a framework for evaluating credit risk: Credit risk is more difficult to model than market risk for several reasons. First, the lack of a liquid market makes it difficult—or impossible—to price credit risk for a specific obligor and tenor. Second, true default probabilities in the market ECONOMIC REVIEW Fourth Quarter 2007 61 F E D E R A L R E S E R V E B A N K O F AT L A N TA cannot be observed. Users must determine these probabilities by either inferring default rates based on observed historical experience of the public credit ratings, using a model such as KMV’s Credit Monitor, or determining the default rate through a subjective credit approval process. Third, default correlations are quite difficult to observe or measure, making it hard to aggregate credit risk. And fourth, to calculate the equity/capital cushion, it is necessary to estimate the tail risk probabilities of asymmetric, fat tailed loss distributions. (1999) There are some substantive difficulties listed here. Jorion (2005) tells us in his Financial Risk Managers Handbook, “Risk management starts with the pricing of assets” (3). Theoretical valuation is the key to understanding and managing risk. It is no surprise that when one looks at aggregate market data on credit derivatives broken down by market participant (whether from BBA, Risk, ISDA, or Fitch), it appears that banks/dealers account for about half of the buying and half of the selling. In short, the market makers are probably acting as market makers. When I was at UBS, it drew a distinction between what it called “flow business” and “structured business”—the former being primarily a market-making operation or market conduit and the latter generating trades that would likely not be backed-to-back (even if they were ultimately hedged using more standardized credit products). If a market maker is running a matched book, the removal of one link in the settlement chain should not be particularly problematic (and certainly shouldn’t bring down a systemic cataclysm). For that reason, as long as the dealers properly manage their credit risk (in the sense of counterparty collateralization and risk capital), one would think a credit event— even on a big name like GM—wouldn’t start a meltdown. One hedge fund trader once lamented that one of the large banks kept asking him to share his positions (which he was adamant he would not do); the trader indicated that he was more than willing to incur capital charges and margining based on his direct exposures to that bank but that he would not disclose his portfolio to that institution; as mentioned earlier, this sort of reluctance on the part of most hedge funds to share their positions is understandable. In one of the less cogent articles suggesting that credit derivatives might result in the systemic breakdown of the financial markets, Chilcote (2006) noted that “Long-Term Capital management did not disclose its risk or positions to investors or counterparties” (1). This observation runs counter to the criticisms of former LTCM principals who claim, in their search to find a source of financial support to allow them to weather the Russian debt crisis, that they were taken advantage of by the larger banks once their positions had become known. Back to valuation, why is credit risk so difficult? When we consider credit events, we are talking about low-likelihood, tail probabilities (which are often recognized as particularly unstable). James (1999) told us, “there is no robust way of finding the fair value of a credit derivative” (1). Partnoy and Skeel (2006b) go one step further; with regard to credit derivatives, they write, “The mathematical precision of the models is illusory” and “If the mathematical models have serious limitations, how could they support a multi-trillion dollar industry?” One of the most difficult modeling issues involves the portfolio risk management of credit risk. At issue is the determination of the correlation between defaults, a consideration particularly important in tranched products (such as synthetic collateralized debt obligations and first-to-default structured products). One hears credit risk modelers talk about such things as fat tails, copulas (elliptical, Archimedean, extreme value, Clayton’s, Frank’s, Gumbel’s), and conditional correlation coefficients. Unfortunately, our intuition, in the context of 62 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA credit derivatives, often fails. For example, financial theory typically advocates the benefits of portfolio diversification, but, as those familiar with first-to-default products understand, a diversified credit pool is not a good thing for the investor.9 The autos: What went wrong? The automobile manufacturer downgrade in 2005 surprised some investors, who lost money and, at first, did not understand why. Let’s look at this event in a bit more detail. On May 5, 2005, Standard and Poor’s, the credit rating agency, downgraded Ford one notch to BB+ and GM two notches to BB (these moves signifying a change in their debt ratings from investment grade to subinvestment grade or “junk”). Although S&P had given anticipatory hints Blaming hedge funds in general for market that these downgrades were possibly to be expected, the market response was immedisasters is like blaming well-fed vultures diate and chaotic. for dead animals; while the two are often By the date of the downgrade, many seen together, it doesn’t mean that one is the hedge funds were engaged in capital structure arbitrage on the automakers. This arbicause of the other. trage often involved trading debt versus equity or one level of debt (such as senior unsecured) versus another level of debt (such as junior subordinated). One trade that many hedge funds used was a long bond and short stock position; the idea was that if the automakers did poorly, the ability to recoup something on the debt was relatively high (given that some recovery value on the bonds was to be expected), whereas if the company went under, the equity would likely be worthless. Many hedge funds, instead of buying the bonds, effectively got long the auto credit risk (synthetically) by selling protection through credit default swaps and then short sold the stock (or effectively got short by purchasing equity put options). What could go wrong? After the downgrade, the price of protection skyrocketed. On their credit trade, hedge funds had a mark-to-market loss. One would think the equity (being subordinate to the debt in case of distress or impending bankruptcy) would lead the bonds, and this presumption is reasonable. But, at that time, in a very public way, an opportunistic corporate raider named Kirk Kerkorian suggested that he might want to acquire a large block (28 million shares) of GM stock at $31 per share (a 13 percent premium over the previous closing price). GM equity took off (causing those who were short the stock—or long the equity puts—to experience a mark-to-market loss); for those using stock as a hedge for their credit derivatives, they lost on their hedge too. There were other trades that hedge funds had on that also blew up. We have not gotten into the details of some of the more structured index credit derivatives, but some hedge funds traded tranches of portfolios of credit risk. In selling protection on the equity tranche (which isn’t equity at all, but debt—though, like equity, this tranche experiences the first losses) and hedging by buying protection on the mezzanine tranche (which takes the next hit) of various structured credit derivatives, essentially these investors were entering into a correlation trade. In the case of the automakers in May 2005, what we saw was a case of correlation gone wrong. The impact on Ford and GM was huge, but other corporate spreads were essentially unchanged. The price of protection on the equity tranche tripled, rising from 16 percent to 50 percent, whereas the mezzanine tranche was unaffected. 9. Any reader interested in the underlying mathematics is directed, as a first step, to Malevergne and Sornette (2006). ECONOMIC REVIEW Fourth Quarter 2007 63 F E D E R A L R E S E R V E B A N K O F AT L A N TA Liquidity. One of Warren Buffett’s favorite similes goes something like this: Derivatives are like hell—easy to enter February 23, 2005 May 15, 2007 and almost impossible to exit. Although this characterization may be GMAC 225–235 GMAC 166–170 glib, it touches a nerve. Liquidity does conGM 285–295 GM 430–435 stitute a systemic concern. There are difFord 230–240 Ford 544–549 ferent definitions and, therefore, measures IBM 12–22 Boeing 8–12 of liquidity, but in this context, it refers to HPQ 20–30 Dow 31–34 the ability to trade continuously in markets Lehman 22–27 Citi 9–11 made up of several competing dealers with reasonable two-sided bid-offer spreads Note: Spreads are shown in basis points. offering conventional trading volumes. There is grave concern that if a number of names simultaneously experienced credit events, the system would grind to a halt. On this count, although one usually thinks of the large investment banks as the market makers, hedge funds may actually prove helpful. Dodd (2006) tells us, “The OTC market in credit derivatives is often cited as a case in point where hedge funds play a critical role in market liquidity. Indeed it is likely the case that market depth and bid-ask spreads are improved by the participation of hedge funds.” One of the simpler (and arguably wanting) measures of liquidity is the bid-ask spread. Incorporated into that basis point differential (quoted annually for standard five-year protection) is a reflection of the number of market participants, the evolving transparency and convergence of valuation, the willingness of the dealers to broker credit risk, the reduction of market maker edge, and the general competitiveness of this area. The table shows a few representative examples of how this bid-ask spread has changed over the past couple of years. Based on these market quotes, the bid-ask spread has tightened (reflecting competition, market participation, and market liquidity) from around 10 basis points in 2005 to a range of 2 to 5 basis points in 2007. Perhaps of greater concern regarding liquidity, though, would be the response of the broker-dealer community—either in their unilateral response to simultaneous largescale credit events or in the treatment of their counterparties under such a scenario. Of course, two-sided trade flow is the lifeblood of a market maker; without that flow, the best a trader can hope for is to dynamically manage his risks as they accumulate—and this point brings us back to hedging, modeling, and valuation. Small number of dealers. In 2004, 81 percent of credit derivatives bought and 75 percent of credit derivatives sold were accounted for by only fifteen large banks. When the New York Fed summoned the credit derivatives dealers on September 15, 2005 (to admonish them for their operational shortcomings in credit derivatives), only fourteen institutions were present. The most recent Fitch Global Credit Derivatives Survey (September 21, 2006) reports that the top ten institutions make up 66 percent of the volume in credit derivatives. Even that figure may be misleading since the majority of the volume in credit derivative trading is done by four counterparties: JP Morgan, Morgan Stanley, Deutschebank, and Goldman Sachs. While the small number of dealers involves industrial organization, it bemoans the fact that a large percentage of the volume of credit derivative trading is concentrated in the hands of a relatively few dealers. Greenspan admits, “One development that gives me and others some pause is the decline in the number of major derivatives dealers and its potential implications for market liquidity and for concentration of counterparty credit risk” (2003). Table Tightening of the Bid-Ask Spread over Time 64 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA There is no doubt that an unexpected departure of any one of these primary dealers would have very negative repercussions on the credit derivatives market. But, not to be dismissive, in light of the fact that credit derivatives account for only about 7 percent of the OTC derivatives volume and cognizant that all these banks are major players in most of the OTC derivatives markets, a number of market participants would probably have more to worry about than credit derivatives. Legal risk. At the end of the day, credit derivatives are almost exclusively unique bilateral OTC contracts and, as such, are only as good as the contractual documentation the attorneys draft. This fact explains the preponderance of lawyers on and around credit derivative desks. In the world of derivatives, some still remember a trade (an interest rate swap contract) entered into by the local U.K. municipal authorities of Hammersmith & Fulham; when interest rates moved against the municipal government’s position, they sought (and eventually obtained from the House of Lords) a formal judgment ordering the nullification of the transaction as illegal. In the world of credit derivatives, a number of issues have ended up in court. Most recently, Bear Stearns bought protection from Aon on a Philippine corporate backed by a government agency. Aon then bought protection from Société Générale on the Republic of the Philippines. When the Philippine agency withdrew its backing of the Philippine corporate, this withdrawal constituted a credit event on the Bear Stearns-Aon CDS but did not trigger the Aon-Soc Gen CDS. The 2nd U.S. Circuit Court upheld the content of the respective contracts, lending support to the process and providing additional confidence in the use of standard documentation. An incredibly disturbing statistic highlights the importance of maintaining proper legal documentation: A September 2004 Fitch report indicated that in some 14 percent of credit derivative claims, there have been subsequent legal proceedings. “In some instances, the disputes have involved assertions that one of the parties breached fiduciary duties owed to its counterparty, the risks associated with the transaction were not adequately disclosed, or the transaction was not suitable for the counterparty.” Caveat vendor! (Seller beware!) Insider trading. Allegations of insider trading have even been made in the credit derivatives market from both the practitioner community (Joint Market Practices Forum 2003; Credit default swaps 2006) and the academic community (Acharya and Johnson 2005). Insider trading rules are well defined for “securities,” but OTC credit derivatives formally fall outside their purview. What this consideration really speaks to is the potential for material nonpublic information flow within the larger broker-dealers. And, really, after consideration of all these risks, what is the worst that could happen? Stephen Ross tells us, As a general rule, regulatory and legislative activity follows any period of financial tragedy, and, however well intentioned, its statutes are often structured in some haste and as much in response to the drama of the events as to the logic. Not unexpectedly, they usually take the form of prohibiting certain activities that were held up by the media as grotesque examples of abuse, and rarely do they take account of the reality that the cure might be worse than the disease. . . . What is remarkable is not the failures, but rather how exceptional they are and how well the market system seems to work overall. (2000) ECONOMIC REVIEW Fourth Quarter 2007 65 F E D E R A L R E S E R V E B A N K O F AT L A N TA Conclusion The mission of the Federal Reserve System falls into four categories (my emphasis added): • • • • conducting the nation’s monetary policy by influencing the monetary and credit conditions in the economy in pursuit of maximum employment, stable prices, and moderate long-term interest rates; supervising and regulating banking institutions to ensure the safety and soundness of the nation’s banking and financial system and to protect the credit rights of consumers; maintaining the stability of the financial system and containing systemic risk that may arise in financial markets; providing financial services to depository institutions, the U.S. government, and foreign official institutions, including playing a major role in operating the nation’s payment system. (BOGFRS 2005, 1) There is no doubt that credit derivatives affect at least three of these duties in a significant way. The probability of systemic risk in the banking industry stemming from macroeconomic events related to credit derivatives is probably much lower than in the past because of the dissemination of default risk among a broader investor base. This claim may not be true, though, of the insurance industry (“insurance companies account for only 1% of protection buyers versus 20% of protection sellers” [Jorion 2005, 523]). For the financial system as a whole—recognizing that hedge funds, on balance, supply and demand comparable magnitudes of credit derivatives to and from the market—hedge funds would appear to provide a buffer for traditional lending institutions. One caveat is the potential for concentration risk if hedge funds all end up taking on the same (losing) positions. The distribution of risk has its downside, though, in terms of control. Some may recall the days when the Fed targeted the money supply. Because banks were so clever at creating money substitutes (regardless of the various definitions of money: M1, M2, M3b), eventually the Fed simply gave up attempting to control or target the monetary aggregates. One wonders whether there is an analogue at work with the control of credit risk (through credit derivatives). The impressive growth of the marketplace for credit derivatives speaks for itself. Recent developments in the settlement procedure, reductions in operational risks, and other advances to improve the clearing, transparency, and liquidity of the market bode well for the continued success of these products. Nevertheless, potential concerns still remain: These include moral hazard associated with the due diligence responsibilities of those involved in the debt origination process; the relatively small number of large broker-dealers; potential conflicts of interest (given the role of the banks as hedge fund counterparties in conjunction with their traditional role in a lending/credit function); the ability to manage credit/counterparty risk; the challenges associated with modeling and hedging the more complex of the credit derivatives; liquidity; and, as always, legal risks. Considering catastrophic or systemic risks brings to mind a quote from Donald Rumsfeld: As we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don’t know we don’t know. (U.S. Department of Defense 2002) 66 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Here is one last conjecture for a potential source of systemic credit risk. Donald Perry wrote: A hot topic of debate has been the financial shock that precipitated the stock market crash of 1929. Following that decline there was increasing unemployment, business bankruptcy, bank failure and deflation. Similar conditions are being mirrored today. . . . We have the largest debt bubble in history at a time when there is growing business failure and unemployment. As continuing growth in debt and the derivatives market weaken the US and world financial condition, some have wondered what future shock could precipitate a massive economic collapse similar to 1929? . . . There is speculation that the shock may come from the combination of Iraq war costs, increasing terrorism, tax cuts, unemployment, weakening pricing power, and travel aversion. Whatever the merits to such speculations, a more sobering shock appears to be on its way. I am referring to Severe Acute Respiratory Syndrome (SARS), an infectious disease that originated in southern China around November of last year and can result in a type of fatal pneumonia. . . . Although I am not a financial expert, I wonder if this disease may become the needle or ‘shock’ that could pop the debt and derivative bubbles. (2003) Will avian flu, after all, be the final straw? REFERENCES Acharya, Viral V., and Timothy C. Johnson. 2005. Insider trading in credit derivatives. London Business School. Social Science Research Network Working Paper, May. Binder, John J. 2006. The expected recovery rate and the probability of default on high yield debt. College of Business Administration, University of Illinois–Chicago Working Paper, December. Altman, Edward I. 2006. Are historically based default and recovery models in the high yield and distressed debt markets still revelant in today’s credit environment? New York University, Salomon Center, Stern School of Business Special Report, October. Board of Governors of the Federal System (BOGFRS). 2005. The Federal Reserve System: Purposes and functions. Washington, D.C. Altman, Edward I., Brooks Brady, Andrea Resti, and Andrea Sironi. 2003. The link between default and recovery rates: Theory, empirical evidence, and implications. New York University, Stern School of Business, Finance Working Paper FIN-03-006, March. Anson, Mark J.P. 1999. Credit derivatives. New Hope, Pa.: Frank J. Fabozzi Associates. Bacon, Louis Moore. 2000. Can institutions afford to ignore hedge funds? Keynote speech, 2000 Hedge Fund Symposium, London, April 27. Banks, Erik, Morton Glantz, and Paul Siegel. 2007. Credit derivatives: Techniques to manage credit risk for financial professionals. New York: McGraw-Hill. Bernanke, Ben S. 2006. Hedge funds and systemic risk. Speech at the Federal Reserve Bank of Atlanta’s 2006 Financial Markets Conference, Sea Island, Georgia, May 16. <www.federalreserve.gov/boarddocs/speeches/ 2006/200605162/default.htm>. Chacko, George, Anders Sjoman, Hideto Motohashi, and Vincent Dessain. 2006. Credit derivatives: A primer on credit risk, modeling, and instruments. Upper Saddle River, N.J.: Wharton/Pearson Education. Chilcote, Edward. 2006. Credit derivatives and financial fragility. The Levy Economics Institute of Bard College, Policy Note 2006/1, January. Credit default swaps may incite regulators. 2006. Bloomberg, October 10. Credit derivative banks cut trade backlog by 80%, adding $17 trillion in risk. 2006. False Flag News, July 26. Curtis, Gerry. 2007. Views from the sidelines #55, January 29. Das, Satyajit. 2005. Credit derivatives: CDOs and structured credit products, 3d ed. Hoboken, N.J.: John Wiley & Sons. Davis, Ann, Gregory Zuckerman, and Henny Sender. 2007. Amid Amaranth’s crisis, other players profited. Wall Street Journal, January 30. ECONOMIC REVIEW Fourth Quarter 2007 67 F E D E R A L R E S E R V E B A N K O F AT L A N TA Dodd, Randall. 2005. Credit derivatives trigger near system meltdown, Special Policy Brief 26. Derivatives Study Center, Financial Policy Forum, August 5. ———. 2006. On the hedge fund question: A program of reckless complacency, Special Policy Brief 29. Derivatives Study Center, Financial Policy Forum, October 30. European Central Bank. 2006. Financial Stability Review (June). Financial Services Authority. 2006. Financial Risk Outlook 2006. Frye, John. 2000. Collateral damage. Risk (April): 91–94. Gallagher, Paul. 2007. The global financial system is burning at both ends. Global Research (March 18). Gastineau, Gary L., and Mark P. Kritzman. 1996. The dictionary of financial risk management, revised ed. Summit, N.J.: Frank J. Fabozzi Associates. Geithner, Timothy F. 2006. Implications of growth in credit derivatives for financial stability. Remarks at the Third Credit Risk Conference—New York University Stern School of Business, May 16. Gladwell, Malcolm. 2005. The moral-hazard myth. The New Yorker (August 29): 2. Goodman, Laurie S. 2001. Synthetic CDOs. In Investing in collateralized debt obligations, edited by Frank J. Fabozzi and Laurie S. Goodman. Hoboken, N.J.: John Wiley & Sons. Gordy, Michael B. 2000. Credit VaR models and risk bucket capital rules: A reconciliation. Board of Governors of the Federal Reserve System working paper. Grant, James. 1992. Money of the mind: Borrowing and lending in America from the Civil War to Michael Milken. New York: Farrar Straus Giroux. Greenspan, Alan. 1999. Financial derivatives. Remarks to the Futures Industry Association Meeting, Boca Raton, Florida, March 19. ———. 2003. Remarks on corporate governance at the 2003 Conference on Bank Structure and Competition, Federal Reserve Bank of Chicago, May 8. ———. 2005. Risk transfer and financial stability. Remarks at the 41st Annual Conference on Bank Structure, Federal Reserve Bank of Chicago, May 5. Jeffolie. 2006. Derivatives will collapse the world’s financial system. iTulip.com, March 17. Joint Market Practices Forum. 2003. Principles and recommendations regarding the handling of material nonpublic information by credit market participants. October. Jorion, Philippe. 2005. Financial Risk Manager Handbook, 3d ed. Hoboken, N.J.: Wiley–Finance. Lowenstein, Roger. 2000. When genius failed: The rise and fall of Long-Term Capital Management. New York: Random House. Malevergne, Yannick, and Didier Sornette. 2006. Extreme financial risks: From dependence to risk management. The Netherlands: Springer. Markit. 2007. Creditex, Markit successfully complete first electronic tradeable tranche fixings. Press release, February 5. <http://markit.com/marketing/press_releases. php?date=05Feb2007>. Mayer, Martin. 1999. Somebody turn on the lights. Derivatives Strategy (November). Neal, Robert S., and Douglas S. Rolph. 1999. An introduction to credit derivatives. In Handbook of credit derivatives, edited by Jack Clark Francis, Joyce A. Frost, and J. Gregg Whittaker. New York: McGraw-Hill. Nelken, Israel. 1999. Implementing credit derivatives: Strategies and techniques for using credit derivatives in risk management. New York: McGraw-Hill. Partnoy, Frank, and David Skeel. 2006a. Credit derivatives play a dangerous game. Financial Times (July 17). ———. 2006b. The promises and perils of credit derivatives. University of San Diego School of Law Legal Studies Research Paper No. 07-74. Perry, Donald. 2003. A needle to pop the debt and derivatives bubble. May 5. Phelan, Ken, and Colin Alexander. 1999. Different strokes. Risk Credit Risk Supplement (October). Platt, Gordon. 2006. Growth in credit derivatives putting stress on controls, despite progress in clearing trade backlog. Global Finance (November). Plender, John. 2006. The credit business is more perilous than ever. Financial Times.com (October 13). Risk. 2005. Greenspan warns of investor backlash. (July): 10. Ineichen, Alexander. 2001. The myth of hedge funds. Journal of Global Financial Markets (Winter 2001): 34–46. Risk, Hamish. 2006. GM bankruptcy risk exposes default swap imbalance (Update 1). Bloomberg.com (March 14). International Monetary Fund. 2005. Annual report. April 6. Ross, Stephen A. 2000. Forensic finance. Presentation at Risk 2000, Boston, June 13–14. James, Jessica. 1999. Credit derivatives: How much should they cost? Risk (October). Ryan, Jennifer, and Hamish Risk. 2006. Credit default swap market whipsawed by Verison (Update 1). Bloomberg.com (August 8). 68 ECONOMIC REVIEW Fourth Quarter 2007 F E D E R A L R E S E R V E B A N K O F AT L A N TA Skinner, Frank. 2005. Pricing and hedging interest and credit risk sensitive instruments. Burlington, Mass.: Elsevier Butterworth-Heinemann. Steinherr, Alfred. 1998. Derivatives: The wild beast of finance. Hoboken, N.J.: John Wiley & Sons. Stoyeck, Richard. 2007. Amaranth hedge fund collapse—friendly banker becomes PREDATOR. ezinearticles.com (February 3). U.S. Department of Defense. 2002. DoD News Briefing— Secretary Rumsfeld and Gen. Myers, February 12. <www.defenselink.mil/transcripts/transcript.aspx? transcriptid=2636>. Weinberg, Ari. 2003. The great derivatives smackdown. Forbes.com (May 9). Tett, Gillian. 2006a. The dream machine: Invention of credit derivatives. Financial Times (March 24). Wellink, Nout. 2007. Risk management and financial stability—Basel II and beyond. Delivered at the GARP Eighth Annual Risk Management Convention and Exhibition, New York City, February 27. <www.bis.org/ review/r070228a.pdf>. ———. 2006b. Lessons for CDS investors. Financial Times.com (November 2). Whalen, Christopher. 2007. Credit derivatives, moral hazard and systemic risk. Seeking Alpha (January 30). ECONOMIC REVIEW Fourth Quarter 2007 69
@FedFRASER