A big step forward in the development of the credit-derivatives market came last month when, after more than a year of preparatory work, the New York-based International Swaps & Derivatives Association (Isda) released standardized documentation for credit-default swaps. But, the market as a whole still lacks universal standards. “The issue of credit derivatives means many different things to many different people,” says Ron Tanemura, head of structured debt at Deutsche Morgan Grenfell in London. “There is one end which is highly structured each transaction may take weeks. At the other end, the default-swap market is becoming increasingly more liquid. There are now nearly half a dozen brokers actively covering the default-swap market.”
While it is impossible to put a precise size on the whole market, unofficial estimates indicate that worldwide the total notional value of all credit derivatives exceeds $100 billion, half of which is written out of London.
So far, a great deal of the market has been characterized by numerous one-off specially tailored trades. However, a number of credit derivatives have been fairly actively traded. There is a parallel market in credit derivatives for all the Latin American Brady bonds and Eurobonds, particularly those issued in Argentina and Brazil. In the US there is at least a nascent parallel market in credit derivatives for corporate borrowers that have at least a few hundred million in total outstanding debt.
Even though some assets could provide a high level of credit-derivative activity, much of the market does not trade on the minute-by-minute basis common for developed derivatives. “You cannot use the word liquidity when describing the credit-derivatives market,” says Philip Borg, head of credit derivatives at Bankers Trust in New York. “If you think that trading a corporate bank loan will take a few weeks, and trading a distressed loan may take a few months, a credit-protection trade falls somewhere between the two in terms of the amount of time it will take to document and complete the transaction.”
Yet at the other extreme the market for default swaps linked to certain emerging debt, such as bonds issued by the Korean Development Bank, is more liquid than the underlying cash market. “There is no physical delivery for these derivatives,” Tanemura says. “Through the default-swaps market you can short the bonds, which is difficult in the cash market. The swaps are traded over the wire – it does not take two weeks to close a transaction.”
The basic structure of credit derivatives isn’t new. It has long existed in the form of loan insurance – although the word insurance is assiduously avoided for fear of US regulatory problems. The typical buyer of a credit derivative is a commercial bank that has made a loan to a corporate client. A great deal of commercial lending has long since become a dead-end business done primarily as a type of sophisticated client-relations exercise. Holding a $10 million commercial loan, for instance, demands a far greater capital reserve requirement than holding a $10 million bond from the same corporate. Furthermore, whatever thin profits are earned from the loan will probably attract a higher rate of taxation than profits earned from a bond.
Client relations
Without access to credit derivatives, a commercial bank would eventually reach a ceiling beyond which it could no longer lend. Although this might not have much adverse effect on immediate profits, it would definitely hamper the client-relations aspect of lending and so might affect future business. Since credit derivatives effectively insure against a loan default, the risk attached to the loan is eliminated. By buying a credit derivative on a given loan, for accounting and risk management purposes the bank can consider it a guaranteed asset – cash in hand, so to speak. The corporate client, though, need never know that the loan has been effectively written off the bank’s balance sheet. And the fact that the loan principal has been guaranteed means that the bank will be happy to extend further loans, thus maintaining the client relationship.
“Banks do a lot of lending for relationship purposes in order to develop more profitable work for their clients,” says Borg. “They have to maintain a lot of credit lines and even extend them from time to time. Otherwise, they wouldn’t get the next mandate their corporate client had to offer. Thus the bank would be more than willing to lose the spread over Libor on a loan, even a bit more than that, in order to buy a credit derivative and be able to write the loan off the balance sheet.”
Credit derivatives written on Latin Bradys have all begun to assume a basic form, but the majority are individually tailored and structured contracts. “Because there has never been anything near standardized documentation, much less standardized terms and conditions, each credit derivative has to be individually and painstakingly negotiated. Thus, there are a lot of transaction costs involved each time a new contract is written,” says Mark Rodrigues, managing director of American Management Systems, a Virginia-based management consultancy that specializes in advising banks on managing financial risk.
“When writing a new credit-derivative contract,” says Borg, “it is as though the seller of the derivative is loaning money – every trade requires due diligence and negotiation of the terms of the agreement. I have so far never seen a standard vanilla trade. There are cases where the transaction works very easily. For example, if the two dealers involved have worked together on many trades before and are already very familiar with the terms they want, then a transaction can be completed relatively quickly. But even these types of trades can still take several weeks to complete.”
The exact terms of each credit derivative will vary from case to case, although with the new Isda standardized documentation there should be a strict set of definitions. The 19-page document “Confirmation of OTC Credit Swap Transactions” is intended to spell out the exact protection provided if certain types of default occur. Isda defines eight different types of “credit events”: bankruptcy, credit event upon merger, cross acceleration, cross default, downgrade, failure to pay, repudiation, and restructuring.
“One of the main issues to be addressed is standardization,” says Borg. “As an example, there can be as many as 12 different definitions as to what exactly constitutes a default. For an efficient market to develop, these type of discrepancies from one derivative to another cannot remain.”
Resale value
The most common type of credit-derivative contract is where the seller agrees to pay the difference between the full face value and the current resale value of a particular bank loan. For instance, when the loan has not actually defaulted but the borrower has been acquired by a less creditworthy corporate, payment would be based on the fall in resale value caused by the increased risk attached to the new acquiring corporate. An example of this would occur if a double-A rated corporate with an outstanding loan were acquired by a double-B corporate. The old loan would carry a greater risk after the acquisition. Should the loan be sold to another bank, it would have a reduced resale value owing to the increased risk. The reduced value is exactly what the credit derivative pays out.
In a total-return swap, which is not yet documented by Isda, the buyer of the derivative, regardless of whether the borrower defaults or not, will assume all economic benefits of the loan and receive absolutely all proceeds arising from it, such as the principal and interest repayments. The original lender effectively sells the entire loan yet, as in the case of credit swaps, still maintains the valuable client relation with the borrower. It is expected that Isda will produce total-return swap documentation by the end of the year.
“Isda is definitely moving in the right direction, but documentation is only one step in developing the market,” says Rodrigues. “What the documentation will do is set a standard but there are still other parts of the problem. The difficult issues are accessing accurate data and using a common benchmark model for pricing risk. There are a lot of numbers that need to be accurately measured – the price of the underlying asset, the credit risk, the probability of default and the degree of liquidity.”
In terms of defining a standard trigger point when the derivative becomes exerciseable, a common one is a “credit event arising from an acquisition” – a fairly cut and dried signal to detect where the double-A rated borrower is acquired by a double-B corporate.
In the case where the credit rating of the borrower is downgraded from single-A to triple-B as part and parcel of a regular business evaluation by Moody’s or Standard & Poor’s, it is unlikely a credit derivative would be triggered unless this event was specifically added to the documentation.
The definition of the trigger point of a credit derivative is usually the non-payment of a financial obligation when it falls due. Strictly speaking, the derivative will be exercised when the underlying loan is in technical default – that is, the day after a scheduled repayment is missed.
Although a credit derivative may be nearly the economic equivalent of a contract for loan insurance, there are several key differences. In terms of documentation, even before Isda released its new documentation last month, market participants all used the Isda master agreement as the starting point to enable trading between counterparties to begin. The banks and individual dealers most active in interest-rate swaps have thus been able to expand easily into credit derivatives.
Credit derivatives are almost always cash-settled. On reaching the trigger point, the seller of the derivative pays the buyer a cash sum. The seller is most unlikely to meet or deal with the defaulting party covered by the credit derivative.
By contrast, if an insurance company has provided loan insurance, it will most certainly want to take an active part in debt rescheduling and insolvency talks in order to salvage as much of the principal as possible. Hence another advantage of using credit derivatives. The commercial bank that first made the loan faces the delicate task of negotiating payment from a financially weak borrower with which it wants to maintain and develop a long-term relationship. The goodwill that will come from a generous debt rescheduling should more than compensate for short-term losses. An insurance company, by contrast, unconcerned with the long-term bank/borrower relationship, might play hardball and force the borrower into liquidation.
Rob Reoch, head of credit derivatives at Bank America in San Francisco, says: “The seller of the derivative who pays out upon default may never be seen: they stay in the background during debt negotiations. This can be a very sensitive issue, especially where there is a debt work-out. One concern is that a bank that has been a long-standing lender will be joined in negotiations by an insurance company that will perhaps not agree upon a rescheduling that the banks want. The insurance company may want to get its money back quickly. The bank, on the other hand may have longer-term considerations.”
Another benefit of credit derivatives compared with loan insurance is time value. Depending on how the documentation is structured, the derivative could pay out to the commercial bank within a week of the borrower missing a payment. With the cash in hand almost immediately, the commercial banker can afford to be far more generous to his client than if he had to go through the tedious process of claiming from an insurance company and conducting insolvency negotiations before being paid.
Culture clash
The client-relations side of commercial banking has undoubtedly been a major driving factor behind credit derivatives. Commercial bankers, however, are also in the perfect position to analyze and price loan risk, thus gaining information on how much they should pay for a credit derivative. So there is a tension in the business between developing client relations and the sales and trading orientation of the derivatives industry.
“There is a very definite cultural clash with credit derivatives,” says Rodrigues. “On one side of the business, commercial bankers would never want to publicly tell their corporate clients that the loan which was so generously made has now been written off by purchasing a credit derivative. On the other side of the business, a trader at an investment bank couldn’t care less about buying and selling loan risk and passing it from one party to another. The real gain for commercial banks will come from institutions benefiting from the synergy created between these two different types of people coming from different cultural environments.”
Since the buyer of a credit derivative is laying off his entire risk, this is likely to cost him all the profits associated with holding the debt. For example, a five-year emerging market bond trading at a spread of US treasuries plus 400 basis points would cost 400bp to insure. The seller, who will receive 400bp for assuming the risks of the bond, could, however, simply buy the bond itself and hold a far more liquid instrument than entering into a relatively less liquid credit derivative and dealing with a 19-page confirmation document on top. So, why would anyone want to buy or sell a credit derivative?
For one thing, if, for example, an investor is holding a 20-year Brazilian Brady and wants only to assume risk for the last 16 years of the bond, he can buy a credit derivative that covers the section of the yield curve he wants to play. The Brady holder would then lose a four-year spread – the derivative will cost him this much – but receive the bigger 20-year spread, which is far higher.
On the other side the seller of a credit derivative may only want to take a four-year view on a particular emerging market sovereign risk but there might be no four-year instruments in the market. So selling a four-year credit derivative on a 20-year bond provides the exposure sought.
The execution of this transaction can probably best be explained using the concept of an insurance policy. The Brady holder wants to take a long-term view on Brazil, but he does not want to be exposed to the risk of default for the first four years. Thus he buys a four-year insurance policy for the Brady. Should there be a bond default during the four-year period, the insurer will compensate for the loss. This amount could be defined as the difference between the specified purchase price of the Brady as opposed to the market price at 5pm New York time one week after the first payment date is missed.
The disadvantaged borrower
The seller of the insurance contract, by contrast, wants to take exposure to Brazilian Brady risk for four years and expects to receive a payment for doing so. He sells the insurance to the Brady holder and receives his payment. Should no default occur during the four-year period he can happily pocket the cash. Should the bond default, he will pay the difference between the specified purchase price and the market price one week later.
Another reason for entering into a credit-derivative contract rather than dealing directly in the cash market occurs in the case of what is called a disadvantaged borrower. An example is an offshore hedge fund that must pay a high rate of interest, for example, US treasuries plus 200bp, for funding the purchase and holding of securities. For the disadvantaged borrower, the net cost of buying a Brady trading at a spread of 400bp, after the cost of carry is deducted, leaves profit of only 200bp. However, an investor holding the Brady could offer to buy from the hedge fund a credit derivative for 300bp. The hedge fund that wants to take a position in the Brady, by selling the credit derivative, will receive a benefit of 300bp – 100bp more than what it would have to pay to buy the bond for cash. The Brady holder that buys the credit derivative pays 300bp yet will receive a guaranteed 400bp return, thus making a net profit of 100bp.
Pricing the trade often has less to do with elaborate models and more to do with simply matching buyers and sellers. According to Borg: “Unlike other types of derivatives, where there is a lot more technical analysis of risk and cashflows, credit derivatives are much more based on asking a simple question: how much money does the seller want to receive in order to assume that risk? Rather than mathematical modelling, the decision at what price to sell is more based on conducting due diligence and making a lending decision.”
Other traders use the same type of analytical tools to price credit derivatives as they use for interest-rate derivatives. “The question how do you price a credit derivative should be rephrased as how do you price between relative benchmarks – such as calculating the nine-year yield on debt where there are only five- and 10-year issues in the market,” Tanemura says. “The techniques we use in pricing interest-rate risk are the same type of techniques that we use to do relative-value pricing in credit derivatives.”
A double-A credit rating is by no means a requirement to deal in credit derivatives. As with interest rate swaps, one party, for example a single-A New York bank, would be willing to conduct trades with a counterparty even if it had a much lower credit rating, such as a double-B rated Australian bank. The New York bank will address its counterparty’s lower credit in one of two ways. First, the deal could be collateralized by the Australian bank depositing securities with the New York bank which would provide back-up should the Australian bank default. Secondly, the New York bank would look to diversify its risk. For a Latin American Brady trade, an Australian bank would make an ideal counterparty since the correlation between Latin Bradys and the Australian banking sector is extremely low.
By comparison, a New York bank would not buy a credit derivative on a Korean sovereign bond from a Korean bank unless the deal was 100% collateralized, because of the high correlation. Through collateral and diversification, it is possible to hedge even double-A credit risk by buying a derivative from a double-B counterparty.
Since every bank has a unique portfolio, there are no hard-and-fast rules about which can and cannot sell credit derivatives. “In looking at potential counterparties, there is a delicate balance between cost of funding, collateral, credit ratings and correlation. It is a complex calculation where you can’t draw easy conclusions as to which is a good counterparty,” says Reoch.
He continues: “If we have little exposure to a particular industry sector, then buying a credit derivative from a client in that sector would be a good idea since it would diversify our portfolio. But this may not be the same for a different bank.”
Through diversification, counterparties in credit derivatives can easily be scattered worldwide, with each pair having little in common. At the height of the Cold War, for example, there were several small east European banks among investors behind the insurance company that covered the Challenger space shuttle and were forced to pay up after it exploded. A Venezuelan bank could possibly act as the guarantor for debt assets held by the highly risk averse Swiss postal workers pension fund.
Hedging the spread risk
The sticky issue of hedging – which when done inaccurately has on several occasions led to heavy losses for derivatives units – may be less problematic with credit derivatives than with other types of structured products. The seller of a credit derivative often has a dual role – as a broker profiting only from a bid-ask spread and as a leveraged investor seeking to take a position with as little cash down as possible. If a German bank writes a credit derivative on a dollar-denominated Russian Eurobond maturing in 2007, for example, it can hedge with a back-to-back trade – that is, buying the same derivative from another party. Here, the only profit for the German bank would be simply the difference between the bid and ask prices. The only risk assumed would be one of non-delivery by the counterparties, which is the standard risk in the brokerage business.
At the same time, the German bank might want to take a proprietary view of the Russian Eurobond and invest in it directly. Selling the credit derivative and holding the proceeds provides the same economic benefit as buying and holding the bond itself. In almost every credit-derivative transaction, there is a combination of some of the risk being hedged by simply conducting back-to-back trades and some of the risk being held as a type of investment that does not require a direct purchase of the instrument in the cash market.
However, some analysts argue that a credit-derivative book can be hedged in exactly the same way as other types of derivatives. “There may be outright proprietary positioning as well as ‘riskless’ broker transactions from time to time,” says Tanemura. “But typically you can take on basis risk. You could decide that it’s good relative value to take a long exposure to a three-year section of the Korean default curve, for example, but, at the same time, you might be trying to short the five-year sector.”
While there still needs to be a vast development in volume before the credit-derivative market can be described as universally liquid, a huge variety of trades have already been successfully concluded. “The market today has seen the full spectrum of credit derivatives. Most things that can be described as credit risk have probably been considered for a credit-derivative structure,” Reoch adds. “You will see the risk from the most illiquid emerging market infrastructure projects hedged through a credit derivative. At the other end of the scale, credit derivatives are actively traded on many liquid instruments.”
From one extreme of the market, derivatives based on corporate loans, to the other, based on liquid bonds, there is a huge leap of liquidity. “Derivatives linked to a portfolio of corporate loans have no liquidity – this is an entirely arranged market where it will take weeks to complete a transaction,” Tanemura says. “But for any issuer with a lot of paper in the market, like the sovereigns, trades are easily done over the wire.”
In terms of the development of the market, while pricing the risk attached to bonds may be as easy as switching on a Reuters terminal, pricing loan risk is still extremely difficult. “There is still a long way to go to standardize and quantify the risks of commercial loans,” Rodrigues concludes. “Not only are there different legal issues from country to country, but there are a wide range of different parties involved – from the commercial bank relationship officer to the quantitative analyst and the sales force. All this makes for an inefficient market.”
Reduction of transaction costs as volume increases has become the tried and trusted tale of the structured products industry. The swaps market for G7 currencies, where margins are now so narrow that banks often lose money on back-to-back trades, is the perfect example. “We will see the credit derivatives market developing more and more like the asset-swaps market, which is now extremely standardized and liquid. The two are economically equivalent,” Borg says.
