Freeing capital: swaps versus CLOs
“There is a common misconception that banks are using credit derivatives because they are worried about the borrower; in fact it’s completely the opposite. Most banks would rather liquidate the cash position than buy a default swap in that situation.” Hermann Watzinger, vice president, credit derivatives at Citibank in London, is keen to dispel the image of the credit derivatives market as a way for banks to dump the toxic legacy of poor lending decisions onto unsuspecting institutional investors. “The most important use will be in loan portfolio management and we already use credit derivatives very extensively as an end-user in our own portfolio. We have done deals from $10 million to $1.1 billion.”
Citibank, in common with other commercial banks, says that it would not lay off risks on a credit that it thought might default. Instead it would liquidate the position or hand the exposure to a specialist work-out team within the bank. Another banker agrees: “We don’t want to sell a default swap and then collect on the default. That would be bad for our clients and no one wants to test that in court. And we wouldn’t buy default protection on risky places either we use emerging-markets trades to manage country limits.”
Most credit-derivative transactions are driven by capital considerations and have their roots in the beginnings of the credit revolution. This began quietly in two activities of the large commercial and investment banking groups. In the lending divisions, losses caused by recession or by external crises forced banks to provision – to mark their loans to market. The economic cycle dictated the timing. Thus in such places as the US, UK, Italy, France and Scandinavia this happened in the late 1980s. Elsewhere, for example in Switzerland, it is still happening. Such large after-the-fact provisioning affected banks in a number of ways that are now driving the development of new means to manage credit risk.
It forced the banks to recognize that they should find better ways to measure and manage credit risk. It damaged earnings to such an extent that it made banks search for a provisioning policy correlated with the predicted losses from default, to avoid earnings volatility. And it gave them a reason to sell loans which in turn made illiquidity in the credit markets an issue. Better provisioning methods – one promise of the new credit risk management models – are also being forced on institutions by more transparent accounting standards. Transparent accounting reduces banks’ ability to use general loan-loss provisions to build hidden reserves and so distort reported earnings figures.
“For banks to look at credit risk afresh they have to realize they are uncomfortable with some of the risks in their portfolio. That will force them to assess them to find out what risks they are comfortable with. And that will force them to manage them. When they realize they have too much risk they will sell it and then, to defray the cost of selling, they will buy. That is what will kick-start the credit derivatives markets,” says Paul Hattori, head of global credit derivatives at Dresdner Kleinwort Benson in London.
Credit risk was also at the forefront of people’s minds in the derivatives group. As banks’ derivatives books grew, traders began to ask credit officers how they calculated the capital charges they set against trades. Realizing that this charge significantly affected the performance of their businesses, and so their bonuses, derivatives chiefs set their quantitative analysts the task of putting credit risk evaluation on a scientific footing. Using the methods they were already adopting to analyze market risk, these quants challenged the credit departments to justify their capital charges.
They also constructed the first credit derivatives – default swaps – and embedded them in bonds to create credit-linked notes. The buyers of these notes assumed the default risk of specific counterparties, thus freeing the sellers’ swap lines and allowing the derivatives departments to originate new business.
These trends coincided with renewed criticism of the Bank for International Settlements’ risk-weighted bank capital requirements, spurred by increasing use of the risk-adjusted return on capital (Raroc) measure both in analyzing performance and in pricing credit. Raroc models measure the amount of economic capital needed to support business lines and help banks more accurately correlate risks and returns in those businesses.
Credit derivatives have benefited from these developments because they are an extremely efficient way to release regulatory capital. Banks use default swaps and total-return swaps to lay off exposures that do not meet their return criteria and use the freed capital to invest in transactions that do. The buyers typically are looking for leveraged or off-balance-sheet exposure, require synthetic assets or are unable to buy the assets they require in the cash market because their funding costs are too high. Credit derivatives are not yet being used much to manage credit risk systematically, though next year bankers expect a boom in the use of these instruments to diversify and reduce concentration risks.
Default swaps are also often structured with maturities different from that of the underlying exposure. This creates synthetic investments that are not available in the cash markets. This can make the swap more desirable than a cash investment and drive down the price.
Hedgers also have to take into account the correlation between the swap counterparty and the hedged exposure. If there is a positive correlation, then default by the underlying borrower may imply that the swap counterparty’s ability to deliver the notional value of the reference security is impaired.
“If there is a high correlation between the provider of protection and the entity on which the protection is written, then there can be a problem,” says Andrew Austin, head of credit derivatives at Nomura International in London. “If Bank A gives me protection against the default of Bank B and both banks are in the same country, I run the risk that whatever event has made Bank B default is also highly likely to have an effect on Bank A which may then be unable to fulfil its side of the swap.”
Credit default swaps can also be structured on a basket of exposures. Trades in excess of $1.1 billion have been done on baskets of between 30 and 50 credits. These are essentially the same as a privately placed unfunded collateralized loan obligation (CLO). “The investor picks the loans from our portfolio and assumes the risk they want from us,” explains one swapper. As in almost all credit derivatives transactions, the lending bank remains the lender of record, ensuring that no damage is done to the bank’s relationship with the borrower.
According to the commercial banks, the main use for these instruments is to manage regulatory capital positions or concentration risk and the underlying loans are high quality. At the moment, when bankers talk about bank loan-portfolio management they mean regulatory capital management. Here credit derivatives offer an instant arbitrage. Corporate loans are weighted at 100% drawn or 50% undrawn. Buy the same exposure via a default swap from an OECD bank and regulatory capital usage falls to 20% and 10% respectively.
Demand for assets via the default swap continues to grow rapidly. Phil Borg, managing director and head of credit derivatives at Bankers Trust in New York, sees default swaps as a straightforward alternative to asset swaps. “When spreads were tight, credit protection trades offered slightly better returns with only a small amount of extra risk – mostly illiquidity risk. So the asset swappers looked at them as a useful alternative.”
Overcapitalized banks and those with low return-on-capital hurdles, particularly continental European institutions, are entering default swaps as risk-takers – taking assets from the balance sheets of banks such as Citibank, SBC Warburg Dillon Read, Chase and JP Morgan and freeing them up to enter new, more profitable trades. These banks, either via their loan departments or through their asset-swap groups, are using default swaps as a loan or bond substitute.
“Very few banks in Europe use credit derivatives for managing their own balance sheets,” says one default swapper. “This will be a booming area in 1998 but they are just not ready yet. However, they are ready to take risk from us via credit derivatives and to use that as a loan or bond substitute. That gets them comfortable with the idea of credit derivatives at a time when they are struggling to identify their risks and quantify them. Then when they are ready, they will use credit derivatives to manage risk. They are just not spending money on protection yet.” Often these are institutions which want access to US loans and which have no presence in the US.
Watzinger, who has talked to may European institutions believes that “the biggest opportunity will be to attract commercial banks into actively managing investment grade portfolios using credit derivatives. Their motivation will be to increase the borrowing capacity of their corporate clients, to reduce portfolio concentrations and to optimize return on regulatory capital.”
Another large source of demand for the risk-taking side of the transaction is from Japanese and other Asian institutions whose funding costs have soared as perceptions of their credit quality have fallen. A highly rated institution with no constraints on balance-sheet size will be able to fund at significantly sub-Libor rates and will be able to buy high-quality assets yielding, say, Libor+20 basis points, in the cash market and generate significant positive carry. A bank funding at Libor+30bp will not be able to buy those assets in the cash market because of the negative carry.
Credit-default swaps allow this type of institution to break out of the cycle of credit weakening that can result from being lesser rated: the cost of funding forces lesser-rated banks to seek higher-yielding assets which can weaken their portfolios and worsen their credit. Because the swap is unfunded – there may be a collateral requirement but it will be much less than the notional amount of the swap – the bank may be willing to assume the exposure for just Libor+15bp. This gives the bank a high-quality asset netting Libor+15bp and the hedging institution cheap cover.
Since unfunded transactions are by definition leveraged, this kind of trade is also popular with hedge funds and the kinds of Prime funds and insurance companies which buy secondary market loans. Demand in that market has grown so strong that the average time to complete a transaction has come down in the last three years from two months to two weeks.
There are also pockets of demand for default swaps as true hedges. Corporations are becoming uncomfortable with the credit risks to which they are exposed. Trade receivables books can contain large exposures to individual buyers of their goods and services. Long-term purchase contracts are a particular source of this type of credit risk. And big multinationals are heavily exposed to the banking sector through long-term trade financings, through loan facilities such as revolving credits where they are exposed to the possibility that a bank default will leave them without funding when they need it, and through swap and option portfolios. They also want to hedge against emerging-market risk where guarantees are not available or export agency cover is expensive. They compare the coupon on a default swap to the cost of other instruments that perform the same economic function, such as letters of credit or guarantees.
A cheap deal
The insurance and reinsurance markets are also potentially large buyers of protection. In some sectors, for example in retailing/wholesaling and construction, corporate demand for insurance is so great that there is no more re-insurance capacity. Not only does this force the corporate insurance buyer to investigate default swaps, it also makes credit derivatives an interesting source of reinsurance for both the insurance and reinsurance industries. By using a default swap to go short the default risk of a basket of, say, construction company bonds, they free their lines in that sector. Again, banks seeking assets in that sector will provide demand for the construction-sector risk and both sides can come away with a cheap deal because of differences in how banks and insurance companies measure their own performance and because of the differences in tax and regulatory capital treatments of the banking and insurance industries.
Robert Scanlon, managing director, credit risk, at SBC Warburg Dillon Read in London also believes that banks’ ability to lay off risk with the insurance industry is crucial. “One of the problems with using credit derivatives to hedge credit risk is that you tend to find yourself dealing with banks and corporates to whom you are already exposed. Better to deal with people who can price off return and not capital and also who have the big benefit of not being big counterparties already.”
The arrangers of bond repackagings are also big users of default swap protection. Suppose, for example, an investment bank can buy Brazilian Brady bonds cheaply in the US markets but cannot sell them into Europe because of their unusual floating coupon and amortization schedules. It places them into a special-purpose vehicle (SPV) and swaps them into fixed-rate Deutschmark bullet bonds. The repackaged securities yield 150bp more than existing Deutschmark-denominated Brazilian bonds and are snapped up by German insurance companies.
Triggers
The risky underlying bonds are in a trust and have been swapped using a cross-currency swap. Most of the structures have triggers that force early redemption of the new bonds and the unwind of the SPV in the event that collateralization levels fall below specified thresholds. So if credit spreads rise, or, in the extreme, the bonds default, that swap will have to be unwound as the structure is dismantled. If the dollar strengthens against the Deutschmark, the bank will lose money unwinding the cross-currency swap. In an extreme case the residual value of the bonds in the trust may not be enough to cover the shortfall, so there is a credit-related contingent foreign-exchange risk. To cover this risk, the arranger can take out default protection on the underlying securities.
Similarly, banks run a quanto credit risk when crossing currencies in asset swaps. For example: a bank takes a dollar-denominated credit and swaps it into yen using a currency swap; if the underlying asset defaults and the swap is unwound there will be a windfall gain or loss on the swap. Bankers recall deals when the dollar/yen traded at 250 that, if they defaulted at today’s exchange rate of 125, would incur windfall losses as large as the bond principal. The way to hedge that is to take credit exposure in dollars and buy credit protection in yen. But the size of this hedge varies with each movement of the exchange rate and must be dynamically managed in the same way as the risks embedded in other quanto instruments.
Commercial banks ought to dominate the default-swap market because of their huge loan portfolios, but the investment banks believe that their culture prevents them from taking advantage. Says John Chrystal, CSFP managing director: “I’ve been focused on credit derivatives since 1991. I thought then that commercial banks would dominate. However they face two hurdles which often hinder business development. First, they are reluctant innovators in the credit world. The most common cause of a bank failure is excess credit losses, and if a firm has a credit approach which is working, it is going to be reluctant to change. Second, commercial banks traditionally separate risk decisions from pricing, rather than considering them as a pair. Many banks have two internal prices for a credit line – free and infinite. The infinite price is hit when the line is full and no trading occurs after that. Investment banks are used to treating risk and return as a pair and their culture is to say, ‘we’ll take any credit risk, as long as the price is right’.”
Also, to succeed in credit derivatives, banks need significant trading capacity. They need to take credit risk in one form – say in a five-year default swap – and find counterparties on whom to lay it off. Often this means carving up the credit risk and repackaging it to suit the requirements of different institutional, corporate and bank clients. Commercial banks have not proved great traders in the corporate bond markets. They are unlikely to be different in credit derivatives.
The total-return swap is also gaining popularity among banks as a balance-sheet management tool. Total-return swaps are agreements in which the coupon and any capital appreciation on an instrument such as a bond are exchanged for a Libor-linked payment plus any depreciation in the capital value of the underlying asset. There is no exchange of principal and ownership and funding of the underlying asset is unchanged. If the swap maturity matches that of the reference asset it is simply a synthetic version of the asset that allows the holder to go long or short easily and without funding. If it is not, then the swap is a synthetic asset.
“A total-return swap is a synthetic long position in an asset and the risk can be compared to a futures transaction when price changes are settled frequently,” says Suresh Anisetti, head of credit derivatives and local currencies at Tokyo-Mitsubishi International in London. “A default swap with the same maturity as the underlying asset is equivalent to a long position in the underlying asset with the interest-rate risk immunized.”
There are similarities with repo, as Anisetti explains: “The difference between a repo and a total-return swap is that repo involves initial and final exchange of securities while a total-return swap is a pure risk transfer transaction and there is no need to exchange underlying securities. We don’t have to own the bonds when we agree to pay our client the total return from them in exchange for our funding/repo costs. We can then either buy the bonds in the market or may do nothing as we have decided to go short.”
From US to Europe
These instruments became popular first in the US because they offer leverage and because they give investors access to markets or parts of the credit curve not normally available. “They are used to obtain exposure to securities or instruments that are difficult to obtain, that they cannot purchase in the form in which they are traded in the cash market or whose form is too complex for them to analyze – for example Vnesheconombank loans or Mexican MYRAs,” says Anisetti. “Or let’s say there is a 10-year Mexican bond at par, trading at Libor+400bp. If you do a default swap at one year at Libor+120 the client is getting something not available in the market.”
Total-return swaps are now gaining popularity in Europe where the branches of many banks cannot buy securities but are allowed to execute off-balance-sheet transactions. They are also increasingly used by banks with balance-sheet constraints.
“We actively use total-return swaps to keep assets off our balance sheet,” explains a US banker. “We want to constrain the size of our balance sheet – don’t ask me why – which means that instead of just buying securities which increases the balance sheet we use balance-sheet providers to warehouse assets for us. So, if we buy bonds we get someone else to buy them and then we do a total-return swap with them which means that we assume the economic exposures of the underlying position. So the counterparty effectively funds our position, we pay the funding costs to them and receive a spread.
“When we trade with a balance sheet provider we post no collateral with them,” the banker says. “They have our highly-rated risk collateralized by a bond and we pay Libor+5. That is better than putting their money in one of our deposits paying Libor less 20 – we don’t want cash coming in, we hate deposits – or buying our commercial paper at Limean. So banks looking for yield enhancement at the short end find this attractive and we usually give them the right to cancel at 30 days.”
Profound implications
Banks do not only use other banks as balance-sheet providers. Large corporations able to raise cash at around Limean are happy to invest in total-return swaps at Libor+5bp with highly rated banks. Corporations are not usually worried about the size of their balance sheets so they will buy bonds, pay the return from the bonds straight on to the bank through the swap, fund that position at Limean and receive Libor+5 instead of depo or repo rates.
Like default swaps, total-return swaps have profound implications for smaller institutions. They offer banks with high funding costs the chance to earn positive carry from high-quality assets. And they offer small banks a way out of what is sometimes called the credit paradox. This is the situation in which small institutions with limited origination capabilities find themselves when they try to build the diversified portfolios they know are most efficient. As their resources and geographical networks are limited, they specialize in particular regions and sectors, building up dangerous concentration risks in their portfolios. They tend not to be asked into loan syndications and, in any case, pricing in that market is so tight that they cannot afford to participate. Total-return swaps allow these banks to lay off risk without damaging relationships and also to diversify their portfolios by gaining access to markets and companies they normally would not have the chance to buy into.
It is good business too for the banks selling risk. They pass only a percentage of the return from the underlying asset to the counterparty. This income is earned at almost no risk: the only risk to the bank is the replacement cost of the swap if the counterparty defaults.
Non-bank institutions have been slower to embrace credit derivatives. However, there are products that can be used to trade credit spreads and to enhance yield. “Institutional investors are using credit derivatives to dissect the credit curve to take specific bets that may not be available as securities in the cash markets,” says Citibank’s Watzinger. “Credit spread options and forwards let you slice up a credit into different tenors. So you can enter into a five-year credit and sell the first two years as a way of rolling down the credit curve.”
Credit spread options allow the buying and selling of an underlying credit-sensitive asset at a pre-determined price for a pre-determined period. Thus an investor might sell to an investment bank an option that gives the bank the right to sell a bond to the investor at a certain strike in the future. That strike would be set as a spread to Libor (or some other benchmark). If on the exercise date of the option the spread at which the underlying bond is trading is lower than the strike price then the option expires worthless and the investor pays nothing. If it is higher, the bank delivers the bond and the investor pays a price whose yield spread over the benchmark equals the strike spread. Investors can use combinations of calls and puts to lock in a current spread. Transactions can also be structured to allow investors to go long or short the spread between two different assets – say two countries’ sovereign debt or benchmark corporate bonds of different ratings or sectors.
As the reference asset is usually an asset-swap – floating-rate assets are favoured because the floating coupon immunizes interest-rate risk – spread options are no more than callable or puttable asset swaps and are sometimes known as asset swaptions. As such they are the oldest credit derivatives. The asset-swap market has always been driven by the desire for yield enhancement and as credit spreads tightened swappers had to come up with new ways to keep yields up. One way was for asset-swap buyers to sell calls exchanging up-front premium income against the loss of upside if spreads tighten.
Refining the science
A recent favourite credit spread trade was a bet that Emu would weaken Germany’s credit quality. The buyer holds an option to enter an asset swap on the 10-year Bund in three years’ time. Under the swap he receives the Bund spread to Libor which, today, would mean having to pay the arranging bank, since 10-year Bunds trade 12bp through Libor.
This is an area dominated by the investment banks, first because they have better relationships with sophisticated clients and second because they have big bond portfolios they want to trade and boost the yield on. Selling spread options is one way to enhance yield. “We don’t see so much of this, partly because we are a commercial bank and so focus on the bank loan-portfolio-related applications. Credit spread options are used by investment banks to manage their bond portfolios and to trade the term structure of credit risk – to roll down the credit curve,” says one commercial banker.
And it is at the investment banks that the science of credit trading and credit arbitrage is being refined and new credit derivative instruments and applications are being developed. Here traders are not so interested in the right price of credit risk – the commercial banks’ holy grail – as in decomposing the credit spread into its constituent parts and trying to predict how those spreads will trade. Asset swaps on callable and puttable bonds add one layer of complexity to the standard structure. Asset swaps on convertibles adds another. Nomura is one house that has decided to build its expertise in this way, creating client opportunities by proprietary trading.
Sophisticated methods
“Unlike a commercial bank we don’t have a big loan portfolio,” says Nomura’s Austin, “so we focus on running a credit spread trading business. This has led us into the development of some very sophisticated models for measuring spread behaviour and some sophisticated methods of trading spreads.” The team tends not to take outright directional views but instead concentrates on arbitraging anomalies in credit spread pricing by creating hedged portfolios of assets and derivatives. For example, default swaps may be used to take advantage of the spread between the subordinated debt and senior secured obligations of the same issuer in the expectation that a more normal spread relationship will prevail.
Arbitrage teams such as Nomura’s look for more subtle anomalies. “We might analyze the spread and look at the implied forward default rate,” says Geoff Chaplin, quantitative analyst, credit derivatives. “When you do these analyses you can find negative implied default rates and rates higher than 100%. These are clearly anomalies.” The bank is also looking at spread correlation trades and at the relative spreads on similar bonds in different currencies.
Liquidity is still a problem, but these arbs will create it. “The liquidity for hedging credit risk is less than you have in other markets.” says Ron Tanemura, managing director at Deutsche Morgan Grenfell in London. “Where you may be able to buy good relative value, arbitrage, or locking profits in, is more difficult. You think you’ve got cheap volatility but you tend to just sit on it because there is no bid for the vol and transaction costs make delta hedging unrealistic. However, liquidity is improving. Default swaps provide a realistic alternative to shorting cash.”
The work of these new arbitrage teams proves the sceptics were wrong. Those – including this writer – who said credit risk was different, that credit derivatives were not true derivatives and that they would never become as widespread as other types of derivative are being overtaken by developments.
