Next-generation credit derivatives - A market forum

How do you see the development of an exchange-traded credit futures market?

Credit market liquidity is certainly growing, with three relatively new market participants boosting end-user demand. Over the last five years collateralised debt obligation (CDO) issuance has grown dramatically, dedicated credit hedge funds have sprung up and pension funds have increased their fixed-income exposure at the expense of equities. Exchange-traded credit futures may seem a natural development given this growth but I feel that success or failure of such a product may be down more to the specific details than the general environment, particularly given that the current suite of iTraxx/CDX over-the-counter (OTC) index products currently serves much of the role of a credit index future.

The obvious reasons for a futures market are liquidity, standardisation, a single counterparty for collateral and ease of clearing and administration. The bid-offer spread on the main (investment-grade) OTC product is now typically one-quarter of one basis point, with over a dozen counterparties quoting. The size in normal market conditions is more than sufficient for most participants’ needs. A futures market would have to trade in tenths of a basis point and have pretty low contract charges to be competitive on cost for most users. Standardisation in the OTC market was an issue before iTraxx/CDX but this has been dealt with, plus the quarterly contract date means liquidity is focused to a specific date.

So far, a futures contract would seem to have little to offer from an end-user perspective, but the idea of managing a single set of collateral and margining requirements is operationally attractive, as would be the minimal trade documentation and rapid confirmation process. Whilst the OTC market seems to offer sufficient liquidity, it currently suffers from a frustrating documentation backlog, and this might be enough to tempt users into an exchange-traded instrument.

What is the attraction of constant maturity credit default swaps (CMCDS)?

The history of derivative market developments from options through swaps to CDS is that the instruments allow better isolation of specific exposures. This in turn facilitates greater precision in risk transfer, hedging and active risk-taking.

CMCDS follows the pattern, allowing the spread-widening risk and default risk embedded in standard CDS contracts to be separated. In many investment portfolios the manager wants to get paid for a research view that Company X is in good shape for a certain number of years and is not expected to default. Sometimes this leads to a value decision based on the current spread level, so a standard CDS or bond investment is the right tool, but at other times the mark-to-market just represents unwanted volatility. Accrual accounting is unacceptable practice in most portfolios for obvious reasons, so CMCDS, with their floating credit spread, can be the solution.

In the same way that exotic options allow different pieces of the expected return distribution to be owned, the CMCDS also allows, for example, a bearish view to be expressed on a company that is not expected to default, without paying for undesired default protection. By receiving a floating CMCDS spread and paying the fixed CDS spread, the position benefits in a spread widening, whilst carrying far less negatively than a straight bought protection position, in exchange for effectively being ’knocked out’ on default.

CMCDS participation rates are driven by the curve slope, so a third ’use’ for the contract is to take a curve view, though for many users this embedded curve exposure represents a risk that may need hedging rather than a feature.

Have investors learned anything from the correlation repricing earlier this year?

More of a reminder than a lesson – liquidity is key! For a number of reasons liquidity in correlation markets will be a problem if tested. Firstly, as a tradeable exposure, it is a relatively new market, so there are only a limited number of players. Secondly, very few real-money investors have the need and/or ability to trade tranched product, so a pool of sizeable potential players are currently sidelined even if they could have seen, at the height of the sell-off, that there was some value in stepping in to support the market. Thirdly, the innovation that resulted in tradeable correlation was born out of bank risk-transfer demand, so at the early stages of the market, as we are in now, flow is mostly one-way. If end users, in this case a small number of mostly leveraged funds who had been the ’natural bid’, suddenly become the offer, there’s no-one there on the other side.

For other participants, such as funds who own CDO equity for straight levered carry, there will be those who fully understand and can manage their correlation exposure and those who at least understand that they have a correlation exposure. There will also be a few for whom, if they don’t accrual-account, the effect on their mark-to-market may have come as a bit of a surprise, with the lesson learnt being to brush the dust off the technical CDO piece they never quite read!

Furthermore, the wider credit markets were impacted via the delta-hedging activity of those with correlation exposures, with spreads moving violently wider before rapidly recovering. Participants in the vanilla credit markets must be more aware in future of developments in the correlation market, and how these may represent a risk to their market positions.

What is the motivation for using credit default swaps on asset-backed securities (ABCDS)?

Like many derivatives, ABCDS are an inexpensive way to get into the market. Many institutions, from monolines to hedge funds, see selling protection as a cost-effective way to gain exposure to asset-backed securities. Also, for smaller tranches where there is less liquidity, selling protection serves as a very good alternative for cash investment.

ABCDS also offer a way to synthetically short the asset-backed securities market. Whereas the cash ABS market is a long-only market, synthetics allow traders to hedge their exposure by buying protection in order to go short specific areas of the capital structure. Previously, investors had to simply sit there in this long-only market. Also, with indices there is the ability to go short and tailor the maturity exactly as required when the cash is not available.

How do you see the ABCDS market shaping up this year?

The main interest will be in the lower-rated issues: triple-B and below. That is where both the yields and volatility are highest. There has been some volume among higher-quality issues, but the bulk has been – and will continue to be – lower down the capital structure. Home equity currently accounts for around 80% of all trades and it will continue to take the largest proportion. Second will be credit-card related issues.

Asset-backed indices (ABX and CMBX for asset-backed securities and commercial mortgage-backed securities respectively) are being developed by a working group of dealers drawn from half a dozen firms. These indices should start trading by the end of the year. This will be an entry point into the market for non-traditional ABS players and should facilitate correlation trading. Another development in the coming months will be investor education. ABCDS are attractive vehicles for both taking and transferring risk and banks are going to be out there educating clients and marketing the benefits.

For our part, we at GFI Group sit in the middle of the inter-dealer market, seeing trades and monitoring flow, and the information that we pass to our clients makes its way out to the buy side and all helps the educational process. The market for ABCDS will continue to benefit from standard documentation. The International Swaps and Derivatives Association published ABCDS documentation in June this year. This new documentation, based on experience in other more mature CDS markets, should boost the market significantly. GFI played a part in developing the documentation through facilitation of the dealer working group and we were pleased to do so as part of our commitment to developing new markets.

Why is the market for collateralised debt obligations (CDOs) of ABS expanding so rapidly?

The attraction of CDOs of ABS is in the granular nature of ABS. They are a good way of getting exposure to consumer risk rather than corporate risk, providing a diversification for CDO investors. This is because CDOs tend to be based on tens of thousands of individual consumers rather than the 100 or so in a corporate CDO. We are also seeing synthetic CDOs selling protection on asset-backed securities themselves. This is a flow that is affecting the market and providing extra liquidity.

Will CDOs of ABS continue to be considered safe havens?

They will continue to be regarded as safe havens as long as there is no deterioration in the consumer cycle. But that is a significant risk. Many market commentators believe, for example, that the housing market is at a peak and if overextended consumers default, spreads will widen dramatically.

What are the challenges of developing efficient straight-through processing (STP) for a market that is becoming increasingly complex?

The first challenge concerns technical and operational issues related to connecting systems and resolving field, entity and counterparty identifiers. Rapid progress is being made industry-wide enabled by increased standardisation, more automated trading and the creation and adoption of third-party STP services. The second challenge is tied up with constant product innovation in credit facilitated by inter-dealer brokers such as ourselves. This innovation creates new instruments and new ways of trading. This makes effective STP a near impossibility as systems cannot keep up – but volumes in this area tend to be low, which reduces the impact of inefficient STP, but does hinder market development.

How do you see the market for credit default swaps on asset-backed securities (ABCDS) developing this year?

The ABCDS market has begun to develop some real shape and direction. Initially it was driven by a few dealers who were using their balance sheets and business flow to buy cheap protection from collateralised debt obligation (CDO) managers and investors who wanted access to assets. It is slowly becoming a two-way market with a development of a correlation component and regular two-way markets.

We think, given the natural illiquidity of the asset class, development of the market will continue to be spotty but as the principal dealers agree on documentation with the International Swaps and Derivatives Association and, more importantly, as and when there are agreed indices of the main asset classes (home equity loans, commercial mortgage-backed securities and perhaps CDOs) then the market will develop more rapidly with more proprietary traders getting involved. Like the correlation markets, the ABCDS market will continue to be driven by the investors and arrangers rather than by traders. However, we think the principal development in the markets over the next 12 months will be the beginnings of a true traded market with the agreement of the components of a traded index.

Why is the market for CDOs of ABS expanding so rapidly?

Over the past few years there has been a fundamental shift of investment into fixed income. This shift in demand has coincided with deleveraging from the corporate sector and a drop in issuance from sovereigns. The result, with demand far outstripping supply, is that credit-based investors have been scouring the universe of available assets for investments. For those investors who are constrained by investment-grade ratings, this has meant the investments have been largely in securitised assets.

The source of such assets is limited. The synthetic market is largely driven by correlation and as we have seen recently the rated market is really driven by the appetite of equity buyers. Recently that demand has fallen as the perception of idiosyncratic risk (principally from the auto sector) has risen. So notwithstanding the appetite of investors for high-rated assets based on investment-grade credit, the ability of arrangers to fill that demand is constrained. The cash market is much larger but the complexity of much ABS is difficult for investors to absorb. The more liquid asset classes have become relatively expensive and so CDO of ABS can be seen as a market where investors pay intermediaries to take out the underlying complexity. The assets are seen as stable both from a spread and rating migration perspective and therefore have created an interesting investment class.

Will CDOs of ABS continue to be seen as a safe haven?

Only for so long as default rates and negative credit migration on the underlying assets remain positive. To that end the biggest impact on safe haven status is the performance of the US sub-prime mortgage market. That market has been the source of much of the collateral seeding CDO of ABS deals. For the past few years default rates have been low, but more particularly loss rates have been very low as the continued rise in US house prices means that most delinquencies or non-payment have been recovered out of asset sales. The market is sensitive to a slowdown in house prices especially in the three areas in which securitisation is most prevalent (Metropolitan North East, Florida and California).

It is our view that as US interest rates continue to rise and disposable income does not keep pace with that rise, so there will be a rise in delinquency and defaults, and house price increases will slow and in some cases reverse with the effect that recoveries will fall and loss rates will rise. This will affect the sub-prime market and will make assessment of tranche quality and the effect of prepayment on tranches and deals very important. It is our view that the effect on the home equity loan market will be to negatively affect the quality of mezzanine tranches (particularly triple-B and double-B) which will widen spreads and in turn feed through to CDOs of ABS. That said it will take the best part of the next 24 months for the effect to be properly felt.

Have investors learned anything from the correlation repricing earlier in the year?

A difficult question. A basic rule of markets is that crowded trades become volatile when they move against you. What happened in the correlation market was that the extent to which the buoyancy of the synthetic CDO market depended on the willingness of traders or investors to sell first-loss protection (buy equity pieces) became very clear in the period from late March to May. Until then the market had defied gravity for the best part of 18 months as dealers created rated assets cheap to the cash market and captured ‘profits’ by finding plentiful equity and super-senior buyers. Because the credit fundamentals were strong, all players in the market behaved as if the returns being generated were of high quality. Unlike the cash CDO market the correlation market marks to market very quickly and when demand dries up or supply exceeds demand values change very quickly.

For investors in rated assets who are not mark-to-market sensitive, the troubles will have been more to do with negative publicity than with losses. For those investors who marked positions to market, the volatility of their positions made it clear that most were not being adequately compensated for the risks they had assumed.

The interesting by-product of the market volatility however was that as the availability of rated mezzanine dried up so those very same investors began buying leveraged super-senior tranches with unwind triggers based on credit spreads. Thus investors who historically have only been directly exposed to losses generated by default have taken positions which could unwind with significant losses if spreads widen. The transaction has been incredibly successful over the past few months and will feed into a massive market volatility if spreads widen sufficiently to trigger unwinds of those positions.

The transactions look a little like the leveraged interest rate transactions done in the early 1990s (Libor-squared and Libor-cubed) which were also about generating returns in a low spread environment. As far as traders are concerned the market has been clearly repriced and the net effect is that the size of the securitised synthetic market will be more constrained than it has been for the past two years.