Now let us look at our last type of synthetic arbitrage CDO. The tranches of the CDS indices described in Chapter 11 are quoted and traded like liquid synthetic CDO tranches. As shown in Exhibit 13.2, the Dow Jones CDX.NA.IG is divided up into 0% to 3%, 3% to 7%, 7% to 10%, 10% to 15%, and 15% to 30% tranches. For the invest- ment-grade indices, equity tranches require an upfront payment from the protection buyer to the protection seller. After that, a fixed 500 bps per annum is exchanged. For the high-yield index, the first two tranches require upfront payments but have no running fee. The higher tranches of the indices trade solely on their running fees. Exhibit 13.2 gives details of tranche structure for various CDS indices.
Investors in standard tranches often engage in various forms of long/short trades. They might sell protection on an equity or first-loss tranche and buy protection on a more senior tranche of the same index.
(In market parlance, they are said to be long the equity tranche and shortthe more senior tranche.) Being long a tranche can be confusing to some investors because one has sold protection on it, but the situation is analogous to being long a bond. When one is long a bond or long a CDO tranche (having sold protection), an investor abhors a default and does not want interest rates, credit spreads, or CDS premiums to rise or widen.
Another popular long/short trade is to sell protection on a tranche in a longer maturity and then to buy protection on the same tranche from the same index in a shorter maturity. Hedge funds are big partici- pants in long/short strategies.
262 SYNTHETIC CDOs
EXHIBIT 13.2 Standard Tranches of CDS Indices CDX NA IG
iTraxx Europe, iTraxx Asia (ex Japan), iTraxx Japan
CDX NA HY
CONCLUSION
Synthetic arbitrage CDOs incorporate ever-evolving structures that have rapidly gained acceptance in the market. Based on the actual amount of credit risk transferred, synthetic structures are more important than those using cash collateral.
Attachment/
Detachment Points
Upfront Payment
Running Premium
Tranche 1 0%–3% Yes 500 bps
Tranche 2 3%–7% No Yes
Tranche 3 7%–10% No Yes
Tranche 4 10%–15% No Yes
Tranche 5 15%–30% No Yes
Attachment/
Detachment Points
Upfront Payment
Running Premium
Tranche 1 0%–3% Yes 500 bps
Tranche 2 3%–6% No Yes
Tranche 3 6%–9% No Yes
Tranche 4 9%–12% No Yes
Tranche 5 12%–22% No Yes
Attachment/
Detachment Points
Upfront Payment
Running Premium
Tranche 1 0%–10% Yes No
Tranche 2 10%–15% Yes No
Tranche 3 15%–25% No Yes
Tranche 4 25%–35% No Yes
Tranche 5 35%–100% No Yes
c13-Synth Arb CDOs Page 262 Monday, March 6, 2006 11:21 AM
One reason for this is the cheap price of unfunded credit protection on super-senior tranches. The economics of synthetic deals are further enhanced by the ability to more easily and quickly ramp-up a portfolio and execute a CDO. Finally, selling protection via credit default swaps is often more rewarding than assuming credit risk through a cash bond.
Synthetic CDO structures are varied. Some synthetic CDOs are static, while others are managed. Some employ standard synthetic archi- tecture (i.e., writing down from the bottom of the capital structure first, no cash traps, and cash settlement). Others employ some aspects of cash flow CDO architecture (i.e., no write-downs until the end of the life of the deal, less equity compensated by cash trapping-coverage tests, and physical settlement).
The rise of single-tranche CDOs affords greater flexibility to inves- tors, including ease and speed in execution and the chance to amend the CDO’s terms later in its life. Standard tranches of CDS indices are the most liquid type of synthetic CDO; they also provide investors with the chance to put on various long-short positions to assume customized risks.
c13-Synth Arb CDOs Page 264 Monday, March 6, 2006 11:21 AM
CHAPTER 14
265
A Framework for Evaluating Trades in the Credit Derivatives Market
s discussed in Chapter 11, the markets for credit default swap (CDS) and synthetic collateralized debt obligation (CDO) have grown tre- mendously, both in terms of trading volume and product evolution. In terms of product evolution, CDSs have developed from highly idiosyn- cratic contracts, taking a great deal of time to negotiate, into a liquid market offering competitive quotations on single-name instruments and even indices of credits. Synthetic CDOs have evolved from vehicles used by commercial banks to offload commercial loan risk to customized tranches where investors can select underlying credits. And the rise of standard tranches of CDS indices has blurred the distinction between credit default swaps and synthetic CDOs.
As a result of these instruments, credit trades offered to participants in the credit derivatives market every day involve assessing trade-offs, such as the following two examples:
■ Is it better to sell credit protection on a single BBB rated corporate name or on the BBB rated tranche of a synthetic CDO?
■ Is it better to sell credit protection on a portfolio of BB names or on the equity tranche of a CDO comprised of A-rated names?
These examples offer us the chance to analyze risk to a single credit, a portfolio of credits, and a tranche of a CDO. They also allow us the opportunity to compare risk to credits with the same underlying rating and to credits with different underlying ratings.
A
266 SYNTHETIC CDOs
In this chapter, we describe an empirically driven methodology that uses historical default and loss-given-default data to answer these types of questions. Specifically, we show how the single-name, portfolio, and tranched positions mentioned would have performed had they been entered into in 1970, 1971, and so on. We begin by examining the his- torical default rate and loss given default data available to investors.
Then we discuss the considerations that go into making a model that uses this data to arrive at a historic perspective. Next, we show historic pro forma results for the positions mentioned above. Finally, we discuss the advantages and limitations of this approach and the type of empiri- cal research that would help improve the analysis.