The number of traditional cash CDO defaults has been widely publicized and, obviously, received widespread damage assessments. But what of credit events that factored into synthetic CDO downgrades over the last two years? Calculating and modeling for events in the world of nontangible assets is an adjustment akin to moving from the roulette wheel to the baccarat tables.
The effects of credit events on a synthetic CDO are layered. Synthetic CDOs harness the credit derivatives market by utilizing credit default swaps tied to reference entities. Although those reference entities are public companies, detailed information on credit events available to the public is next to none. Sources state that synthetic CDO investors track and record their own independent information.
Yet, for all the mystique, synthetic CDOs comprise a sizeable portion of CDO new issuance this year. According to ASR, 11 synthetic CDOs are currently in the pipeline and await pricing. In total, they represent $14.5 billion in referenced collateral.
Banc of America Securities is reportedly marketing a $1 billion synthetic CDO, again comprised of IG debt, called Westmoreland CDO 2003-1. The collateral manager of the CDO is Westmoreland Capital and, according to sources, $40 million triple-A rated A class, with a five-year average life is being marketed at 65 basis points over the three month LIBOR.
Currently, an eye-popping $5 billion synthetic CDO called Fibonacci 2003-1 is in the pipeline, say market sources. The vehicle is a culmination of investment grade (IG) debt that has been packaged together by Deutsche Bank, which serves as both the agent and asset manager on this behemoth. The arbitrage cash flow-styled deal contains a super senior tranche that comprises 86% of the pie, or $4.3 billion. However, there was no price talk available on the $175 million B tranche, which is triple-A rated.
Overall, according to the British Bankers' Association, the credit derivatives market is expected to reach $4 trillion notional by next year. Investors who delve into synthetics state they enjoy the relatively cheaply priced tranches when compared to traditional cash CDOs. A recent figure from Goldman Sachs estimated that $38.2 billion in static synthetic CDOs were issued in the U.S. last year; Europe saw EUR$53.8 billion. Both those figures are up tremendously from 2001 when $21.8 billion in the U.S. and EUR25.4 billion in static synthetics were issued.
Modeling for the cache of credit events that effect collateral within traditional cash CDOs is a challenge on its own, said one source familiar with synthetics. But mitigating for the effects of credit events on synthetic CDOs is an entirely different ballgame, he said.
Last week Fitch Ratings published a study analyzing the credit events, called to date, involving synthetic CDOs. Entitled "Credit Events in Global Synthetic CDOs" the study appears to be the first one that analyzed data from investment banks and asset managers.
"We are trying to offer more information in an attempt to increase transparency in the market," said Jill Zelter, managing director at Fitch.
The study found that a record number of credit events for synthetic CDOs were generated during the 2001-2002 period, as a difficult credit environment combined with fallen angels and fraud-related cases. And from 2001 to Jan. 31, 2003, roughly 112 credit events were called, totaling $885 million in notional exposures among 47 Fitch-rated transactions.
Out of the top 28 reference entities that triggered credit events, WorldCom Inc., Enron Corp. and Teleglobe Inc. grabbed the most with 25, 17, and 11 events, respectively, which occurred between 2000 to January 2003. Digging further through the list, during the same period, Marconi plc accounted for seven credit events, Conseco Inc. two, Global Crossing four, Kmart Corp. three, and Xerox Corp. three.
The study offers a catalog of credit events, some familiar and some not. Some 94.5% of credit events were called under bankruptcy or failure to pay. Not surprisingly, an "overwhelming majority" of reference entities with credit events were companies with ratings that were highly speculative or in default at the time of the event. Some 67.3% of credit events were attributed to reference entities with DDD' to D' ratings. Only 11.6% of the reference entity ratings are B-' or higher.
The highest level of volatility was experienced in 2002 as Fitch found that 73 credit events occurred, attributable to 16 reference entities, compared to 34 credit events involving nine reference entities in 2001. For the first month of 2003, that level appears a bit more similar to levels in 2000; January 2003 saw three credit events involving one reference entity, whereas all of 2000 recorded two credit events involving two reference entities.
Furthermore, there was little difference in recoveries due to variations in the obligation valued. Most credit evens valued senior unsecured bonds or loans that were pari passu, the study found.
"Recoveries in credit events and the number of bids obtained in the valuation process were positively related," said Andrew Jackson, senior director at Fitch. "The higher number of total bids obtained, the higher the observed recoveries. (Furthermore) the length of the valuation process also affected the recovers, which were higher when the valuation process lasted from 30 to 50 days and 90 to 180 days, while recoveries for credit events settled from 50 to 90 days were below average," he said.
In all, the volume of credit events for synthetic CDOs was reflective of investment-grade bond defaults. Fallen angel defaults totaled, $24 billion in 2001 and $35.7 billion in 2002.