Super Senior Sophistry: The AAA Disappearing Act
Book: Structured Finance and Collateralized Debt Obligations | Author: Janet M. Tavakoli | Publisher: John Wiley & Sons (2008) | ISBN: 978-0-470-44344-6
What if I told you that the biggest, most “safe” tranche in the whole CDO structure had no standard definition, no standard price, and no standard way to mark it to market?
That’s what Chapter 13 is about.
The Super Senior Was Born Out of a Problem
Banks had a problem. They were making investment-grade loans at tight spreads. When they tried to securitize those loans, the economics didn’t work. Issuing a large AAA tranche was expensive. The cost of the securitization would exceed the income from the underlying loans.
So banks invented a solution: the super senior tranche.
Here is how it works. A rating agency tells you how much subordination you need to get a AAA rating. Say it’s 12 percent. That means 12 percent of the deal absorbs losses first before the AAA gets touched.
In a regular cash CDO, the AAA tranche would be everything above that 12 percent line. It would be a big chunk of the deal.
In a synthetic CDO, structurers added a step. After solving for the AAA attachment point, they split that big AAA piece in two. A smaller AAA tranche sits just above the 12 percent subordination. Above that sits something even bigger: the super senior. Often 80 to 90 percent of the whole deal.
The claim was that the super senior was so safe it barely needed any spread. Banks would pay 6 bps on it instead of 50 bps.
For a $5 billion transaction, that was $17.6 million per year in savings.
The Cash Flow Magic Trick
Tavakoli calls this “the greatest triumph of illusion in twentieth-century finance.” And she says it is perfectly legal.
Here is the trick. Banks moved 44 bps of spread from the big old AAA tranche into profit for the deal arranger. The risk did not disappear. It just got repriced as if it had.
The super senior income was so cheap that it created a slush fund. Banks that retained both the super senior and the equity tranche could shuffle that synthetic income around in ways that made equity returns look amazing. A 60 percent IRR on equity! Incredible deal!
Except a large chunk of that IRR was just cash stripped off the super senior and reclassified.
The AAA That Is Not Really AAA
Here is the part that should make you uncomfortable.
Rating agencies do not rate the super senior. They solve for the AAA attachment point and then structurers invent the super senior above it. The rating agencies have nothing to do with this step.
Standard & Poor’s position is actually clear and damning: S&P does not recognize anything above AAA. So if you buy the AAA tranche of a synthetic CDO that has a super senior above it, S&P would tell you that tranche is not really AAA. It is the first-loss piece of what was formerly a AAA tranche.
Tavakoli walks through the math. Two identical $25 million investments, same portfolio, same rating. One is the AAA of a regular cash CDO. The other is the AAA of a synthetic CDO with a super senior above it. If 2 percent of the portfolio defaults (say WorldCom and Enron, each 1 percent of the deal), about 2.3 percent of the first investment loses its AAA quality. About 40 percent of the second investment loses its AAA quality.
Same rating. Very different risk.
Nobody Agrees on What Super Senior Even Means
There is no market standard. One bank uses a one-in-a-million probability of default as the cutoff. Another requires a 5 percent AAA buffer below the super senior. Some deals Tavakoli saw had no AAA tranche at all – they went straight from AA to super senior.
Ambac reviewed a deal and said they wanted 10 percent subordination (15 names out of 150 before they got hurt). Another counterparty said they’d do it with only 12 names. Both called what they held a “super senior.”
The super senior is whatever a structurer says it is.
Ratings Shopping
Tavakoli watched this happen in real time. Arrangers would shop deals to whatever rating agency gave them the most lenient subordination. S&P was widely perceived as less strict than Moody’s on corporate portfolios. Less subordination helps the CDO arbitrage, so arrangers preferred S&P’s model.
She describes one deal where she flagged the subordination as insufficient under Moody’s. The arranger said they were using S&P’s benchmark. She declined. They found other buyers at wider spreads.
Her advice: always ask for Moody’s benchmark tranching. It generally gives you more protection.
Regulators Asleep
Tavakoli tells a story about a senior bank regulator she discussed this with in late 2002. He had never heard of the issue. As she explained it, she watched him talk himself out of caring.
His response: “I’ve read that the market figures out all new products in around eighteen months, so this will soon be generally known.”
She points out: at the time of that conversation, BISTRO deals (JPMorgan’s super senior prototype) had been around for five years.
Junior Super Seniors and the Rest
When monolines didn’t want the full super senior, structurers created yet another tranche: the junior super senior. Between the AAA and the super senior. A tranche with no definition, above another tranche with no definition.
Tavakoli’s summary is dry and accurate: “The term junior super senior makes a mockery of the entire concept of the super senior tranche. Why stop at two super senior tranches? Why not have senior, junior, sophomore, and freshman super seniors?”
Leveraged Super Seniors and CPPI
For investors who wanted even more pain, structurers packaged super seniors into leveraged super senior (LSS) products. You put up 10 percent of the notional in cash and get exposure to 100 percent of the super senior’s risk. A 10 percent move in the value of the underlying wipes you out completely.
These were popular from 2004 to 2006 before the obvious problems became impossible to ignore.
Some LSS product got embedded in constant proportion portfolio insurance (CPPI) products. CPPI embeds a zero-coupon bond so your principal looks protected. Tavakoli describes this as a parking lot for risky products designed to lull investors into a false sense of security. If you need to sell before maturity, you almost certainly take a loss.
The Punchline
The head of structuring at a U.S. investment bank once justified holding enormous super senior positions by asking: “Has a super senior ever experienced a default?”
The answer was no. Until 2007, when CDOs backed by subprime collateral unwound and investors got back less than 25 cents on the dollar.
Tavakoli had been writing about this problem for five years before it happened.
Previous: SME Securitization and Basel II: Europe vs America