Why did all those super-seniors exist?

By Felix Salmon
April 28, 2010
Pablo Triana asks, via email, about all that super-senior risk that was sloshing around the financial system when everything imploded in 2008. Did most of it actually need to exist? Or was it a conscious and voluntary decision on the banks' part to create it and, often, to hold on to it?

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Pablo Triana asks, via email, about all that super-senior risk that was sloshing around the financial system when everything imploded in 2008. Did most of it actually need to exist? Or was it a conscious and voluntary decision on the banks’ part to create it and, often, to hold on to it?

Clearly with cash CDOs, the super-senior tranche has to exist. And equally clearly, after looking at the Abacus deal, it does not have to exist in synthetic CDOS. This was news to me, but conceptually it makes sense — since a synthetic CDO is a zero-sum bet between two parties, they can happily bet just on the performance of some tranche without there needing to be a parallel bet on the performance of some other tranche.

It’s worth adding here that technically synthetic CDOs are not zero-sum bets: because investors buy them with cash, that cash has to be invested somewhere, and the interest earned on it makes the CDO a positive-sum game overall. Think of the senators who said yesterday that Goldman Sachs was just running a casino: it’s one of the more popular of the Goldman metaphors these days. In a casino, if you’ve made a big bet, the house will happily comp your room and your drinks. And if Goldman sold you a synthetic CDO, it will give back to you the interest that your money earns before it gets shipped off to John Paulson.

But the fact is that when cash CDOs gave way to synthetic CDOs, it was possible for the banks creating these new instruments to solve the problem of super-senior tranches by simply never creating them in the first place. Instead, they did create them, and when they couldn’t insure them through AIG any more, they just kept them on their books instead. Why?

I think there are two main reasons. The first is that the people buying the short side of the deal had done a lot of work on mortgage-backed securities, and knew which kind of securities they wanted to go short. When banks put together synthetic CDOs, they would fund the long side by selling bonds to investors like IKB, and they would fund the short side by selling subprime CDS to investors like Mike Burry or Andrew Lahde. And in the middle, the synthetic CDO structure allowed them to turn the interest payments from the shorts into a income stream that they could then sell to the longs.

But if you do it that way, you’re bound to end up with a super-senior tranche. The only way to avoid having a super-senior tranche is to fund the short side not by selling protection on subprime bonds to people like Burry and Lahde, but rather to sell protection on certain tranches of the synthetic CDO itself to a sophisticated investor like Paulson who understands exactly how the structure works because it was his idea in the first place.

In Michael Lewis’s book, Burry is surprised when he finds out that his short positions are being bundled into synthetic CDOs: he was never an active participant in structuring these things, partly because he never had $15 million to pay to Goldman as a fee. If Goldman had tried to sell Burry protection on the triple-A tranche of an Abacus CDO, he wouldn’t have been interested: the structure was extremely complicated, and if he wasn’t involved in creating it, he couldn’t be sure that his analysis would play out as expected. People like Burry just wanted to look at bonds in the market and buy protection on them.

And it seems that even Paulson felt the same way: for all intents and purposes, his role in the Abacus deal was to buy protection on a bunch of subprime bonds. He did so in two different transactions: the central Abacus deal with ACA and IKB was one, and then there was a separate super-senior deal. And the equity tranche never existed or got funded. But essentially what Paulson ended up with was very similar to what he would have got if he had simply bought protection on the bonds. Maybe that’s why Goldman ended up with that 45-50% tranche: Paulson didn’t really have much interest in buying protection on this slice or that slice, he just wanted protection on the whole thing.

So while in theory it was possible to put together synthetic CDOs where the super-senior tranche didn’t exist, in practice it seems that such things were pretty much impossible to sell to the relatively small universe of protection buyers.

And then there’s the second reason why banks felt comfortable holding on to super-senior risk. John Cassidy writes today about how the banks failed utterly to deal well with model risk:

The risk models that were commonly used on Wall Street failed abysmally. Not only did they fail to protect their users from a bad outcome, they made such an outcome far more likely. In short, the risk models added to systemic risk.

In part, this was a failure of statistical modeling. The techniques that the risk modelers used weren’t up to the task they set for themselves. But it was also a problem of how the models were used. Rather than looking on them as a useful but limited tool, banks and other institutions used them as a substitute for proper risk management, and as a justification for taking on more leverage and more risk. This explains how the risk models made the entire system more risky.

The risk models all considered the super-senior tranches to be significantly safer than triple-A, and the banks’ internal risk-management systems loved that kind of thing, especially when those tranches paid out even a tiny amount over Libor. After all, Libor is only double-A, and if a bank can fund itself at Libor and make a profit by investing in what the models said were completely risk-free securities, then why not do that? As Cassidy explains,

Model-based risk management seduced the regulators, too. Under the Basel system of international banking regulation, big financial institutions were allowed to use their own risk models in setting their capital reserves. Alan Greenspan and many other policymakers insisted that the development of “scientific risk management” had made the system a lot safer.

Finally, of course, there’s the old question about incentives and executive pay. No one ever got paid a seven-figure bonus for refusing to do a bad CDO deal, while lots of people (like Fabrice Tourre and the other Goldman types we saw getting grilled yesterday) got paid millions for doing them. Including a super-senior tranche undoubtedly made it much easier to do more and bigger synthetic CDOs, and so that’s what the structured-product desks did. You can call it regulatory arbitrage if you like, or you can just call it the natural consequence of the incentives built into the bonus system. Either way, banks like Citigroup and UBS lost billions on liabilities that senior executives never really even knew that they were holding. Because no one bothered to question the models underneath them.

6 comments

Comments are closed.