There was a flurry of predictable coverage on Wednesday when the news broke that a subprime mortgage bond called Springleaf Mortgage Loan Trust 2011-1 was being priced with a triple-A rating from S&P. The WSJ’s Al Yoon has the best detail on how the bond is structured; basically, 5,629 performing subprime loans are being put into a pool, and holders of triple-A bonds get the first 48.85% of the cash flowing from that pool. Because you can be pretty sure that at least 48.85% of the mortgages will be making their payments, the triple-A tranche is considered ultra-safe; Daniel Indiviglio has attempted to defend the deal and its rating on those grounds.
But as Bloomberg reports, it’s not as simple as that. There are at least five main arguments for subprime bonds not to have a triple-A rating when the US is only double-A:
- The US is not revenue-constrained in the way that a structured product is: the US can raise taxes, but Springleaf can’t raise mortgage rates. In extremis, the US can act to avert a default; the Springleaf deal can’t.
- If the US were to default, the repercussions would be so huge and unpredictable that it’s very hard to see how anybody can have faith in the Springleaf deal continuing to pay out in full. That’s the thinking behind the institution of the sovereign ceiling: you can’t really have more faith in an American bond deal than you can in America itself.
- A triple-A rating is pretty useless if there’s a good chance that it’s just going to get downgraded in the near future. S&P downgraded $30 billion of re-remics shortly after giving them post-crisis triple-A ratings; there’s no good reason to believe the same thing won’t happen here.
- The reason for the triple-A is that mortgage payments are uncorrelated: even if some people default, they won’t all default at the same time. But as we learned so painfully in 2008, correlations tend to spike dramatically during a crisis, and the whole point of a triple-A bond is that it’s meant to be crisis-proof.
- Rating structured bonds is impossible. These things are simply too complex and unpredictable to be able to apply a triple-A rating to. Haven’t we learned anything from the financial crisis?
And even putting all those arguments aside, the very existence of this bond is depressing, for a couple of reasons.
Firstly there’s that 48.85% number — much, much lower than the kind of numbers we saw during the subprime boom, when well over 90% of a subprime pool would regularly receive a triple-A rating. S&P is putting up unprecedented barriers to structured products getting a triple-A rating — and companies like Fortress (NYSE:FIG), which owns Springleaf, are still going through all the hoops they need to go through to get one. What that says to me is that we really haven’t learned one of the main lessons of the financial crisis — that triple-A ratings are treacherous and largely meaningless things, and that it’s silly to go to great lengths to get one, or to confine yourself to investing only in bonds with that particular seal of ratings-agency approval.
Secondly, there’s the fact that the Springleaf deal is made up of old, seasoned loans, rather than new mortgages. It’s financial engineering, in other words, and isn’t driving any capital to the housing market. During the bubble, there was too much money going from structured finance into the housing market; right now, however, the market needs all the help it can get, and it’s getting none at all from that quarter.
Finally, it’s worth noting what people are buying, in terms of bonds backed by brand-new mortgages. They’re buying agency bonds, issued by Fannie Mae and Freddie Mac, which carry a double-A credit rating. (And which yield much less than the 4% on the Springleaf deal.) If we ever want to transition to a world where the private sector is funding American mortgages and the federal government is getting out of it — if we ever want to realize the stated aim of politicians on both sides of the aisle to make Frannie much smaller — then we’re going to have to find some kind of palatable alternative to agency bonds. But there’s no indication that any such alternative exists.
There’s no way that the Springleaf model can be extended to new mortgages. The overcollateralization in Springleaf is too great: it would be uneconomic to put such a deal together with new mortgages, since it would be impossible to find investors to take the other 51.15% of the deal. And of course investors would demand even more overcollateralization with new mortgages than they do with seasoned ones.
The story of Springleaf, then is ultimately a depressing one. It shows that people still care about triple-A ratings when they shouldn’t; it shows that mortgage finance is not going to move to the private sector any time soon; and it shows that the paradigm of ratings agencies needing to ratify structured products and give them a triple-A rating is alive and well. Whether you think this particular rating is justified or not, there’s no good news in its existence.