If you thought that the rating agencies had cleaned up their act in the wake of the crisis, think again. Our Richard Smith reported on a couple of black eyes by Moody’s, one a rather implausible 180 degree turn on its take on the US tax deal, the other a suspiciously flattering take on whether Countrywide had indeed transferred notes (retaining them, as an executive testified they did on a routine basis, would confirm our suspicions about widespread problems in the securitization industry.
Now we have a big blooper by S&P, this one in the form of mass rerating, based on an admitted faulty analysis. That is code for “big error in the model that everyone missed.”
The subject of this screwup is 129 so-called re-remics, consisting of about $85 billion of bonds which were devised by repackaging existing residential mortgage backed securities. Never heard of re-remics? There’s a good reason why. This market has been very active since early last year, but third party purchases were limited. Given the scarcity of third party buyers, which means the deals were largely retained by banks, one has to assume that the object of re-remics was to manipulate capital levels. Just like CDOs, which are now understood to have been over-rated, re-remics achieve the seemingly impossible task of increasing ratings of junky RMBS. From an October Bloomberg report:
Bank of America Corp., seeking to reduce risk and meet new capital standards, upgraded billions of dollars of distressed mortgage bonds by repackaging them into new securities using a variation of a Wall Street technique that failed during the credit crisis.
The transactions, known as re-remics, are designed to add a layer of protection to residential mortgage-backed securities that sustained losses, enabling them to regain investment-grade ratings. The strategy helped the bank pare its RMBS holdings by $5.2 billion in the second quarter, or about 15 percent, according to a company filing.
The most questionable part of this exercise is that 308 re-remic tranches had been downgraded by S&P from AAA to CCC, which is junk, back in May. Earth to base, any instrument that has so much embedded leverage that it can be downgraded that dramatically in the absence of accounting fraud means it should never have been rated AAA in the first place.
One has assume that the reason they were so popular is that the issuers and banks knew they were rated incorrectly.
That of course means the banks will be hoist on their own petard as downgrades proceed apace. There may not be enough re-remic paper at any one bank for these downgrades to make a meaningful dent in their balance sheets. Nevertheless, this is an indicator that even with asset-value-flattering super-low interest rated, banks still carry more dreck on their balance sheets than most investors realize.
At least one source for the Bloomberg story on S&P’s latest pratfall was suitably pointed:
“An admission of guilt by a rating agency: How refreshing, and also what a wonderful Christmas-time present,” said Sylvain Raynes, a principal at R&R Consulting in New York and co-author of “Elements of Structured Finance,” published in May by Oxford University Press. “What I want to know is, is anyone going to get fired over this?”








I do structuring. I take S&P at its word that this was a mistake. But considering what they were rating(I think of these as more CDO’s than Re-Remics) the mistake was also an example of sloppiness and not reviewing the waterfall of their models.
These structures were all about Time tranching a questionable bond and asking how many of the flows were AAA. S&P clearly never reviewed their model and relied on an old, outdated and incorrect waterfall.
At the end of the day, we learned once again, that the rating agencies don’t bring value to the table. Incidentally a fair number of these types of transactions one, only a year ago or so have been downgraded.
Oh one other point, and this may not be obvious to all your readers. These transactions were pure and simple regulatory arbitrage for a large bank who had non-investment mbs paper that no longer qualified for zero risk weighting. All this
badly done engineering had one purpose, to salvage some amount of paper as being zero risk for regulators, who clearly blessed this non-sense or were alseep at the wheel(again).