Mike raised an excellent and common point in the comment section on the last post.
He objected to my “free pass”, so to speak, for the rating agencies on the MBS mess. He felt, as many do, that they had an obligation to identify the bubble, and warn the public.
What Mike and others are demanding is that S+P abandon their tradition of using statistics and balance sheets, and base their ratings on subjective opinion instead. In a sense, that’s what they did by lowering the long-term rating for the United States’ debt to AA+. As they said in their commentary when they issued that rating downgrade, they do not doubt the ability to pay, just the willingness.
With a bubble like the housing market experienced, the problem was more one of future ability to pay, especially for the class of loans called “alt-A”. Those loans to people with FICO scores north of 700 (definitely NOT subprime) have turned out to be worse than subprime on hundreds of deals. Those alt-A deal losses were initially modeled to be less than a quarter the severity and frequency of subprimes.
That said, my response to Mike’s comment follows the break.
Mike spoke for many of us, when he said:
But you are just wrong about S+P not recognizing a bubble. The Holland Tulip Bubble looked like high quality investments, the South Sea Land Bubble similarly. And the High Tech bubble. The Mortgage Derivatives debacle was not the first of its kind. I do not believe that S+P were not permitted to see outside the bubble to see that it was in fact not an expanding universe, but a bubble near bursting. Around 2002 I had information indicating that the mortgage derivative market was a scam, and I do not believe that I alone had that insight. The company that had responsibility to rate the bonds certainly knew everything I did, and ignored it all. S+P made a lot of money selling (lies) good ratings, and the hypothetical honest guy who blew the whistle would have been fired or killed. Maybe there is a pile of bones in a closet at S+P, of honest people who were going to blow the whistle.
I’m sure I could have been clearer, or more emphatic, when I didn’t lay into S+P (nice choice to use plus sign instead of ampersand) as much as you thought I should for the failure of their AAA MBS ratings.
They certainly failed to do a number of things they should have, including
- Recognizing the widespread fraud they were “rating” — though this is tough because they are essentially hamstrung, forced to assume that the data is correct and true that they get from the underwriters and bond issuers (the same people from the Wall Street firms controlled both during the bubble era).
- Disregarding their own standard for historical data and statistical proof when they created and pushed the “technology” of bundling a hundred BBB MBS bonds into a single CDO deal, because they assumed those bonds wouldn’t be very well correlated, the way a hundred BBB corporate bonds would be relatively uncorrelated.
- Monitoring MBS deals for delinquency and signs that defaults were going to exceed their projections, and following with early and repeated downgrades of bonds that statistically warranted high initial ratings.
Still, I take exception with the Monday morning quarterbacks like Mike or Congressman Frank when they assert that S+P should not have given those MBS bonds AAA ratings in the first place.
Before anyone gets crazy on me, let me say it categorically: S+P should not have issued the ratings it did for CDO (Collateralized Debt Obligation) deals that consisted primarily of subprime MBS bonds as underlying collateral.
As I’ve said before, the performance expectations for subprime mortgages were established by decades of experience and millions of mortgages. Ironically, many of those mortgages were the CRA loans the truly dishonest (or terminally ignorant) love to blame for the meltdown. They performed so well that a pool of “typical” subprime loans that looked like CRA loans on paper could be reasonably expected to produce losses over time that amounted to 5% or maybe 10% of the original pool balance.
The MBS deal structures had hundreds of basis points of excess interest available each year to enhance the credit support, so until 2006 it was extremely rare for even BBB bonds in MBS deals to experience payment shortfalls. In fact, it was more likely for a subprime BBB bond to be upgraded to AAA than to have it downgraded to even top-rated “junk”, or BB+.
Those were the facts. The data. The statistics.
S+P offers statistical ratings. To do otherwise, to offer opinions not supported by the data, would have been a serious deviation from their stated mission, and a misrepresentation of what they were selling.
As a user of their ratings, I would have had a real problem if they simply threw out the history, and offered opinions instead.
I would have had no problem if they had offered political opinions or economic projections and labeled them as such. I would have had a big problem if they simply changed the rules midstream.
What about bubbles?
I bought a hundred Krugerands in the 90’s for around $30K. They would have fetched $180K last week (if I still had them). Am I going to say the price of gold is in a bubble? No. But it’s gone up more that three times as much as US housing went up, and many observers today say that was a “bubble”.
Hmmm. There are whole countries whose housing costs more, as a ratio to household income, than America’s housing cost at the very peak. Does that mean all those countries are in a “super bubble” in housing? I think not.
Every market is different, and the forces that pushed housing prices down in the
US weren’t simply reactions to fraud, or high prices, or easy credit.
I’m going to play a bit of a teaser here. To see all the moving parts, and how they came together to create the mortgage debt disaster, you’ll have to buy my book.