In Depth Analysis: CalculatedRisk Newsletter on Real Estate (Ad Free) Read it here.

Saturday, July 07, 2007

It's Not the Default, It's the Deleveraging

by Tanta on 7/07/2007 01:38:00 PM

More proof that Australians are smarter than us Americans: here's an article from the Sydney Morning Herald on CDOs that is not stupid. I found the discussion of credit spreads a bit foggy, it is true, but since the issue is either absent or entirely mangled in every other mainstream media piece I've seen on the subject, I can accept foggy. In any case, I recommend the whole thing if you're new to the leveraging issue.

Ibbetson says there are three parties who need to shoulder blame for the current imbroglio, and all are equally culpable: Bear Stearns for leveraging the fund so highly, banks for lending the money to enable the leverage, and investors for buying into an investment that they really should have known could be exceptionally risky.

I guess they haven't figured out down in Oz that it's all the rating agencies' fault. John Mauldin has some ideas on that in this week's newsletter:
When a corporation gets a rating there are audits, not to mention regulators that are there overlooking the data upon which the ratings are based. But no one was looking at the data used to create the ratings on RMBSs and CDOs, to make sure there was some type of reasonable similarity or standard of the securities being rated and the databases used to do the risk analysis.

Subprime loans made in 1999 or 2002 were significantly unlike those made in 2004-2006. At the end of 2006, many subprime loans were defaulting in only one or two months from the date of the loan. No-documentation "liar's" loans were common. Adjustable Rate Mortgage (ARM) loans, made where the applicant could clearly not make the payments when the interest rate reset, were common. Thus, the past performance the ratings were based on was significantly different than for the crop of then-current loans, and was substantially misleading. We are talking an apples and cumquats type of difference here.

So, the rating process was not the same as the ratings that were used in the corporate world. But the problem is that the ratings used the same designations. Instead of creating a whole new type of rating standard (say, using numbers like "CDO rank 1-10"), they used the same designations that bond investors were used to.

I think it is disingenuous for a rating agency to explain the difference in paragraphs 457-503 in 7-point type and dense legalese in their disclosure document. Investors had (and should have) a certain level of expectation when the designation "AAA" is used. GE and Exxon types of expectations.

I am not expecting infallibility here. Let's make it clear that the rating agencies have made mistakes in the past and will make them in the future. You do your best, and in general I think they do an excellent job given the pressures and the vagaries of the business. (I certainly have made a few mistakes here and there in my career that I would not make today. And I will make more in the future. We live and learn.)

The problem is in allowing the confusion of rating a CDO as you would a GE or Exxon. I think Dann has a point when he says the rating agencies aided and abetted the investment banks. And that point gets even bigger when we are talking about CDOs.

When you pool BBB tranches into a CDO and now turn 80% plus into AAA at the touch of an algorithm, based on faulty assumptions, someone somewhere should have raised an eyebrow.

This is not going to end up pretty. You can bet that 20-20 hindsight is going to kick in here as the regulators and various attorneys general get involved. The rating agencies may be able to justifiably say that they were doing exactly what they said they were doing in the disclosure documents. But then someone at the investment banks (especially those that owned subprime mortgage brokers and should have been able to see the deteriorating quality of the loans) should maybe have thought that the default rates would change and therefore should have used different assumptions.

But then, that would have killed what was a very profitable business. And everyone was doing it, so to unilaterally disarm when every other investment bank and agency was doing the same thing evidently did not cross the mind. Last year there were $500 billion in CDOs sold, and half of it subprime. In June, there was only $3 billion. And you can bet there was no subprime in them.

As an aside, the rating agencies are starting to downgrade the CDOs. Of the pool of securities created from 2006 subprime mortgages, Moody's has downgraded 19% of the issues they've rated and put 30% on a watch list. That will grow.

And let's look at the investment banks. Creating and selling CDOs was a particularly juicy business. I have heard, but not verified, that sales commissions were running 5%. You can bet the banks were making at least as much. Put together a $250 million CDO and sell it to institutions, pension funds, insurance companies, and hedge funds, put some of the equity portion into your own portfolio, and you could generate substantial profits and commissions. Rinse. Lather. Repeat.