OT…. and FYI…

http://www.nytimes.com/2009/11/20/business/global/20rating.html?ref=business

Moodys former compliance officer testified to the House Oversight and Reform Committee last month that Moodys did no surveillance of its muni ratings. They issued them and then left them static although they continued to bill the issuer.

This means that retail investors who were sold those securities had no specific or general sense of the accuracy of the rating or the pricing that derived from the rating.

I’m sure that once the ratings are available for analysis we will see lots of market and academic scrutiny of ratings accuracy. Long overdue.

The SEC is requiring the raters to expose the histories in XBRL for more ready comparison. Here is a presentation that I gave at a conference at the FDIC last month on the topic:

http://shopyield.com/2009/10/06/xbrl-and-credit-ratings/

You’ve asked several excellent questions. And you hit the nail on the head regarding some of our strongest issues with the current ratings system. The current ratings system has no way of representing the magnitude of losses on a security. The ratings provided by the rating agencies reflect the likelihood of a bond not paying 100% of principal, but make no representation of how much principal they are projecting will be lost once that bond is determined not to pay back 100% of principal. For example: Are you going to lose all of your money or just 1%? The current system does not tell us this. This becomes a serious issue when dealing with multiple-obligor securities. The example you provided refers to a multi-obligor security (in this case a Mortgage-Backed Security) which is comprised of a large amount of excellent borrowers who are still paying their mortgages as contracted. Unfortunately, in your example, the rating on the bond would be downgraded to CCC or lower, even if all 900 paying borrowers continued to pay their mortgages until full payment is made.

For an individual investor, the rating may not be prohibitive. However, for banks that own this security, once it’s been downgraded to below investment grade (and possibly purchased originally with a AAA rating), the entire balance of their investment will be classified as substandard. You read that right. Even the “performing” portion of the security will be classified since the entire security has been rated CCC (or worse).

As you might suspect, this in turn creates a much larger problem in the securitization market since many banks will not want to continue to own these securities after they have been downgraded since they will become classified as substandard (and now perceived as “toxic”). And certainly because of the substandard classification issue, other banks are extremely hesitant to buy more of these securities. So what we are now left with is a huge decrease in demand for these securities, and a lot of banks being forced to sell these securities because of downgrades (increasing supply in the market). This, in a nutshell, is why prices have plummeted for mortgage securities and the secondary mortgage market is no longer functioning as it used to.

I hope that helps. Your example and questions touch on the very reason we feel so strongly about the need to revamp the current ratings system. This must happen before we can realistically expect to get the secondary mortgage market moving again, which in turn must happen before our economy truly recovers.

We DO need to protect the FDIC fund. One way is to re-impose Glass-Stiegel. I prefer a sliding capital scale:
If you are a $100mm bank, 8% capital.
If you want to be a $500mm bank, 10% capital.
If you are a $10B bank, 20% capital.
If you want to be a $1T bank, 50% capital.

Capital requirements should increase with risk to the insurance fund. But use some sliding scale that makes banks build reserves in the good times, and let them pick how big they want to be vs. the capital requirement.

However. Securitization didnt cause the “credit crisis”; leverage did. We should not allow any securitized asset from getting a government guarantee unless it meets strict (80/20, prooof of income) lending standards and the Trust agrees to electronic full disclosure at least annually (like munis).

Also restrict deposit taking banks from proprietary trading…

Securitization is the tough one…

http://freerisk.org/wiki/index.php/Securitization

Government Ratings would be susceptible to political / social goal setting.

ie: Is it “fair ” for Detroit to pay more to borrow than DesMoines?

The conflict in interests arose because the raters had no stake (beyond reputation) in the quality of their ratings. Obviously, reputation matters – it accounted for the high credibility that the agencies enjoyed before they went south. But the late unpleasantness has put people in a state of mind where the actual dollar incentives of facing actual human beings will be given the greatest weight by the market.

My UL model is not a “user pays” approach. Issuers pay for credit enhancement. I’m merely suggesting that the rater have a financial stake in investment results, and, of course, that the people who make the call get paid on those results and not on issuer-paid revenues.

All of these changes will be judged by the bond markets, which will believe the ratings or won’t. The spread between corporate and Treasury rates will measure the confidence that the raters have earned.