Bond Insurer Transparency; Open Source Research

If anyone has not read Bill Ackman’s letter regarding MBIA & AMBAC, it is a good read. http://ftalphaville.ft.com/blog/2008/01/31/10612/monoline-flatline-bill-ackmans-letter-to-the-insurance-regulators-in-full-and-his-models/ Basically it is about estimating the bond insurers exposure and potential losses due to the numerous CDO’s they have insured. If, as Ackman claims, the credit rating agencies do not have the same level of detail as his model, why have many ratings given by the agencies been taken at face value. Their long standing history? If a well respected agency is providing grades or analysis on a situation, they should have the most accurate and complete information available to them. Right? Who is there to check the rating agencies and tell them they are lacking information or their assumptions are wrong? Personally, I like the idea of open source information and not just taking the word of the credit agencies. I understand the necessity of the agencies, but at the cost of their ratings not being accurate and thus beneficial?

The agencies were pretty up front about their models (in a finance-relative kinda way). They even knew they were very limited.

your outputs are only as good as your inputs…there were no inputs for CDOs.

I guess this is fairly obvious in hindsight, but if you have limited models and poor inputs then how can you think your models have any realistic accuracy.

Bond Insurers dont get compensated by ongoing performance, they get compensated up front. There’s no realistic way to compensate them on a continuing basis. Just as investors got lazy, insurers got lazy too. No one really thought 2nd liens were going to run at 20% CDR (20% annualized default / year). Cum losses for subprime are running at a rate 1.5 - 2x estimates…many estimates were extrapolations off of a very, very small subset of performance data from very different conditions in history…investors are upset that monolines / ratings agencies weren’t quick enough to realize what was going on and adjust their ratings, but what incentives did they have? What’s unbelievable is how inferior insured structures were to uninsured structures. Ratings agencies basically went off of AAA ratings from insurers and didn’t care what structures the AAAs came in. Investor lazyness contributed to a lot of the problem.

“ratings agencies weren’t quick enough to realize what was going on and adjust their ratings, but what incentives did they have?”…incentives? Hmmm, how about their reputation?

Most times, money is a better incentive than reputation. If they would have downgraded deals within the first month or two after rating them, they would have lost out on a ton of new business…reputational risk is apparent, but I’d say it’s pretty low down the totem pole. I don’t understand why people would think that a lowly analyst making 60-70k couldn’t and wouldn’t get outsmarted by CDO managers that were making deep 7 figures. I’m not really sure why people thought that the ratings agencies were the end all, be all of knowledge. I don’t see many money managers and real money people knocking on their doors for a job…

Did you think ratings agencies were the “end all, be all of knowledge”? There’s discussions on AF going back several years about how the rating agencies assessments of these things was largely irrelevant to the risk associated with them. If you suggested at any of the risk conferences I went to that “the ratings agencies say that [blah], so …” everyone would have scoffed at you. Even the market pricing of most of these suggested uncertainty with rating agencies; we haven’t heard yet of any arb hedge funds blowing up from say, shorting agency debt and buying senior CDO debt to “arb” the yield differential because the ratings agencies say they are the same thing. I don’t think ratings agencies are the villains here. I think the villains are all the people who thought that since they were selling off the loans, they didn’t need to do good due diligence.

Here’s something I can’t quite judge if its wrong behavior or not. Goldman made a lot of money in fees structuring BAD CDO’s. Then with another department they go ahead and short the type of assets backing these CDO’s making a killing. I know they haven’t legally done anything wrong…but come on. Its like no one cares about reputation risk as everyone’s reputation is sh!t. As for the rating agencies…what realistically could they have done. Sure they should have been a little more pessimistic in their ratings but would that have changed much in what’s going on now?

Not really wrong…many times different trading desks have different calls…why should that be wrong? Point blank, the deals would not have gotten done if they weren’t AAA.

^ actually that’s sort-of a decent argument that the rating agencies did something wrong. They had this excessive level of trust and they should have said something like “this class of securities contains risk that does not allow them to be part of our traditional taxonomy so we are going to rate them rekative to others in the class and let you decide if the asset class is appropriate”. Instead people saw AAA and thought it was US govt debt when it was really home equity loans on house trailers (I just made that last part up, I think).

I don’t get the need for rating agencies? The way I see it the yield on a bond should determine its rating - the rating shouldn’t determine the yield.

JoeyDVivre Wrote: ------------------------------------------------------- > Instead people saw AAA and thought it was US govt > debt when it was really home equity loans on house > trailers (I just made that last part up, I think). I agree completely. But I think instead of house trailers, they were loans to unemployed individuals that became SIV’s. (http://www.youtube.com/watch?v=SJ_qK4g6ntM)

TJR Wrote: ------------------------------------------------------- > I don’t get the need for rating agencies? The way > I see it the yield on a bond should determine its > rating - the rating shouldn’t determine the yield. Well if the rating agencies keep behaving like this, that’s what you’ll get. In the meantime, credit ratings help solve agency problems by allowing investors to specify default risk, separate credit risk from other aspects of pricing, use the threat of ratings downgrades from allowing managers to exploit bondholders for the benefit of stockholders, and provide credibility to the bond market for investors who trust shoe-leather due diligence more than efficient markets.

Many pension mandates also specify a minimum level of AAA assets. Regulatory risk capital applies different discount rates to AAA assets against even AA. It absolutely matters that agencies rate appropriately.

There is no way for agencies to rate appropriately if the underlying information is not available. That is why I do not understand why the agencies could not have done something similar to what JDV mentioned (special rating class or similar). They should have avoided making a definitive rating on assets where they acknowledge that full information was not available. Apparently it was too much work to change their rating structure and instead decided to give a rating that is essentially worthless (in hindsight of course).