Arnold Kling  

A Rating Agency Example

Lectures on Macroeconomics, No... Too Much Deregulation?...

Gretchen Morgenson writes,

The safest slice of the security held $165 million in loans. When it was issued on Aug. 17, 2006, Moody's and S.& P. rated it triple-A. Just eight months later, Moody's alerted investors that it might downgrade the top-rated tranche. Sure enough, it dropped the rating to Baa, the lowest investment-grade level, on Aug. 16, 2007.

Then, on Dec. 4, 2007, Moody's downgraded the tranche to a "junk" rating. On April 15 of this year, Moody's downgraded the tranche yet again; today, it no longer trades. The combination of downgrades and defaults hammered the securities.

One question that I would like to see asked at next Tuesday's hearing on Freddie Mac and Fannie Mae is what those two firms were thinking while the rating-agency nonsense was going on. I imagine that they ran some of these securities through their stress tests, the way any manufacturing company would test a competitor's products. Surely, they saw that these securities would blow up if house prices started falling. Why did they not call baloney sandwich?

Comments and Sharing

COMMENTS (14 to date)
Craig Bardo writes:

Why didn't the fox alert the farmer that the entry to the chicken coop was unguarded?

The more compelling question to me is why do the people who looked the other way when the OFHEO regulator told the House and Senate panels that there were problems, get to ask questions now? More simply, why aren't Chris Dodd and Barney Frank wearing orange jump suits?

Greg writes:

I would love to hear that. Unfortunately, I don't have a lot of faith in Congress's investigational skills, particularly in the form of public hearings. They seem more interested in hearing themselves talk.

Unit writes:

Is it true that the rating agencies had a semi-monopoly guaranteed by the regulations?

Jeffrey Friedman writes:

It is true. They had a legal oligopoly, instituted in 1970, in response to the "Penn Central crisis."

A quick scan of the NYT article today indicates that it glides right past that fact, but it is only the legally created oligopoly--and the legal requirement that institutional investors invest in AAA-rated securities from one of the oligopolists--that allowed S&P, Moody's, and Fitch to begin *charging issuers* for their ratings "in the 1970s," as the NYT mysteriously puts it.

For free-marketeers, I think this may be the missing piece of the puzzle of this so-called crisis of capitalism. Apart from the possible role of the CRA, Fannie, and Freddie in mandating and buying them, why did private investors buy subprime mortgages?

Answer: they were misled by previously gold-standard ratings from credit agencies that either knew they were underestimating the risk, but had the incentive to do so for the sake of their customers; or did not know, because they used overoptimistic risk models, and were protected from competition from rating agencies that might have arisen to blow the whistle by using more conservative models (perhaps Arnold Kling might have started such an agency, if he hadn't been legally barred from doing so...?).

The key point: the investors were ignorant of the legal oligopoly, or of its implications, just as the NYT reporter seems to be.

Jeffrey Friedman writes:

Correction: The Penn Central crisis was in 1970; but the SEC's response was not until 1975. At that point the SEC required brokers to count only investment-grade bonds, as designated by a "Nationally Recognized Statistical Rating Organization," as part of their minimum legally required capital--and named S&P, Moody's, and Fitch as the only three such Organizations. Other regulatory agencies followed by requiring high NRSRO ratings for other institutions' holdings.

Bill Woolsey writes:

I think that the failure of the ratings agencies is important.

The "shadow" banking industry held little capital (was highly leveraged) and it was very much based upon what the ratings agencies said about securitized loans.

For example, money market mutual funds were invested in asset backed commerical paper. This was all based upon high ratings for the underling securitized loans.

When they were proven wrong, the demand for securitized loans fell, and people fled to the conventional banking system.

That the SEC approved only certain ratings agencies, does suggest that the government was making sure that the shadow banking system was safe.

On the other hand, I don't believe that it would be illegal for a competing, more-conservative ratings agency to sell its services to investors.

"The "big three" who are being paid by the issuers to certify their securities are giving high ratings to risky securities. Yes, the SEC allows various financial institutions to buy high risk, but "highly rated" securities. But should you buy one? And, should you lend to a financial institution that is heavily invested in them? If you want to keep you money safe, buy our Acme
ratings book. Find out what securities really are safe."

It would even be possible for issuers to pay for
this addional ratings-- but only if many investors used these ratings. Such an issuer would, of course, have to pay one of the approved three to get government approval. But to get
skittish investors to buy them, they would have to buy the services of our more conservative agency.

But, the investors must wonder if the more conservative agency is "right" or too alarmist.
Of course, we can trust the SEC...

stanfo writes:

Under the new Basel II, international banks have the options of determining their own risks and ratings on assets, as a method to determine their own appropriate level of risk-weighted capital.

This introduces some competition into the ratings game, and allow the industry to develop ratings "best practices". Surely this will be important going forward...

Les writes:

I don't dispute that the ratings firms were sloppy or tardy or worse in some cases.

But it seems naive to rely on their ratings when it was plain that:

1) They were a three-firm oligopoly,
2) They were paid by the issuers of the bonds they rated and therefore were hardly objective,
3) Lowered ratings will damage the credit and reputation of the particular issuers - so ratings firms can logically be expected to be slow to lower ratings.

So it seems to me that a "caveat emptor" approach would clearly be indicated.

Andrew Garland writes:

The Moody's Blues - Credit agency negligence

A post about a WSJ article criticising the ratings agencies. An eye opener to how cavalier and stupid the ratings agencies were about their business and responsibilities.

Lord writes:

Overreliance on national historical house prices never declining made the decline a self-fulfilling result.

Jeffrey Friedman writes:

I think Les's point may be easy to deal with. He is assuming that everybody knew in, say, 2004 what he now knows. My guess is that very few retail investors knew, or even now know, the legal status of the Big 3. (They certainly won't find out from reading the front page of today's NYT.) As for institutional investors, the NYT article, for all its faults, has emails from Pimco's, Vanguard, BlackRock, and Fortis screaming at Moody's in July 07 for having misled them and everyone else about the safety of the MBSs. Vanguard explicitly says that they used to rely on the Big 3's ratings but won't do so in the future if they can avoid it.

Bill's point is more difficult to deal with. Maybe it's fatal to my hypothesis. Here's my initial shot at it:

Prior to 1975, the advent of the Xerox machine had changed the economics of the ratings biz. Investors could now Xerox the information in the ratings books without subscribing to the services that produced the books. So selling ratings to investors was out; selling ratings to the issuers became the new model.

So here's my reformulation of Bill's point: Once the new model required issuers to pay ratings agencies, why would the issuers of bum securities pay upstart, stringent Kling's to rate them when established, lax Moody's would already do so? Kling's would have to see specific information about a given MBS in order to rate it; why should the issuer of a bad MBS release this information to stringent Kling's (let alone pay Kling's for the privilege) if it could get an AAA rating from Moody's instead? So how would Kling's even get off the ground?

I think the answer is that absent the post-1974 regulations, Moody's could maintain this competitive advantage over potential upstarts only by preserving its reputation among investors--by not misleading them (intentionally or otherwise).
Even after Xerox machines forced the Big 3 to start selling ratings to issuers rather than investors, the ultimate market for these ratings was still investors: If the Big 3 screwed up, investors would not trust the Big 3 any more, so issuers would have no reason to pay the Big 3 to rate their bonds. Once that happened, the issuers would have every reason to solicit ratings from Kling's; or perhaps it would have to be Buffett's--someone with an established reputation who could step in to fill Moody's shoes.

After 1974, however, bond issuers were forced to pay the Big 3 *regardless* of the Big 3's reputation with investors, inasmuch as investors were now legally barred from buying bonds that didn't have a high rating from one of the Big 3. So 34 years later, it doesn't really matter that investors are pissed off at the Big 3 for misleading them: a bond issuer still has to get a AAA rating from one of the Big 3, even if investors now come to view the rating as worthless. The Big 3's reputation doesn't matter to their income any more, so they can scew up with impunity.

Maybe in 2009, upstart agencies will take advantage of this situation and, being dependent on their future reputation, be likelier to prevent future screwups--that is, as likely as the Big 3 would have been, prior to 1975. Meanwhile, though, (arguably) the current screwup was enabled by the post-1974 regulations' removal of pressure against lax practices that might ruin the Big 3's reputations. The Big 3 will still get their ratings money even if their reputation is now ruined and upstarts induce issuers to pay them for more trustworthy ratings, because issuers will continue to be required by law to pay for a worthless Big 3 AAA rating even when selling a bond that also has an upstart's trustworthy AAA rating.


I'm only a political scientist, so help me out.

Ned writes:

"Lord" makes an important point (although the phrasing is awful :-)

If one looks at the history of nominal house prices (nominal prices are what matters for mortgagors' ability to repay their loans by selling the house), the prices have never declined on the nationwide basis, going back to the great depression. And, during the great depression, they declined not because of bursting of some housing bubble, but because of depression itself. Therefore, if you have to construct a probability distribution of HPA (House Price Appreciation) to project different future scenarios, based on the available data, you may as well assume that is it Log-Normal!

Yes, I know, everybody is smart now. But the thinking was that one characteristic of the housing market is that people avoid selling at a loss, so they take their houses off the market, which they are normally able to do, since the houses are neither perishable goods nor do they suffer from change in model/style/technology (unlike cars, for example).

My larger point is that the rating agencies, investment and commercial banks, and other actors, were behaving in a way that was close to rational ex-ante. One does not need to assume inordinate greed, stupidity, or malice to explain their behavior.

Les writes:

I'd like to thank Jeffrey Friedman for helping to prove my point that it was naive to rely on rating agencies.

First, he "guesses that very few retail investors knew, or even now know, the legal status of the Big 3." Unfortunately, guesses don't decide arguments. And there was widespread knowledge in the financial community 30 years ago about the ratings agencies, if not among political scientists. In any case, retail investors are dwarfed by institutional investors, so what retail investors knew or knew not, is of little relevance.

As for institutional investors, the fact that they relied upon ratings does not mean they relied exclusively upon ratings, to the exclusion of other information. Certainly institutional investors knew - or should have known - the hazards of relying upon ratings. And Vanguard's announced rejection of reliance upon ratings simply confirms my point that reliance upon ratings is unwarranted.

So thanks again, Mr. Friedman for helping to prove my point that it was naive to rely on rating agencies.

Lord writes:

Yes, that was awkward. Try: Belief in its impossibility made it inevitable.

Comments for this entry have been closed
Return to top