Home > finance > Reputational risk is insufficient for ratings agencies

Reputational risk is insufficient for ratings agencies

April 18, 2012

I’ve had a few conversations recently with intelligent, informed people about the failure of the ratings agencies during the housing bubble to keep up standards on their ratings of CDO‘s. You can discuss all day whether it was individual ratings they got wrong (at the level of the MBS‘s) or whether it was the correlation of defaults they were underestimating. It was both. But in the end the fact is they sold AAA ratings.

Nobody really argues against that. What fascinates me, though, is that people sometimes still argue against the idea that the revenue model of ratings agencies, whereby the issuers of debt pay the ratings agencies for ratings, is fundamentally flawed.

Their explanation is something like this.

That system worked fine for a long time, because for a given rating they wouldn’t sacrifice their reputation on a ridiculous rating for some small-fry issuer. And the system would have continued to work fine except that the issuers became huge and the amount of money involved became too tempting and so they ended up whoring themselves. But there’s nothing fundamentally wrong with the incentive system, we just need to keep reputational risk the driving force.

What?? That’s argument kind of reminds me of the so-called dental insurance which pays for cleanings but when it comes to dental emergencies with root canals and surgeries you’re shit out of luck. That’s not insurance at all, in other words.

I see the need for ratings agencies – it’s a way of crowd sourcing due diligence, which makes sense, but only if we can trust the ratings agencies as an impartial third party. And I don’t want to seem like someone who doesn’t have faith in humanity, but my trust isn’t won by a system of perverse incentives that has already failed. Let’s just say I have hope for humanity but I also acknowledge our weaknesses.

And just to be clear, the new bond rating agency Kroll, which has been getting a lot of attention, also uses an issuer-pays revenue model. But I guess Egan-Jones doesn’t, it uses a subscription-based revenue model. I still prefer the concept of an open source ratings agency - I’ve been in touch with Marc Joffe, who is doing just that for sovereign debt, which I will talk more about in a separate post.

Categories: finance
  1. lawrence castiglione
    April 18, 2012 at 10:17 am | #1

    Profit provides the motivation for rationalization. There are endless good reasons to give the client what the client wants.

  2. Mike Maltz
    April 18, 2012 at 10:56 am | #2

    What’s wrong with CalPers and other pension (and mutual fund) organizations paying for the ratings? They’re the ones that take a beating when the ratings are falsified. [But would the raters then be able to get the necessary information from the companies they are rating?]

  3. Ray
    April 18, 2012 at 1:26 pm | #3

    I don’t profess to be an expert on this – to the contrary, I’m not even confident I fully understand what these securities even are. So to me, this also raises some interesting questions about what the due diligence process even is for a mortgage-backed security. Do we actually have analysts from Moody’s scratching at the drywall of some McMansion in Florida trying to figure out whether it’s worth $300,000? Does S&P ask John Q. Patsy’s boss for last year’s paycheck stubs and whether there are any layoffs planned in the next 10-30 years? If not, it seems like the ratings agencies would have to rely on information provided by upstream parties, which in essence doesn’t seem that different from asking the upstream parties to police themselves.

    I guess I just don’t necessarily see any need to posit bad faith on the part of the ratings agency when the result is just as easily explained by bad faith – or even “mere” negligence – on the part of the mortgage lender or other upstream parties. Of course, this doesn’t call into question the very well-founded underlying point you’re making about the comically obvious conflicts of interest that arise out of the issuer-pays model.

  4. April 26, 2012 at 11:27 am | #4

    Hi Cathy

    Among certain parties I’ve dealt with since the subprime blowup, there is definitely a move toward more in-house research to supplement, or in certain cases, replace dependency on ratings agencies. The subprime debacle wasn’t the first time the agencies have gotten it wrong and at least for those who are managing enough money and to an extent more active than simply “setting it and forgetting it”, the consensus is that if you want proper due diligence, you have to do it yourself. And if you don’t have the manpower to do the due diligence and you lack the funds to hire that manpower, then you shouldn’t be in that market to begin with. In short, there’s more than enough evidence out there to suggest that ratings agencies are more reactive than predictive, and any investor thinking clearly is acting accordingly.

    That’s only as far as sophisticated investors are concerned though. When it comes to the big pension managers and other such quasi-public sector counterparties, I really don’t know what the best answer is. Taking the sophisticated investor approach to due diligence is kind of outside the scope of their mandates but they clearly need something beyond taking the ratings agencies at their word for the very conflicts of interest you mention that by now are widely understood, no matter how much ratings agencies try to distract attention away from them.

    As far as this open source ratings idea, this is the first I’ve heard of it, so thank you for posting. I don’t anything about it beyond what’s available from the links you posted here but I can already see that this methodology is going to have serious limitations in the sovereign credit space–namely, that once you get out of North America, Western Europe and Japan, its utility in determining anything of value is going to diminish dramatically for a variety of reasons.

    Nevertheless, I’m still interested in where this open source ratings movement goes next, so if you see or find out more on that, please keep me posted.

    Thanks and regards

    • April 26, 2012 at 11:32 am | #5

      Thanks! I will hopefully be contacting you soon.


      • Pat
        May 4, 2012 at 10:48 am | #6

        Hi Cathy!

        I have watched your interviews on the PBS and I believe you want to do good. I don’t want to discourage you neither, but I have a question to ask you.
        Do you realize what exactly are you against?
        Do you realize that your group is trying to tell the goverment to crack down on those that feed [lobby] them money? The goverment knows there’s a shady business going on inside the banks, but since “Money Talks”, they leave them alone and even help them to get more. They all have a shady business going on that’s why they’re slow to criticize them in the public.

  5. May 4, 2012 at 12:53 pm | #8

    The idea of open source ratings got a huge boost this week when the Financial Times Alphaville blog reported on our sovereign/municipal rating tool. See http://ftalphaville.ft.com/blog/2012/05/02/983041/monte-carlo-simulated-sovereign-credit/ for their coverage.

    Open source rating methodologies should be linked to the idea of forming a not-for-profit rating agency. As I will argue in a forthcoming opinion piece, Consumers Union is a great template for what a rating agency can and should be. A very detailed proposal for a not-for-profit rating agency was published last month by the Bertelsmann Foundation. See http://www.bfna.org/publication/blueprint-for-incra-an-international-non-profit-credit-rating-agency. I don’t agree with much of their implementation, but it is good that they have put the idea out there.

    I actually believe the reputational argument. It broke down in structured finance because each underwriter (Goldman, Morgan Stanley, Lehmann, etc.) accounted for a large share of the rating agency’s revenue. When that happens the demands of a single client can outweigh concerns over loss of reputation, especially when you take discounting into consideration, i.e. if you sell your soul for an issuer you get money today and embarrassed tomorrow.

    With respect to Ulysses point, I agree that our solution is best suited to “advanced economy” governments. There are models better suited to emerging nations. Kamakura has one. An additional open source tool that implements a multiple regression on a few indicators may be the best option for these countries.

    • May 12, 2012 at 5:53 pm | #9

      Marc, don’t you think data reliability issues would swamp even a simple model in certain countries?

      • May 14, 2012 at 12:11 am | #10

        The model is directed at “advanced economy” sovereigns and sub-sovereigns (e.g., US states, Canadian provinces, UK, France, Germany, etc,) that are unlikely to be cooking their books. US states and Canadian provinces, for example, provide audited financials each year.

        Cathy, I hope that you and others on this board realize that this is the real thing. Up until now, there has been a lot of talk about open source credit rating models but little in the way of tangible results. In 2009, freerisk,com ceased operations before publishing a model. Open Models Valuation Company has made substantial noise under the auspices of Macrowikiomics author Don Tapscott, but has not produced anything.

        A blogger at Wall Street Daily totally got it when he wrote “Are We Witnessing a Ratings Revolution?” at http://www.wallstreetdaily.com/2012/05/08/are-we-witnessing-the-start-of-a-ratings-revolution-psc/. He properly connected PSCF directly to the Nate Silver article that you guys discussed last August.

        I know some people would have preferred to see a corporate or structured open source model first. But, if we understand open source to be a solution to socially important problems, we should agree that government ratings is the first place to go. These ratings affect the level of services a government will be able to provide and they can even determine whether a country will face civil disorder, so they are a lot more important than other types of ratings.

  6. May 7, 2012 at 12:55 am | #11

    fyi Global Finance has a cover story about politics and the ratings agencies:


  7. May 12, 2012 at 5:47 pm | #12

    I see the parallel, but the counterargument actually sounds fine. Not all threshold arguments are wrong.

  8. May 15, 2012 at 5:29 am | #13

    and the slaughter continues…Moody’s just cut ratings for 26 italian banks to four steps above junk. What I wonder is: how long will it take Moody’s to downgrade Italian banks to junk?

Comments are closed.

Get every new post delivered to your Inbox.

Join 886 other followers

%d bloggers like this: