Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.
Display:
but at least two Nobel prizes have gone to individuals (Stiglitz and Arrow) involved in proving that objective credit analysis by lenders (or anyone else) is effectively a mathematical impossibility, which means we can't avoid, under any imaginable economic system -- not just markets or capitalism -- providing greater wealth and power to the few individuals that end up capable of compelling others in society of their credibility (and power) to determine the credibility of others.

Rating agencies are just the banker's attempt to solve the inescapable principal-agent problem given conditions of imperfect information (read: incentive to lie, even to yourself) by trying to establish an institutional wall between lenders and risk analysts by outsourcing to a pooled resource.  It makes a lot of sense and is an improvement over what existed before, but it's a problem that has no permanent institutional solution.

by santiago on Mon Apr 26th, 2010 at 12:05:48 PM EST
[ Parent ]
One of the solutions Krugman cites is for the SEC to hand out the rating jobs to the agencies rather than allowing the institutions pick their agencies.

I suppose you could imagine, in the long term, the SEC rating the rating agencies on the accuracy of their forecasts and removing certification from ones who failed to do a good.

by Colman (colman at eurotrib.com) on Mon Apr 26th, 2010 at 12:11:44 PM EST
[ Parent ]
It's a good solution for the present situation and one that I support, but preventing self-interest from becoming a part of risk and credibility analysis is, according to the compelling work of Stiglitz and Arrow, always going to be a whack-a-mole game. It's not possible to get to Krugman's "idiot proof" lending environment for more than a short time because power (compelling individuals to surrender their own interests to those of the group) and self-interest are fundamental variables in any lending process.

For example, how well has the SEC actually performed its less critical job responsibilities up to now?  So why would we expect it to do much better job with greater power and responsibility over rating agencies?

by santiago on Mon Apr 26th, 2010 at 12:25:37 PM EST
[ Parent ]
But it's all always whack-a-mole: the system has to constantly adapt to technological innovation. This is the current problem with a lot of our systems - we design institutions in 1710 or 1810 or 1910 or 2010 and we expect it to not need changing in pretty fundamental ways as the world changes.
by Colman (colman at eurotrib.com) on Mon Apr 26th, 2010 at 12:31:25 PM EST
[ Parent ]
technological innovation and unexpected or expected results of long-term iterations.

In this case, perhaps the tree of credit needs to be periodically watered with the blood of rating agencies.

by Colman (colman at eurotrib.com) on Mon Apr 26th, 2010 at 12:32:39 PM EST
[ Parent ]
Yes, adaptation is an inherent element, which is the main argument against Krugman's thesis of "idiot-proofing" the system instead of relying of skilled actors.  (I tend to support Krugman on this, however, but there is a valid counter-argument to his case.)
by santiago on Mon Apr 26th, 2010 at 01:01:50 PM EST
[ Parent ]
Also, Krugman's solution doesn't resolve the issue that most large securitized credit offerings like the big MBS deals that got us into trouble, ALREADY get rated by more than one, if not all, of the agencies, so it isn't picking the agency where a lot of the corruption occurs.  It's that there are costs and professional risks involved for any analyst and his or her employers, past, present, and future, in making a prediction that deviates too much from the common wisdom or expectations of the day. Look at the outrage and accusations of political manipulation that occurred when the agencies downgraded Irish debt, for example.  People can get fired or are otherwise punished for rating something bad that turns out to be good (and lots of times probably should be, just like doctors who operate when something less dangerous could have been done), so they tend toward status quo.  
by santiago on Mon Apr 26th, 2010 at 12:37:55 PM EST
[ Parent ]
Anonymize the ratings.
by Colman (colman at eurotrib.com) on Mon Apr 26th, 2010 at 12:42:27 PM EST
[ Parent ]
It's a good idea, but one that comes through sacrificing much ability to hold bad rating analysts accountable and making it harder to tell if there is rampant insider corruption among their SEC regulators. That might be an acceptable cost right now, where the pressures of conformity have proven so great, but a few years down the road, the crisis narrative of the day could instead return to the more familiar story of government officials in cahoots with raters and former/future employers in banks to manipulate securities values, and everyone will be wanting to hang the criminals who advocated reducing the transparency of ratings agencies. Prediction: if anonymity and more control of ratings by government were to occur, a new set of non-governmental privately-hired rating consultancies would just develop in their place and eventually hold more credibility among market participants than the low-paid "government" ones like Moodys and S&P.  
by santiago on Mon Apr 26th, 2010 at 12:56:17 PM EST
[ Parent ]
Sure, but they wouldn't hold any mandates under Basel II or the like - and the government could link the requirements for pension funds and the like  to the official rating agencies.
by Colman (colman at eurotrib.com) on Mon Apr 26th, 2010 at 12:59:43 PM EST
[ Parent ]
But there's the rub.  Why would the market respond favorably to rules set up by the government?  After a few years, those rules will be said to be too "old school" and inappropriate for new developments in the industry, and too politically cumbersome to update as well, leading to a new private industry doing the same thing rating agencies do now -- qualitative, privately compensated advice to very wealthy investors.  And people will believe the ones paid by the billionaires with real skin in the game before the government bureaucrats, like they do now in every filed regulated by government ratings and analysis.
by santiago on Mon Apr 26th, 2010 at 01:07:01 PM EST
[ Parent ]
Anonymize and wikify the ratings.

And make all financial transactions public with standard non-vague on-book accounting definitions, so that anyone with a mind to can check a rating for themselves.

by ThatBritGuy (thatbritguy (at) googlemail.com) on Mon Apr 26th, 2010 at 02:41:25 PM EST
[ Parent ]
Good idea.
by santiago on Mon Apr 26th, 2010 at 04:19:45 PM EST
[ Parent ]
Why would the market respond favorably to rules set up by the government?

That's not the point. The point is the market only exists within the rules set up by government.

The brainless should not be in banking -- Willem Buiter

by Carrie (migeru at eurotrib dot com) on Mon Apr 26th, 2010 at 02:56:00 PM EST
[ Parent ]
Yes, but the government only exists within the rules set up by other social forces, one of which today is a shared, deep core belief that markets are both the default and preferred way of organizing things.
by santiago on Mon Apr 26th, 2010 at 04:19:11 PM EST
[ Parent ]
And people will believe the ones paid by the billionaires with real skin in the game before the government bureaucrats, like they do now in every filed regulated by government ratings and analysis.

"People" and "wealthy investors" can believe whatever they want. The point here is to prevent the banks from counterfeiting, not to prevent market-worshippers from being separated from their money.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Mon Apr 26th, 2010 at 05:15:36 PM EST
[ Parent ]
Agreed, but my point is that people willing to risk their money -- including governments in many cases -- will just ignore the cumbersome US government paid rating agencies anyway and go with the private ones, like they do now.
by santiago on Tue Apr 27th, 2010 at 03:39:47 PM EST
[ Parent ]
It's possible to pseudo-anonymise the ratings: Have a central database, Alice, where every registered rater and rating agency are assigned random alphanumeric codes, another database, Beatrice, where every report is assigned tied to the alphanumeric codes representing the author(s) and their affiliation(s).

N months after a rating has been issued (where N possibly depends on the type of rating), the rating is compared to the actual reality. This information is then input into the Beatrice database.

The people who run the Beatrice database can then mine the now blinded data for noteworthy patterns. Any suspicious patterns discovered can be analysed independently, and only if they are actually damning is the Alice database contacted for retrieval of the identity of the suspicious author(s) and/or institution(s).

That's roughly analogous to how you'd blind a medical trial. It's not idiot-proof (no system is, and even if it were, the universe is continually working on inventing more creative idiots), and it's certainly not fraud-proof either (no institution in this universe can be made completely fraud-proof), but it would be head and shoulders above what we currently have.

(Assuming you even need rating agencies at all...)

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Mon Apr 26th, 2010 at 05:09:34 PM EST
[ Parent ]
Rating agencies are just the banker's attempt to solve the inescapable principal-agent problem given conditions of imperfect information (read: incentive to lie, even to yourself) by trying to establish an institutional wall between lenders and risk analysts by outsourcing to a pooled resource.  It makes a lot of sense and is an improvement over what existed before, but it's a problem that has no permanent institutional solution.
This is a great insight, thank you.

The brainless should not be in banking -- Willem Buiter
by Carrie (migeru at eurotrib dot com) on Mon Apr 26th, 2010 at 12:28:39 PM EST
[ Parent ]
I don't get it. Why can't the lender and the risk analyst be the same person/institution?

Peak oil is not an energy crisis. It is a liquid fuel crisis.
by Starvid on Mon Apr 26th, 2010 at 01:06:07 PM EST
[ Parent ]
Becuase they can short and/or insure their own loans if the insurance and/or short is in danger of becoming more profitable than the potential loan loss.

And without proper regulation and oversight they might be encouraged to do so.

Hypothetically.

by ThatBritGuy (thatbritguy (at) googlemail.com) on Mon Apr 26th, 2010 at 02:51:38 PM EST
[ Parent ]
Hypothetically? Hasn't the Great Vampire Squid already done just that?

The brainless should not be in banking -- Willem Buiter
by Carrie (migeru at eurotrib dot com) on Mon Apr 26th, 2010 at 02:57:05 PM EST
[ Parent ]
I was being ironic. :)
by ThatBritGuy (thatbritguy (at) googlemail.com) on Mon Apr 26th, 2010 at 03:18:29 PM EST
[ Parent ]
So, uh, can't we just ban that?

And how in gods name can they short a non-securitised loan they have issued themselves?

Peak oil is not an energy crisis. It is a liquid fuel crisis.

by Starvid on Mon Apr 26th, 2010 at 04:04:01 PM EST
[ Parent ]
One way is to swap the interest and principle payments for the loan bank A made for the interest and principle payments of a different loan that bank B made. And banning that, or regulating it very severely rather, is precisely what is being proposed in the US banking reform legislation in Sen. Blanche Lincoln's version. The banking industry is resisting it full bore, however, since that kind of insurance trading is how so many US banks posted their most profitable years on record last year, during the worst banking crisis in 70 years.
by santiago on Mon Apr 26th, 2010 at 04:32:05 PM EST
[ Parent ]
Becuase they can short and/or insure their own loans if the insurance and/or short is in danger of becoming more profitable than the potential loan loss.

This, however, requires that the sucker who is on the other end of the deal hasn't done his risk analysis properly.

Which really speaks more to the need for keeping widows and orphans out of the capital markets than to any need to rein in the capital markets' tendencies to separate suckers from their money. Because the former is possible - the latter may very well not be.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Mon Apr 26th, 2010 at 05:19:37 PM EST
[ Parent ]
For the same reason your accountant and your auditor cannot be the same person.

The brainless should not be in banking -- Willem Buiter
by Carrie (migeru at eurotrib dot com) on Mon Apr 26th, 2010 at 02:55:13 PM EST
[ Parent ]
Rating agencies are just the banker's attempt to solve the inescapable principal-agent problem given conditions of imperfect information (read: incentive to lie, even to yourself) by trying to establish an institutional wall between lenders and risk analysts by outsourcing to a pooled resource. It makes a lot of sense and is an improvement over what existed before

The problem isn't, in and of itself, that the credit rating is outsourced.

The problem is that the credit rating - which, like all risk analysis is inherently (partly) political suddenly gets treated like an objective fact simply by the invocation of an Our Standards Are Poor "AAA" rating.

In that respect, it makes less sense than keeping it in-house: As long as it is done in-house, everybody who has eyes to see with will realise that the risk analysis is at least partly a matter of political fiat. By outsourcing it to a third party, it becomes easier to convince yourself that the resulting rating is an objective fact.

Even if we take as read that performing the political process of risk evaluation in a different organisational entity gives you objectively better data quality (something that is not altogether self-evident), it is perfectly possible that this superior data quality comes at a price of less competent and cautious data processing. And the quality of a decision is limited by the worst of the data and the protocols that deal with the data.

But I may be biased by the fact that I prefer to work with data whose pedigree I know...

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Mon Apr 26th, 2010 at 04:06:47 PM EST
[ Parent ]
Quite true, but often it comes down to a trade off between trust and worst quality data.  If you believe that your counter-party in a deal has an incentive to withhold some of the great data they have from you in strategic ways, then it's probably a good idea to just assume they are withholding key data. A somewhat more independent appraiser, even if flawed, is likely to be better than the one you must assume, a priori, is lying to you.
by santiago on Mon Apr 26th, 2010 at 04:51:03 PM EST
[ Parent ]
If you're not allowed to see the underlying data, you don't buy the piece of paper. If they don't show you all the data you need, they are either lying or incompetent at data collection, and you don't buy the piece of paper.

Seems simple enough.

(Yeah, that would kill the entire securitisation market dead. That is a feature, not a bug.)

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Mon Apr 26th, 2010 at 05:30:10 PM EST
[ Parent ]
Yeah, it's called due diligence...

The brainless should not be in banking -- Willem Buiter
by Carrie (migeru at eurotrib dot com) on Mon Apr 26th, 2010 at 05:42:10 PM EST
[ Parent ]
But due diligence is not what rating agencies do.  It's what investors are supposed to do. Rating agencies are just supposed to predict the odds of default relative to alternative investments, assuming that what they've been informed about underlying collateral is true, which is just one piece of due diligence that a responsible investor is supposed to take into account.  
by santiago on Tue Apr 27th, 2010 at 10:37:23 AM EST
[ Parent ]
But due diligence is not what rating agencies do.

Well, that's quite clearly true.

Whether they should have been doing it is, hopefully, something that will be decided in the course of a fair an honest criminal trial.

Rating agencies are just supposed to predict the odds of default relative to alternative investments, assuming that what they've been informed about underlying collateral is true,

Then rating agencies are worth fuck all, if you'll excuse my French.

You need to see the underlying data to determine whether it is plausible. If the rating agencies are permitted to take the truth of the underlying data as read, then the investor still has to perform an independent analysis of the data. And once you're looking at the data yourself anyway, you might as well run the full battery of tests.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Tue Apr 27th, 2010 at 10:43:15 AM EST
[ Parent ]
I agree completely. Rating agencies aren't worth much by themselves, so the idea propogated in Basel or in current policy reforms that we can regulate risk by fixing it somehow to the judgments of analysts at leading rating agencies is inherently flawed. Basic due diligence means that investors themselves should have their own look a things before they buy, regardless of what third party bookmakers might opine about odds.
by santiago on Tue Apr 27th, 2010 at 11:22:11 AM EST
[ Parent ]
To clarify, rating agencies are supposed to provide independent analysis of risk in order to solve the problem of people misrepresenting to others what their own data mean by creating a common means of comparison. They are supposed to prevent people from lying in the interpretation of data, not prevent people from actually lying about the data themselves.  Stopping people from cooking the books is what auditors and other regulators do. But cooking the books is not what is given as the cause of the Great Recession -- bad and biased interpretation of the real risks of failure are.  It's the risk ratings that failed, not the auditors.
by santiago on Tue Apr 27th, 2010 at 11:44:02 AM EST
[ Parent ]
No, usually it's just too expensive to pull large data sets like that, so you only do it if you know the person who requested it has the capability of making sense of it.  Few do, since it requires a lot of IT as and finance skills and specialized equipment to analyze large financial data sets. The risk is that someone will look at it and come to the wrong conclusions because they don't know what the data mean.  That said, however, the better purveyors of things like mortgage backed securities, such as GMAC,  did provide useful, detailed data for public inspection accessible by website on each security.  It may have made their securities more valuable in the marketplace, but those securities went bust just like everyone else's, so it wasn't lack of underlying data that kept rating agencies in the dark -- they were selling these things as "sub" prime and "alternate" prime securities after all -- completely open about the fact that these things were backed by junk.  The AAA ratings came from the way bond tranches were organized and insured so that even if things went bad, some investors would still get paid before others (and mostly still are) so that their risk was minimal.  More underlying data about the loans could have added nothing to a rating agencies' judgment, since the rating has almost nothing to do with underlying data about individual loans.
by santiago on Tue Apr 27th, 2010 at 10:26:13 AM EST
[ Parent ]
If you're unable to conduct an independent risk assessment because the piece of paper is too complicated, then you do not buy the piece of paper.

If you're an institutional investor then you should have that capability, unless the security in question is excessively complicated (in which case you don't want to buy it). If you're not an institutional investor, then you shouldn't be playing in the capital markets with money you can't afford to lose.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Tue Apr 27th, 2010 at 10:32:02 AM EST
[ Parent ]
Rating agencies like S&P don't buy anything and their ratings have almost nothing to do with checking underlying data on securities.  They just set the odds based on reported information.  Institutional investors, on the other hand, should have the capability of analyzing the data before they buy anything, and most did.  But it still didn't save them from getting burned because lack of data wasn't the problem.  It was their beliefs and models about how the world worked that were wrong.
by santiago on Tue Apr 27th, 2010 at 10:43:57 AM EST
[ Parent ]
Institutional investors, on the other hand, should have the capability of analyzing the data before they buy anything, and most did.

Did have the capability, or did use the capability?

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Tue Apr 27th, 2010 at 01:13:56 PM EST
[ Parent ]
You can drag a horse to water ...
by santiago on Tue Apr 27th, 2010 at 03:41:45 PM EST
[ Parent ]

Display:

Occasional Series