The Rating Agencies and Complexity, Conflicts, and Complacency – Additional Note

This is a further comment to my two posts below on the rating agencies, here and here. I probably should have titled the post on regulators investigating the rating agencies differently.  Although the post starts out talking about US government civil investigations into the rating agencies, it is mostly about theories of liability that might be raised, not just by US authorities, but by other jurisdictions, particularly as Europe reacts to sovereign debt downgrades.

I agree with Glenn (and thanks for the Insta link) that the US government investigations are not likely to go anywhere, partly because the relations between regulator and regulated are likely too close for political comfort, and partly because of the optics of actually pursuing S&P so soon after the US sovereign debt downgrade.  In principle these are all separate and the investigations commenced before the downgrade, etc., etc.. But in our current environment of intertwined regulators and regulated, Wall Street and Washington (whether you call it a salutary blending of the “public” and “private” interests to the benefit of everyone, or just plain crony capitalism and rent-seeking), I find it hard to see an investigation as being anything other than a nod to the Senate committee and its hearings.

That said, the fundamental point I seek to make below, despite the post title, is that although fraud – including fraud in some highly technical, securities-law version with its own peculiar standards of intent and other elements – is the main and likely the only realistic basis for going after the rating agencies in the United States, and is not likely to win for the government, other jurisdictions might see things quite differently.  US law sees this as having to fall into the fraud paradigm, however specifically worked out in statute, regulation, and case law, and not as a question of standards of due care by the rating agencies with regard to the markets as a whole.

Other jurisdictions might well believe that there is a standard of care both in the construction of risk models and in their specific applications, and that this is a duty of care to the markets as a whole, if one chooses to be a participant in a general way – not merely as some uncompensated commentator or blogger, but accepting fees, expressing views as a commercial player, etc.  The US has many federal and state standards as to when commercial actors owe each other duties and of what kind and degree, but no one should assume that those rules obtain in other jurisdictions.  And what might not be achievable as a legal result in the US, might well obtain – with some of the same practical consequences – in the EU.  For that matter, I do not regard the chances of successful litigation against the rating agencies on some theory of standards of care owed to the market as a whole by a market player as zero even in the US, though those theories are much more likely to succeed somewhere in Europe.

My post title takes off from the current investigations by US authorities, but is really about the range of possible theories of liability, in the US or elsewhere.  I should note that this is on the assumption that no one finds evidence within the rating agencies or the relationships with Wall Street that comes much closer to a smoking gun of conflicts of interest resulting in actual deception resulting in actual fraud.  The theories in the post are premised on the rating agencies not having committed fraud in its traditional sense.

The easiest way to support that assumption is to assume what many financial journalists have said (though disputed by some of the comments to the post below), viz., that the complexity of the financial instruments was such that the rating agencies did not understand them sufficiently to model the risks, and instead accepted the banks’ own (highly flawed but deeply self-serving, we all now understand) models as a basis for making their ratings decisions.  Many of the commenters emphasize not complexity as befuddling the rating agencies, but instead conflicts of interest arising from who pays the rating agency for the rating – the issuer.  This surely plays a part, and so does “complacency” – the assumption that someone remote in time and space, pensioners in Germany (to use an example from a Wall Street banker that Roger Lowenstein cites) or somewhere, will be left holding the bag.  It is true that market players who are skeptical of the ratings won’t much care so long as the risks are offloaded on someone else, but someone at the end of the road either has to be mistaken as to the uncertainties but willing to rely on the ratings; or else believes mistakenly that there is a further party in the Ponzi scheme on whom to offload the instrument; or else believes that the size of the market stamped with AAA ratings is so large and has already been partly blessed through the US government NSRO brand that the US government will ultimately insure against failure, i.e., the final party in the Ponzi scheme onto whom the risk is offloaded.

That’s looking with hindsight.  But seen from the front end, I would stress that those coming up with these instruments were in a complex cognitive and psychological world in which it was not overwhelmingly evident that this was the house of cards all can now see.  They were in a sea of hopes, uncertainties about the future, insufficient worry about other people’s money, the ready cash available upfront, the fact that many smart people with much mathematical skill were telling them this was okay, the comforting presence of the herd, the gently subversive and lulling deep water current of an anticipated Bernanke put to save things, and many other things that cloud the waters.  There is a narrative flaw in Michael Lewis’s The Big Short, in other words, despite its many virtues – the flaw is that the heroes of that book are the people who indeed had the foresight to short the mortgage market, but they are not people who could have been picked in advance to be the Wise Voices.  These guys were brilliant, contrarian, and in this case right – and in any other circumstance might have been just as brilliant and contrarian, and utterly wrong.

In those circumstances, the conflict of interest that the rating agencies have as to who pays them is easy to focus on as the root source of corruption.  It is a problem, and needs a fix.  But issuers have been paying rating agencies for ratings on corporate bonds for a long time and that does not appear to have been a problem despite the genuine conflict.  Issuers retain trustees in a securities issuance and the securities laws impose various duties on the trustee despite being selected by the issuer; the potential problems from this conflict are significant, but don’t appear to have led to market meltdown fraud.  What makes those different, however, is a level of transparency as to what might constitute bad, fraudulent, and deceptive behavior, such that both investors and regulators have a pretty good model for it.  It’s not so complicated.  Combine uncertain about the future, complexity in the model, and many promises from very smart people that we have turned a corner on risk management and we can have a much freer lunch than before, and we have “this time is different.”

Well, plus ca change … my underlying point is that future financial regulation needs both to separate out the elements of complexity, conflicts, and complacency, as well as how they intertwine to reinforce each other.  And to do so in a way that is something that can be captured by the common sense of a prudential regulator.  Prudential regulation by following the herd is not a great heuristic; but neither is, the contrarians must always be right.  Nevertheless, financial engineering (I’m not sure the term is all that useful anymore) seems to have taken as its metaphor systems “optimization” and the assumption that there’s a way to capture of all possible benefits.  Mechanical engineering for a robotic craft trying to navigate around the Martian landscape, with much built-in slippage and assumption of many unanticipated failures – limits – might be a better engineering metaphor.