S&P’s decision to downgrade US Treasury bonds from AAA to AA+ has elicited various reactions, some of which I’ll doubtless repeat here. Obviously, S&P has no particular expertise (apparently it couldn’t even get the arithmetic right) and based on its historical and continuing performance, its opinions ought to carry no particular weight with anybody (they say so themselves, when under pressure over obvious cases of misrating, asserting that they are merely offering an opinion).

On the other hand, it’s also pretty obvious (and even more so after the Repubs successful use of the debt ceiling to force Obama to abandon any call for tax increases along with the cuts they both wanted) that the US has some fairly intractable problems in dealing with its (technically quite manageable, but still substantial) public debt. Finally, as I said last time I discussed this, a decision of this kind (including a decision to maintain AAA ratings) is inherently political

There are two reasons why S&P’s choice of rating matters more than, say, my own opinions on the matter
* First, a lot of investors still pay attention to ratings agencies, for whatever reason
* Much more importantly, agency ratings are embedded in global regulations concerning prudent management of investment. If a second major agency were to join S&P in downgrading, large numbers of institutions would be debarred, under existing rules, from investing in Treasury bonds

That’s clearly unsustainable, so what will happen?

All sorts of fudges are possible, but the only clean response is to remove reference to private agency ratings from regulations prescribing prudent management of investment. Under current rules, provided managers invest in AAA-rated assets they are normally regarded as having discharged their duty of care. On the other hand, the agencies insist that their ratings are mere opinions, and that they have no duty of care whatsoever in offering them. Dodd-Frank was supposed to fix this, but as I just discovered (H/T Anders Widebrant), the SEC has abandoned any attempt to implement this part of the law.

But what is the alternative. I (and others) have previously floated the idea of a public ratings agency, but I now think this is a roundabout and inappropriate solution to the problem of defining a safe harbor for institutional investors. What is needed, simply, is a list of approved investments drawn up by the relevant regulators. Investors who choose assets outside this list would do so on the basis of their own judgements and would have to defend that judgement in the event of default.

There are some crucial differences between this and the existing system. First, regulators take responsibility for their own decisions, rather than outsourcing them (though some may choose to use a common list). Second, and most importantly, this approach reverses the onus of proof regarding financial innovation. If someone comes up with a new financial instrument, and a theory as to why it is (almost) risk-free, they have to persuade regulators (who will carry the can if something goes wrong) to approve it. Under these circumstances, the kinds of institutions that are required to make prudent investments will be effectively excluded from investing in innovative instruments. Given the history of financial innovation, that’s a good thing.

There are still problems here. Lots of governments have AAA-rated debt and lots more would like their debt regarded the same way, even if (or especially if) they are not pursuing particularly prudent fiscal policies. Obviously they would like, and will pressure, regulators to take a favorable view, and will be particularly miffed if their debt is dropped from an approved list. There are several cases here. One is the case where the regulator is responsible to the government concerned. In that case, I think nothing can be done or should be done. The government will declare that someone who invests in its own securities has done nothing wrong – institutions regulated by that government will have to make up their own minds whether or not to do so. The second case is that of the debt of foreign governments – here a downgrading might involve some embarrassment, but as long as it is done by independent regulators. The third, and trickiest case is that of regulation in a federal or confederal system (US states or EU member countries). I haven’t thought through this one yet, so I’ll leave it for comments.

What would happen to the ratings agencies if they were cut out of the regulatory loop? I’d guess that they would continue the business of rating ordinary corporate debt, much as before, but they would lose a lot of business associated with rating innovative financial instruments (given their record of failure in this area) and government debt (given that they would be competing both with regulators and with CDS markets).