Steve Reads

Does anyone know how we got into a position where a few large institutions (Standard and Poor’s, Moody’s) tell us how reliable a given security is? John Quiggin makes the point that after the subprime collapse, those companies have got to go.

My (undeveloped) instinct would say that “in the long term,” the interest rate should reliably signal the risk of a security; hence the rating agencies would be superfluous. I’m sure there are reasons why that’s wrong, and Quiggin may in fact hint at some of those reasons in his post. If your organization is required to invest only in AAA securities, you’re presumably going to be very cautious; you won’t, for instance, invest in any investment vehicles until they’ve been tested. So any of the exotic derivatives would presumably be ruled out. Maybe one justification for a debt-rating service is that it keeps money flowing into securities that would otherwise be illiquid. It’s not clear that this is a virtue.

I’d just like to know how we ended up here in the first place. Why rely on what Moody’s considers safe? If I’m a rational investor who’s concerned about the safety of my money, why not pay attention to where, say, Calpers or Warren Buffett are investing? Is this too much of an informational burden to expect most investors to bear?

If, on the other hand, I pay attention only to interest rates as a measure of risk, I really want to be looking at the standard deviation of interest rates over a sufficiently long period; the return on something like a blue-chip stock would be less variable than the return on a penny stock, and likewise for a Treasury bond versus a derivative. Maybe it’s unreasonable to expect people to pay attention to beta coefficients and whatnot.

So then if boundedly rational agents can’t be expected to process all the data that’s in front of them, someone is going to come in and simplify for them. Whether it’s Consumer Reports or Calpers or whoever, someone is going to publish a simple measure of a security’s risk. What’s not clear to me is why we’d then expect there to be so few risk-measurers. If we just “naturally” ended up with Moody’s and S&P, this would seem to suggest that reputation is a good with massively increasing returns; monopolies would fall right out of the process.

That may well be the case, but it strikes me as odd. Do any other goods have so few guarantors of their reputation? If you’re looking to buy a car, there are limitless places you could check: you can see if it’s a good model on Consumer Reports; you can check its blue-book value; you can check its serial number on the web to see if it’s ever been in any accidents; you can do a quick web check on the vendor to see if other people have labeled him a scam artist; etc. All of that would take you approximately 10 minutes, if you’re a slow typist.

Or consider a restaurant’s reputation. Part of that reputation is backed by the law: you can be reasonably sure that there will be no feces in your food. There’s an argument that lawsuits and capitalism would buy you everything that sanitation departments do: a persistent pattern of soiled food would quickly put a restaurant out of business, and you could sue them if someone you love dies. That seems to make the point more strongly: there’s a web of institutions putting some lower bound on a given restuarant’s reputation. There’s no Standard & Poor’s for restaurants. So why is there an S&P for securities?