The "regulator's dilemma"

I often save clippings from newspapers and magazines, to refer back to at a future date. I just discovered one from the August 12, 2011 issue of the WSJ: it's from their "op ed" section, titled "S&P 500 and the 'Regulator's Dilemma.'"

The article discusses the U.S. Senate Banking Committee's examination of Standard &Poor's decision to downgrade the long-term U.S. debt. The writer points out something ironic: "S&P's judgements carry such weight because Washington [i.e., the federal government, or perhaps more correctly, the United States Congress] told the markets to pay attention to them. Federal regulators have embedded credit ratings into countless financial rules." Not surprisingly, as a result of Congress' habit of excluding various parties from their regulations (even themselves, on occasion), Treasury debt was exempt.

The author points out that "a critical ingredient in the 2008 financial crisis was the encouragement that regulators gave banks to hold mortgage-backed securities rated by S&P, Moody's and Fitch – the government-created oligopoly of credit judges."

And while Dodd-Frank instructed bureaucrats to remove credit rating references from their rules, bank regulators resisted, because of their struggle to devise better standards to judge an asset's safety.

A key statement in this piece: "as counterintuitive as it may be to politicians, having no federal standard on risk is the best standard of all." [emphasis added]


Those interested in reviewing the article's points regarding banking regulations are welcome to do so. My reason for mentioning this piece is the reference to trying to "standardize risk." It's illusive, ambiguous, impossible to classify with any degree of agreement, and impossible to measure in a way that all would find acceptable.

In my recent post that highlighted the 10 things I like best about GIPS(R) (Global Investment Performance Standards), I (with some hesitation I might add) applauded the introduction of the requirement for a three-year annualized standard deviation. Not because I think it's an ideal measure, because I'm on record objecting to it. However, I also realize that there is no measure that all would agree with, and that this is a formula that is quite easy to calculate and interpret. Firms can include additional risk measures; the new requirement simply aims to have something that prospective clients can see.

In the Standards' 2010 edition exposure draft, the GIPS Executive Committee suggested mandating risk disclosures in composite presentations; this was objected to by most who took the time to comment, and the EC wisely withdrew it. Risk is SO difficult to get one's arms around. Trying to regulate it, much more than this simple requirement, would probably be unwise. And, create additional dilemmas we don't need.