On the New Statesman blog today, David Skelton of Policy Exchange argues [3] that “We must free ourselves from the tyranny of the credit rating agencies”. Citing the failure of the big three ratings agencies (Standard and Poors, Moody’s and Fitch) to anticipate the subprime mortgage crisis, he says that the agencies now “hold enormous power over democratically elected governments … often enough to force a government turn away from the democratic mandate on which they were elected.” Because they’ve been so badly wrong before, we should “stop thinking that their declarations should be decisive”.

I think Skelton has missed the most important point. He asks whether the "anonymous, powerful experts deserved the credibility and the exalted position they are given by the media and politicians?", apparently unaware of the fact that the ratings agencies owe their position not to media (or even politicians') trust, but to a complex thicket of US financial regulation.

As Professor Lawrence J. White of Stern Business School has written [4], it was “the regulatory structure that propelled these companies to the center of the U.S. bond markets”, and what has stopped them from going down following their colossal failure in 2008.

In 1936, financial regulators eager to impose discipline on the banking sector introduced rules that banned banks from holding bonds rated below BBB standard by one of the “recognized ratings manuals” – S&P, Moody’s or Fitch. In White’s words, “the creditworthiness judgments of these third-party raters had attained the force of law.”

Over the following decades, insurance regulators followed suit, so that eventually all fifty US states had rules requiring insurance companies to hold amounts of capital commensurate to the riskiness of their bond holdings as judged by the ratings agencies. Federal pensions regulators did the same in the 1970s.

Finally, in 1975, the Securities Exchange Commission (SEC) created a category of “nationally recognized statistical rating organizations” (NRSRO) to risk-rate the bonds that the SEC now required broker-dealers to hold. According to White: “The other financial regulators soon adopted the SEC's NRSRO category and the rating agencies within it as the relevant sources of the ratings that were required for evaluations of the bond portfolios of their regulated financial institutions.” The SEC only granted NRSRO status to four more organizations over the next 25 years, but by 2000 these had merged with the original three, once again leaving only three agencies on the market.

Whether through error or design, the NRSRO application process was remarkably opaque, with no formal application or review processes. As White argues, these regulations created (and continue to create) an enormous barrier to entry for any new ratings agency: “Without the NRSRO designation, any would-be bond rater would likely be ignored by most financial institutions; and, since the financial institutions would ignore the would-be bond rater, so would bond issuers.”

With the ratings agencies as insulated from competition as this, it is hardly surprising that they all made the same errors in the run-up to 2008. Nor is it a surprise that their market dominance has continued, despite that massive failure. It was only in 2006 that the SEC’s barriers to entry were reformed at all, and even then the reforms were limited, at best.

These facts are not well known. Jeffrey Friedman has argued [5] that widespread ignorance of the ratings agencies’ status was a significant contributory factor in the subprime mortgage bubble – bankers were unaware that the ratings they were getting were not the product of a competitive marketplace (where risk-taking that turned out to be correct would be rewarded) but of a quasi-governmental oligopoly.

I may disagree with Skelton more broadly – when he accuses bond markets of holding governments to ransom, he is really just attacking them for being wary about who they lend to. And only governments relying on borrowing to fund spending could be forced to change policy to satisfy bond markets. But his assessment of the ratings agencies appears to miss the most important fact of all: that these agencies are creatures of regulation, not competition.