Measuring the measurers

IF YOUR biggest shareholder is Berkshire Hathaway, the chances are you are a firm with a future. Moody's, a rating agency in which Warren Buffett's investment company owns a 17% stake, is certainly in fine fettle. Its operating margin last year was 54%; revenue is growing at close to 20% a year. This is thanks largely to the explosion in issuance of “structured finance” products: bonds backed by mortgages, car loans and the like, as well as the derivatives linked to them. Rating these products now accounts for almost half of the firm's business. The other giant of the industry, Standard & Poor's (S&P), is also in rude health. It has helped to drive its parent company's shares up by more than 30% in the past year.

Moody's and S&P have a combined market share of 80%. Together with Fitch, the number three, they have over 95%. This concentration has long raised questions about competition—and regulators are, at long last, on the case. On May 23rd America's Securities and Exchange Commission (SEC), prompted by Congress, voted to help new agencies gain faster approval. Only half a dozen companies have made the grade since the approvals system began in 1975. Toronto-based Dominion Bond Rating Service had to wait 13 years.

Glenn Reynolds, head of CreditSights, a research firm, says it is time the barriers fell. The big rating agencies are “as close to Shangri-La as you can get, at Microsoft-plus margins,” he says. They wield great power—a downgrade can tip a company into bankruptcy—but lack accountability. After crises, such as Enron's collapse, they have deflected legal assaults by claiming their ratings were “opinions”, constitutionally protected as free speech. That stance may be harder to defend next time.

Even if the SEC is serious about promoting competition, it may not succeed. Many firms that sell bonds are content to stick with Moody's and S&P, rather as buyers of technology once felt safe choosing IBM. And many funds have charters stipulating that their investments be rated by one of the big two. Steve Joynt, Fitch's boss, says he has had to work hard to get firms to change their guidelines.

This leaves some wondering if credit ratings are a natural oligopoly, with new entrants offering a level of choice that investors simply do not want.

Quis custodiet ipsos custodes?

The big agencies bill the issuer, not the investor. This leaves them conflicted, say critics, as they have an incentive to give the issuer-client the rating it wants so as to win repeat business. Investigations have been launched into whether downgrades of American utilities by Moody's, S&P and Fitch were influenced by the companies, which may have been seeking lower ratings to justify price rises. The agencies say they let clients see their reports before they are published, but only to root out factual errors or non-public information.

The biggest concern is structured finance, where the agencies now depend on a few large investment banks for much of their revenue. A recent study found that the two now work so closely together on complex products such as collateralised debt obligations (CDOs)—tranched pools of debt from different sources—that the agencies could be seen as helping to underwrite them. This may expose them to lawsuits from disgruntled investors. It is an area being studied by IOSCO, an association of securities regulators.

The other question is whether the rating agencies are up to the task. Structured issuance ballooned to $2 trillion last year and grew in complexity. The agencies may be struggling to keep up—partly because they are losing some of their best people to investment banks. The banks offer recruits princely sums to reveal how to “pierce the armour of the computer models they themselves created”, says one observer.

Add to that the weaknesses of the models. Janet Tavakoli, a consultant, believes that standards have slipped as business has boomed. In subprime lending, she says, the models misread the level of correlation between different types of assets—a crucial variable—and ignored signs that risks were greater than historical data suggested. Securitised assets using this flawed methodology were used as collateral in other deals, compounding the error. It is hardly a good sign that most hedge funds pay little attention to ratings, assuming their own models to be better. Even Mr Buffett, as happy as he must be with his investment in Moody's, crunches his own numbers when using derivatives.

The agencies have been shifting their ground, downgrading mortgage-backed bonds by the dozen and promising to keep a close eye on CDOs. But they rebut allegations that they messed up. Brian Clarkson, head of structured finance at Moody's, says that much criticism is based on the misunderstanding that ratings gauge the risk of trading losses, when in fact they look only at the probability of default.

Yield-hungry investors have failed to heed repeated warnings by the rating agencies: “We've published reams of reports, but no one reads them until something happens,” says Mr Clarkson. Martin Fridson, a credit analyst, argues that the agencies are often used as scapegoats.

Politicians and state attorneys-general want someone to blame, too. They may soon have more ammunition. Joseph Mason, co-author of the critical study on structured ratings, thinks losses on recently assembled CDOs could top $100 billion. That would hurt pension funds, which bought them, and send congressional leaders into a spin. Worse for the industry, widespread losses might cool the fervour for structured products, which command much higher fees than corporate ratings do.

But if the Jeremiahs are wrong and volumes hold up, Moody's and S&P can expect more handsome returns. It would take the entry of a Bloomberg or a Thomson-Reuters to stir up real competition, but neither is thought to be preparing an assault. So for the moment the incumbents' dominance is likely to remain, as Ms Tavakoli says, a “gift that just keeps giving.”