Criticised for failing to predict the banking crisis and for the simplicity of their financial strength models, has the illusion of power wielded by the rating agencies finally been broken?

In early June, billionaire Warren Buffett leapt to rating agencies’ defence during his testimony to the fifth hearing of the Financial Crisis Inquiry Commission in New York. “I think they made a mistake that virtually everybody in the country made,” he said of agencies’ failure to call the bursting of the US housing bubble and reflect this in the rating of mortgage-backed securities.

But Buffett, whose company, Berkshire Hathaway, is the largest shareholder in Moody’s, is the exception rather than the rule. Everyone else, it seems, is queuing up to tear strips off the agencies, from journalists to politicians, and even the issuers and institutions that pay rating agencies to rank their bond issues, creditworthiness or claims-paying ability.

Those looking to take shots at rating agencies will find no shortage of ammunition. Aside from failing to spot the housing bubble, some argue that they had over-rated what turned out to be weak financial institutions. Lehman Brothers, for example, was rated A by Standard & Poor’s (S&P) directly before its collapse.

Model behaviour

The insurance industry has its own specific examples of what appear to be poor rating calls. Both Moody’s and Fitch cut American International Group’s ratings two notches in September 2008 after the collapse of Lehman Brothers triggered the wider financial crisis, and S&P dropped the insurer three notches. The deep cuts suggest that the original rating assumptions were incorrect.

One critic from the reinsurance industry points out that many of the insurance and reinsurance companies that hit trouble during the financial crisis were in the double-A range. “I can’t think of an A- company that had a problem,” he says. “It calls into question the accuracy of the ratings and the robustness of the rating models.”

Part of the reason for the crisis-triggered downgrades, argues the critic, is that rating agencies’ insurance financial strength rating models had failed to take into account the asset side of reinsurers’ balance sheets, and so did not spot the risks they were taking on there.

He also says the financial strength models themselves leave a lot to be desired. “They continually update them, but they struggle. Capital adequacy modelling is extraordinarily sophisticated and the rating agencies are never going to truly know what’s on the inside of a reinsurance company. The ability to make those fine gradations between the different rating notches is not there.”

A further concern for some is that the ratings are not solely based on capital models. Rating agencies will typically incorporate opinions on a company’s management and business strategy, among other qualitative measures, into their verdicts. The reinsurance executive recalls asking why a downgrade had been made, only to be told that doing so had made one of the agency’s rating committee feel more comfortable. “What do you do with that?” he asks.

Above all, rating analysts are not infallible. “There were times when I came to realise that I had blown the rating – where my rating conclusion should have come in lower and I was very unhappy with the result,” one former rating analyst says. “Ratings are opinions and they are not going to perfectly predict risk for any number of reasons.”

Even when they make the right call, the rating analysts’ actions can be delayed by executives eager to keep their company afloat. A downgrade below the A-range can effectively put an insurer or reinsurer out of business because insureds or cedants will no longer consider it secure enough to place business with.

It seems some executives will stop at nothing to forestall this, even when a downgrading is justified. One chief executive actually reportedly stormed into a rating agency office to confront an analyst about a decision.

Cutting a company’s rating below the critical A-range can also be delayed because analysts need to be certain about decisions that could put a company out of business. “Downgrading a company’s rating is not a simple matter, as the downgrade is more than just an opinion; it can become a self-fulfilling prophecy as the market reacts to this new information,” says the former rating analyst. “For me, downgrading a company required a higher burden of proof – sort of like sitting in the jury box where the judge instructs the jury not to convict if there is reasonable doubt.”

A possible indicator of the waning confidence in ratings is that some reinsurance cedants are now using credit default swap spreads as early-warning systems, because financial strength ratings do not respond quickly enough.

Missing the point?

While acknowledging that rating agencies are not perfect, the former rating analyst believes their performance over time is good. He says that judging by the default rates published by S&P and Moody’s, which go back 30 years, three in every 100 triple-A-rated companies would fail over a 10-year period, compared with between four and five for double-A and between five and six for single-A.

Some rating agencies say complaints about their models and methodology are rare. “Generally, we have received very favourable feedback from investors and other users of our ratings on how we responded with respect to insurance and reinsurance company ratings as events of the financial crisis unfolded,” Fitch group managing director and global head of insurance ratings Keith Buckley says.

Others feel that criticising model capability misses the point. While reinsurance executives may argue that their own internal capital models are more sophisticated than those of the rating agencies, for example, rating agencies contend that this can only be expected given their role.

“When you look at what reinsurance executives are trying to do with their model, you would expect it to be more sophisticated and specific to fit their needs,” says senior vice-president of global property/casualty ratings at AM Best, Matt Mosher. “Rating agencies are looking at a broad spectrum of companies and we have to be able to look at them on a consistent basis.”

While acknowledging that there are more sophisticated economic capital models available than those typically used by rating agencies, the former analyst says: “For comparison purposes, a static, risk-based capital model in my mind is superior to the most sophisticated economic dynamic financial analysis.”

He adds that the number of company-specific assumptions made in a dynamic model can make comparisons between company projections challenging.

Despite complaints about rating agencies’ use of softer measures, such as management strength, the agencies themselves argue that this is essential for rating companies. Mosher explains that, while a purely model-driven approach works for rating financial instruments, the relatively changeable nature of companies renders such an approach is useless for rating them.

“A company could buy another that completely changes its risk profile. That doesn’t happen with an asset-backed security – it doesn’t change overnight,” he says. “That is why management and qualitative factors are as important in a corporate rating.”

Staying relevant

While ratings are not perfect, and often need to be used in conjunction with other tools, the guidance they offer can be a boon for security committees at brokers or cedants as they provide an independent opinion on a reinsurer’s financial strength, and offer broader, deeper coverage than a committee alone.

Yet despite their usefulness as a guide, some feel rating agencies’ days are numbered. Financial centres around the world have introduced or are introducing advanced solvency regimes, such as the European Commission’s Solvency II directive.

“Rating agencies run the risk of becoming irrelevant for reinsurance companies,” the reinsurance executive says. “If a public, governmental entity whose only task is to monitor insurance solvency says these reinsurance companies pass every test they can devise, what is the point of an A versus and A+ versus a AA-?”

The former rating analyst argues that previous attempts by governments to set up rating agencies have failed, however, and a single central agency would not be able to offer the diversity of opinion that multiple agencies can.

“I don’t think there is a perfect solution. I think you can improve the current solution, but I don’t think you can find something better than having multiple rating agencies covering the same companies,” he says, adding: “The devil you know is probably better than the devil you don’t, as rating quality can be handicapped by the intelligent user.” GR