Mr. Kramer wasn't alone in his assessment of rating agency power at the Forum–or even a week earlier, when PXRE, announcing a $300 million boost to its hurricane losses, said it was considering strategic alternatives, including a sale, in anticipation of downgrades to follow.

Then, during an earnings conference call for Aspen Holdings that coincidentally took place the same day as the PXRE ratings cuts, Aspen CEO Chris O'Kane was asked for his take on the market impact of PXRE's downgrade. “It's another powerful reminder of the power of the rating agencies,” he said. “Most people assume that an 'A' in the rating reassures that you can trade successfully. With a rating with a 'B' in it, it's very, very tough to keep a portfolio of clients together.”

In recent months, Rosemont Re, Quanta, Alea and PXRE were all forced into “difficult situations” when their ratings fell.

I will not argue–nor do I have any evidence to suggest–that any or all of these companies should not have been pushed into these situations. Indeed, in some cases, there's market chatter to suggest businesses might not have been managed at levels appropriate for the volatile lines written.

I will argue, however, that “power” extends beyond the rating agencies, and that situations such as these should serve as reminders that the power of other market players–namely, brokers who react to rating agency actions–needs to be wielded with careful consideration. The inevitable questions about whether rating agencies acted too much, too little, too soon or too late also apply to users of ratings.

As I heard company executives discuss rating agency clout over the last few weeks, I recalled an interview I did with analyst Steve Searby of Standard & Poor's back in September 2004–about the time when Converium's ratings were being cut by S&P and other agencies.

“Brokers have long used ratings as one of the filters to be applied when selecting reinsurers for their clients,” he said, noting that the minimum hurdle for short-tail business is often as high as “A-minus.”

“S&P has long argued that this is not an effective use of ratings, as it creates arbitrary rating cliffs, whereby the reinsurer can find business drying up when it would still be considered creditworthy in the absence of the hurdle. Indeed, a 'triple-B'-range rating is considered 'good,'” he said, going on to comment on the danger of downgrade clauses in reinsurance contracts.

At that point I urged Mr. Searby to clarify. Does S&P believe ratings should not be used to measure quality? How should brokers measure reinsurer quality for ceding company clients?

He responded that “ratings are designed to measure financial strength on a relatively smooth continuum–varying from 'triple-C' to 'triple-A.' For a broker to set a minimum security level, we think represents the wrong use of ratings.”

He added that “a sensible way to [measure quality] is to say, I want some 'double-A,' 'single-A' and 'triple-B' reinsurers on my program, but I'm prepared to pay more for 'double-A' security.”

Drawing a parallel to the way a debt market works, he said that “if you buy a 'triple-A' bond, you expect to receive a lower coupon than [for] a 'triple-B,' and we would argue that [brokers] should be using ratings in the same way.”

A.M. Best analyst Robert DeRose had a different concern–about brokers' use of downgrade clauses–back in September 2004. “Some lower-rated reinsurers have, I think, wrongly maintained a position in the marketplace–in part due to [brokers'] ability to sell these” clauses, he said, arguing they provided “a false sense of security” that an “A-minus” with a downgrade clause was as good as a “A-plus.”

Somehow, witnessing all the fingers being pointed at rating agencies recently (and catastrophe modeling firms, and “the weather!”) to explain the difficult straits that some companies find themselves in now brought these comments back to mind.

Putting the specifics of the PXRE and Quanta situations aside, I wonder whether these analysts' words deserve consideration–whether we need to examine the power of the users of ratings in the insurance industry.

Reinsurance brokers have done a great deal of work in recent years to develop technology and teams and models to independently assess the ability of reinsurers to pay claims. Yet the business still turns on those fateful announcements from rating agencies of cuts to “B-plus-plusses” and “triple-B's.” Is that appropriate?

Recent events also speak to the increasing difficulties faced by ratings agencies, brokers and any of us who try to assess the financial strength of insurers and reinsurers. Didn't life seem easier when we could attribute downfalls to inadequate loss reserves for casualty lines?

“Easier” is an oversimplified way of saying that however hard it is to assess the adequacy of liability loss reserves (and, yes, it is difficult and requires some judgment), it is still a quantitative measure–and quantitative measures are easier to justify and debate.

Increasingly, we are becoming aware of the need for rating analysts to meet the challenges of qualitatively assessing the ability of company managers to measure and manage the risks of their enterprises and the interplay of those risks. It's a difficult exercise for the companies themselves, and a much more difficult task for raters, brokers and customers looking in from the outside.