I was flipping through a report about insurance company financial statements the other day. I like such things. They tell you what's going on in enormous detail. This particular report I found sitting on an airplane seat, left behind by its owner.
(It wasn't my greatest airplane find. I once picked up a prospectus for an as-yet unannounced insurance company IPO that someone had left on the plane. I toyed with breaking the news worldwide, but then something got the better of me. I called the company and told them what I had. They knew who I was and so denied that they could spell IPO, let alone had one in mind. "So I can run the story," I asked? "Do so, and we will sue you into the 4th century B.C. for stealing the document you don't have," the woman said. Facts confirmed. I didn't write the story, which was just as well, since the IPO never occurred.)
The company report I found recently analyzed many of the largest reinsurance companies. I was especially struck by a comparison of estimates the companies first released for the 2005 hurricanes Katrina, Rita, Wilma, and the final amount of losses they paid.
Here's how it works. A hurricane hits on Oct. 1. You are under immediate pressure to issue a press release advising investors of how much you think you'll have to pay. Based on your computer output and industry estimates from the modeling companies, it looks like your share is, let's say, $75 million. You put out a press release to that effect. So does everyone else in the industry that might have to meet claims.
Three weeks later, the modeling agencies revise their estimates. You have more information by this point, and now, one way or another, it looks as if you're in for $120 million. You put out a press release. So does everyone else.
Then year-end arrives. You have to book a hard number for the auditors to pretend to look at while they're busy drafting letters of nonresponsibility. You have still newer data, and so do the modeling agencies. You figure you're on the hook for $150 million. You issue your annual report. So does everyone else.
Next up: the first quarter figures for the following year. That's six months after the hurricane hit. The real number is starting to coalesce. It's $113 million. You deduct $37 million from your reserves, and issue the quarterly earnings press release. So does everyone else. The value of your stock goes up by 10 percent. So does everyone else's.
By now, you can see where I'm heading with this. The "invisible hand" that's supposed to drive the markets assumes a perfect flow of perfect information.
The insurance industry, namely the property-catastrophe folks in this example, have put out erroneous figures for the third and fourth quarters of last year, and now the first quarter of this year. No accusations here: this is how it works.
You paid your underwriters their year-end bonuses based on wrong figures. Because the bonuses were less than the underwriters had anticipated, they all left and got jobs elsewhere. At the end of this year, their bonuses will also be inaccurately calculated because of last year's reserve adjustments booked this year by their new employers.
Is this any way to run an ant farm?
ROGER CROMBIE is a Bermuda-based columnist for Risk & Insurance®.
November 1, 2009
Copyright 2009© LRP Publications