Say what? I didn't quite catch that statistic until I saw the image of cardboard cylinders piling up into the shape of the Empire State building. The next time I saw it, I listened to the voiceover and now that fact is etched into my brain.

This isn't another column about sustainability initiatives. What I am more interested in is the creative presentation of numbers in a way that captures our attention and leaves us wondering how on earth they could possibly be accurate.

In our industry, reinsurance brokers have been convincing us about the enormity of the amount of excess capital in a similar fashion. A market turn won't happen unless the industry incurs losses equivalent to “three Hurricane Katrinas”—roughly $50 billion, a broker announced two years ago.

With the images of the suffering Louisianans carved my memory, the idea was mind-blowing.

Two years later, with worldwide insured catastrophe losses mounting to $22 billion in 2009 and $36 billion in 2010, there's been no turn. And as we reported last week, a reinsurance broker believes that “even a $50 billion hurricane loss event would only give the market pause for a year or so.”

“At $100 billion, we believe 'outlier' reinsurance entity failures could occur, while a $150 billion insured loss event would create a decided and sustained turn,” Guy Carpenter said on its blog, gccapitalideas.com, earlier this month.

That's nine Katrinas!

Such predictions are typically developed from comparisons of today's industry ratio of premiums-to-surplus to levels that prevailed during hard markets. There's sound reasoning behind the calculation, but given the magnitude of the results it seems logical to ask if the yardstick is entirely appropriate.

Guy Carpenter gives the answer that it is not. “The element of surprise is important,” broker executive David Flandro said in an article we published last week. A $50 billion terror event or earthquake could have a larger market impact than a $50 billion hurricane, he reported.

Indeed, it's not hard to imagine a liability issue—a flood of cyber losses or a torrent of wage-and-hour claims—hobbling one or two high-profile insurers enough to instill fear in casualty competitors.

What if such losses were incurred at a time of major investment losses?

On Sept. 17 of last year, I was surprised to watch tornado-shaped icons approaching my hometown depicted on a map on the Weather Channel, and I was sure a market turn was imminent. Two tornadoes and a macroburst touched down in the boroughs of Queens and Brooklyn in New York that day.

What if a stronger event occurred nine miles west, in midtown Manhattan, causing damage near a building that would take 34 billion one-dollar bills to fill lined up end to end?

The industry may have enough capital, but how much is really excess capital?