Here's how engineering works: Machinery operates until it fails. The particular cause of the failure is then fixed, and it does not happen again. The machinery then runs until it fails for some other reason. And so on, until a newer model is introduced, which in turn runs until it fails.
Barring the discovery of perpetual motion, or imperfect human beings suddenly becoming able to design perfect machines, it could not be any other way. In time, everything fails. Call it the Roger Principle (unless it's already called something else).
Bridges, apparently, stand for about 30 years, and then they collapse. Their replacements are constructed in the knowledge that they, too, will fail at some point, for completely different reasons. Airplanes fly thousands of long-distance journeys, then a bolt gives out, the fleet is grounded and that problem doesn't happen again. O-rings on the Shuttle; the inevitable blue screen of death on your computer; even human beings--everything fails.
For many years, I have watched the insurance industry fail, achieving only one annual underwriting profit in 25 years. Every new major event, it seems, is cause for head scratching and cries (actually, off-the-record comments) of: "We didn't see that coming." Never more so than lately, as a consequence of the learning curve in property-catastrophe insurance.
I was initially highly critical of this record of failure. Now that I understand that everything fails, I have a little more sympathy. If the insurance industry had been able to see such catastrophes coming, adequately priced policies would have covered them, and insurance magazines would have been scratching for headlines.
But . . . a subsidiary principle of the Roger Principle is that when something fails, it must not be allowed to fail again in the same way. Is that happening in insurance?
In 2000, a 13-year-old soft market prevailed in many property/casualty lines and seemed set to last forever. Then came Sept. 11, 2001. What failed there was the insurance industry's ability to predict the inhumanity of suicide bombers crashing airplanes into buildings. The insurers "fixed" that problem by ceasing to offer meaningful terrorism insurance and letting governments carry the weight.
Then came the hurricane season of 2004. One, two, three, four--Charley, Francis, Ivan and Jeanne wracked Florida and the Caribbean, causing untold difficulties centered on "second-event" outcomes. Before that could be properly fixed came 2005. Katrina was hardly unprecedented, but its severity across property/casualty lines was. Rita and Wilma followed, underlining the intensity of the Katrina effect with further second- and third-event issues.
Insurers have been "fixing" those problems ever since with new models and new capital requirements, innovative financing creations such as sidebars and collateralizations, and a bunker mentality in the face of the 2006 hurricane season.
As sure as Del Monte makes apple juice, something else is going to happen sooner or later--may have happened, by the time you read this--that will be unprecedented and industry-changing, an event that will alter the shape of risk as we understand it. In due course, the particular danger of the financial damage it causes will be "fixed" and avoided forever. What are insurers and reinsurers, if not financial engineers?
The day on which we have thought of, and provided for, every possibility will be a dull day indeed for columnists, and for the shareholders of insurance companies. Luckily, that day will never come. Everything may fail, but we know not the hour of its failing. Therein lies the essence of risk: not so much what, but when.
is a Bermuda-based columnist for Risk & Insurance®.
August 1, 2006
Copyright 2006© LRP Publications