The rating-agency practice of making arbitrary adjustments to a company's loss estimates in an attempt to reflect an update that may not be credible for that company is also troublesome.

An insurance company's rating is a very important indication of financial health. For non-catastrophe information, rating agencies have long had the benefit of consistent, independently audited financial information that is derived using well-known and accepted standards. Similar standards must be used for catastrophe-loss information.

Cat risk should not be opaque or mysterious for any company. The process by which natural hazards cause damage is straightforward, and so is the modeling process.

For hurricanes, for example, there is a wind footprint for each modeled event, and that footprint is superimposed on the company's insured-property values by location. Based on the wind speed and the type of property, damage functions are applied to estimate damage to the properties.  Policy conditions are then applied to the damage to estimate the insured losses from the event. 

The process is the same for every model. It is the model frequency and severity assumptions that differ, resulting in differences in model outputs.

So how can the process for gauging cat-risk exposure be made more transparent and more consistent for rating agencies? The most important step is to develop a consistent, robust set of transparent scenarios to represent cat risk for each region. The same benchmark scenarios can then be applied to every company's portfolio so the rating agencies are truly comparing apples to apples.

Let's use the Northeast as an example. Most scientists agree that the 1938 Great New England Hurricane is the region's worst and most-damaging event on record.  Most sources put the total insured wind damage at $20-$30 billion in today's dollars, and the return period at 100 years, plus or minus. This is a very plausible 100-year benchmark event.

To appease skeptics who will say this same event is not likely to happen again, we can anticipate additional scenarios by varying the track and other characteristics of the wind footprint to develop credible 1938-like events. 

There are many advantages to this approach over the current process of using model-generated 100-year probable-maximum-loss-point estimates. Among the most important is that companies and rating agencies can clearly monitor effects of exposure changes. Currently these cannot be separated from model changes and nuances.

Cat models are blunt tools, not surgical instruments. Over the past 20 years, the models have improved significantly, advancing one might say from a hand saw to a chain saw. But a saw is still a saw. It is not appropriate for the delicate work of brain surgery, and a cat model is not appropriate, as currently used, for rating-agency purposes.