Ratemaking Research in 2005
By John Pedrick

Just in time for the Seminar on Ratemaking in New Orleans, Louisiana, the 2005 Call for Ratemaking Papers has come to a close. In January 2004, the CAS Committee on Ratemaking issued a call for papers to be published in 2005, just prior to the seminar. We reviewed proposals in March and drafts throughout the summer and fall. After working with the authors, offering advice and editorial comments, we are pleased to see these papers published in the Forum and presented to the actuarial community.

So, what new ideas can you expect after your morning beignets and café au lait? What will actuaries discuss over dinner, besides the unresolved debate over okra versus filé for thickening gumbo? Will the oyster shuckers at Felix's or ACME care?

Here are the answers.

Generalized Minimum Bias Models (.pdf)
We may all be familiar with minimum bias methods and generalized linear models. Now, in "Generalized Minimum Bias Models," or GMBM, LuYang Fu and Peter Wu present their research into a comprehensive and flexible approach to these methods. In their paper the authors show that all the multiplicative minimum bias models along with commonly used Generalized Linear Modes are special cases of GMBM.

The authors present the following advantages of their approach: There is no assumption of a specific form of distribution, increasing application appropriateness and model-selection flexibility. It improves accuracy and goodness of fit for classification rates. It is easy to understand and does not require advanced statistical knowledge.

Anyone who regularly employs minimum bias methods or Generalized Linear Models, or who wants to gain further understanding of them, will find this paper helpful and interesting.

Loss Trend (.pdf)
Chris Styrsky demonstrates an underlying problem in using calendar year data to estimate trends in, "The Effect of Changing Exposure Levels on Calendar Year Loss Trends." He demonstrates that as exposures increase, calendar period severity will decrease because the older, higher value claims from lower exposure years are compared to more recent, higher exposures. The opposite is also true when exposures are decreasing.

The author then proceeds to demonstrate a method for overcoming this problem by calculating adjusted pure premium, pairing past exposures with the losses they produced.

Tort Reform (.pdf)
Gail Tverberg explores ten problems actuaries should consider when analyzing changes to a state's tort law in, "Pitfalls in Evaluating Proposed Tort Reforms." Actuaries who file medical liability rates have encountered requirements from states (like mine—Ohio) to reflect changes in tort law in their rate filings. Company management also needs to gain an appreciation for the potential changes in future losses due to these changes in law.

The author gives the reader important tips and advice for approaching this problem. Regulators and actuaries involved in lines affected by tort reform will gain insight into the difficulties associated with the many changes we've seen recently.

Captives (.pdf)
Ann Conway's paper, "Ratemaking for Captives and Alternative Market Vehicles," provides an excellent introduction to captives of various sorts and the problems encountered in ratemaking for them. The author shows the various structures and entities involved in captives, risk retention groups, and sponsored cells, the usual sources of capital for them, their major domiciles, reasons for forming these entities, as well as the many issues encountered in estimating rates. Actuaries who need to gain a quick understanding of what captives are and how to approach ratemaking for them will find this paper helpful.

D&O Reinsurance (.pdf)
Those who are involved in directors and officers liability reinsurance will want to read "D&O Reinsurance Pricing—A Financial Market Approach," by Athula Alwis, Vladimir Kremerman, and Junning Shi. The authors start with some interesting data showing settlements in the largest securities class action suits, along with the most successful plaintiffs' law firms. They continue with a discussion of the history of this market and current ratemaking practices. They then present a proposed methodology that incorporates market capitalization, frequency of lawsuits, loss as a function of market capitalization, and correlation between and within industry sectors.

These five papers will be available online and published in the 2005 Winter Forum—just in time for you to join the fun in New Orleans.

So, if you find yourself standing at the ACME oyster bar and the shucker seems interested in your thorough knowledge of the latest actuarial research, don't be shellfish—leave a nice tip!

Bad jokes and technicalities aside (are mollusks considered shellfish?), all five papers will be presented during the concurrent sessions at the Seminar. After you've relaxed with coffee and beignets, come sit in on one of these sessions. Your questions, comments, and ideas for future research are always welcome.

What about the gumbo debate? Use both, just to be safe.