Special Section sponsored by TMPAA

Underwriting Managers

Mid-year meeting session addresses achieving consistent profitability

This year’s mid-year meeting of the Target Markets Program Administrators Association in April featured a session on how underwriting managers can achieve consistent profitability with their programs. Prior to that discussion, Ken Robinette, president of Bellingham Underwriters, Inc., in Seattle, Washington, gave a detailed account of how underwriting managers (UMs) deal with program exposures versus how managing general agencies (MGAs) deal with them. Of course, Robinette was speaking in general terms because not all MGAs operate on the same basis, nor do UMs.

With that in mind, Robinette made the following observations:

• The UM must research the product and program premise and sell them to a carrier, spending considerable time and development expense. In the case of the MGA, the carrier usually brings the program to the MGA or helps with the development.

• The UM usually develops the rating plans while, in the case of the MGA, the carrier usually defines the rating plans.

• The UM is responsible for all policy administration. Generally, the MGA and the carrier split the policy administration, with the MGA generally doing the quote and binding phase and the carrier generally issuing the policy.

• The UM is responsible for loss control and safety services. With the MGA, the carrier generally is responsible for loss control and safety services.

• The UM is responsible for the administration of claims and the philosophy by which they are adjusted. Generally the carrier controls all aspects of claims management.

• The UM develops and maintains full systems for rating, policy issuance, endorsement processing, claims administration, statistical coding, data collection, premium accounting and management reporting. Carriers generally share their system with the MGA and provide the MGA with pre-agreed amounts of information.

• The UM is wholly responsible for profitability of program. The carrier and the MGA share responsibility for profitability, though true responsibility lies with the carrier.

• Broad authority is given to UM. Limited authority is given to MGA.

Having listed those differences, Robinette then went on to discuss the methods and requirements of an underwriting management company. He said that in the development of an underwriting plan, one needs to consider the operational characteristics versus the qualitative risk characteristics.

“Qualitative risk characteristics are those usually found in rating plans,” said Robinette. “They are generally considered ‘judgment rating’ factors. Whether a property or liability coverage, these would include such hazard categories as: condition of premises or housekeeping; management experience; safety attitude or plans; cooperation with carrier; and employee experience.

“These factors tend to be fuzzy by nature and are used to determine the quality of a risk when compared to others in the same class. The presumption is the class on which the risk is rated determines the correct price for the average risk and the judgment credits,” he explained.

The problem, according to Robinette, is that, for general liability, there are 1,115 class codes or 1,010 SIC codes. For workers compensation there are 615 classes. “For commercial auto liability we have only 39 primary class codes followed by eight secondary groupings, so we end up with a total of 288 possible classifications. In every grouping of risk, there exist hundreds of thousands, if not millions of variations of operations,” said Robinette.

Noting that most of the insurance industry is rated using a class rating structure of some kind based on a single rating variable, Robinette said that, in many classes, the rate variable is not the best determination of actual exposure. “You begin to build an underwriting plan by researching your class or product to determine the primary causes of loss frequency and severity,” said Robinette.

“In that class or product, what are the best indicators of causes of loss? How can they be used to measure exposure? Look first at how the industry rates the product and ask yourself if the rating basis is truly the best measurement of actual exposure. In most cases, you will find it is not. Find what is, and build your selection criteria around the true risk measurement.”

Robinette defined “operational risk characteristics” as those “which measure and define the difference in operations without regard to quality.” They measure the quantity of risk or exposure generated from a specific operation.

Said Robinette: “Operational risk characteristics might be number of customers, board feet or cubic volume of material produced or miles traveled. Or, they might be a laundry that does pick up and delivery, a contractor who provides free engineering service, or a landscaper who installs sprinklers or builds fountains or just mows lawns. Or, they might be physical differences such as a grain elevator with an open head house or a swimming pool with a diving board.”

He continued: “These factors do not deal with quality, but describe unique characteristics, which define the amount of risk generated from a class by differentiating it from other risks within that same class. Build your underwriting plan based on operational risk characteristics.”

Robinette then turned his attention to loss ratio management techniques. “The primary goal of loss ratio management is to create indicators that allow you to determine whether your pricing and selection models will yield the results you need to achieve,” he said. “Build data collection capabilities and spend the money to collect the data to analyze your business. Set benchmarks in advance and review your book at each benchmark.

“For casualty lines,” continued Robinette, “the best benchmark is a limited loss model. When writing policies with a limit of $1 million, we use a loss limitation of $100,000 because 50% of gross expected loss will occur within the first $100,000 and 50% will occur in the $900,000 excess of $100,000 layer. Because of this division, if we want to achieve a 60% ultimate loss ratio, we must allocate our premium so that 40% is available for expenses, 30% for losses in excess of $100,000 and, consequently, we must keep limited losses less than 30% of premium.

“This relationship is not understood by most underwriters and is the primary cause for carrier and program failure,” Robinette explained. “If a program regularly writes business that has never had a large claim, but runs a routine loss ratio of say 45%, then over time the loss ratio on the book of business will become 90%, since the limited loss will represent half of the total projected loss.

“The other great advantage of using a limited loss model is that it does not develop significantly at the end of the earning period, or even during the earning period, so it allows us to know usually before the end of the first program year whether our pricing and selection model is yielding the targeted results.”

Finally, Robinette advised his audience: “Always know your business better than the company you represent. Never allow a carrier’s actuaries to ever give you information you don’t already know. Always tell the carrier’s actuaries why the results you achieved have been achieved. And, know any bad news before they do, and provide an analysis and solution before they are aware of the problem.” *