The following pertain to actuarial pricing. For a full list of articles on this site, please use the Article Index.

New Product Pricing

By Jeff Adams, November 12, 2013

New product pricing can be significantly more difficult than the renewal rating of a current set of benefits. In renewal rating, the actuary generally uses historic experience to estimate future costs by adjusting this experience for any known or expected change in circumstances. In new product pricing, there may be no historic claims experience upon which to base future cost estimates, thus making pricing very difficult.

This article discusses some of the steps used in pricing new benefits. Some of the steps are consistent with renewal rating, but new product pricing can add a couple of steps to the process. The actual process used in pricing new benefits is dependent on the benefits being priced, the data available, and the desired methodology with which the actuary is comfortable. There is generally more than one method for pricing, and the actuary has some flexibility in choosing the best method for the given situation.

Steps in New Product Pricing

The following are general steps used in pricing, although these steps may vary significantly based on the product being priced and the data available:

  1. Determination of specifics surrounding product being priced.
  2. Determination of data to be used in pricing.
  3. Determination of, and valuation of, differences in circumstances between data to be used and actual pricing scenario.
  4. Trending experience to pricing period.

The end product should be well-documented and allow all persons reviewing the documents to follow the process, identify all underlying assumptions, and understand the final rates or rating factors obtained. The document should adhere to all applicable state and federal regulations and requirements of applicable Actuarial Standards of Practice.

Specifics of Product Being Priced

For more accurate pricing, the actuary must understand the implications of the circumstances surrounding the product being priced on the services’ cost and utilization. The following is a partial list of elements that would need to be considered:

  1. Is it a stand-alone product offering or an additional benefit that will be sold in conjunction with other benefits? If it is a stand-alone benefit, it is probable that the impact of adverse selection would need to be added to the rates as those who believe they would need the benefit would be more apt to choose to buy it. The level of selection to be added is dependent on the extent to which this service could be anticipated in advance.
  2. A comparison of benefits and rates to competitors’ benefits and rates can also assist in assessing selection that might occur. If a new product has higher benefits and higher premiums than competitors’ benefits and premiums, persons purchasing this higher benefit would likely, on average, need this benefit and use the service at a higher rate than those purchasing the competitors’ benefit, thus increasing the utilization assumption and possibly the cost per service assumption. Likewise, suppose the benefits and rates are less than the competitors’ benefits and rates. In that case, the utilization and cost per service assumption might be reduced to reflect the fact that those choosing the benefit might be more interested in a cheaper product than a high level of benefits.
  3. The provider network and payment method play a major role in pricing. If there are in-network and out-of-network benefits, the adequacy of the provider network in-network and out-of-network need to be analyzed. A network barely meeting adequacy requirements may mean utilization may be less than a network with an overabundance of providers. In-network versus out-of-network adequacy should be analyzed as a factor of cost along with the payment method in-network versus out-of-network, such as capitation in-network and percent of charges out-of-network.
  4. Utilization management and utilization review protocol are also important as part of the cost analysis for many services. Often, utilization management and utilization review can substantially reduce costs, although it may slightly increase administrative costs. The actual impact may vary substantially by service.
  5. Other elements such as the claims processing system and legal department and upper management mentality may affect costs. These elements may be significantly different between carriers resulting in differences in payments even if everything else is the same. For example, a legal department in one company may not want to put forth effort fighting certain claims payment disagreements and require its company’s claims department to pay claims that a more restrictive company may not. Another example is a company that is more “customer-centric” than another and would tend to pay more to keep members and providers satisfied.

This is only a brief list of items that should be analyzed even before progressing to the step of deciding what data to use as a basis of the rate calculation. Many other items may affect pricing, even items that may be overlooked due to apparent insignificance at first look.

Determination of Data to Be Used

The choice of which data to use is critical in the pricing process. Data should be chosen with risk characteristics similar to that of the population for whom the product will be sold. For example, PPO data for an older population should not be used if the product is to be offered to an HMO population and is expected to draw younger adults.

The risk characteristics of the population from whom the data is taken need not be identical to the population that will buy the new product, but the actuary should be able to adjust for differences mathematically. For example, continuance tables from a consultant should not be used to build rates from scratch for a new product, but these continuance tables can be used to calculate an adjustment factor from the base experience of the health insurer to the benefits being valued, or demographic factors can be used to adjust for different demographics. Continuance tables have an underlying set of risk characteristics, such as a PPO population, and other attributes such as a certain quality of utilization management that is different from that of any insurer. Using a consultant’s continuance table to calculate the rates from scratch may produce misleading results due to these differences.

Sources of data that can be used include, but are not limited to, insurer’s actual claims experience, publically available databases, consultant databases, related studies available on the Internet, related studies available from more private sources such as American Hospital Association, etc., and actuarial studies available from sources such as the Society of Actuaries.

Adjustments to the data may need to be made for any or all of the differences in risk characteristics or other differences described in the sections above.

Determination of, and Valuation of Differences from Data to Target Population

Much of the previous portion of this article discusses variations from the data to the target population. The next step is for the actuary to value any difference in underlying risk characteristics from these two populations. This may be a very difficult task, or even impossible, in some situations. The actuary must know and understand the strengths and weaknesses of his or her assumptions during the calculations. For example, if using continuance tables from a consultant, the underlying benefits upon which these continuance tables are based may not be known. This does not make the benefit cost relationships unusable, but the actuary must understand that the underlying benefits may affect the final factors obtained.

The actuary must also understand the sensitivity of the varying assumptions on the pricing. For example, suppose 2012 non-Medicare population data is used to estimate 2014 rates for a Medicare population for a particular integration with Medicare. In that case, much attention needs to be paid to the impact of the type of Medicare integration. Using an incorrect trend might impact the rate by up to 10% or so, but an incorrect assumption on the impact of the Medicare integration might impact the rate by 250% or more. A rigid Exclusion of Benefits integration with Medicare might have an integration factor of 0.250 or a reduction of the appropriately aged experience by 75%, while a loose Coordination of Benefits integration may result in an adjustment factor of 0.900 or a reduction of 10%. If the actuary uses the 0.250 when the 0.900 should have been used, then the impact would be a ( (0.900 / 0.250 – 1) x 100% =) 360% understatement in the rate. Obviously, this is an extreme example, but there are many other scenarios with less dramatic scenarios where not enough attention is paid to factors that significantly impact the pricing process.

Adverse selection may need to be included if the selection for the product to be sold differs from the selection that would be inherently in the base data used for the rating. For example, if a stand-alone dental product is to be sold, the use of dental experience for dental benefits that can only be bought alongside a medical benefit might require an adjustment of 8% to 10% to reflect the additional utilization inherent in a dental stand-alone product. Selection adjustments may need to be applied if a product is to be offered against certain competitor products and, as a result, draw buyers with risk characteristics different than the base rating data used.

Trend and Other Adjustments from the Experience Period

Even if data with identical risk characteristics to the target pool risk characteristics are chosen, adjustments need to be made to the claims experience. These adjustments include trends from the experience period to the rating period. Adjustments need to be made for any other changes that might affect the costs. These changes would include state mandates or changes in utilization management, utilization review, or claims processing which occurs between the end of the experience period through the end of the rating period.