When 'speed' is missing from a speed-to-market approach

Antiquated processes hamper insurance innovation. A new approach can put insurers in the pole position.

Real-world insurance policy pricing models often have 50 to 100 variables such as gender, age, credit score, claim history and more. (Photo: iStock)

Speed to market is a high priority for many property and casualty insurers, but creating quote-ready rate tables remains a major roadblock for personal line insurers to realize that priority. This stubborn issue applies to any line of business that has numerous rating variables and coverages. If it takes two years to update rating plans, the underlying rate factors may already be stale by the time the rate plans are used.

Auto and homeowner policies are two examples. If an underwriter were asked, “Where do rates come from?” he or she might respond, “The actuaries.” In fact, factors created by actuaries are but the first step. The process of creating quote-ready rate tables is complex and time-consuming — truly among an insurer’s most complex architecture processes.

Rate creation

Actuaries build statistical pricing models that use factors to determine rates. A 16-year-old male will be charged higher rates than a 30-year-old female, all things being equal. That is because the model uses factors such as: male = 1.06 vs. female = 1.00; and age 16 = 1.52 vs. age 30 = 1.12.

In this simple example, using a base rate of $1,000 as the premium for our two prospective customers, the results are:

Of course, this example is a simplification. Real-world pricing models often have 50 to 100 variables such as gender, age, credit score, claim history and more. Auto and homeowner policies frequently have many different coverage choices. Some auto coverage examples include liability, comprehensive, collision, under insured, uninsured, towing and rental. For an insurer that writes in multiple states, there may be 50 to 100 different rating factors for multiple coverages for up to 50 states.

Competition, regulation and marketing

Actuaries are responsible for creating pricing models that best reflect anticipated claims. But these models do not account for competition, regulation or marketing. Insurers will often change a factor to remain competitive or gain market share in a specific segment.

If marketing requests that the factor for expensive vehicles be lowered, an insurer could become more competitive in the high value vehicle segment. State regulators might chime in regarding specific variables (gender or credit score, for example) that can’t be used for rate quotes. Ultimately, the process of creating quote-ready factors becomes an iterative circle comprised of actuaries, underwriters and regulators.

The big question

If factor values that come out of the pricing model are unacceptable, then what should the factor values be? Answering that question is more art than science.

To determine the “final” factor, an analyst reviews current factors, new factors, impact on current premium, supplemental exhibits (such as graphs for every factor at a coverage/state level), and other relevant data. Current premium is required to see the impact of changing the factor. If the current factor is changed from .72 to .69, the impact on current premium also needs to be anticipated.

The process of selecting a “final” factor value is a balance of desired goal, impact on premiums across the book of business, and building a consensus that can include actuarial, underwriting, marketing and regulators.

It’s a traffic jam

Insurers often use spreadsheets for these complex calculations, as they are well understood and accepted. It’s common practice. But should it be?

Let’s again use auto insurance as an example, which can be 50 to 100 factors and seven coverages across 50 states. Often a coding language capable of interacting with spreadsheets will be introduced for automation, shaping data and mimicking business rules. That’s a lot of spreadsheet files and code that needs to be updated and maintained. The result is a labor-intensive, error-prone, highly complex process that few people in the organization truly understand.

Consider the potential size of the problem: An insurer writes $100 million in auto coverage, using a factor in every auto coverage and state. If a clerical error causes a factor to be off by just .01, the result is a $1 million premium loss or premium overcharge. The task of measuring the impact of factor changes at a coverage or book level is iterative, and the labor hours required to measure the impact of changing factors are significant.

Putting speed in your speed to market

So, how can an insurer remove the factor management dead end and accelerate speed to market? What steps are required? The three-part road map involves:

Commitment. Realize that moving factors from model output all the way through quote-ready tables can be a highly complex process. It is essential that senior management understands this and supports efforts to modernize the factor management process.

Measured steps. Develop a plan to start small and scale out to larger segments of your book. This allows for learning, feedback and tweaking the process.

Process and technology. Even as technology advances over the last 40 years have completely transformed the industry, spreadsheet-driven factor management processes remain stalled. Don’t get me wrong. Actuaries can keep their spreadsheets. Spreadsheets are indispensable for tasks like displaying data, creating formulas and allowing users to update values, but they should not be used to shape data and mimic business rules. Newer technologies bring substantial advantages.

What does a modernized factor management process look like? Business rules can be placed inside a business rules engine that supports versioning, governance and ease of use. Final factors get married to business rules by version, in turn making re-rating policies as simple as selecting a version and clicking the run button. Finally, a process flow guides data creation, business rules and factor selection. This allows the analyst to make a factor selection without the drudgery of maintaining spreadsheet files.

A solid platform for managing factors and measuring the impact of factor changes opens the door to enhanced pricing analytics like retention, lifetime value and optimization.

The road to modernized factor management leads to better efficiency, accuracy and governance. Most importantly, it will put real speed back into your speed-to-market campaign. Start your engines!

Matt Bailey (Matt.Bailey@sas.com) is an insurance solutions architect at SAS. He has worked more than 20 years for P&C insurance companies, specializing in actuarial, underwriting, fraud analytics, data and IT as they relate to data management, enterprise analytics and predictive modeling.

These opinions are the author’s own.

See also:

Recommended Stories