Article

Analytics and Technology: Taking Underwriting to the Next Level

By Steve Callahan
Practice Development Director

Although insurers have made real progress in automating risk assessment processes, there is significant room for improvement in the quality, speed and consistency of risk decisions.

Advances in healthcare, longevity, medical practices and information collation, combined with an increasingly diversified population, are bringing greater complexity to the underwriting risk assessment process. This is despite solid progress made over the past decade in automating risk assessment processes. Electronic applications, straight-through processing, workflow systems and document management have converged to greatly improve the flow of information to and from underwriters. Notably less progress, however, has been made in gathering supplemental information or improving the quality, speed and consistency of risk decisions.

Risk assessment remains a black-box consolidation of underwriter experience, supporting information and intuition. Is untapped revenue and opportunity being blocked by the rigor of “old school” underwriting?

Given good claims experience, many companies are disinclined to “mess with perfection” in fear of loosening the knob too much. The risk of worsening the claims experience without necessarily offsetting premiums is a strong inhibitor to adopting new risk assessment models. Consider life insurance, where both the frequency of claims and the ability to adjust premiums are measurably less than annually renewable property/casualty lines.

With life, the challenge is exacerbated by underwriters’ skepticism that technology can approach the effectiveness and sensitivity of their decision making for even the more routine cases. A Society of Actuaries study indicated that only 1% of North American life insurers were utilizing predictive modeling in the underwriting process. This is despite proven successes in the property/casualty industry, where risk selection and the buying process have been effectively transformed through the use of scoring and modeling. In the life industry, opportunities abound to significantly expedite the underwriting process while saving millions of dollars in labor and laboratory test costs.

Days vs. Weeks for Underwriting Decisions
By identifying candidates who can be issued policies without ordering lab tests and medical records, consistently high-quality underwriting decisions can be made in days instead of weeks. Cost-effective models are already being used to supplement carefully defined application information with third-party data that includes personal history interviews, Medical Information Bureau (MIB), motor vehicle records (MVR), prescription databases, credit reports, career information and family histories. With 30% to 50% of the business going straight through without any additional tests or records, service levels improve, underwriting resources are freed up, expenses are reduced and proper focus can be put on the other 50% to 70% of complex cases requiring human oversight.

Moving beyond individual decisions, predictive modeling can be used to analyze the risk profiles of entire market segments. Predictive indicators can provide marketers with insights into the likely profitability of a given product set in a given market segment prior to introduction. Then pricing can be fine-tuned to create greater competitive advantage. The result is profitable products that have unique appeal with a given market segment.