New credit-based insurance score taps two powerful resourcesBy Tim Grinde | February 4, 2020
Credit-based insurance scores deliver strong, recognized correlations with policyholder risk. But, as with many things, the tool is only as good as the materials used to create it. Without a deep reservoir of consumer data modeled and refined by industry-leading data scientists into insightful analytics, credit-based scores may not reach their full predictive potential. And it may be harder to clear regulatory hurdles for their use.
The challenge is big enough to command the resources of two leaders in data and analytics. This has brought together Verisk, a leader in insurance analytics, compliance, and actuarial science, and Equifax®, a global data, analytics, and technology leader. The result is an inflection point, a change of direction toward a more powerful form of credit-based insurance scores.
Verisk’s Inflection™ Insurance Score (“Inflection”) uses a trended-credit model powered by Equifax to help personal auto and property insurers improve risk segmentation, pricing, and customer service with applications across the policy life cycle. Analysis using Equifax and ISO data shows that, compared with control models that don’t include credit-based attributes, Inflection can provide 2.7 times lift between the lowest and highest risk bands in predicting losses.
By combining Equifax credit data on some 220 million consumers with Verisk’s vast repository of premium and loss data and advanced predictive modeling, Inflection provides a balanced framework in which to evaluate policyholders, improve competitiveness, and help avoid adverse selection. Inflection can integrate with Verisk products and services such as the ISO Personal Auto program and LightSpeed® Auto, Verisk’s one-rate quoting platform.
Through it all, Verisk’s deep knowledge and experience with insurance laws and regulations help ensure that credit-based tools are compliant as well as effective. Inflection will be available to insurers in the second quarter of 2020.
Tagged with: ,