Calibrating premium to risk is crucial to your business – and the business of daily life for your customers. What if improving accuracy also meant uncovering profitable opportunities and increasing your bottom line? Well, that’s the power of a credit-based insurance score.

The More Variables and Tools, the Closer to Crystal-ball Status

Determining exposure involves a variety of variables. New tools and nontraditional, behavioral data attributes can help create a more accurate result. When it comes to automobile insurance, usual suspects include geography, driver class, coverage history, loss history, vehicle type, and motor vehicle record (MVR) information. Meanwhile, emerging technologies such as telematics deepen behavioral insight by gathering data on how drivers operate vehicles, enhancing the overall risk assessment.

Another factor, consumer credit history, makes a sizeable impact on risk and loss prediction for both personal auto and property insurers. But a credit-based insurance score focused on credit reports alone is limited when compared with new available attributes and modeling techniques.

A more nuanced picture of a consumer’s financial history – coupled with powerful analytics – can supercharge the predictive power of a credit-based insurance score moving forward.

What’s in a (Credit-based Insurance) Score? 

Started in the early 1990s, the credit-based insurance score initially relied solely on credit report attributes. Even with a limited number of attributes, the score quickly demonstrated a high correlation with predicting loss. A 2003[1] study of more than 2.7 million automobile policies ranked the credit-based insurance score as one of the top three most important risk factors within each major coverage area.

Today, use of the credit-based insurance score is accepted practice among regulators and consumers. It’s also considered one of the most powerful pricing and underwriting predictors for automobile and homeowner insurance.[2]

Auto insurance policies in the lowest insurance score range are two times more likely to submit an insurance claim than policies in the best insurance score range.

An Inflection™ Point Where New Data Meets Predictive Modeling

Including more data attributes in the score improves predictability and fully assesses the risk of more potential new customers. For the millions considered “credit invisible” by the Consumer Financial Protection Bureau,[3] analysis of credit history can result in no-hit or thin files. Alternative data like utility and telecom information supplements credit report information. As a result, it paints a clearer picture of a wider market for personal auto and property insurance.

Powered by data from Equifax(R) and developed by Verisk™, the Inflection™ Insurance Score combines predictive analytics and actuarial expertise with powerful data and attributes, including trended credit and alternative data. It’s an industry-leading, credit-based rating solution for personal property and casualty insurers. The result? It offers 2.6 times the lift in pure premium relativity from the lowest to the highest risk band. That's when compared to a control model that did not include credit-based attributes.

Whether you’re looking to refine a personal property and casualty program or reduce adverse selection, learn how Inflection™ gives you the starting point for balanced risk assessment.  

 

[1] Michael J. Miller and Richard A. Smith, “The Relationship of Credit-Based Insurance Scores to Private Passenger Automobile Insurance Loss Propensity.” EPIC Actuaries, LLC. June 2003.

[2] Linda L. Golden, Patrick L. Brockett, Jing Ai and Bruce Kellison, “Empirical Evidence on the Use of Credit Scoring for Predicting Insurance Losses with Psychosocial and Biochemical Explanations.” North American Actuarial Journal, 20:3. 233-251.

[3] Kenneth P. Brevoort, Philipp Grimm and Michelle Kambara, “Data Point: Credit Invisibles.” The Consumer Financial Protection Bureau Office of Research. May 2015. https://files.consumerfinance.gov/f/201505_cfpb_data-point-credit-invisibles.pdf