Verisk Analytics announces a new addition to the ISO Risk Analyzer suite of predictive modeling tools for more accurate pricing of physical damage coverages for personal autos.

The ISO Risk Analyzer Personal Auto Vehicle Module is an analytic tool that uses vehicle characteristics to predict collision and comprehensive losses at a highly refined level. The vehicle module represents a significant evolution in the refinement and accuracy of vehicle Symbols and vehicle rating.

"For decades, Verisk's ISO subsidiary has provided tools to understand how the car a person drives affects insurance risk, says Kevin Thompson, senior vice president of insurance services at ISO. "ISO is now revolutionizing these tools with the introduction of powerful analytical methods and extensive data on vehicle characteristics. The vehicle module will provide unparalleled information to insurers about vehicle risk."

The vehicle module contains two levels of output—ISO Risk Analyzer personal auto symbols and more detailed component-level output.

Personal auto symbols provide more than 400 symbols and corresponding rating relativities for comprehensive and collision coverage. The new symbols are two characters in length, composed of a "price new" character and a "vehicle attribute" character.

ISO derives the ISO Risk Analyzer Personal Auto Symbols using individual vehicle attributes, such as body style, and performance and safety features, as well as information related to the manufacturer's suggested retail price (MSRP). The new symbols represent the combined predictive power of hundreds of vehicle characteristics distilled into a compact symbol representation.

"ISO Risk Analyzer Personal Auto Symbols are significantly more granular than ISO's current Symbols, in part due to the more refined treatment of vehicle series and the use of actual vehicle characteristics in the modeling," says John Kollar, vice president of personal lines, increased limits, and rating plans at ISO. "ISO Risk Analyzer Personal Auto Symbols bring the power of predictive analytics to physical damage vehicle rating and allow you to easily enhance your vehicle rate-segmentation capability."

ISO plans to file ISO Risk Analyzer Personal Auto Symbols with insurance regulators in most states. The filings will introduce an alternative rating rule that insurers may license and adopt.

In addition to the auto vehicle symbols, the vehicle module provides outputs to aid insurers in their loss prediction and internal modeling. Module output consists of loss relativities and components, with separate frequency and severity models built by coverage for collision and comprehensive. Comprehensive coverage is also modeled by peril.

For each coverage, the set of relativities provides loss indications for the three major facets of each model: price new (or MSRP), model year, and vehicle characteristics. The relativities can be used in conjunction with the new symbols to provide refined estimates of insured losses.

Components are groupings of input variables constructed from VIN information and a number of data sources. They represent powerful combinations of specific loss-related vehicle attributes. The components can be thought of as "super variables" that an insurer can use directly as inputs in their own predictive modeling efforts.

"One of the best things about how we have constructed the product is that you can use the output of the vehicle module as stand-alone analytics to help classify, segment, and price your vehicle physical damage risks with confidence," says David Cummings, vice president and chief actuary of ISO Innovative Analytics. "Or you can use the components as input to jump-start your own customized model, taking advantage of ISO's extensive data acquisition, preparation, and variable selection work that is necessary to build a reliable model."