Technology improves risk-related decision making

Technology improves risk-related decision making

By Steve Pietrzak and Lamont Boyd

Physical property inspections remain a primary source of risk management information gathered for insurance companies, for accurate underwriting of new applications as well as renewal policies.

The importance of a renewal review cannot be overlooked, as the information obtained from an original inspection does change over time. While floods or tornadoes can certainly quickly alter the state of a property, they are only the tip of the iceberg when it comes to fully understanding renewal policy hazards and current home values.

And it is far from ideal to try to generate an accurate quote for property insurance at the time of application without access to physical inspection data. In many cases, property inspections uncover significant hazards that the applicant didn’t disclose, such as a trampoline or swimming pool, which may require a material alteration of the terms of the quote or written policy.

Underwriters should use property inspection information as a guide and should be allowed to make underwriting decisions based on all reasonable underwriting evidence presented. In reality, most properties are free of undue hazards, and most consumers properly maintain their properties and are not likely to present an inordinate number of claims. As a result, these are the risks that would and should receive more favorable underwriting and pricing consideration.

These good risks should not have to pay higher prices to help insurers cover poorer risks. Consumers stand to benefit if insurance companies can do a better job of pricing according to risk, making insurance more available and affordable for the majority of consumers.

Predictive analytics technology, which has been recognized as a top trend to watch by many industry experts, is a tool that improves decision-making, enabling underwriters to refine the information procured from an inspection by appropriately weighting the observed attributes according to their predictive value, producing a score that rank-orders applicants and policyholders by their expected loss relativity.

Predictive analytics methodologies have been used to statistically develop an insurance risk model based on inspection data. They are also able to help an insurer more effectively segment property risks at the point of application for more objective and consistent underwriting and pricing decisions.

Predictive analytics tools are enabling insurers to utilize the relationship of a set of known outcomes (loss ratio performance) and corresponding risk attributes (property inspection or application factors) to predict the next outcome if similar risk attributes are presented.

This is a powerful advantage for insurers, because risk data analytics offer an objective, impartial and consistent tool to assist in their underwriting efforts. Part of the underwriter’s job is to determine who is more likely to submit a claim and what rate to charge based on that exposure to loss.

If your organization is considering a predictive analytics solution for managing risk, here are five tips that insurers should consider:

1.   Predictive analytics rely on the collection of accurate information, both during model development and at decision point. The more accurate the information, the greater the predictive accuracy, the better the decisions being made.

2.   It is important to have a clear understanding of the performance being measured (e.g., loss ratio relativity, profitability) by predictive analytics to assure significant added value to your decisions.

3.   Predictive analytic models provide objective and consistent indications of risk that can be effectively used in conjunction with other underwriting and pricing variables.

4.   Predictive analytics allows for more effective deployment of resources toward applications, policies or books of business of indicated concern, allowing for automated underwriting and pricing of greater and greater numbers of risks using rules-based decision tools.

5.   It is important to monitor results over time as you may want to adjust your use of predictive analytics to play an earlier or enhanced role in your decision-making.

Steve Pietrzak is president and CEO of Itasca-based Millennium Information Services, Inc., a company he founded in 1991. Lamont Boyd, the insurance market director for global scoring solutions at FICO and is recognized as one of the industry’s leading experts in predictive scoring technology.

Posted on Thursday, August 06, 2009 (Archive on Thursday, August 13, 2009)
Posted by jstoltz Contributed by jstoltz

Return