“There is more emphasis on risk management as catastrophes are getting bigger and worse,” says Morgan Davis, independent director of Montpelier Re Holdings Ltd. “I've heard from many CEOs who worry about their company's ability to survive tough years like what we experienced in 2011. CEOs are asking themselves, 'Do we know how much risk we have?'”

According to Ernst & Young's Global Insurance Outlook, the No. 1 thing carriers can do to improve their combined ratio is to achieve superior underwriting. Yet carriers face challenges in underwriting as they struggle to understand the risk profile of the properties they write and renew from year to year.

The magnitude and volatility of historical combined ratios support the notion that unacceptable Property and Liability risks are being systematically accepted and retained through the underwriting process. Additionally, properties are being renewed at insurance-to-value levels well short of 100 percent, causing premium leakage and negatively impacting combined ratios.

This confluence of market realities leaves carriers in search of different methods to improve the performance of their Homeowners' portfolio. Some carriers are implementing rate increases and geographic retraction, and many are also turning to the use of predictive analytics as a promising technology to help improve underwriting, claims and marketing.

|

What Is Predictive Analytics?

Predictive analytics uses statistical and analytical techniques to develop models that enable accurate predictions about future event outcomes. Predictive models can take various forms, with most models generating a score that indicates the likelihood a given future scenario will occur. Mining data in order to identify trends, patterns or relationships is a key component of building predictive models. Predictive analytics enables powerful and sometimes counterintuitive relationships among data variables to emerge that otherwise may not be readily apparent, thus improving a carrier's ability to predict the future outcome of a policy.

How Predictive Analytics Impacts Homeowners'

The use of predictive analytics in Homeowners' insurance has steadily increased over the past several years. A majority of Homeowners' insurers have utilized predictive analytics for many years in rate-making (i.e. loss-cost relativities, by-peril rating), rating-plan assignment (sometimes called “underwriting scorecards” for tier or company placement, declines and nonrenewals) and price-discount qualification. A widely known practice is the use of credit-based insurance scores in many states.

Homeowners' carriers are also integrating predictive models to help with claims handling, marketing and new methods of underwriting. Claims analytics models are being used by forward-thinking carriers in several ways, including to:

• Predict overall claim severity
• Focus on early identification of fraud
• Target the likelihood that a claimant will litigate a particular claim

Marketing departments are taking the lead from financial-service organizations to analyze consumer purchasing patterns and other behavioral attributes through the use of predictive models. This additional insight allows marketers to improve both their hit ratio (the percentage of prospective customers who buy a policy) and retention ratio.

Leading carriers have recently discovered better, more cost-effective ways of underwriting risks beyond the use of underwriting scorecards. The fundamental problem many carriers face is the lack of reliable data on the homes they insure, which diminishes an underwriter's ability to estimate the expected future loss performance of a policy. Using predictive models, carriers are able to redesign costly information-gathering efforts, such as property inspections, and re-focus underwriting efforts on homes that are most likely to have insurance-to-value deficiencies or condition and liability hazards. This allows a carrier to charge the amount of premium commensurate with the risk on the policy.

|

Best Practices in Predictive Modeling

Predictive analytics has many benefits to offer the insurance industry, and there are important factors that will enable a carrier to be successful deploying predictive models:

• Good, clean data is key to delivering results: Carriers collect considerable amounts of data. When this data is properly structured, carriers can build accurate predictive models. In the past, larger carriers were better able to take advantage of predictive models because of their large datasets and ample modeling and IT resources. However, there are now several third-party data and modeling resources available to all carriers that allow the small and midsize companies to compete effectively. Experts who have been using predictive models for decades testify to the importance of accessing representative and sufficiently sized data samples. Mark Gorman, CEO of The Gorman Group agrees, “Data, especially external data, has become even more important. Carriers have to access third-party data in order to be competitive.”

• Predictive analytics enables consistent decision-making: Top carriers utilize predictive models to make data-driven decisions consistently across their underwriting staff. These models allow carriers to use evidenced-based decision-making rather than relying solely on heuristics or human judgment to assess risk. Underwriters thrive when they have reliable information to aid their expertise, and accurate predictive models have proven to outperform traditional decision-making techniques.

• Predicting future performance increases profits: Incorporating predictive analytics allows a carrier to anticipate the future performance of a policy, which is vital to managing risk and removing the uncertainty for what they insure. Gorman emphasizes the fundamental changes that occur when adopting predictive analytics: “This represents a complete evolution of the industry in terms of people, process and technology. At the same time, what we know from a revenue standpoint is that carriers who adopt predictive analytics earlier gain market share—profitable market share. Carriers who aren't moving in that direction are dealing with adverse selection.”

• Using predictive analytics helps carriers know what they insure: Using predictive analytics enables underwriting, claims and marketing to make evidenced-based decisions that lead to better risk selection, pricing and claims handling. With recent data and technology advancements, carriers of all sizes can compete on a more level playing field. Predictive analytics offers a better, faster and cheaper method to allow underwriters to assert more control over their portfolio's performance. It is possible to counteract the unpredictable Homeowners' market, gain a competitive advantage in underwriting and improve operating performance. While it is impossible to control the weather, it is possible to understand more about the homes carriers insure and price them accordingly using predictive analytics.