Risk. Reward. Balance.

Few industries attempt to forecast risk and reward more fervently than the insurance industry. An obvious domain for implementing predictive technologies where analysis, forecasting and planning play integral roles in its members’ success, or failure. Notably, this domain continues to be dramatically under served by modern analytical tools in an environment where the smallest of savings can amount to billions of dollars for each company. Insurers are faced with a host of decisions including the types of programs to offer in their “book” (e.g. auto insurance, home, life products, etc.), the amount of premiums to charge for these programs, the underwriting criteria used to accept applications for particular programs, the amount of reserves to set aside for potential claims, and the evaluation and settlement of claims. All these decisions have a significant impact on a company’s bottom line. Nevertheless, the insurance industry has historically approached its decision-making process by taking past amounts realized and adding arbitrary safety, or contingency, factors.  The standard comment in the industry when talking about current and historical analytical evaluation capabilities is, “it’s like driving by looking in the rear-view mirror”. It is no wonder that, although insurance companies are making substantial gains on their investments, they are losing money on their book.

PredictionProbe has a complete business intelligence software suite to bring Predictive Technology to the insurance market. This solution will provide the needed insight and understanding to make the right decisions and avoid costly hunches and guesses and ensure your company’s rewards are far greater than its risks, which is the name of the game.

Insurance Products and Services

Improving the prediction of outcomes has enormous potential benefits in the insurance industry. First, insurance products by definition are financial vehicles designed and priced based on the prediction of future losses. Guessing high on predicted losses means higher premiums and lost sales; guessing low means reduced profits. Secondly, fleets of underwriters and clerks analyze incoming applications and must predict if any are submitted fraudulently and whether any of the risk should be reinsured. Guessing incorrectly could result in incurring losses from applications that should not have been accepted or by being exposed from a lack of reinsurance. Third, claims staff needs to predict how much should be paid for covered losses, based both on true costs and on what can be negotiated. Incorrect loss cost predictions literally give away profit and makes it difficult for a company to compete long term. Fourth, teams of actuaries perform detailed analyses to identify how much cash must be “reserved” to pay for future losses. These “reserves” are kept in highly liquid, low return investment vehicles so the incentive is to reserve the minimum required. However, a variety of regulatory and rating agencies vigilantly watch reserve development and can aggressively penalize a company whose losses regularly exceed the reserves. And finally, insurance companies have a variety of needs that cut horizontally across industries: predicting customer behaviors; evaluating management performance; and testing various strategies before extensive costs are incurred.

Applications and Opportunities

The opportunity for implementation of Predictive Technology is infinite, but to demonstrate crystal clear examples, here are 14 uses of PredictionProbe’s Predictive Technology in the insurance domain. These examples represent opportunities in claims, underwriting, product design, product profitability improvement and even Six Sigma implementations. Product distribution methods vary widely from “shrink-wrapped” applications to customized solutions developed in conjunction with strategic partners.

Product Design and Management

• Insurance program modeling software for the creation of new products, expanding existing products to new markets, and modeling changes to existing programs for pricing or profit analysis.
• Software that allows for an understanding of the sensitivity of profit to each rating criteria creating a deeper understanding of an insurance company’s book of business.

Actuarial Analysis

• Reserve adequacy monitoring and self-correcting software that reduces reserve inadequacies allowing for better information management, lower capital costs and fewer regulatory audits.
• Loss forecasting software that increases loss estimates for better capital management.

Claims Decision Support

• Comparative negligence assessment decision support software that optimizes decision consistency and reduction of loss payments.
• Repair facility fraud identifier that improves the prediction of fraudulent activity by service provider.
• Bodily injury evaluation software that improves prediction of claims or litigation settlement based on the predicted value of injury, reducing litigation expenses and overall costs.
• Predictive fraud software that augments existing relationship-based fraud management tools reducing fraudulent payments.

Underwriting Decision Support

• First-look toolkit that identifies fraudulent applications as they come through the door.
• Compliance toolkit that identifies program compliance percentages and loss ratios.
• Reinsurance book modeler that allows an underwriter to review a book of business and identify the benefits and risks of reinsurance.
• Acquired book evaluation software that allows for the review and analysis of a book of business for potential acquisition.

Horizontal Prediction Management

• Staff evaluation software that allows management to control for all other factors in evaluating claims, product or underwriting to identify staff success associated with defined products.

• Customer behavior software that creates a behavior model to be used to test new products, rates, or approaches.

Insurance Business Intelligence Product Line

• Reserve Monitor and Self Corrector
• Insurance Prediction Toolkit and Program Modeler
• Predictive Fraud Modeling
• Executive Information System Based on Six Sigma or Incumbent Methods

Process Description

Potential Predictive Technology insurance products starts with a Product Description. This description includes:

• List of all variables known to impact the process
• Target to be solved for, or what is being predicted
• List of all available uncertain variables which may or may not influence the target
• Explanation of the Outputs Required
• Set of complex mathematical equations and algorithms