Have workers compensation insurance companies, who rely on predictive modeling and smart systems, lost their touch for underwriting?

Evidently I’m stuck on discovering the role underwriting plays in todays modern workers compensation insurance company and I’ll go ahead and apologize for that now! In the last post I touched on the use of predictive models and smart system underwriting models that insurance companies have developed or paid to have developed and use to filter preferred risk clients into their pricing systems. Lets continue with this discussion but focus todays blog a little more on the role workers compensation underwriting plays in todays insurance company.

Insurance companies are in business to make money. Sitting in front of your TV you may have come up with the idea that an insurance company wants to become your best friend. That they send someone over as soon as you call to help remove your tree off your neighbors car or house. Sure they’ll pay claims, well most of them. (That’s another topic all together!) But one thing you should always remember, an insurance company is in business to make money! For a stock company, they need to perform for their stockholders and return a profit to their bottom line for distribution in the form of dividends. For a mutual company, they need to perform for their policyholders and administration so they can stay in business and continue to provide service to policyholders.

For many large workers compensation insurance companies their first line of defense from adverse risk selection is their predictive model. It’s a tool that allows an insurance company to use their own data to assist them by identifying those above average clients who present the best possible opportunity to make a profit for their insurance company. Predictive modeling is a system developed using statistical techniques to analyze and manage current and historical data into a model capable of predicting future outcome. Think about it, for an insurance company, an organization that constantly struggles for profit, the ability to predict the future from a given set of current and past circumstance, the ability to identify the best clients based on their own past experience would be a gold mine!

Predictive models utilize a system of algorithms working together to discover similarities in data. And one thing an insurance company has at its finger tips is data! In a 2013 survey conducted jointly and published by Earnix, Inc. and Insurance Services Office, Inc. entitled “2013 Insurance Predictive Modeling Survey” they indicate that 82% of the respondents to their survey use predictive modeling in one or more insurance products they provide. It shouldn’t surprise you that larger insurance companies are more vested in developing and implementing predictive analysis. The survey goes on to indicate that 52% use predictive modeling for underwriting. 

What’s all this have to do with workers compensation insurance?

Insurance companies are constantly on the look for methods to improve performance. Identification of risk, identification of exposure to loss, improved pricing based on risk characteristics, improved risk selection and consistent underwriting profit are all areas within the workers compensation insurance product delivery system that are touched by predictive analysis.

A common belief is that traditional underwriting techniques have become an inadequate method of delivering consistent underwriting results. When in fact a successful predictive modeling system will incorporate the exact defined, repeatable exposure variables that underwriters have always worked with to pinpoint good and bad risk characteristics.

When looked at as an enhancement to those otherwise mundane underwriting tasks, previously worked by a human, you start to buy into the belief that predictive modeling can be a useful tool. Why not remove or mechanize those basic tasks and free up your human underwriting team so they may concentrate on the more subjective part of underwriting? Let your predictive model or smart underwriting system filter out undesirable risks, improve processing of new and renewal policies and let your system smooth out the bumps that human underwriters will inevitably allow to creep into your book of business.

Never forget that for an insurance company the underwriting process is the foundation for securing, managing and maintaining workers compensation clients. Incorporation of predictive modeling systems within workers compensation underwriting departments can help streamline work flows freeing underwriters to work more closely with those risks that require special attention. 

Used as a tool and backed up with tried and true underwriting skills, performed by competent workers compensation underwriters, predictive modeling will no doubt meet its goal of enhancing underwriting performance. Allowing an insurance company to better perform during the up and down of market cycles should be good for all involved.

 However, I have concerns. Will predictive models take over the job of underwriting? Since more and more insurance companies are jumping on the smart system band wagon, who’s training the human underwriters? Will we see fewer trained workers comp underwriters in the future? Will they be trained to rely on their smart systems and predictive models rather than basic and advanced underwriting techniques? What about the rules that govern workers compensation? Who will oversee their compliance? Will new underwriters even know how a workers compensation policy is constructed, how and when to add endorsements to the policy, how a workers compensation policy is rated or the individual rating factors that go into developing the correct premium? Will they even know how to properly classify a business and what about the use of proper classification codes? Or will they just be there to blame the “system” when things go wrong? 

Has big brother really arrived?

Hope this help you out and thanks for reading!