CIA (e)Bulletin/(e)Bulletin de l'ICA
Past Issues |  
October 2017

An Actuary in an Analytics World

Print Print this Article | Send to Colleague

By Jeffrey Baer, FCIA

Not long ago, I realized that actuarial science was a highly specialized form of predictive analytics—an insight that didn’t occur to me throughout my undergraduate education in actuarial science, or in six years of studying for actuarial exams. It wasn’t until I had been leading an analytics team for over a year that it dawned on me: my two professional identities, actuary and analytics professional, are cut from the same cloth.

Practical Applications

Consider the two most common actuarial functions: pricing and valuation. We’re taught to apply specific actuarial techniques to determine indicated rates—but any analytical process that produces rates that are not excessive, inadequate, or unfairly discriminatory could be a valid pricing approach. Determining reserves that are sufficient to cover unpaid claims and expenses is a forecasting problem that can be attacked using a host of different predictive modelling methods.

As manager of advanced analytics at Economical Insurance and Sonnet Insurance, I am privileged to lead a diverse and accomplished team of analytics professionals with backgrounds in statistics, operations research, geospatial analytics, and machine learning. I’m responsible for collaborating with business stakeholders in areas like sales, underwriting, marketing, and claims to develop data-driven solutions to business problems, and leading research into innovative applications of predictive modelling within our organization. My job description doesn’t require an actuarial background. Yet time and again, I’ve found that being an actuary serves me well in the analytics world.

Actuarial Concepts in Non-actuarial Domains

In some cases, my team takes traditional actuarial concepts and applies them to non-actuarial domains. For example, we recently engaged our claims department to address an interesting business problem at Economical. We have a great team of 20 field adjusters who are on the ground supporting our customers during the property claims process. But to ensure we achieve strong customer service levels while balancing adjuster capacity, travel time, and other constraints, we have to determine the size of the territory for which each adjuster is responsible. To do this, we can apply actuarial concepts to predict the number of property claims that will require field adjusting over the next year. We can then use these projections as an input to an optimization model that accounts for our objective and constraints and determines optimal adjuster territories.

In other cases, I apply the same skills I learned as an actuary to analytical solution design. Like actuarial science, predictive analytics requires prudence in assumption setting, expertise in programming and data wrangling, and strong business knowledge. From a soft-skills perspective, problem solving, teamwork, and the ability to communicate technical concepts to non-technical audiences are all necessary competencies.

Improving Customer Experience, Internal Processes, and Business Decisions

With these skills in place, applications of predictive analytics to improve the customer experience, internal processes, and business decisions are endless. We’ve used predictive analytics to determine target growth areas, triage accounts for underwriting, detect opportunities for claims recoveries, and optimize the selection of properties for risk inspections.

A Bright Future

The future for the combined actuary/analytics professional is bright. Predictive analytics is now being introduced into the core actuarial syllabus, which is a great development. I look forward to welcoming the next generation of hybrid actuary/analytics professionals to Economical!

Jeffrey Baer, FCIA, is the manager of advanced analytics for Economical Insurance and Sonnet Insurance.


Back to CIA (e)Bulletin/(e)Bulletin de l'ICA

Share Share on Facebook Share on Twitter Share on LinkedIn