The world is constantly changing, and as an actuary, I probably view these changes differently than most people. In my world, all of the advancements in new and innovative technology that have made our lives more convenient also present more complex risks.
From an insurance perspective, new and more prevalent technologies like mobile payments and drones require more complex risk-management tools. Previous methods for quantifying and managing risk — such as using past data to price insurance products — may no longer be sufficient. At the same time, the digital revolution, led by smartphones and wearable devices, is giving us more data than ever before. Insurers need to embrace and mine the increasing volume of data, finding new techniques to evaluate and produce insights.
The good news is that a lot of new data is readily available — the not-so-good news is that insurers and their analytical teams may not know what to do with this data. Everything from the sheer volume of data to the nature of how it is stored and processed can make it hard to sift through and find information that will be useful.
Due to this influx of data, the industry has seen the partnering of actuarial work with data science to perform predictive modeling. Despite the arrival of new techniques, however, insurance remains a highly specialized and highly regulated industry — those handling the data in insurance companies need to fully understand the business context in which it lives. One variable of data may represent something that is not legal or socially acceptable to actually use in practice, or, data may say something that makes no sense at all – for example, that women with red hair have more auto accidents (when anyone can dye their hair). So those working on data analytics teams need to have a strong sense of causality when evaluating data, knowing how it plays into the larger business context of the problem they're trying to solve.
|Bridging the communications gap
There is no prescribed composition for an effective data analytics team — it can have a mix of data scientists, actuaries, statisticians, and others. Each professional brings something to the discussion, and increasingly the "team" approach to analytics results in success. However, those same professionals need to understand each other's perspectives — they need to be able to speak the same language in order to communicate and collaborate. Ideally, members of the team will have a certified set of predictive analytics skills, which can help set a standard and bridge the communication gap that exists.
For employers, this lack of common "language" in the predictive analytics environment can also affect their recruitment. Position titles such as "data scientist" or "modeler" do not have a consistent description or industry standard. Last year, when the Casualty Actuarial Society (CAS) conducted market research with insurance company executives on the subject, employers cited recruiting/hiring as one of their greatest challenges in predictive analytics. In fact, 76 percent of those surveyed noted that a certification would be beneficial to employers seeking to hire specialists in predictive modeling.
Becoming a certified predictive analytics specialist
This is one of the many reasons that The CAS Institute, a subsidiary of the CAS, recently launched its Certified Specialist in Predictive Analytics (CSPA) credential. The credential, created for data professionals with several to many years' experience, requires that candidates demonstrate evidence of applied knowledge in predictive analytics by passing a series of four assessments. The program draws from the history and strength of the CAS, whose high-quality educational standards and credentialing programs for actuaries have been recognized globally for over 100 years.
The curriculum of the CSPA credential is overseen by an expert panel comprising industry specialists working in predictive analytics. The four required assessments cover:
- The fundamentals of property and casualty insurance;
- How data works, including the forms it can take;
- How to present and work with data, including building models; and
- How to apply these skills to a real-life scenario.
The final assessment also asks candidates to complete data analysis and a report based on an assigned scenario. The candidate is required to integrate and apply all knowledge from the previous three assessments in order to achieve success.
Final projects will vary so as to reflect real-world-type predictive analytics scenarios. For example, one project might have candidates working to improve claims department operations, such as identifying potential high-severity claims, or controlling claims department costs. A marketing-focused project could ask candidates to improve sales through methods such as matching product offerings to customer type, or targeting new or optimal customer segments. CSPA candidates may also use their predictive analytics skills in scenarios involving underwriting, pricing, or even operations. This "case study" project helps round out the CSPA curriculum by testing the candidates' ability to use their predictive analytics skills in the workplace.
Related: Insurance industry education is more than just letters after your name
|A new professional community
CSPA credential holders are also required to complete an ethics course and adhere to a standard of professionalism and code of conduct, something not previously required of those in analytics roles.
After traveling all over the U.S. sharing information about our new CSPA credential with employers, we can say that the response has been overwhelmingly positive. Employers are enthusiastic to see a program that can provide professional education and certification to members of their team who have previously been without these types of dedicated resources. Employers now have a reference point when they decide to add predictive analytics professionals to their staff. The CAS Institute also provides its members with a professional community, where those working in this specialized field can connect.
Ultimately the expansion of predictive analytics within the insurance industry has opened doors for new opportunities to improve business performance. In order to maintain momentum and keep up with changes, predictive analytics teams need to make sure they are well-equipped and collaborating effectively to adapt to new technologies and new data. It's only through the improvement and standardization of analytical skills, coupled with the willingness to learn, that we will remain ready to respond to the technological (and societal) changes that still await us.
Nancy Braithwaite, FCAS, MAAA, CPCU, is a Second Vice President and Actuary in the Excess Casualty Department at Travelers Insurance Co. She currently serves as president of the Casualty Actuarial Society (CAS). Opinions are the author's own.She can be reached at [email protected]. The CAS can be reached at (703) 276-3100.
Want to continue reading?
Become a Free PropertyCasualty360 Digital Reader
Your access to unlimited PropertyCasualty360 content isn’t changing.
Once you are an ALM digital member, you’ll receive:
- Breaking insurance news and analysis, on-site and via our newsletters and custom alerts
- Weekly Insurance Speak podcast featuring exclusive interviews with industry leaders
- Educational webcasts, white papers, and ebooks from industry thought leaders
- Critical converage of the employee benefits and financial advisory markets on our other ALM sites, BenefitsPRO and ThinkAdvisor
Already have an account? Sign In Now
© 2024 ALM Global, LLC, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to [email protected]. For more information visit Asset & Logo Licensing.