In September the CAS sponsored the Chief Data Officer Forum Insurance 2016 in Chicago, an event that gathered insurance-specific chief data officers (CDO), chief analytics officers (CAO) and other senior analytics professionals. The purpose of the Forum was to foster discussions about the evolving demands of big data and analytics in the insurance space. Highlighting data’s growing importance to insurance companies and services, the event also explored topics such as fraud analytics, predictive modeling, customer data management, disruptive innovation and data quality.
One panel, “Data Scientists and Actuaries — Partnering for Success to Maximize Business Outcomes,” explored the growing intersection between actuarial and data science roles at insurance companies. Chris Monsour, FCAS, MAAA, vice president, analytics, CNA Insurance Companies, shared his personal experiences as both an actuary and data analyst, illustrating how the two functions can best collaborate within an insurance company.
Monsour serves as a subject matter expert for The CAS Institute, a new CAS subsidiary that recently launched the Certified Specialist in Predictive Analytics (CSPA) credential. The CSPA will recognize expertise in the specialized area of predictive analytics for property and casualty insurance applications. The CAS Institute will also provide resources and a broader practice community for many specializations of the insurance industry’s quantitative professionals.
Eric Huls, FCAS, SVP, Chief Data Science Officer, AllState, also presented a keynote At the September’s Forum. In “Building Internal Bridges and Creating a Culture of Mature Analytics,” Huls talked about data as a business function, explaining that the value of data is not just from having it, but from using it. Huls also defined the reality of becoming an analytics-driven organization within today’s insurance market, and the change management that can foster a deeply embedded culture of analytics.
Actuary on the Street: Adler’s Take on the CDO Forum
Avraham Adler, FCAS, CERA, MAAA, isn’t a CDO, but he was intrigued by the content of the CDO Forum Insurance 2016. Adler, who is senior vice president, GC Analytics™ for Guy Carpenter in Chicago, attended the event and shared some of his thoughts from the day with Actuarial Review.
Actuarial Review: What were some of your take-aways from the event?
Avraham Adler: For me, some of the take-aways from the event were that to perform efficient, high-throughput data analytics, especially on large data sets, requires planning and investment in the underlying hardware infrastructure as well as the software. Also that there is so much information “in the wild” that for an organization to keep track and remain current may require more than one person in the data analytics support role — separate and distinct from classic IT.
AR: What are some of the main challenges facing data analytics professionals right now?
AA: I would say that one of the main challenges now is the relative newness and thus somewhat chaotic nature of the field. So many ideas and products are competing for the limited attention and resources of the data professional that it is hard to decide where to focus. Should investment be in hardware? Do we invest in people with knowledge and push off the inanimate (hard/software) investing?
AR: How does being an actuary play into data analytics skills?
AA: I think one of the main advantages being an actuary offers is training and experience in understanding the data with which one is working. Not every data set can be equally considered “a bunch of numbers.” Knowing the context in which one is working helps the actuary understand the limits of the data and thus the limits of any analysis to which it will be subject. Moreover, knowing the context of the problem and the data helps the actuary determine which method may be best under the circumstance. Similarly, understanding the context of the data and the problem allows the actuary to recognize when a result, even if valid, is inappropriate.
AR: What do actuaries need to know to succeed in data analytics? How do you feel about professional education for this expanding field?
AA: I believe that the FCAS needs to know the basics of predictive modeling, and, more importantly, where to go to develop any needed expertise, but not that every FCAS needs to be a predictive modeling expert. Also, not every person who wants to spend their career in predictive modeling for the insurance industry wants to become an FCAS. Having the CAS, through The CAS Institute, develop, maintain, and test a curriculum geared to predictive modeling for risk ensures that the syllabus and testing will be recognized throughout the industry as comprehensive, rigorous and fair.
