To truly be Data Scientists other industries require marketing and business statisticians to broaden their software skills beyond SAS and R to Hive and Python. We are starting to see the insurance industry ask more of their actuarial hires in their software and database skills.
Historically, actuaries haven’t had the need for sophisticated programming skills. Many of the vendors to the risk industry provided custom built and custom supported software. AON/Symphony with eSolutions, Marsh & McLennan with Stars, Milliman with ReservePro and Towers Watson with Radar and MoSes for example. An academic study of actuarial software use found Excel was the preferred product of 98% of actuaries with 95% of them using it daily. Excel was considered, “absolutely wonderful, computationally limited, difficult to control.” Excel’s flexibility is also its curse as this same study found errors, usually human cutting and pasting or selecting the wrong cell, in 100% of the spreadsheets reviewed. The limitation of 256 columns and 65,636 rows means that large data sets can’t be handled by the much-loved Excel.
For the same reason other industries are turning to software able to manage unstructured data, we are seeing much broader utilization and demand in the insurance industry for more sophisticated data management software like SAS and R. These packages have great graphical output but a steeper learning curve if the use is intermittent as probable for actuaries. SAS has great data management and great support but is quite expensive. R is an open source tool so essentially free but the support can be spotty.
“Actuaries who have SAS or R, or experience with even more sophisticated tools like Pig or Python will find many more opportunities in their job search, and will be more innovative once they are at those companies using these resources,” says Rory Hauser, Executive Recruiter at Smith Hanley Associates. “I’m seeing more and more clients show specific interest in advanced programming skills. With data sets becoming ever larger, organizations want to find better ways to analyze their data and programs like SAS are increasing in popularity.”