- Aquaculture, Equine & Livestock
- Architects & Engineers
- Aviation & Aerospace
- Consumer Goods & Services
- Education & Public Entities
- Entertainment & Leisure
- Financial Services
In the Numbers: Analytics Support Human Judgment
December 10, 2014
As pervasive as analytics and technology are today, it’s hard for many people to grasp that we didn’t always have the capability to capture and analyze the volume of data that we do today. In the commercial insurance industry, advanced analytics is becoming main stream. But that is a relatively recent innovation.
As a math major in college, I always liked “unsolvable” problems. I have always believed that we can solve part of the problem or at least understand it better. That philosophy has really shaped my career in analytics.
Like a lot of people I know, my path into insurance was accidental. At the time I graduated, math majors had few career choices – teaching or becoming an actuary were two of the most common. I chose actuarial work because it sounded interesting. And it really is. My job at XL Group perfectly suits my skills and interests.
At XL, I lead a team that develops and implements advanced analytical tools which help our underwriters better understand risks. What’s so fascinating about analytics? Exploring and analyzing new risk data provides constant surprises in what’s possible. XL Group specializes in insuring complex risks and we have a mandate to push analytics beyond what is traditional in the commercial insurance space. When XL formed its Strategic Analytics department in 2010, we envisioned developing advanced analytical models which use new risk data to support underwriting.
Challenging conventional wisdom
Part of what we do is challenge conventional wisdom. An example of how analytics can change people’s thinking comes from the mortgage industry. When my parents sought their first home loan about 50 years ago, lenders commonly tracked how long applicants worked for their employers. Lenders’ conventional wisdom in that era held that a long-tenured worker was a better risk. Then analysts started looking at the data and the performance of home loans. It became clear that a person who has a good track record of finding new jobs has more marketable skills and will have an easier time finding new employment should the need arise than does the long-serving employee, which makes them a better mortgage risk. Using analytics to debunk conventional wisdom has enabled mortgage providers to be more effective at offering mortgage terms which, in turn, increases profits. In order to benefit from this kind of innovation, the banking industry had to challenge an old way of thinking and embrace the new way.
Innovation is part and parcel of Strategic Analytics at XL Group. We are constantly striving to find new ways to understand risks so we can better underwrite them. There is a long history of conventional wisdom in how insurance companies underwrite risks and conventional thinking has dominated up until now. But it won’t work in the future as businesses evolve and new risks emerge. At XL Group, we believe that one of the best ways to outperform is to use new approaches to analyzing risk. And my team is helping to do that today.
There are common misperceptions about analytics. For example, many people are afraid that models will replace underwriting judgment. That is not the case, at least not at XL. We’re not building models to replace underwriters. Human judgment will always be critical to our underwriting. Models can’t access business, interpret and apply model insights and manage portfolios.
I often get asked “What if the model turns out to be wrong?” The model will be wrong some of the time because no model is perfect. We just want it to be less wrong than what we used before. Our goal is to have incremental improvement, not perfection.
Improving on what we’ve developed is a matter of staying open to new ideas. The easiest thing in the world to do when somebody offers a new idea is to say “No.” We make a conscious effort to find the positives in an idea and build on those. If we want to be successful, we have to find new ways to do things.
About the AuthorKim Holmes is senior vice president of Strategic Analytics at XL, where she is responsible for expanding XL Group’s use of leading-edge risk analytics throughout the company. She joined XL in 2011 from Endurance Risk Solutions, where she was senior vice president and chief actuary. Her career experience includes a variety of actuarial and underwriting roles.