During your underwriting career you have worked in both a “slow evidence” and an e-data environment. How would you describe the transition?
For many years we really didn’t see any change. Underwriting shops used the application, paramedical exams, fluids, physical measurements, EKG, MIB, MVR, medical records, inspection reports, etc. to complete the risk assessment. Then we saw the introduction of prescription drug data, but for most underwriters, this evidence simply became another item to underwrite. In some instances, it replaced the need for medical records.
Today we see a range of e-data sources being used in both traditional underwriting and automated underwriting engines or in some stage of R&D. I would categorize them as …
- Well-established e-data – Rx, MIB & MVR
- New e-data – clinical lab data (CLD), health claim information, criminal records, credit based mortality risk scoring, other medical based mortality risk scoring, etc.
- Emerging e-data – propensity models, electronic health records (EHR), wearables data, facial analytics, etc.
While evidence traditionally was used to assess the risk, today the data is applied in multiple ways from input into risk assessment and algorithms to alert misrepresentation and determination of non-e-data requirements.
What underwriting R&D initiatives are you currently working on?
We’re working on studies to validate new data such as CLD and health claims information and its protective value as underwriting evidence. We’re also looking at EHR data providers and other health data.
We think CLD is a real game changer in terms of providing digital medical information on individuals. It’s a step towards electronic health records, which is the holy grail for instant underwriting decisions.
Each additional e-data source provides protective value with no degradation in the speed of the decision. In addition, it allows carriers to get smarter about when “slow evidence” is really required. The new e-data sources and removal of fluids/exams can, however, cause movement between risk classes of the existing insured/applicant pool.
Most of the new, emerging e-data sources can be used to predict/stratify mortality. The challenge is to determine what the remaining or incremental value is when combined with other data sources.
Do you see a common approach to how e-data sources are used?
There is not one e-data strategy that will fit all carriers. The strategy will be driven by a carriers’ risk appetite, client base, distribution channel, desire to be an early adopter, etc.
The SCOR Life R&D team and Velogica team have been instrumental in evaluating these new data sources for our carriers. We have completed both protective value and mortality improvement studies and provided valuable input to carriers in the design of accelerated underwriting programs and use of e-data.
What can underwriters do to develop skills to better understand and be engaged in the new frontiers of underwriting?
Most life companies today are involved to some degree in accelerated underwriting. Some are strategic adopters; some are all in. My suggestion is to volunteer for related project work, attend industry events and network with underwriters and vendors active in this space. In my experience, I’ve learned the most when several disciplines were involved in the project.
For those looking to branch out, I would also recommend developing skills in programs like Excel and Tableau and gaining knowledge of predictive modeling and artificial intelligence.
Any final thoughts?
The changes we see in underwriting create exciting new opportunities in our profession, but there’s no typical way to prepare for these hybrid roles. Make sure you communicate your desire to try something new, then be proactive and take the initiative wherever you can.