March 2022 – Super-Charging Past Data

WARNING: This paper contains opinions that may shock, annoy, dismay and generally offend valuation actuaries!

With this headline warning we did not deliberately set out to cast aspersions on the science of setting disability life reserves nor the professionals that are charged with this responsibility.

Who are we with only our claims management expertise to encroach on another discipline?

Needless to say, we felt with all humility that we had some views to share that might be beneficial to valuation actuaries, pricing actuaries, group underwriters and claims experts in the management of Long Term Disability (LTD) business. So while we know this will be somewhat provocative, we hope all can be open to other views and stimulate constructive dialogue.

The Edge of Actuarial Science
For many years at The Claim Lab, we have worked on the edge of actuarial science trying to make a case for individual reserving or at least better individual informed portfolio reserving for LTD claims.

We learned a long time ago that data items such as:

• diagnosis 
• age 
• gender

Are just not enough to predict the duration of an LTD claim, and also, we understand that some carriers are not yet using diagnosis in their reserve calculations. 

In addition, the use of historic data can let us down when the unexpected happens. At the beginning of last year such an event occurred with an unprecedented pandemic that has sent the life and disability insurance industry reeling.

The problem with using an historic average duration is that the average is only as good as the data from which it is calculated. For that average to change, the influx of ‘new’ claims data has to run its course through the lifecycle of the claim. In other words, by the time you spot something is wrong, it’s too late to do anything about it. 

It is possible to get greater insight at an earlier stage from claims data, and some enlightened claim managers have been screaming ”we could see this coming!!!” but at the time those termination tables were not proven to be incorrect. Typically, termination studies are conducted only annually, and sufficient claims experience needs to be seen to ensure credibility.

The Influence of the Pandemic
In our client’s claim data, we could see the influence of the pandemic across all aspects of the claim’s lifecycle. We saw it first in Short Term Disability (STD) claims, about 18 months ago, when claim volumes were not necessarily higher, but the claims were getting more complex. It was no longer just a rotator-cuff or a back sprain, but in addition there was “My employer is not taking correct precautions in the workplace, I’m anxious about returning to work” OR “I have a very bad lower back pain, my company keeps laying off more and more employees, my pain is unbearable….!”

Then gradually the STD claims started moving ever closer to becoming LTD claims, more and more claims were transitioning. Some actuaries were waiting, standing by their reserving methods until more time passed. However, these new LTD claims did not close, termination rates were getting worse, and the claims management were asked to explain what was happening….“What have you done wrong?”

Well, what we have been witnessing is a tidal wave of psychosocial issues caused by the pandemic, aside from the COVID-19 virus itself. Claims operations had no way of quantifying the added complexity of these claims at an early stage. When they did raise the alarm, actuarial science was not nimble enough to react.

Now going forward, maybe we can learn that we need a more sensitive measure of the fluctuations across a block of claims, which are not just based on ‘diagnosis’, ‘age’, and ‘gender’. We should also calculate reserves on an individual claim basis (or at least consider modifying factors) rather than at an aggregate level, and we need to have the confidence to use early-stage claim data to quantify claim complexity based on a full spectrum of influencing factors.

Is There a Better Way?
The industry is not doomed, losses will be recouped with revised pricing, and we are now getting to work with enlightened actuaries who are interested in the work we are doing. However, moving forward perhaps the voice of the claim managers saying, “I told you so!” will not be so easily ignored! 

Perhaps we can collectively, across disciplines, use more early-stage claim data to “predict” better, which will lead to more precise and timely estimation models. 

To learn about the power of Psychosocial Data in claim management, email us at info@claimlab.org

Thank you!

Ian Bridgman
Executive Director