RGA’s Actuarial Validation of Irix® Risk Score 3.0
An independent reinsurer study proves the heightened value of our latest model, with and without Medical Data.
To facilitate this actuarial study, we provided RGA with a dataset following more than 42 million insurance applicants since 2005, encompassing nearly 250 million exposure years and 2 million deaths. RGA found that:
- Risk Score 3.0 with Prescription Data and Medical Data delivers improved risk selection by broadening the distribution of scores while maintaining the predictive power in each score bucket. This was found across sectors with both low and high risk (life insurance vs. final expense) and held true over all durations (significant predictive power continued to exist even over 11+ years)
- Risk Score 3.0 shows improved segmentation compared to Risk Score 2.2, even when both models are limited to the same Prescription Data-only input, so the model itself is superior.
- In cases where both Prescription Data and Medical Data were available, the marginal effect of adding Medical Data was significant. In about 8% of such cases, adding Medical Data resulted in score changes >1. In addition, adding Medical Data increases the likelihood of receiving a non-zero score to approximately 90%
RGA also noted that unlike competing models that merely rank risk, the output of the Risk Score 3.0 model is relative mortality; if the score doubles, mortality also doubles, making Risk Score an intuitive tool carriers can use to spot both risks and opportunities.
Please enter your email to keep reading. Don’t worry—we promise we won’t start hounding you with calls and emails. We just need to make sure our valuable content is delivered to the right audience.
