The Research Behind MOAT
Independent research from leading insurers, reinsurers, and researchers on how continuous behavioural and wearable data sharpens mortality risk assessment, improves persistency, and reduces costs.
Reclassifying Mortality Risk
WTW × Klarity · 2025
Reinsurer validationRead the source
- Traditional risk factors — cholesterol, blood pressure, BMI, tobacco use, family history — only tell part of the story, and routinely miss individualized signals like resting heart rate, heart-rate recovery, sleep, and activity levels, causing applicants to be misclassified.
- Adding wearable-derived activity data and the Klarity risk score reveals that 34% of “second-best” nonsmoker risks and 16% of residual standard-class risks actually look as healthy as the best nonsmoker class — while 6-13% of applicants in the top two preferred classes actually belong in a lower, non-preferred class.
- Step count and activity duration correlate with mortality even more strongly than several traditional underwriting markers, including cholesterol, BMI, and family history of heart disease or diabetes.
- Though trained mostly on UK data, the model performs effectively on a US population, and continues to strengthen as more US insured data is added.
Can a Digital Health Programme Reduce Mortality Rates?
Swiss Re Institute · 2026
Reinsurer researchRead the source
- Across roughly 30,000 participants tracked from January 2022 to December 2023, those engaged with the programme for five months or more increased their physical activity by an average of 22% versus their pre-programme baseline.
- Sustaining that activity gain into old age translates into a future mortality-rate reduction of roughly 3-13% across participants.
- The biggest gains came from groups usually left behind by digital health tools: 60% of participants had an elevated BMI, 20% were 50+, and 24% were sedentary before joining — yet the over-60s and previously-inactive users showed the highest engagement and the largest projected mortality improvement.
- More app usage tracks directly with more exercise and better projected mortality outcomes — engagement itself is the mechanism, not just enrollment.
- The estimated mortality reduction corresponds to potential term life claims-cost savings of SGD 2-8 per adult per month per SGD 500,000 sum assured.
Physical Activity Data from Wearables
Munich Re · 2025
Reinsurer validationRead the source
- Daily step count segments mortality risk across age, smoking status, BMI, and gender, and in a full model is more predictive of mortality than nearly every traditional measure — second only to age itself.
- Mortality risk for people walking under 5,000 steps/day is more than four times higher than for those walking 15,000+, with the steepest gains occurring between 5,000 and 7,000 steps.
- Active “current” smokers doing 7,000+ steps/day have better mortality outcomes than “never” or “former” smokers doing under 5,000 steps — behaviour can outweigh a static risk label.
- A step-count-augmented underwriting model outperforms a traditional lab-based full-underwriting model at identifying preferred-risk applicants, and step count ranks as the second most important mortality predictor overall, behind only age.
- Physically active people currently deemed “uninsurable” by traditional criteria have mortality risk comparable to insurable people with low step counts — pointing to an expanded insurable pool for carriers using this data.
Digital Tools for Health and Wellness in Insurance
OECD · 2024
Policy & regulatoryRead the source
- Rising healthcare costs — more claims, higher medical and drug costs, more hospitalisations and diagnostics — are a central pressure point for health and long-term care insurers, making prevention economically attractive, not just a wellness nice-to-have.
- There is growing evidence that chronic disease can be prevented through healthier lifestyles, particularly by addressing behavioural risk factors like smoking, alcohol use, and poor diet.
- Wearables and mobile apps with built-in sensors can monitor physical, mental, and emotional states, and some tools use AI on large datasets to flag disease risk early, making users more aware of — and more likely to change — their own health behaviour.
- Insurer-run digital wellness tools are generally a “win-win”: better health for policyholders and reduced claims for insurers, with many programmes borrowing directly from behavioural economics — using rewards, incentives, and occasionally premium discounts to nudge activity.
- This trend has expanded beyond health and long-term-care insurers into life insurance, as life insurers make their products more relevant to younger, digitally-native customers.
- Data privacy and security, safety, data quality, and overall effectiveness remain key issues insurers need to manage when adopting these tools.
The Impact of a Lifestyle Behaviour Change App on Healthcare Costs
Agachi, Mierau, van Ittersum & Bijmolt — Preventive Medicine · 2024
Peer-reviewed studyRead the source
- Quasi-experimental difference-in-differences design with propensity-score matching across 15,506 participants, using real healthcare claims data from 2015-2019 rather than a controlled trial setting.
- Total healthcare costs fell about 4.9% in the first year after app launch and 5.3% in the second year, averaging roughly a 5% annual reduction per participant.
- The reduction — including lower general-practitioner spending — corresponds to average savings of about €154 per user.
Want the full technical validation packet?
We'll walk you through the underlying models, the data sources, and how they map onto your book.