And somewhere, a carrier with better data already knows exactly who they are.
Here is the thing actuaries know but most boards have never fully sat with: your pricing model does not price your policyholders. It prices groups they happen to belong to. A 35-year-old SME owner in Chennai is priced on the average loss experience of 35-year-old SME owners in Chennai. Not on how she actually runs her business, manages her cash flow, or handles risk.
Within every pricing segment you have, there are good risks who are overcharged and bad risks who are undercharged. The good risks are subsidising the bad ones. This has always been true. What is new is that someone can now tell the difference.
The adverse selection problem is not theoretical. It is happening right now in every segment where a competitor has better behavioural data than you do.
How it plays out
The mechanism is simple and it is brutal. A carrier with better data identifies the good risks within your pricing segment — the safe drivers in your motor book, the well-managed SMEs in your commercial property portfolio, the healthy individuals in your group health scheme — and offers them a correctly priced, lower premium. They leave. Your book degrades. Your average risk quality worsens. Your loss ratio climbs. You raise rates. More good risks leave.
This is not a hypothetical future scenario. In the UK motor market, telematics-based insurers spent a decade quietly extracting the best young driver risks from traditionally priced books. The traditionally priced books got worse. The telematics books got better. The gap between the two is now large enough to be a structural competitive disadvantage, not a pricing adjustment.
In Southeast Asia, Grab has GPS, braking, acceleration, and routing data on every driver across its network. That is a more granular motor risk signal than any policy application form has ever captured. The carriers who establish data partnerships with Grab and platforms like it are building a pricing edge that competitors cannot access without making the same partnerships. The ones who don't are pricing on proxies while their competitors price on behaviour.
Telematics did not make traditional motor insurance unprofitable. It made the residual book — the one left behind after the good risks switched out — progressively harder to price at a sustainable margin.
This is coming for more than motor
Health is next. Discovery's Vitality model — operating through AIA across Singapore, Hong Kong, Thailand, and Japan — links health insurance premiums to verified health behaviours. Gym attendance, step counts, health screenings. Members who engage show better loss experience and stay longer. The carriers running Vitality programs are not just pricing health better. They are attracting and retaining healthier policyholders while their competitors get the ones who chose not to engage.
Commercial property is following. Satellite imagery and aerial analytics now allow carriers to assess individual property risk at a resolution that annual surveys cannot approach — roof condition, vegetation proximity, drainage infrastructure. Carriers using these signals are identifying within-segment risk variation that traditional commercial underwriting prices as homogeneous. The ones who are not are charging the same premium to a well-maintained facility and a deteriorating one, and the market will eventually separate them.
The question is not whether behavioural data improves pricing. It demonstrably does. The question is whether you are building the data infrastructure to use it before your competitors finish building theirs.