An insurer cannot know which policyholder will claim. Its business does not require that knowledge, because the predictability it depends on emerges only at the level of the group.
Individual outcomes are unknowable
Whether a particular house burns or a particular driver crashes in a given year is close to unpredictable, and no amount of underwriting changes that.
What can be estimated is how often such events occur across a large number of comparable situations, which is a different question with a stable answer.
Insurance is built on that difference, converting an individual uncertainty into a group average that can be priced.
Why scale stabilises the result
As the number of similar and independent policies grows, the proportion that claim converges towards the expected rate, and variation around it shrinks relative to the total.
A small insurer covering few policies faces wide swings from year to year, while a large book produces results close to expectation most of the time.
This is why growth is not merely commercial for an insurer. Scale directly improves the reliability of the mechanism the business depends on.
Correlation breaks the arithmetic
The stabilising effect requires losses to be largely independent. Events that strike many policyholders simultaneously violate that condition.
Windstorms, floods, earthquakes and widescale economic events produce claims that arrive together, so the pool provides no offsetting.
Insurers manage this by limiting concentration in any one area or peril, and by transferring part of the exposure to reinsurers who pool across regions.
What segmentation is actually doing
Rating factors sort policyholders into groups expected to claim at different rates, so each group pays closer to its own expected cost.
Without segmentation, lower-risk policyholders subsidise higher-risk ones and tend to leave, which raises the average risk of those remaining.
Taken far enough, segmentation approaches individual pricing and undermines pooling, which is why many markets restrict the factors that may be used.
Why capital exists behind the premiums
Premiums are set against expected losses, and actual losses in any year can exceed expectation. Capital is what allows the insurer to pay anyway.
Regulators require capital sufficient to withstand severe scenarios, because the promise being sold is worthless if it fails in exactly the conditions that trigger it.
The cost of holding that capital is part of every premium, alongside expected claims and expenses, which is why premiums exceed average claims in the long run.