Why insurance pools losses
Imagine that you own a small bakery. You can budget for flour, rent and wages, but a fire could destroy equipment that would take years of profit to replace. Insurance lets you pay a known premium in exchange for the insurer’s promise to meet covered losses under the policy. You still need to understand the limits, exclusions and any deductible, which is the part of a covered loss you must bear yourself.
How the pool helps. An insurer collects premiums from many policyholders. Most will not experience the same serious loss during the same period. Their contributions help finance the claims of those who do. The insurer also needs enough money for expenses and other costs, so the expected cost of claims is not the whole premium.
Predicting a group. The law of large numbers helps explain why a larger pool of comparable, sufficiently independent risks produces more reliable estimates of average loss. It cannot tell the bakery owner whether a fire will happen next Tuesday. It also does not promise that the insurer’s actual claims will equal its forecast in any particular year.
Suppose a teaching example gives 1,000 comparable shops a 1% annual chance of one $20,000 loss each. The expected number of losses is 1,000 × 0.01 = 10. Expected claims are therefore 10 × $20,000 = $200,000, or $200 per shop. These are averages used for planning. Exactly ten shops need not have a loss, and $200 is not a quoted insurance premium.
Why the mix matters. If all those shops are on one flood-prone street, one flood could damage many at once. Adding nearby shops increases the size of the pool without removing that shared exposure. Insurers must consider how risks are related as well as how many they insure. Pooling also does not require equal premiums: a shop with a different risk of loss may contribute a different amount.
Comparing experience fairly. A larger portfolio can have more total claims even while its average loss becomes more predictable. Compare similar exposures over comparable periods: 1,000 properties observed for a month do not provide the same time at risk as 1,000 properties observed for a year. Copying records adds no new experience. A meaningful estimate needs information about how often losses occur and how large they are.
When a question mentions a larger insurance pool, look for improved predictability across the group. An answer promising that each member becomes safer, or that every member should pay the same premium, is making a different claim.
NAIC, A Regulator’s Introduction to the Insurance Industry, pp. 6–7, 9.