Risk pooling: your premium is mostly other people's claims

What follows is an argument about risk pooling, and about where the received version of it stops being true.

The basic mechanism

A large group each contributes a small, affordable amount, and the collected fund pays the few who suffer a large loss. No individual could fund a house fire, and a hundred thousand households together can fund several a year comfortably. Your premium is not held for you; it is spent on claims made by strangers, plus expenses and capital costs.

Expecting to get back what you put in misunderstands the transaction entirely.

Why the law of large numbers matters

Individual losses are unpredictable, but the aggregate loss of a large group is predictable within a reasonable range. That predictability is what allows a premium to be set at all. The larger and more homogeneous the pool, the tighter the prediction and the lower the margin needed for uncertainty.

This is also why very small or very unusual risks are expensive to insure.

What breaks the pool

Pooling fails when losses are correlated, meaning many members claim simultaneously, as in a flood or a windstorm. It also fails when losses are certain rather than probable, which is why chronic conditions and wear and tear are excluded. Insurers manage correlation through reinsurance, geographic spread and exclusions.

In practice, recognising these two failure modes predicts most of the exclusions in most policies.

Adverse selection and why insurers ask questions

If everyone paid the same premium, low-risk people would leave and high-risk people would stay, raising the average cost. That spiral is called adverse selection, and underwriting questions exist to prevent it. This is why a question that feels intrusive is usually a pricing question rather than a moral one.

It is also why non-disclosure is treated so seriously: it moves cost onto everyone else in the pool.

Moral hazard and why excesses exist

Once insured, people may take marginally less care, which raises claim frequency. Excesses, no-claims discounts and co-payments all exist to keep some financial consequence with the policyholder.

These are not penalties; they are what keeps the premium payable for everyone. A policy with no excess and no discount scale would cost noticeably more.

Some of this will suit you and some will not, and that is the point.

What this means for you

Insure the losses you could not absorb, and self-insure the ones you could, because pooling small predictable costs adds administration without adding value. Judge a policy by what it pays for the catastrophic case, not by whether you got your money back. A year with no claim is a good year, not a wasted premium.

How much risk your household should retain is a personal financial question and one for a regulated adviser where the sums are significant.

The takeaway

Insure what would ruin you, not what would annoy you. That is what pooling is good at.

Pick the one that costs you least, and let the rest wait.