Case: Sam's family protection checklist

With a fixed budget, spend it on the losses the family can't absorb, before the small ones. Rank each option by the size of its loss against what savings could cover, then add premiums from the top until the budget runs out.

For Sam, the biggest losses are the two incomes (1,860,000 and 1,500,000), then belongings and flooding the neighbours at home, plus the war add-on because debris hit the building last year. Those four:

7,800 + 6,000 + 1,500 + 900 = 16,200 a year, or 1,350 a month — inside the 18,000 budget, with 1,800 to spare.

Adding the accident cover (2,400) would push it to 18,600 and break the budget. Note what order does not depend on: how likely a seller says a loss is. Rank by how much a loss would hurt, not by a sales pitch.

A policy's exclusions list what it never pays for, whoever is to blame. In Sam's table, the war_risks column is the one to read: 5 of the 7 policies exclude war risks (both life covers, both home versions and the accident cover). Only home_war_option is marked covered — and it pays only for damage to belongings, not life or injury.

An exclusion doesn't make a policy worthless. Life cover still pays for the cases it does cover, such as illness, and the income loss is too big to leave open. But you have to know the gap before you sign:

• ask each insurer whether a war-risk option exists and what it costs;
• check which places and kinds of damage an add-on actually includes.

One row reads differently: a gadget warranty's war_risks is n/a, because it insures a breakdown, not a cause like shelling.

The job_loss row is the odd one out: it has no premium and no insurer, but it's the only big loss with a chance filled in — 30% a year. Turn that into an expected yearly loss:

30% × 93,000 = 27,900 a year

That's larger than all the premiums in the table combined. A loss this likely and this big can't be bought away — it's the most probable hit the family faces, and no policy covers it. The only defence is the cushion.

So the spare budget, and the money not spent on gadget covers, goes to savings, not to one more policy. Sam has 10,000; three months of his pay is 93,000, leaving 83,000 to build. This is the module's line between insurance and saving: insure the losses you can't absorb and can buy cover for; self-insure the rest with a cushion.

Sam's worry — what if both devices break at once? — is the standard argument for a gadget cover. The answer is the one from the first lesson: self-insure small, rare losses, and the cushion does the job of many little policies at once.

The two covers cost 2,525 a year for 880 of expected repairs, so 2,525 − 880 = 1,645 a year is pure overpayment on average. Put that 1,645 into the cushion instead and within a few years it comfortably covers even a double breakdown (4,000 + 6,000 = 10,000) — and the same savings also stand ready for a job loss, a leak or anything else small and rare. One pot, many uses, no premium.

Compare ratios to confirm: the laptop cover is 1,400 / 480 ≈ 2.92 times its expected repairs, worse than the phone's 2.81. Neither is a good deal when the loss fits in savings.