Insure Catastrophes, Not Inconveniences

One principle, most of the answers

Insurance causes more confusion per dollar than almost anything else you can buy, and one idea clears up most of it: insure what would ruin you, and cover the rest yourself.

Here's why that works. Insurance is a losing bet on average, and it has to be. Your premiums pay for other people's claims, plus the insurer's costs, plus its profit. Over a lifetime you will almost certainly pay in more than you get out.

That makes it worth buying for exactly one kind of risk: the loss you could never absorb — the one that would break you (The First $1,000 Does the Most Work: how much buffer you actually need follows how a shortfall cascades; this is the same thing at a scale you don't recover from). For that, losing money on average is a fair price for putting a floor under how bad it can get.

For anything you could pay for out of savings, you're paying that same premium and getting nothing you needed. Insuring small risks is a slow, quiet leak — the same economics as $230 Billion a Year Is the Price of Inattention: the fees worth moving accounts over, just wearing protective clothing.

What to insure and what to skip

Insure the big things.

Liability — being sued has no natural ceiling, which is exactly what insurance is for. Your car and umbrella liability limits deserve more attention than most people give them.

Your income — this is the most under-bought cover relative to its risk. Roughly 1 in 4 of today's 20-year-olds will be disabled for a year or more before they reach retirement age. Fewer than half of civilian workers have disability cover through work, and only about 1 in 5 adults hold a policy of their own.¹ For most households the ability to earn is the largest thing they own, and Borrowed More Than You Meant To? the 3 biases behind it explains why it goes unprotected: we assume it won't be us.

Life — term insurance, sized to what the people depending on you would need. When Whole Life Is Sold, Not Bought settles the term-versus-whole-life question.

Health — a serious medical bill is the classic amount nobody can absorb.

Your home — enough to rebuild it.

Cover the small things yourself. The phone, the appliance, the trip, the scraped bumper. This is where extended warranties live (Buying a Car: negotiate the price and the financing as 2 separate deals covers the finance-and-insurance office at the car dealership), and your savings are the better insurer: they charge no premium, apply no excess, and cover everything.

There's a useful link here. Every dollar of savings raises how much you can cover yourself. So as your buffer grows, it makes sense to raise your excess — you accept more small risk and pay a lower premium, which is a straightforwardly good trade once you could comfortably pay the excess. And you can drop the small add-ons entirely.

The audit

Once a year, do the Shopping Auto Insurers Saved a Median of $461 a Year: what else is negotiable sweep on your cover. List every policy and add-on with its yearly cost. Check the big five above for anything you're missing. Check the small stuff for anything you're wasting money on. Then re-shop whatever can be shopped (Shopping Auto Insurers Saved a Median of $461 a Year: what else is negotiable explains why loyalty is priced against you).

What the audit usually turns up is the principle upside down: warranties on gadgets, a low excess, comprehensive cover on an old car — and no disability cover, with thin liability limits. Paying retail to insure what savings would cover for free, while the things that could actually ruin you go uncovered.

The takeaway

Insurance is for catastrophes. Cover the losses with no ceiling — liability, your income, life cover if people depend on you, health, your home — and let your savings handle the inconveniences. Raise your excess as your buffer grows, skip the warranties, and check it once a year with every premium stated as a yearly figure. Losing money on average is fine, as long as what it buys you is a floor under ruin.