The care bill most retirement plans quietly ignore

Roughly half to two-thirds of people who reach 65 will need some form of paid care, and residential care runs well into six figures a year. It is the largest single financial risk in later life and the one people are least likely to have costed.

You

yrs
yrs
$
Investments and savings, plus home equity if you'd be willing to use it.
$
Pension and annuity income. This offsets the fees before you touch capital.

The care

$
Costs vary enormously by region — by a factor of five or more between the cheapest and most expensive areas. Use a local figure if you have one.
yrs
Most stays are short; a minority run five years or more. The tail is what does the damage.
%
Estimates range from about 56% to 70% depending on how "care" is defined. Women's odds run meaningfully higher than men's.
% p.a.
Care is labour, and labour costs rise faster than general prices. Using headline inflation here understates the bill.

Insurance option

$
$
yrs

Means-tested support

$
Most systems only step in once your assets fall below a set level. Set your own jurisdiction's figure.
% p.a.

What it costs, and when

Insure or self-fund?

How fast the money goes

Assets from the day care starts, with fees running and your income offsetting part of them.

Self-funding With insurance Support threshold

If it runs longer than expected

The average stay is not the risk. The five-year stay is.

Making sense of this one

This is an uncomfortable subject, and skipping it is what makes it dangerous. A few minutes here is genuinely worth it.

The scale of the thing

Somewhere between half and two-thirds of people who reach 65 will need some form of paid care before they die. Most of those episodes are short and manageable. A minority run for years, and residential care costs well over $100,000 a year in most developed countries — often considerably more.

What makes this different from every other retirement risk is the shape of it. Most risks are moderate and likely. This one is unlikely-to-even-odds and potentially enormous, which is exactly the profile that wrecks plans built around averages.

An example: Anne, 66

Anne has $800,000 and $34,000 of continuing pension income. She assumes she might need care from around 83 — seventeen years away.

Care costing $115,000 today will, at 4.5% care-cost inflation, cost about $243,000 a year by the time she needs it. Care inflates faster than general prices because it's mostly wages, and wages rise faster than goods.

Her pension covers part of that, so roughly $209,000 a year comes out of capital. Her savings will have grown by then, but at that burn rate they don't last long.

A two-and-a-half year stay is survivable. A five-year stay is not. That's the real question — not "what will care probably cost" but "what happens to my spouse if I need five years of it".

Reading the insurance comparison

The table shows whether the policy or self-funding comes out ahead on expected value. Be careful with that number, because expected value is the wrong lens for a catastrophic risk.

Insurance is rarely a good bet on average — if it were, insurers wouldn't sell it. What you're buying is the removal of an outcome you couldn't survive. So the question isn't "will I come out ahead" but "would a five-year stay ruin us". If the answer is yes, a policy can be worth having even at a poor expected value. If your assets are large enough that even a long stay leaves you comfortable, you're wealthy enough to be your own insurer and the premiums are money you don't need to spend.

Practical notes

General information only — not financial, insurance, legal or tax advice. Care costs, eligibility rules, asset thresholds, look-back periods and the treatment of the family home differ enormously between countries and even between regions, and change frequently. The insurance comparison ignores policy exclusions, elimination periods, benefit inflation riders, and the real possibility of premium increases. Anyone actually making this decision should get advice specific to their jurisdiction.