The Retirement Risk Nobody Wants to Model

Why most retirement plans go quiet on this one question

Ask most retirement calculators — and a fair number of professionally built financial plans — what happens if you need extended nursing home or memory care for three or four years starting at 80, and you'll get silence. Not a bad answer. No answer. The tool wasn't built to ask the question.

That's not really a technical limitation. Long-term care is hard to model for three separate reasons, and it's worth naming them honestly instead of blurring them into one vague "it's complicated": the cost is large and genuinely uncertain, the probability and duration vary enormously person to person, and — the uncomfortable one — most people, understandably, don't want to spend planning time imagining themselves needing help with basic daily tasks. All three are real. Only the first two are reasons a model should be cautious. The third is a reason people avoid the topic, not a reason the risk is small.

The result is a quiet mismatch: a huge share of retirement plans carry meaningful exposure to a cost that could reshape the whole plan, and nobody ever built a way to show the owner of that plan what it looks like if it happens.

Why this isn't the same kind of risk as market risk or longevity risk

It's tempting to lump long-term care in with "other retirement risks" and assume a plan that's robust against market downturns and a long lifespan has implicitly covered it. It hasn't, because the shape of the risk is different.

Market risk and longevity risk are both about duration — can the money last as long as it needs to, given ordinary variation in returns and lifespan. A plan absorbs that kind of risk by having enough margin spread across many years. Long-term care risk isn't about duration in the same sense. It's a large, concentrated, unplanned expense stacked directly on top of a retiree's normal cost of living, arriving with little warning, for a period that could be a matter of months or could stretch past a decade. It doesn't erode a plan gradually the way inflation or a soft market does — it can hit like a second household's worth of spending, layered on the first, at exactly the point in life when there's the least room left to adjust by working more or cutting back elsewhere.

A plan can pass every stress test you throw at it for market crashes and long lifespans, and still have zero answer for the one scenario that actually breaks retirements in the real world.

This is why "my success rate is 90%+" doesn't automatically mean long-term care is covered. A high success rate against simulated market returns and mortality is a real, meaningful signal — it's just answering a different question than "what happens to this plan if a multi-year care need shows up in year twelve of retirement." Both questions matter. Conflating them is how a plan ends up feeling more prepared than it actually is.

"Unmodeled" doesn't mean "doesn't matter"

There's a natural but mistaken inference people make: if a planning tool doesn't ask about long-term care, and the plan still looks solid, the risk must be small or already baked in somewhere. It isn't. Unmodeled means exactly what it says — the tool never asked the question, so the plan's output carries no information about it either way. That's different from a risk that's been evaluated and found to be minor, and it's different again from a risk that's been deliberately accepted with eyes open.

Being silently exposed and being knowingly exposed are not the same position, even though the dollar risk is identical in both cases. The first means a real event could arrive as a total surprise, with no plan in place and no time to build one. The second means a decision already got made — self-fund it, insure it, accept the risk within existing margin — and everyone involved knows what that decision was. Getting from the first state to the second doesn't require solving long-term care. It just requires actually looking at it once, on purpose, instead of the topic quietly falling through the cracks of every tool that happened not to ask.

The honest range of ways people actually handle it

There's no single correct answer here, and any framing that pretends otherwise is selling something. The real options, each with real tradeoffs:

• Self-funding from a larger cushion. Building enough general portfolio margin that a multi-year care need can be absorbed without a separate insurance product. This works best for larger portfolios where the cost, while painful, doesn't threaten the whole plan — and it means carrying the full risk yourself rather than transferring any of it.
• Traditional long-term care insurance. A real risk-transfer tool, but one that's gotten harder to buy affordably over time — premiums can be underwritten based on current health, they're not guaranteed to stay flat, and qualifying gets more difficult (or impossible) the later someone waits, especially after a health event. It's most straightforward for people who buy in relatively early and in good health.
• Hybrid life/LTC policies. Life insurance with a long-term care rider or accelerated benefit, so premiums aren't "wasted" if care is never needed — the policy still pays a death benefit either way. The tradeoff is usually a smaller LTC benefit pool than a dedicated standalone policy would provide for the same premium.
• Medicaid as a last resort. Medicaid does cover long-term care, but only after assets are spent down to program limits, and the planning window for doing this in an orderly way (rather than a crisis) is measured in years, not months, because of look-back rules on asset transfers. It's a real backstop, not a strategy to lean on casually or late.
• Accepting the risk and building general flexibility. Not every plan needs a dedicated LTC strategy. Some plans have enough slack — home equity, flexible spending, family support, a smaller expected gap — that the honest choice is to accept the exposure rather than pay for a formal product against it. That's a legitimate choice too, as long as it's made knowingly rather than by default.

None of these is obviously "the smart one." They trade certainty of cost (insurance premiums, paid whether or not care is ever needed) against uncertainty of cost (self-funding, which could turn out to cost nothing or could turn out to be the single largest expense of retirement). The right tradeoff depends on health, family history, the size of the portfolio relative to potential care costs, and honestly, risk tolerance for a very specific kind of risk.

A worked example: an unplanned care need in year ten

Consider a plan that looks genuinely solid on paper — a comfortable success rate against market and longevity risk, spending that fits comfortably within the portfolio's capacity, retirement age and Social Security claiming both reasonably optimized. Ten years into retirement, one spouse needs an extended stay in memory care.

Nothing about the household's original planning was wrong. The market did roughly what was modeled. Nobody overspent. The plan simply was never asked the one question that turned out to matter most in year ten. That's the specific failure mode this risk represents — not a bad plan, a plan with an honest, unaddressed blind spot.

What actually helps

The goal isn't eliminating the risk — nobody can do that. It's making sure the plan has a real answer instead of silence:

• Know your own exposure before you need to. The single highest-leverage step is simply finding out whether your plan has any margin for this at all, while there's still time to do something about it — buy insurance while still insurable, build a larger cushion, or have the family conversation about what happens if care is needed.
• Match the response to the portfolio size. Very large portfolios can often self-fund comfortably. Very tight plans may need Medicaid planning started early, with real lead time. The broad middle is exactly where the insurance-versus-self-fund tradeoff is genuinely close, and worth working through deliberately rather than defaulting into either one.
• Act early if insurance is part of the plan. Traditional LTC insurance gets more expensive and harder to qualify for with age and health changes — this is one of the few retirement decisions where waiting has a real, compounding cost.
• Build in flexibility everywhere else. A plan with real slack elsewhere — flexible discretionary spending, a paid-off home, family capacity to help — absorbs an LTC shock better than a plan that was already spending to the edge of its margin under ordinary assumptions.

None of this requires solving long-term care perfectly. It requires not letting it stay invisible.

What this doesn't capture

Even a plan that deliberately accounts for long-term care exposure is working from a projection, not a guarantee. Actual care costs, care duration, and care needs vary enormously by individual, region, and circumstance, and no model can tell you in advance whether you personally will need care, for how long, or of what kind. A state-aware classification of exposure is a real signal about margin — it's not a forecast of what will happen to any one household.

Common questions