A client in his late sixties came in last fall holding a letter from Genworth. His annual premium on a policy he'd carried since 2003 was going from $3,180 to $5,420, a 70% jump, the third increase since he bought it. He wanted to know what I'd do in his shoes. I told him the truth: I would have asked the same question, and there is no clean answer.
That letter is not unusual. It's the long-term care insurance industry in miniature. A product seniors were told to buy in their fifties has, over twenty-five years, become one of the most volatile lines in personal finance. The carriers that built the market either left it or are still digging out from underwriting assumptions that didn't survive contact with reality.
If you're shopping for LTC coverage in 2026, or sitting on a legacy policy with a premium that keeps climbing, the landscape looks almost nothing like it did when the product was first pitched to your generation. Here is what I tell clients now.
The market that collapsed
At the peak in the early 2000s, roughly 125 carriers wrote standalone long-term care insurance. Today, fewer than fifteen do, and the active list is short: Mutual of Omaha, New York Life, Northwestern Mutual, MassMutual, Thrivent, OneAmerica, Securian, and a handful of others. Genworth, which once dominated the market, stopped selling new traditional policies in 2019 and now services legacy books while seeking premium increases from state regulators almost continuously.
What happened is straightforward in hindsight. The original policies assumed lapse rates (the percentage of policyholders who would drop coverage before claiming) that resembled life insurance. Life insurance lapses run high; LTC policyholders, it turned out, almost never lapse. They keep paying because they intend to use the coverage. The actuaries also underestimated how long claimants would draw benefits and overestimated investment returns on reserves during a decade of near-zero interest rates. Three bad assumptions stacked on top of each other.
The result: carriers either exited the market, sought rate increases of 50% to 100% on legacy blocks, or did both. Genworth alone has settled multiple class-action suits over rate hikes. In those settlements, policyholders were typically offered a choice: pay the higher premium, accept a reduced benefit (shorter benefit period, lower daily benefit, or weaker inflation rider) and keep paying the old premium, or take a paid-up nonforfeiture benefit. None of those options feel like a win. They are damage control.
If you hold a traditional standalone policy from before 2010, assume more rate increases are coming. The blocks are still underpriced relative to claims experience, and state regulators have generally approved hikes when carriers can document the gap.
What you can actually buy now
There are still four real products on the shelf in 2026, and they sit in different places on the cost-versus-flexibility spectrum.
Traditional standalone LTC. Use-it-or-lose-it coverage with a daily or monthly benefit, an elimination period (the deductible measured in days, typically 30, 60, 90, or 180), a benefit period (usually 2, 3, 5, or 6 years; lifetime is essentially extinct), and an inflation rider (3% or 5% compound). Premiums are not contractually locked. They can rise subject to state insurance department approval, and the recent history says they will. For a healthy 60-year-old, expect $2,500 to $4,500 a year for a middle-of-the-road policy. For a 70-year-old, double that if they'll write you at all.
Hybrid life insurance with an LTC rider. This is now the dominant new sale in the LTC space. You buy a permanent life insurance policy, and a rider lets you accelerate the death benefit to pay for qualified long-term care. If you never need care, your heirs get the death benefit. If you do, the policy pays for care up to the face amount, sometimes more if the rider includes an extension-of-benefits feature. Premiums are typically fixed and guaranteed not to rise. Lincoln Financial's MoneyGuard, Nationwide's CareMatters, OneAmerica's Asset Care, and Pacific Life's PremierCare are the products most commonly quoted. The economics work best as a single premium or short-pay structure (one-time $100,000 or ten years of $12,000) rather than lifetime payments.
Hybrid annuity with an LTC rider. Less common, but worth knowing about. Pension Protection Act of 2006 made qualified LTC withdrawals from these annuities tax-free. The annuity grows on a tax-deferred basis, and if you need care, the LTC benefit can be a multiple of the account value. These tend to suit people who already have annuity money parked somewhere and want to repurpose it. They are not a great fit for new money.
Short-term care insurance. A one-year benefit, sometimes 360 days, with simpler underwriting and lower premiums. Useful as a gap product if you have some assets but want a buffer against the most likely scenario, a recovery period after a stroke or fall that doesn't extend into years. I've recommended these for clients in their seventies who could no longer qualify for traditional coverage but wanted something in place.
A fifth category, critical illness riders attached to life insurance, gets pitched as LTC adjacent. It is not. Those riders trigger on a diagnosis (cancer, heart attack, stroke) and pay a lump sum. They don't cover custodial care or the slow grind of dementia. Don't substitute one for the other.
The triggers nobody explains until you claim
Every qualified LTC policy pays benefits when a licensed professional certifies one of two conditions. First, you cannot perform two of six activities of daily living without substantial assistance: bathing, dressing, eating, toileting, transferring (getting in and out of a bed or chair), and continence. Second, you have a severe cognitive impairment, usually defined as requiring substantial supervision to protect yourself or others, that lasts at least 90 days.
The ADL trigger is the gate most claimants pass through. The cognitive trigger matters most for dementia patients whose bodies still work but whose judgment doesn't. Read your policy's exact language. Some older policies require three ADLs instead of two. Some define "substantial assistance" more narrowly than others. A carrier I won't name once denied a claim because the claimant could technically dress herself if someone laid out the clothes. The policy required "hands-on" assistance, not "standby." That distinction was buried on page 31.
The 90-day elimination period is the deductible most policies use, and it is measured in days requiring care, not calendar days. A client of mine spent four months in and out of a rehab facility after a hip fracture before her policy paid its first dollar. She'd assumed 90 days meant 90 days from the day she filed. It didn't.
What it costs to skip the insurance
The Cost of Care Survey, which the industry treats as gospel even after Genworth's own retreat from the market, is now run by CareScout, a Genworth subsidiary, and its 2025 edition was released in March 2026. The national medians: $6,200 per month for assisted living ($74,400 a year), and $315 a day for a semi-private nursing home room, which CareScout reports annually as $114,975 and works out to about $9,580 a month. In-home help ran $35 an hour — and note that CareScout has folded the old homemaker and home health aide categories into a single "non-medical caregiver" rate, so a quoted home health aide figure from an older survey is not measuring quite the same service. Memory care runs higher. Costs in the Northeast and on the West Coast run 30% to 50% above those medians. In Connecticut, where I practice, a semi-private nursing home room is closer to $15,000 a month.
The planning numbers most often cited come from the Department of Health and Human Services, and the versions that circulate at free seminars are not the ones HHS actually publishes. HHS ASPE projected in 2022 that 56% of Americans turning 65 today will develop a disability serious enough to require long-term services and supports, and that 22% will live with one for more than five years.
HHS publishes no median duration at all. That matters, because a median is the figure most often quoted back to me, and it is usually somewhere near two and a half years. The closest real number is the average duration of disability for men, 2.5 years, against 3.6 for women, which is a different measurement of a different thing. What HHS does report cuts against the sales pitch: averaged across everyone turning 65, people need 0.8 years of paid care, because 55% will never use paid care at all. Families absorb it instead.
So run the math on the tail, not the typical case, because the typical case is not what ruins anybody. Among people who do develop a significant disability, care runs 5.1 years on average for men and 5.6 for women, and about 4% of older adults use five years or more of paid care. A five-year dementia care stay at $11,000 a month is $660,000. To self-insure with any confidence, I tell clients they should have $300,000 to $500,000 earmarked specifically for care, not their general retirement portfolio, but a sleeve they can mentally wall off. Below that threshold, the math for insurance gets more compelling. Above $2 million in liquid assets, self-insuring usually wins.
When to buy, and when the window closes
Underwriting tightens fast with age. The sweet spot for new applicants is mid-50s to early 60s. Premiums are still reasonable, and most chronic conditions haven't surfaced yet. After 70, most carriers will not write a new traditional policy, and the few that will charge premiums that are difficult to justify against the same money invested. After 75, the standalone market is essentially closed, though some hybrid products will still consider applicants up to 80.
Underwriting itself is more invasive than most clients expect. Blood draws, paramedical exams, MIB and prescription database checks, and a cognitive screen, usually a short word-recall test administered by phone or in person. Cognitive concerns are the fastest decline. Diagnoses of Parkinson's, multiple sclerosis, recent stroke, or memory issues will close the door at almost every carrier.
The tax treatment helps a little. Premiums on federally qualified LTC policies count as a deductible medical expense, subject to age-based caps. The IRS adjusts these annually; for 2026 the caps are $1,860 for ages 51 to 60, $4,960 for 61 to 70, and $6,200 for over 70. The deduction only matters if your total medical expenses exceed 7.5% of adjusted gross income, which they often do for retirees.
Alternatives worth knowing about
A handful of paths sit alongside or instead of traditional coverage.
State Partnership Programs let qualifying LTC policies protect assets dollar-for-dollar against Medicaid spend-down rules. If you buy a Partnership-qualified policy with $200,000 in lifetime benefits and exhaust it, you can keep $200,000 in countable assets and still qualify for Medicaid. Most states participate. The mechanics are technical; an elder-law attorney can map it for your state.
Continuing care retirement communities (CCRCs) function, in part, as prepaid long-term care. You pay an entrance fee (typically $200,000 to $800,000 depending on the contract type and the community) and a monthly fee, and the contract obligates the community to provide higher levels of care (assisted living, memory care, skilled nursing) as you need them. Type A "life care" contracts offer the strongest LTC protection. Type C "fee-for-service" contracts shift more risk to you. The financial review on a CCRC contract is significant and deserves outside professional eyes before signing.
Veterans benefits can fill part of the gap for those who qualify. Aid and Attendance pension benefits can pay over $2,000 a month for a single veteran needing care, and the VA also operates state veterans homes. The application process is slow and document-heavy but worth pursuing for any veteran or surviving spouse.
Medicaid itself remains the default long-term care payer in this country, covering more than half of all nursing home residents. Qualifying generally requires assets under $2,000 (with significant variation by state and program) and triggers look-back rules on prior asset transfers. Medicaid planning done five years before a crisis looks different from Medicaid planning done five months before a crisis. The earlier you start, the more options you have. For state-specific guidance, I've written separate breakdowns of Florida, California, Texas, New York, and Michigan Medicaid rules.
What I'd do in 2026
If you're 55 to 65, healthy, and have $1 million to $2 million in retirement assets, look hard at a hybrid life-with-LTC policy on a 10-pay structure. The premium is guaranteed, the benefit isn't lost if you don't need care, and you've taken the rate-increase risk off the table.
If you're 65 to 75, the math gets harder. Most clients in this band end up with either a small standalone policy (just enough to defer Medicaid spend-down by a few years), a hybrid funded with existing IRA money via a 1035 exchange, or a deliberate self-insurance plan with a designated sleeve of the portfolio.
If you're over 75 or have health issues that will close standard underwriting, short-term care, state Partnership planning, and Medicaid pre-planning with an elder-law attorney are the practical paths. The honest answer is that the optimal time to buy passed without you knowing it. That's not a moral failing; the product was sold poorly to a whole generation.
If you hold a legacy policy and the premium just jumped, do not drop it without running the comparison. A rate increase is not the same as a bad deal. Even at the new premium, the coverage may still be cheaper than self-funding, especially if you've already paid in for a decade. Ask the carrier in writing for the reduced-benefit options (most are required to offer them) and run the numbers with someone who isn't paid on the result.
The industry that sold these policies didn't get the actuarial math right. Your job is to make the best decision you can with what's actually on the shelf now, not the version that was promised in 2003.
