The expat health-insurance market has a design property worth stating before any number. It is affordable in precisely the years you are least likely to claim, and it reprices or refuses you in precisely the years you are most likely to. That is not a malfunction. It is the product working as priced. The actuary who built it did nothing wrong. The brochure that sold you the move simply declined to show you the curve the actuary used.
Lead with the curve, because the curve is the argument.
The table
Indicative cost of single expat health cover in Thailand, 2026. The ranges are wide because they span Thai-domestic plans (cheaper, narrower) and international plans (comprehensive, portable), and because this market prices on individual health. Every figure is dated and should be re-quoted before it is trusted.
| Age band | Monthly, typical | Annual, single | What is happening |
|---|---|---|---|
| 50–59 | ~$60–200 | ~$0.7k–2.4k | The product looks cheap. You rarely claim. |
| 60–64 | ~$120–350 | ~$1.5k–4k | First steepening. Still “manageable”. |
| 65 | — | Pacific Cross ~$2.5k–4k; international ~$5k–8k | The split between domestic and international widens hard. |
| 70–74 | ~$167–1,000 | ~$2k–12k | Entry now restricted at several insurers. The spread is the tier, not the applicant. |
| 75+ | ~$500–900+ | ~$6k–12k+ | New cover largely closed. Renewals reprice five-fold-plus territory. |
And the rule that table encodes, sourced separately: premiums rise roughly 30–50% from 50 to 60, double or triple from 60 to 70, and can rise five-fold or more after 75. Read those three multipliers as one sentence. The cost of the thing accelerates fastest exactly as the probability of needing it does.
The acceptance curve, which is the other half
Price is only one wall. The other is whether the door is open at all, and it closes at different ages by insurer.
| Insurer | New applicants to | Renews to | The catch |
|---|---|---|---|
| Cigna Global | New applicants to No limit | Renews to Lifetime | The catch The most open door, and priced for it — comprehensive, international. |
| ACS | New applicants to No limit | Renews to Lifetime | The catch Senior cover written from 60 with no upper cap. |
| Pacific Cross | New applicants to 75 | Renews to 99 | The catch Latest affordable entry; renews far, but reprices up the age curve the whole way. |
| IMG | New applicants to 74 | Renews to 75+ (Senior Plan) | The catch Terminates at 75 unless you held cover continuously from before 65. |
| BCBS Global Solutions | New applicants to 74 | Renews to 83 if in force | The catch No new policy over 74; only continues one already in place. |
| VUMI | New applicants to 74 | Renews to — | The catch New entrants stop at 74. |
| Luma (Thai-domestic) | New applicants to 70 | Renews to 99 | The catch The earliest door and the narrowest cover — but it renews as far as any of them. |
Lifetime renewal is real, but conditional on continuity from before 65. In this set the cheaper the door the earlier it shuts, which is a property of these seven insurers rather than a rule of the market.
Source: Insurer terms (Cigna, Pacific Cross, IMG/BCBS Global Solutions via expatfinancial, VUMI, Luma, ACS) · checked 2026-08-16
The market is not uniformly shut at 70. It is a set of doors closing at staggered ages, and in this set the cheaper the door, the earlier it shuts. Read that as a description of these seven insurers rather than as a rule of the market: across the wider set this site audits, published entry caps run from the mid-fifties to eighty, and the order they close in does not track price. Counted rather than read one insurer at a time, though, the stagger here is tighter than it looks: three of these seven close in the single year between seventy-four and seventy-five.
Step, not trend: a door counts as open at its stated maximum entry age and shut the year after it, so Luma's cap of 70 closes at 71 and the three caps of 74 close at 75. Entry only — renewal of a policy already in force is the separate column above.
Source: Insurer entry caps as tabled (Cigna, ACS, Pacific Cross, IMG, BCBS Global Solutions, VUMI, Luma); doors counted · checked 2026-08-16
There is one structural fact buried in that table that decides everything, and it deserves its own line. Almost every lifetime-renewable policy is only lifetime-renewable if you never leave it. Continuity from before 65 is the hinge. Miss it, lapse it, or try to enter late, and “cover for life” was never an option that was open to you. The market does not sell most people lifetime cover. It sells lifetime cover to people who bought it early and never stopped paying, and it sells the idea of it to everyone else.
Lapse by attrition: the failure nobody models
Here is where the competing pages stop and this one does not. The search results, the broker blogs, the comparison sites all answer one question: can a 70-year-old get insured? They treat entry as the problem, “apply early” as the solution, and end. That framing hides the actual failure mode, which is not at the door. It is years past it.
Take the person who did everything the responsible content told them to. Cover bought at 62, in good health, a proper international policy, renewable for life. They are, on every comparison site’s logic, safe. Now run it forward, which no comparison site does.
The renewal premium compounds on two stacked curves at once: the age-banding above, and medical-cost inflation on top of it, running around 11% a year across Asia-Pacific (the drawdown model carries that figure and its sources). Their income does the opposite. If it is a UK State Pension in Thailand it is frozen outright, for the reasons the frozen-pension arithmetic sets out in full; if it is a private drawdown it rises, at best, with home inflation, which is a fraction of medical trend. Two lines on a chart. One accelerating, one flat. They cross.
The year they cross is the lapse year, and it is the whole point. The policy never denies the holder anything. It does not have to. Somewhere in the mid-to-late 70s the renewal invoice simply exceeds what the income can service, and at that moment the holder discovers the part the word “renewable” concealed: they cannot shop. They are now old and carry a claims history, so the open doors from the acceptance curve are shut to them. The policy they can no longer afford is also the only policy they can no longer replace. They do not get refused at the cliff. They are quietly walked off the back of it by their own renewal schedule — at the exact age the entire plan was supposed to start paying out.
Put rough numbers on it, illustratively, so the crossing is not abstract. A comprehensive international policy bought at 62 in good health runs on the order of $4,000 a year. Hold the conservative end of the sourced escalation: the age curve roughly doubles it per decade, before adding medical trend. That is about $4,000 at 62, $8,000 by the early 70s, $16,000-plus by the early 80s, and the five-fold-after-75 territory beyond. Now set that against a UK State Pension, about £241 a week and near £12,550 a year for 2026/27 and, in Thailand, frozen at that figure for life. The premium line passes a quarter of that income in the mid-60s, half of it around 70, and the entire income somewhere in the late 70s — after which the policy costs more than the person receives to live on. Add medical-cost inflation on top and every one of those crossings comes forward, not back. The exact year is yours to compute with your real quote and your real income. That there is a year is not optional. It is what two diverging exponential-against-flat lines do.
And the first serious claim accelerates it. A real diagnosis at 76 does not just trigger a payout. It triggers the next renewal’s repricing or an exclusion written around the new condition, which means the cover thins precisely as it is first heavily used. The product is at its most expensive and least generous in the year it is first genuinely needed. That is not the exception. On this curve it is the modal outcome.
The mandate that closes the trap
There is a final turn, specific to Thailand, that converts a financial squeeze into a residency one. The O-A long-stay visa carries a health-insurance mandate, commonly cited at US$100,000 / THB 3,000,000 of cover through the immigration-approved insurer scheme. Hold that against the lapse year. The visa can require continuous cover at the same age the market is repricing the holder out of it. The cost line and the legal line converge on the same person in the same years.
The widely-used escape is the non-O retirement route, which does not carry the insurance mandate. Read what that escape actually is. It is choosing to be deliberately uninsured in your highest-risk decade because cover became unaffordable, and calling it a visa strategy. It is not a loophole anyone should feel clever about. It is the lapse year wearing a different form. (These rules move.)
The exclusion that hollows the policy out
Premium is the visible wall. Underwriting is the invisible one, and it does more of the damage. Most international policies are written one of two ways. Full medical underwriting prices and excludes your conditions at entry: anything you already have is named and carved out, permanently, before you pay a penny. A moratorium policy is cheaper and looks kinder, covering you unless you needed treatment or advice for a condition in a defined look-back window, typically the last few years; it quietly re-excludes anything that recurs. Either way the structure points the same direction. The conditions a 75-year-old most needs covered are exactly the ones the policy is engineered to have already excluded, because by 75 almost everyone has a history, and history is what underwriting removes.
This is why “I have insurance” is a weaker statement than it sounds at the age it matters. The policy can be in force, the premium paid, and the specific thing now killing you written out of the contract by a clause agreed when it was hypothetical. The cliff is not only that cover gets unaffordable. It is that the cover you can afford is most hollowed out precisely where you will lean on it. A sourced premium table cannot show you that, which is partly why the cheerful comparison content stops at the premium.
The Philippines: the opposite trap, not the absence of one
Thailand’s failure runs through a mandate. The Philippines runs through its absence, and the absence is not mercy. There is no O-A-style insurance requirement on the common Philippine retirement routes, so the market never forces the issue, which sounds like freedom and functions as a default into being uninsured. The same international insurers apply the same age curve and the same acceptance caps to a Manila address as to a Chiang Mai one; what changes is only that nothing external compels you to hold cover, so the modal Western retiree in the Philippines is uninsured by drift rather than priced out by event. The cliff is the same height. One country pushes you off it with a rule. The other simply does not build a railing — and lets the drawdown do it. Neither is the safe option the brochure for either implies.
What would have to be true for the cliff not to catch you
The honest section, because some people clear it, and they are nameable.
The cliff does not catch you if your income rises faster than medical trend, which in practice means a large, genuinely inflation-linked pension or a portfolio explicitly provisioned for a health line growing at low double digits, not a State-Pension-scale fixed income. It does not catch you if you bought a lifetime-renewable international policy before 65 and can fund its compounding renewal indefinitely, which is a sentence most people cannot finish honestly past 80. It is softened if you can absorb a major claim in cash without the policy, which is self-insurance, viable only at a level of capital that makes the insurance question moot anyway. And it is structurally different, though not absent, in a system where you retain home-country cover you can return to, which is the geographic-cure question wearing an insurance hat.
For everyone whose retirement income is fixed, frozen, or modest, which is most people the move is marketed to, the cliff is not a risk to manage. It is a scheduled event with an estimable date. The responsible advice, “get insured early,” is correct and is also not a solution. It postpones the lapse year. It does not remove it.
Contest the model
The credibility is that you can see the inputs and move them.
- Premium ranges and escalation. Source: Thai-market aggregators and broker pages (insurance-thailand.com, PacificPrime, ExpatCompares), 2026, marked illustrative. Wide by nature; re-quote for your age, health, and plan. Lower entry via a Thai-domestic plan delays the lapse year but caps what is covered, which trades the cliff for a coverage gap.
- Acceptance and renewal caps. Source: insurer terms (Cigna Global, Pacific Cross, IMG via expatfinancial, BCBS). The hinge variable is continuity from before 65. Miss it and “for life” was never yours.
- Medical trend on top of age-banding. Cross-referenced to the drawdown model’s sourced ~11% Asia-Pacific figure. It compounds on the age curve, not instead of it; halving it moves the lapse year out but does not abolish it.
- Income path. Flat or frozen pulls the lapse year in hard; a large uprated income pushes it out. Few of the people sold the move hold the second kind.
Rebuild it with a real quote and your real income. If the renewal premium stays under a survivable share of that income into your late 80s with the medical-trend assumption switched on, the cliff is a line item for you. If it does not, you have found the year, and you have found it while you can still act on it instead of being walked off it.