[HC-05] The Non-ACA Alternatives, Honestly Assessed

COBRA, healthshares, short-term plans, and direct primary care are all cheaper than a marketplace plan for reasons — and every reason is something they do not cover.

Pillar: Healthcare Navigation · Applies to: Anyone leaving employer coverage before 65, and anyone who has been quoted a number that looks too good next to a marketplace premium Last verified: August 2026 · Refresh cadence: Event-driven — short-term-plan rules are under active rulemaking (§4, FN-03); figures via FN-02 Related: HC-01 ACA Bridge · HC-02 Medicare 101 · HC-03 Medicaid Floor · HC-04 HSA Mastery · TR-05 Exit Logistics · ST-04 State Quick Reference · FN-03 Policy Watch

Not advice. Three of the four products on this page are not insurance and are not regulated as insurance, which means a state insurance department cannot help you when a claim is denied. The premiums and deductibles in the worked examples are stated inputs, not market figures — substitute your own. Read your own plan documents, in full, before replacing comprehensive coverage with any of these.

---

TL;DR

• Only COBRA is insurance. Healthshares, short-term plans, and direct primary care are cheaper because they exclude, cap, underwrite, or simply are not coverage — and each is priced accordingly.
• COBRA’s case is narrower than its reputation and mostly comes down to one question: are you getting a subsidy anyway? COBRA premiums are never eligible for the premium tax credit ✅, so for a subsidised household the arithmetic in §3 is brutal — $39,474 over 18 months against $11,520.
• The met-deductible argument has a ceiling, which makes the decision computable without forecasting your health. If the premium gap exceeds the new plan’s out-of-pocket maximum minus what is left on the old one, COBRA cannot win at any level of spending. In §3 the gap is $9,318 and the ceiling is $8,400.
• Above the 400% FPL cliff, COBRA becomes genuinely competitive — there is no credit left to forfeit. Same household, same arithmetic, and COBRA is $5,642 ahead.
• Direct primary care is the one alternative that got better in 2026: a qualifying arrangement up to $150/month self or $300/month family is no longer disqualifying for HSA purposes and can be paid from the HSA ✅ (FN-02).

---

1. Why an alternative is cheaper

Every product here undercuts a marketplace plan, and the discount always comes from one of five places. Learning to name which one is most of the skill:

Source of the discount	What you are actually buying
Underwriting	A price available only while healthy — and withdrawn at renewal if you stop being healthy
Exclusions	No pre-existing conditions, or no coverage of a benefit category the ACA requires
Caps	Annual or lifetime limits, which ACA plans may not impose ✅
No risk transfer	A promise that is explicitly not a contractual obligation to pay
Not being coverage at all	A membership for a service, which is fine, as long as you know that is what it is

Only one thing on the list is not a discount: a subsidy you did not claim. That is the most expensive mistake on this page, and the most common — a household comparing COBRA’s sticker price to a marketplace sticker price without computing the premium tax credit is comparing the wrong two numbers (HC-01).

2. COBRA: the only one that is real insurance

COBRA continues the same plan — same network, same formulary, same deductible progress. It is the only alternative here that is comprehensive coverage with legal recourse, and the only one whose weaknesses are purely about price.

The mechanics worth knowing exactly:

Cost	Up to 102% of the plan’s total cost — the employer’s share plus 2% ✅
Standard duration	18 months after job loss or reduction in hours ✅
Disability extension	29 months total, with months 19–29 chargeable at up to 150% of plan cost ✅
36-month events	Divorce, death of the covered employee, the employee’s Medicare entitlement, or a child ageing off ✅
Election window	60 days from the later of coverage loss or the election notice ✅
First payment	45 days after election ✅
Premium tax credit	Never available for COBRA premiums ✅
Medicare	COBRA is not active-employer coverage, so it does not protect you from Part B late-enrolment penalties ✅ (HC-02)

The 60-and-45-day structure is a free option almost nobody uses. COBRA elected within the window is retroactive to the date coverage ended. So a healthy household can decline to elect, go uncovered on paper for up to 60 days, and elect only if something expensive happens — paying nothing if nothing does. It is genuinely free optionality, and it is also genuinely a gamble on the calendar: elect late in the window and you owe every back premium at once.

The trap on the other side: electing COBRA does not preserve your marketplace special enrolment period. Dropping COBRA voluntarily is not a qualifying event; exhausting it is ✅ (HC-01 §7). Elect COBRA in July because it seemed simpler, change your mind in September, and you are locked out of the marketplace until January.

3. The arithmetic, twice

A couple retiring 30 June. Employer family plan costs $2,150/month in total, so COBRA is $2,193. The family deductible is met; $8,400 of a $12,000 out-of-pocket maximum is used. A marketplace plan at their target MAGI nets $640/month after the credit (HC-01), with a $5,500 deductible restarting at zero. Six months remain in the plan year. All figures are stated inputs.

	COBRA	Marketplace
Premiums, six months	$13,158	$3,840
Gap COBRA must earn back	$9,318	—

COBRA’s only structural advantage is the carried-over deductible and out-of-pocket progress. That advantage has a ceiling, and the ceiling is what makes this decidable without predicting anyone’s health: the most it can ever be worth is the new plan’s remaining exposure minus the old plan’s — here $12,000 − $3,600 = $8,400.

$8,400 is less than $9,318, so COBRA loses at every possible level of medical spending, from zero to catastrophic. Not “probably loses.” Cannot win.

The rule. Compare the premium gap against (the new plan’s remaining out-of-pocket maximum − what is left on your old one). If the gap is larger, stop calculating — the met deductible cannot pay for itself.

Now move the same household over the cliff

Change one thing: their MAGI is above 400% FPL, so after the 2026 expiry of the enhanced credits there is no premium tax credit at all (HC-01 §4, FN-03). The marketplace plan costs its full $2,400/month. Nothing else changes.

At $10,000 of remaining spend	Total
COBRA	$15,158
Marketplace, unsubsidised	$20,800
COBRA ahead by	$5,642

This is the real COBRA case, and the return of the subsidy cliff made it much larger. Over the cliff there is no credit to forfeit, so the met deductible, the intact network, and the mid-treatment continuity are free wins rather than expensive ones.

And the 18-month view

“COBRA is only temporary” is the most misleading sentence in this area:

18 months	Cost
COBRA at 102%	$39,474
Subsidised marketplace	$11,520
Difference	$27,954

That is a Roth conversion’s worth of tax, or two years of a decent travel budget, spent on continuity of network. Sometimes it is the right purchase. It should never be an accidental one.

4. Short-term plans: the product whose rules are moving

Short-term, limited-duration insurance is medically underwritten, excludes pre-existing conditions, need not cover essential health benefits, may impose annual and lifetime caps, and can be rescinded on post-claim review.

Federal duration limits, as of August 2026: the 2024 rule caps plans sold on or after 1 September 2024 at an initial term of three months and four months including renewals ✅. But this is unsettled. In August 2025 the Departments of Labor, Health and Human Services, and the Treasury announced they would open new rulemaking on the definition and, in the meantime, would not prioritise enforcement against issuers not meeting it ✅. The rule is formally in force and functionally in flux — track it in FN-03 before relying on any duration.

The pricing logic, with stated inputs:

	Monthly	Annual
Short-term plan	$210	$2,520
Subsidised Bronze alternative	$560	$6,720
Premium “saving”		$4,200
One excluded pre-existing episode		$38,000
Net position		−$33,800

The saving is real and the exposure is larger than the saving by roughly an order of magnitude. And for the audience this wiki serves, the duration cap ends the discussion before the risk analysis begins: a plan that cannot legally run more than four months is not a bridge to Medicare. It is a gap-filler between a retirement date and a January 1 — which is a legitimate, narrow use, and the only one.

5. Health care sharing ministries: not insurance, in the legal sense

Healthshares are member cost-sharing arrangements, usually with a religious statement of belief. The important facts are structural rather than rhetorical:

• They are not insurance and assume no risk. The organisation is not contractually obligated to pay your bills ✅.
• Roughly 30 states explicitly exempt them from insurance regulation ✅ — so a denied “claim” is not something a state insurance commissioner can act on. There is no external appeal right.
• They may exclude pre-existing conditions, impose annual and lifetime caps, and decline to share costs they consider inconsistent with their standards ✅.
• They are not minimum essential coverage ✅.

The cautionary history is not hypothetical: the Aliera-affiliated collapse produced multi-state enforcement actions, and state attorneys general — California’s issued a renewed consumer alert as recently as July 2026 ✅ — continue warning that these products are marketed as if they were ACA-equivalent coverage.

Where they can legitimately fit: a household that understands it is joining a charitable cost-sharing community rather than transferring risk, that can absorb a large denied bill without damage, and for whom the religious dimension is the point rather than the marketing. That is a real but small group, and it is not “a healthy 58-year-old who wants a cheaper bridge to Medicare.” For that person the product is at its cheapest precisely while it is least needed and unavailable at renewal exactly when it becomes needed.

6. Direct primary care: not insurance, and better in 2026

DPC is a monthly membership paid directly to a primary care practice — typically same-day access, long appointments, and direct messaging, with no insurance billing. It has never claimed to be insurance and does not cover hospitalisation, specialists, imaging, or surgery. Pairing it with a catastrophic or high-deductible plan is the only coherent way to use it.

What changed on 1 January 2026 ✅: under OBBBA, a qualifying DPC service arrangement is no longer a disqualifying health plan for HSA purposes, and the fee itself is a qualified medical expense payable from the HSA. The caps are $150/month self and $300/month family, indexed (FN-02) — $1,800 a year at the individual cap. Arrangements are disqualified if they include procedures requiring general anaesthesia, prescription drugs other than vaccines, or most laboratory services ✅.

Why this matters more than it sounds: before 2026, a DPC membership could void HSA eligibility, so the natural pairing — DPC for routine care plus a cheap high-deductible plan for catastrophe — carried a tax penalty. It no longer does, and 2026 also made Bronze and Catastrophic exchange plans HSA-eligible ✅ (FN-02, HC-04). The two changes compose: Bronze plan, HSA, DPC membership on top, all inside the tax shelter.

7. The honest ranking

Product	Is it insurance?	Regulated?	Pre-existing conditions	Sensible use
Subsidised marketplace plan	Yes	Yes	Covered ✅	The default for nearly everyone (HC-01)
COBRA	Yes	Yes	Covered ✅	Over the subsidy cliff; mid-treatment; narrow network; employer-paid in severance
Direct primary care	No — a service	Varies by state ◻️	N/A	Supplement to real coverage; now HSA-compatible ✅
Short-term plan	Technically, excepted from ACA rules	Lightly, and states differ ✅	Excluded ✅	A few weeks between a retirement date and 1 January
Healthshare	No ✅	Mostly exempt ✅	Excluded ✅	A deliberate, informed, religious choice with reserves behind it

Two more options belong on the list and are covered elsewhere: Medicaid, if your MAGI is low enough (HC-03), and a spouse’s employer plan, which is usually the cheapest answer available to anyone who has one.

Three Scenarios

A. Retiring 30 June at 58, subsidised. He assumes COBRA is the responsible choice and the marketplace is for people without options. The §3 arithmetic says COBRA cannot win at any level of spending, because the gap exceeds the ceiling. He takes the marketplace plan, and the 60-day COBRA election window costs him nothing to keep open in the meantime — he declines to elect, banks the option, and lets it lapse in September.

B. Retiring at 61 with $310,000 of income from a deferred-comp payout (TR-05). He is far above the 400% FPL cliff this year, so there is no credit to lose and COBRA’s continuity is nearly free — the $5,642 case. The right plan is COBRA this year and a subsidised marketplace plan next year, once the payout is behind him and MAGI is managed (ER-05). The insight is that the answer changes year to year with MAGI, and it should.

C. The 56-year-old quoted $210/month by a broker. The quote is a short-term plan, and it is 62% cheaper than the Bronze plan she was looking at. Three facts end it: it excludes the condition she already has, it cannot legally run more than four months ✅, and it is not minimum essential coverage. She buys the Bronze plan, adds a $150/month DPC membership for the access she actually wanted, and pays the DPC fee from her HSA ✅ (HC-04).

💡 Pro-Tips

1. Compute the credit before comparing anything. Two sticker prices are not a comparison when one of them is subsidised (HC-01).
2. Use the ceiling rule in §3 before modelling scenarios — it usually decides COBRA in one line.
3. Do not elect COBRA on day one. The 60-day retroactive election is free optionality; electing early only starts the clock and the payments.
4. Ask whether severance includes employer-paid COBRA — it is one of the more winnable asks (TR-05), and it inverts the arithmetic entirely.
5. Re-run the decision every year. The right answer at $310,000 of one-off income is the wrong answer the following January.
6. If you elect COBRA, calendar its exhaustion date, because exhaustion is a qualifying event and voluntary termination is not ✅.
7. Treat DPC as a supplement with a receipt — from 2026 the fee is HSA-qualified up to the caps ✅, so pay it from the HSA rather than from cash.
8. Ask a healthshare for its written denial and appeals process before joining. If there is no external appeal, that is the product working as designed, not a defect.

⚠️ Common Pitfalls

1. Comparing COBRA’s price to an unsubsidised marketplace price when you qualify for a credit — the single most expensive error here.
2. Electing COBRA and then wanting out mid-year. Voluntary termination is not a qualifying event ✅.
3. Believing COBRA delays Medicare enrolment. It is not active-employer coverage, and the Part B penalty accrues ✅ (HC-02).
4. Missing the 45-day first-payment deadline after electing, which voids the election.
5. Treating a healthshare as insurance. No risk transfer, no state regulator, no external appeal ✅.
6. Assuming a healthshare or short-term plan satisfies a state individual mandate. Neither is minimum essential coverage ✅ (State notes).
7. Planning a bridge around a short-term plan. Four months is the federal cap, and five states prohibit the product outright ✅.
8. Buying DPC instead of coverage rather than alongside it — it covers no hospitalisation, specialists, or surgery.
9. Exceeding the DPC fee caps and disqualifying your HSA contributions for the year ✅ (FN-02).
10. Forgetting that underwritten products are withdrawn at renewal. The healthy price is not a price you get to keep.

✅ Actionable Checklist

Before your last day (TR-05)

• Get the plan’s total monthly cost in writing — COBRA is 102% of that, not of your payroll deduction
• Record deductible and out-of-pocket progress as of your termination date
• Ask whether severance can include employer-paid COBRA months
• Note the exact dates of the 60-day election and 45-day payment windows

Choosing

• Compute the subsidised marketplace premium at your planned MAGI (HC-01, ER-05)
• Apply the ceiling rule: premium gap vs (new out-of-pocket max − remaining old exposure)
• Check whether any provider or ongoing treatment is out-of-network on every marketplace option
• If you are over the 400% FPL cliff this year, price COBRA properly — it may well win

If you are considering a non-insurance option

• Confirm in writing whether pre-existing conditions are excluded and whether caps apply
• Confirm whether your state regulates the product at all (State notes, ST-04)
• Confirm it is not being relied on to satisfy a state individual mandate
• For DPC: confirm the fee is within the caps and the arrangement excludes no disqualifying services ✅

State notes (→ ST-01, ST-04): Three of the four products on this page are state-governed. Five states — California, Illinois, Massachusetts, New Jersey, and New York — prohibit the sale of short-term plans outright ✅, and in roughly nine more plus DC the rules are strict enough that no insurer offers one ◻️, so the federal duration cap is a ceiling rather than the binding constraint in much of the country. Around 30 states explicitly exempt health care sharing ministries from insurance regulation ✅, which is precisely why there is no regulator to appeal to. California, Massachusetts, New Jersey, Rhode Island, and DC enforce individual mandates with penalties, and Vermont has one without a penalty ✅ — neither healthshares nor short-term plans are minimum essential coverage ✅, though Massachusetts operates a limited mechanism recognising some healthshare membership ◻️. DPC is authorised by statute in most states, and the statutes differ on whether a DPC agreement is insurance ◻️ — that question is separate from the federal HSA treatment, which is now settled ✅. COBRA itself is federal, but state “mini-COBRA” laws extend similar rights to employers below the 20-employee federal threshold ◻️, which is the one place a small-employer retiree may have more options than this page implies.

Sources & further reading (verified August 2026)

• U.S. Department of Labor, FAQs on COBRA Continuation Health Coverage and 26 CFR § 54.4980B — the 102% and 150% premium limits, the 18/29/36-month durations, the 60-day election window, and the 45-day first-payment deadline.
• Departments of Labor, HHS and Treasury, final rule on Short-Term, Limited-Duration Insurance (effective for plans sold on or after 1 September 2024) — the three-month initial term and four-month maximum including renewals; and the tri-agency statement of 7 August 2025 announcing new rulemaking and non-prioritisation of enforcement in the interim. Status tracked in FN-03.
• NAIC, “Not All Products are Health Insurance: Health Care Sharing Ministries, Discount Plans and Risk-Sharing Plans”, with state consumer alerts including the California Attorney General’s July 2026 alert — that HCSMs assume no risk, are exempt from insurance regulation in roughly 30 states, and are not minimum essential coverage.
• One Big Beautiful Bill Act and IRS Notice 2026-05 — direct primary care service arrangements are not disqualifying health plans for HSA purposes for months beginning after 31 December 2025, subject to the $150/$300 monthly caps and the exclusions for general anaesthesia procedures, non-vaccine prescription drugs, and most laboratory services.
• KFF, short-term plan availability by state — the five prohibiting states and the further states where no product is marketed.
• tools/hc05_worked_examples.py — every figure in §3 and §4, computed and checked against this article by CI. Premiums and deductibles are stated inputs.
• HC-01 owns the premium tax credit this page keeps deferring to · HC-03 owns the Medicaid floor below it · HC-04 owns the HSA mechanics · TR-05 owns the exit-week logistics.