[HC-01] Navigating the ACA Health Insurance Bridge Before Medicare

Maximizing Subsidies and Cost-Sharing Reductions in the Post-2025 World

Pillar: Healthcare Navigation · Applies to: Anyone retiring before 65 without employer/retiree coverage Last verified: August 2026 · Refresh cadence: Annual (November) + event-driven (pending Senate legislation) Related: HC-02 Medicare 101 · HC-03 The Medicaid Floor · HC-04 HSA Mastery · ER-02 Roth Conversion Ladder · TX-01 0% Capital Gains

Not advice. Educational reference. Verify current figures at Healthcare.gov / your state exchange and with a tax professional before acting — the rules below changed materially on January 1, 2026, and may change again.

---

TL;DR

• If you retire before 65, the ACA marketplace is usually your bridge to Medicare. Your premium subsidy is a function of MAGI — and as an early retiree, you get to choose your MAGI by deciding which accounts you spend from and how much you convert to Roth.
• The world changed on 1/1/2026. The enhanced (COVID-era) subsidies expired. The 400% FPL “subsidy cliff” is back: one dollar of income over the line can cost you $15,000–$25,000/year in lost credits. Required contributions rose at every income level, and the caps that used to limit how much excess subsidy you’d repay at tax time are gone — clawback is now unlimited.
• The three income lines that matter most in 2026 (48 states, household of 2): $29,187 (138% FPL — the Medicaid floor in expansion states, and 200% FPL in the four Basic Health Program jurisdictions, HC-03), $42,300 (200% FPL — the last strong cost-sharing-reduction tier), and $84,600 (400% FPL — the cliff).
• Master three levers: (1) fund spending from sources that don’t create MAGI, (2) place Roth conversions and gain harvesting deliberately (December, once income is known), (3) pick your metal tier based on which side of the CSR lines you’ll land on.

---

1. The 2026 Landscape: What Just Changed

Three separate policy events stacked on top of each other. Understand them before any planning:

Change	What it means	Source of change
Enhanced premium tax credits expired 12/31/2025	The 400% FPL eligibility cliff returned; required premium contributions rose at every income band (e.g., a household at 200% FPL paid ~2% of income for the benchmark plan in 2025; it pays ~6.6% in 2026). Average subsidized enrollee premium payments roughly doubled.	ARPA/IRA sunset; Congress did not extend
Unlimited APTC clawback	Repayment caps on excess advance premium tax credit are eliminated starting with the 2026 plan year. Previously a sub-400%-FPL household repaid at most ~$375–$3,250; now you repay every dollar of subsidy you weren’t entitled to.	OBBBA §71305 (July 2025)
Enrollment tightening	Pre-enrollment income/eligibility verification phases in, effectively ending passive auto-renewal; the continuous low-income (<150% FPL) special enrollment period is terminated. The CMS rule also shortened open enrollment to Nov 1–Dec 15 — but a court vacated that provision in June 2026 and it is on appeal (§7).	OBBBA §§71303–71304 + CMS Marketplace Integrity rule (2025), partly vacated

Legislative watch (as of August 2026): The House passed a three-year extension of the enhanced credits in January 2026 (230–196); the Senate rejected a similar bill in December 2025 and has been negotiating a narrower two-year compromise (income caps, minimum premiums). Nothing has been enacted. Plan under current law; treat any restoration as upside. → Track in FN-03 Policy Watch.

---

2. How the Premium Tax Credit Actually Works

The formula

Your annual PTC = (Benchmark plan premium) − (Your required contribution)
Your required contribution = MAGI × applicable percentage

• Benchmark plan = the Second-Lowest-Cost Silver Plan (SLCSP) available to your household in your county. Not the plan you buy — just the yardstick.
• You can apply the credit to any metal tier (except catastrophic). Buy cheaper than benchmark and you keep the savings, down to a $0 premium; buy pricier and you pay the difference.
• The credit is advanceable (APTC — paid monthly to your insurer) and reconciled on Form 8962 with your tax return. See §6 — reconciliation just got dangerous.

Step 1: Compute your ACA MAGI

ACA MAGI is not the same as other MAGIs in the tax code:

ACA MAGI = AGI
         + tax-exempt interest (muni bonds)
         + non-taxable Social Security benefits
         + excluded foreign earned income

What counts (creates MAGI): wages, self-employment income, interest, dividends, capital gains, traditional IRA/401(k) withdrawals, Roth conversions, pension income, rental profit, all Social Security (even the untaxed portion).

What doesn’t count: Roth qualified withdrawals, withdrawals of your own Roth contribution basis, spending down cash savings, the return-of-basis portion of taxable-account sales, HSA withdrawals for medical expenses, loan proceeds, gifts, and (up to the §121 exclusion) home-sale gains.

What reduces it: deductible HSA contributions, deductible traditional IRA contributions (if you have earned income), self-employment deductions, capital losses (up to $3k against ordinary income).

This is the core early-retiree superpower: a couple spending $80k/year might show $35k of MAGI — or $95k — depending entirely on which pocket the money comes from.

Step 2: Find your % of FPL

2026 coverage uses the 2025 HHS poverty guidelines (48 contiguous states; AK/HI higher). Every figure in this table tracks FN-02 §7 ✅ — that page is the numbers authority and is refreshed each January when HHS releases new guidelines; if this table and FN-02 ever disagree, FN-02 is right.

Household size	100% FPL	138%	150%	200%	250%	300%	400% (cliff)
1	$15,650	$21,597	$23,475	$31,300	$39,125	$46,950	$62,600
2	$21,150	$29,187	$31,725	$42,300	$52,875	$63,450	$84,600
3	$26,650	$36,777	$39,975	$53,300	$66,625	$79,950	$106,600
4	$32,150	$44,367	$48,225	$64,300	$80,375	$96,450	$128,600

Step 3: Apply the 2026 applicable percentage (Rev. Proc. 2025-25)

MAGI as % of FPL	You’re expected to pay (% of MAGI) toward the benchmark
< 133%	2.10%
133% – <150%	3.14% → 4.19%
150% – <200%	4.19% → 6.60%
200% – <250%	6.60% → 8.44%
250% – <300%	8.44% → 9.96%
300% – 400%	9.96% (flat)
> 400%	Not eligible — $0 credit

(Percentages interpolate linearly within each band. For 2025 comparison: the enhanced schedule ran 0%–8.5% with no cliff.)

Worked example: the bridge couple

Two 60-year-olds, no dependents, 48-state metro where the SLCSP for the household is $2,400/month ($28,800/year) — a realistic 2026 figure for that age.

Planned MAGI	% of FPL	Required contribution	Annual PTC	Effective premium for benchmark
$45,000	213%	$45,000 × ~6.9% = $3,105	$25,695	$259/mo
$60,000	284%	$60,000 × ~9.4% = $5,640	$23,160	$470/mo
$80,000	378%	$80,000 × 9.96% = $7,968	$20,832	$664/mo
$85,000	402%	n/a — over the cliff	$0	$2,400/mo

The last two rows are the headline: $5,000 of extra MAGI destroyed $20,832 of credit — an implicit marginal rate of over 400% on that increment, before regular income tax. A single 60-year-old faces the same cliff at $62,600.

---

3. Cost-Sharing Reductions: The Second Subsidy Most People Miss

CSRs lower your deductible, copays, and out-of-pocket max — but only if you buy a Silver plan. They survived 2025 intact and are now relatively more valuable since premiums rose.

MAGI (% of FPL)	Silver plan actuarial value becomes	What that feels like
100–150%	94%	Better than a Platinum plan — deductibles often $0–$500 ◻️
>150–200%	87%	Better than Gold — deductibles often $800–$1,500 ◻️
>200–250%	73%	Barely better than standard Silver (70%)
>250%	70% (none)	Standard Silver; for reference, the average 2026 silver deductible is ~$5,300 ◻️

The actuarial-value tiers (94/87/73%) are statutory ✅ (FN-02 §7). The deductible figures are illustrative market averages, not entitlements ◻️ — actual deductibles vary by carrier, plan, and county, and the only number that binds you is the one on the plan you buy.

The 200% FPL line is a genuine cliff of its own. Crossing from 199% to 201% of FPL can raise your family’s deductible by $3,000–$4,000 ◻️. For a household of 4, that line sits at $64,300 of MAGI in 2026 ✅ (FN-02 §7).

Silver loading (why plan choice is state-dependent): insurers load the cost of CSRs onto silver premiums. Consequences:

• If you qualify for CSRs (≤200% FPL especially): buy Silver — the loaded premium is subsidized away and you keep the rich AV.
• If you don’t (>250% FPL): loaded Silver is often the worst value on-exchange. Compare Gold (sometimes cheaper than Silver with a lower deductible) and Bronze.
• Over the cliff entirely (>400% FPL): also price off-exchange silver plans, which may skip the CSR load.

---

4. Metal Tier Strategy for 2026 (Including the New HSA Play)

Your situation	Usually strongest play	Why
MAGI ≤ 200% FPL	Silver, on-exchange	94/87% AV CSRs are unbeatable
200–250% FPL	Silver vs. Gold — run both	73% CSR is weak; Gold sometimes wins after silver loading
250–400% FPL, healthy	Bronze + HSA	Big PTC applied to a cheap plan → low/zero premium; and see below
250–400% FPL, heavy utilization	Gold	Trade premium for lower deductible
> 400% FPL (over cliff)	Bronze/Gold, compare off-exchange; consider MAGI engineering to get back under	Silver load makes on-exchange silver poor value at full price

New for 2026 — the Bronze HSA double-dip: OBBBA expanded HSA eligibility so that all marketplace Bronze and Catastrophic plans qualify as HDHPs starting with the 2026 plan year. For early retirees this is a rare compounding move: the HSA contribution (2026 limits: $4,400 self / $8,750 family, +$1,000 catch-up each at 55+) is deductible → reduces MAGI → increases your PTC → and grows triple-tax-free for retirement healthcare. A 60-year-old couple contributing $10,750 lowers MAGI by the same amount, which near the cliff can be the difference between $0 and ~$20k of credit. → Full treatment in HC-04.

---

5. MAGI Engineering: The Early Retiree’s Playbook

The spending-source hierarchy (subsidy-preserving order)

1. Cash / money market already taxed — $0 MAGI per dollar spent (interest still counts).
2. Taxable brokerage, high-basis lots — only the gain is MAGI. Selling $40k with $30k basis = $10k MAGI.
3. Roth contribution basis (contributions come out first, tax- and MAGI-free at any age).
4. HSA reimbursements for receipts you’ve shoeboxed — $0 MAGI.
5. Traditional IRA/401(k) withdrawals & Roth conversions — 100% MAGI. Use these deliberately to hit your target, not accidentally.

The Roth-ladder tension (the FIRE community’s central tradeoff)

Every dollar you convert to Roth during the bridge years is a dollar of MAGI. In 2026 the true marginal cost of a conversion is:

income tax on the conversion
+ PTC lost (roughly 10–15¢ per $1 within the phase-out, from the rising applicable %)
+ any CSR tier you fall through (step losses at 150/200/250% FPL)
+ the entire credit, if the conversion pushes you past 400% FPL

Practical resolutions:

• Pick a lane each year. Big-conversion years (accept full-price insurance or a lean subsidy) alternated with subsidy-max years is often better than splitting the difference every year.
• Convert in December, after mutual-fund distributions post and your income picture is final. Convert up to a pre-chosen line (e.g., 199% FPL, or $1–2k under the cliff as a buffer). A conversion cannot be undone — recharacterization was abolished — so undershoot.
• Remember conversions started before ~age 60 are also feeding your ladder’s 5-year clocks (ER-02) — the ACA cost is partly buying future tax-free access.

The floor: don’t go too low (HC-03 preview)

PTC eligibility starts at 100% FPL. Below it (or below 138% FPL in Medicaid-expansion states like WA, CA, NY, etc.), the exchange routes you to Medicaid instead. Medicaid is real coverage, but many early retirees want marketplace plans (provider networks, travel coverage, asset-freedom optics). If your natural MAGI is too low, you may need to manufacture income — a small Roth conversion or gain harvest — to stay above the floor. Good-faith income estimates that come in low are generally not clawed back, but repeated years of under-floor actuals invite verification problems under the new rules.

---

6. Reconciliation Just Became High-Stakes

You estimate income at enrollment; the IRS trues it up on Form 8962. Two regime changes:

1. No more repayment caps (2026 forward). Overshoot your estimate and you repay the entire excess APTC — potentially $20k+ — when you file. The old $375–$3,250 safety net is gone.
2. Underestimating is still refundable. If you took less APTC than you deserved, you get the difference back at filing.

The asymmetry dictates the strategy: when income is uncertain (consulting income, a possible home sale, lumpy capital gains), take less than your full credit in advance — even $0 — and collect the true-up as a refund. You’re floating the premium for a year in exchange for eliminating clawback risk.

Classic clawback triggers to model before they happen: a December Roth conversion done “by habit,” large mutual-fund capital-gain distributions in taxable accounts, a spouse’s untaxed Social Security (it’s ACA MAGI!), severance or deferred comp landing in year one of retirement, home-sale gain beyond the $250k/$500k exclusion, and mid-year inheritance income (IRD like inherited-IRA withdrawals).

---

7. Enrollment Mechanics (2026 Rules)

• Open enrollment — unsettled, but the operating rule is stable. The CMS integrity rule shortened Healthcare.gov’s window to Nov 1 – Dec 15. On June 12, 2026 the District of Maryland vacated that provision in City of Columbus v. Kennedy, so the 2027 window is currently expected to run Nov 1, 2026 – Jan 15, 2027 ◻️ — and the government has appealed to the Fourth Circuit, which could restore the short window on short notice. Operate as if Dec 15 is your deadline anyway: in every state, enrolling by Dec 15 is what buys January 1 coverage; Dec 16 – Jan 15 starts you February 1. ✅ The January backstop is a safety net you may not have, not a plan. State exchanges set their own windows regardless (WA’s Healthplanfinder, Covered CA, etc.). → FN-03 §3
• Auto-renewal is effectively dead. Verification requirements mean you should actively re-enroll and re-verify income every year or risk losing your credit; $0-premium auto-enrollees can be charged a small monthly amount until they confirm eligibility.
• Special Enrollment Periods (SEPs): losing job-based coverage (retirement!) gives you a 60-day window — this is how mid-year retirees enroll. Also: moving, marriage, birth. The continuous SEP for <150% FPL households is terminated. Voluntarily dropping COBRA mid-stream is not a qualifying event; COBRA expiring is.
• COBRA vs. ACA at retirement: COBRA keeps your exact network/deductible progress for up to 18 months at ~102% of full cost with no subsidy. ACA restarts deductibles but subsidizes. Common pattern: COBRA to finish a treatment year or hit an already-met out-of-pocket max, then ACA at the next Jan 1. Beware: once your 60-day SEP lapses, you’re locked to open enrollment.
• State subsidy layers exist. Several states add their own premium help on top of federal PTCs (e.g., Washington’s Cascade Care Savings, plus programs in CA, NJ, NM, VT, CO, MA). Check your exchange even if you’re over the federal cliff.

---

8. The Endgame: Handing Off to Medicare at 65

• Your PTC eligibility ends when you become eligible for premium-free Medicare Part A — generally the month you turn 65 — not when you get around to enrolling. Keeping APTC past that point = repayment.
• Enroll during your 7-month Initial Enrollment Period (3 months before your birthday month → 3 after). Missing Part B triggers a lifetime 10%/year late penalty (2026 base Part B premium: $202.90/mo).
• Cancel the marketplace plan yourself effective the day Medicare starts — it does not auto-terminate, and unsubsidized ACA premiums for a 65-year-old are brutal.
• Stop HSA contributions at least 6 months before enrolling in Medicare after 65 (Part A enrollment is retroactive up to 6 months; overlapping contributions are excess contributions).
• The age-63 trap: Medicare’s IRMAA surcharge uses a two-year MAGI lookback. Your income at 63 sets your premium at 65 — the 2026 first IRMAA tier starts at $109,000 single / $218,000 joint (2024 income), adding up to ~$487/mo per person to Part B. Big bridge-year Roth conversions at 63–64 can buy you two years of surcharges. → HC-06.
• Married with an age gap: when the older spouse exits to Medicare, the household ACA math changes — same MAGI, smaller household premium, and the FPL table still uses full household size. Re-run the numbers; the younger spouse often keeps a strong subsidy.

---

9. Three Real-World Scenarios

A. Traditional bridge couple (both 60, retiring this year, $85k/yr spending). Spending funded: $45k from cash + taxable basis, $10k from high-basis stock sales (~$3k gains), $32k IRA withdrawal. MAGI ≈ $36k (170% FPL) → Silver plan with 87% AV CSR, required contribution ≈ $1,900/yr against a $28.8k benchmark → ~$26.9k credit, sub-$200/mo premium with a ~$1,200 deductible. They defer Roth conversions to ages 65–72 (post-ACA, pre-RMD), accepting future IRMAA management instead (TX-02).

B. FIRE family of four (parents 45, kids 12 & 14, $70k spending). Target: 199% of FPL = $63,980. They harvest exactly enough gains + convert enough Roth to land at $63,500 (buffer), staying under the 200% CSR line. Bronze-vs-Silver test: at 87% AV the Silver wins easily with kids’ sports injuries in the mix. Their Roth ladder runs at reduced size for now; they schedule two “fat conversion” years for when the kids age off the plan.

C. The cliff casualty (single, 62, $61k salary-replacement plan + surprise). Planned MAGI $58k (93% of the way to the $62,600 cliff). In November her bond fund distributes $6k of gains — actual MAGI $64k. Under 2025 rules she’d have repaid ≤$1,575 of her APTC. Under 2026 rules she repays all ~$14,000. The fix that would have saved her: holding distributing funds in tax-advantaged accounts, tracking estimated distributions in October, and keeping a 5–8% MAGI buffer below the cliff.

---

💡 Pro-Tips

1. Plan to lines, not vibes: write down your three numbers each January — Medicaid floor, 200% FPL, 400% FPL — for your household size, and steer MAGI to a target with buffer.
2. December is decision month: fund distributions post, income is known, and Roth conversions/gain harvests can be sized to the dollar.
3. Take partial APTC when income is lumpy — the reconciliation asymmetry (unlimited clawback vs. refundable shortfall) makes conservatism free except for cash-flow.
4. Bronze + HSA is the new cliff-adjacent play: the contribution both shelters money and raises your subsidy by lowering MAGI.
5. SLCSP ≠ your plan: always fetch your actual benchmark from your exchange before modeling — it varies wildly by county and resets annually.
6. Check state-level subsidies even if you fail the federal test.
7. Age 64 is a tax-planning year for Medicare, not just ACA — the IRMAA lookback is already running.
8. Keep proof of income estimates (spreadsheet + enrollment screenshots). The new verification regime makes documentation valuable.

⚠️ Common Pitfalls

1. $1 over the cliff = total subsidy loss. There is no phase-out above 400% FPL anymore.
2. Assuming the old repayment caps still protect you. They don’t, as of the 2026 plan year.
3. Forgetting the MAGI add-backs: muni interest and the untaxed portion of Social Security count. A bridging spouse + a Social-Security-claiming spouse is the classic surprise.
4. Falling below the floor and getting routed to Medicaid you didn’t want — or into the coverage gap in non-expansion states.
5. Missing the 60-day SEP after leaving work, then discovering open enrollment is months away.
6. Letting the plan auto-anything: auto-renewal is unreliable now, and marketplace plans don’t auto-cancel at Medicare.
7. HSA contributions overlapping retroactive Part A after 65.
8. Crossing 200% FPL by a hair and silently losing the 87% AV plan — the deductible jump often outweighs the premium math people fixate on.
9. Trusting a January enrollment deadline — it is currently restored by litigation, under appeal, and never bought January 1 coverage anyway.
10. Treating a Roth conversion as reversible. It isn’t. Size with a buffer.

✅ Actionable Checklist

12+ months before retiring

• Build the MAGI model: list every income source and which are steerable
• Pre-position 1–2 years of spending in cash/high-basis taxable (subsidy fuel)
• Check whether your state expanded Medicaid and has a state subsidy program
• If 63+: model the IRMAA lookback before any big conversion

At retirement / SEP window

• Mark the 60-day SEP deadline from your coverage-loss date
• Price COBRA vs. subsidized ACA (include deductible progress already banked this year)
• Pull your county’s actual SLCSP; compute credit at your target MAGI
• If ≤200% FPL: default to Silver for CSRs; if 250%+: test Bronze+HSA and Gold
• Decide APTC level (full / partial / none) based on income certainty

Every year on the bridge

• Actively re-enroll during open enrollment — assume Dec 15, whatever the litigation says (FN-03)
• Re-verify income documentation proactively
• October: estimate fund distributions; December: size conversions/harvests to your target line
• File Form 8962 reconciliation; adjust next year’s estimate from actuals

Approaching 65

• Calendar the 7-month Medicare IEP
• Stop HSA contributions ≥6 months before Medicare enrollment (if enrolling after 65)
• Cancel the marketplace plan effective Medicare’s start date
• Re-run ACA math for a younger spouse staying behind on the exchange

---

State notes (→ ST-01, ST-04): Nearly every load-bearing number on this page is state-shaped. Your benchmark SLCSP is set at the county level and varies enough between counties that national or state averages are useless for planning — pull your own (§2, and the checklist below). Medicaid expansion status sets your floor: 138% FPL in expansion states, versus the coverage gap below 100% FPL in non-expansion states, where PTCs simply don’t begin (§5). ✅ For the 2026 plan year, 21 state-based exchanges — 20 states plus DC — run their own eligibility and enrollment platforms; 30 states use Healthcare.gov. Those exchanges set their own enrollment deadlines, which do not track Healthcare.gov’s (§7). Several also layer state premium subsidies on top of federal PTCs, and those can survive above the 400% federal cliff (§7). Model your county, not the national averages here.

Sources & further reading (verified Aug 2026)

• IRS Rev. Proc. 2025-25 (2026 applicable percentage table)
• CRS R48290, Enhanced Premium Tax Credit and 2026 Exchange Premiums: FAQ
• KFF: What We Know So Far About 2026 ACA Marketplace Enrollment, Premiums, and Deductibles; Peterson-KFF Health System Tracker on 2026 premium/deductible tradeoffs
• OBBBA (P.L. 119-21) §§71303–71305; healthinsurance.org OBBBA coverage summary
• CMS Marketplace Integrity and Affordability final rule (June 2025) — note ongoing litigation (City of Columbus v. Kennedy)
• 2025 HHS Poverty Guidelines (govern 2026 plan-year coverage)
• Healthcare.gov / your state exchange for live SLCSP quotes