Roth Conversion vs ACA Subsidy: $1 Over Can Cost $14K

Maria is 58, single, and retired early in Columbus, Ohio. She lives on $30,000 of taxable brokerage income plus a small amount of interest, and she buys her health insurance on the ACA marketplace. Her 2026 MAGI lands near $48,000 — about 307% of the federal poverty level (FPL), using the 2025 single-person guideline of $15,650 that 2026 marketplace coverage is priced against — which earns her a premium tax credit worth roughly $14,400 a year on the benchmark Silver plan in her area. She has $1.1 million in a traditional IRA and a financial planner who keeps telling her to “convert before RMDs.” So she asks: should she convert $90,000 this year to get ahead of required minimum distributions?

In her case, the answer is no — not this year. Here is why, with the full math.

Why a Roth conversion is income for the ACA

When you move money from a traditional IRA or 401(k) to a Roth, the pre-tax portion you convert is ordinary income in the year of the conversion. It shows up on Form 1040 and flows into your adjusted gross income, then into modified AGI (MAGI) — the figure the marketplace uses to size your premium tax credit under IRC §36B(d)(2), the statute that both creates the credit and defines the income that sizes it.

For ACA purposes, MAGI is your AGI plus a few add-backs: tax-exempt interest, the excluded portion of Social Security benefits, and excluded foreign income. A Roth conversion has no special carve-out. Convert $90,000 of pre-tax IRA money, and your MAGI rises by $90,000. The marketplace does not care that the money never left your control — it only sees the income line.

This is the trap. The premium tax credit is one of the most generous benefits in the code for early retirees, and it is sized entirely off MAGI. A conversion that looks cheap when you only count the income-tax bracket becomes expensive once you count the credit you lose.

The worked example: a $90,000 conversion in a subsidy year

Start with Maria’s baseline and layer the conversion on top.

The federal tax alone is the part most people model: about $19,800 at a 22% blended rate. But the $14,400 credit clawback nearly doubles the bill. Maria pays roughly $34,200 to move $90,000 into a Roth — an effective rate near 38%, not the 22% bracket she was looking at. For a 58-year-old, that is a steep price to pre-pay tax on money she will not be forced to touch until her required-minimum-distribution age of 75 under SECURE 2.0 §107 (she was born after 1959).

Reconciliation makes it concrete. Maria received the credit in advance — it was paid directly to her insurer to lower her monthly premium. When she files and reports the conversion, Form 8962 reconciles the advance payments against the credit she actually qualified for. Because her MAGI jumped, she has to repay the excess on her return. The clawback is not theoretical; it is a line on the 1040.

The 400% FPL cliff — and why 2026 is uncertain

Under the original ACA structure, the premium tax credit ended entirely once MAGI crossed 400% of FPL (roughly $62,600 for a single filer in 2026). That was a true cliff: $1 of MAGI over the line could cost a four- or five-figure credit. A conversion that nudged you from $62,500 to $62,700 of MAGI could vaporize the whole subsidy.

The American Rescue Plan Act (2021) and the Inflation Reduction Act (2022) suspended that hard cliff through plan year 2025 and capped the benchmark premium at 8.5% of MAGI at any income. Under those enhanced rules there is no 400% cliff — the credit phases down smoothly — but a large conversion still erodes it dollar by dollar.

Verify before you rely on either version. Whether the enhanced subsidies continue for plan year 2026 is unsettled and politically contested. Before you treat 400% FPL as a hard cliff — or assume the smooth 8.5% cap still applies — confirm the current rule for your plan year at healthcare.gov or your state exchange. The conversion math swings hard depending on which regime is in force.

The decision rule: model the credit clawback as marginal cost

The mistake is treating a conversion’s cost as just the tax bracket. The correct frame: your true marginal cost on each conversion dollar = income tax + state tax + lost premium tax credit on that dollar.

In a subsidy year, that combined marginal cost can sit in the 30s or 40s as a percentage even while your nominal bracket is 12% or 22%. The credit phase-out acts like a hidden surtax stacked on top of the bracket. So the rule splits by year type:

• Subsidy year (you are enrolled in marketplace coverage and getting a credit): cap conversions below your subsidy-loss point. Convert only the amount that keeps MAGI under the level where the credit drops sharply — or skip conversions entirely that year.
• Non-subsidy year (on COBRA, on a spouse’s plan, on Medicare, or simply not enrolled): convert aggressively. With no credit at stake, your marginal cost collapses back to the bracket alone — fill the 12% and 22% brackets to their ceilings.

This is why early retirees who plan well often convert in bursts: small or zero conversions in subsidy years, then a heavy push in the COBRA window after a layoff or in the gap year before age 65 when they may briefly drop coverage or carry COBRA.

The COBRA-year strategy

COBRA premiums are not means-tested. You pay the full group rate regardless of your income, so your MAGI has no effect on your health-insurance cost. That makes a COBRA year a clean conversion window.

Suppose Maria had been laid off and elected COBRA for 18 months. In that window, the same $90,000 conversion costs only the federal tax — about $19,800 — plus Ohio state tax. There is no $14,400 clawback because there is no credit to lose. Her effective rate drops from ~38% back toward ~22-25%. Same conversion, $14,400 cheaper, purely because of the year she chose.

Partial conversions: filling room up to the breakpoint

You do not have to choose between “$90,000” and “nothing.” In a subsidy year, the smart move is a partial conversion sized to stop just below the point where your credit drops sharply.

Maria could, for example, convert $10,000 — lifting MAGI from $48,000 to $58,000. Under the enhanced-subsidy phase-down, the credit erosion on that $10,000 might be a few hundred dollars rather than the full $14,400. She moves money to Roth at a blended marginal cost in the low-to-mid 20s instead of ~38%, and she repeats the modest conversion every subsidy year. Over a decade those small conversions add up without ever triggering the cliff.

The mechanics to respect:

1. Project full-year MAGI first. Add brokerage income, interest, dividends, capital gains, and any other income before you size the conversion.
2. Find your subsidy-loss point. Identify the MAGI level where your specific credit drops steeply (under the enhanced rules) or hits the 400% FPL cliff (if the hard cliff is back).
3. Convert up to — not past — that point by the conversion deadline of December 31 (IRA contributions allow until April 15; conversions do not — see stats above). There is no recharacterization to undo an over-conversion since TCJA eliminated it.

What most people miss

Three traps catch even careful early retirees:

• The clawback is not a phase-out you can ignore — it is real cash at filing. The advance credit went to your insurer all year. A late-year conversion that blows past your subsidy-loss point means you owe that money back on Form 8962. People who convert in December without re-projecting MAGI get a nasty surprise in April.
• Capital gains and the conversion compound. If you also realize long-term gains, those count in MAGI too. A conversion can push ordinary income up while simultaneously shoving stacked capital gains from the 0% LTCG bracket (up to $48,350 taxable income, single, 2026) into the 15% bracket. You can lose the subsidy and the 0% gains rate in the same move.
• The 5-year clock starts per conversion. Each conversion has its own 5-year seasoning period before the converted principal can come out penalty-free if you are under 59½. If you convert at 58 and need the money at 61, the 5-year conversion clock — not just the age test — can trigger a 10% penalty on early withdrawal of converted principal.

There is also a quieter point: this is the early-retiree mirror image of the IRMAA problem. Before 65, the binding MAGI cliff is the ACA premium tax credit. At 65, coverage shifts to Medicare and the binding cliff becomes the IRMAA surcharge, which starts at $103,000 MAGI (single, based on MAGI two years prior). Many people aggressively convert at 63-64 to clear the traditional balance before either cliff binds — threading the narrow window between losing the ACA credit and triggering IRMAA.

The decision lever

Sort your years before you convert a dollar. In any year you are collecting an ACA premium tax credit, treat the credit clawback as part of the conversion’s marginal cost — not just your tax bracket — and cap the conversion below your subsidy-loss point (or convert nothing). In any year the credit is not in play — COBRA, a spouse’s plan, an unenrolled gap year — convert hard and fill the 12% and 22% brackets to their ceilings. The lever is not whether to convert; it is which year you do it in. For Maria, a $90,000 conversion costs ~$34,200 in a subsidy year and ~$19,800 in a COBRA year. Same $90,000 to Roth, $14,400 difference — decided entirely by the calendar.