Roth Conversions vs ACA Subsidies: The Bridge-Year Tradeoff

The years between your last paycheck and age 65 are the strangest of your financial life. Your income is not something that happens to you — it is something you choose. You decide how much to convert from a traditional IRA to a Roth, how much gain to realise, and which accounts to spend from.

That freedom creates a genuine conflict, and it has no universal answer.

Convert aggressively and you move money into a tax-free account at a low rate, shrinking the required minimum distributions that will otherwise arrive at 73 and possibly push you into a higher bracket for life.

Convert conservatively and you keep Marketplace MAGI low, preserving premium tax credits and cost-sharing reductions that can be worth thousands of dollars a year.

You cannot do both. A Roth conversion is ordinary income, so it raises MAGI dollar for dollar. Every dollar you convert makes your health insurance more expensive.

Why the conflict is sharper than it looks

If this were only about tax brackets, it would be easy: convert up to the top of the 12% or 22% bracket and stop. The bracket boundaries are wide, and the marginal cost of one more dollar is knowable.

ACA subsidies do not behave like brackets. Two different mechanisms are in play.

Premium tax credits phase out gradually. An extra dollar of income costs you a fraction of a dollar in credits — annoying, but smooth and easy to price.

Cost-sharing reductions are step functions. They quietly upgrade the actuarial value of a silver plan — a lower deductible, a lower out-of-pocket maximum, better copays — and they are tied to thresholds commonly set at 150%, 200%, and 250% of the federal poverty level. Cross one by a dollar and the upgrade is gone for the whole year.

So the marginal cost of the 10,000th converted dollar can be near zero, and the marginal cost of the 10,001st can be several thousand. That is not a curve you can eyeball.

Pricing the two sides

What a conversion is worth

The benefit is the difference between the rate you pay now and the rate you would have paid later, applied to the converted amount, plus the value of removing that money from future RMD calculations.

Convert $30,000 at an effective 12% instead of an eventual 22%, and the direct saving is roughly $3,000 — plus decades of tax-free growth and a smaller future RMD.

Conversions are most valuable when:

• You have a large traditional balance and a long runway before RMDs
• Your current bracket is genuinely low and will not stay that way
• You expect a pension or Social Security to raise your later baseline
• You are married and thinking about the survivor filing at single rates

What the subsidies are worth

Harder to price, because it depends on the plan year and your household, but the components are:

• The premium tax credit forfeited
• The cost-sharing reduction lost by crossing a threshold — often the larger number, and the one people forget
• Any state-level assistance tied to the same income measure

The cost-sharing piece is the part that makes this decision non-obvious. Losing a CSR tier does not raise your premium; it raises your exposure if you get sick. That is a probabilistic cost, which makes it easy to under-weight and expensive to be wrong about.

The two-ceiling problem

Here is the situation in concrete numbers — a single filer, 2026 coverage, holding a 250% FPL guardrail:

Item	Amount
Earned income	$12,125
Ordinary dividends and interest	$1,000
Qualified dividends	$3,000
Long-term capital gains	$15,000
Planned Roth conversion	$8,000
Marketplace MAGI	$39,125

The 2025 FPL for a household of one is $15,650, so a 250% guardrail sets the ceiling at exactly $39,125. Headroom: $0.

Now the tax view of the identical situation:

• Taxable income: $39,125 − $16,100 standard deduction = $23,025
• 0% long-term capital gains ceiling (2026, single): $49,450 → $26,425 of room left
• Room to the 24% bracket: $105,700 + $16,100 − $21,125 = $100,675

Two ceilings, wildly different answers. On brackets alone this person looks like they have six figures of conversion capacity. Against the guardrail they have nothing.

The usable answer is always the smaller of the two:

additional_roth_room = min(magi_headroom, ordinary_bracket_room)
additional_ltcg_room = min(magi_headroom, zero_percent_ltcg_room)

Here both evaluate to zero. The headroom planner computes exactly this, which is why it reports a single binding number instead of three encouraging ones.

Conversions or gain harvesting?

Both consume the same MAGI headroom, so in a constrained year it is a choice, not a sequence.

Roth conversion moves money permanently into a tax-free account, shrinks future RMDs, and starts a five-year clock after which the converted amount can be withdrawn penalty-free. Long-horizon, compounding benefit.

Tax-gain harvesting — selling appreciated holdings inside the 0% bracket and immediately repurchasing — raises your cost basis at no federal tax cost, reducing tax on a future sale. Smaller benefit per dollar, but repeatable every year, and no five-year lockup. Worth knowing: wash-sale rules apply to losses, not gains, so you can repurchase immediately.

A reasonable heuristic: if you have a large traditional balance and a long runway, conversions usually dominate. If your traditional balance is modest, or you are close to 65 and need the basis for near-term spending, harvesting often wins.

A framework for the decision

1. Estimate MAGI before you do anything. All of it: earned income, dividends, interest, planned conversions, realised gains, plus the add-backs for tax-exempt interest and non-taxable Social Security.
2. Pick a guardrail deliberately. Not the highest one you can tolerate — the one whose benefits you actually want. Run your numbers at two different guardrails and compare what you give up.
3. Compute both ceilings — MAGI headroom and bracket or 0% LTCG room — and take the smaller.
4. Price the tradeoff for your year. Get an actual premium and CSR quote at two income levels rather than assuming.
5. Decide which lever gets the room, then leave the rest alone.
6. Re-run in November. Dividends, interest, and any surprise income will have moved the number since your January estimate.

Three ways this goes wrong

Converting in December without re-checking. People plan in January, forget about capital gains distributions in November, and convert into a MAGI that has already moved.

Optimising only against brackets. The single most common error. Bracket room is not the same as usable room in an ACA year, and the tools most people reach for only model the former.

Treating the guardrail as permanent. It is a per-year decision. A year with a large one-off gain might be the right year to abandon the guardrail entirely and convert aggressively, precisely because the subsidy is already lost.

Work out your own number

The Roth Conversion & Capital Gains Headroom Planner estimates your Marketplace MAGI, expresses it as a percentage of FPL, and reports how much additional conversion and harvesting fit under whichever ceiling binds — plus a sensitivity table for $1,000 and $5,000 of extra conversion and $10,000 of extra gains.

For background on what actually lands in that MAGI figure, see what counts toward Marketplace MAGI. If you are still sizing the bridge itself, the Can I FIRE Yet? Calculator shows how many years your cash can carry, and the Couples FI Planner handles households where the two of you stop working in different years.

This is an explanation of how the calculation works, not tax advice. The interaction between conversions, subsidies, brackets, and state tax is genuinely intricate and the rules change — confirm any decision with a qualified tax professional.