Roth Conversion Ladders and the Four Times Not To Convert

In the FIRE playbook, the Roth conversion ladder has hardened from tactic into ritual. Retire with a seven-figure pre-tax 401(k), convert to the top of a low bracket every year before Social Security, season each batch for five years, and withdraw tax-free for the back half of a 40-year retirement. The popular guides, including the White Coat Investor's conversion guide, argue at length about how much to convert to Roth in gap years before Social Security. Almost nothing in standard early retirement tax planning asks whether this year's conversion should happen at all, which is why a Bogleheads thread titled Anyone NOT Roth Converting? reads like dissent. The posters declining are not confused, and their reasons double as a checklist for when not to do Roth conversions in FIRE. Each situation has a dollar price, and in all four the correct amount for the year can be zero.

The logic underneath is a single comparison. A conversion adds wealth only when the rate you pay today is lower than the blended rate those same dollars would otherwise face later, and later means more than the tax table at age 75. The real later rate bundles RMD-era brackets, lost health-insurance subsidies, Medicare surcharges, and eventually your heir's bracket. Price those four vetoes in dollars per year and the endless how-much debate mostly resolves itself, which is the point.

What the Roth Conversion Ladder Default Actually Buys

The mechanics fit in one paragraph. You move traditional IRA money into a Roth and pay ordinary income tax at this year's rates. Each conversion opens its own five-year clock; once a batch seasons, its principal comes out tax-free and penalty-free even before age 59½, which is the entire point for early retirees. Earnings wait for age 59½ and the account's own five-year clock. Repeat annually and the ladder bridges the years between an early retirement and penalty-free access.

Now the default in dollars. A couple, both 58, living on $30,000 of taxable-account dividends and gains can convert roughly $95,000 in 2025 while staying inside the 12% bracket for joint filers, which holds about $96,000 of taxable income once you add the roughly $30,000 standard deduction. Blended, the conversion costs around 11 to 12 cents on the dollar.

The other side of the trade arrives at 73, when required minimum distributions begin under current RMD rules, age 75 for those born in 1960 or later. A $2 million pre-tax balance forces roughly $80,000 of taxable income at age 75 before counting two Social Security checks, which plausibly stacks the couple into the 22% bracket with Medicare surcharges layered on top. Filling the brackets at 12% now instead of absorbing 22% or worse later, across a few hundred thousand dollars of conversions, is a spread worth tens of thousands. That is the default, and for most FIRE savers it holds.

Every veto below is one component of that later rate spiking above the gap-year rate. The fill-the-brackets ritual assumes future rates are always higher. Usually true here. Nameably false in four cases.

Veto 1. Roth Conversions and ACA Premium Tax Credits

Does a Roth conversion affect ACA premium tax credits?

Yes, and in full. Conversion income is ordinary income, and marketplace subsidies are keyed to MAGI with no carve-out for conversions, as the HealthCare.gov premium tax credit rules make clear. The credit works by capping your premium contribution as a share of income, so this is where Roth conversions and ACA subsidies collide: extra MAGI either raises what you must pay toward a benchmark plan or, past a threshold, wipes the credit out.

Size the exposure before touching a bracket chart. Unsubsidized silver premiums for adults in their early 60s often run $1,200 to $1,500 per person per month, so a retired couple's annual credit can exceed $15,000. Two mechanisms attack it:

1. The contribution cap. Where the 8.5% cap binds, each conversion dollar effectively costs its tax rate plus up to about 8.5 cents of lost credit. A conversion you priced at 12% actually runs near 20.5% all-in, uncomfortably close to the 22% you were deferring it past.
2. The cliff. The enhancements that currently smooth the 400% federal poverty line, about $84,600 of MAGI for a two-person household in 2025, are scheduled to expire after 2025 absent new legislation, and KFF's expiration analysis prices what returns: above the line, no credit at all. A single dollar over the threshold can forfeit five figures of subsidy. Verify whether the enhancements were extended before leaning on the smoother math.

If advance credits were paid on an estimate you blow through with a December conversion, reconciliation lands on Form 8962, and IRS Publication 974 walks the repayment arithmetic. The practical veto: before 65, cap conversions below the binding threshold, and when the credit is large, the right conversion for the year is zero. Near a threshold, the clawback can exceed the entire tax advantage of converting. This is the most time-sensitive veto on the list, because subsidy law is in motion while the rest of the math is stable.

Veto 2. The IRMAA Lookback at Ages 63 and 64

Medicare bills you on old income. IRMAA and Roth conversions collide on a two-year delay through the two-year MAGI lookback: assuming Medicare starts on schedule at 65, premiums at 65 bill the age-63 tax return, and premiums at 66 bill age 64. A conversion made at 62 never reaches a Medicare bill. Conversions at 63 and 64 absolutely do, and so does every conversion made after Medicare starts.

Each tier is a cliff with no phase-out, per the 2025 IRMAA brackets. The first tier for joint filers begins near $212,000 of MAGI and adds roughly $1,050 per covered spouse for the year across Part B and Part D, about $2,100 for a couple.

The dollar trap follows directly. A couple converting toward the top of the 22% bracket pushes MAGI near $236,000, crossing that first tier by roughly $25,000. Spread across an overage that large, the surcharge adds about 8 percentage points to those marginal dollars, a defensible price if the later rate is higher. Cross the tier by $1,000 and the same roughly $2,100 rides on $1,000 of income, a marginal cost north of 200% on the stragglers. IRMAA cliffs are lethal in proportion to how small the overage is.

So model ages 63 and 64 as their own years, separate from the rest of the ladder. Divide the couple's surcharge by the size of any intended overage, add that to the bracket rate, and compare the sum to the later rate those dollars face. Below the tier, convert as planned. Above it, either cross deliberately with a large overage or stop at the tier line. Small stumbles over the edge are the expensive ones.

Veto 3. The Pro-Rata Rule and Stranded After-Tax Basis

One poster in that Bogleheads thread skips conversions for exactly this reason: years of nondeductible IRA contributions left the basis records murky, and the pro-rata rule froze the whole strategy. The fear holds even with clean records, because the rule forbids the one move people assume is available.

You cannot convert only the after-tax money. Under Publication 590-B and the Form 8606 instructions, the pro-rata rule taxes every conversion in proportion to your after-tax basis across all your traditional, SEP, and SIMPLE IRA balances at year-end. Hold $90,000 of pre-tax money and $10,000 of basis, and you are 10% basis, full stop. Convert $10,000 and $9,000 of it is taxable income. The free conversion you hoped for costs about $1,980 at a 22% gap-year rate, purely to relabel money you already owned. The same aggregation is what poisons backdoor Roth contributions for anyone with stranded pre-tax balances, the exact worry in the thread.

The escape hatch into an employer plan

Dollars rolled into a 401(k), 403(b), or governmental 457(b) leave the Form 8606 denominator entirely. Roll the pre-tax IRA balance into an employer plan that accepts IRA roll-ins, including a solo 401(k) if you have any self-employment income, and what remains in the IRA is your after-tax basis. Convert that remainder and the taxable share approaches zero. This veto is procedural rather than permanent: resolve the basis before the ladder starts, or accept that every conversion carries its pro-rata taxable share and price that into the year's cliff map. If the basis is murky, reconstruct the Form 8606 history first, because basis only exists to the extent prior returns tracked it.

Veto 4. Heir Brackets Under the 10-Year Rule

The SECURE Act ended the stretch IRA, and its inherited IRA 10-year rule requires most non-spouse heirs to empty the account within ten years. Under the final 10-year regulations, when the owner dies after required minimum distributions have begun, heirs generally must also take annual distributions inside that window. Spouses, the owner's minor children, and disabled or chronically ill beneficiaries keep better options.

That turns every unconverted dollar into a bracket transfer. The question readers ask, should I convert if my heirs are in a lower tax bracket, is really a cleaner one: whose marginal rate on these dollars is lower, yours this gap year or theirs during the emptying decade? Run both directions against a 24% gap-year conversion rate on a $500,000 pre-tax balance.

• The high-earning heir. A child earning $180,000 absorbs most of each year's $50,000 distribution at 32%. Converting at 24% during gap years wins outright, and the ladder is doing its job.
• The cheaper heir. A child earning $45,000, married, in a state with no income tax, absorbs the same $50,000 almost entirely at 12% once the standard deduction applies. Converting at 24% burns roughly 12 cents on the dollar, and skipping the conversion on those dollars leaves the estate larger after tax.

Two nuances keep this honest. Dollars you will actually spend in your RMD era still compare against your own later rate; the heir comparison governs the surplus you will never touch, which for many FIRE estates is most of the account. And money you converted but did not spend passes as Roth, which the heir empties on the same ten-year clock but tax-free, so the comparison stays clean on both sides.

The Fill-or-Defer Rule for Every Gap Year

Collapse everything above into one instruction:

Fill to the nearest cliff and stop. Bracket tops, premium tax credit thresholds, and IRMAA tiers are all cliffs. The cheapest dollars sit before the first one, and the right conversion amount is whatever reaches it, including zero.

The per-year checklist

1. Pull Form 8606 first. Total basis, total IRA balances. The pro-rata math changes the taxable share of everything downstream, so it leads.
2. Write down this year's cliffs in MAGI dollars. The bracket top you will pay, the ACA threshold if you are under 65, the IRMAA tier if you are 63 or older.
3. Price the next $1,000 all-in. Bracket rate, plus lost credit cents, plus the couple's surcharge divided by any intended overage, plus the pro-rata taxable share.
4. Compare it to the later rate. Your own RMD-era stack for dollars you will spend, your heir's blended rate for dollars you will not.
5. Convert up to the nearest cliff where today's price still wins, and not a dollar past it.
6. Re-run every single year. The answer moves with the thresholds, the subsidies, and the calendar, and a Roth conversion ladder run on autopilot quietly assumes none of them changed.

What changes every year

The numbers here are 2025 vintage. Bracket tops, poverty-line multiples, IRMAA tiers, and the subsidy structure all move annually or with legislation. The method does not: compare an all-in price today against a blended later rate, treat every threshold as a cliff, respect the two-year lookback calendar, and check your heir's bracket before converting surplus. The 58-year-old couple from the first section sails through all four checks. The same couple at 63, one tier below a cliff and with a daughter absorbing an inheritance at 12%, may rationally convert nothing that year. That is the ladder working, not failing. A repeatable rule earns its keep precisely because it can say no.