Roth Conversion Ladder: The 11% Break-Even Rate

Key takeaways

• A couple both 65 or older can report $148,300 of income in 2026 and stay inside the 12% bracket, once the $32,200 standard deduction, the $3,300 age addition and the $12,000 senior deduction come off.
• Filling that whole band from a $40,000 base converts $108,300 and costs $11,600 of federal tax. That's an average rate of 10.7% on the converted money, not 12%.
• Ten years of that ladder converts $999,000 by age 75 at a cumulative cost of $119,102. Measured on tax paid alone, it doesn't break even until age 87.
• Measured on after-tax wealth instead, the future rate that leaves the household indifferent is 10.98% at a ten-year horizon and 7.90% at twenty.
• Required distributions begin at 3.77% of the balance, from the Uniform Lifetime Table divisor of 26.5, and reach 11.24% by age 95 at a divisor of 8.9.

Does converting early actually pay, and after how many years

You've got a large traditional IRA, a gap between retiring and required distributions, and someone has told you to fill your low tax brackets with conversions. You want to know whether the arithmetic supports it, and how long you'd wait to find out.

Two honest answers exist and they point in opposite directions. On cumulative federal tax paid, a ten-year ladder in the household modelled below doesn't cross over until age 87. That's 22 years of being behind. On after-tax wealth, the same ladder is ahead almost immediately, because the break-even future tax rate turns out to be 10.98% at a ten-year horizon and 7.90% at twenty. Both numbers come from the same model. They disagree because they measure different things, and the second one is the one that answers the question.

A Roth conversion ladder is just that trade repeated. Here's the whole mechanism in a sentence. A conversion moves money out of a traditional IRA into a Roth and you pay ordinary income tax on the amount moved that year. After that, Publication 590-B is blunt about what happens: you don't include qualified distributions in gross income. What you're buying is the removal of a future tax liability of unknown size. What you're paying is a tax bill of known size, today, in exchange.

How much bracket space a 2026 Roth conversion ladder actually has

Start with the space, because the space is the whole design. Revenue Procedure 2025-32 sets the 2026 figures. For a married couple filing jointly, the 10% bracket runs to $24,800 of taxable income and the 12% bracket ends at $100,800. Above that, the next dollar is taxed at 22%, and the jump from 12% to 22% is the largest proportional step in the schedule.

Deductions decide how much gross income that $100,800 represents. The 2026 standard deduction for a couple is $32,200. Each spouse aged 65 or over adds $1,650, so $3,300 for two. And for tax years 2025 through 2028 only, the enhanced deduction for seniors adds $6,000 per qualifying person, or $12,000 for a couple where both are 65 or older, phasing out above $150,000 of modified adjusted gross income.

Add those and you get $47,500 of deductions. A couple can therefore report $148,300 of income in 2026 and still land exactly on the top of the 12% bracket. Our piece on withdrawal order uses $136,300 for the same ceiling, and both figures are right: $136,300 is the ceiling without the senior deduction, which is how the arithmetic looks again from 2029 under current law. The temporary provision widens the window by $12,000 a year per couple, and only for three more tax years.

Filling the whole 12% band costs 10.7%, because the bottom of it is free

Take a couple who both turn 65 in 2026, hold $1,200,000 in a traditional IRA, and have $40,000 a year of pension and interest income. They haven't claimed Social Security yet. Their headroom to the top of the 12% bracket is $148,300 minus $40,000, which is $108,300 of conversion.

The tax on that isn't 12% of $108,300. The first $7,500 of it is absorbed by unused deductions and taxed at nothing. The next $24,800 is taxed at 10%. Only the remaining $76,000 meets the 12% rate. Total federal tax: $11,600. Divide by the amount converted and the average rate on the conversion is 10.71%.

That gap between the posted rate and the paid rate is the thing bracket-filling exploits, and it shrinks as your other income rises. A household already reporting $100,000 before conversions has no deduction gap and no 10% band left to use. Their headroom is $48,300 and every dollar of it costs 12%. Same bracket, different price.

The posted rate is rarely the rate a conversion dollar pays

Three things sit on top of the published schedule, and each of them is triggered by the conversion itself.

The first is Social Security. Publication 915 sets the base amount for a couple at $32,000 and the adjusted base amount at $44,000. Above the higher figure, up to 85% of benefits become taxable. Those are flat dollar amounts, unchanged in the 2015 edition of the same publication, so they don't move with inflation the way the brackets in Revenue Procedure 2025-32 do. Inside the phase-in range, each dollar of conversion also drags 85 cents of benefit into tax, which turns a posted 12% into 22.2%.

The second is capital gains stacking. The 0% long-term capital gains ceiling for a couple is $98,900 of taxable income in 2026, and the 12% ordinary bracket ends at $100,800. Those two cheap zones almost coincide, so ordinary income from a conversion pushes gains that were free into the 15% band. A dollar converted can cost 12 cents of income tax and displace 15 cents of gains, for 27 cents.

The third isn't a rate at all. Medicare's income-related monthly adjustment is a cliff. CMS set the standard 2026 Part B premium at $202.90 a month, and a couple whose modified adjusted gross income exceeds $218,000 pays $284.10 instead. Social Security's own manual confirms the 2026 tier is set from income reported for 2024, two years earlier. That $81.20 monthly step, applied to two people for twelve months, is $1,948.80 of extra premium triggered by a single dollar of conversion. Add the $14.50 Part D surcharge at the same tier and it's $2,296.80. There's no taper, so a ladder built to a bracket ceiling and a ladder built to an IRMAA threshold are different ladders.

None of these bite the household in the model, whose income tops out at $148,300 and who haven't claimed benefits. That's not luck. It's what the window is for.

What required minimum distributions would have taken anyway

The ladder only makes sense against a counterfactual, and the counterfactual is the required distribution. The IRS requires withdrawals from a traditional IRA starting at 73 for most people alive today. Congressional Research Service tabulation of SECURE 2.0 puts the applicable age at 75 for anyone born on or after 1 January 1960, which the same table defines as the cohort that turns 73 after 31 December 2032. Someone turning 65 in 2026 was born in 1961, so their first required distribution falls in 2036. The window is ten years, not eight, and that matters more than any single-year decision inside it.

The size of the distribution is a division. Publication 590-B's Uniform Lifetime Table gives an applicable denominator of 26.5 at age 73, 24.6 at 75, 16.0 at 85 and 8.9 at 95. As a share of the balance that's 3.77%, 4.07%, 6.25% and 11.24%. Vanguard's own description matches: slightly more than 4% at 75, more than 11% at 95.

Run the model forward without conversions. The $1,200,000 grows at 5% for ten years to $1,954,674 by the end of 2035. The first required distribution in 2036 is $79,458. On top of $40,000 of other income that's $119,458 of gross income, comfortably inside the 12% bracket even without the senior deduction. Run it with the ladder and the traditional balance at 75 is $631,555, the required distribution is $25,486, and the tax that year is $3,102 rather than $9,579.

So the ladder does what it says. It just doesn't rescue this household from a rate disaster, because they didn't have one coming.

A ten-year Roth conversion ladder, year by year, against doing nothing

Both paths below start from the same $1,200,000 traditional IRA, the same $40,000 of other income, the same 5% growth, and a $250,000 taxable account that pays the tax and receives the distributions. Deductions drop from $47,500 to $35,500 in 2029 when the senior deduction expires, which is why the conversion falls from $108,300 to $96,300.

Ten years of conversions move $999,000 and cost $119,102 in tax by age 75. The household that did nothing has paid $12,729 over the same period. Cumulative tax paid doesn't equalise until age 87, at $168,283 for the ladder against $178,863 for doing nothing. If the question is which household hands more money to the Treasury over a lifetime, the ladder is behind for 22 years.

That framing is popular and it's the wrong ledger. Tax paid isn't the objective. What the household has left is.

The break-even tax rate falls as the horizon lengthens

Vanguard's July 2025 paper on the break-even tax rate makes the argument formally. Their BETR is the future rate at which the after-tax withdrawal value is identical whether you convert or not. If your expected future rate sits above it, converting comes out ahead. Their headline result is that the break-even rate sits well below the current marginal rate whenever the conversion tax is paid from outside the IRA. At a 35% current rate and a 20-year horizon, paying from a tax-efficient taxable portfolio puts the break-even at 30.1%, paying from a tax-inefficient one puts it at 23.5%, and paying from cash puts it at 14.1%. Their case study, Jill, converts $100,000 at 35% expecting a 24% future rate, and her break-even works out at 23.3%.

Applying the same logic to the ladder gives a sharper number, because the ladder converts at a low rate rather than a high one. Value each household's position as the Roth balance, plus the taxable account, plus the traditional balance net of a single flat future rate. Solve for the rate that makes the two equal. At age 75 it's 10.98%. At 80, 10.22%. At 85, 7.90%. At 90, 2.03%.

Read that carefully, because it's the finding. The ladder pays off unless the household's future tax rate is below roughly 11%, and it falls further with every year the Roth compounds untaxed. The lowest posted bracket is 10% and it only runs to $24,800. A retired couple would have to end up with almost no taxable income at all for the ladder to have been a mistake on these assumptions. That is a much weaker condition than "do you expect rates to rise", which is the question people usually think they're answering.

Where a Roth conversion ladder loses

The strongest objection isn't about rates. It's that some traditional IRA dollars were never going to be taxed at all.

A qualified charitable distribution is, in Publication 590-B's words, generally a nontaxable distribution made directly by the IRA trustee to an eligible organisation, available from age 70 and a half, with a maximum annual exclusion of $108,000 per person. Money that leaves the account that way is never taxed at all. Converting it first and paying 10.7% turns a zero into a positive number. A household with genuine charitable intent has a portion of the balance for which every conversion is a loss, and the size of that portion is the size of the mistake.

Heirs matter for the same reason in reverse. Publication 590-B puts beneficiaries of a traditional IRA under an obligation to include any taxable distributions in their own gross income, so if they're in a lower bracket than the converter, the balance is taxed at their rate rather than yours. If they're in a higher one, the Roth wins by more than the model shows.

Then there's the irreversibility. Publication 590-A is explicit that there are no recharacterizations of conversions made in 2018 or later, so a conversion done in a year that turns out badly can't be undone. And Publication 590-B attaches a five-year clock to each conversion: take a distribution of converted money inside that window and the 10% additional tax on early distributions can apply to the amount you had to include in income. A separate five-year period runs for each conversion, which makes a ladder a stack of overlapping clocks rather than one.

Finally, dying early. The ladder prepays tax to buy decades of untaxed compounding. A household that doesn't get the decades has bought something it didn't use. The break-even table above is also a mortality table read sideways.

What this arithmetic cannot tell you

The model is federal only. State income tax on conversions varies enormously and can reverse the answer on its own, particularly for someone converting in a high-tax state and drawing in a low-tax one, or the reverse.

Every figure past 2026 assumes today's statute holds, which is an assumption and not a forecast. Congress has changed the RMD age twice since 2019 alone, and the direction of the next change isn't knowable from here. The senior deduction expires after 2028 by its own terms, which the model does capture. Nothing else is scheduled, which isn't the same as nothing else changing.

The 5% growth assumption does real work here. A higher return makes the Roth's untaxed compounding worth more and pushes the break-even rate lower; a lower one does the opposite. And sequence matters, not just the average. Converting into the year before a large drawdown means paying tax on a balance that then falls, which is the same asymmetry that drives sequence risk on the withdrawal side.

The model also assumes the conversion tax is paid from a taxable account rather than withheld from the IRA. Vanguard's Scenario 1 shows why that matters: when the tax comes out of the IRA itself, the break-even rate is simply the current marginal rate, and the entire advantage disappears.

One more limit. Everything here is a single deterministic path. It isn't a distribution, it carries no probability, and it says nothing about how a household would feel about a $119,102 bill for a benefit that arrives 20 years later. LedgerTouch models the account balances and the bracket boundaries; it doesn't model that.

What would change the conclusion

If your traditional balance is small relative to spending, the case largely evaporates. The ladder's value comes from a required distribution that eventually exceeds what you wanted to withdraw. If 3.77% of your balance is less than you were taking anyway, there's no forced income to pre-empt and the conversion is just an early tax payment.

If a meaningful share of the balance is earmarked for charity, the break-even rate calculation applies to the wrong denominator. Those dollars have a future tax rate of zero, and no positive current rate beats zero.

If the 0% capital gains band or the taxation of Social Security is restructured, the interaction effects change shape. The 27% and 22.2% effective rates in this piece are products of two rules touching, not features of the conversion itself, and either rule moving alters which brackets are genuinely cheap.

The variable to watch isn't the tax bracket you're in now. It's the ratio of your tax-deferred balance to your spending, because that ratio determines whether required distributions ever arrive as income you didn't want. Everything else in the ladder arithmetic is a second-order adjustment to that one number, and it's the same number that decides whether a safe withdrawal rate has any slack in it, and whether the pro-rata trap in a backdoor Roth is worth stepping around in the first place.