Is a Roth Conversion Worth It? Break-Even at Age 78 vs 85

Margaret and Tom Halvorsen are 66 and 64, married filing jointly, retired in Arizona with a $1,200,000 traditional IRA, $400,000 in a taxable brokerage account, and no pension. Their advisor told them to “convert as much as you can” to a Roth. Margaret’s question was sharper: will we actually come out ahead, and when? The honest answer is a number. If they convert in big lumps, they break even around age 85. If they convert smaller amounts each year and pay the tax from the brokerage account, they break even closer to age 78 — and since the actuarial planning horizon for a healthy 66-year-old couple runs into the mid-90s, that earlier break-even turns a maybe into a yes.

The naive break-even is the wrong question

Most people frame the Roth-conversion decision as: “At what point do my total Roth dollars equal what my traditional IRA dollars would have been?” That comparison is misleading because it treats a traditional IRA dollar and a Roth dollar as equal. They are not.

A traditional IRA carries an embedded tax liability. Every dollar inside it is taxed as ordinary income when withdrawn. A Roth dollar is fully spent or inherited tax-free. So comparing nominal balances is like comparing a price before sales tax to a price after — the comparison only works once you put both on an after-tax footing.

On a tax-adjusted basis, a $900,000 Roth is worth more than a $1,000,000 traditional IRA — by $150,000 — if you would otherwise withdraw the traditional at a 25% effective rate. The raw-dollar comparison says the traditional is $100,000 bigger and gets the decision exactly backwards. The right frame is: does converting move dollars from a higher future tax rate to a lower current tax rate?

The single input that decides it: rate today vs. rate later

Strip away every spreadsheet and the Roth-conversion decision reduces to one comparison:

• Convert if you expect your future withdrawal rate to be higher than your conversion rate today.
• Don’t convert if you expect your future rate to be lower, or you’ll need the money within a few years.
• It’s a wash (before second-order effects) if the rates are equal — though the tie still breaks toward Roth because of no RMDs, tax-free heirs, and IRMAA control.

For 2026, the 22% MFJ bracket runs from $96,951 to $206,700 of taxable income, and the 24% bracket extends to $394,600 (IRS Rev. Proc. 2025-32). If converting today fits inside your 22% bracket but your future RMDs would land you in 24% or 32%, the conversion pays. If you’re currently in 24% and your future self will be in 12%, paying tax now is value destruction.

Why your future rate can be higher even if Congress does nothing

OBBBA (July 2025) made the TCJA brackets permanent, so the old “rates will spike when TCJA sunsets in 2026” argument has weakened — you can no longer count on a statutory rate increase as your reason to convert. But your personal rate can still climb for three reasons that have nothing to do with the statutory schedule:

1. RMDs on a large IRA. At age 73 (born 1951–1959) or 75 (born 1960+), SECURE 2.0 §107 forces distributions. On a $1.2M IRA the first-year RMD divisor at 73 is 26.5 (IRS Pub. 590-B, Table III), forcing out roughly $45,000 of taxable income whether you need it or not — on top of Social Security.
2. Survivor single-bracket compression. When the first spouse dies, the survivor files single. The 22% bracket for a single filer ends at $103,350 (2026) — barely half the $206,700 MFJ ceiling. The same income suddenly faces 24% and 32%.
3. Medicare IRMAA cliffs. Part B surcharges begin at $103,000 single / $206,000 MFJ MAGI (2026, based on 2024 MAGI). A large RMD can trip an IRMAA tier and add hundreds per month per person.

The break-even age: lump vs. annual conversions

“Break-even” here means the age at which the after-tax value of having converted overtakes the after-tax value of having left the money in the traditional IRA. The crossover depends mostly on conversion size and how you pay the tax. The numbers below are the rules of thumb the planning literature consistently produces, assuming you pay tax from outside funds and rates are roughly equal:

Read the table against your own longevity expectation. A healthy 65-year-old non-smoker today has roughly a 50% chance of living past 85 and, in a couple, a strong chance at least one spouse reaches the early 90s. A break-even at 78 is a comfortable yes. A break-even at 88 is a coin flip you take only if you have specific reasons to expect long life or you’re explicitly converting for heirs (a Roth is the best asset to inherit — tax-free, and beneficiaries get 10 years to drain it tax-free under the inherited-IRA rules of IRC §401(a)(9)(H)).

The biggest lever: pay the tax from outside the IRA

This one choice can move break-even by 7–10 years. Suppose the Halvorsens convert $100,000.

• Pay tax from the brokerage account: the full $100,000 lands in the Roth. They’ve effectively shrunk their taxable account and grown the tax-free wrapper — the equivalent of an extra contribution. Every future dollar of growth on the full $100,000 is tax-free.
• Withhold tax from the IRA: at a 22% rate, only $78,000 reaches the Roth and the other $22,000 leaves the tax-deferred system entirely. If either spouse is under 59½, that $22,000 also triggers a 10% early-withdrawal penalty. The tax-free balance is smaller, so growth compounds on less.

The Halvorsens have $400,000 in taxable brokerage — ample to cover conversion taxes from outside. That is precisely why their break-even sits at the favorable end. A couple without outside cash, forced to withhold from the IRA, often never reaches break-even before their planning horizon ends.

What most people miss

Three things consistently get overlooked, and each one changes the answer:

1. The widow’s tax cliff is usually the strongest reason to convert. Couples model their joint MFJ brackets and conclude rates are flat. But the first death flips the survivor to single brackets — the 22% ceiling drops from $206,700 to $103,350 (2026). Converting while both spouses are alive and filing jointly uses the wider brackets that disappear at the first death. This single factor can pull break-even years earlier than any longevity assumption.
2. Conversions raise this year’s MAGI — watch the second-order bills. A conversion is ordinary income. It can push you into a higher IRMAA tier two years out (2026 IRMAA uses 2024 MAGI), increase the taxable share of Social Security (up to 85% above $44,000 combined income MFJ), and affect ACA premium subsidies if you convert before Medicare. Convert up to a threshold, not blindly past it.
3. The Roth has no RMDs — the traditional’s forced withdrawals are the hidden cost. The break-even math understates the Roth’s edge because it rarely prices the flexibility of never being forced to withdraw. A Roth lets you control MAGI in every future year. A large traditional IRA hands that control to the RMD divisor table.

A worked decision for the Halvorsens

They’re both retired, Social Security delayed, and their only taxable income before RMDs is about $30,000 of brokerage dividends. That leaves enormous room under the 22% MFJ ceiling of $206,700. Their plan:

• Convert roughly $150,000 per year from 66 to 72 — enough to fill the 22% bracket without spilling into 24% — paying the ~$33,000 annual tax from the brokerage account.
• Over 7 years that moves more than $1,000,000 into the Roth at a blended ~20% effective rate.
• At 73, their remaining traditional balance is small, so RMDs are modest and they avoid the 24%/32% spike a $1.2M IRA would have forced.
• If Tom predeceases Margaret, she inherits a largely tax-free Roth and never faces single-bracket compression on forced RMDs.

Because they’re converting annually and paying tax from outside funds, their break-even lands near age 78 — well inside their planning horizon. The decision is a clear yes.

When the answer is no

Be honest about the cases where converting loses:

• You’ll genuinely be in a lower bracket later. Modest IRA, modest Social Security, no pension — if your RMDs will keep you in the 12% bracket, paying 22%+ to convert now is a loss.
• You’d pay the tax from the IRA. No outside cash means shrinking the very wrapper you’re trying to grow. Break-even often slips past your horizon.
• You’ll need the converted money within 5 years. Each conversion starts its own 5-year clock (the conversion 5-year rule); withdraw converted principal sooner and, if you’re under 59½, you face a 10% penalty.
• A conversion would trip a costly cliff. Pushing MAGI just over an IRMAA tier or an ACA subsidy threshold can cost more than the bracket arbitrage saves. Size conversions to the threshold.

The decision lever

Stop asking “am I too old” and start asking “is my future withdrawal rate higher than today’s conversion rate, and do I have outside cash for the tax?” If both answers are yes — and they usually are for retirees with a $1M+ IRA, a long horizon, and a surviving-spouse risk — then convert annually to the top of your current bracket, pay the tax from your brokerage account, and stop before any IRMAA or subsidy cliff. That sequence is what moves break-even from age 85 down to age 78 and turns the math decisively in your favor. The conversion that fails is the rushed lump sum funded from the IRA itself; the conversion that wins is the measured, bracket-filling, outside-funded one.