The Conversion Runway

There is a deadline coming that does not move, cannot be extended, and has no grace period.

December 31st.

Most tax moves give you until April. You can fund an IRA for last year in April. You can fund an HSA in April. You can file an extension and push a return to October. The tax code is surprisingly forgiving about timing.

A Roth conversion is not. If the money does not leave the traditional account and land in the Roth by the last business day of the year, it is a next year conversion at next year's numbers. And since 2018 you cannot undo one. Recharacterization of conversions is gone. What you convert, you own, at the tax cost it created.

Ninety seven days from today.

I want to spend this issue on the part of conversions that gets skipped. Everybody understands the concept: pay tax now at a known rate to avoid tax later at an unknown one. Fine. That is the easy half.

The hard half is sizing. How much should actually move this year? And that question has five separate answers, because there are five different ceilings sitting above you, and the lowest one wins.

For years the conversion pitch ran on a simple clock: rates are scheduled to go up in 2026 when the old law sunsets, so convert before then.

That clock is gone. The 2017 brackets were made permanent. The seven brackets with a top rate of 37 percent are not expiring. If your entire conversion thesis was "rates are going up by law," you need a new thesis, and a lot of people have not updated.

Here is the new one, and it is better.

The law that made the brackets permanent also created a set of deductions that are explicitly temporary, running through 2028. A deduction for tip income up to 25,000 dollars. A deduction for overtime premium pay up to 12,500 dollars, or 25,000 dollars for joint filers. A bonus deduction of 6,000 dollars for filers 65 and older, 12,000 dollars for a couple where both qualify. A state and local tax cap raised to 40,400 dollars for 2026, up from 10,000, scheduled to revert in 2030.

Every one of those is temporary, and every one of them phases out based on income.

That combination is the whole ballgame. Temporary deductions create artificially low tax years, and artificially low tax years are exactly when you want to recognize income. But income based phaseouts mean that recognizing too much income in one of those years claws the deduction back at a brutal effective rate.

So the window is real and the window has a shape. Let us map it.

This is the one everybody knows, so I will be quick.

You are filling a bracket. Find your projected taxable income for 2026, find the top of the bracket you are sitting in, and the gap between them is your first ceiling.

A couple filing jointly with 180,000 dollars of taxable income is in the 22 percent bracket. The 24 percent bracket starts somewhere in the mid two hundreds. That gap is conversion room at 22 cents on the dollar.

The mistake here is using last year's return as your projection. Do not. Use a real projection: year to date pay stubs, expected bonus, business income through Q4, dividends and capital gains distributions that have not been declared yet, and any interest income that jumped this year because cash yields went up.

That last one bites people. If you moved 200,000 dollars into Treasury bills and money market funds over the last two years, you generated meaningful ordinary income that was not on your prior return. Rates went up again last week. Interest income is real income and it eats conversion room.

Here is where the effective rate stops matching the bracket, and where most conversion planning quietly fails.

When a deduction phases out as income rises, every extra dollar of income does two things. It gets taxed, and it shrinks a deduction, which exposes another dollar to tax. Your true marginal rate is higher than your bracket, sometimes dramatically.

Take the senior deduction. It phases out at 6 percent of every dollar of income over 75,000 dollars single, 150,000 dollars joint. So for a couple in that range, each additional dollar of conversion income costs the bracket rate plus 6 percent of the bracket rate in lost deduction. A 22 percent bracket becomes roughly 23.3 percent. A 24 percent bracket becomes roughly 25.4 percent. Not catastrophic, but it is not 24 percent and you should not be planning as if it is.

The tips and overtime deductions phase out starting at 150,000 dollars of modified income for single filers and 300,000 dollars for joint. If you qualify for those, a conversion that pushes you through the threshold can cost you a deduction worth up to 25,000 dollars. That is not a nudge to your effective rate, that is a wall.

And the state and local cap phases down above 505,000 dollars of modified income for 2026, with the phaseout applied at a 30 percent rate. Inside that range, a dollar of conversion income can carry an effective cost well north of 45 percent. If you are a high earner in a high tax state, this is the single most expensive zone on the map and it is completely invisible if you are only looking at brackets.

The discipline: before you size a conversion, list every deduction and credit you are claiming that has an income threshold attached. Write down each threshold. The lowest one above your projected income is a real ceiling, not a suggestion.

If you are 63 or older, this one is probably the binding constraint and it is the one most people discover two years too late.

Medicare Part B and Part D premiums are income adjusted through something called IRMAA. The income they use is your modified adjusted gross income from two years prior. So your 2026 income sets your 2028 premiums.

IRMAA is not a phaseout. It is a cliff. Cross a threshold by one dollar and you pay the entire next tier, for both Part B and Part D, for both spouses, for the full year. Depending on the tier, one dollar over can cost a couple well over 2,000 dollars.

So if you are anywhere near 63, your conversion ceiling is not a bracket edge. It is the IRMAA tier below you, with a safety buffer of a few thousand dollars because you cannot perfectly predict December interest and dividend income.

There is an appeal process if you have a life changing event, which includes retirement, so if you convert in the year you stop working, know that form exists. But do not plan around winning an appeal.

Conversion income is not net investment income. It does not get hit by the 3.8 percent surtax directly. People hear that and relax.

They should not. The surtax applies to the lesser of your net investment income or the amount your modified income exceeds 200,000 dollars single, 250,000 dollars joint. Conversion income raises your modified income. So a conversion can drag your existing dividends, interest, and capital gains into the surtax even though the conversion itself is exempt.

If you have meaningful taxable investment income and you are sitting just below those thresholds, a conversion has a hidden 3.8 percent surcharge attached to it. Model it.

A short list of cliffs that do not bend, they snap.

Premium tax credits. If you are on a marketplace health plan, conversion income can reduce or eliminate your subsidy, and reconciliation happens at filing. This is often the largest single cost of a conversion for early retirees and it is routinely missed.

The zero percent capital gains bracket. Long term gains are taxed at zero up to a taxable income threshold. If you are sitting inside that band, you have a choice to make: harvest gains at zero percent, or convert at 10 or 12 percent. You generally cannot do both with the same dollars of room. Gain harvesting usually wins for younger investors with appreciated taxable accounts. Conversion usually wins for people staring at large future required distributions.

Financial aid. If you have a kid two years from a FAFSA, conversion income lands in the base year and it counts. Coordinate or wait.

Now the operational part. This is the order I run it in.

Step one, project in October. Build your income projection now, with the year still open enough to change. Year to date pay, expected Q4 income, business distributions, interest, dividends, realized gains.

Step two, list every threshold above you. Bracket top. Every phaseout in the stack. IRMAA tier if relevant. Surtax threshold. Subsidy cliff. Write them in ascending order. The lowest one above your projection is your ceiling.

Step three, subtract a buffer. Take 5 to 10 percent off your available room. December mutual fund distributions are declared late and you do not control them. Interest accrues. Someone sells something. The buffer is what keeps a good plan from becoming an IRMAA cliff crossing by 400 dollars.

Step four, convert in December, not now. Everything above is planning. The execution should happen in the first half of December, when your actual income is nearly fully known and December distributions have been declared or estimated. Converting in September on a projection is how people convert too much.

Step five, pay the tax from outside the account. If you have 100,000 dollars to convert and you withhold the tax from the conversion itself, you did not convert 100,000 dollars. You converted less, and if you are under 59 and a half, the withheld portion is a distribution subject to penalty. Pay from taxable cash. Always. If you do not have taxable cash to pay the tax, the conversion is usually smaller than you think it should be.

Step six, handle the estimated tax correctly. A December conversion creates a tax liability the IRS considers due throughout the year, which is how people end up with underpayment penalties on an otherwise smart move.

Two clean solutions. Hit a safe harbor by paying in 100 percent of last year's total tax, or 110 percent if your adjusted gross income was over 150,000 dollars, and you are protected regardless of what you convert. Or, take the tax payment as withholding from an IRA distribution in December, because withholding is treated as paid evenly across the year no matter when it actually happens. That second one is a genuinely useful trick and almost nobody uses it.

I am not in the business of selling you a move you do not need.

Skip the conversion if you expect to be in a meaningfully lower bracket later, which is most people with a big income now and a modest retirement plan. Skip it if you will need the converted money inside five years, because each conversion carries its own five year clock before earnings come out clean. Skip it if paying the tax requires selling appreciated assets that generate their own gains. Skip it if you are on a subsidized marketplace plan and the credit loss exceeds the benefit, which it very often does.

And be honest about one more thing. Higher rates cut slightly against conversions. When cash earns 4 percent, a dollar of tax paid today has a higher opportunity cost than it did when cash earned nothing. That is a real consideration at the margin. It does not kill the case, but anyone selling you conversions as a no brainer in a 4 percent world is not doing the math.

Two things I use.

For modeling the phaseout stack, I do not trust mental math and neither should you. I run the scenarios through Galaxy.ai with the thresholds and my projection laid out, asking it to compute effective marginal rate at 10,000 dollar conversion increments. Seeing the effective rate curve, rather than a bracket label, is what makes the ceiling obvious.

For execution, I have a Make.com scenario that fires December 1st with the full checklist, pulls my year to date numbers into a summary, and does not stop reminding me until I mark it done. A deadline with no reminder attached is a wish.

Want the model?

The Conversion Runway Calculator is the workbook I run this in. Income projection sheet, the full threshold ladder with every 2026 phaseout and cliff pre-loaded, the effective marginal rate curve, the buffer logic, the safe harbor calculator, the December execution checklist, and the withholding trick written out step by step.

Reply to this email with the word RUNWAY and I will send it.

And if you want help running it.

The Grid Inner Circle is the paid community where this stuff gets implemented instead of admired. We are doing a live conversion sizing session in early December, which is exactly when this decision has to get made, and members bring their actual projections.

29 dollars a month. Lifetime access is 499 dollars and stays open while the room is small.

Reply with CIRCLE for the details.

Sunday's issue is The Edge, and it is about something that has been sitting under all three issues this week.

Alex Rivera, Wealth Architect at Wealth Grid