A high-tax state makes converting worse, not better. The threshold is 5.58 points.

This article first appeared in The Nest Egg, my daily email, on 2026-10-08. Read the original issue. It is educational, not a recommendation for you. The household in it is a hypothetical example, and market figures are as of the date shown and will have changed.

Every figure below names its document. Every calculation is shown. Nothing here is a forecast.

Good morning. A broad retreat, the Russell 2000 worst at 1.31% and gold off 1.35%, with the VIX barely moving. Then the state tax layer, which turns out to run backwards from the way almost everyone carries it.

In today's issue

The tape. Everything lower, gold down hardest, and a selloff with no fear in it. Two minutes.

3.99%. North Carolina's rate, from the state's own page, and what it does to the conversion. Four minutes.

5.58 points. The only number that decides whether geography matters at all. Five minutes.

One more thing. The exemption that can set a child's rate to zero, and the trigger that may cut these parents' rate again. Two minutes.

The tape

Close of October 7, 2026. Index and futures levels from CNBC US Markets, pulled at 4:15pm ET. Treasury yields are as of that moment, not the official close. Gold is the December 2026 contract and WTI the November 2026 contract, which expires October 20.

A broad retreat. All four indexes finished lower, the Russell 2000 worst at 1.31% and the Dow down 0.66%, while the VIX added only 0.47% to 15.08. A selloff with no fear in it is usually positioning rather than news. Gold fell 1.35% to $4,130.60, its largest single-session drop on this page this month, and crude eased 0.51% to $88.98 on the November contract, now two weeks from expiry. The 30-year rose to 5.669% and the 10-year finished where it started, at 5.286%.

The story

Yesterday's letter ended on a promise: that the child's state of residence could flip the conversion decision, and that the swing might be larger than the 5.58-point federal margin that carried the answer.

The first half is true. The second half is true for some households and not for this one.

And the direction is the opposite of what almost everyone assumes. A high state tax rate does not make converting more attractive. It makes it worse.

The position

Carried forward and reconciled: a $165,000 conversion costing $27,100.40 in federal tax and Medicare surcharge, a break-even of 16.42%, and a child aged 48 who inherits at 58 and faces a blended federal rate of 22.00% if they level the withdrawals and 23.79% if they take only the required minimums.

The parents are in North Carolina. This letter is written there, and until today every figure in it has been federal only.

The arithmetic

Step one. The rate, from the state's own page.

From the North Carolina Department of Revenue's tax rate schedules, read yesterday, quoting Session Law 2023-134:

"For Taxable Years beginning in 2025, the North Carolina individual income tax rate is 4.25% (0.0425). For Taxable Years after 2025, the North Carolina individual income tax rate is 3.99% (0.0399)."

So the 2026 rate is 3.99%, flat, with no brackets. The same page adds that further reductions may apply from 2027 under revenue triggers, which matters later.

Step two. The state taxes both sides, not one.

This is where the common intuition goes wrong. People picture state tax as something the Roth escapes, and stop there.

A Roth conversion is ordinary income in North Carolina in the year it happens. So the parents pay 3.99% of $165,000, or $6,583.50, today.

The conversion now costs $25,952.00 of federal tax, $6,583.50 of state tax, and the $1,148.40 Medicare surcharge billed in 2028. Total $33,683.90. The cost rate rises from 16.42% to 20.41%.

And the inherited traditional distributions are ordinary income in the child's state, whenever and wherever that is. So the state rate lands on both sides of the comparison.

Step three. Which means it mostly cancels.

Write the decision out. Convert if the heir's total rate beats the total cost rate:

federal heir rate + child's state rate, against federal cost rate + parents' state rate.

Move the state terms together and the whole thing collapses to one line:

(federal heir rate less federal cost rate) must exceed (parents' state rate less child's state rate).

The state rates survive only as a difference. If the parents and the child live in the same state, that difference is zero and the answer is exactly yesterday's federal answer, whatever the rate happens to be. A family in a 9% state and a family in a 4% state get the identical decision so long as nobody moves.

Step four. So there is one threshold, and it is 5.58 points.

The left side is the federal margin from yesterday: 22.00% less 16.42%, or 5.58 points on the levelled schedule. On the required-minimum schedule it is 7.37 points.

So geography flips this decision only when the parents' state rate exceeds the child's by more than 5.58 points.

The threshold is the federal margin from yesterday, 22.00% less 16.42%, or 5.58 points. Converting wins wherever the parents’ state rate exceeds the child’s by less than that. North Carolina’s 3.99% is from the Department of Revenue’s tax rate schedules, read October 7, 2026; the comparison rates are illustrative.

Step five. For these parents, it cannot happen.

North Carolina is 3.99%. The lowest any child's rate can be is zero.

So the largest gap this household can ever face is 3.99 points, against a threshold of 5.58. Even in the worst case, a child who pays no state income tax at all, converting still wins.

What changes is the size of the win, and it changes a lot. In dollars at the end of the child's tenth year:

Child in North Carolina too: converting is worth $51,712.19.

Child in a state with no income tax: $35,339.95.

Child in a state at twice North Carolina's rate: $68,084.44.

Yesterday's federal-only figure was $52,807.95. The state layer moves the answer by up to $17,467 in either direction without ever changing what the household should do.

The direction everyone gets backwards

Look at the bottom three rows of the table again.

A household in a 7.00% state, with a child who pays no state tax, should not convert. A household in a 9.30% state, same child, should not convert either. The higher the parents' state rate, the worse converting gets.

That is exactly backwards from how state tax is usually discussed. The pitch is always some version of "you live in a high-tax state, so get the money into a Roth before it costs you more." But the conversion is the event that triggers the state tax. Leaving the money in the traditional account defers that tax to a person who may live somewhere that never charges it.

A high state rate is a reason to convert less, not more, whenever the next owner is likely to pay a lower rate than you do. The only households for whom a high state rate argues for converting are the ones whose heirs will pay an even higher one.

What the four of them would say

The portfolio manager. Note what today did not change. Not one position moved, not one yield, not one balance. A single state rate entered the model and the answer swung by $17,467 without the recommendation changing at all. That is the signature of a robust decision: the inputs move the size of the prize, not the direction. Most of what this letter has corrected in three weeks has been the opposite, where one input moved the direction. It is worth knowing which kind you are holding.

The CPA. The cancellation is the thing to carry out of here. Any tax that falls on both sides of a comparison at the same rate drops out of the comparison, and only the difference matters. That is true of state rates here, and it is why "my state has high taxes" is almost never an argument on its own. The question is never the level. It is always the gap between the rate you pay and the rate the other party pays.

The financial planner. There is a planning move hiding in the 5.58 points, and it is not a tax move. If the parents expect to move, the order matters. A conversion done before relocating to a state with no income tax costs the full North Carolina rate. The same conversion done after costs nothing to any state. Nobody schedules a Roth conversion around a moving van, and on $165,000 at 3.99% that scheduling is worth $6,583.50.

The estate attorney. And the child's state is not a fact you get to know. It is a guess about where a 48-year-old will be living in 2037, which is the same class of guess as their income, and the letter is now resting two separate assumptions on the same unknown person. That is an argument for the thing I said yesterday, which is to split the beneficiary designations rather than bet the whole account on one forecast.

One more thing

Two moving parts that the table above holds still.

The first sits on the child's side. Everything here assumes the child's state taxes an inherited distribution as ordinary income. Several states exempt some or all retirement income, and a state that exempts it sets the child's rate to zero no matter what its headline rate says. A child in a nominally high-tax state with a generous retirement exclusion looks, in this arithmetic, exactly like a child in a state with no income tax at all. That is a question for the child's state and it moves the gap by the full width of the exemption.

The second sits on the parents' side, and it runs in their favour. The North Carolina page does not stop at 3.99%. It says additional rate changes may apply from 2027 under revenue triggers, pointing to the same session law. A lower North Carolina rate narrows the gap, which makes converting better, not worse.

So the one input here that is genuinely scheduled to move is moving the right way. That has not been true of much this month.

Tomorrow

Today put a state rate on the conversion and on the inheritance. It did not put one on the ladder.

The $1,587,583.15 municipal ladder at the centre of this month has been treated as tax-exempt in every letter, and federally it is. What North Carolina does with municipal interest from bonds issued somewhere else is a separate question, with a separate answer, and nobody has asked it here.

Tomorrow we ask it, at the source, and find out whether the ladder that solved the Medicare problem has been quietly carrying a state tax bill the whole time.