Locking in the 2017 rates turns Roth conversions into a hunt for one-off deductions

Set the marginal rate at conversion equal to the marginal rate at withdrawal and the two paths land on the same after-tax dollar, because it does not matter whether you multiply by growth first or by one-minus-the-rate first ^[1]. That identity is the whole reason permanence in the 2017 brackets matters here ^[1]: it deletes the term that used to justify converting the whole balance, without making conversions bad in themselves. What remains has to be found inside the client. Velazquez's own list is the honest inventory: not just the tax arbitrage, but stealth taxes and, in his framing, a longevity question about how long the Roth compounds after the bill is paid ^[5], with the first few years feeling worse than having done nothing ^[6].

Now price the offsets. Houchins points to a bigger standard deduction, a bigger senior enhanced deduction, and for 2026 a deduction of up to $1,000 per person for cash charitable gifts ^[8]. Two people, so $2,000 of converted income sheltered ^[2]; against a hypothetical $500,000 traditional balance that is four tenths of one percent of it ^[3]. The charitable line changes the tax on a slice; the decision itself stays the same. The item with actual scale is the one Velazquez leads with, a large carryover loss from a business sale ^[3], and the source does not say whether that loss is ordinary or capital, or how much of it can be applied against income in a single year, which is precisely the number that decides how much conversion it can absorb.

The counter-thesis is in the same article, and it is stronger than the permanence story. Mahaney says required distributions are starting to bite savers who have not been drawing down their IRAs, and that larger mandatory RMDs will force them into higher brackets ^[13]. That is a forecast about one household's effective rate rising inside frozen statutory brackets, and it never depended on Congress at all. Smith's window sits in the same place: for clients whose needs are covered by pensions and Social Security, the years before distributions begin are the conversion runway ^[12], and those distributions start somewhere between 70 and a half and 75 depending on birth year ^[11].

This is probably wrong in one direction, so name it: permanent describes current law rather than the future, and if a later Congress moves the top brackets, the advisor who converted only in loss years will have prepaid too little. The constituency for converting on volume regardless is anyone doing it for heirs, since Roth assets generally pass without income tax on distributions ^[14] and the relevant rate belongs to a stranger decades out.

The allocation consequence is the part that shows up on a payroll. A blanket rule is one decision applied across a book; a case-by-case rule needs the client's tax preparer in the loop on each one, which Velazquez says explicitly ^[7], plus a read on the heirs' brackets ^[10] and a calendar built around gap years between retirement and Social Security ^[9]. That is per-household, seasonal work. The rate forecast once justified converting the whole balance. The 2026 charitable deduction only covers $2,000 of it ^[2].