The October decisions that set the withdrawal rate

Suppose a client walks into the office on October 14 with three documents on the desk and not one of them about markets: the receipt for a 2025 IRA contribution she made in January, in what she and her advisor expected to be a low-income bridge year until a deferred payout landed in the fall and turned it into a contribution the year will not allow; a premium notice for next year's health coverage quoting a monthly figure that rests on a federal subsidy; and a fraud file covering a loss she paid out of her retirement account in a year when the deduction for that kind of loss did not exist.

An hour later the two of them have not discussed asset allocation once, and that is roughly the shape of the work now: the last mile of retirement is becoming a sequence of tax and regulatory checkpoints, each dated and each capable of moving a client's sustainable spending number further than a quarter point of advisory fee. The business spent three decades learning to stress-test portfolios; it has not learned to stress-test the calendar.

The contribution goes first for the simple reason that it is the only item on the desk with a date certain attached: October 15 is the last penalty-free exit for a 2025 IRA contribution the year's income will not support, so unwinding it by then ends the matter. Leave it, and the 6% penalty repeats annually as long as the excess remains in the account — not a one-time toll but a recurring charge on money that was never eligible to be there.

Put the excess at $10,000, an illustration rather than this client's number, and the charge runs $600 a year, $6,000 across a decade, and onward for as long as the money sits — a guaranteed negative return that nothing else in a retirement plan carries. For a client who does not expect the contribution to become legal in some later year, and the coverage describing the deadline lays out no route by which it does, unwinding before the fifteenth dominates every other path, and it is not close.

The meeting happens on the fourteenth instead of in April for another reason: the same income figure answers both of the first two questions on the desk. Whether the contribution was permitted and what she pays for coverage next year are both settled by the year's final income number, which in October is still being assembled from the W-2, the K-1, and the payout that arrived after the projection was made — wait for the completed return and the conversation happens the following spring, one missed deadline late.

The deadline also happens to be the cheapest client-contact event on this month's calendar: every client who made a 2025 contribution has a fact pattern worth five minutes of review, and the review has a binary answer — the contribution stands, or it comes out before the fifteenth. Practices that run the list early in October will find the one or two files that need the conversation; practices that wait will meet those same files a year from now, with the first 6% already owed.

The Tax Foundation's fiscal arithmetic argues for stress-testing the federal healthcare subsidy and the employer-coverage exclusion the way advisors already stress-test equity returns, and this publication has taken that argument a step further: healthcare's fiscal trajectory belongs in every decumulation model. For the client at the desk, the base case is that the credit holds and next year's premium stays where the notice says, while the stress case — the one almost nobody builds — is the credit shrinking or the employer-coverage exclusion biting somewhere in the household and the premium climbing by more than a bad year in the portfolio costs her.

The asymmetry matters because a premium is funded before travel, before gifting, before anything the client would recognize as a choice, so an increase lands on the part of the plan with the least room to absorb it. A practice that stress-tests equities while holding the premium at a fixed number is being rigorous about the smaller risk.

The practical form of the stress case is a second premium line in the plan, in dollars, so the client can see what coverage costs if the credit does not survive contact with the year, and that number has to come out of discretionary spending rather than out of the withdrawal rate because the withdrawal rate is the output of the plan while the premium is an input to it. Advisors who build only one premium line are quietly assuming the more favorable of two futures on the client's behalf.

The same discipline extends to the tax items on the desk: a waiver delivered on a known timetable and a penalty that begins on a known date can both be put into the plan in dollars, which is more than can be said for next year's return on the S&P 500. The tax code is currently publishing more reliable planning inputs than the capital markets are, and the practices that notice will price them.

The fraud file is the one most advisors would shelve under history and forget, but the House voted 408-17 to restore a deduction that fraud victims lost in 2018, and the measure pairs it with a penalty waiver and a one-year repayment window — the two pieces that turn a closed year into a planning year. A House vote is not enactment, and the coverage does not say where the Senate stands.

For a client who took a distribution from her retirement account to cover a fraud loss, the sequence under the measure would run roughly this way: the deduction comes back, the early-withdrawal penalty is waived, and the repayment window gives her twelve defined months to put the money back where it came from. What that calls for in the meantime is unglamorous and specific — keep the file open rather than archived, keep the police report and the loss paperwork in one place, and calendar the repayment window from the date the measure takes effect rather than from the date the client first heard about it.

The file is worth reopening for every client who withdrew money after a fraud loss, because a one-year window pays the client whose documentation is already assembled and leaves everyone else arguing about a calendar they have never seen before; the one to worry about is the client who was too embarrassed to bring the loss up when it happened.

The fourth item is not on the desk, and it should be: BlackRock's tailored 401(k) default carries guaranteed income and private assets alongside the usual lineup, and once a plan's default includes a guaranteed-income sleeve the rollover stops being where the decumulation conversation starts. A participant who moves the balance to an IRA is walking away from an income decision the plan has already made on her behalf, which shifts the advisor's task from picking a destination to producing a comparison.

For the client in the chair, that comparison is concrete: her former employer's plan against an IRA, run explicitly rather than assumed, and where a plan lacks a guaranteed-income default the preference the market has already expressed still applies. LIMRA's final second-quarter tally came in $2.7 billion below its preliminary annuity total, and our read of that gap is that buyers were not chasing yield; they were buying protected participation, the same instinct the new in-plan defaults are built around, which is why an annuity conversation belongs after the plan comparison rather than in place of it.

For most of the past two decades the rollover was the automatic answer to nearly every plan balance, and against a menu of funds it was rarely worth a second look; against a default that already carries guaranteed income, however, it is a decision about income rather than where the money sits, and the comparison has to exist before the forms get signed. Advisors who still process it as paperwork are the ones most likely to be asked, years later, what happened to the guarantee they gave up.

None of this is portfolio work, and all of it is the kind of work a client can verify: a penalty avoided is a dollar the client can point to on a form, a plainer proof of value than a basis point of outperformance that has to be taken on faith. Practices that treat the fall calendar as a service line rather than an administrative nuisance will find the file work compounds, each deadline handled in the open becoming the reason the client answers the phone in July.

Two of the three items on the October 14 desk will be settled by events no one in the room controls — what happens to the fraud measure and what her coverage costs once the year's income is final — but neither is a reason to wait on the parts that can be finished now. Build the premium stress case in dollars, run the plan comparison, and close the contribution before the fifteenth, and let the withdrawal rate be the last number the plan settles.