The Roth conversion screen should start with taxable assets

The Roth conversion conversation at the desk tends to open with a forecast — the client's bracket today against the bracket they expect in retirement. The second installment of Ed Slott's conversion series, published September 14 on IRAHelp, works the question from the other direction, and its list of five situations where a conversion may not be the ideal way forward is the document an advisor can put to work this week. Slott's own framing is that a conversion is not a universally beneficial transaction, so each case has to be evaluated on its own facts. Two households with identical balances and identical bracket projections can reach opposite answers, and the deciding facts may have nothing to do with tax rates.

The first disqualifier is cash flow: converting an IRA or a workplace plan generates a higher tax bill for the year of the conversion, and a household living on a fixed income is adding an expense it did not have. Slott's test is blunt: the money the client needs to live on cannot be the money that settles with the IRS. A conversion funded out of the household's spending money is not a planning strategy, and the desk should treat the question of where the tax gets paid from as the first gate rather than the last.

The second is the bracket itself: a client who expects to take distributions later at a materially lower rate is prepaying tax at the higher rate now; the converted balance then grows and comes out tax-free, but the lifetime comparison carries the rate paid at the front end. This is arithmetic an advisor can model, and it is usually the piece of the decision the client arrives already thinking about.

The fourth situation deserves to be a rule rather than a judgment call, because the failure is mechanical rather than behavioral: Slott's recommendation is to pay the tax due on a conversion from another source of funds instead of withholding it from the IRA. Withholding leaves less capital inside the Roth to accumulate tax-free, which is reason enough by itself. Pay the conversion tax from the IRA and the client has prepaid tax on a smaller balance.

Age turns that preference into a trap: for an IRA owner under 59½, the dollars withheld for taxes are a withdrawal, not part of the conversion — an early distribution, subject to a 10% penalty on the money sent to the IRS. The client pays the tax and a penalty on the tax. Slott's guidance is that an IRA owner under 59½ should almost never have taxes withheld from the IRA on a conversion, and the hedge is his; the practical reading is that the exception is narrow.

Withholding is the easy path: a conversion processed with taxes withheld requires no second account, no check, and no conversation about which pot of money is which. Those are precisely the conditions under which the penalty attaches to a payment the client assumed was a formality. Raise the funding source in the first conversion meeting and the problem never arrives.

The fifth entry reads less like a client profile than a screening question for the first meeting—anyone who asks how soon they can take the money out—and Slott flags the question itself. For a client over 59½, converted funds are accessible right away, so the ask can be entirely innocent. In a household that needs the money soon, though, converting means financing a tax payment with dollars already headed out the door, which is the first disqualifier arriving under a different name.

Line the five up and the rate forecast stops being the organizing principle. Three of the five — the fixed-income senior, the client with no non-retirement assets to pay the tax, and the client asking about access — are funding questions, and they are answered with a balance sheet rather than a projection. Only the second entry is a pure rate call, and it is the one that fits most neatly into a single projection slide.

Set the balance sheet test ahead of the projection, then, and sequence the work around it. As this publication has argued, tax planning is the ground where advisors can still distinguish themselves as portfolio construction commoditizes, and the Roth conversion is the highest-frequency version of that work: it recurs annually, it carries a deadline, and it produces a number the client can see. What the second Slott installment adds is the discipline of the no — a documented screen for who should not convert, applied before the modeling rather than after it.

The two entries bracket the year. The August 31 piece laid out where conversion can make sense; this one lays out where it doesn't, and the cheaper screen is the negative list, because it takes one statement to run. Three questions do most of the work: can the household pay the conversion tax from taxable assets, does it need the IRA for spending this year, and does it expect a lower bracket later. A no on the second ends the analysis whatever the answer to the third, and the income year the client converts is the year that matters.