Ask most people about Roth conversions and you'll get a yes/no answer: "Should I convert my Traditional IRA to Roth?" That framing is already the wrong question. The real decision isn't binary. It's a bracket-fill problem, solved year by year, against a tax code with several traps most people don't see coming.
The Real Goal: Fill the Bracket, Don't Overflow It
A well-run Roth conversion strategy isn't "convert everything now" or "convert nothing, ever." It's converting just enough each year to use up room in a favorable tax bracket without spilling into the next one, and without triggering secondary costs that don't show up on the tax bracket chart at all.
That last part is where most DIY conversion plans fall apart, because the visible marginal tax bracket is not the only cost of an extra dollar of conversion income.
The Hidden Traps
IRMAA (Medicare premium surcharges). Medicare Part B and D premiums jump in tiers based on MAGI from two years prior. Convert too much in one year, and your Medicare premiums increase two years later, a delayed cost that's easy to miss when evaluating a conversion in the moment.
NIIT (Net Investment Income Tax). Once MAGI crosses a threshold, a 3.8% surtax applies to investment income. A large conversion can push you over that line, adding a tax that has nothing to do with the conversion itself but is triggered by it.
The Social Security tax torpedo. If you're already claiming Social Security, additional income from a conversion can push more of your benefit into taxable territory, up to 85% inclusion, meaning the effective marginal rate on a conversion can be significantly higher than the bracket you think you're in.
ACA MAGI (during the pre-65 bridge). As covered in a previous post, a conversion that seems small can be the exact dollar that trips a subsidy cliff.
Any one of these can turn a "smart" conversion into a net loss. Stack two or three together, and the optimal conversion amount can look nothing like what a simple bracket chart would suggest.
Why This Requires Real Modeling, Not Rules of Thumb
"Convert up to the top of the 22% bracket" is a rule of thumb. It ignores IRMAA look-back timing, NIIT thresholds, Social Security inclusion math, and ACA MAGI ceilings, all of which interact with each other and change year to year as thresholds shift and your income sources change.
A proper conversion strategy models this as a greedy optimization problem: for each year, find the conversion amount that fills available bracket room while checking every one of these constraints simultaneously, then re-solve for the next year given the new account balances. That's a fundamentally different exercise than checking a tax table once and picking a number.
The Takeaway
The question was never "should I convert." It's "how much, in which years, sequenced against IRMAA lookbacks, NIIT thresholds, Social Security torpedo math, and, if you're retiring early, the ACA subsidy cliff." Get the sequencing right, and conversions meaningfully reduce lifetime tax and RMD exposure. Get it wrong, and the conversion you thought was smart quietly costs more than it saved.