Skip to content
Back to IRMAA

Education / IRMAA

Sizing Roth Conversions Around IRMAA Cliffs

A decision-worksheet for sizing Roth conversions when projected MAGI is near an IRMAA tier boundary. Covers the two-year lag, the cliff cost calculation, and a "convert to the boundary" rule of thumb.

Updated: Sat May 02 2026 00:00:00 GMT+0000 (Coordinated Universal Time)

If you're considering Roth conversions in the gap years between retirement and Required Minimum Distributions, IRMAA tier boundaries are one of the biggest hidden constraints on your conversion size. This worksheet walks through how to model the interaction.

The setup — why IRMAA matters for conversions

A Roth conversion adds the converted amount to your taxable income for the year. That increases your AGI, which increases your IRMAA MAGI two years later, which increases your Medicare premiums two years later.

For most households the federal-tax cost of a conversion is the dominant consideration: convert into the lowest available bracket, stop before crossing into a higher one. But for Medicare-eligible households (or those who will be Medicare-eligible within two years of the conversion), IRMAA tier sizing is a second binding constraint on top of the federal-bracket constraint.

The two often disagree:

  • The federal 24% bracket might extend up to MAGI of ~$220,000 (single).
  • The IRMAA Tier-1 threshold is at MAGI of ~$106,000 (single).

A conversion sized to fill the 24% bracket would push the household into IRMAA Tier 1 (or higher) and trigger a two-years-later premium surcharge. The two constraints don't move together; you need to model both.

Step 1 — Project MAGI without the conversion

Start with your projected MAGI for the conversion year, before the conversion. Add up:

  • Wage / self-employment income (if still working part-time)
  • Pension income
  • Taxable Social Security benefits (apply the SSA taxability formula)
  • Traditional IRA / 401(k) RMDs (if applicable)
  • Capital gains and dividends (planned)
  • Interest income (including tax-exempt muni interest — IRMAA adds this back)
  • Any other taxable items

Call this number MAGI_base. Write it on the worksheet:

MAGI_base = $___________

Step 2 — Identify the IRMAA tier boundaries above MAGI_base

Look up the current-year IRMAA brackets at Medicare.gov. Identify the tier boundary just above your MAGI_base. That's the cliff you need to size around. Also identify the boundary above that one (in case you're considering a multi-tier conversion).

Current IRMAA tier:         Tier ___
MAGI_base:                  $___________
Next IRMAA boundary at:     $___________
Boundary after that at:     $___________

Step 3 — Calculate the headroom

IRMAA_headroom = next_boundary - MAGI_base

This is the maximum conversion that stays in the current tier.

For a married couple with MAGI_base of $250,000 and Tier-1 boundary at $266,000, IRMAA_headroom = $16,000.

Step 4 — Calculate the cliff cost

If you exceed the boundary, you pay the next tier's surcharge for the entire year. Calculate the annual cliff cost:

cliff_cost_per_person = (next_tier_Part_B - current_tier_Part_B) × 12
                        + (next_tier_Part_D - current_tier_Part_D) × 12

cliff_cost_household = cliff_cost_per_person × people_on_Medicare

For a married couple where both are on Medicare, going from Tier 0 to Tier 1 (illustrative): cliff_cost = ($74 × 12 + $13 × 12) × 2 = $2,088.

This is the dollar penalty for exceeding the boundary — paid two years later, but locked in by the conversion year.

Step 5 — Compare cliff cost to conversion benefit

The conversion's federal-tax benefit depends on the rate-arbitrage — how much your future tax rate would be without the conversion vs. how much it costs to convert now.

A simplified model:

conversion_benefit = converted_amount
                     × (future_marginal_rate - current_marginal_rate)
                     × years_until_withdrawal_or_RMD

Compare:

  • Path A — convert up to the IRMAA boundary ($IRMAA_headroom). No cliff cost. Federal benefit captured up to that conversion.
  • Path B — convert into the next tier ($IRMAA_headroom + extra). Federal benefit captured on the larger conversion, minus the IRMAA cliff cost.

If extra × (rate_arbitrage) × years > cliff_cost, Path B wins. Otherwise Path A wins.

Worked example

Married couple, age 64, both retiring in the conversion year. Pre-conversion MAGI for the year: $250,000. Both will be on Medicare in 18 months. Married- filing-jointly federal brackets straddle 22% / 24% near $206,000.

  • MAGI_base: $250,000
  • IRMAA_headroom: $266,000 − $250,000 = $16,000
  • Tier-1 cliff cost: $2,088/year (Part B + D, both spouses)
  • Tier-2 boundary: $334,000 (so room to fill Tier 1: $84,000 inside it)
  • Tier-2 cliff cost (additional, on top of Tier 1): about $2,664/year

Three conversion paths:

PathConversionMAGIIRMAA tierCliff costApprox. federal tax @ 24%Net cost
A$16,000$266,000Tier 0$0$3,840$3,840
B$50,000$300,000Tier 1$2,088$12,000$14,088
C$84,000$334,000Tier 1$2,088$20,160$22,248
D$100,000$350,000Tier 2$2,088 + $2,664 = $4,752$24,000$28,752

The IRMAA cliff cost is paid one year (when 2028 Medicare premiums hit). The federal tax cost is paid this year. Both reduce the net long-run benefit of the conversion.

For this couple, the break-even rate-arbitrage to justify Path B vs Path A on the additional $34,000 of conversion is about:

2,088 / 34,000 = 6.1% rate arbitrage required

If their future marginal rate when the converted money would otherwise have been withdrawn is at least 6.1 percentage points higher than today's rate, Path B is positive-EV despite the cliff. If not, Path A (stop at the boundary) is the better move.

Rule of thumb — "convert to the boundary"

For most pre-retiree couples, the practical rule is:

Size your conversion to fill the headroom up to the next IRMAA boundary but not exceed it, unless you have a documented reason to believe future rates will be substantially higher than today's.

Reasons to deliberately exceed:

  • You're confident the post-RMD bracket will be 30%+ vs. today's 22-24%.
  • You have a long horizon (15+ years before the converted money would naturally be withdrawn) so the tax-free compounding offsets the cliff cost.
  • You're consolidating multi-year conversions into one big year (so future years stay below the boundary, avoiding compounded cliffs).

Reasons to stop firmly at the boundary:

  • You're already in your low-tax window — current rate is at the floor.
  • You're within 5 years of large RMDs starting; you'll have less rate arbitrage than you think.
  • You're modeling charitable-giving as the destination for some pretax assets (Qualified Charitable Distributions are a higher-leverage move than conversions for those dollars).

What if I'm not yet on Medicare?

If neither spouse is Medicare-eligible in the conversion year, IRMAA still kicks in two years later. So a 63-year-old converting in 2026 affects their 2028 Medicare premium starting at 65. The two-year lag means you're paying for the conversion year's IRMAA tier in the year you turn 65.

If only one spouse is on Medicare in the relevant year (e.g., one spouse is 65 and the other is 62), the cliff cost is on one spouse's premium, not both. That cuts the effective cliff cost in half.

What if I have a one-time MAGI spike that's not from conversion?

Common situations:

  • Sale of a business or appreciated stock
  • RMD from an inherited IRA
  • Concentrated severance / employment-settlement payment
  • Real-estate sale with significant depreciation recapture

In a year with a one-time MAGI spike, the IRMAA cliff is already triggered by the spike. If you're going to exceed the next boundary anyway, the marginal IRMAA cost of an additional Roth conversion is near zero — you've already paid the cliff. Conversions made in such a year are unusually well-priced.

This is a meaningful planning insight: a year with an unavoidable spike is also a great year for a discretionary Roth conversion into the same tier. The conversion fills the rest of the tier without adding cliff cost.

Primary sources and tools

This worksheet is published by NestPilot Foundation Inc. — a nonprofit (501(c)(3) filing in progress). The dollar amounts in the worked example are illustrative based on recent-year IRMAA structure. For a personalized analysis that integrates with your full retirement plan, the free Roth Conversion Decision Explorer and the authenticated planning workspace at app.nestpilot.org provide deeper math.