Top to bottom
Explore the page
- 01
Set the context
Step 1 asks which federal bracket is worth filling with a Roth conversion. Each row is the whole plan run again with that bracket as its maximum, under the same guardrails. Six bracket targets are tried: 10%, 12%, 22%, 24%, 32% and 35%.
Every answer is measured against the plan converting nothing and shielding nothing, even when the plan already holds a conversion; a banner says so, and applying a row replaces what the plan holds rather than adding to it. A Today’s money switch reads every figure at today’s purchasing power without recomputing anything.
- 02
Step 2: Capital gains harvesting
Gain harvesting models selling and immediately repurchasing appreciated taxable investments so market value stays invested while cost basis rises. It compares 0%, 15%, and 20% federal gains targets under the same selected brokerage accounts, date window, annual gain cap, and net investment income tax guardrail.
Review the cost basis raised, taxes paid, lifetime tax difference, and ending net worth, then inspect the annual income chart to see the remaining gains room and what stopped each year. The applied Roth schedule normally runs first; Step 2 also lets you ignore that schedule for its comparison while keeping it saved. Retirement, cash, HSA, and education accounts are not eligible.
The model uses annual proportional cost basis, does not select individual tax lots or execute trades, and a 0% federal target can still incur other taxes.
- 03
When a comparison is unavailable
A 401(k), 403(b) or traditional IRA gives the page something to move; without one it says nothing here converts, and notes that an HSA (health savings account) or an inherited account does not qualify. With pre-tax money present, two other empty states are told apart.
“These limits leave no room” means pre-tax money exists but the guardrails are already reached before a conversion starts, in every year of every bracket, and the page quotes the amount that is blocked. “Nothing left to convert” means the pre-tax accounts are already empty by the years a conversion would happen in.
- 04
Set strategy limits
Seven cards hold in every year of every bracket, so the columns stay comparable.
Five are ceilings: a capital-gains rung (0% or 15%); the 3.8% net investment income tax (NIIT) threshold; a Medicare income-related monthly adjustment amount (IRMAA) cliff from 1 to 5, where cliff 1 is the line below which no surcharge is paid; the Affordable Care Act (ACA) line at four times the poverty level; and an annual cap on the conversion itself, starting at $50,000 in today’s money and climbing with the plan’s inflation.
Two are rules: lower-rate first fills every feasible lower bracket across the projection before using the next, and withdrawal shielding pays for the year from Roth and cash before a pre-tax withdrawal would cross a line. A card the plan gives nothing to hold is shown disabled with its reason.
- 05
Read the recommendation
Above the table, one paragraph names the recommended bracket and how the conversion helps: it pays ordinary-income tax earlier, moves money into Roth, and shrinks the pre-tax balance behind future required minimum distributions; where the conversion’s average tax rate is below the modeled rate on later pre-tax withdrawals, the two rates are set side by side.
A bracket is recommended only if its full-plan rerun passes a safety check, meaning it does not bring the first shortfall forward, add unmet spending, or add a forced property sale, and it improves ending after-tax wealth, or relieves unmet spending when the plan already runs short. When no bracket qualifies, the page says so rather than picking the least bad.
Selecting an excluded column explains why it is excluded.
- 06
Compare conversion brackets
The table starts with a no-conversion baseline: ending after-tax wealth, the future pre-tax withdrawal rate, and the year required withdrawals begin. Each column is a conversion bracket, with markers for the best result and strategies that weaken the plan.
The conversion rows show how much moves, when conversions happen, spending funded from tax-free money during shielding, and the effective conversion tax rate compared with the later withdrawal rate.
What it changes compares each strategy with the baseline: ending wealth, tax, beneficiary net, required withdrawals, and Medicare surcharges. Ending after tax shows the difference and the change per dollar converted. Net worth after tax at the end shows the absolute ending value. If the baseline runs short, Ending after tax instead reports unmet spending relieved by the strategy.
Ending balances shows the tax-deferred and Roth balances, including their changes. The beneficiary-net comparison uses the Estate page’s default tax and cost assumptions. Select any measure’s info button for its definition and a related guide.
- 07
Inspect the selected strategy
Selecting a column draws three charts. “Conversion schedule” stacks each conversion year: income already using the bracket, required withdrawals, withdrawals for spending, and the conversion on top, under a dashed step for the bracket limit; it totals the money converted and the added tax, and deliberately leaves later effects out.
“Impact by year” plots one chosen measure as the conversion plan minus the no-conversion plan, tax liability by default, or net estate, net worth, effective tax rate, RMDs, IRMAA, ACA subsidies, the tax-deferred balance or the Roth balance, with conversion years shaded. Net estate uses the Estate page’s calculation under its default assumptions in every year.
“Lifetime impact” shows when the full plan pulls ahead and whether later tax reductions offset the early added tax.
- 08
Review and apply a schedule
Below the charts is the year-by-year plan the simulation actually ran: for each year, the amount, the account it came from, ordinary income before and after, the tax and its rate, and which limit stopped the year, or that the account ran dry first.
Each year opens into the source-account amounts, the ordinary-income stack, the conversion tax split and its effective-rate arithmetic, and the room left under every modeled limit. The first 8 years show; a button unfolds the rest. “Apply to plan” saves the schedule as the plan’s Roth conversion setting, together with the shielding it was priced beside, replacing whatever conversion the plan holds.
Once saved, every projection drawn from the plan runs it. The button is disabled while a sweep is running or its result is out of date, and reads “Applied” when the plan already holds the schedule on screen.
What changes what
Controls & settings
Check what each control changes and whether it recalculates your forecast.
Discounts every figure on the page to today’s purchasing power, or shows the dollars of each year. A way of reading the answer, not of computing it.
Selects which bracket’s schedule, charts and Apply button are shown. The page moves to the recommended bracket whenever a new comparison lands.
Holds a year’s conversion where it would push realized gains and qualified dividends off that rung; picking a rung switches the limit on. A year with no gains is not held.
Stop a conversion short of the 3.8% investment surtax threshold, and of the 400% of poverty line where the marketplace credit ends. The ACA card is disabled when the plan buys no marketplace cover.
Holds income under the chosen Medicare surcharge cliff, read two years ahead of the premium it sets. Disabled when nobody reaches Medicare inside the horizon.
Caps the conversion itself at a set amount a year in today’s money, $50,000 when first switched on, indexed with the plan’s inflation.
Keeps the row’s bracket as the maximum but reserves every feasible lower federal rung across the whole projection before the next rung is used.
Every candidate pays for its years out of Roth and cash before a pre-tax withdrawal would cross the bracket or a limit above; where tax-free money runs out the year spends over the line rather than going short. The baseline is never shielded.
Reading the result
Read the results
A “best” mark is the winner on that one measure alone. A bracket that empties the tax-deferred accounts fastest is rarely the one that ends furthest ahead, so read the ending row and the weakens flag before the rows it wins.
Green and red follow the direction a measure improves in, not its sign: less tax, fewer required withdrawals, a smaller tax-deferred balance and a lower effective rate are all shown as good news.
“Rate now vs later” sets the conversion’s effective tax rate against the rate the plan already pays when pre-tax dollars come out on their own; “worse” means paying more today to avoid less tomorrow.
Every amount, tax and income figure in the table and the schedule is what the settled plan did, not what the solver proposed. A year converting less than its bracket allows is explained in its “Stopped by” cell.
The figures on the limit cards are this year’s. The brackets and IRMAA cliffs are indexed and stand higher in later years; the NIIT threshold is the one that never moves.
One plan, several readings
Related pages
Projection →
The projection’s “Roth Conversion” button opens a compact dialog that runs the same sweep and the same guardrails, turning the comparison into six buttons; its Save writes the same schedule this page’s Apply does.
Taxes →
A saved conversion shows up on the taxes page as a “Roth conversions” income line in each conversion year, and that page’s next-dollar chart stands the Medicare and marketplace cliffs where the projection charges them, which is what the guardrails here hold under.
Estate →
Because the saved schedule runs inside every projection drawn from the plan, the ending estate, the milestones and the resilience test all reflect it.
The plan’s own sections
The guardrails are an account preference shared across plans, saved as they are switched; the applied schedule is the plan’s Roth conversion setting, and the IRMAA and ACA cards only wake for a plan whose Health section gives them something to hold.
Keep following the math
Key concepts
Keep in mind
Model limits
The engine uses simplified annual federal and state schedules. It does model each conversion’s own five-year clock, and the separate five taxable years a Roth’s growth waits on where the account’s opening year is given; it does not model every tax credit or future changes in law, and the page is an educational projection, not tax advice.
The spending order is not a variable here. The page answers how much to convert and when; it never rearranges the order the household spends its accounts in.
A limit is held where the solver can see it, but a year can cross one anyway, because the tax on a conversion has to be funded and funding it from a pre-tax account is itself income; the schedule says which limit stopped each year.
The comparison is one deterministic projection per bracket, not the plan resilience test; a bracket that passes the safety check here has not been run against shuffled market history.
Withdrawal shielding never leaves a year short: once the Roth and cash run out, the year spends over the line rather than going without.
This guide documents the page as built. The figures on it follow the plan’s own assumptions; none of it is individualized tax, legal, Medicare, or investment advice.