FIREproof has two tools for deciding how much to convert. Auto mode sizes each year's conversion on its own, as the plan runs. The Tax Strategy Explorer searches whole-plan strategies, compares them with your current plan across every start year, and lets you save the one you want.

Auto mode

Auto mode is fill to bracket on a Roth Conversion adjustment. Instead of converting a set amount, the simulation sizes each year's conversion as it runs, to fill a federal bracket or stay under an IRMAA tier. Pick the goal mode on the adjustment, run the plan, and each year's conversion appears in the year detail with the reasoning behind its size. Roth conversions are a Pro feature.

Goal modes

Fill to bracket

The classic mode. The optimizer finds the next federal income tax bracket above your current taxable income and converts just enough to reach the top of it. Anything more would spill into the next bracket and pay a higher marginal rate; anything less leaves cheap bracket headroom on the table.

Fill to IRMAA

For households on Medicare, the cliff that matters most is often the next IRMAA tier. The Medicare premium surcharge that kicks in when modified AGI crosses certain thresholds. fill_to_irmaa targets the IRMAA threshold instead of a federal bracket, so each year's conversion respects the surcharge cliff.

The NIIT threshold cap applies here too

With the NIIT cap turned on in your Tax Strategy settings, an auto-sized conversion is also held so the year's MAGI stays at or below the Net Investment Income Tax threshold. It runs after the bracket and IRMAA ceilings, so whichever leaves the least room is the one that sizes the conversion, and the year detail's "Why this conversion?" panel names which one bound. A conversion you entered as a fixed amount is never trimmed by it - that number is what you asked for.

If your income is already past the threshold before converting, the cap leaves no room and the conversion does not fire that year.

The Tax Strategy Explorer

Auto mode answers "how much fits under this ceiling this year?". It cannot tell you whether filling that ceiling every year leaves you better off. The explorer answers that, for Roth conversions and, before 59½, 72(t) SEPPs. Open it with Explore tax strategies at the top of the Tax Strategy section of Inputs, on any Roth Conversion, or on the Taxes tab of Proof. It is a Pro feature.

The explorer is three steps: choose the strategies, set the hard limits, read the results. The ribbon at the top shows where you are and a one-line summary of each step, and you can go back to any step without losing what you entered.

1. Choose the strategies

Three doors. Find one for my goal is the search: FIREproof builds the candidate strategies from your plan's own dates (now, retirement, the end of ACA coverage, the year before required distributions begin), runs each one and ranks them against the one goal you pick. The explorer never blends goals into a hidden score.

• Make the plan last: succeed in more start years.
• Leave more after tax: the largest ending portfolio after the tax your heirs would owe on pre-tax money.
• Largest ending portfolio: the largest ending balance, before any heir tax.
• Pay less lifetime tax: the lowest total tax across the plan.

Three more goals sit behind "More goals": lower Medicare surcharges (IRMAA), smaller RMDs, and a lower effective tax rate. The goals that minimize a cost only consider strategies that succeed at least as often as your current plan, because a plan can pay less tax by running out of money.

Pick a strategy lays the common strategies for your plan out as a grid: each row is a target (the top of a tax bracket, or an IRMAA ceiling with its income limit), and each column a timeframe from your plan's dates, headed by its years and the two milestones it runs between, such as Retirement to RMDs start. The milestone colors match the timeline on the Hard limits step. A box is one strategy; a dot means that combination isn't offered. Tick the ones you want to see and every one is run in full on every market path, alongside your current plan and no conversions. The goal you pick here only decides the order of the results. Build my own lets you set the windows and brackets yourself and compare up to three strategies of your own; nothing else is generated.

In Build my own, each strategy is a section with its name and Add a window at the top, then its windows, one row each: its years, the milestones they fall on, and what it converts up to. Edit opens it as a form right under it, and nothing changes until you press Save window, which checks that the window starts before it ends and doesn't overlap another; Cancel leaves it as it was. One window is open at a time. In the form, the first and last year show everyone's age, and the lists open with your plan's milestones (retirement, the end of ACA coverage, the start of RMDs). Picking a milestone fills in its year; the window keeps that year if the milestone later moves. To convert up to a tax bracket, pick it from the row of brackets, each showing its top in taxable income. For a window that starts in or after retirement, the bracket your yearly spending lands in is marked suggested. To convert up to an IRMAA tier, each tier shows its income ceiling and the yearly Medicare surcharge for landing in it.

2. Set the hard limits

Limits are hard: a strategy that breaks one is never recommended, and nothing is relaxed to make a search fit. You can choose which accounts convert and where the money goes (the destination must belong to the same person), keep an existing conversion as it is, name the accounts the tax must fit in and a cash reserve to leave alone, and cap the extra tax in any year. Employer plans such as a 401(k) are not converted unless you include them, because FIREproof cannot know whether your plan allows it. A success floor that even your current plan does not reach ends the search as having no answer, with the comparison still shown so you can see how close each strategy came. Each limit under More limits is off until you turn on its switch, and says in one line what it rules out.

The heir tax rate is an assumption, not a limit: the rate your heirs would pay on inherited pre-tax money. It is what turns a gross ending balance into a tax-adjusted one. A dollar in a Traditional IRA is worth less to an heir than a dollar in a Roth, so the comparison always shows both figures.

3. Read the results

Starting a search takes you straight to this step, and its progress shows here until the results arrive. Cancel it before anything has been compared and you return to the limits.

A strip at the top restates what you asked for: the goal (or, for the other two doors, what the rows are ranked by), who converts into what, which accounts pay the tax and the reserve, and the heir tax rate, with Edit limits and Search again beside it. Under it, one line gives the result in two numbers: what the best strategy does for your goal and what it does to your ending wealth after heir tax, both against your current plan.

The table has one row per strategy. The first two rows are the baselines, your current plan and no elective conversions; "No conversions" can be the best found, and that is a valid answer. Each row names the strategy with its window underneath ("starting now, until RMDs begin"), and a strategy that breaks one of your limits says which, inline.

The four columns each stack related results: Lasting (how many periods the plan lasts, and wealth after heir tax on poor paths), Ending wealth (after and before heir tax), Tax (lifetime, the first 5 years, and the effective rate) and Money moved (converted, required distributions, and any Medicare surcharges, ACA premiums or early-withdrawal penalties). Each column header names the statistic (median or average) and that it is in today's dollars. Your goal's column is marked, and your goal's figure sits at the top of it. The top figure in each column shows its change against your current plan, green when that is the better direction; hover any other figure to see its change. The current plan row has no changes, and "same" means the two are equal. Figures that are zero for every strategy are left out.

4. The three tabs

Results opens on the comparison table. Click a strategy to read it: the page switches to a list of every strategy on the left and, on the right, the one you picked, with four headline figures against your current plan, Save, and three tabs. Step to the next strategy from the list, the arrows above it, or the up and down arrow keys; the tab you are on stays put, so you can compare the same view strategy by strategy. Back to compare all returns to the table with that strategy highlighted.

How it holds up. Averages hide losers, so this tab draws every start year as a bar: how much better or worse the strategy ended than your current plan, sorted from the worst start year to the best, with the counts and the worst, middle, average and best change beneath. When your plan already has Roth conversions, a switch reads the same chart against making none at all, which asks whether converting is worth it, not just whether this strategy beats your plan. Below that are any failures it avoided or introduced, and when the up-front tax pays off: the year your wealth after heir tax pulls ahead of your current plan and stays ahead, as the soonest, typical and latest year across start years (with everyone's ages), plus how many never do. An amber dot on the tab marks a strategy that introduces failures or, with Monte Carlo returns, did not hold up on the second batch of paths (see below).

Why this strategy. What the strategy changes against your current plan, each a figure in the table; the best strategy at other heir tax rates, from the same runs; and what would change the answer: what was tested (the heir tax rate, lower returns, higher inflation) and what was not.

To see a strategy year by year, save it as a new scenario and open it in Proof. To change a strategy's years or bracket, use Build my own.

Funding timing. A conversion's tax is paid the following year, from your normal withdrawal order. The funding accounts and reserve you name limit how much is converted, so the tax fits in them; they do not change which account the payment is drawn from. The year detail lists the withdrawals of the payment year.

With Monte Carlo returns, strategies are chosen on one batch of simulated paths and then scored on a second batch the search never saw. A strategy that led on the first batch and not on the second is labeled as not holding up, is never recommended, and cannot be saved.

5. Save

The Save button sits beside the strategy's name when you read it, and names the selected strategy. Save as a new scenario copies your plan and swaps its conversions for the strategy, leaving the original untouched. Replacing the conversions in your current plan shows exactly what will be removed and added first. Nothing changes in your plan until you save. The button is disabled, with the reason, for a strategy that breaks one of your limits and for a result the plan has changed underneath; run the search again. When your plan has changed, the warning above the results has a Re-run this search button that runs the same search again with its original goal, picks and limits.

If you change anything in Steps 1 or 2 after a search (the way you're choosing strategies, the goal, your picks, your own strategies, a limit), Results stops showing the old answer and lists what changed instead. Run the search with your changes, or choose Go back to the settings these results used to see the old results again straight away, with nothing re-run.

72(t) strategies before 59½

If someone on the plan is under 60 and has a Traditional IRA or similar pre-tax account, the Pick a strategy door adds a group, Penalty-free income before 59½. Each row is a 72(t) SEPP: a fixed yearly payment from a pre-tax account with no 10% early-withdrawal penalty. The catch is the lock: until five years pass or the owner reaches 59½, whichever is later, nothing else may come out of that account and nothing may be converted from it.

The choices are a small table, grouped under each eligible account with its balance. Each row is a share: 25%, 50%, 75% or all of the account. Only that share is locked into the 72(t); the rest stays free to spend or convert. Each share comes in the three IRS methods (amortized and annuitized pay a fixed amount at a 5% interest rate, the RMD method is recalculated each year), and the row quotes the first year's payment. The IRS allows up to 5% or 120% of the federal mid-term rate, whichever is greater, so 5% is always allowed, and a higher rate means a larger payment from the same share. The amortized rows show first; Show all adds the other two methods. An early-withdrawal penalties figure appears under Money moved in the results whenever a strategy pays one, and the No early-withdrawal penalties limit, under More limits, rules out any strategy that pays one on any path.

A 72(t) row in Pick a strategy keeps your conversions exactly as they are in your plan, so it answers "given my conversions, which 72(t) is best?". To weigh the two together, use Find one for my goal: on a plan like this the search also tries each conversion strategy with an amortized 72(t) on half and on all of the largest eligible account, and names both in the result ("Fill the 12% bracket, with a 72(t) on 50% of Riley's Traditional IRA"). With all of an account in the 72(t), conversions from it wait until the 72(t) ends. If your plan already has a 72(t), a 72(t) strategy replaces it; to keep it instead, turn on its Keep switch under More limits, and no 72(t) is tried on its account.

Saving a 72(t) row creates a new scenario with the account split in two: the share moves to a new account named "(72(t))" with the 72(t) entry on it, and the rest stays in the original. Both keep the original's owner, investments and settings. Replacing in your current plan is not offered for a 72(t), because it changes your accounts.

What the explorer supports

Supported: US plans for a single person or a married couple, with historical or Monte Carlo returns, on any spending plan except Risk-Based Guardrails. The explorer tells you when a plan is not supported and why. The cases are:

• The plan's tax country is not the United States.
• The plan has two adults who are not married.
• The spending plan is Risk-Based Guardrails.
• The plan has no spending plan.
• No account can be converted (no Traditional IRA or other eligible pre-tax account, and no employer plan you have included).
• There is no Roth account owned by the same person as the account to convert.

Not modeled: state tax in the funding estimate, estate tax, and your heirs' actual tax situation beyond the single rate you enter.