A plain-language guide to using the income modeling tool.
This calculator answers one main question: if you withdraw a certain amount from your retirement accounts (401K, IRA, etc.) each year in retirement, perhaps based on RMDs, spending needs, or both, does your money hold up for your estimated lifespan?
You tell it what you have, what your income requirements are, and what you expect your investments to earn (or use default values). It then walks through your retirement year by year, using the IRS rules that apply to you — required withdrawals (RMDs), federal taxes, Social Security — and models what happens: which assets are exhausted first and when, when your portfolio could be exhausted, and how much you'd owe in taxes along the way.
It's a resilience check — a way to see how a plan holds up under a set of assumptions you control, so you can adjust before a significant shortfall shows up, not after.
This is not financial or tax advice. It's a tool for exploring "what if" questions with your own numbers to help you model and plan. Talk to a financial advisor or tax professional before making decisions or taking action.
The page has two parts. On the first page, you enter your numbers. When you run the simulation, the results appear on a different page.
You don't have to fill in everything. Most sections are optional — they start collapsed and closed with a small switch beside the heading. Leave a section closed, and it simply has no effect on your results. Only four sections are always open and matter for every run: Starting Point, Starting IRA Balances, Spending & Income, and Return Assumptions. Every dollar field starts at $0, so nothing is pre-filled with someone else's numbers — you're building this from scratch with your own figures.
When you're ready, click the blue "Run the Simulation" button at the bottom of the form. Results appear on the results page — and update instantly if it’s already run — every time you change a number and click again.
If you want to start completely over, there's a "Reset All to Defaults" button just below it. Click it once, and it'll ask you to click again to confirm, since it clears everything you've entered; clicking anywhere else on the page instead cancels it without changing a thing.
One more thing before diving in: this manual walks through the full Tool's twenty sections, numbered to match that page exactly. There's also a Quick Retirement Check, a shorter page covering just the essentials, balances, spending, a return preset, and (if you're married) a second spouse with their own age and Social Security, running the same underlying calculation with a lot less to fill in. Its four sections use their own separate 1–4 numbering, so if you're using that page instead, don't expect "Section 3" here to line up with "Section 3" there. Most of what follows describes detail the Quick Check doesn't expose at all.
Many fields and sections also have a small ? button beside them — tap it for a quick explanation and, where it's relevant, a plain note on when you'd actually want to use that feature. This manual covers the same ground in more depth, for anyone who wants the fuller picture.
Three numbers: your current age, how far out you want the plan to run (95 by default — a reasonably long lifespan to plan for), and Retirement age, which defaults to match your current age as you type it, meaning no gap at all, until you change it.
If you're not retiring right away, still working and not concerned about portfolio income until then, set Retirement age to whenever that actually is. Every year between Current age and Retirement age skips ordinary living-expense spending entirely, a paycheck is assumed to be covering it, so the portfolio simply grows at its assumed returns with nothing withdrawn for spending. Once Retirement age arrives, everything behaves exactly as it always has.
A few things deliberately don't wait for Retirement age, and keep working on their own configured ages regardless. Lump-sum expenses and charitable giving still fire whenever you've set them to, independent of this field, since you may genuinely want a lump sum, a new roof, say, while still working. Social Security, pension, and QLAC income all still start on their own separately-configured ages too. And RMDs remain mandatory by their own age rule no matter what: if one falls due while you're still working, before Retirement age, the IRS-required amount is shown in the RMD column. As-needed Savings and Brokerage balances can reduce how much actually has to come from the IRA that year — the RMD column still shows the full IRS amount, but the cash may arrive from taxable accounts instead. Whatever's beyond what you actually need that year becomes extra after-tax income, the same way any other income surplus already does elsewhere in this tool.
Note — employer plans while still working: The rule above applies to traditional IRAs, which always require distributions once you reach your RMD start age regardless of employment. Participants in a current employer’s 401(k) or similar workplace plan may be able to defer distributions until they actually retire, unless they own 5% or more of the business. Because this tool combines IRA and employer-plan balances into a single pool, it cannot separately apply that deferral option. If a meaningful portion of your pre-tax balance is in a current employer plan and you are still working past your RMD start age, the tool may show a larger required distribution than you would actually owe in the short term. Once you separate from that employer, normal RMD rules apply to that account as well.
This decides how every other dollar in this tool gets spent down. Four approaches are offered, presented as four buttons: Capital Preservation (selected by default), Time Segmentation, Total-Return: Proportional, and Total-Return: Rebalancing. All four are established approaches real retirees and advisors actually use; none is inherently safer or better than the others, they trade off differently, and picking one is really a question of how you already manage, or want to manage, your own money, not a right-or-wrong choice this tool is nudging you toward.
Capital Preservation is an income-first strategy. Interest and dividends sweep to cash and are spent first. When income alone doesn't cover the gap, principal is drawn in bucket order — cash first, then short-term bond, then core bond, then TIPS, then any custom funds, then equity last. Bond principal stays flat as long as income covers the need. The goal is to live off the "fruit" (income) and preserve the "tree" (principal) for as long as possible. The trade-off is opportunity cost: income swept to cash doesn't compound; it's available to spend, not invested for growth.
Time Segmentation organizes the portfolio by time horizon: cash covers near-term needs, bonds cover the medium term, equity funds the long term. Nothing sweeps to cash; every holding compounds in place within its own bucket. Withdrawals still follow the same bucket order (cash first, then short-term bond, then core bond, then TIPS, then any custom funds, then equity), but the money in each bucket grows as a total return rather than trickling income to cash along the way. Think of it as "spending the nearest bucket while the rest grows."
Total-Return: Proportional keeps everything invested and compounding together rather than segmented; each year's need is split across every holding matching its current share of the total, whatever that share has drifted to since you started, up or down. Within the "everything else" side (the holdings other than equity), the largest one is sold first, then the next largest, so cash and smaller positions are generally the last things touched, not the first. This approach never corrects drift, it just proportionally shrinks whatever allocation already exists; it also has no way to avoid selling equity in a year equity is down, its current share still gets sold regardless of whether that share just fell.
Total-Return: Rebalancing also keeps everything invested, but instead uses the split implied by your starting balances as a target, and sells preferentially from whatever's grown above its own target share, funding the withdrawal by trimming winners rather than taking a flat percentage of everything. This is a commonly cited practice among financial advisors and DIY investors managing their own total-return portfolio. If nothing is far enough above target to cover the whole withdrawal that year, whatever's still needed is funded the rest of the defensive sleeve first, equity only as a last resort, the same protective order the bucket approaches use for their own shortfalls. In practice, this means Rebalancing avoids selling equity in a year it's underweight, typically right after it's fallen, a situation Proportional has no way to avoid.
Roth stays protected as the last resort in all four. This setting also works with Monte Carlo (Section 20) and Historical Simulation (Section 21) — whichever approach you select here is the one their randomized or historical returns are applied to.
This is where you tell the tool what you actually have. It splits your combined tax-deferred retirement balance into five general asset categories — equity (stocks), core bond fund (a composite of intermediate- and long-term bonds), short-term bond funds, TIPS bond funds (inflation-protected), and cash. This isn't limited to a traditional IRA: a 401(k), 403(b), or any similar employer plan subject to RMDs belongs here too, combined into the same five buckets. If you're married, these are your combined household totals across both spouses' accounts, not just one spouse's (you describe the split later on).
If you hold a CD or bond ladder rather than a fund, it still fits here, no need to break out individual rungs. Put its total value in whichever of Short-Term Bond or Core Bond matches its typical duration, and its blended average yield across all the rungs in the matching field in Return Assumptions. Under Capital Preservation (Section 2), that balance holds steady year to year and only its interest sweeps out, the same stable-principal, no-forced-selling behavior a real ladder is built to provide, so this ends up modeling a ladder reasonably well without tracking each rung separately. A TIPS ladder works the same way, though TIPS' real-yield-versus-accretion split (see the column glossary under Reading the Results) makes it a bit more involved.
Important note: Under IRS rules, you must calculate the RMD separately for each traditional, rollover, SEP, and SIMPLE IRA you own. However, you may add those IRA RMDs together and withdraw the combined amount from any one or any combination of those IRAs. And that's precisely what this tool does.
Multiple 403(b) tax-sheltered annuity accounts have a similar—but separate—rule: calculate the RMD for each 403(b) account, then satisfy the combined 403(b) amount from one or more of your 403(b) accounts. This tool does not do this; it aggregates them together with Traditional accounts.
In contrast, RMDs from 401(k)s and most other employer-sponsored defined-contribution plans generally cannot be aggregated with an IRA, a 403(b), or another 401(k). You must calculate each plan's RMD and withdraw that amount from that specific plan. Again, this tool calculates your RMD in aggregate, not on an account-by-account basis.
If you hold more than one type of Traditional account, the tool calculates your overall annual RMD to illustrate cash flow. But for IRS compliance, you'll need to separate that amount into the applicable withdrawal groups: the IRA group, the 403(b) group, and each individual 401(k) or other non-aggregable workplace plan. Consult a qualified tax professional or financial adviser for circumstances such as inherited accounts, annuity features, or employer-plan-specific rules.
RMD start age and the 1959 boundary: Under SECURE 2.0, the Required Minimum Distribution start age is 73 for anyone born in 1959 or earlier, and 75 for anyone born in 1960 or later. The tool infers your birth year from your current age and retirement start age, and assigns age 73 or 75 accordingly. If you were born in 1959, a technical ambiguity in the original SECURE 2.0 legislation briefly created uncertainty about whether you’d start at 73 or 75; The 2024 final RMD regulations (T.D. 9991) reserved this question; IRS proposed regulations issued alongside them proposed age 73 for 1959 births, but that proposed rule has not been finalized. RetireFlow uses age 73 pending a final resolution. The tool reflects that correctly. There is no override for this field — the RMD start age is fully determined by birth year.
Two more things are in this section:
Two already-taxed accounts, held apart from the Traditional and Roth balances above, and not subject to RMDs, since neither was ever tax-deferred retirement money.
A separate, already-taxed account, savings, money market, or CDs. Its interest is taxed as ordinary income every year it's earned, whether or not you actually withdraw any of it, the way real bank and CD interest generates a 1099-INT regardless of activity; withdrawing from the account itself isn't a second taxable event, since that money, principal or previously-taxed interest, has already been taxed once.
Turning this on reveals a starting balance, an interest rate (its own field, separate from Cash or Short-Term Bond, since real savings and CD rates can move differently from a bond fund's), an age the account becomes available, and a choice between two ways of drawing it down.
As-Needed puts this account first in line, ahead of every other holding, cash and bonds included, once its available age is reached, the standard tax-efficiency principle of spending already-taxed money before triggering a new taxable event elsewhere. It only comes out when there's an actual spending gap to cover that year — and only for the portion of that gap above any Required Minimum Distribution. RMDs must be distributed from the Traditional retirement account itself, based on its own prior December 31 balance; Savings & CDs cannot satisfy that obligation. In an RMD year, the tool ensures the Traditional account distributes at least the full RMD amount, and Savings & CDs can cover only additional spending need beyond that.
Scheduled instead draws a fixed amount, in today's dollars, growing with inflation, every year once available, whether or not the full amount is actually needed that year. If it's more than that year's gap, the excess doesn't disappear, it simply becomes extra spending money, the same way a pension or Social Security surplus already does elsewhere in this tool; if the remaining balance can't cover the full scheduled amount, whatever's short falls through to the rest of the plan the normal way.
Either way, "available starting at age" has no expiration once reached, the account stays eligible for the rest of the plan. Setting that age to your retirement start models a Social-Security bridge, spending this account down (Scheduled) or leaning on it first (As-Needed) through the years before Social Security begins, so the benefit itself can be delayed for a larger monthly amount. Setting it to whenever RMDs begin instead models a later-life reserve, available as a backstop once the mandatory-withdrawal years start, without ever touching it before then. Both use the exact same mechanism, just a different age.
This account sits outside the main portfolio pool entirely, and behaves identically no matter which Withdrawal Approach (Section 2) is selected; it's a dedicated, special-purpose holding, not part of the general investment mix.
A regular, non-retirement brokerage account, stocks or funds, taxed differently from anything else in this tool: only the gain portion of a withdrawal is taxable, not the whole amount, since the rest is already-taxed principal. Unlike Savings & CDs, its taxable portion is taxed at federal long-term capital gains rates, 0%, 15%, or 20% depending on your total income for the year (2026 thresholds, IRS Rev. Proc. 2025-32), not the ordinary brackets everything else in this tool uses; the 3.8% Net Investment Income Tax surtax that applies to some higher-income households is not modeled.
Turning this on reveals a starting balance, a cost basis, a price return and a dividend yield (separate fields, since price growth and dividends are taxed differently in timing, gains only when sold, dividends every year they're paid), an age the account becomes available, and the same As-Needed/Scheduled choice Savings & CDs uses.
Cost basis is what you originally paid for what you're still holding, the part that isn't taxed again when withdrawn. It defaults to match your starting balance as you type it, which assumes no gain has accumulated yet; type your own figure into the cost basis field directly if you know this account already has embedded, unrealized gains; a lower basis than balance means a higher taxable share of every future withdrawal. This tool tracks one blended basis for the whole account, the average-cost method, not individual purchase lots, so a withdrawal realizes gain and returns basis in the same proportion the account holds them in that year, not by which shares were bought when.
Dividends are taxed as earned every year, whether or not shares are sold, the standard treatment for qualified dividends, and reinvest into the account, becoming new, already-taxed basis going forward the same way any other after-tax contribution would.
This account sits second in line, right after Savings & CDs and before any Traditional holding, in As-Needed mode: already-taxed money, brokerage included, covers spending need before ordinary-income withdrawals from Traditional accounts — but only spending need above any Required Minimum Distribution. The RMD itself must still be distributed from the Traditional account (IRS rules base it on that account's own prior December 31 balance), so in an RMD year the Brokerage can reduce spending above that floor, not the mandatory distribution itself. Scheduled mode works exactly as it does for Savings & CDs, a fixed amount, growing with inflation, drawn every year once available, any surplus beyond that year's need simply becomes extra spending money. Like Savings & CDs, "available starting at age" has no expiration once reached, and this account sits outside the total-return pool entirely, behaving identically no matter which Withdrawal Approach (Section 2) is selected.
How much gross income you need each year, before federal taxes, in today’s dollars (the tool adjusts for inflation automatically), along with what you expect inflation to run at and what Social Security pays. Enter a pre-tax figure: the tool calculates taxes on top of whatever you enter here, not out of it. If your target is a specific after-tax spending level, you’ll want to gross it up — for a rough estimate, divide your desired after-tax amount by (1 minus your expected effective tax rate). Keep this to ongoing living expenses only. A QCD, cash charitable giving, and lump-sum expenses are entered separately later in the form, and each one adds to this figure rather than coming out of it. (If you're planning a $20,000 QCD, for instance, don't lower your spending need by $20,000 to allow for it; enter your actual spending need here, and the QCD amount in that section, and the tool adds them together correctly.) If you'll retire before Social Security actually starts, which is common if you're claiming later to grow the benefit, enter the age it actually begins in "Social Security starts at age." Enter the benefit in today’s dollars — the tool automatically inflates it forward to your claiming age at the same inflation rate you’ve set for the rest of the plan, so you should enter what Social Security’s estimator shows as your benefit in current purchasing power, not a future nominal projection. (Social Security’s own estimator, which gives results in today’s dollars, is the right source for this number.) In the simulation, every year before that age shows $0 from Social Security, so the portfolio or other income must cover the full spending gap, if there is one, until it kicks in. One thing to clarify: this tool doesn't calculate Social Security benefits itself; it has no knowledge of your earnings record, full retirement age, early or delayed claiming reduction and increase percentages, spousal or survivor benefit rules, etc. It simply takes whatever dollar amount and start age you enter and uses them that year, with COLA-style inflation growth from there. Before using this tool, determine the actual benefit for the claiming strategy you're considering using Social Security's own estimator or a financial professional, then enter the result here.
Testing the 4% Rule. If you have heard that withdrawing 4% of your portfolio in the first year of retirement is a reasonable starting point — the so-called "4% rule," drawn from research on sustainable withdrawal rates — you can test that assumption directly here. Take your total portfolio balance, multiply it by 4%, and enter that as your annual spending target. The rule assumes you then increase that dollar amount each year by inflation, which is exactly what this tool does when you set your spending in today's dollars and let it adjust automatically. What the rule cannot tell you on its own is how your specific plan holds up — your income streams, your tax situation, your asset mix, your timeline. Running Monte Carlo or Historical Stress Test with that spending figure is how you find out whether 4% actually works for your situation, or whether a slightly higher or lower rate fits better. Once you run the tool, the results page shows your actual year-one withdrawal rate in the summary numbers — useful confirmation that your spending entry translates to the rate you intended, and a quick benchmark against the 4% starting point.
If you're married, turning this on lets you enter a second age and life expectancy, a second Social Security benefit (or spousal benefit) with its own start age, independent of Spouse 1's, and how the Traditional balance is split between the two of you. This matters because Required Minimum Distributions are based on each person's own age, not a household average. A younger second spouse can meaningfully reduce how much has to come out early on.
This section also handles what happens if one of you predeceases the other. Because you entered each spouse's own expected lifespan, the tool compares the two and works out who's actually likely to go first and at what age, then handles the aftermath correctly: Social Security keeps the larger of the two benefits rather than a rough percentage guess, and every Required Minimum Distribution from that point forward is based entirely on the survivor's own age, since a surviving spouse legally inherits the account. If you'd rather not model a predecease at all, set each spouse's expected lifespan comfortably beyond your plan's own end age — just matching the two fields to each other isn't quite enough if there's an age gap between you, since the tool is comparing when each of you would reach your own number. One limitation worth knowing if you're also using delayed Social Security claiming: a spouse who dies before their own start age shows $0 toward the "keep the larger benefit" comparison, since they'd never actually claimed. In real life, a surviving spouse can still claim a survivor benefit based on the deceased's own record even if the deceased never filed; this tool doesn't model that, so a household where one spouse dies young and unclaimed may see a lower survivor benefit here than Social Security would actually provide.
There's one more choice built into this same section. By default, the tool models a surviving spouse as rolling the account into their own IRA, the common, usually simplest path. A checkbox lets you instead have the tool check every year after a death whether staying an IRA beneficiary would produce a smaller required withdrawal than a rollover would that year. It stays on the beneficiary path for as long as that's true, then switches permanently to the rollover the first year it isn't, since a rollover can't be undone once made. Because of that, turning this on can never leave you worse off than the plain rollover; either it finds an advantage and holds onto it for as long as it lasts, or it finds nothing and behaves exactly like the default from day one. One nuance: if the deceased spouse dies before reaching their own RMD start age (73 or 75 depending on birth year), IRS rules give the surviving spouse beneficiary a deferral window — no beneficiary RMDs are required until the later of the year after death or the year the deceased would have first reached their own RMD age. The tool applies this rule automatically; during that window, no beneficiary RMD is shown, and the annual comparison resumes once the window closes.
Federal tax is always calculated from the real IRS brackets, using no assumptions from you; state tax works differently, it's optional, and simpler. Turning this on reveals a choice between two options. Pick "No state income tax" for the states that don't tax income at all (Alaska, Florida, Nevada, New Hampshire, South Dakota, Tennessee, Texas, Washington, and Wyoming), or "Other" to enter your own rough effective rate, applied to the same taxable-income figure used for the federal calculation, a rough approximation rather than that state's own precise rules, since actual state tax treatment of things like Social Security and retirement-account withdrawals varies widely and isn't modeled state by state here. Leaving this off, the default, simply omits state tax from every result, the way this tool always has.
If either of you has a pension already coming in, or an existing annuity already in payout, whichever type it is, enter it here; the tool treats both the same way. For each spouse: an annual amount, whether it's inflation-adjusted or fixed in nominal dollars for the rest of that spouse's life, and a survivor continuation percentage, anywhere from 0% (stops entirely) to 100% (continues in full) if that spouse dies. This income offsets the year's withdrawal need directly, the same way Social Security does, but it's taxed differently: it counts as ordinary income and isn't run through Social Security's special partial-taxability formula. Each spouse's pension or annuity follows its own survivor rule, independent of the other spouse's, and independent of how Social Security's own "keep the larger benefit" rule works, and the reduction takes effect the year a spouse dies, not the year after. One thing to know if you're entering an annuity specifically: a "non-qualified" annuity, one purchased with after-tax money rather than IRA funds, normally has part of each payment excluded from tax as a return of your own principal. This tool doesn't model that exclusion, it treats the full amount as taxable ordinary income, so a non-qualified annuity's real tax bill will likely run a bit lower than what shows up here.
Important note: If the annuity you're entering was purchased with Traditional IRA or 401(k) money, a "qualified" annuity, and it's already annuitized, the payment stream you're entering here generally satisfies that contract's own RMD requirement automatically; no separate balance-based RMD applies to it. Enter its income in Section 8 only. Don't also include its original purchase value anywhere in Section 3's Traditional balance: an annuitized contract typically doesn't carry a remaining balance the way a regular IRA does, and including it there would double-count the same money, once as a balance inflating your calculated RMD, and again as income here.
One newer wrinkle, under SECURE 2.0 since 2023: if a qualified annuity's payment exceeds what's needed to satisfy its own RMD, the excess can now count toward RMDs on other accounts eligible to be aggregated with it. This tool doesn't model that credit, so if it applies to you, your actual required withdrawal from other accounts may run a bit lower than what's shown here.
The annual return you expect from each part of your portfolio. Two preset buttons — Moderate and Conservative — provide reasonable starting points; you can then adjust any individual number, which automatically switches the label to "Custom." If either bond field represents a CD or bond ladder rather than a fund, this is where its blended average yield across all the rungs goes. Custom Fund 1 and 2's own return assumptions are here too, each split into a yield (like bond interest) and a price return (like stock appreciation), so you can model any mix.
If the pension hasn’t started yet — a deferred benefit from a former employer, for example, that begins at 65 when you plan to retire at 62 — enter the start age in the “Pension starts at age” field for that spouse. The tool will add that income beginning in the year that age is reached and not before. Leave the field blank if the pension is already coming in; it will be counted from the first simulation year.
A QLAC is an insurance product: you hand over a lump sum (the tool assumes from your core bond fund), and in exchange you get guaranteed income starting later in life (often around 85). Buying a QLAC may also reduce your near-term RMD requirement, whether purchased from a Traditional IRA or a 401(k). Turn this on to see how buying one changes the picture—reducing your RMD, but also showing that you're moving money from one place to another, not adding new money to the plan. The IRS caps the total amount you can hold in QLACs at $210,000 across all your IRAs and 401(k)s combined (2026 limit); the tool enforces this cap automatically. The results narrative will show exactly how much of your core bond fund goes toward it once you've entered a premium. If you already own a QLAC and it's already paying out, don't model a purchase here, enter its income in the Pensions and Annuities section instead, the same way you would any other annuity already in payout.
A DIA is an annuity you purchase with after-tax money outside your IRA. Like a QLAC, it pays guaranteed income starting at a future age you choose; unlike a QLAC, it doesn’t reduce your IRA balance or affect your Required Minimum Distributions in any way. Enter the annual income you expect to receive in today’s dollars, and the age at which payments begin. If your contract includes a cost-of-living adjustment tied to inflation, check the inflation-adjusted box and the payments will grow each year with the plan’s assumed inflation rate. The premium is not tracked here — the tool simply adds the income stream from the start age onward. DIA income is treated as fully taxable ordinary income for federal and state tax purposes, consistent with the way pensions are modeled.
If you're 70½ or older, you can send money directly from your IRA to a charity, and it never counts as taxable income. This section lets you set an annual amount and see how it affects your Required Minimum Distribution and your taxes. It's added on top of the spending need you entered in Section 5, not subtracted from it — the two fields stay independent.
QCDs must come from an IRA, not a 401(k) or 403(b): Under IRS rules, Qualified Charitable Distributions are only available from traditional IRAs (and inactive SEP or SIMPLE IRAs). They cannot be made directly from an employer plan such as a 401(k) or 403(b). This tool combines all of your pre-tax retirement money into a single Traditional pool for modeling purposes and does not distinguish between IRA and employer-plan dollars within that pool. If a meaningful portion of your Traditional balance is still held in a 401(k) or 403(b), the QCD amount the tool projects may exceed what is actually eligible under IRS rules until you roll those funds into a traditional IRA. If you plan to use QCDs, confirm that sufficient IRA assets are available to cover them.
Age 70 vs. 70½: The QCD rules technically become available at age 70½, not exactly 70. The tool activates the QCD in the first full calendar year you are age 70 or older, which means someone whose birthday falls in the second half of the year could see a QCD credited up to six months before they’re technically eligible under IRS rules. The dollar impact is generally small — one year’s QCD amount at most — but if precision matters for your planning, set your retirement start age a year past your 70th birthday year, or simply note that the first year’s QCD may be slightly early.
Different from a QCD: this is giving cash after withdrawing it, which is fully taxable when withdrawn. Starting in 2026, cash gifts to a qualifying public charity get a modest deduction even if you don't itemize — up to $1,000 (single) or $2,000 (married) — and it stacks right on top of your standard deduction rather than competing with it. Give more than that cap, and the extra doesn't reduce your taxes any further in this tool, since it doesn't model full itemizing. Gifts to a donor-advised fund don't qualify for this deduction; it must be a direct cash gift to a public charity. Same rule as QCD: this adds to your spending need, it isn't part of it.
A one-time cost that only matters for a single year — a car, a home purchase or down payment, a gift to a family member, anything sizable enough to affect that one year's withdrawal. Each of the three slots has its own optional description, an amount, and the age it happens at; leave an amount at $0 to skip a slot you're not using. A lump sum is funded and taxed the same way as any other withdrawal, and it could count toward your RMD or cause you to exceed it. It also adds to your spending need rather than being part of it — a $40,000 car doesn't mean lowering your regular spending by $40,000 that year, it means an extra $40,000 comes out on top of it.
Research consistently shows that real retirement spending doesn't stay flat — it declines meaningfully over time. David Blanchett's 2014 "retirement spending smile" study found real spending falls roughly 26% before bottoming around age 84, and JP Morgan's Guide to Retirement data shows a similar 30% decline from the early 60s to the mid-80s. Most retirement calculators ignore this and model flat inflation-adjusted spending for 30 years, which overstates what you'll actually need in the later years of retirement.
This section lets you model up to three one-time permanent reductions in spending at specific ages. Each step-down fires once when that age is reached — the tool cuts the spending base by the percentage you set, and from that point forward the lower amount continues to inflation-adjust normally. Step-downs set before your retirement age are ignored.
The defaults (10% at 75, 10% at 82, 5% at 88) are a reasonable approximation of the research, but you can adjust them freely. Two step-downs instead of three is fine; just leave the third set to a percentage of 0 or leave it unconfigured. This works on your total spending need, so it affects everything the spending need covers — it isn't a separate healthcare line or a carved-out category.
The practical effect is meaningful: modeling spending that actually declines over time rather than assuming you'll spend the same (in real terms) at 88 as you did at 65 can raise your safe withdrawal rate from around 4% to closer to 4.7%, according to the Blanchett research. It doesn't mean lowering your spending need entering retirement — it means the tool will project lower real spending later, which translates to a more realistic picture of how much your savings actually need to carry.
In real life, the economy and financial markets can be volatile. To account for major swings in a given period, you can insert a bad stretch — a crash — or a good one – a rally – at any age you choose, using either a preset based on 2008-2009 (the real estate "crash") or 2022–2023 (both stocks and bonds down), or your own numbers. You can stack up to three of them at different ages if you want to see how your plan holds up against more than one bad patch, or a really good one, over a long retirement. Custom funds aren't affected by shocks or rallies; they always use their own assumed rates.
One thing to be aware of when combining market shocks with Capital Preservation: in a shock year, the modeled price change applies to bond and TIPS buckets as well as equity — sweeping stops and each bucket’s balance adjusts by the shock percentage you entered. Capital Preservation’s normal promise (bond principal stays flat unless spending needs it) applies to ordinary simulation years; it isn’t a guarantee that bonds are immune in an extreme stress scenario, where real bond funds can and do lose principal. If you want to model a shock that hits only equity, simply leave the bond and TIPS shock fields at 0%.
The three shock sections don't have to represent three separate, unrelated events. Each one supports up to two years of its own magnitude, but nothing stops you from giving them consecutive start ages instead of scattering them across retirement, and when you do, they chain together into one longer, evolving story rather than three isolated crashes. A decline followed by a gradual recovery is a good example: Shock 1 covers the down years, Shock 2 picks up right where it left off for the next phase of recovery, and Shock 3 finishes it off if you need a fifth year.
Here's a worked example. Say a $1,000,000 core bond holding drops $100,000 in year one, another $50,000 in year two, then recovers $25,000 a year for three years after that. Since the shock fields want a percentage, not a dollar amount, you'd convert each year against whatever the balance already is at that point:
That's very different than just picking one flat "market drops 20%" shock, and it's worth the extra work for a specific reason: the trajectory of the path can matter more than where it ends up. If nothing gets withdrawn from that bucket while it's depressed, because cash or another bucket is covering the need instead, a full recovery really does behave about the same as no shock at all. But if withdrawals come out of that bucket during the down years, they permanently remove principal that never gets the chance to participate in the recovery, even when the bonds' return, taken in isolation, nets out close to what a smooth, steady path would have produced. Modeling the decline and recovery as separate phases, rather than collapsing them into one average, is how you'd actually see that effect show up in your results instead of just an average.
A way to model a future change to federal tax rates—a "what if" for a different tax environment in the future, such as a rate increase or a reversion to lower rates. Tax bracket changes are rarely a single uniform shift across every bracket; historically they've almost always been concentrated at the top or the bottom of the ranges, so this section uses two separate adjustments rather than one: a taxable income threshold, and a rate adjustment for taxable income below it and another for taxable income at or above it, each in percentage points, positive for an increase or negative for a decrease. Leave one of the two adjustments at 0 to change only the other side.
Taxable income is a smaller number than you might expect, not the same as AGI (adjusted gross income), the figure that governs things like the senior deduction's phase-out. Take a couple with $54,902 in RMD withdrawals and $50,000 in Social Security, of which $36,517 is taxable; their AGI comes to roughly $91,000. But their taxable income, after the standard deduction and the 65-and-older senior bonus are subtracted, is only about $44,000, nearly $47,500 lower, essentially half of AGI. If you're picturing this threshold in AGI terms, you'll want a meaningfully lower number as the taxable income threshold you set.
Set an age too; from that age forward, both adjustments apply for the rest of the plan, and nothing before that age is changed. A bracket that straddles your threshold is split at the exact dollar line; the portion below the threshold uses the low-side adjustment, and the portion above it uses the high-side one, so nothing gets rounded to the side closer to it.
This shifts tax rates only. The bracket thresholds themselves—the income levels where each rate kicks in—don't move, and neither does the standard deduction or the 65-and-older senior bonus described under Reading the Results, which is set to expire in 2029.
The bottom of the form is where you choose how returns are modeled and then run the tool. Three modes are available, selected by clicking the button for the one you want before clicking Run the Simulation.
Stable is the default. It uses the return assumptions you entered in Section 9 — one smoothed average per year, every year — and is the most straightforward way to test a specific scenario. Market shock overlays (Sections 16–18) and the tax-rate change (Section 19) apply only in this mode.
Monte Carlo replaces the fixed return path with thousands of randomly generated year-to-year return sequences, then shows the range of outcomes and the probability your money lasts. When you select it, a settings panel opens below the mode buttons. Like Return Assumptions, it has Moderate and Conservative presets, though the numbers are different in kind: Monte Carlo needs the raw historical arithmetic mean and standard deviation, not the smoothed geometric average that Section 9 uses, so don’t expect the two to match directly. Adjusting any field switches it to Custom. Market shock overlays are replaced by the ongoing randomness built into the simulation itself; it runs using whichever withdrawal approach Section 2 is set to.
Historical replays every overlapping stretch of actual market history that fits your plan’s length — one full run per starting year — using real S&P 500 total return, real 10-year Treasury total return, and real inflation from that year forward (1928–2018, sourced from Aswath Damodaran’s NYU Stern dataset). A 30-year plan gets 62 starting years tested; a 40-year plan gets 52. No inputs are needed: the real relationship between stocks and bonds in each year is whatever actually happened. Results name the specific best and worst starting years, so you can see what 1929 would have done to your plan, not just how likely a bad outcome is in the abstract. Worth knowing: overlapping windows share most of their years with their neighbors, so the trials aren’t as independent as Monte Carlo’s, and the dataset reflects only this country’s market history — no guarantee the future resembles the past. Market shock overlays are replaced by the historical sequence itself. If your plan is longer than 91 years, there isn’t enough history to run it and the tool will tell you so; use Monte Carlo instead.
Once you click Run, several things appear on the results page, top to bottom:
If you also ran Monte Carlo, its own version of the headline, summary numbers, and milestones appears in a separate box at the top of the page, replacing the ones described above rather than sitting alongside them, so you're never looking at two different answers to the same question. Its stats and milestones are true medians computed independently across every trial, not the numbers from any single one, so a milestone reads as, for example, "equity first tapped, age 94, in 72% of trials," telling you both when something typically happens and how often it happens at all. Its own chart shows the full range of outcomes across all trials, not just the median. Below that box, this card still appears, but now showing only the two pieces Monte Carlo doesn't already cover: the surplus/deficit chart and the four year-by-year tables, both driven by the one trial whose ending balance landed closest to the median, a real, internally consistent path, not your fixed-rate returns scenario. A note there says so, since its exact numbers may not match the medians reported above it.
Income & spending
Full bucket balances
Interest & dividend income
Federal tax & after-tax income
Double-check the number you typed first — a stray zero, or a field left at its default, is the most common cause of a surprising result. If the math still looks off after that, flag it and look into it further. If you're fairly certain there is a problem with the tool, contact us.