Under the hood
What the tool does, what the simulation models, what the results show — and what's outside the scope of the analysis.
What the tool does
Will It Last — Scenario Explorer is a retirement cashflow simulation tool. You enter your scenario; the model runs it. Nothing is crowned the right answer — you adjust the inputs, see how the results move, and draw your own conclusions.
Every result describes what the model returned for the exact inputs you entered. The tool shows outcomes; it does not recommend a path. Decisions about what to do with those results are yours.
Designed for U.S. residents — the model uses U.S. tax rules, Social Security, and U.S. retirement account types (401k, IRA, Roth IRA).
The scenario form collects:
After running, the spend slider lets you scrub the success rate live across a precomputed grid — no re-simulation needed as you move the slider. Changing any structural input (ages, balances, SS) marks the result stale and requires a new run.
The simulation
The assumptions and data sources the simulation is built on. Transparent — so you can evaluate whether the model fits your situation.
| Component | Approach |
|---|---|
| Simulation engine | Markov three-state regime model (bull / bear / recovery) with persistence controls. Each year transitions between states based on historical probabilities, capped to prevent unrealistically long streaks. |
| Equity return assumptions | Drawn from institutional capital market assumptions (Vanguard, BlackRock, Fidelity presets available). Total return basis — dividends included. |
| Historical calibration | S&P 500 annual returns, 1995–2024 (FRED data series). Used to calibrate regime transition probabilities and return distributions. |
| Bond returns | Bloomberg US Aggregate Bond Index (intermediate-duration aggregate) as the bond proxy. |
| Inflation | CPI-U (all urban consumers, FRED series CPIAUCSL). Note: the model uses CPI-U, not CPI-E (the elderly-specific index), which may slightly understate inflation for retirees with above-average healthcare spending. |
| Equity allocation | Glides from a higher equity allocation pre-retirement to a lower allocation in retirement over a configurable transition period. |
| Component | Approach |
|---|---|
| Federal income tax | Current IRS brackets (married filing jointly or single, matched to your household). Standard deduction applied. Brackets are not adjusted for future inflation in the model — a conservative assumption. |
| Long-term capital gains | 0% / 15% / 20% brackets, stacked on top of ordinary income per IRC §1(h). |
| Net Investment Income Tax | 3.8% federal surtax on investment income above the MAGI threshold (not inflation-indexed). |
| Social Security taxation | Provisional income rules per IRC §86. Up to 85% of benefits taxable above income thresholds. Thresholds are not inflation-indexed — meaning more of your benefit becomes taxable over time in real terms. |
| State income tax | Modeled as a flat rate approximation for the selected state. Progressive states use an estimated effective rate at typical retirement income levels. Changing state requires a new run. |
| Medicare IRMAA | Current CMS Part B/D surcharge tiers applied to projected income each year. Uses a simplified current-year income proxy rather than the actual two-year look-back. |
| Component | Approach |
|---|---|
| Required Minimum Distributions | IRS Uniform Lifetime Table (post-SECURE 2.0). RMD start age follows SECURE 2.0 rules: age 73 for those born 1951–1959, age 75 for those born 1960 or later. |
| Withdrawal sequencing | Roth last — taxable accounts and traditional first, Roth preserved. Adjusted in bear years and when the cash buffer falls below target. This is a fixed model rule, not a per-run setting. |
| Cash buffer | Model maintains a target cash buffer (a multiple of annual spending) to avoid forced selling in down markets. Buffer is rebuilt in normal market years. |
| Social Security | Entered as the expected annual benefit at your claimed age. Early and delayed claiming factors follow SSA rules. Spousal and survivor benefits are not separately modeled. |
| Working-years savings | While still earning salary: lifestyle spend is modeled at 90% of gross salary (household total when both spouses work). 10% is saved — typically 6% to traditional IRA/401(k) via salary deferral and the remainder to cash/brokerage. After retirement, your entered monthly spend goal applies instead. |
These settings are the same for every run. They reflect industry-standard retirement planning practices and are not configurable.
| Assumption | Value |
|---|---|
| Cash buffer | 1.5× annual lifestyle spend maintained in cash at all times. Rebuilt in normal market years; not force-rebuilt in bear years. |
| Withdrawal order | Roth last — taxable accounts (brokerage, cash) drawn first, then traditional IRA/401(k), then Roth. Adjusted in bear years and when cash falls below buffer target. |
| Filing status | Married filing jointly for couple plans; single filer for solo plans. Thresholds follow IRS brackets for the matching status. |
| Equity glide path | Equity allocation steps down from a higher pre-retirement percentage to a lower in-retirement percentage over a 5-year transition ending at the primary retirement age. |
| Regime model | Markov three-state (bull / bear / recovery) with semi-Markov streak caps. Maximum consecutive bear years and bull years are capped based on empirical 1995–2024 S&P 500 data. |
| Flexible spending guardrail | When flexible spending is active, the default guardrail is withdrawal-rate based. Full lifestyle spending is modeled as long as the annual portfolio draw stays at or below a ceiling (about 6% of current balance); only a draw above the ceiling is trimmed, to the dollar amount that holds the withdrawal at that ceiling — an absolute level tied to current balances and income, not a fixed fraction of the target. |
| Investment fee drag | 0.50%/yr applied to portfolio returns each year, reflecting moderate fund and platform costs. |
What the results show
Each surface in the walk-through describes a different slice of what the model returned for your inputs.
| Output | What it describes |
|---|---|
| Success rate | The share of 500 simulated paths where the portfolio funded the full plan to the end of the planning horizon. A 75% rate means the plan ran out of money in 25 of 100 paths — not that any specific future is predicted. |
| Median ending portfolio | The portfolio balance at the end of the plan on the median simulated path — the 50th percentile outcome, in today's dollars. |
| Modeled lifetime tax | Total modeled federal and state income tax across all years of the plan, in today's dollars. Reflects the tax trajectory the model followed for your inputs. |
| Income mix over time | Where spending money comes from each year on the median path — work income, Social Security and pension, other income, and portfolio draw. Shown as a stacked bar by year. |
| Account balances over time | How each bucket (traditional IRA/401k, Roth, brokerage, cash) changes across the plan on the median path, in today's dollars. |
| Range of outcomes | The 10th–90th percentile band of portfolio values across all paths. The spread reflects market uncertainty in the model — not model error. A wide band means the outcome is highly sensitive to market sequence. |
| Spend sensitivity | How the success rate changes across different monthly spend levels, from the grid computed at run time. The live slider scrubs this grid client-side — no re-simulation needed. |
| Withdrawal rate sensitivity | What the model returns when the draw rate from traditional accounts is adjusted up or down — testing different bracket-fill strategies on your inputs. |
| Retire-later sensitivity | What the model returns if the retirement date is extended by one, two, or three years — holding all other inputs equal. Shown only when baseline success is below 85% and later retirement is meaningfully different. |
| State comparison | What the model returns if the state of residence changes — re-running the simulation under another state's tax rules, everything else equal. |
Know the limits
Every model has boundaries. Here's an honest accounting of what's outside the scope of this analysis — and what that means for how you use it.
The PDF record
After running a scenario, you can export a landscape PDF documenting exactly what the model returned for your inputs — the same content as the walk-through, formatted for printing or sharing. It runs up to 12 pages; the sensitivity pages appear only when they apply to your scenario.
Cover
Run identity
Tool name, run date, and an inputs fingerprint that pins down the exact scenario.
Inputs
The scenario you ran
Every input exactly as entered — ages, balances, Social Security, pensions, state, and spend target.
Headline
What the model returned
Success rate, median ending portfolio, and modeled lifetime tax, across all simulated paths.
Charts
Balances, income & range
Account balances and income mix over time, plus the 10th–90th-percentile outcome band — one page each, on the median path.
What moved it
Spend & sensitivities
Monthly spend — the largest lever — and, where they apply, withdrawal rate, a later retirement date, and a state comparison, one page each.
Close
In short & disclaimer
Plain-language observations from your run, then the full legal disclaimer.
Enter your numbers and see what the model returns.
Run your scenario →For educational and illustrative purposes only — not financial, tax, or legal advice.