Full plan · Monte Carlo · 1,000 trials
Enter your plan on the left. Every change re-runs 1,000 randomized market histories and updates the odds, the range of outcomes, the stress tests, and the trade-off table.
Probability of success
—%
Treat this as a conversation number, not a pass/fail grade — many advisors consider roughly 75–90% a workable planning band, and a plan with flexible spending can run lower safely.
Portfolio value over time (today's dollars, all accounts)
The shaded fan is the spread across all 1,000 trials at each age — the top edge is the 90th-percentile path, the bottom edge the 10th-percentile path. Where the bottom edge touches zero, at least 10% of trials have run out of money.
Stress tests
| Scenario | Result | vs. base |
|---|---|---|
| Running… | ||
Historical replays are single deterministic paths: the named sequence of approximate real returns hits in your first retirement years, with expected returns before and after — they show sequence-of-returns risk, not probabilities. The correlation and Social Security rows re-run all 1,000 trials.
What moves the odds
Educational model, not financial advice. Everything is real (inflation-adjusted), so results read in today's dollars.
Annual returns are drawn iid from a normal distribution (or Student-t in research mode). The planning default for stocks is 6.8% real mean / 17.2% SD — a deliberately conservative forward capital-market assumption, below the roughly 7.8% historical arithmetic mean of U.S. equities, which is available as a research toggle. Bonds 2.0% / 6.5%; cash 0.5% / 1.0% (stylized low-vol); stock–bond correlation 0.10 in the base, 0.40 in the stress row. The annual fee is subtracted from returns before cash flows. iid draws understate long slumps and momentum; the historical replays exist to show what iid sampling hides.
Three buckets (taxable, tax-deferred, Roth) share one allocation and rebalance annually. Withdrawals are grossed up for tax: deferred at your ordinary rate, taxable at your capital-gains rate assuming 50% of each withdrawal is gain (basis is not tracked), Roth tax-free. RMDs start at 73 or 75 by birth year (SECURE 2.0, computed automatically) using the Uniform Lifetime Table; net RMD proceeds beyond spending are reinvested in taxable. Real tax brackets are nominal and progressive; flat rates in a real model are an approximation.
Social Security is a real annuity from the claiming age; the model assumes full CPI-W COLA, and CPI-W does not perfectly match retiree household inflation. The optional trim scenario multiplies benefits by 0.85 from a chosen year. For couples, the survivor keeps the larger check (once past their own claiming age — an approximation) and spending falls to the survivor factor. Guardrails cut spending by the slider amount when the current withdrawal rate breaches 120% of its starting level and restore below 80%; cuts are sticky until recovery and never take spending below 70% of plan. Because guardrails adapt, ruin gets rare — read the “trials that cut spending” stat alongside the success rate. The smile option trims real spending 1%/yr past the chosen age. Healthcare is a flat extra line from 65 growing faster than CPI — IRMAA brackets are not modeled. Cash flows land at year-end, which is slightly optimistic.
Trials are seeded deterministically from your inputs, so the same plan always shows the same result — use re-roll to see sampling variation. Historical replay data are approximate real total returns.