— · All stress scenarios use your actual numbers ·
Base ending should match Dashboard ✓
Showing Sample Data
Connect your sheet to stress-test YOUR numbers. Set up →
Retirement Intelligence · Stress Testing
What happens to your plan when markets crash?
Your Dashboard assumes a smooth 7% every year. Real markets don't work that way — so this page runs your actual saved plan through bad markets in four ways: real crashes from history, a crash you design, every crash level at once, and 1,000 random futures. Nothing here changes your saved plan.
Pick a Crash Scenario
Choose a preset (2008, stagflation, COVID) or dial in your own with the sliders below.
This is a safe what-if sandbox. Everything here is layered on your saved numbers just for testing — it never changes your real plan. To change your plan for keeps, edit it on the Inputs page and Save.
Your Plan Fingerprint — Confirm This Is Your Data
Portfolio
—
Starting
Legacy Goal
—
Target end
Base Ending
—
Matches Dashboard ✓
Plan Through
—
Both partners
Must-Pay Floor
—
base + health + debt
These are your real saved plan numbers, pulled straight from your sheet — give them a quick look to confirm this is your data before you test.
Why Year 1 is the worst timing:
A crash early in retirement forces you to sell at the bottom while withdrawing — those shares never recover.
The same crash in Year 20 barely dents your plan because you had 20 years of growth first.
Test 1 · If History Repeats
What if a real crash from history happened again?
Runs your actual saved plan through the real year-by-year losses of history's worst bear markets — as if that exact crash hit at the very start of your retirement — and shows the most extra you could still safely spend each year and still reach your legacy goal.
Your calm-weather safe extra spend is —/yr. Each row shows what still holds if that crash hit at the worst possible time.
The exact numbers behind these: each scenario uses the actual S&P 500 total returns (with dividends reinvested) for that period — 2008 was −37% in a single year; the dot-com bust fell three years running; 1973–74 dropped −15% then −26% and is modeled with that era's ~9% inflation eating your spending; the Great Depression is the worst on record at roughly −64% over four years. Multi-year crashes apply their real losses in order, with the worst year where it actually landed.
Why isn't 2020 here? The COVID crash fell −34% in five weeks but fully recovered the same year (+18% for 2020). On a yearly basis it wasn't a sustained crash — and a plan that holds through 2008 shrugs off a fast dip like that.
Test 2 · Crash Test
One test at a time. The Crash Test (▶ Run New Test) and the Scenario Sweep both run on your live Google Sheet, so they take turns — if you start one while another is still running, it waits a few seconds for its turn instead of running alongside it. Kick them off one at a time. The Monte Carlo further down runs entirely on your own computer, so it’s independent and never has to wait.
Pick a preset or dial in your own downturn with the sliders, then press ▶ Run New Test.
Scenario Parameters
▶ Your test starts at your base plan (0% — no crash). Move the sliders below to model a downturn. The −30% Crash card up top is a fixed worst-case reference, not the current setting.
Year 1 Market Return 0%
-50%0%+30%
A crash in your first retirement year is the worst scenario — you sell at the bottom to fund living expenses.
Years of Poor Returns 1
1 yr3 yrs5 yrs
How long the downturn lasts. Even 2008 fully recovered within 5 years.
Inflation Rate 3.0%CRASH TEST ONLY
1%3%10%
Higher inflation erodes purchasing power every year. 1970s saw 8-12%. This slider only changes the crash test above — not the Monte Carlo below, which uses your saved plan's inflation.
Spending Adjustment 0%
-30%0%+20%
Your most powerful lever. Cutting 10% in a downturn dramatically changes outcomes.
SS COLA Rate 2.5%
0%2.5%5%
Zero models a worst-case SS reform scenario.
+ Add a second crash later in retirement (optional)
STARTS (YEAR)
YR-1 RETURN %
LASTS (YRS)
Leave blank for no second crash. Models a downturn later in retirement, on top of the early one set by the sliders.
Runs your real engine on these settings — give it ~20–30 sec (thousands of calculations). No estimates: this is a true run of your plan.
Portfolio Over Time — Base Plan vs. Crash Scenario
── Base Plan── With Crash- - Legacy Goal▓ Crash Period
Return Assumptions — Year by Year
Red bars = crash years · Your plan must survive these
Stress Test Result — at a glance
Base Plan Ending
—
no crash
With This Scenario
—
—
The Hit
—
base − scenario
Extra You Can Still Add
—
above your floor, this crash
Drops Below Legacy
—
year, or holds
Your Legacy Target
—
tested against
—
Top Actions
Want to change your plan based on what you see here? Edit it in Inputs, Save, then click Reload Data above.
Saved Tests REAL ENGINE
Every time you press ▶ Run New Test above, that result saves here so you can compare your scenarios side by side — including any with a second, later crash.
What this shows. A crash early in retirement is one of a plan's biggest risks — you're forced to sell investments while prices are down. Each test reveals which lever bends first: your spending (the plan trims to protect itself) or your legacy goal (the damage lands there instead). Holding your legacy goal through a steep, multi-year crash means real resilience is built in.
Saved Tests — compared to your base plan · newest fills the next slot, 6th overwrites Test 1
Base plan:load your plan to see base numbers
Test 3 · Scenario Sweep
Scenario sweep — safe extra spend at every crash level
Uses your real plan data. Shows what Phase 1 extra you could safely spend and still hit your legacy goal — even if that crash hits in year 1. Other phases scale proportionally.
Running real engine — scenario 1 of 8…Each scenario runs on its own — rows fill in as they finish · keep tab open
How this works: The crash test asks one question — "if the market drops X% in year 1, what's the most extra I can safely spend and still hit my legacy goal?" This sweep runs that question at every crash level automatically.
▶ Sweep All Scenarios — calls your live engine for each crash level using full Master formula calculations: taxes, RMDs, and SS staggering all reflected. Each scenario runs on its own, so rows fill in one by one as they finish. Keep the tab open while it runs.
Other phases scale proportionally to Phase 1. All scenarios assume normal returns after year 1.
Test 4 · Monte Carlo
The Monte Carlo below is independent of everything above.
The sliders up top — including the inflation slider — only shape the one crash scenario in the crash-test chart.
The Monte Carlo ignores them completely: it builds its own 1,000 random futures and takes inflation from your saved plan.
So to change the inflation the Monte Carlo uses, edit it on the
Inputs page and reload — not with the slider here.
TEST 2 · MONTE CARLO
Monte Carlo Simulation
The scenarios above each test one specific crash you define.
But which sequence of returns would you actually get in retirement?
Monte Carlo answers that by running 1,000 randomized futures and counting how many your plan survives —
giving you a true probability of success.
Reading this against your Dashboard: the average of these runs should land near your Master projection — a good sign it's using your real portfolio. But the median and success rate come in lower, on purpose — that gap is real-market volatility your smooth projection can't show. Plan around the median and the success rate, not the average: half of all futures are worse than the average.
How These Tests Differ
Crash Scenarios
Monte Carlo
How it works
You set the crash
Random each run
Output
1 outcome
1,000 outcomes
Tells you
"What if 2008 hits?"
% chance of success
vs Dashboard
More control
Most rigorous
After running:✅ 90%+Very strong·⚠️ 75–89%Acceptable·❌ Below 75%Needs adjustment
Monte Carlo Simulation
1,000 randomized market scenarios · Runs entirely in your browser · No re-query needed
Simulations
Return Model
Spending Flexibility
Year 1 Return (stress-condition)
ⓘ New to this? What Monte Carlo and these numbers mean
Monte Carlo means we don't bet on one future — we replay your plan through 1,000 different random market histories (calm years, crashes, long flat stretches) and count how it turns out. It answers "how often does my plan work?" rather than "what happens if returns are exactly average?"
Never ran out of money — the share of those 1,000 futures where your money lasted your whole life. This is the number most planning tools lead with.
Success rate — a stricter bar: futures where you not only didn't run out, but finished with your full legacy goal still intact.
Median — the middle outcome: half the futures did better, half did worse.
P10 / P90 — your unlucky and lucky edges. P10 is the result only the worst 10% of futures fell below; P90 is the result only the best 10% beat. A wide gap means your outcome rides heavily on market luck.
—
Won't Run Out of Money
—
of 1,000 futures — if you trim discretionary spending in down years (what most retirees do)
—
never ran out at full spending
—
reached full legacy goal at full spending
—
Median Ending Balance
—
Worst 10% Scenario (P10)
—
Best 10% Scenario (P90)
—
Scenario Breakdown
✅ Exceed goal—
⚠️ Survive (≥$0)—
❌ Run out—
What if you treat your "extra" spending as optional?
Extras rigid — spent through crashes
base + debt + extras · worst case for behavior
—
Cut to this → sustainable
—
Extras flexible — paused in rough years
base + debt only · what most retirees actually do
—
Same 1,000 market histories — the only thing that changes is your extra (discretionary) spending. Your base living + debt is the non-negotiable floor and is never cut here. The left number keeps spending every extra dollar even through crashes; the right number pauses the extras in rough years. Your real resilience sits between the two — the closer to the flexible number the more willing you are to trim extras when markets are down. This is a what-if only; it doesn't change your saved plan.
Portfolio Range Across All Scenarios
Teal line = your middle-of-the-road outcome · Shaded = range of all scenarios · Orange = worst 10% · Top of chart capped for readability (a few lucky runs go far higher)
Chance You Can't Cover Must-Pay (by Year)
Ending Balance at Full Spending
■ ran out ■ below legacy goal ■ at/above goal
Running Monte Carlo Simulation
Testing 1,000 random market futures against your actual plan...
0 / 1,000
0%
This takes 5–15 seconds depending on your device
About these numbers:
All 82 columns behind your plan are calculated fresh in your browser and checked line-by-line against your Google Sheet — and match it to the exact dollar on almost every one. On long-range running balances you may occasionally see a difference of a few dollars — that’s normal rounding between two calculators compounding over decades, not an error. Your Google Sheet stays the official record.
