Your plan
Change anything; the simulation re-runs instantly.
Watch the buckets
You spend from Bucket 1. Each year the chain refills toward its targets, front to back — except during a downturn, when nothing is refilled: Bucket 1 drains first, then the buckets behind it, in order. The reservoir is never refilled. All buckets share one dollar scale, so their heights show where your money actually sits.
Did the buckets earn their keep?
Same dollars, same withdrawals, same downturn — your buckets against the alternatives most people actually consider. Tap any line to show or hide it. Sometimes the buckets win, sometimes they don't; the tool's job is to show you which, for your numbers.
Safe / Moderate / Growth are buy-and-hold blends built from your safe and growth rates — 30, 60, and 85% in growth — held without rebalancing, withdrawals taken proportionally so the mix holds steady. The rebalanced 60/40 is a fixed benchmark built from your safe and growth rates — 60% growth, 40% safe, rebalanced yearly, and deliberately independent of your bucket sizes and growth shares so it stays a stationary yardstick as you tune the buckets. It's worth understanding on its own: what it is and why it behaves the way it does →
Year-by-year table — show your work
How the math works
Order of operations
Each year: withdraw from Bucket 1 (reaching further down the chain — Bucket 2, then the next, then the reservoir — if it's dry) → cascade-refill the front buckets toward their targets, front to back, unless the refill rule below says not to → apply each bucket's return: its growth share at your growth rate, the rest at your safe rate. Refilling before growth means transferred money earns the destination bucket's return that year — with the default growth shares, the lower one — a slightly conservative choice, made deliberately.
The refill rule
Each front bucket has a target measured in years of spending, re-priced to the current year's withdrawal so targets rise with inflation. In normal years the cascade tops every front bucket back to target, Bucket 1 first, each drawing on the buckets behind it. During a downturn, with Don't refill during a downturn on, nothing is refilled at all — not even Bucket 2 into Bucket 1. Bucket 1 simply pays the bills until it is empty; then the withdrawal comes out of Bucket 2, then the next bucket, and finally the reservoir — a forced sale at depressed prices, shown, not smoothed over. Switch the rule off and the cascade refills every year as if there were no downturn, selling at the bottom to do it.
Coming out of a downturn your front buckets are drained, and refilling them to target in a single year means liquidating years of spending out of the reservoir in one January — which nobody actually does. With Rebuild the buffer gradually on, in any year the refill runs the reservoir gives up no more than that year's spending plus one extra year — counting both what the withdrawal pulls from it and what it sends forward. The cascade still runs front to back, so Bucket 1 gets the extra year first, and money moving between front buckets, Bucket 2 into Bucket 1 say, is not capped. Take a two-year Bucket 1 and a two-year Bucket 2, both emptied by a long downturn. The first year after it, the year's spending comes out of the reservoir and Bucket 1 gets the extra year: Bucket 1 holds one year, Bucket 2 none. The second, Bucket 1 spends its year and takes two back: two and none. The third, Bucket 1 spends one and takes one back, and Bucket 2 gets the extra year: two and one. The fourth: two and two. A four-year buffer takes four years to come back, and no year sells more than two years of spending out of the reservoir. Switch the rule off and watch what the lump-sum refill costs — with small front buckets it is a rounding error, and it grows with the size of the buffer.
What the downturn does
You set three things: how far the market drops, how long it stays down, and which year it starts. The drop is applied once, then the market holds flat for the full length — struck assets earn nothing until the window ends, at which point normal returns resume from wherever the balances landed. There is deliberately no recovery leg: the shape and timing of a rebound is a forecast, and inventing one would be exactly the kind of false precision this tool avoids. Each bucket's exposure is its growth share: that part takes the drop and then sits flat, while the rest keeps earning your safe rate straight through — so Bucket 1 at 15% growth barely moves and the reservoir at 85% takes most of the hit. A flat, un-recovering downturn is a pessimistic scenario on purpose; real crashes usually do recover eventually, and a retiree who wants to model that can shorten the "stays down" figure to represent the depth at a chosen point in time.
Where the growth rate comes from
The growth asset's default rate is not a forecast. It is solved from two things: history, and the crash you set. Since 1928 the S&P 500 with dividends compounded at 10.0% a year (NYU Stern / Damodaran data, 1928–2025). That figure already contains every crash in the record — 26 losing years, between 5 and 12 in any 30-year stretch, 7 in a typical one — so it cannot be used for the years outside a crash you have modelled separately without counting crashes twice. But the crash you place here is one event, not seven. Using the market's rate between bad years (about 15.7% if you strip out every year that lost 10% or more, 20.4% if you strip out every losing year) would treat your single crash as the only bad patch in the whole plan. The tool resolves that exactly: it finds the rate r such that (1 + r)^(N − k) × (1 − drop) = 1.10^N, where N is your plan length in years and k is the number of years your downturn occupies. With the defaults — 30 years, a 30% drop held for 3 years — that is (1.10^30 ÷ 0.70)^(1/27) − 1 = 12.65%. Your timeline as a whole then compounds at history's 10.0%, and what you are testing is the placement and shape of the crash, not its total damage. Change the crash and the rate re-solves; move the slider and the note beside it tells you which way you have left history. With the downturn off there is nothing to calibrate against, and the rate is simply 10.0%.
The comparison lines
Every line spends the same dollars on the same schedule through the same downturn; tap any to show or hide it.
Safe, Moderate, Growth are buy-and-hold blends — 30%, 60%, and 85% in growth, the rest in cash — built from your safe and growth rates. They're bought once and never rebalanced, and each year's withdrawal is taken proportionally from both sleeves so the mix stays put. This is what "just leave it in a balanced fund and don't touch it" actually does. Note that the blend's cash portion isn't protected the way Bucket 1 is — it only dilutes the blended return. Same asset, different job: your Bucket 1 cash is ring-fenced and spent first, so it shields income; the blend's cash just averages the pain. The Growth blend, 85% exposed, craters in a crash nearly as hard as being all-in on stocks.
Rebalanced 60/40 holds a fixed 60% growth / 40% cash mix (built from your safe and growth rates) and rebalances to it every year — buying stocks all the way down and trimming them on the way up. It's fixed on purpose: it does not follow your bucket sizing, so when you resize your front buckets you watch the bucket line move against a benchmark that stays put. Expect it to finish low here, and understand why before you draw a conclusion from it: rebalancing's payoff is buying cheap shares that later become dear, and this model's downturn has no rebound to collect on. What's left is the other half of the mechanism — trimming your fastest-compounding asset every year to buy your slowest — which over thirty years is expensive. Compare it to the Moderate line, which holds the same mix and never rebalances: the gap between them is the price of rebalancing in a market that doesn't bounce. If you believe crashes recover, this line is understated. It's the least intuitive benchmark and the one whose result you should question hardest, so it gets its own page.
How big should the front buckets be?
Bigger is not safer. The safe asset that makes up most of the front buckets has historically earned almost nothing after inflation — US T-bills returned about 0.3% real over the last century — so every year of spending you park in the front is mostly money not compounding, and against withdrawals that rise with inflation, that bleeds. A large buffer also takes longer to rebuild after a downturn, which is its own drag. Size the front buckets to the length of the downturn you actually want to sit through, and understand that each extra year of cushion is bought with real money. The tool will show you the price; it won't pick a number for you.
A drift worth knowing about
Because your front-bucket targets grow only with inflation while the reservoir moves at its own rate, the strategy's mix between safe and growth assets shifts over time on its own — no glidepath knob required, and none offered, because adding one would just fight the rule already in place. Which way it drifts depends on you: if your growth outruns your withdrawals, equity share rises with age; if withdrawals are eating into your wealth — the more common case — it falls. Either way it's a consequence you can watch in the year-by-year table, and steer by changing your bucket targets.
What this tool is not
No taxes (planned for a later phase — assets are treated as one pool), no Monte Carlo success percentage, no verdict badge. A success probability is a green checkmark that hides its own machinery; a downturn you place on the timeline — depth, length, and start year all yours — is something you can interrogate. Everything here is in nominal dollars: returns as people quote them, withdrawals escalated by your inflation number where the scheme calls for it. One honest caveat about the comparison: which line wins depends heavily on assumptions this model makes explicit rather than hides. A permanently repriced market flatters cash-heavy strategies and punishes rebalancing; a market that rebounds does the reverse. Read the lines as a map of consequences under your assumptions, not as a verdict. What they buy isn't the highest ending balance; it's income that never depends on selling something that just dropped, and a plan you can actually stick to when the screen is red. This tool exists to show you that trade with the price tag attached, not to talk you into either side.
Glossary
- Reservoir
- The last bucket. It has no target and is never refilled — it's the growth engine that feeds the chain, and the only bucket you never take income from directly.
- Growth share
- How much of a bucket sits in the growth asset; the rest is in the safe asset. It sets the bucket's return — growth share × growth rate plus the rest × safe rate — and how much of a downturn it takes. The defaults run from 15% in Bucket 1 to 85% in the reservoir.
- 4% rule (Bengen)
- Withdraw 4% of your starting assets in year one, then give yourself a raise for inflation each year regardless of what markets do. Fixed income; can deplete.
- % of current balance
- Withdraw a fixed percentage of whatever the portfolio is worth that year. Income swings with markets; mathematically never fully depletes.
- Sequence-of-returns risk
- Why a downturn's start year matters more than its depth: the same drop early in retirement, while withdrawals are eating a shrunken portfolio, does far more damage than the identical drop fifteen years in.
- Rebalancing
- Periodically selling whatever grew past its target weight and buying whatever fell below it, to restore a fixed mix. It mechanically buys low and sells high — but demands you buy stocks during a crash.
- .buk file
- Your plan saved as a plain JSON file on your device. Load it later to pick up where you left off. Nothing is ever sent anywhere.