Start here
The basics
The rule keeps planned retirement spending unchanged until the portfolio grows enough above its starting value, measured after inflation. It then gives the budget a permanent percentage increase.
After a raise, the rule waits several years before considering another. It does not cut those raises when markets fall, so the plan must still be able to fund the higher spending.
Illustrative numbers
A strong decade, then a 30% fall
Portfolio the first retired year (the base)$1,000,000
Base grown twelve years at 2.5%$1,344,889
The bar to clear (+50%)$2,017,333
Portfolio twelve years on$2,086,219
The cards’ bill that year$80,693
Spending after the step (×1.10)$88,763
Portfolio the next year, after −30%$1,460,353
For the first twelve retired years nothing happens — the portfolio is comfortably ahead, but never 50% ahead of a base that is itself growing with inflation. In the thirteenth the bar is cleared and the budget steps once, from $80,693 to $88,763. The 30% fall the following year takes the portfolio to $1,460,353, far under the bar, and the rule does not answer it: the raised scale of 1.10 is kept, so the next year’s bill is $90,982 and spending carries on rising with inflation from its new level. No clamp bit in this run — the ceiling only appears on a fifth step, where a compounded 1.61 is cut back to the default 1.50.
Calculation transparency
How it works in MoneyWhatIf
- 01
The base is the portfolio in the first retired year the rule is live in — the salary has stopped, the strategy’s window is open, the cards carry more than $0.50 of spending, and the accounts hold more than $0.50. That year spends exactly as written, at a scale of 1, and it also starts the cooldown clock, so the earliest possible first raise is the base year plus the cooldown.
- 02
Portfolio means last year’s closing account balances, each floored at zero and added up — the same statement a required minimum distribution is worked out from. Only a funded run has such a statement, so a run without one keeps its spending as written whatever the setting says.
- 03
The bar the portfolio must clear grows with the plan’s own inflation: the base is compounded from the base year to this one, with any negative year floored at zero growth, and then raised by the threshold — 50% by default. A decade of nominal drift therefore cannot be mistaken for prosperity.
- 04
A raise needs both halves of the test in the same year: the portfolio at or above that bar, and at least the cooldown’s years since the last step. When both hold, the scale is multiplied by one plus the step — 10% by default — and the year it stepped is recorded. The scale is never multiplied down, whatever the portfolio does afterwards.
- 05
This is the rule that reads least: the portfolio, the calendar, and the plan’s own inflation for the bar.
The other four size the portfolio’s share of the bill — the cards’ spending less freeCash, the year’s other money, with income and rent in and the deferrals, the after-tax contributions, payroll and income taxes, the roof, the debts and the cards the rule may not bend out.
On a plan that settles its income tax the following April, the income tax in that list is only the share the paycheck withheld, and last April’s settlement is netted beside it: handed over when the year owed, added back when it was a refund, because the carry is signed and an over-withholding household really does get the money back.
The ratchet never forms that figure at all: freeCash reaches it and is ignored. Spending plays exactly one part, as a gate — a year with $0.50 or less on the cards is skipped entirely, at a scale of 1 and with the rule’s memory untouched.
- 06
The scale is clamped between the floor and ceiling every rule shares — 60% and 150% of what the cards say by default. Steps compound rather than add: 1.10, 1.21, 1.331, 1.4641, and then a fifth step’s 1.61 is cut back to 1.50. The floor is inert here, because the scale starts at 1 and only ever climbs.
- 07
The scale multiplies only the expense cards the strategy named — naming none means all of the household’s own cards, and a card left out is simply a bill the raise does not reach.
It is applied to the rows before they are posted, so the ledger, the flow chart, the draw, the cash buffer, the tax on it and the Medicare surcharge two years later all quote the raised figure.
A working year, or a retired year outside the window, passes through at a scale of 1 with the memory kept, and a window that opens late anchors the base where it opens rather than at the retirement it skipped.
Keep in mind
Model limits
The test is made once a plan year, against last year’s closing balances: a portfolio that touched the bar in June and fell back by December is never seen, and a raise takes effect for a whole plan year rather than from a month.
The published rule ratchets a withdrawal rate inside a historical backtest; here the same three dials bend a live plan’s expense cards, the base is set once and is not re-based after a step, and the bar is grown by the plan’s own inflation so that nominal drift alone cannot earn a raise.
The model assumes the household both takes every raise it earns and holds it through whatever follows. It cannot know that a real household might decline one, or quietly cut in a bad decade — so a plan that fails under this rule fails at full spending.
The threshold, step and cooldown are dials, and their defaults are the published figures rather than a recommendation for any particular household. This is educational output, not advice.
This explanation documents the planning model. It is educational, not individualized tax, legal, Medicare, or investment advice.
Where it appears
Where to use it
See this concept in context, with a guide to each page and its controls.
The words behind it
Related financial terms
Plain-English definitions, with 2026 figures and worked examples, from the financial terms glossary.