My expenses, cut back in bad years is the middle choice in Expenses → Spending strategy: a guardrail, in the Guyton-Klinger sense. Real retirees don't spend a fixed amount no matter what markets do. They ease off in bad stretches and loosen up in good ones. This models that flexing instead of pretending spending is rigid.
It sits second on a ladder of three: your expenses as entered never react, this nudges your flexible spending when your withdrawal rate strays, and a share of your portfolio recomputes the number from the balance outright. It's the option for someone who wants to keep a plan and still answer to the market.
How it decides
Each simulated year compares your current draw rate, meaning this year's budget less Social Security and other guaranteed income, over the portfolio, against two bands. With Rates rise with age on, the bands you enter describe your first spending year and rise as the years left shrink, on the same curve the withdrawal-rate check uses: a 5.5% cut rate for a thirty-year plan is about 6.9% with twenty years left and 8.5% with fifteen. A fixed rate at every age trims a plan in its eighties for a draw that a short horizon carries easily; Guyton and Klinger's own version of this is simply no cuts in the final fifteen years. The two bands:
- Above the upper guardrail (spending is stretching the portfolio) → trim the flexible slice of spending by the cut step (e.g. 10%).
- Below the lower guardrail (you clearly have room) → raise it by the raise step.
- A floor and ceiling bound how far cumulative cuts or raises can drift from your original plan, and cuts can heal (reversible) when markets recover.
What it may touch
Only the flexible portion of Recurring expenses that you declare: dining out, travel, hobbies. The rest of Recurring, plus Housing and Medical, is treated as essential and is never trimmed. At 100% the whole Recurring budget flexes; lower it if part of your budget is untouchable.
A mortgage payment is a contract, so the rule never touches it: it is charged on its own schedule, at the amount your note says, and it is unchanged by a cut, by a raise, and by a first death if you model one. The home goal seekers protect each candidate's rent and ownership costs the same way, as rent gives way to ownership.
Paying extra principal is one-time spending, not a raise or a cut: money your plan has committed this year is set aside before the rule reads your balance, and the rule never multiplies it. It can still change the year's decision, because a smaller balance is a higher withdrawal rate. That is the rule working, not the payment being trimmed.
That field sits outside the rule's own knobs because it isn't only this rule's. The same split is the essential floor a share rule can never spend below. It is one budget fact, read two ways.
You can see the split year by year in Lifetime spending projection, beside the Expenses panels: hover any year and only the Recurring line moves with the rule. Housing, Medical and Medicare stand at the figures your plan schedules, however deeply the rule has cut.
The presets
The four presets under Adjust the rule are one axis: how hard the rule protects the portfolio versus your spending.
- Cuts sooner, deeper reacts at a lower rate and cuts to a lower floor. It protects the portfolio, and your spending swings more.
- Balanced is the standard setting, and the app's long-standing defaults.
- Cuts later, less reacts at a higher rate and holds a higher floor. It protects your spending, and leaves more risk on the plan.
- Custom is your own six numbers.
Each preset states how low it would take your flexible spending, in dollars per month, rather than as a floor percentage, because a percentage of a slice of a budget is not a thing anyone can picture. These labels used to read Conservative / Moderate / Aggressive, which had it backwards where it counted: "Conservative" cuts sooner and deeper, since what it protects is the portfolio, while a conservative person expects their spending protected.
A higher success rate bought by spending cuts is only a good deal if you can see the cost. Reduction below entered expenses is that cost, in monthly dollars, and the line beside it says how often the rule actually fires and when the first reduction typically lands. The shortfall line under it is a different fact and is deliberately worded as one: "Some expenses go unfunded in 8.0% of outcomes." That is what the portfolio could not pay, not what this rule chose to take off. Trimming is supposed to make that number smaller; it is never the same number.
Why you can't add it to a share rule
You pick one rule, not a rule plus a modifier. A share of your portfolio already recalculates your spending from your balance every year, so layering this on top would react twice to the same signal. You could no longer tell which one moved your spending. Choosing the share rule is choosing not to trim.
To see what you gave up, run Explore → Compare spending strategies: the cut-back row is scored with the knobs you set here, whichever rule you currently have picked, and with the flexible portion you can set beside that table.
Following the rule for real
The chart and the tiles above are the simulation. To actually run this rule on your own balances, name this plan as the one you follow (the Scenario menu, the calendar icon beside a saved plan) and use the spending check-in on the Overview.
It applies exactly the arithmetic described here, once a year, to what you actually have. Two consequences worth knowing before you start:
- The rate it tests is your current budget over your portfolio, so after a trim the numerator is the trimmed figure, not the plan. That is what stops a rule from cutting a second time for a stretch it already answered.
- Money your plan commits to a one-time expense this year (a house bought for cash, a car) is set aside before the rate is taken. It is not a cushion under the rate, and the card says how much was set aside when it happened.
- Your recorded budget carries into the projection on this page. Editing the thresholds above keeps it; only Restart this rule puts it back to the plan as typed.