Real retirees don't keep spending the same amount while their portfolio falls. They travel less, delay the car, eat out less often. Adaptive Spending models that: when withdrawals climb past a threshold you set, the simulation cuts back the spending you've marked discretionary - by a percentage you choose, not all of it.

It's off by default, and turning it on will usually raise your chance of success - because spending less in bad years genuinely helps. The question worth asking is whether you'd actually make those cuts.

How to set it up

1. Set Essential % on your recurring expense rows. This is the substrate: it's the only thing that tells the rule what may be cut. A row left blank counts as fully essential, so a plan with nothing marked has nothing to trim and the rule will do nothing at all.
2. Turn on Adaptive spending on the Recurring Expenses page.
3. Click the tier summary to adjust the thresholds. The panel shows what each tier would trim in dollars from today's spending.

How the tiers work

Each tier is a pair: a withdrawal rate that triggers it, and how much of your discretionary spending to cut when it does. The defaults are 5% / cut 10% and 7% / cut 25%.

The tiers replace each other rather than stacking. Crossing the second threshold cuts 25%, not 10% plus 25%. The second threshold must be higher than the first.

The withdrawal rate is measured against your investable portfolio - taxable, tax-deferred and tax-free accounts. Cash and health savings accounts aren't counted, because the cash you hold for a bad year is exactly what a bad year spends first, and including it would hide the pressure the rule exists to detect.

One-time expenses don't count toward the rate. A new roof or a car is a one-off, not a rate of drawdown, so it's left out of the trigger - though it's still fully funded and still withdrawn. Without this, a single $60,000 expense against a $1.5 million portfolio adds six percentage points to that year's rate on its own, which is enough to clear the second threshold from a comfortable starting point and cut your discretionary spending in a year the plan is fine. The spending pressure isn't ignored, only deferred to when it's real: the money has genuinely left the portfolio, so the smaller balance raises the rate from the following year onward.

Easing the thresholds as the plan nears its end

A fixed threshold is blind to how much of your plan is left. Drawing 6% a year would drain a 30-year retirement, but over your final five years it's entirely comfortable - the money only has to last five more years. A rule that keeps asking for cuts at 6% late in the plan is asking you to underspend for no reason.

The optional Ease the thresholds as the plan nears its end setting fixes that. Both thresholds start at exactly the rates you entered and rise as your remaining years shrink. On a 20-year plan, a 5% first tier is 5% at the start, about 6.2% with 15 years to go, and about 8.7% with 10 years to go.

Two details are worth knowing:

• It's anchored to your plan, not a standard one. The rates you type are the rates used at the start of your retirement, whether you're planning for 20 years or 45. Your number is never quietly reinterpreted.
• The easing is capped at double. The arithmetic alone would say a 5% threshold should become 21% with five years left. We don't go there, because your planning age is an assumption rather than a known date - if you plan to 95 and live to 100, spending as though the money genuinely runs out at 95 is how you arrive at 96 with nothing left. The cap keeps the rule working through the final years instead of switching itself off.

It only ever eases, never tightens. If you're not yet retired, the thresholds stay exactly as you entered them until retirement begins.

Why the cut is smaller than the percentage suggests

Cuts apply only to the discretionary portion, never to what you've marked essential. If your spending is 90% essential, a 25% rule changes your total spending by about 2.5%. That's correct, and it's why the settings panel shows dollars: an aggressive-looking rule on a mostly-essential plan is a rule that does almost nothing, and it's better to see that while configuring it than to wonder later why your results barely moved.

Reading the result

After a run you'll see something like "In your last run, 34% of simulations trimmed at least once, typically in 2 years." That's a distribution across all the simulated futures, not a schedule.

We don't tell you which years you'd cut in, and that's deliberate. Every simulated future is tested on its own, so different futures trim in different years - a single answer would mean picking one future and presenting it as the forecast. That's a specific-looking number with nothing behind it, and it's the kind of false precision this tool tries to avoid. The honest statement is how often it happened and for how long, not when.

For the same reason, the year-by-year table can show no cuts at all while a third of your simulations trimmed. The table follows one representative path; the percentage describes all of them.

Limits worth knowing

• Taxes are not re-cut. A simulation that trims its spending still pays the tax bill calculated before the trim, so it's very slightly over-taxed. The error runs one way - it makes results marginally more conservative, never less.
• One-time expenses sit outside the rule entirely. They have no Essential % field, so they're never trimmed, and as above they don't count toward the withdrawal rate that triggers a cut.
• Nothing carries between years. Each year is tested on its own: a simulation that trims one year and recovers the next simply stops trimming. There's no memory of having cut before.

Adaptive Spending and the funded ratio

These two are a natural pair, and they use the same Essential % input. The funded ratio's essentials figure tells you where you'd land if you cut all discretionary spending - the floor. Adaptive Spending tells you how often you'd actually need to move toward it. One is a deterministic bound, the other a probability.