Your spending plan is what the portfolio exists to fund. How you model it — regular vs one-off, inflation-linked vs fixed, phased over time — directly shapes every outcome metric.
Go to the Spending pattern tab. Essential expenses are fixed or near-fixed outgoings expected to continue throughout retirement. Discretionary expenses are lifestyle spending that may taper. Click a category pill from the left panel to add it, then set the amount and frequency. Tick Infl. to link the expense to CPI each year.
Every expense card has Starts and Ends controls. Starts can be Now, At retirement, or a Custom age. Ends can be At retirement, End of plan, a Custom age, or a fixed number of times. This lets you model expenses that stop (mortgage paid off), start later (care costs from age 80), or only run for a limited period (school fees for 5 years).
Retirement spending often follows a smile pattern — higher in active early retirement, lower in the quieter middle years, then higher again for care costs. Model this by adding multiple expense lines with different Custom age ranges rather than a single average figure. The engine handles each independently.
The engine treats both categories identically in the simulation. The distinction is for you — when you look at a stressed scenario, essential expenses are the ones you cannot cut. Discretionary ones show you where the plan has flex. Use + Custom... to add any expense category not in the preset list.
Each month, the engine deducts active expenses from the cash buffer. If the buffer is insufficient, it triggers a portfolio withdrawal to replenish it. The Summary tab’s Expense & Income Timeline shows every expense as a horizontal band across your age range — colour indicates magnitude (red = highest, green = lowest), shape indicates frequency (bar = monthly, dot = annual or one-off). Use this to spot gaps, overlaps, or unexpected durations.
Any metric that averages spending across the full plan is distorted by large one-off items. A £100k gift at age 75 adds £4,200/year to a 24-year average — suggesting higher ongoing spending than actually exists. This is why the app never shows a single spending number and always exposes the full timeline instead.
Every £500/month reduction in spending reduces portfolio withdrawals and lets it compound longer. The effect is larger in bad scenarios (P10) than good ones (P90) — because in bad scenarios the portfolio is already under pressure. The cumulative success probability curve shifts up across the entire horizon.
A £100k expense at age 62 (2 years into retirement) is far more damaging than the same expense at age 72, because it depletes the portfolio while returns are still needed to fund the remaining horizon. Model large expected expenses at their actual ages — do not smooth them into the monthly figure.