Variable Spending

By default every expense you entered is paid in full each year. Variable spending adds a yearly spending budget on top of them. A policy computes the total spending the portfolio can sustain. Anything above your entered expenses is spent as extra discretionary spending. When the budget falls short, only expenses you marked flexible are cut. Your entered expenses remain the plan’s floor.

A plan holds one or more spending phases. Each covers a stretch of years, and the default phase runs from retirement to the end of the plan. Years outside any phase have no budget. When two phases cover the same year, the first wins. Each phase’s budget strategy is either Amortized, which this page describes, or a Custom formula, which sizes the budget from a formula you write. A custom formula is how you express a guardrail rule such as Guyton-Klinger.

The amortized budget

The amortized budget is a variable percentage withdrawal (VPW) rule with taxes built into the arithmetic. Each year, in every projection and every Monte Carlo trial, FinZot recomputes the budget from that year’s actual state. The budget is your spendable wealth multiplied by a payment factor, plus the Social Security and pension income being paid that year, after an assumed tax. The payment factor depends on the years remaining, the expected return, and the spending tilt.

The rule has no memory on purpose. It does not remember last year’s budget or forecast future years. Next year re-amortizes whatever balance actually results, so a bad decade lowers spending instead of exhausting the plan, and a good one raises it. What a trial can still fail to sustain is the floor beneath the budget, meaning the expenses the budget may not cut. That is what the success rate measures (Meeting the budget).

The wealth the factor multiplies is the start-of-year balance. That is the number you would read off your own statement, restated for the year’s inflation so the budget is in that year’s dollars. The current year’s market move, income, and distributions do not enter on purpose. They reach the budget the following year, through the balance they leave behind.

Spendable wealth

The base is your invested accounts. Real estate and other property are never counted. Debt is not counted either. Nothing is subtracted for the loans you owe, and loan payments are paid outside the budget. Balances enter at their after-tax value, so a traditional dollar supports less spending than a Roth dollar.

Holding	Counts as
Traditional (pre-tax retirement)	75% of the balance. This assumes a 25% ordinary income rate on withdrawals.
Taxable (brokerage)	Cost basis, plus 85% of the unrealized gain. This assumes a 15% capital gains rate at sale.
Roth, HSA, 529	Full value.

The two rates are fixed assumptions of the strategy, not measurements. The actual tax on every withdrawal is still simulated in full (How a Year Works). A taxable account worth less than its cost basis is valued slightly above market, because the unrealized loss is a future tax shield.

The payment factor

The factor spreads spendable wealth over the years left to the end of the plan. It is the arithmetic of a loan payment run in reverse, drawing the portfolio down to zero over the horizon, or down to the legacy target. In the final year the factor spends the whole remaining balance. Two settings shape the schedule.

• Expected return. How fast the schedule assumes wealth grows. You state it as a stock and bond blend, 50/50 by default. FinZot prices the blend off the long-run averages of the catalog’s total-market stock and bond funds, after your return-rate overrides, and above inflation. It is an assumption only. It never changes how your accounts are invested, and the simulation uses whatever your holdings actually earn. Assume too much and spending starts high and is cut later. Assume too little and you underspend and leave more behind. The rate is also trimmed by an estimate of yearly tax drag on the taxable share of wealth, since a brokerage account’s distributions are taxed every year.
• Spending tilt. The planned slope of the schedule, after inflation. Level, or 0, is the default: the same real spending every year. Tilting toward spending more now plans each year slightly below the one before. Tilting the other way plans each year slightly higher. It changes the shape, not the total.

Income that raises the budget

Social Security and defined benefit pensions add to the budget dollar for dollar from the first year payments arrive. A lifelong inflation-adjusted stream needs no present value math. Its yearly amount simply joins the budget. Each stream is counted after an assumed tax. Social Security counts at about 79% of the benefit, because at most 85% of it is taxable and the assumed rate is 25%. Taxable pension payouts count at 75%. Tax-free payouts count in full. A FERS-style supplement that ends at 62 is left out, because counting it would schedule a known future cut.

Nothing else raises the budget directly. Benefits not yet claimed count as zero until they start. Wages, rental income, and other income are never counted. The valuation only errs toward caution, so a stream starting is always a raise and never a scheduled cut. This is why a phase normally starts at retirement. During working years, when a salary funds most spending, the budget would sit far below what you can afford. Uncounted income still lands in the portfolio, so it raises future budgets through next year’s balance.

The legacy target

A plan-level Legacy target makes the schedule draw down to a remainder instead of zero. You state the target in today’s dollars as a net amount, meaning what your heirs keep after the plan’s assumed heir tax rate. That rate applies only to the traditional share of the portfolio, so a legacy held in traditional dollars reserves more balance than the same legacy in Roth, and an all-Roth portfolio reserves exactly the target. Each year FinZot sets aside the target’s present value, discounted at the phase’s expected return, before the payment factor applies. The reserve ranks above variable spending but below your expense floors. A bad year cuts the discretionary layer before it touches the reserve, and only mandatory expenses can invade it. Only invested assets count toward the target. Property passes to heirs on top of it. The target only takes effect while an amortized phase is active.

Meeting the budget

The budget is a total, and your entered expenses are funded first.

• Budget above expenses. The surplus is spent as fully discretionary spending on top of them.
• Budget below expenses. Flexible expenses are reduced toward their Spending Floor in cut order, lowest cut order first (expenses and healthcare). Mandatory expenses are always paid in full. When mandatory expenses plus floors alone exceed the budget, they are still paid. Next year simply re-amortizes from the resulting balance.
• Outside the budget entirely. Debt payments and property costs, such as mortgage, taxes, and maintenance, are always paid in full, regardless of the budget. A loan whose principal never comes due inside the plan is therefore never reserved for, so the schedule can spend the portfolio to zero with the balance still outstanding. A plan that ends that way is reported as balance-sheet insolvent rather than as a success (projections & monte carlo).

Your entered expenses are therefore the plan’s floor. What you mark mandatory, plus the floor kept on each flexible expense, is the spending the plan must sustain. The Monte Carlo success rate measures that floor, not the budget. A variable spending plan’s rate is not comparable to a fixed spending plan’s, as covered in Projections & Monte Carlo.

Reading the budget

The budget is recomputed fresh each year with no smoothing. The year something changes, the budget moves in a step, not a glide. A spouse’s Social Security starting late in the plan raises the budget that year, dollar for dollar. The valuation’s tax rates are fixed assumptions while the simulated taxes are not, so the spending line can tilt or kink rather than holding perfectly level. The effect is mild when the assumptions sit close to a plan’s real rates and larger when they do not. Every step has a cause visible in that year’s detail.

In today’s dollars the budget also carries the year’s own inflation. The wealth base is restated to the year’s close, while today’s dollars index spending at the year’s start, so in a trial a year of 8% inflation reads about 5% richer than one of 3%. The dollars actually spent are unaffected.

A single path is therefore the wrong place to judge this strategy. The detailed projection shows how and why the budget moved along one path. Judge the strategy itself by the sustained-spending range across the Monte Carlo trials, which shows the budgets the rule produced over the whole range of market histories.