How much can you safely spend in retirement?

Last updated: August 8, 2026

It is the question every retirement plan exists to answer, and most tools answer it badly: with a rule of thumb, a flat average return, or a plan that quietly ignores taxes. The honest answer is not a percentage you look up — it is a measurement of your actual situation: the accounts you hold, the taxes those accounts will generate, when your income streams start, and how markets behave in the order they actually behave in.

This page explains what that measurement involves, so you can judge any tool — including ours — that offers to give you a number.

Why rules of thumb fall short

The famous shortcut is to withdraw 4% of your starting portfolio, adjusted for inflation each year. It came from real research, and as a first sanity check it is not crazy. But it answers a question that isn't quite yours. It was derived for one portfolio mix, one 30-year horizon, and — critically — for pre-tax withdrawals with no modeling of where the money sits. Two households with the same $1 million can have very different spendable incomes if one holds mostly Roth assets and the other holds mostly a traditional IRA that the IRS will eventually force into taxable income.

A rule of thumb also can't see your timing. Retiring at 62 with Social Security claimed at 70 means eight years where the portfolio carries the entire load — the years when market losses hurt most. A flat percentage has no way to notice.

What actually determines your number

A real measurement has to account for, at minimum:

• Which accounts hold the money — not just the total. Taxable, tax-deferred, and Roth dollars are not the same dollars. Withdrawal order across those buckets changes your tax bill every single year, and the tax bill changes what is left to spend.
• Taxes as an output, not a guess. Federal brackets, capital-gains stacking, the taxation of Social Security benefits, state tax, and Medicare IRMAA surcharges all move with your withdrawals. A plan that applies one flat "effective rate" is guessing at the largest expense of your retirement.
• Required minimum distributions. From your mid-70s the IRS dictates a minimum draw from tax-deferred accounts whether you need the money or not — often pushing you into higher brackets exactly when you'd prefer to coast. RMDs reshape the plan's back half.
• When income streams start. Social Security claiming age, pensions, and any part-time income change how hard the portfolio works in each specific year — and the early, unprotected years dominate the risk.
• The order of market returns, not the average. Two futures with the same average return can end in very different places, because losses early in retirement — when withdrawals are amplifying them — do damage that later gains cannot undo. This is sequence-of-returns risk, and it is the reason an "average 7% growth" projection flatters almost every plan.
• How long the money must last. Planning to life expectancy means a coin flip's chance of outliving the plan. A measured answer should let you see the cost of planning to 95 or 100 instead, and say so out loud.

How a measured answer is built

The mechanics are conceptually simple, even though the arithmetic is heavy. You simulate the plan year by year: each simulated year applies market returns to the portfolio, pays that year's spending, computes that year's actual taxes given which accounts the money came from, takes any required distributions, and adds any income that has started. Then you repeat the whole exercise across many possible market futures — including bad ones — and ask: what is the highest starting spend where the plan still survives nearly all of them?

That highest-surviving spend is your number. It is a property of your situation, discovered by search, not a percentage applied to a balance.

There are two respectable ways to generate the market futures: statistical simulation (Monte Carlo) and historical replay — walking your plan through every actual market sequence on record, including retirements that began in 1929, 1966, and 2000. Each has blind spots, which is why a careful tool shows both and tells you how many of the tested futures your plan survived, rather than a single false-precision verdict.

What to look for in any tool that gives you a number

Whatever software you use — ours or anyone's — the same standards apply:

• It should compute taxes year by year, from your actual account types, rather than asking you to guess an effective rate.
• It should model sequence risk — thousands of market paths or full historical replay, not one smooth average line.
• It should state its survival standard. "Your plan survives ~90% of tested markets" is a claim you can interrogate. A green checkmark is not.
• It should disclose every assumption — returns, inflation, longevity, tax law year — and let you see how the answer moves when they do.
• It should measure, not sell. Be wary of a "free plan" whose real product is routing you to an advisor's book of business, or a tool that needs your data on its servers to work. The math above runs perfectly well on your own computer.

Measure yours

ZBee™ is a Mac app built to do exactly this measurement: it connects to your accounts read-only (or takes manual entry), computes taxes, RMDs, and Social Security year by year, stress-tests the result across a thousand simulated markets and every real market sequence since 1928, and reports the most you can sustainably spend — in today's dollars, after tax, with every assumption disclosed. Everything computes locally; your holdings never leave your Mac.

The whole app is free to explore on realistic sample data, no account required — download ZBee for macOS and see how the measurement works before entering a single number of your own. The full method, source by source, is published in the methodology.

ZBee is a measurement and planning tool, not financial, tax, or legal advice. Projections are hypothetical estimates that will not match actual results. See the Disclaimer and Terms of Use.