A retirement withdrawal changes the amount left to experience the next market return. That is why two portfolios with the same set of gains and losses can end at different balances when spending continues along the way. Follow the balance after each payment—not just the average return—to see sequence risk working.
This article is for education, not personal investment, retirement or tax advice. The calculations below are hypothetical and do not predict market returns or prescribe a withdrawal rate.
Follow four years, one line at a time
Start two imaginary portfolios at $1,000,000. Both withdraw $50,000 at the end of each year, after that year’s return. There are no contributions, taxes, fees or inflation adjustments. Returns apply to the whole remaining portfolio.
Portfolio A experiences −20%, −10%, +10% and +20%. Portfolio B experiences the same returns in reverse order. The calculation for every year is:
Year-end balance = opening balance × (1 + annual return) − withdrawal.
| Year | A: return | A: balance after withdrawal | B: return | B: balance after withdrawal |
|---|---|---|---|---|
| 1 | −20% | $750,000 | +20% | $1,150,000 |
| 2 | −10% | $625,000 | +10% | $1,215,000 |
| 3 | +10% | $637,500 | −10% | $1,043,500 |
| 4 | +20% | $715,000 | −20% | $784,800 |
The difference is $69,800. Both portfolios supplied $200,000 of withdrawals. The different closing balances come from the order of returns interacting with those withdrawals, not from one portfolio spending more.
For A’s second year, the arithmetic is $750,000 × 0.90 − $50,000 = $625,000. For B’s second year, it is $1,150,000 × 1.10 − $50,000 = $1,215,000. Writing those lines exposes the changing base on which the next return operates.
The thumbnail’s $340,000 and $1,600,000 labels are an illustration of contrasting outcomes. They are not the results of this four-year example or forecasts for two actual retirees.
Remove the withdrawals and compare again
With no cash moving in or out, the two sequences give the same final balance:
$1,000,000 × 0.80 × 0.90 × 1.10 × 1.20 = $950,400.
Changing the order of those multiplication factors does not change their product. This result assumes the same four portfolio returns, no other cash flows and no additional costs. It does not say that every real investment strategy behaves identically when markets arrive in a different order.
The arithmetic average of the four returns is zero. Yet the no-withdrawal balance falls by 4.96% over the period. Gains and losses compound; a zero arithmetic average does not imply a zero cumulative result.
With withdrawals, even matching the compounded return of the underlying return sequence is insufficient to match the final account balance. Each payment removes money before some of the remaining returns arrive.
Why selling during a decline can matter
Charles Schwab’s explanation of sequence risk describes the mechanism: raising a fixed amount of cash from a falling portfolio can require selling more investments, leaving fewer assets to participate in a later recovery. Its retirement illustrations use their own assumptions; the table here is separately calculated.
A simple share example makes the cash requirement visible. At $100 per share, a $5,000 withdrawal needs fifty shares. At $80, it needs 62.5 shares, assuming fractional sales, no fees and one holding. This is an illustration of the selling arithmetic, not a complete portfolio model.
The money withdrawn has funded spending; it has not simply vanished. But it is no longer invested in that account. If prices later recover, the recovery applies to the assets still held. An index returning to an earlier level does not therefore establish that a withdrawing investor’s account has returned to its earlier balance.
Change one assumption before trusting a projection
The four-year table keeps the annual withdrawal constant. Real spending may differ, and inflation can change how much cash buys the same goods. Vanguard’s retirement-savings guide identifies expenses, returns, inflation and unexpected costs as separate planning factors, while warning that past performance does not guarantee future results.
For an illustrative spreadsheet, test those assumptions separately:
- Move the withdrawal to the beginning of the year.
- Increase withdrawals annually by a stated hypothetical inflation rate.
- Add a one-time expense in a specified year.
- Include explicit investment costs rather than treating returns as cost-free.
These changes answer different questions. A beginning-of-year withdrawal leaves less money exposed to that year’s return. An inflation adjustment changes the cash requirement. A one-time bill changes the path even when the return sequence stays fixed.
Record the convention beside the calculation. Two projections with different cash-flow timing should not be presented as if they differ only in market assumptions.
Evaluate a response with its cost attached
Lowering a withdrawal preserves more money in the account under the same assumptions. The practical constraint is whether spending can actually fall. Identify essential bills separately from optional purchases before treating flexibility as an available solution.
A reserve can provide another source for a payment, but account for it within the whole plan. If a projection simply adds extra cash outside the starting portfolio, it has increased starting resources. That would not be a fair test of changing the allocation alone.
Schwab discusses reserves and reduced spending as possible responses, with investment-risk caveats. They do not ensure a market recovery or prevent losses. The article does not adopt its suggested allocation as a personal recommendation.
For a planning conversation, bring the opening assets, outside income, essential spending, optional spending and the assumptions behind each projection. Ask what would change after a poor first year and which payments could not be postponed.
The useful result is a plan you can interrogate year by year. You should be able to locate the withdrawal, reproduce the next balance and explain the assumptions. A favorable average return, on its own, cannot do that work.
