See why the order of gains and losses matters when money is being withdrawn, even if long-run average returns are identical.
Average return does not describe every retirement path
During accumulation, two return sequences with the same compounded return can lead to similar outcomes when cash flows are identical. During retirement, withdrawals change the mathematics. A large loss early in retirement can force withdrawals from a depressed portfolio, leaving fewer assets to participate in a later recovery.
This is sequence-of-returns risk: the order of returns matters because money is entering or leaving the portfolio while prices change.
A useful way to study sequence-of-returns risk in retirement is to separate the calculation from the decision. The calculation answers a narrow question using stated inputs; the decision also depends on timing, liquidity, uncertainty, fees, taxes, contractual terms and what happens if an assumption is wrong. In this lesson, the central idea is see why the order of gains and losses matters when money is being withdrawn, even if long-run average returns are identical. Treat that statement as a framework to test rather than a one-off rule to memorise.
Retirement planning is best framed as a future cash-flow problem rather than a target portfolio number in isolation. Estimate the spending the household wants to support, separate essential from discretionary spending, and then subtract income expected from pensions, annuities, state benefits, work or other relatively stable sources. The remaining amount is what the investment portfolio or other flexible assets need to fund.
Inflation and longevity should be explicit assumptions. A nominal withdrawal that stays unchanged for thirty years can lose substantial purchasing power, while increasing withdrawals with inflation places greater pressure on the portfolio. No one knows the exact retirement horizon, so test several lengths rather than one assumed age. A plan that survives only under a short horizon or low inflation has less margin than the headline balance suggests.
Sequence-of-returns risk matters when withdrawals are occurring. Large losses early in retirement can be more damaging than the same losses later because assets are being sold at the same time the portfolio is depressed. A constant average-return model cannot show that path dependency. Stress-test a poor first five years, higher inflation, lower long-run returns and combinations of those conditions to see how much flexibility the spending plan has.
Review withdrawal rules, asset allocation and cash reserves together. Holding some near-term spending in less volatile assets can reduce the need to sell growth assets after a decline, but holding too much cash can increase inflation risk over a long retirement. The balance depends on reliable income, spending flexibility and the ability to reduce withdrawals temporarily. Treat any withdrawal percentage as a planning assumption that should be reviewed, not as a permanent guarantee.
A practical exercise is to build a baseline using today's best-known numbers, then change one important input at a time. Keep the other assumptions fixed so the effect is visible. After that, combine two adverse changes to see whether the conclusion is still robust. This method is deliberately simple: it does not predict the future, but it shows which variable has the greatest leverage and where a small amount of extra margin could materially improve resilience.
Finish by writing a short decision note: what was assumed, what evidence supports those assumptions, what could invalidate them, and when the calculation should be reviewed. That habit is especially useful for sequence-of-returns risk in retirement because the inputs can change while the original reasoning is easily forgotten. A model becomes more valuable when someone can return later, update the changed facts, and understand why the earlier conclusion moved.
A simple example
Imagine two portfolios start with the same balance and experience the same set of annual returns in opposite order. Without withdrawals, the final values can be similar because multiplication is commutative. Add a fixed withdrawal each year and the outcomes can diverge sharply. The portfolio that suffers losses first sells more units at low prices to fund spending, permanently reducing the capital base.
This is why a retirement model based only on one constant average return can hide an important risk.
Inflation-linked spending increases the pressure
If retirement spending rises with inflation, the currency amount withdrawn can increase over time even when markets are weak. High inflation and poor returns occurring together can therefore be particularly difficult. The portfolio may be falling while the spending requirement is rising.
A robust model separates nominal returns, inflation, and withdrawal rules rather than assuming a fixed nominal withdrawal forever unless that is the intended scenario.
Flexibility changes the result
A retiree who can temporarily reduce discretionary spending after a severe market decline has a different risk profile from someone whose spending is almost entirely fixed. Flexible withdrawals can reduce the amount sold after losses, preserving more capital for a recovery.
This does not mean every retiree should cut spending after a decline. It means withdrawal flexibility is a variable that can be modeled instead of pretending spending is perfectly rigid.
Cash reserves can change which assets are sold
Holding near-term spending in cash or lower-volatility assets can reduce the need to sell risky assets immediately after a market fall. But cash has its own inflation and opportunity-cost risks. The appropriate reserve size depends on spending, income sources, risk tolerance, and the rest of the portfolio.
A model can test one, two, or more years of planned spending held outside the growth portfolio and compare how different market sequences behave.
Guaranteed or stable income reduces portfolio dependence
Pensions, annuity income, social-security-type benefits, rental income, or part-time work can cover part of spending. The less spending that must come from the investment portfolio, the lower the withdrawal pressure. Those income streams can have their own inflation, credit, policy, or longevity characteristics, so they should be modeled separately.
Do not treat every income source as identical or perfectly guaranteed without checking its terms.
Fees compound the sequence problem
Ongoing fees reduce returns in every year, including recovery years. When the portfolio is already being drawn down, the same percentage fee is charged against a shrinking asset base but still reduces the money available for spending and future growth. Small annual differences can materially affect long horizons.
Use net-of-fee returns or subtract fees explicitly, but not both.
Test many paths, not one forecast
A useful retirement stress test includes poor early returns, poor late returns, high early inflation, long life, and combinations of these. Monte Carlo simulation is another technique, but its results depend heavily on the assumed return distribution, correlations, inflation process, and withdrawal rules. A precise-looking probability is not automatically accurate if the assumptions are weak.
Deterministic scenarios remain valuable because they show exactly what caused the result.