Rules describing how much a retiree can withdraw each year are widely quoted and rarely examined. Each is the output of a specific test with specific assumptions attached.
The method is historical simulation
The standard approach takes a portfolio, applies a withdrawal in the first year, increases that withdrawal with inflation each subsequent year, and checks whether the money lasted.
The test is then repeated starting in every available historical year, so the portfolio is run through booms, crashes and inflationary periods in the order they actually occurred.
The reported rate is the highest starting withdrawal that survived every tested period, which makes it a worst-case figure from one dataset rather than a typical outcome.
Assumptions do most of the work
The result depends heavily on the assumed retirement length, because a plan that must last considerably longer cannot support the same starting withdrawal.
Asset allocation matters just as much. Portfolios holding too little in growth assets fail to outpace inflation over long horizons, while portfolios holding too much are more exposed to poor early years.
Costs are frequently omitted from the underlying tests, so investment charges and platform fees come directly out of whatever margin the rule appeared to provide.
Inflation adjustment is the demanding part
The rule assumes the withdrawal rises with prices every year, which means the real income stays constant while the nominal amount grows steadily.
Sustained high inflation is more damaging than a market fall in this framework, because it raises the required withdrawal permanently rather than reducing the portfolio once.
Retirees whose spending falls in later years, as many do, face a lighter requirement than the rule assumes, which is one reason strict application can be unnecessarily conservative.
The geography of the data matters
Most widely cited figures come from a single market with an unusually strong long-run record, and that market's history is not representative of investing outcomes generally.
Studies using a broader set of countries produce lower sustainable rates, because they include markets that suffered prolonged disruption and slower recovery.
A rule derived from the most fortunate available history should be read as an upper bound rather than a neutral estimate.
Fixed rules ignore available information
The framework assumes a retiree sets a withdrawal on day one and never reacts, which nobody actually does. Real spending responds to how the portfolio is performing.
Approaches that adjust withdrawals within bands, or recalculate annually against the current balance, support higher average income at the cost of a variable one.
Choosing between them is a question of whether stable income or a larger total matters more, and it is the kind of decision where individual circumstances warrant professional guidance.