Guardrails, and what they cost
Decision rules turn 'you might run out of money' into 'you might have to spend less for a while'. It is a much better trade. It is not a free one, and this site will not pretend otherwise.
Jonathan Guyton and William Klinger published a set of rules in 2006 that do something the 4% rule refuses to: they let the retiree respond. The version this site implements has three parts.
The inflation rule
Normally your spending rises with inflation each year. But after a year where the portfolio lost money, and only if you are already withdrawing at more than your starting rate, you skip the raise. You are not cutting the check — you are just not increasing it. The rule also caps the raise at 6% in any one year, which is a brake that engages only when inflation is doing real damage.
The capital preservation rule
If your withdrawal rate drifts more than 20% above where it started — a plan begun at 5% now running at 6% — you cut spending by 10%. This is suspended in the last fifteen years of the plan, because there is no sense slashing spending at 88 to protect a balance you have three years left to spend.
The prosperity rule
The one nobody talks about, and the one that matters most for the argument on this site. If your withdrawal rate falls more than 20% below where it started — because the market did well and you didn't notice — you give yourself a 10% raise. It is a rule whose entire job is to stop you from underspending, and it fires far more often than the cut does.
What it buys
On this site's data, a flat 4% plan fails four of the sixty-nine complete 30-year windows since 1928. A 5% plan with these rules fails none of them — a full percentage point more spending, every year, with a better survival record.
That is not magic. It is what happens when you stop requiring the plan to survive 1966 without ever acknowledging that 1966 is happening.
What it costs
Here is the part most guardrails advocacy skips. Take that same 5% plan and run it through 1966 specifically. It survives — and at its worst point, real spending is roughly two thirds below where it started.
Two thirds. Four separate 10% cuts, plus a decade of skipped inflation raises during the worst inflation in modern American history, compounding into a standard of living that is a fraction of the one the plan opened with. The plan did not fail. The person living inside it had a very different retirement than the one they budgeted for.
A plan that survives by starving you has not really survived. This is why the spending chart sits directly under the portfolio chart, at the same size, and why it is not optional.
So the honest framing is not "guardrails make you safe". It is that guardrails change the currency the risk is paid in. Instead of a small chance of catastrophe, you accept a larger chance of a real, and occasionally severe, reduction in spending. For almost everyone that is the better deal — you can live on less, and you cannot live on nothing — but you should agree to it with your eyes open, having seen the bottom of the band and asked whether you could actually live there.
The rules are a starting point, not scripture
The 20% width, the 10% cut, the 6% inflation cap, the fifteen-year suspension: these are Guyton and Klinger's published figures, and they are defaults on this site rather than commandments. Someone with a large Social Security floor can afford a wider guardrail and deeper cuts, because their spending band has a floor under it the model doesn't know about. Someone whose entire budget is discretionary-free cannot. Change them and re-run it.