Guyton-Klinger Guardrails

The problem it addresses

The 4% rule assumes a retiree keeps withdrawing the same inflation-adjusted amount for thirty years regardless of what the portfolio does. That assumption is what makes the rule testable, and it is also what makes it conservative: the safe rate has to survive the worst historical sequence while the retiree does nothing in response.

Guardrail systems, introduced by Jonathan Guyton and William Klinger in the mid-2000s, start from the observation that nobody actually behaves that way. If the rules governing the response are set in advance rather than improvised, the improved outcome can be modelled — and the starting rate can be higher.

The decision rules

The system is usually described as a set of rules applied each year. The two that do most of the work are the guardrails themselves.

Current rate = This year's withdrawal ÷ Current portfolio value

Guardrails are set either side of the initial rate,
conventionally at ±20% of it:

  Initial rate 5%  →  lower guardrail 4%   upper guardrail 6%

• Capital preservation rule. If the current rate rises above the upper guardrail — the portfolio has fallen, so the same withdrawal is now a larger share of it — the withdrawal is cut, conventionally by 10%.
• Prosperity rule. If the current rate falls below the lower guardrail — the portfolio has grown faster than spending — the withdrawal is increased, conventionally by 10%.
• Inflation rule. Withdrawals rise with inflation in most years, but the increase is skipped after a year in which the portfolio lost value. The raise is not made up later.
• Portfolio management rule. Withdrawals are taken from whichever asset class has performed best, so that depleted assets are not sold at their lows.

The capital preservation rule is commonly suspended in the final years of the plan, on the reasoning that a portfolio only needs to last as long as the retirement does.

Why flexibility buys a higher rate

The damage in a bad sequence comes from withdrawing a fixed amount out of a shrinking portfolio, which forces the sale of more and more units at low prices. Guardrails cut the withdrawal precisely when that is happening, so fewer units are sold at the bottom and more remain to participate in any recovery.

Because the worst historical sequences are the binding constraint on any safe withdrawal rate, softening them raises the rate that survives. This is why guardrail research reports sustainable starting rates above the 4% associated with fixed withdrawals — the rate is higher because the retiree is doing something the fixed rule assumes they will not.

What it costs

• Income variability. Spending can fall by 10% after a bad year, and the skipped inflation increase is permanent rather than deferred. Over a long retirement with several poor periods, the compounding of skipped raises can be substantial in real terms.
• Complexity. Four rules applied annually, with a recalculated withdrawal rate, is materially harder to follow than one inflation adjustment — and the system only works if it is followed.
• Floor risk. The rules do not distinguish between discretionary and essential spending. A retiree whose budget is mostly fixed costs may find a triggered cut difficult to actually implement.
• Parameter sensitivity. The guardrail width, the size of the adjustment, and the starting rate are all choices. Results reported in the research apply to the specific parameters tested, not to any variation of them.

Related approaches

• Simple percentage withdrawal. Taking a fixed percentage of the current balance each year. The portfolio can never be exhausted, but income swings with the market and can fall sharply.
• Variable percentage withdrawal. A percentage that rises with age, drawn from a table, so the portfolio is deliberately spent down over a known horizon.
• Floor-and-ceiling rules. Guardrails with hard limits, so that spending is never allowed below or above set amounts regardless of what the rules indicate.
• Bucketing. Segregating a few years of spending into cash, which addresses the same sequence problem through asset location rather than through spending rules.