Fixed, variable percentage, or guardrails: three withdrawal strategies on the same 1,000 markets

By DecisionSheet · Updated September 22, 2026 · 2026 tax figures

A withdrawal rule decides how much to take from a portfolio each year. The fixed rule takes the same inflation-adjusted amount regardless of markets. The other two adjust: a variable-percentage rule takes a share of whatever the portfolio is worth, and a guardrails rule, after Guyton and Klinger, cuts or raises spending when the withdrawal rate drifts too far from where it started.

The comparison below holds everything else fixed: a $1,500,000 portfolio, 80% stocks and 20% bonds, $60,000 of first-year spending (4%), 40 years. Only the rule changes. Survival rises from 91.3% (fixed) to 98.1% (variable) to 100.0% (guardrails). The median ending balance falls in the same order.

The three rules as the model applies them

All amounts are in today's dollars. Each year's withdrawal comes out at the start of the year, before that year's return.

• Fixed real withdraws $60,000 every year until the money runs out.
• Variable percentage withdraws 4% of the balance at the start of each year, but never less than $42,000 or more than $78,000 (70% and 130% of the starting spending). Because of the floor, it can still run out.
• Guardrails starts at $60,000. Each year it divides last year's spending by the current balance. Above 4.8% (20% over the 4% target), it cuts spending by 10%. Below 3.2% (20% under), it raises spending by 10%. Otherwise spending stays where it was. Each change carries into the following years. There is no floor on the cuts and no ceiling on the raises.

The 20% triggers and 10% adjustments are those of the "capital preservation" and "prosperity" rules in Guyton and Klinger's 2006 paper. The paper also stopped the cuts in the last 15 years of the planned horizon, skipped some inflation raises after a losing year, and set an order for which assets to sell. The calculator implements none of those. What it calls guardrails is the two triggers alone, applied every year.

Results over 1,000 simulated markets

Open each: fixed real · variable percentage · guardrails. The three scenarios differ only in the rule.

The calculator seeds its random draws from the market assumptions alone, and each run has its own stream of annual returns. So the three links run on the same 1,000 simulated markets: run by run and year by year, every rule meets the same returns, and only the withdrawals differ.

The figures still depend on which 1,000 markets were drawn. DecisionSheet simulated the same assumptions under 10 other seeds (runs of 1,001 to 1,010, since the run count is part of the seed). Across those and the linked draw, survival ranged from 88.3% to 91.7% for fixed and 97.0% to 98.7% for variable, and guardrails lasted in every run on every draw. The survival order and the reverse order of median balances held on every draw. The fixed rule's 10th percentile was the lowest of the three each time. The two adjusting rules' 10th percentiles stayed above $900k, but they sit too close together to rank.

The fan chart for the guardrails link shows the 10th, 50th and 90th percentile balances of its 1,000 runs, year by year.

What each rule trades away

A success rate counts a run as a success if money remains, whatever was spent along the way. The adjusting rules differ from the fixed rule only in how much they spend each year, and the calculator reports balances, not spending. So the 1,000-run results cannot show how often the adjusting rules cut or how deep the cuts went.

A single deterministic market can show it. Set both returns to 0% after inflation and both volatilities to zero, and every run follows the same path. Each year's fall in the balance is exactly that year's spending. The model's balances give these withdrawals (rounded to $100):

Open each: fixed · variable · guardrails.

• Fixed real keeps spending whole and runs out in year 25.
• Variable percentage is at its $42,000 floor by year
  10. Because the floor stops further cuts, it runs out in year
• Guardrails makes its first cut in year 6 and 16 in all. It lasts the full 40 years, ending with $234,868, by spending $11,100 in the last year, 19% of where it started.

The calculator's tiles for that last case show 100% survival. Nothing on the page shows the spending behind it.

The trade runs the other way in strong markets. The variable rule stops rising at $78,000. The guardrails rule has no ceiling and keeps raising. The fixed rule never raises spending at all.

So the choice among the three is between different risks. The fixed rule risks running out, the variable rule risks a spending cut of up to 30% and, once at its floor, running out anyway, and this guardrails rule risks cuts with no set limit. The model reports only the first of those risks directly. The other two have to be read from paths like the one above.

Assumptions and limits

• Returns are assumptions, not forecasts: the calculator's defaults of 6.8% real for stocks (16.0% volatility) and 2.2% for bonds (6.0%). They are compound growth rates, converted to average annual returns by adding half the variance before each year's independent normal draw. Stocks and bonds have a fixed 0.15 correlation, and the portfolio is rebalanced every year.
• The flat market is a device for reading spending, not a forecast.
• Everything is in real dollars. No fees, taxes, Social Security or other income. There is no input for any of them.
• The guardrails rule measures its triggers against the safe-withdrawal-rate input. Here that equals the first-year rate, because the portfolio is exactly 25 times spending. On a portfolio that starts above its target the two differ.
• A run with a few dollars left counts as a success, as does a run whose spending has been cut to a fraction of the plan.

Method and sources

The model is calculateFire in the FIRE & safe withdrawal calculator, with the calculator's default allocation and returns. The data file checks that the three rules' year-1 balances match exactly, which confirms that the rules start on the same draws.

• Guyton and Klinger, "Decision Rules and Maximum Initial Withdrawal Rates", Journal of Financial Planning, March 2006: the capital preservation rule (a 10% cut when the withdrawal rate rises more than 20% above the initial rate, lifted 15 years before the end of the plan) and the prosperity rule (a 10% raise when it falls more than 20% below), along with the paper's withdrawal and portfolio management rules.
• "Multi-period Returns", Macro-Investment Analysis (Stanford): the approximation that the geometric return is the arithmetic mean less half the variance, which the simulator uses in reverse.

Open this scenario in the calculator

All figures on this page come from the FIRE Simulator calculator. Change any input there and the numbers update.