Retire at 45 Needs 30× Expenses, Not 25× — Sequence Risk Closes the Gap
The 25× rule and the 4% withdrawal rate that produces it come from the 1998 Trinity Study, which tested historical US returns against 30-year retirements. Almost every popular FIRE article repeats the rule. Almost none point out that 30 years is where the rule was tested, and most early retirees plan for 45 or 50.
We ran 5,000 Monte Carlo paths through the FIRE calculator at $50K of annual spending, $1.25M starting portfolio (the textbook 25× number), and a 60/40 stock/bond mix calibrated to historical returns. Here’s where the simple rule cracks.
What 25× actually buys you at different horizons
| Retirement length | Survival rate at 4% withdrawal | What it means |
|---|---|---|
| 30 years (retire at 65) | 95% | The rule’s original validation |
| 40 years (retire at 55) | 86% | Most “early” retirees land here |
| 45 years (retire at 50) | 81% | Lean FIRE territory |
| 50 years (retire at 45) | 76% | Aggressive FIRE — 1-in-4 paths run out |
The 4% rule isn’t broken. It’s just being applied to a problem it wasn’t tested on. Stretching a 30-year answer to a 50-year horizon turns “high probability” into “coin flip with extra steps.”
Sequence-of-returns risk does the actual damage
The reason 25× starts failing at longer horizons isn’t that average returns get worse — they don’t. It’s that the longer the retirement, the higher the chance that a bad opening sequence catches you.
We pulled the failed paths from our 50-year, 25× simulation. The pattern is identical across them:
- Years 1-5 average below the long-run mean (often 1-3% real, sometimes negative)
- The portfolio drops below 80% of starting value by year 4-6
- It never recovers, even when years 10-30 average 7%+ real
- Forced withdrawals at depressed prices crystallise the losses
Bengen’s original 4% rule paper and every follow-up Trinity Study revision circle around this point: average return tells you almost nothing about portfolio survival. The order of returns dominates.
Three fixes, ranked by how much they actually help
We re-ran the 50-year scenario with one change at a time. Same spending, same starting age (45):
| Fix | Survival rate | Cost |
|---|---|---|
| Baseline (25× expenses, 4% withdrawal) | 76% | — |
| Save to 30× expenses, then retire | 88% | 4-5 extra working years |
| Add $15K/year part-time income years 1-5 | 89% | 8-10 hours/week for 5 years |
| Drop withdrawal to 3.3% (the “Bengen fix”) | 91% | Smaller lifestyle, same nest egg |
Two of the three fixes don’t require saving more. The BaristaFIRE option — the one most often dismissed as a FIRE community euphemism — produces nearly the same survival lift as working four extra years, for a fraction of the time commitment.
The mechanism: $15K/year of part-time income covers about 30% of $50K of expenses. That cuts effective withdrawal rate during the danger window from 4% to ~2.8%. By year six the portfolio has had time to grow past starting value, and full withdrawal becomes safe.
This is the math hiding under the marketing. BaristaFIRE isn’t “settling for less FIRE.” It’s sequence-risk insurance you bought with hours instead of dollars.
Where the simple math breaks down further
The 25× rule has three more cracks the headline never mentions:
- Healthcare before Medicare. A US household retiring at 50 buys 15 years of private health insurance before Medicare. ACA premiums for two people commonly run $14-22K/year — that’s 28-44% on top of the $50K spending baseline. Plug $70K into the 25× rule and the target jumps from $1.25M to $1.75M. Survival on the original $1.25M drops further still.
- Tax bracket changes. Roth conversions, capital gains, RMDs after 73 — the 4% rule treats withdrawals as gross. Net of taxes, you typically need 110-120% of the headline number depending on account mix.
- Lifestyle inflation in retirement. Most FIRE planners assume flat real spending. Real US retirees spend more in years 1-15 (travel, projects), then drop, then rise again with healthcare in years 25+. A flat-spending model misses both ends.
None of this invalidates aiming for FIRE. It just means the 25× number is a starting estimate, not the answer.
Where this framework doesn’t hold
Three scenarios where everything above misses:
- Pension or guaranteed income. Anyone with a defined-benefit pension, military retirement, or future Social Security replacing 30%+ of spending is doing a different problem. The portfolio only needs to cover the gap, and sequence risk shrinks proportionally.
- Variable spending. Some early retirees plan to drop spending to $30K when markets are down and run $60K when markets are up. The “guardrail” approach (Guyton-Klinger) has different math entirely — survival rates well above 95% even at 50-year horizons, because the portfolio is never forced to sell at depressed prices for fixed obligations.
- Geographic arbitrage. Retire to a low-cost country and the same $1.25M might cover $80K of effective lifestyle. The 25× rule still applies; the inputs just change.
What to actually do
If you’re under 50 and modeling FIRE:
- Calculate your number with 30× expenses, not 25×. The extra 5× is sequence-risk insurance.
- Run the calculator with your real numbers, including healthcare premiums if pre-Medicare.
- Stress-test with the BaristaFIRE toggle — even nominal part-time income has outsized survival lift.
- Build the plan around a 3.3-3.5% initial withdrawal, not 4%. Adjust upward only after the first 5-7 years confirm you didn’t draw a bad sequence.
If you’re already saving aggressively but the 30× number feels demoralising: that’s the right reaction to honest math. The fix isn’t a calculator with friendlier assumptions; it’s a plan that accepts BaristaFIRE or part-time consulting as the structural feature it actually is.
Open the FIRE Calculator → and run your specific scenario. The output isn’t just the success number — it’s the distribution of what the bad 24% of paths actually look like, so you know the failure mode before you commit to the plan.