QWE AI Academy Can you actually retire on what you’ve saved?
You’ve got a number in your head – maybe $900k, maybe $1.8M – and you’re wondering if it’s enough to stop working without running out of money. That’s the exact question the 4 percent rule in retirement was built to answer with one back-of-the-envelope calculation.
Short version people remember: take 4% of the portfolio in year one, then raise that dollar amount with inflation every year after. US market history says a balanced mix often lasted 30+ years that way. The longer version has teeth – and traps glossy explainers skip.
What the 4 percent rule actually says
SAFEMAX sat near 4.15% for tax-advantaged accounts – not a clean 4%. William Bengen’s October 1994 Journal of Financial Planning piece tested rolling 30-year retirements on US large-cap stocks and intermediate Treasuries from 1926. First-year withdrawal around that level, then pure CPI adjustments, never wiped the portfolio in under roughly 33 years in that record. The street “4% rule” is a rounded shortcut.
Two mechanics matter more than the headline:
- Year 1 only: withdraw 4% of the starting balance.
- Every year after: drop the percentage. Apply a COLA to last year’s dollars from prior-year CPI – like Social Security. You are not taking 4% of the new, possibly smaller balance each January.
Same math, flipped: the rule of 25. Need $50,000 a year from the portfolio? Target about $50,000 × 25 = $1.25 million. The 1998 Trinity study later backed the same ballpark success rates for stock-heavy mixes at 3-4% initial rates.
| Portfolio at retirement | Year-1 withdrawal (4%) | Implied annual need covered |
|---|---|---|
| $750,000 | $30,000 | $30,000 |
| $1,200,000 | $48,000 | $48,000 |
| $1,500,000 | $60,000 | $60,000 |
| $2,500,000 | $100,000 | $100,000 |
That’s portfolio spending only. Guaranteed income (Social Security, pension, rentals) sits on top and can shrink the nest egg you actually need – calculate the portfolio slice separately, then stack the rest.
Practical setup: run your own numbers in 10 minutes
Skip the abstract. Use your real figures.
- Estimate annual portfolio need. Living costs minus guaranteed income. Include healthcare and ugly one-time hits.
- Multiply by 25 (or divide by 0.04). Rough nest-egg target under classic 4% assumptions for a ~30-year horizon.
- Check asset mix. Original work leaned on real equity exposure (often illustrated around 50/50 to 75/25 stocks/bonds) with annual rebalancing. All-cash or all-bonds changes the math.
- Stress the first five years. In a sheet, apply a 20-30% drop right after retirement while you still take the planned withdrawal. Path scares you? Lower the starting rate or hold 1-3 years of expenses in cash/short bonds.
- Write the year-2 rule down. “Next year I take last year’s dollar amount × (1 + CPI).” Not “4% of whatever is left.”
Example: You want $48,000 from investments after other income. $48,000 × 25 = $1.2 million target. At $1.2M you start with $48,000. CPI runs 2.5% year one → year two is $49,200 – even if the market is down 15%.
Here’s a quieter question the spreadsheet won’t ask: if the math “works” only when markets behave and you never need a new roof the same year your portfolio is down 25%, is that still “enough” – or just enough on a good path?
Advanced moves when 4% feels too blunt
Bengen never sold a one-size forever rule. His 2025 book A Richer Retirement (as of that release) broadened asset classes – small/mid/micro caps, international – and put a diversified “Universal Safemax” near 4.7%. He’s also said many retirees today can support something in the mid-5% range depending on valuations, inflation, and personal factors, still watching the worst historical paths.
Pro tip: Pair a higher initial rate with guardrails. Portfolio drops hard? Cut the COLA or the dollar withdrawal for a year or two. Flexible methods (constant percentage of current balance, endowment-style averages, guardrails) support higher starting rates than a rigid inflation-adjusted floor – you trade that for more variable spending.
Forward-looking base case for fixed real spending and 90% success over 30 years: about 3.9% starting withdrawal for moderate equity mixes, per Morningstar’s 2025/2026 research (as of that update; up from 3.7% the prior year). Lower than pure historical SAFEMAX because it uses expected future returns and valuations, not 20th-century US averages. Flexible approaches in the same work push initial rates higher.
Actually – if you’re in FIRE and planning 40-50 years, treat classic 4% as aggressive. Longer horizons raise the odds of a worse sequence; for multi-decade early retirement, ~3-3.5% is the cautious band unless you’ll cut spending or earn later.
Honest limitations (the parts that break the simple story)
No fees. No taxes. That was the lab setup. Real expense ratios, advisor fees, and ordinary-income tax on traditional IRA/401(k) withdrawals cut what you can spend. A 4% gross line can land closer to 3-something% after costs and tax – bracket and account type decide how far.
Turns out US 20th-century returns were the friendly case. Wade Pfau’s international work (Journal of Financial Planning / related papers) found a 4% real rate historically “safe” in only a small minority of developed markets – roughly 4 of 17 under optimistic assumptions. Many countries cleared far lower sustainable rates.
Sequence-of-returns risk is the killer: bad markets plus withdrawals in the first 5-10 years permanently lower the base later recoveries work with. High early inflation compounds it. Bengen’s toughest path was the late-1960s retiree who got hit by stocks and inflation together – fixed COLAs on a falling portfolio accelerate depletion even when long-run average returns look fine on paper.
RMDs at 73 or 75 (SECURE 2.0: born 1951-1959 vs 1960+) can force larger distributions than your planned COLA path. The strategy doesn’t auto-fail – but “forget the percentage” collides with IRS rules. Bracket-plan and consider Roth conversions in the years before RMDs start.
Is a future worse than 1968-1982 possible? History doesn’t hand you a ceiling. That’s why this is a planning starting point, not a guarantee.
FAQ
Does the 4% rule include Social Security?
No. Portfolio withdrawal first; add Social Security (and any pension) after. Delaying SS often lifts the guaranteed floor so you can take less – or more flexibly – from investments.
I have $800k. Can I retire next year under the 4% rule?
Strict 4% → $32,000 year one from the portfolio. If that plus expected SS and other income covers your real budget, you’re in classic-rule territory for a ~30-year horizon with equity risk – still run the first-five-years crash test. Need $45,000 from the portfolio alone? $800k misses the 25× target ($1.125M). You’d start near 5.6%, more aggressive than the original SAFEMAX.
Is 4% still safe in 2026, or should I use Morningstar’s 3.9% or Bengen’s higher numbers?
Different questions. Bengen’s updated historical SAFEMAX (~4.7% with broader diversification, as discussed around the 2025 book) asks: what never failed in the past US record under stated assumptions? Morningstar’s ~3.9% base case (as of the 2025/2026 retirement-income work) asks: given today’s forward return and inflation assumptions, what fixed real starting rate still hits ~90% success over 30 years? Neither knows your SS claim age, health, home equity, or willingness to cut spending. A lot of people land between 3.5% and 5% after stress-testing, then adjust yearly. One percentage is a first draft.
Next action: blank sheet. Current investable assets. Estimated annual portfolio need. Need × 25. Note the gap. Then simulate a 25% drop in year 1 while you still take the planned withdrawal for five years – watch the ending balance. That exercise beats another generic article.
Deeper sources: Bengen’s plain-English explanation, the 1994 paper reprint, and current Morningstar safe-withdrawal research for a forward-looking check. Pfau’s international findings are on SSRN.