What the 4% rule actually says

Ask the internet how much you can spend in retirement and you will get the same answer everywhere: withdraw 4% of your portfolio in the first year, raise that dollar amount with inflation every year after, and a 30-year retirement probably will not outlive the money. It is a genuinely useful rule of thumb. It is also narrower than the way it usually gets quoted.

The claim, precisely

In 1994, financial planner William Bengen replayed historical US stock and bond returns through every 30-year retirement in the record and asked: what is the highest initial withdrawal rate that would have survived even the worst starting year? For portfolios of roughly half to three-quarters stocks, the answer came out at about 4%. The later Trinity study reframed the same idea as a table of success rates across withdrawal rates and stock/bond mixes, and the shorthand stuck.

Note what the rule actually prescribes:

1. 4% is the first year only. Every later withdrawal is that dollar amount adjusted for inflation - you are not re-taking 4% of whatever the balance happens to be.
2. Spending is fixed in real terms. The rule models a retiree who never cuts back in a crash and never splurges in a boom.
3. Success means not hitting zero by year 30. Ending with $3 and ending with $3 million both count as success.

In dollars, with a $1,000,000 portfolio and 3% inflation, the schedule looks like this:

Year	Withdrawal (nominal)	In today's dollars
1	$40,000	$40,000
5	$45,020	$40,000
10	$52,191	$40,000
20	$70,140	$40,000
30	$94,263	$40,000

The nominal number nearly triples; the purchasing power never moves. That is the whole trick of the rule - and why quoting "4% a year" without the inflation adjustment misstates it.

What it quietly assumes

• US history, and only US history. The dataset behind the rule comes from the single best-performing major stock market of the twentieth century.
• Exactly 30 years. Retire at 40 and the horizon is closer to 50 years; the safe rate drifts down as the horizon stretches.
• No fees, no taxes. A 1% advisory fee is, to a first approximation, a 1% permanent headwind the studies did not model.
• A robotic retiree. Real people cut spending in bad markets - which is exactly the flexibility that rescues most borderline plans.

The 4% rule is a summary of one dataset under rigid assumptions - a fine starting point, and a poor place to stop.

Test the number, don't inherit it

Your retirement is not a 50/50 index portfolio with a robotic withdrawal schedule - it has a mortgage that ends, Social Security that starts, a spending pattern that is anything but flat. Model the plan you actually have in the planner, turn on the historical backtest, and see which starting years break it. Then let "Solve for…" find the spending level that hits the success rate you can sleep on - that number is your withdrawal rule, and it did not come from a rule of thumb.