Why this one input dominates everything

Compounding turns small differences in the assumed rate into enormous differences in the outcome. Take $500,000, leave it alone for 20 years, and watch what the return rate for FIRE projections does to the “answer”:

Same money, same 20 years — a 3.6x spread between the cautious answer and the bull-run answer. Every downstream decision (when to coast, when to quit, whether you are “behind”) inherits whatever you typed into that one field. That is why it deserves more thought than the thirty seconds most people give it.

The bull-run trap

Here is the pattern we see constantly in FIRE forums: someone opens their brokerage statement, sees that their accounts returned 12–14% a year over the past eight years, and plugs that into their projection. The math is done honestly — but the input is a rear-view mirror. A strong recent stretch tells you what markets did, not what they will do; if anything, long periods of above-average returns tend to be followed by leaner ones, because part of those gains came from valuations getting more expensive rather than from businesses earning more.

Over the last century, US large-cap stocks have returned roughly 10% a year in nominal terms — commonly cited as about 7% after inflation. Those long-run figures already include the crashes of 1929, the 1970s, 2000, and 2008. A projection built on 12–14% is not “using historical data” — it is assuming the best stretch of the historical record repeats on schedule, with your retirement date riding on it.

The practical damage runs in a specific direction: an inflated return rate makes your projected portfolio look bigger, which makes early retirement look closer, which invites you to quit sooner than the money justifies. The error only reveals itself years later, when it is expensive to fix.

Real vs nominal: the other silent error

Before choosing the number, be sure which kind of number you are choosing. A nominal return includes inflation; a real return strips it out. The rule that keeps your math honest: if your spending target is in today’s dollars (it almost always is), your return assumption must be real. Using a 10% nominal return against today’s-dollar expenses quietly ignores decades of inflation and overstates your future purchasing power by 40% or more. We cover the mechanics — and how two people can get different answers from the same inputs — in our guide to real vs nominal returns.

The numbers worth using

There is no single correct rate, but there is a defensible band. For a diversified, stock-heavy portfolio (a global index fund, an S&P 500 fund, a target-date fund early in its glide path):

• Base case: 4–5% real (about 7% nominal with 2–3% inflation). Below the long-run historical average, which is the point — it leaves room for an ordinary decade rather than requiring a great one.
• Optimistic case: 6–7% real. Roughly the century-long historical average for all-stock portfolios. Reasonable as an upside scenario; risky as the plan.
• Conservative case: 3% real. What a weak-but-not-catastrophic decade looks like, and closer to what balanced portfolios with significant bonds should assume anyway.

If your portfolio holds meaningful bonds or cash, shade every number down — a 60/40 portfolio has historically earned notably less than all-stock. And if you plan to de-risk as you approach retirement (most people should), your assumed rate should step down with the allocation, not stay pinned at the all-equity figure. The US SEC’s Investor.gov compound interest calculator is a neutral place to sanity-check any growth figure against your own.

A worked example: the $6M that was really $4M

A couple in their mid-40s has $3,000,000 invested and six years until their planned retirement at 50. Their accounts returned 12% a year over the last eight years, so that is what they project:

The bull-run assumption says almost $6 million; the sober, inflation-adjusted assumption says $4 million of purchasing power. That is not a rounding difference — at a 4% withdrawal rate it is the gap between $237,000 and $161,000 a year of sustainable spending. If their retirement budget only works at the top number, they do not have a plan; they have a market forecast with a plan attached. The 4% rule can only be as honest as the portfolio number you feed it.

The three-rate stress test

Instead of arguing about the “right” number, run your plan three times — it takes two minutes in any decent tool:

1. Run it at 5% real. This is your plan. If the retirement date at this rate is acceptable, you are done deciding.
2. Run it at 3% real. This is your downside. You do not need to like this date — you need to be able to live with it, because you cannot control which decade you get.
3. Run it at 7% real. This is your upside — the case where you get to retire earlier than planned, not the case you plan around.

The discipline this enforces is simple: set your date with the middle number, survive on the low number, and treat the high number as a bonus. You can run all three scenarios in our Coast FIRE Calculator or the full FIRE calculator, and test raw growth math in the compound interest calculator — all of them let you set the return and inflation directly, alongside the rest of our free financial calculators.

One last nuance for the withdrawal phase: once you start spending from the portfolio, the order of returns matters as much as the average — a crash in the first years of retirement does damage that a later boom cannot fully repair. That is sequence-of-returns risk, and it is the reason retirement-phase planning uses safe withdrawal rates rather than simple average-return math. During the coast phase, while you are not withdrawing, a bad stretch only delays you; after you retire, it can wound you.

Frequently asked questions

Last updated: July 2026 · Historical figures are commonly cited long-run US averages; examples use round numbers for clarity.