How to Calculate Your FI Number

The FI number is the amount of invested assets where, in theory, you never have to add another dollar of income to keep living the way you already live. The shortcut everyone quotes is "25 times your annual spending." That number isn't arbitrary: it's simple algebra applied to a withdrawal rate.

Where 25x comes from

If you assume you can withdraw 4% of a portfolio each year and have it last through a long retirement, then the portfolio needed to support any given spending level is:

FI number = Annual spending ÷ Withdrawal rate

At a 4% withdrawal rate, that's the same as annual spending × 25, because 1 ÷ 0.04 = 25. The 4% figure itself comes from two related pieces of research: William Bengen's 1994 study of historical U.S. market returns, and the 1998 "Trinity study" by Cooley, Hubbard, and Walz, which is generally regarded as the source of the "safe withdrawal rate" framing that popularized the idea (see bogleheads.org's summary of both). Neither study set out to invent a rule of thumb for FIRE. The FIRE community, most visibly through Mr. Money Mustache's savings-rate math, adopted 4% as a planning shortcut, and 25x fell out of it as the inverse.

That's the whole calculation. The hard part isn't the arithmetic; it's getting the "annual spending" input right, and knowing what the 25x figure does and doesn't guarantee.

Building the spending number that feeds the calculator

The 25x math is only as good as the annual spending figure you plug into it, and most people's first guess is off in one of two predictable directions.

The first mistake is using today's spending unchanged. Some current costs disappear in retirement (commuting, work clothes, retirement account contributions themselves), while others show up for the first time or grow (full-price health insurance before Medicare eligibility at 65, more travel, more time for hobbies that cost money). A number copied straight from a current budget tends to overstate work-related costs and understate the free-time costs that replace them.

The second mistake is treating retirement spending as one flat number for the rest of your life. Spending in the first decade after leaving work (often more active, more travel-heavy) frequently looks different from spending in your 70s or 80s, when travel tends to slow and healthcare costs tend to rise. A single 25x figure smooths over that shape, which is fine as a planning shortcut but worth remembering isn't the full picture.

A reasonably reliable approach: track actual spending for six to twelve months if you're within a few years of your target date, since your current spending pattern is the best predictor of your near-term retirement pattern once a few known categories (commuting, contributions) are removed and a few known additions (health insurance, more leisure spending) are added back in.

The assumptions baked into 25x

Treat these as the fine print, because each one can move your real number meaningfully:

• It assumes a 4% withdrawal rate is appropriate for your situation. That rate came from research modeling roughly 30-year retirement horizons. A 35-year-old retiring at 40 has a much longer horizon than a 65-year-old retiring at 65, which argues for a lower initial withdrawal rate (and a bigger FI number) for very early retirees. See the 4% rule, explained for the full derivation and caveats.
• It assumes your spending is roughly flat and predictable in retirement. A lumpy spending pattern (a paid-off mortgage that ends housing costs at 62, kids finishing college, health costs rising over time) doesn't fit a single flat number well. Multi-phase retirement spending needs a multi-phase FI number, not one 25x figure.
• It doesn't separately account for one-time or irregular costs. A major home repair, a wedding, replacing a car: these aren't "annual spending," and if they're not planned for separately, they'll eat into a portfolio sized only for routine expenses.
• It's based on historical U.S. market data. Bengen's data set and the Trinity study both look backward at periods that happened to occur, not a guaranteed distribution of future returns. Past success rates in backtests describe what happened; they don't promise what will happen to a portfolio starting today.
• It ignores taxes on the withdrawal itself in its simplest form. If your spending number is after-tax, you may need to withdraw more than that figure from a taxable or tax-deferred account, which raises your real target above a clean 25x.

Decision framework: setting your own number

Work through these in order, since each step changes the number before it:

1. Start with a real spending estimate, not your current budget. Build it from your actual likely retirement lifestyle: housing, food, insurance, travel, hobbies. If you don't have a full budget yet, start with how to build a budget.
2. Choose a withdrawal rate that matches your time horizon, not a default 4%. A 30-year horizon supports something close to 4%. A 40- or 50-year horizon (someone retiring in their 30s or 40s) is more often modeled at 3% to 3.5%, which raises the multiplier from 25x toward 29x–33x.
3. Add a buffer for irregular costs. Many planners use 5% to 15% on top of the base number, sized to how lumpy your expected expenses are.
4. Check the number against a tax-aware withdrawal plan, especially if most of your savings sit in a traditional 401(k) or IRA, where withdrawals are taxed as ordinary income.
5. Revisit annually. Your FI number isn't a one-time calculation. Spending changes, family situations change, and health insurance costs in particular tend to move faster than general inflation.

Worked example

Say your realistic retirement spending is $55,000 a year in today's dollars, and you're planning to retire at 45, a roughly 45-year horizon if you expect to live to 90.

• At a 4% withdrawal rate (25x): $55,000 × 25 = $1,375,000.
• Given the long horizon, you decide 3.5% is a more conservative starting point (about 28.6x): $55,000 × 28.6 ≈ $1,573,000.
• You add a 10% buffer for irregular costs and rising health insurance premiums before Medicare eligibility at 65: $1,573,000 × 1.10 ≈ $1,730,000.

That's a $355,000 gap between the naive 25x shortcut and a horizon- and buffer-adjusted target, a meaningful difference in how many more years of saving it represents, and a good illustration of why 25x is a starting point, not a finished plan.

Limits and exceptions

The 25x/4% framework assumes a relatively passive, buy-and-hold portfolio drawing down over a long but bounded horizon. It fits less well if:

• You expect Social Security, a pension, or rental income to cover part of your spending later. In that case your portfolio only needs to bridge the years before those income sources start, which is a smaller and different calculation (see bridge account planning for early retirement).
• You're planning Coast FIRE rather than a full stop, where the relevant number is what your current balance will grow into by a target age, not what you need today.
• Your spending is expected to change substantially and predictably (a mortgage payoff, kids leaving the house), in which case a single flat FI number oversimplifies a plan that's really two or three numbers stitched together.

One number, or several

Most people benefit from calculating more than one FI number rather than treating 25x as a single verdict. A "lean" number based on bare-essentials spending tells you the earliest point at which quitting a job becomes structurally possible, even if uncomfortable. A "base" number, built from your realistic expected spending, is the one worth actually planning around. A "fat" number, built from a more generous spending level, shows what a few extra years of saving buys in terms of cushion. Seeing all three side by side does more to inform a real decision than a single output ever will; the Lean, Fat, Coast, and Barista FIRE framing exists precisely because "your FI number" isn't one-size-fits-all.

What this number can't tell you

An FI number tells you how much you need; it says nothing about how the sequence of returns in your specific retirement will actually play out, which matters more than most people expect. Two people who retire with the same portfolio size can have very different outcomes depending on what the market does in their first five years, a risk covered in sequence-of-returns risk. The FI number is the map. It isn't the terrain.