How Much Do I Need to Retire: The Real Number

A 35-year-old who expects to spend 40,000 a year in retirement needs a pot of roughly 1,000,000 to stop working at 65. That's the 25x rule in one sentence. Hitting that target takes around 820 a month for 30 years at a 7% real return — and that monthly number is usually more useful than the headline figure, because it's the one you can do something about.

The real question isn't how much do I need to retire. It's how the answer shifts when you change your spending, your start age, your expected returns, or how long the pot has to last. The retirement calculator handles those inputs; this post explains the framework underneath, so the output makes sense rather than feeling like a guess. All figures are in your home currency — the maths works the same in dollars, pounds, euros, or rupees.

What is the retirement number?

The retirement number is the size of the investment pot that can sustain your spending for the rest of your life once you stop earning. It assumes the pot is invested in a diversified portfolio and that you draw down a portion each year. The most common shortcut is the 25x rule: take your expected annual retirement spending and multiply by 25.

That multiple is the inverse of a 4% withdrawal rate. The rule withdraws 4% in year one, then rises with inflation each year after, and historical market data suggests the pot has a high chance of lasting around 30 years. It's a starting estimate, not a promise — and it moves around as assumptions about returns, inflation, and longevity change.

Why this matters

Retirement is the single largest financial goal most people ever fund. It's bigger than buying a home, because a home pays for itself in shelter. A retirement pot has to replace decades of income with nothing fresh coming in. OECD data shows that average life expectancy at 65 across member countries is now roughly 19 years for men and 22 for women, and both figures have been rising for decades. A pot that runs out at 85 is one that runs out too early.

Two effects make your starting age the dominant variable. The first is compounding — returns earned in year one keep earning returns for every subsequent year. The second is contribution leverage: starting ten years earlier roughly halves the monthly contribution required to reach the same target. The retirement number itself doesn't change with age, but the path to it does.

How the retirement number is calculated

The calculation has two layers. The first sizes the target pot. The second works out the contributions needed to reach it.

The target pot uses the inverse of your chosen withdrawal rate:

Target pot = Annual retirement spending / Safe withdrawal rate

At a 4% withdrawal rate, that simplifies to annual spending multiplied by 25. At 3.5% it becomes 28.6 times. At 5% it drops to 20 times. The withdrawal rate is the lever most affected by expected returns and how long the pot has to last.

The contribution layer uses the future value of an annuity formula:

FV = PMT × [((1 + r)^n − 1) / r]

Where:

FV — the target pot you're saving toward
PMT — the contribution per period (usually monthly)
r — the periodic real return (annual real return divided by 12 for monthly)
n — the number of periods until retirement, in months

Rearranged to solve for the contribution: PMT = FV ÷ [((1 + r)^n − 1) / r]. The retirement calculator handles both layers in one step and lets you stress-test the withdrawal rate and return assumption side by side.

A worked example with real numbers

Maya is 35. Her annual retirement spending, in today's money, is 40,000. She wants to retire at 65 and assumes a long-run real return of 7%, which sits within the range that equity-heavy portfolios have historically produced after inflation.

Step one sizes the pot. Using the 25x rule:

Target pot = 40,000 × 25 = 1,000,000

Step two works out the monthly contribution. Maya has 30 years, or 360 months, until retirement. The monthly real return is 7% ÷ 12, or about 0.5833%.

Annuity factor = ((1.005833)^360 − 1) / 0.005833 ≈ 1,220

Monthly contribution = 1,000,000 / 1,220 ≈ 820

So Maya contributes around 820 a month for 30 years at a 7% real return to hit a 1,000,000 pot. If returns disappoint and she actually earns 6% real, the annuity factor falls to roughly 1,004 and her required contribution jumps to about 996 a month — a 21% increase from a single percentage point of returns.

This is why the assumption matters. The target pot is stable; the contribution rate that gets her there is highly sensitive to expected returns. Isolating the growth side in a compound interest calculator shows how a single percentage point of return reshapes the end pot. Folding the withdrawal rate back in gives you the full retirement projection.

How to use the retirement calculator

The retirement calculator takes five inputs. Current age and target retirement age set the time horizon. Annual retirement spending sets the target pot through the withdrawal rate you select. Current savings reduces the gap that fresh contributions need to close. Expected real return drives the compounding.

The output shows the target pot, the monthly contribution required to close any gap, and a year-by-year projection of the pot's growth. The chart is usually the most revealing part, because compounding looks underwhelming in the early years and only becomes obvious in the final third of the timeline.

Stress-testing is the point. Running it at 7% and then 5%, at a 4% withdrawal rate and then 3.5%, and at retirement ages of 65 and then 60, maps out a range of scenarios. The spread between those scenarios is the planning range, which tells you far more than any single point estimate.

How much do I need to retire by age

The 25-year-old early starter

A 25-year-old aiming at the same 1,000,000 target has 480 months to compound. At 7% real, the annuity factor comes out to about 2,625, putting the required monthly contribution at roughly 381. Starting a decade earlier than Maya more than halves the monthly bill. That's the compounding effect in concrete terms — not a slogan, an arithmetic fact.

The 35-year-old steady saver

Maya's case. About 820 a month at 7% real, or 996 at 6% real, for 30 years to a 1,000,000 pot supporting 40,000 a year. This shape — a 30-year runway and a defined retirement age — is the most common case in projections, because it captures the middle of working life. The contribution is meaningful but not crushing.

The 45-year-old late starter

A 45-year-old starting from zero with 20 years of runway faces a steeper climb. The annuity factor at 7% real over 240 months is around 521, giving a required contribution near 1,920 a month. That's more than double the 35-year-old's bill for the same target. Late starters typically lower the spending target, work a few years longer, or accept a higher withdrawal rate. All three reduce the contribution, and the calculator shows the trade-off explicitly.

The dual-earner household

For two earners with combined spending, the calculation runs once on the combined pot rather than twice on individual pots. Splitting contributions across two earners often improves tax efficiency through tax-sheltered retirement accounts, but the underlying maths doesn't change.

The variable-income earner

Self-employed earners and contractors face the same target with irregular contributions. The framework still works if the average annual contribution is roughly stable — variance around that average matters less than the long-run rate.

Frequent oversights

Anchoring on current income instead of retirement spending. The 25x rule multiplies projected retirement spending, not pre-retirement income. Conflating the two inflates the target by 15 to 30 percent and overstates the contribution needed.
Using nominal returns instead of real returns. A 7% nominal return with 3% inflation is a 4% real return, and the difference between the two compounded over 30 years is enormous. The calculation only works if returns and the withdrawal rate are both expressed in real terms.
Ignoring sequence-of-returns risk. The 4% rule has held historically, but a poor returns sequence in the first decade of retirement can materially raise the failure rate. A 3.5% withdrawal rate is more defensive, at the cost of a larger target pot.
Treating the number as fixed for life. A retirement spending estimate from age 30 isn't the same as one from age 55. Rerunning the calculation every few years catches the drift between assumptions and reality before it becomes a problem.
Skipping the longevity question. A pot sized for 30 years and a retiree who lives 35 years is a pot that runs out. OECD longevity figures have trended upward for decades; ignoring that trend produces a target that's chronically too small.

The retirement number sits inside a wider planning stack. Three adjacent tools cover the parts it leaves out.

Compound interest calculator — isolates the growth side of the equation. Useful for sanity-checking the return assumption underpinning your projection.
FIRE calculator — runs the same framework with aggressive savings rates and earlier retirement ages, showing the trade-off between contribution intensity and time to financial independence.
Savings rate calculator — works backward from your current income and contributions to show what proportion you're actually saving, which is the lever most directly under your control.

Frequently asked questions

How much do I need to retire at 65?

The most common shortcut is 25 times your expected annual retirement spending. For annual spending of 40,000, that produces a 1,000,000 target. For 60,000, it produces 1,500,000. The multiple comes from a 4% withdrawal rate, which historical market data suggests sustains a diversified portfolio for around 30 years. It's a starting estimate, not a guarantee. Real outcomes depend on the actual sequence of investment returns, on inflation, on longevity, and on any guaranteed income — state pensions, workplace pensions, social security — that reduces the spending your portfolio itself has to cover.

Is the 4% rule still reliable?

The 4% rule traces back to William Bengen's 1990s research using US market data, and subsequent peer-reviewed work has tested it across other markets and time periods. The headline finding has held up reasonably well in most jurisdictions, but research from the CFA Institute and others suggests modest revisions. A starting rate of 3.5% improves longevity in lower-expected-return environments. Dynamic withdrawal strategies — ones that adjust spending in poor return years — tend to outperform fixed-percentage rules. The 4% figure remains a useful benchmark; treating it as a ceiling rather than a floor is the more defensive read.

How much should a 30-year-old have saved for retirement?

There's no single benchmark, but a common framework is one times current annual income by age 30, three times by 40, six times by 50, and ten times or more by 60. These are heuristics rather than targets. A 30-year-old earning 50,000 with 50,000 saved sits at the benchmark, but the same person with higher expected retirement spending faces a bigger gap than that multiple suggests. Running the actual calculation — with current savings, expected spending, and the years left until retirement — gives a much better answer than any rule of thumb.

What return assumption should I use?

Long-run real returns for equity-heavy portfolios have historically landed in the 5–7% range, depending on the market and time period. Mixed portfolios with a meaningful bond allocation produce real returns more like 3–5%. Using 7% as a planning assumption is reasonable for an equity-heavy portfolio, but it creates a false floor if returns disappoint. Running the calculation at two assumptions — a central estimate, and one a full percentage point lower — produces a planning range instead of a single point estimate, which is much harder to be wrong about.

Does inflation change the retirement number?

The retirement number expressed in today's money doesn't change with inflation. The nominal pot required at the retirement date does. If your target spending today is 40,000 and inflation averages 3% over 30 years, the same lifestyle costs about 97,000 a year by year 30 and requires a nominal pot of roughly 2,400,000. Most retirement calculations are run in real terms, which strips inflation out of both the return and the target. Working in real terms is cleaner, but it helps to understand that the nominal numbers you'll see at retirement will look much larger than today's figures.

What if I already have a workplace or state pension?

Any guaranteed income in retirement reduces what the investment pot has to fund on its own. If your annual spending target is 40,000 and you expect 10,000 a year from a state or workplace pension, the portfolio only needs to cover 30,000. Under the 25x rule, that drops the target from 1,000,000 to 750,000. The retirement calculator handles this through the existing savings and additional income inputs. Including pension income often materially lowers the contribution the calculation requires.

Sources and methodology

The calculations in this article and the linked tool rely on the future value of an annuity formula and the inverse-withdrawal-rate framework. Both are standard finance constructs documented in peer-reviewed sources rather than proprietary methods.

OECD Pensions at a Glance provides comparative data on retirement ages, replacement rates, and life expectancy across member countries. Used here for the longevity figures.
CFA Institute Research Foundation publishes peer-reviewed work on safe withdrawal rates, including extensions and stress tests of the original Bengen 4% rule across global markets and asset allocations.

The 25x rule traces back to William P. Bengen's 1994 paper in the Journal of Financial Planning, which established 4% as a historically sustainable initial withdrawal rate for a balanced US portfolio over a 30-year retirement. Subsequent academic work has tested the rule across other markets and time periods.

The bottom line

The retirement number resolves to one equation and three honest assumptions: projected annual spending, an expected real return, and a withdrawal rate. For a 35-year-old targeting 40,000 of annual spending, that produces a 1,000,000 pot and contributions of around 820 a month over 30 years at a 7% real return. Stress-testing those assumptions — instead of treating any one of them as fixed — is what separates a useful retirement plan from a paper one. The retirement calculator makes that stress-testing visible. Running it once a year, with updated spending estimates and contribution levels, keeps the gap between plan and reality small enough to close.