How Much Money Do You Need to Retire?

Start With Spending, Not Income

Almost everyone begins this calculation with their salary. That is the wrong input.

What you need to fund in retirement is your spending, and the two are rarely equal. Retirement typically removes commuting costs, work clothes, the mortgage if it is paid off by then, and whatever you were saving for retirement itself. It adds healthcare, and often more travel in the early years.

A common planning shortcut puts retirement spending at 70% to 80% of pre-retirement income. Treat that as a placeholder until you have looked at your actual outgoings, because the shortcut hides enormous variation. Someone who owns their home outright and someone still paying rent have very different numbers at the same salary.

Work out what a year of your retired life costs, in today’s money. Everything else in this article scales off that figure.

The 25x Rule and Where It Came From

The 4% figure traces to William Bengen, who published research in 1994 testing how much a retiree could withdraw each year without running out of money. Using US market data going back to 1926, he found that withdrawing 4% of the starting portfolio, then adjusting that amount for inflation annually, survived every 30-year period in the record. The Trinity Study reinforced the finding in 1998 using similar data.

So the rule has genuine research behind it. It also has three specific boundaries that get dropped whenever it is repeated: it was calibrated on US markets, over a 30-year retirement, with a portfolio of stocks and bonds held throughout.

Worked Example

Situation: You expect to spend 40,000 a year in retirement, in today’s money. These figures work in any currency.

The spending figure never changed. Only the assumed withdrawal rate did, and the target moved by two thirds.

Account for other expected income on the same tax and inflation basis as spending. In the simplified example, a pension that covers 12,000 of the annual 40,000 need leaves 28,000 to fund from the portfolio: at an assumed 4%, the arithmetic target is 700,000. Confirm when the benefit starts and whether its value changes over time.

The Number That Changes Everything

Given how much the rate drives the answer, it deserves more scrutiny than it usually gets.

In a 2010 historical study, Wade Pfau examined 17 developed countries. A 4% real withdrawal met the study’s safety criterion in four countries under its optimized allocation assumptions; a fixed 50/50 stock-and-bond allocation failed in at least one tested period in every country. These are historical model results, not a current recommended withdrawal rate.

This matters directly if you are not investing in US markets, or if you are retiring somewhere the currency and market history look nothing like the American record. The 4% rule is not a law of finance. It is a description of what happened in one country’s unusually successful century.

Two further adjustments are worth knowing. Retiring early lengthens the horizon well beyond the 30 years Bengen tested, which argues for a lower rate. And sequence-of-returns risk means a market crash in your first few retired years does far more damage than the same crash a decade later, because you are selling assets to live on while they are cheap.

Use several withdrawal-rate scenarios to see how sensitive the target is. A lower assumed rate produces a larger target, but no single percentage establishes that a plan is safe. Retirement length, asset allocation, fees, taxes, spending flexibility and the sequence of returns all matter.

Working Out What to Save Each Month

Having a target is only half the job. The other half is the monthly figure that reaches it.

This is compound growth running forward, and the variable doing most of the work is time rather than the amount. Someone starting at 25 with a 40-year horizon needs to contribute far less each month than someone starting at 45 with 20 years, because the early contributions spend decades compounding. Our guide on how compound interest works covers the mechanics.

Keep inflation treatment consistent. CalcRange already adjusts the income goal for inflation, so its return inputs should be nominal, not inflation-adjusted real returns. A separate model expressed entirely in today’s money can use real returns if its cash flows use the same basis. In either model, set contribution growth explicitly: a fixed monthly amount and contributions that rise with salary are different assumptions.

What the Simple Model Ignores

• Inflation over decades. At 3% annual inflation, today’s 40,000 of spending needs roughly 72,000 of nominal income in 20 years.
• Tax on withdrawals, which varies enormously by country and account type and can consume a meaningful share of what you draw.
• Healthcare costs, which tend to rise with age and are the single largest source of unplanned retirement spending in countries without comprehensive public cover.
• Longevity. Planning for 30 years is standard, but a couple retiring at 60 has a reasonable chance of one partner living past 90.
• Investment fees. A 1% annual fee compounds against you exactly as returns compound for you, and over 30 years it can consume a large fraction of the final balance.
• Property. A home you own reduces required spending; a home you plan to sell is a lump sum with timing risk attached.

If the Number Looks Impossible

For most people running this calculation for the first time, it does. That reaction is normal and the target is more movable than it appears.

Reducing planned spending is the most powerful lever, because it works twice: every 1,000 you cut from annual spending removes 25,000 from the target while also freeing money to save today. Working two or three years longer is the second most powerful, since it adds contributions, adds growth, and shortens the horizon the portfolio must cover.

Beyond that: check what guaranteed income you are already entitled to, since state pensions are frequently overlooked and directly reduce what the portfolio must produce. Reduce investment fees. Consider part-time work in early retirement, which cushions exactly the sequence-of-returns risk that does the most damage.

What does not help is abandoning the calculation because the answer is uncomfortable. A partial plan started today outperforms a perfect plan started in five years, and the arithmetic of compounding is unforgiving about delay in a way it is forgiving about amount.

Frequently Asked Questions

Is the 4% rule still valid?

The 4% rule describes a historical withdrawal approach, not a guarantee or a universal recommendation. Its original U.S. stock-and-bond research used an initial withdrawal followed by inflation adjustments. A different retirement horizon, investment mix, tax treatment or market experience can change the result.

How much do I need to retire at 40?

Retiring at 40 can require funding many more years than a 30-year example. Test the longer horizon, realistic spending changes and several withdrawal assumptions. Dividing spending by 3% gives about 33.3 times annual spending as arithmetic, but it does not establish that 3% is suitable or safe for you.

Should I use gross income or spending to set the target?

Start from expected spending, including taxes and healthcare where applicable. Some work-related costs and contributions may end, while other expenses remain or rise. A salary replacement percentage can be a rough placeholder, but it does not replace an individual retirement budget.

Does the state pension count toward my target?

Expected pension income can reduce the amount spending must draw from a portfolio. Check its start date, tax treatment, inflation adjustments and eligibility. Under the simple 4% illustration, an annual 12,000 reduction in the portfolio-funded spending gap reduces the arithmetic target by 300,000; that is not a universal valuation or guarantee of pension benefits.

What return should I assume?

Test several return assumptions appropriate to the assets being modeled. CalcRange expects nominal returns before fees and asks for inflation and fees separately. Do not enter a real return into that field. Historical returns and a smooth constant-return projection do not predict future results.

What is sequence-of-returns risk?

The risk that poor market returns arrive early in retirement rather than late. Two retirees can experience identical average returns over 30 years and end in completely different positions depending on the order, because withdrawals during a downturn permanently remove shares that would have recovered. It is the main argument for flexible spending.

Is my house part of my retirement fund?

Only if you intend to sell it or borrow against it. A home you live in reduces your spending rather than generating income, which lowers the target without contributing to the portfolio. Counting its full value as retirement savings while also planning to live in it counts the same asset twice.

The Bottom Line

Estimate retirement spending, account for other income on a consistent tax and inflation basis, and test more than one withdrawal rate. The spending-divided-by-rate formula is a starting illustration. In the retirement calculator, use nominal returns and a separate inflation input, then compare several scenarios. A target that works under one set of assumptions is not a guarantee.

Sources and Further Reading

• Bengen WP. Determining Withdrawal Rates Using Historical Data, Journal of Financial Planning, 1994.
• Cooley PL, Hubbard CM, Walz DT. “Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable” (the Trinity Study), 1998.
• Pfau WD. An International Perspective on Safe Withdrawal Rates: The Demise of the 4 Percent Rule?, 2010.
• OECD – Pensions at a Glance, for country-level retirement income systems.