Why the same target appears twice
A pot of $2.4 million at age 65 sounds like a different problem from a pot of $1.1 million, but at 2.5 percent inflation over thirty years they are the same problem described in different currencies. Everything above is quoted both ways for that reason. The nominal figure is what the statement will say. The today's-money figure is what it buys, and it is the only one of the two you have any intuition about, because it is denominated in prices you pay this week.
Most retirement targets circulating in conversation are nominal figures with the inflation quietly removed, or today's figures with the inflation quietly added, and it is rarely stated which. If someone gives you a retirement number without saying which of the two it is, the number is not usable.
What the calculation actually does
Spending is entered in today's money and inflated to the retirement year. Indexed income is inflated the same way, so the gap between them keeps its real size. The pot has to fund that gap for the whole retirement, with the gap itself rising with inflation each year, while the remaining balance earns the return. The present value of a stream rising at inflation and discounted at a nominal return is an annuity at the real rate, which is why the real return, and not the headline return, is the figure that dominates the answer.
Fixed income is handled separately because it behaves differently. A level pension of $1,000 a month is $1,000 forever in nominal terms, which is a shrinking amount in real terms: at 2.5 percent inflation it buys about half as much after 28 years. Discounting it at the nominal return, rather than inflating it, captures that decline correctly. Lumping indexed and non-indexed income together, which many calculators do, overstates the non-indexed part substantially over a long retirement.
The real return is the whole result
Change the return from 5 to 7 percent and the pot needed falls by a large fraction. Change inflation from 2.5 to 4 percent and it rises again. What matters is the difference between them, compounded, and small changes in that difference move the answer more than almost anything else you can enter.
| Return / inflation | Real return | Effect on the pot needed |
|---|---|---|
| 7% / 2.5% | 4.39% | Lowest requirement of these rows |
| 5% / 2.5% | 2.44% | Noticeably higher |
| 5% / 4% | 0.96% | Higher again, approaching the no-growth case |
| 3% / 3% | 0% | The pot is simply the total real spending, undiscounted |
Notice the last row. When the real return is zero, the arithmetic collapses to something anyone can check: the pot equals the number of years multiplied by the annual real shortfall. That is a useful sanity check on any retirement figure you are given, including this one.
What this deliberately leaves out
Spending is treated as flat in real terms across the whole retirement. In practice it tends to fall through the seventies as travel and activity reduce, then rise again if care is needed, which is a shape no single monthly figure captures. Healthcare costs are inside your monthly figure or they are missing entirely, and in countries where they are not covered they are the largest single uncertainty in the whole calculation.
Also absent: any property you might sell or release equity from, any inheritance, any part-time income after the stated retirement age, and any tax at all. And the monthly saving figure is a level nominal amount, which is harder in the first year than the last, since inflation erodes what it costs you. Increasing contributions with your income each year gets to the same place with less strain at the start. To see what that path looks like month by month, the compound interest calculator shows the accumulation directly, and the monthly need calculator comes at the same problem from the withdrawal side.
Questions people ask
Which figure should I actually aim at, the nominal one or the today's-money one?
Aim at the real one and track the nominal one. The today's-money figure is the one you can judge, because it is denominated in prices you understand, and it is the one to use when deciding whether the plan is plausible. The nominal figure is what your account balance will need to read on the day, and it is the one to compare against a projected statement. They describe the same target and neither is more correct, but quoting one while thinking of the other is how people end up badly over or under saving.
Is 5 percent a reasonable return to assume?
This page will not tell you, because the honest answer depends on what the money is invested in, over what horizon, net of what fees, and in which country. What can be said is what the assumption does: it is the single most influential input on the page, more than your age or your spending, and every result shown moves with it. The useful exercise is to run the page two or three times across a range you would be willing to defend, and look at how wide the answers get. That spread is a more honest output than any single number.
Why does planning to age 92 rather than 85 change things so much?
Because those seven extra years are funded entirely by the pot, at a point when it is at its smallest, and each one adds a full year of the real shortfall to the requirement. Planning to a longer age is the conservative direction, since the cost of leaving money behind is much lower than the cost of running out at 88. Average life expectancy is the wrong number to plan to for exactly this reason: roughly half of people exceed it, and it rises the longer you have already lived.
Why is the answer lower than the 4 percent rule suggests?
Usually because this page funds a defined number of years and allows the pot to be exhausted at the end, while the 4 percent rule was derived to survive a fixed 30-year period across historically bad sequences with something usually left over. If your plan is shorter than 30 years, or your assumed real return is generous, this page asks for less. If your horizon is long or your real return is low, it asks for more. Both are approximations of the same question, and the gap between them is a fair measure of how much the assumptions are doing.
How do I handle a workplace pension that is partly indexed?
Split it between the two fields in the proportion that is indexed. A pension where the first portion rises with prices and the rest is level can be entered as two amounts, one in each box, and the arithmetic will treat each correctly. If the indexation is capped, which is common, the capped part behaves like the fixed field over a long retirement because the cap binds in exactly the years when inflation matters most. Erring toward the fixed box is the conservative choice.