Time vs Rate: What Actually Drives Compound Growth
A worse investor with a longer runway beats a better investor who started late. Here is by how much.
Two people put $400 a month into a retirement account. Ana starts at 25 and earns 7% a year. Ben starts at 35, ten years later, and is a better investor — he earns 9%. Both stop at 60.
| Ana | Ben | |
|---|---|---|
| Monthly contribution | $400 | $400 |
| Years contributing | 35 | 25 |
| Annual return | 7% | 9% |
| Total contributed | $168,000 | $120,000 |
| Balance at 60 | $720,422 | $448,449 |
Ben beat Ana by two full percentage points a year for twenty-five straight years — an enormous edge that almost no one sustains — and finished 38% behind. For Ben to match Ana starting ten years late at the same contribution, he would need a return of about 11.76% annually, every year, for a quarter century.
Run it the other way and the point gets sharper. A saver who starts ten years earlier and earns only 5% ends with $454,437. A saver who starts ten years later and earns 7% ends with $324,029. The worse investor with the longer runway wins by 40%.
Why the trade is so lopsided
Time and rate both appear in the compound growth formula, but not symmetrically. The rate sits in the base and time sits in the exponent, and exponents win. More concretely: extra years do not just add another year of contributions, they add another year of growth on every dollar already in the account, including the growth those dollars have already produced.
There is a second, less mathematical reason the trade is lopsided. Time is something you control and a higher return is something you hope for. Starting a decade earlier is a decision; it happens with certainty the moment you make it. Earning two extra points a year for thirty years is a forecast, and it is a forecast that most professional managers fail to deliver. Comparing a certain gain against an uncertain one at face value overstates the uncertain side badly.
The Rule of 72, and exactly where it stops working
Divide 72 by the annual percentage return to estimate the years it takes money to double. At 8%, 72/8 = 9 years. It is the single most useful piece of mental financial arithmetic, and it is worth knowing its error profile:
| Annual return | Rule of 72 estimate | Actual doubling time (annual compounding) | Error |
|---|---|---|---|
| 2% | 36.0 years | 35.0 years | +2.9% |
| 4% | 18.0 years | 17.7 years | +1.9% |
| 6% | 12.0 years | 11.9 years | +0.9% |
| 8% | 9.0 years | 9.0 years | −0.1% |
| 10% | 7.2 years | 7.3 years | −1.0% |
| 20% | 3.6 years | 3.8 years | −5.3% |
| 30% | 2.4 years | 2.6 years | −9.2% |
Note the basis in that third column: these are annual-compounding doubling times. Compound the same 8% monthly and the double arrives at 8.7 years rather than 9.0, which is the figure the Compound Interest Calculator quotes. Both are right; only the compounding frequency differs.
The rule is tuned to be nearly exact around 8% and degrades in both directions. Below 4% it is slightly pessimistic; above 15% it becomes optimistic fast, understating the doubling time at 30% by more than nine percent. If someone uses the Rule of 72 to justify a high-return projection, the shortcut is working in their favor.
The larger limitation is one people rarely mention: the Rule of 72 describes a lump sum and says nothing about contributions. Most people are not sitting on a lump sum; they are adding money monthly, and each contribution has its own doubling clock starting on the day it arrives. A dollar contributed in year 34 of a 35-year plan doubles zero times. That is why total contributions and final balance diverge so wildly — Ana above put in $168,000 and finished with $720,422, but almost none of that leverage came from her last five years of deposits.
The last decade does most of the work
Take a single $10,000 investment at 7% and watch it across four decades:
| Year | Balance | Growth during that decade |
|---|---|---|
| 0 | $10,000 | — |
| 10 | $19,672 | $9,672 |
| 20 | $38,697 | $19,025 |
| 30 | $76,123 | $37,426 |
| 40 | $149,745 | $73,622 |
Total growth over forty years is $139,745. Of that, $73,622 — 52.7% — happens in the final ten years. The first decade contributes $9,672, or 6.9%. More than half the result arrives in the last quarter of the time, and the opening decade, the one that felt like it was doing nothing, is what made the last one possible.
Two consequences follow, and they cut in opposite directions. The first is the familiar one: adding years at the front is worth far more than it feels like at the time, because you are not adding a low-growth decade, you are shifting every subsequent decade one step further up the curve. The second is the one that gets ignored: interrupting the account near the end is disproportionately destructive. Withdrawing $30,000 in year 32 does not cost you $30,000. It costs you $30,000 compounded for the remaining eight years, taken out of the steepest part of the curve. This is a real argument for holding an adequate cash reserve elsewhere — the point of an emergency fund is partly to protect the compounding you have already paid thirty years to earn.
Fees are a rate reduction that never has a bad year
Every fee comes straight off the exponent. The Department of Labor's guidance on 401(k) plan fees puts a number on it with an example worth reproducing: a participant with 35 years to retirement, a $25,000 balance, and 7% average annual returns ends with about $227,000 if fees and expenses run 0.5%, and about $163,000 if they run 1.5%. The Department's conclusion is that "the 1 percent difference in fees and expenses would reduce your account balance at retirement by 28 percent."
Running the same arithmetic independently: $25,000 at a net 6.5% for 35 years is $226,556; at a net 5.5% it is $162,846 — a 28.1% reduction. The example holds.
With ongoing contributions the effect is similar in proportion and much larger in dollars. Contributing $500 a month for 35 years at a net 7% produces $900,527; at a net 6% it produces $712,355. One percentage point of fees costs $188,172, or 20.9% of the final balance. The Securities and Exchange Commission's investor bulletin on fees makes the mechanism explicit: "not only is your investment balance reduced by the fee, but you also lose any return you would have earned on that fee." A fee is a small negative return that compounds with perfect consistency, in every market condition, forever.
This is also the clean resolution to the time-versus-rate framing. You cannot reliably buy two extra points of return. You can reliably buy back most of one point by choosing a lower-cost fund, and unlike a return forecast, that improvement is contractual.
What these projections assume, and what they get wrong
Every figure above assumes a constant annual return. Real markets do not deliver 7% a year; they deliver something wildly different each year that may average out to something like 7% over a long enough stretch. That matters in two ways. Volatility drags the compound result below the arithmetic average of the annual returns, so a portfolio averaging 7% arithmetically compounds at less than 7%. And the order of returns matters once you are adding or withdrawing money — a bad decade at the end of accumulation does far more damage than the same decade at the start, which is the mirror image of the table above.
The 7% assumption itself is contested. It approximates long-run US equity returns before inflation over the last century, but there is genuine disagreement among researchers about whether that is a reasonable forward-looking figure or an artifact of an unusually favorable period for one country's stock market. Some argue for materially lower expected returns from current valuations. Nobody knows. That uncertainty is another argument for the time side of the trade: the return assumption is the part of the projection that might be wrong, and the number of years is the part that is not.
One structural note if you are starting late. The IRS elective deferral limit for 401(k)-type plans is $24,500 in 2026, with an additional $8,000 catch-up contribution available at age 50 and over, and a higher $11,250 catch-up for ages 60 through 63 under SECURE 2.0. The catch-up provisions exist precisely because the arithmetic in this article does not work for late starters — when you cannot add years, the only lever left is contributing more per year.
Model your own version with the Compound Interest Calculator and the Retirement Calculator: hold the contribution fixed, then change the start date by ten years, then change the return by two points, and see which one moves the answer more. For nearly every realistic set of inputs, it is the start date.
Related tools
- Compound Interest Calculator — vary time, rate, and contributions independently
- Retirement Calculator — project a balance and a drawdown from your own numbers
- Savings Goal Calculator — find the monthly amount that reaches a target on schedule
Sources and further reading
Figures and definitions on this page are drawn from the following primary sources. If you find something out of date, tell us and we will correct it.