Compound interest is growth earned on both the original money and the growth already added. That second layer is why a flat line can eventually bend upward—but only when time, contributions, and returns are allowed to keep working.
Compound interest in plain English
Start with $1,000 and assume it earns 10% in one year. The balance becomes $1,100. If the next year also earns 10%, the gain is $110 because the return is applied to the full $1,100, not only the original $1,000. The extra $10 is growth on earlier growth.
Future value = principal × (1 + periodic rate)number of periods + compounded contributions
For recurring contributions, every deposit has a different amount of time to grow. The first deposit gets the longest runway; the last deposit gets the shortest. WealthClock models monthly deposits at the end of each month, which is a common and slightly more conservative convention.
A worked example: $100 per month
The table below uses no starting balance, a fixed $100 monthly contribution, end-of-month deposits, and smooth annual returns converted to monthly rates. Markets do not move this smoothly; the numbers are deliberately clean so the effect of time is easy to see.
| Assumed return | 10 years | 20 years | 30 years | Total contributed in 30 years |
|---|---|---|---|---|
| 5% | $15,528 | $41,103 | $83,226 | $36,000 |
| 7% | $17,308 | $52,093 | $121,997 | $36,000 |
| 10% | $20,484 | $75,937 | $226,049 | $36,000 |
The wide spread is the lesson. A small change in the assumed return becomes a very large change over 30 years. That is why a calculator should be used to compare scenarios—not to select the prettiest output and call it a forecast.
The assumptions hiding inside the curve
Returns are shown as smooth
Real investments rise and fall. Two portfolios can have the same long-term average return but feel completely different along the way. A smooth calculator line hides bad years, recovery years, and the emotional pressure that can cause people to stop contributing.
Inflation is usually missing
A future balance is normally shown in future dollars. If prices rise over time, those dollars will buy less than the same amount buys today. A useful second pass is to reduce the assumed return by an inflation estimate or compare the result with a goal stated in today’s purchasing power.
Fees and taxes reduce what remains
Expense ratios, advisory fees, trading costs, taxes, and account rules can all reduce the amount that compounds. Even a fee that looks tiny in one year can remove both the fee itself and the future growth that money could have earned.
Contributions are rarely perfectly flat
People pause, increase, reduce, or withdraw contributions. A calculator is still useful, but it should be rerun when the plan changes. The model is a snapshot of the inputs—not a permanent promise.
Four common calculator mistakes
- Using one optimistic return. Run lower, middle, and higher cases.
- Ignoring the contribution total. Separate money deposited from modeled growth.
- Comparing future dollars with today’s goal. Consider inflation.
- Treating a result as advice. The output describes an assumption, not what you should buy.
Frequently asked questions
Does compounding work if returns are negative?
Yes, but in the wrong direction. Losses reduce the base available for later growth. A 50% loss requires a 100% gain to return to the starting value.
Is monthly compounding always better than annual compounding?
With the same stated annual rate, more frequent compounding can produce a slightly higher result. In real investing, however, asset returns are not credited like a fixed bank rate, so the difference between modeling conventions may matter less than the return, fees, and time horizon.
What return should I enter?
There is no universally correct number. Use a range that reflects the asset, fees, taxes, and your uncertainty. The conservative case is usually the most informative.
Sources and further reading
Reality check
This guide explains a planning model, not a recommendation or forecast. Change the inputs, run a downside case, and compare the result with an independent source before acting on it.
Run the number