Compounding describes growth applied to a balance that already includes previous growth. The consequence is that the shape of the curve matters more than the rate attached to it.
The base grows along with the balance
Simple growth applies a rate to the original amount each period. Compound growth applies it to the current balance, which includes everything earned so far.
Each period therefore starts from a larger base than the one before, and the absolute gain per period rises even when the rate stays constant.
Over a handful of years the difference is modest. Over several decades the two paths separate to a degree that is difficult to estimate intuitively.
The curve is heavily back loaded
Because gains build on gains, the largest absolute increases occur at the end of the period rather than the beginning.
An investor can hold a position for many years seeing unremarkable progress, then see the balance move substantially in the final stretch without the rate changing at all.
This is why cutting a long horizon short is more costly than it appears. The years removed are the ones that would have contributed most.
Reinvestment is what makes it work
Compounding in an investment context depends on income being put back to work. Dividends and interest withdrawn as cash do not compound.
Two otherwise identical holdings, one reinvesting distributions and one paying them out, diverge steadily over time, and the gap is entirely attributable to that choice.
The distinction matters when comparing quoted returns, since a total return figure assumes reinvestment while a price return figure does not.
Costs compound in the same direction
An annual fee is deducted from the balance that would otherwise have grown, so its effect compounds exactly as returns do.
A charge that sounds trivial as a yearly percentage removes a meaningful share of the final balance across a working lifetime, because it removes the growth on the amount taken as well.
The same logic applies to tax drag on realised gains, which is why the timing of realisation affects long-run outcomes independently of the rate applied.
Time dominates the arithmetic
Comparing contributions started early against larger contributions started later, the early start often ends ahead despite less money being put in.
The reason is that the early money is exposed to more compounding periods, and periods multiply rather than add.
Nothing later in the sequence recovers a lost decade at the start, which is the practical reason horizon is treated as the most important input.