The power of compounding
The answer
Compounding is what happens when returns earn returns. It has 3 inputs: the amount you invest, the rate you earn after costs, and the number of years.
Almost everybody tries to improve the second input. The first and third are easier to control, and costs subtract from the second every year without ever appearing on a statement as a loss.
Why this costs you money
Two percent a year sounds small. Here is what it does.
Invest a fixed amount every month for 20 years and earn 7% a year after inflation. Now run the same amount again at 5% a year, because 2% went to fees, commissions, taxes on frequent selling and the gap between a regular plan and a direct plan.
You end with about 21% less money. Not 2% less. Not 20 times 2%. About 21% of the final amount, gone.
Run it for 30 years instead and you end with about 32% less.
The reason is that the missing 2% is not taken from your contribution. It is taken from the compounding. Every rupee or dollar of fee is a rupee or dollar that never earns anything again, and the money it would have earned never earns anything either.
This is the most expensive thing in most people's financial lives and it is invisible, because a percentage deducted from a fund's value does not appear on any statement as a payment. You never write the cheque. You simply end up with less, and you have nothing to compare it against.
The second cost is time, and it is worse. Money invested in year 1 has 20 years to compound. Money invested in year 15 has 5. A delay of 3 years at the start cannot be repaired later by investing more, because the input you removed was years and you cannot buy them back.
How it works
Simple interest earns on the original amount. Compound interest earns on the original amount plus everything earned so far. Over short periods the difference is trivial. Over long periods it is the entire result.
The rule of 72 gives you the arithmetic in your head. Divide 72 by the annual percentage return, and the answer is roughly the number of years for money to double. At 8% a year, money doubles in about 9 years. At 6%, about 12 years. At 12%, about 6 years.
That rule shows why the last years matter most. Money that doubles 5 times becomes 32 times the original. The fifth doubling adds 16 times the original amount by itself, and it is the same 9 years as the first doubling, which added only 1. Nothing changed except how much was already there.
Here is the arithmetic in a form you can use directly. These are multipliers for a fixed monthly investment. Multiply your own monthly amount by the factor to get the value at the end.
| Years of monthly investing | Factor at 7% a year | Factor at 5% a year | Share of the final amount lost to 2% of costs |
|---|---|---|---|
| 10 | 173 | 155 | about 10% |
| 20 | 521 | 411 | about 21% |
| 30 | 1,220 | 832 | about 32% |
Read the last column again. The damage done by a fixed rate of cost grows with time. Costs are not a flat tax on a long-term investor. They are a tax that increases every year you stay invested, which is exactly the opposite of how they are usually described.
Three practical points follow.
Use real returns, not nominal ones. A 12% return with 6% inflation is a 6% increase in what you can buy. Planning in nominal numbers makes every projection look better than it is. For a long horizon, 6% to 7% a year after inflation is a defensible assumption for a diversified equity portfolio in either country, and anything above that should be justified rather than assumed. Interruptions are more expensive than they look. Stopping contributions for 2 years during a bad market removes the contributions and removes the compounding on them. Worse, the contributions you skip are usually the cheapest ones, because prices are low when people stop.
Nothing compounds if the average return is negative. Compounding is a multiplier, and it multiplies whatever the return actually is. A 50% fall requires a 100% rise to get back to level. This is why the position sizing article is a prerequisite for this one: avoiding a large permanent loss is worth more than adding 1% to the return.
What it tells you, and what it does not
Compounding tells you that the 2 inputs you fully control — how much and how long — are worth more than the input you do not control, which is the return. Doubling the monthly amount doubles the result exactly. Adding 10 years multiplies it by far more than 2.
It does not tell you that returns will arrive smoothly. The tables above use a single number every year. Real markets do not. A portfolio can go nowhere for 7 years and then produce most of a decade's return in 18 months. The arithmetic still holds over the full period, but there is no year in which it feels like it is working.
Compounding also does not guarantee a positive result. It describes the shape of growth, given a return. It does not supply the return. A country's stock market can go 34 years without a new high, and the arithmetic of compounding is completely silent about that.
And a projection is not a plan. A table saying you will have a certain amount in 20 years assumes 240 consecutive months of behaviour that most people do not manage. The behaviour is the hard part, and it is the subject of the last 2 articles in this cluster.
The decision rule
Work on the inputs in this order, because that is the order of how much control you have.
- Years. Start now with a small amount rather than later with a large one.
A delayed start cannot be repaired by a larger contribution.
- Costs. Every 1% of annual cost removed is a permanent increase in the
compounding rate. Direct plans over regular plans, index funds over high-fee active funds unless you can defend the fee, and fewer transactions.
- Amount. Increase the monthly contribution with your income, every year,
automatically.
- Return. Last. It is the input you control least and the one that
attracts all the attention.
Try this now
Five minutes, with your own actual number. Not an example.
- Find what you actually invest each month. Add up your systematic investment plans, your monthly transfers, and any regular purchases. Use the real figure, not the one you intend to reach.
- Multiply it by 521. That is roughly what it becomes after 20 years at 7% a year above inflation.
- Multiply the same monthly figure by 411. That is the same 20 years with 2% a year of costs removed.
- Subtract the second number from the first. Write that number down.
- Now find your actual costs. For each fund, find the expense ratio on the factsheet, and check whether you hold the direct plan or the regular plan of the same scheme. For individual stocks, add up the brokerage and taxes you paid last year and divide by your portfolio value.
What you should see. The number in step 4 is usually large enough to be uncomfortable, and it is the cost of things you have never been billed for.
Step 5 tells you how much of that 2% is really yours to lose. If you hold direct plans and index funds and you rarely trade, your total cost may be well under 0.5% a year, and the number in step 4 is mostly hypothetical. If you hold regular plans of active funds and you trade often, 2% is optimistic.
Do the same calculation with 30 years if you are young. The share lost to costs rises from about 21% to about 32%, and seeing that shift is the point of the exercise.
Three real cases
1. Berkshire Hathaway, 1965 to 2023 (United States) — what a rate difference does over a lifetime Berkshire Hathaway compounded at about 19.8% a year in market value from 1965 to 2023, against about 10.2% a year for the S&P 500 including dividends over the same period. The gap in the annual rate is about 9.6 percentage points, which sounds like roughly double. Over 58 years it is not double. It is a difference of several thousand times in the final amount. That is the whole argument for why a small annual difference, in either direction, is not small.
2. Morningstar's "Mind the Gap" study, 2026 edition (United States) — the gap between the fund and the investor Morningstar compares the return a fund produced with the return its investors actually received, which differs because investors buy and sell at different times. In the 2026 edition, covering the 10 years ended 31 December 2025, the average dollar invested in United States mutual funds and exchange traded funds earned 8.7% a year, against 9.9% a year for the funds themselves. The gap of 1.2 percentage points a year is about 12% of the total return, and no fee caused it. Timing caused it. The behaviour is the cost.
3. SEBI's study of individual traders in equity derivatives, 23 September 2024 (India) — costs, compounded against you SEBI's press release of 23 September 2024 reported that over FY22 to FY24, individual traders in the equity futures and options segment paid roughly ₹50,000 crore in transaction costs, and that in FY24 the average individual trader paid about ₹26,000 in such costs. A later SEBI release of 20 August 2026 put cumulative transaction costs paid by individuals over FY22 to FY26 at approximately ₹1 lakh crore. Compounding works in both directions. Money paid out in costs compounds for the person who received it.
The question that resolves it
A novice looks at an investment and asks: what return will this give me?
An expert asks: what will I keep, after costs and tax, and for how many years will I actually leave it alone?
The first question has an answer nobody knows. The second has an answer you can calculate this afternoon, and it decides more of the result.
What would make this wrong
If costs did not matter much, then low-cost funds and high-cost funds with the same strategy would produce similar results for investors over long periods. They do not, and the difference is close to the size of the fee difference, consistently, across markets and decades.
The honest limits.
The multiplier tables assume a constant rate. Real returns arrive in an uneven order, and the order matters when you are taking money out. Two portfolios with the same average return can end very differently if one had its bad years at the start of a withdrawal period. The tables also assume you keep contributing, and most people do not.
And a low fee is not automatically the right choice. A fund charging 1% that you will hold through a 40% fall is better than one charging 0.1% that you will sell in month 3 of that fall. The Morningstar gap above is larger than most fee differences. Cost is the first thing to fix because it is certain, not because it is the only thing that matters.
In India
The most common avoidable cost in India has a name. Every mutual fund scheme has 2 versions: a regular plan, which pays a commission to the distributor who sold it to you, and a direct plan, which does not. The portfolio is identical. The difference is deducted from your returns every year, forever. Check whether the scheme name on your statement contains the word "Direct". If it does not, you are paying the commission version.
Fund charges are capped. SEBI sets maximum total expense ratios that fall as a scheme grows, and index funds and exchange traded funds sit far below the cap. Systematic investment plans are the Indian mechanism for the "years" input. A fixed monthly amount is debited automatically into a fund. The value is not that it improves returns; it is that it removes the monthly decision, which is where most of the damage happens.
Tax reduces the compounding rate whenever you sell. Long-term capital gains on listed equity are taxed at 12.5% above ₹1.25 lakh of such gains in a financial year; short-term gains at 20%. Every sale converts a compounding balance into a smaller one. That is an argument for holding, quite separate from any view about the company.
In the United States
The American version of the same lesson is dollar-cost averaging: investing a fixed dollar amount on a fixed schedule, usually automatically from salary.
The structural advantage is the tax-advantaged account. A 401(k) at work and an Individual Retirement Account allow investments to grow without tax on gains along the way, which raises the compounding rate directly. Many employers also match a percentage of contributions, which is an immediate return before any market return at all. Annual contribution limits apply.
Costs are unusually low and unusually visible. Large index funds and exchange traded funds charge a few hundredths of a percent a year, disclosed in the prospectus and on every fund page. The difference between a 0.03% index fund and a 1% advised product is roughly half of the 2% used in the table above, and it is available to anyone who reads one line of a factsheet.
The American cost that is easy to miss is advice. A common fee for a managed account is around 1% of assets a year, on top of the funds' own charges. That may be worth paying for somebody who stops you selling in March 2020. It is not free, and it belongs in the 2% calculation.
Where they differ, and what that tells you
The arithmetic is identical. The place where the cost hides is different.
In India, the largest avoidable cost for most investors is the regular plan commission built into the fund's expense ratio. It is not billed, not itemised and not visible on a statement. You find it by reading the scheme name.
In the United States, fund costs are usually already low, and the largest avoidable cost is more often the advisory fee, the tax paid on trading in a taxable account, and the money given up by not using the employer match or the tax-advantaged account at all.
What that tells you is where to look first in each country. An Indian investor's first hour is best spent checking whether every scheme held is a direct plan. An American investor's first hour is best spent checking that the employer match is fully used and that long-term holdings sit inside a tax-advantaged account.
Both actions take under an hour, are permanent, and are worth more over 20 years than any stock either investor will pick this year.
Carry this
- Compounding has 3 inputs. You control 2 of them completely: the amount and the years.
- Two percent a year of costs removes about 21% of the final amount over 20 years, and about 32% over 30.
- Costs are a tax that grows the longer you invest, not a flat one.
- Rule of 72: divide 72 by the return to get the years to double.
- Nothing compounds if the average return is negative, so avoiding a large permanent loss is worth more than adding 1% to the return.