Reading Investment Returns: CAGR, Fees and the Trap in 'Average' Returns

Every advertised return number is a choice made by whoever wrote it. Learning to read the choices takes about ten minutes and will save you more than most investment advice.

Four adjustments turn a headline figure into something closer to what you would actually experience: compounding, volatility, inflation and fees.

Compounding: why “average” is the wrong word

Suppose a $10,000 investment gains 50% in year one and loses 50% in year two.

• The arithmetic average of the two annual returns is 0%.
• The actual result is $10,000 → $15,000 → $7,500, a 25% loss.

This is not a trick. Multiplication does not average the way addition does, and the order of returns matters when the starting amounts differ. The larger the swings, the wider the gap between the average and the outcome — which is why the same average return can produce very different results depending on how volatile the path was.

The number that describes what actually happened is the compound annual growth rate, or CAGR. For the example above, a starting value of $10,000 ending at $7,500 after two years is a CAGR of about −13.4% a year — nothing like the 0% that the arithmetic average suggests. That is the number to look for. When you see “average annual return” without “compound” or “annualized”, assume it may be the arithmetic version, which is always equal to or higher than the compound figure.

Inflation: nominal versus real

A 6% nominal return during a period when prices rose 4% is a 2% real return. That distinction matters enormously over decades, because what you can buy is the only thing that matters.

Two details that catch people out:

• Taxes are calculated on the nominal gain. You can owe tax on a return that lost purchasing power.
• Inflation compounds too. At 3% a year, prices roughly double in about 24 years, so a nominal plan that ignores inflation quietly assumes a much lower standard of living in the later decades.

The Bureau of Labor Statistics publishes the Consumer Price Index, which is the standard reference for this calculation.

Fees: the certain cost against the uncertain return

Expense ratios are charged on assets every year, whether the fund rises or falls. Over long periods the compounding of that fee is substantial — the table in our companion piece on index funds works through it. The key asymmetry is that the fee is guaranteed and the excess return it supposedly purchases is not.

Look for the net expense ratio, which reflects fee waivers, rather than the gross figure. And note that a fund’s reported return is conventionally shown net of its expense ratio, but not net of any sales load, platform fee or advisory fee that you pay separately.

The return you earned versus the return the fund earned

This is the adjustment that most affects individual investors and the one least often discussed.

Funds report time-weighted returns: the performance of $1 left in the fund for the whole period. Your personal return depends on when you put money in and took it out, which is a money-weighted figure. If you bought in heavily after a strong run and sold during a decline, your personal return will be worse than the fund’s reported figure — sometimes dramatically so.

No fund can fix this. It is the arithmetic consequence of investing more at higher prices. It is also why the honest measure of your own investing is the one your brokerage statement reports, not the one on the fund’s fact sheet.

Survivorship and window selection

Three more ways a performance figure can mislead without being false:

1. Survivorship bias. A track record that only counts funds still in existence excludes the ones that closed after performing badly.
2. Window selection. A “since inception” figure from a favorable starting date, or a “ten-year” figure that happens to exclude a bad decade, is legitimate and still misleading. Look at rolling periods, or at several different windows.
3. Benchmark mismatch. A fund compared against an index it does not actually resemble can look superior for reasons unrelated to skill.

A ten-minute check

Before acting on any return claim:

1. Is it compound and annualized, or an arithmetic average?
2. Is it net of all fees you would pay, including advisory and platform fees?
3. Is it nominal or inflation-adjusted?
4. What period does it cover, and what would a different period show?
5. How does it compare with a low-cost index fund tracking the same asset class?

If the answer to the fifth question is “roughly the same, before fees”, then the cheapest broadly diversified option you can hold for decades is very likely the better deal. The market’s return is available to anyone at a few basis points of cost. Anything above it has to be earned, and the evidence that it is earned persistently is thin.