Geometric Mean for Investment Returns

Investment returns compound. That means a simple average of yearly percentages can make performance look better than the money path really was.

Geometric mean fixes that problem by averaging growth factors instead of raw percentages. It answers a practical question: what steady return per period would produce the same final result?

Quick answer

Use geometric mean for investment returns when you have a list of period returns, such as yearly or monthly percentages. Convert each return into a growth factor, calculate the geometric mean of those factors, then subtract 1 to convert the result back into a return.

| Return | Growth factor | | --- | --- | | +6% | 1.06 | | +7% | 1.07 | | +8% | 1.08 | | -35% | 0.65 | | +10% | 1.10 |

For this sequence, the arithmetic average return is positive, but the geometric average return is negative. The loss year reduces the base, so the final investment value falls.

Why arithmetic average can mislead

Suppose an investment gains 20% in one year and loses 20% the next year.

| Period | Start value | Return | End value | | --- | ---: | ---: | ---: | | Year 1 | 1,000 | +20% | 1,200 | | Year 2 | 1,200 | -20% | 960 |

The arithmetic average is 0%, but the investment ends at 960. The actual average compound return is about -2.02% per year.

That gap appears because returns do not add to the same base. The second year’s 20% loss happens after the first year’s gain.

Step 1: Convert returns to factors

Before calculating geometric mean, convert every percentage return into a factor:

| Return type | Conversion | Factor | | --- | --- | ---: | | 8% gain | 1 + 0.08 | 1.08 | | 10% loss | 1 - 0.10 | 0.90 | | 0% return | 1 + 0.00 | 1.00 | | 35% loss | 1 - 0.35 | 0.65 |

Do not enter raw values like 8, 10, -35 when you mean returns. Enter factors like 1.08, 1.10, 0.65.

Step 2: Calculate the geometric mean

For five return factors, multiply the factors and take the fifth root:

Geometric mean factor = (1.06 x 1.07 x 1.08 x 0.65 x 1.10)^(1 / 5)

The result is the average growth factor per period. If the factor is above 1, the average return is positive. If the factor is below 1, the average return is negative.

Step 3: Convert the factor back to a return

Subtract 1 from the geometric mean factor:

Average compound return = geometric mean factor - 1

If the factor is 1.0256, the average return is 2.56%. If the factor is 0.975, the average return is -2.5%.

Geometric mean, CAGR, and XIRR

These three ideas are related, but they are not interchangeable.

| Metric | Use it when | Input style | | --- | --- | --- | | Geometric mean | You have period-by-period return factors | 1.06, 1.07, 0.65, 1.10 | | CAGR | You have one beginning value, one ending value, and years | 1,000 to 1,750 over 5 years | | XIRR | Money was added or removed on specific dates | Dated cash-flow rows |

Use the CAGR Calculator when you only know start value, end value, and years. Use the CAGR to XIRR Converter when deposits, withdrawals, or SIP-style contributions happened during the period.

For the formula relationship, see CAGR vs Geometric Mean. For cash-flow timing, see CAGR vs XIRR.

Common investment-return mistakes

Averaging raw percentages directly

Raw percentages hide the compounding base. Convert returns to factors first.

Treating a loss and equal gain as canceling out

A 50% loss needs a 100% gain to recover. Equal positive and negative percentages do not cancel when they compound.

Using CAGR when cash flows happened

CAGR assumes one beginning value and one ending value. If money moved in or out during the period, XIRR is usually the better annualized return.

Ignoring volatility

Geometric mean tells the average compounded return. It does not show how bumpy the path was. Pair return analysis with drawdowns or spread measures when risk matters.

Why investors should care about the path

Investment performance is not only about the list of yearly percentages. It is about what those percentages do to the account value. A gain in one year changes the base for the next year. A loss does the same, and a large loss requires a larger future gain just to get back to even.

For example, a portfolio that gains 50% and then loses 50% does not end where it started:

100 x 1.50 x 0.50 = 75

The arithmetic average return is 0%, but the investor has lost 25% of the starting value. The geometric mean explains the actual experience:

(1.50 x 0.50)^(1 / 2) = 0.8660

That is a compound average return of about -13.4% per period. The number feels harsher than the arithmetic average because it follows the account value, not the simple average of the percentages.

A full portfolio example

Suppose an investment has these annual returns:

| Year | Return | Factor | Value from 10,000 | | --- | ---: | ---: | ---: | | 1 | +12% | 1.12 | 11,200 | | 2 | +6% | 1.06 | 11,872 | | 3 | -18% | 0.82 | 9,735.04 | | 4 | +14% | 1.14 | 11,097.95 | | 5 | +9% | 1.09 | 12,096.77 |

The ending value is 12,096.77. The investment gained about 20.97% overall.

The geometric mean factor is:

(1.12 x 1.06 x 0.82 x 1.14 x 1.09)^(1 / 5) = 1.0389

The compound average annual return is:

1.0389 - 1 = 3.89%

That means the same start and end values would be produced by a steady return of about 3.89% per year for five years.

Arithmetic average versus investor reality

The arithmetic average of those returns is:

(12 + 6 - 18 + 14 + 9) / 5 = 4.6%

The arithmetic average is not useless. It describes the simple average of the five yearly percentages. But it does not tell you the steady rate that reproduces the final account value. That is why performance reports usually need geometric mean, CAGR, time-weighted return, or money-weighted return depending on the data.

Use this rule:

| Question | Better metric | | --- | --- | | What was the simple average yearly return? | Arithmetic average | | What steady yearly return matches the compound path? | Geometric mean or CAGR | | What return did my personal deposits and withdrawals earn? | XIRR or money-weighted return | | How volatile were returns? | Standard deviation |

Time-weighted versus money-weighted thinking

Geometric mean of period returns is close to time-weighted thinking. Each period return gets equal treatment if each period has the same length. This is useful for judging a strategy or fund manager because it focuses on the return path, not on when an individual investor added money.

XIRR is money-weighted. It cares about when cash enters or leaves. If you invest most of your money right before a bad year, your personal return can be much worse than the strategy’s time-weighted return. If you invest more before a strong recovery, your personal return can be better.

Neither view is automatically superior. They answer different questions. Use geometric mean when you have clean period returns. Use CAGR to XIRR Converter when dated cash flows matter.

How to prepare return data

Before calculating, clean the data:

Make sure every row covers the same period length.
Convert returns to factors.
Check that every factor is greater than zero.
Keep fees, dividends, and reinvestment rules consistent.
Do not mix monthly and annual returns in one list.

If a return is -100%, the factor is zero. That means the investment has lost all value for that path, and ordinary geometric mean becomes zero. Treat that case as a complete-loss scenario, not as an ordinary average.

How to explain the answer

A helpful investment-return explanation should include both the factor and the percent return:

The geometric mean factor was 1.0389, so the compound average return was 3.89% per year.

If you compare it with the arithmetic average, explain why they differ:

The arithmetic average return was 4.6%, but the compound average was 3.89% because the negative year reduced the base for later gains.

That wording helps users understand the result without making the geometric mean sound mysterious.

When to show a chart or table

Investment-return articles, reports, and dashboards are easier to understand when the geometric mean is paired with the value path. A small table showing starting value, return, factor, and ending value makes the compounding visible.

Use a table when:

The return sequence includes at least one loss.
The arithmetic average and geometric average differ meaningfully.
You need to explain why a positive simple average did not create a positive final result.
The audience may not be comfortable with factors.

For example, showing 10,000 -> 12,096.77 alongside the geometric mean makes the average return feel grounded. The user can see that the rate is not just a formula output. It is the steady rate that connects the actual start and end values.

Interpreting risk alongside return

Geometric mean is a return measure, not a complete risk measure. Two investments can have the same geometric mean and very different experiences. One may grow steadily, while another swings sharply between gains and losses.

Pair the compound average with context such as:

Worst period return
Best period return
Maximum drawdown
Standard deviation of returns
Number of positive and negative periods

This keeps the geometric mean in the right role. It summarizes the average compounding path, while risk metrics describe how uncomfortable or uncertain that path was.

FAQ

Is geometric mean the same as average return?

It is the average compounded return when your inputs are period return factors. A simple arithmetic average is a different statistic and can misstate performance.

Can geometric mean handle negative returns?

Yes, if negative returns are converted to positive factors. A 10% loss becomes 0.90. A return of -100% becomes zero, which breaks ordinary geometric mean because the investment went to zero.

Should I use geometric mean or CAGR?

Use geometric mean when you have a list of period factors. Use CAGR when you have a beginning value, ending value, and time period.

Use the Geometric Mean Calculator for return factors. Use the CAGR Calculator for start-to-end investment growth. If you added or withdrew money on different dates, use the CAGR to XIRR Converter.