How to Calculate Geometric Mean for Negative Returns

The geometric mean is a type of average that accounts for compounding effects, making it particularly useful for financial returns that include negative values. Unlike the arithmetic mean, which can be misleading when dealing with percentages, the geometric mean provides a more accurate representation of average growth rates.

What is Geometric Mean?

The geometric mean is a statistical measure that calculates the central tendency of a set of numbers by using the product of their values. It's particularly useful when dealing with rates and ratios, as it accounts for compounding effects.

For a set of positive numbers \( x_1, x_2, \ldots, x_n \), the geometric mean \( G \) is calculated as:

\( G = \sqrt[n]{x_1 \times x_2 \times \cdots \times x_n} \)

When dealing with negative returns, we first convert the returns to their multiplicative factors (1 + return) before calculating the geometric mean.

Why Use Geometric Mean for Negative Returns?

When working with financial returns that include negative values, the arithmetic mean can be misleading because it doesn't account for the compounding effect of returns. The geometric mean provides a more accurate representation of the average growth rate over time.

For example, if you have two periods with returns of +50% and -33.33%, the arithmetic mean would be 8.335%, but the geometric mean would be 0%, correctly reflecting that the investment would be back to its original value after these two periods.

How to Calculate Geometric Mean

Convert each return to its multiplicative factor by adding 1 to the return (e.g., 50% becomes 1.50, -33.33% becomes 0.6667).
Multiply all the factors together to get the product.
Take the nth root of the product, where n is the number of periods.
Subtract 1 from the result to convert it back to a percentage.

Geometric Mean Return = \( \left( \prod_{i=1}^{n} (1 + r_i) \right)^{1/n} - 1 \)

Where \( r_i \) represents each individual return.

Example Calculation

Let's calculate the geometric mean for three periods with returns of 20%, -10%, and 15%.

Convert returns to factors: 1.20, 0.90, 1.15
Multiply factors: 1.20 × 0.90 × 1.15 = 1.236
Take the cube root: \( \sqrt[3]{1.236} \approx 1.072 \)
Convert back to percentage: 1.072 - 1 = 7.2%

The geometric mean return is approximately 7.2%.

Note: The arithmetic mean for these returns would be (20% - 10% + 15%)/3 = 11.67%, which overestimates the actual average return.

Comparison with Arithmetic Mean

The table below compares the geometric mean and arithmetic mean for different sets of returns:

Returns	Geometric Mean	Arithmetic Mean
10%, 20%, 30%	19.15%	20%
50%, -33.33%, 0%	0%	8.33%
-5%, -10%, -15%	-10%	-10%

As shown, the geometric mean provides a more accurate representation of the average return, especially when negative returns are involved.

FAQ

When should I use geometric mean instead of arithmetic mean?

You should use geometric mean when dealing with rates of return, growth rates, or any situation where the compounding effect is important. It provides a more accurate representation of the average growth rate over time.

Can I use geometric mean for negative returns?

Yes, you can use geometric mean for negative returns. The key is to first convert each return to its multiplicative factor (1 + return) before calculating the geometric mean.

What's the difference between geometric mean and harmonic mean?

The geometric mean is used for rates and ratios, while the harmonic mean is used for rates and averages of reciprocals. The geometric mean accounts for compounding effects, making it more appropriate for financial returns.

Is the geometric mean always less than the arithmetic mean?

Not necessarily. The geometric mean can be greater than, equal to, or less than the arithmetic mean depending on the distribution of the data. For positive numbers, the geometric mean is always less than or equal to the arithmetic mean.