Skip to main content

Arithmetic Mean vs. Actual Investor Return

The gap between what a fund reports and what you actually earned is not a rounding error—it's a systematic and predictable mathematical divergence. When a fund says "10% average annual return," there are at least three different things "10%" could mean, and each one tells a radically different story about your actual wealth.

An arithmetic mean is simple: add all returns and divide by the number of periods. A geometric mean is what actually happened to your money: how much it compounded, period by period. The difference between these two metrics is the difference between a lie (however well-intentioned) and the truth about your portfolio.

Quick definition: Arithmetic mean is the simple average of returns across periods. Geometric mean (annualized return) is the constant growth rate that would produce the same final value if applied each period. Geometric mean is always lower when volatility is present.

Key Takeaways​

  • Arithmetic mean return is simple to calculate but disconnected from reality
  • Geometric mean return (also called CAGR or annualized return) reflects what actually happened to your money
  • The formula relating these is: Geometric Mean ≈ Arithmetic Mean - (Variance / 2)
  • This relationship holds across all investment types and time periods
  • Funds report arithmetic means because they're higher; actual investors care about geometric means
  • The larger the volatility, the larger the gap—and the more important it is to know the difference
  • Understanding this distinction is essential for comparing investments and predicting future wealth

The Two Ways to Calculate Returns​

Let's say you have a $100,000 investment with these year-by-year returns:

  • Year 1: +15%
  • Year 2: +5%
  • Year 3: +20%

Flowchart​

Method 1: Arithmetic Mean (Simple Average)

Add all returns and divide by number of periods:

(15% + 5% + 20%) / 3 = 40% / 3 = 13.33% average annual return

Method 2: Geometric Mean (What Actually Happened)

Calculate the ending value and solve for the constant growth rate:

  • End Year 1: 100,000 × 1.15 = 115,000
  • End Year 2: 115,000 × 1.05 = 120,750
  • End Year 3: 120,750 × 1.20 = 144,900

The geometric growth formula is:

Geometric Mean = ((Final Value) / (Initial Value))^(1/n) - 1
Geometric Mean = (144,900 / 100,000)^(1/3) - 1 = 1.449^(1/3) - 1 = 1.1315 - 1 = 13.15%

In this example with stable returns, the difference is tiny: 13.33% (arithmetic) vs. 13.15% (geometric). But watch what happens when volatility increases.

When the Gap Becomes Dramatic​

Now consider the same ending value but with different sequencing:

Volatile Sequence A:

  • Year 1: +50%
  • Year 2: -25%
  • Year 3: +31.25%

Arithmetic mean: (50% - 25% + 31.25%) / 3 = 18.75%

Let's calculate the actual ending value:

  • End Year 1: 100,000 × 1.50 = 150,000
  • End Year 2: 150,000 × 0.75 = 112,500
  • End Year 3: 112,500 × 1.3125 = 147,656
Geometric mean: (147,656 / 100,000)^(1/3) - 1 = 1.47656^(1/3) - 1 = 1.1384 - 1 = 13.84%

Here's the gap: arithmetic mean of 18.75% vs. geometric mean of 13.84%. That's a 4.91 percentage point difference.

Over 30 years, the impact is massive:

Using arithmetic mean: $100,000 × 1.1875^30 = $4,837,273
Using geometric mean: $100,000 × 1.1384^30 = $3,161,447

The incorrect arithmetic mean projects 53% more wealth than will actually occur. This is a $1.6 million error, stemming entirely from volatility.

The Formula: Connecting the Two​

The precise relationship between arithmetic mean (AM), geometric mean (GM), and variance (σ²) is:

GM = n-th root of (product from i=1 to n of (1 + r_i)) - 1

For a simpler approximation when volatility is moderate:

GM ≈ AM - (σ² / 2)

This approximation is elegant: every unit of variance (squared volatility) costs you half that much in annual geometric return.

Let's verify this with actual numbers. Suppose a portfolio has:

  • Arithmetic mean return: 10%
  • Standard deviation: 8%
  • Variance: (0.08)² = 0.0064

Predicted geometric mean: 10% - (0.0064 / 2) = 10% - 0.32% = 9.68%

Let's verify by simulation with a portfolio that experiences these actual monthly returns over a year:

  • Months 1-6: +1.5% each
  • Months 7-12: +0.5% each

Arithmetic mean: (6 × 1.5% + 6 × 0.5%) / 12 = (9% + 3%) / 12 = 1% per month = 12% annualized

Actual ending value: $(1.015)^6 × (1.005)^6 = 1.0934 × 1.0304 = 1.1261$ or 12.61% growth

The geometric mean is: $1.1261^1 - 1 = 12.61%$ (this is a single-year example, so the formula is simpler)

Over a full year, the volatility drag is: 12% (arithmetic) - 12.61% (geometric) = -0.61%.

Wait—the geometric is higher? That happens because the volatility helped you (larger gains came when the portfolio was smaller, smaller gains came when it was larger). More commonly, volatility hurts, and geometric mean trails arithmetic mean.

The Real-World Investor Impact​

For a typical investor, here's what the arithmetic mean vs. geometric mean distinction means:

Scenario: You read that a fund has "12% average annual return" over 10 years.

If you assume simple compounding (arithmetic mean logic), you think:

100,000 × 1.12^10 = $310,585

But the fund's actual geometric mean was 11.5% (due to volatility drag):

100,000 × 1.115^10 = $293,150

The difference: $17,435—about 5.6% less wealth.

Over 30 years, the cumulative error explodes. If you're planning for retirement using arithmetic means when you should be using geometric means, you might plan for $2 million when your portfolio will only reach $1.6 million.

How Funds Exploit This Gap​

Many funds deliberately report arithmetic means because they're higher. Here's why:

  1. Arithmetic means are always equal to or higher than geometric means (when volatility exists)
  2. Marketing advantage: A fund with actual geometric return of 9% can advertise "9.5% average return" (arithmetic) without lying
  3. Complexity advantage: Most retail investors don't know the difference, so the higher number goes unchallenged
  4. Regulatory wiggle room: The SEC allows arithmetic mean reporting, though it requires funds to disclose standard deviation as well

The SEC's Form N-1A (prospectus for mutual funds) does require funds to show both average returns and standard deviation. But most investors don't look at the standard deviation column—they just see the top-line return number.

The Three Ways Returns Can Be Reported​

There are actually multiple definitions of return, all technically correct, all potentially misleading:

1. Arithmetic Mean Return (AM)

Simple average of periodic returns. Easy to calculate, highest number, most commonly reported.

Example: (15% + -5% + 10%) / 3 = 6.67%

2. Geometric Mean / Compound Annual Growth Rate (CAGR / Geometric Mean)

The actual annualized growth rate, accounting for compounding. Always lower than or equal to arithmetic mean.

Example: (1.15 × 0.95 × 1.10)^(1/3) - 1 = 5.89%

3. Money-Weighted Return / Internal Rate of Return (IRR)

Accounts for the timing and amount of cash flows. If you added money right before a downturn, your IRR will be lower than the fund's CAGR.

Example: If you invested $100,000 on 1/1 and added $50,000 on 6/30, and ended with $150,000 on 12/31, your IRR is lower than the simple CAGR calculation.

Most investors mistakenly use arithmetic means (#1) when evaluating performance, but they should use geometric means (#2), and if they've added or withdrawn money, they should calculate their actual money-weighted return (#3).

A Worked Example: The Full Comparison​

Let's compare three hypothetical funds over a 5-year period:

Fund A (Stable Growth):

  • Returns: 10%, 10%, 10%, 10%, 10%
  • Arithmetic mean: 10%
  • Geometric mean: 10%
  • Variance: 0
  • Final value of $100,000: $161,051

Fund B (Moderate Volatility):

  • Returns: 20%, 5%, 15%, 5%, 10%
  • Arithmetic mean: 11%
  • Standard deviation: 6.4%
  • Variance: 0.004096
  • Geometric mean ≈ 11% - (0.004096 / 2) = 10.8%
  • Actual calculation: $(1.20 × 1.05 × 1.15 × 1.05 × 1.10)^0.2 - 1 = 10.75%$
  • Final value of $100,000: $158,789

Fund C (High Volatility):

  • Returns: 50%, -20%, 40%, -15%, 20%
  • Arithmetic mean: 15%
  • Standard deviation: 29.6%
  • Variance: 0.0876
  • Geometric mean ≈ 15% - (0.0876 / 2) = 15% - 4.38% = 10.62%
  • Actual calculation: $(1.50 × 0.80 × 1.40 × 0.85 × 1.20)^0.2 - 1 = 10.56%$
  • Final value of $100,000: $157,365

Look at the arithmetic means: Fund A (10%), Fund B (11%), Fund C (15%). Fund C looks best.

Look at the geometric means: Fund A (10%), Fund B (10.75%), Fund C (10.56%). Fund B is actually best, despite having a lower arithmetic mean than Fund C.

And look at the actual final values: Fund A ($161,051), Fund B ($158,789), Fund C ($157,365). The order is maintained, but the gap between C and B is only $1,424—much smaller than arithmetic means would suggest.

If you made your fund choice based on arithmetic means, you'd pick Fund C and get $157,365. If you made your choice based on geometric means, you'd pick Fund B and get $158,789. The difference ($1,424) is small on a 5-year timeline, but compounds to tens of thousands over decades.

Why Geometric Mean Is the Right Number​

The geometric mean is the "right" return metric for investors because:

  1. It reflects actual compounding: If you invested $100,000 and it grew to $158,789, the return that actually happened is the geometric mean (10.75%), not the arithmetic mean (11%).

  2. It enables prediction: If a fund's geometric mean is 9% and its volatility is 8%, you can predict its future distribution of outcomes. The arithmetic mean tells you almost nothing about future performance.

  3. It's consistent across time periods: The geometric mean of a 3-year return is the same whether you calculate it as one 3-year return or three 1-year returns. The arithmetic mean is not.

  4. It accounts for volatility automatically: You don't need to separately adjust for volatility; the geometric mean already incorporates it.

Common Mistakes​

Mistake 1: Using arithmetic mean to project future values

If a fund has a 10% arithmetic mean and you project $100,000 growing at 10% per year, you're ignoring volatility drag. The geometric mean is likely 9.2-9.8%, which will compound to significantly less over 30 years.

Mistake 2: Comparing funds on arithmetic means without volatility data

Fund A: 9% arithmetic mean, 5% standard deviation Fund B: 9% arithmetic mean, 15% standard deviation

Without looking at standard deviation, they seem identical. But Fund A's geometric mean is likely 8.985% and Fund B's is likely 8.988%—tiny difference. But the psychological experience is vastly different.

Mistake 3: Forgetting that geometric mean compounds correctly

If a fund had a geometric mean of 9% over 10 years, and you want to project 20 years, you calculate:

1.09^20 = 5.60x growth

You do NOT calculate $(1.09 \times 2) = 1.18x$ growth (that's arithmetic). The geometric mean compounds on itself.

Mistake 4: Ignoring your own money-weighted return

A fund might have a 10% CAGR, but if you invested at the peak and withdrew at the trough, your money-weighted return is probably 4-5%. Your actual return depends on your personal cash flow timing, not the fund's published return.

FAQ​

Q: Is geometric mean always lower than arithmetic mean?

A: Geometric mean is always equal to or lower than arithmetic mean. They're equal only when there's zero volatility.

Q: Can I calculate geometric mean from just arithmetic mean and volatility?

A: The approximation Geometric Mean ≈ Arithmetic Mean - (Variance / 2) works well for moderate volatility. For very high volatility or precise calculations, you should calculate geometric mean directly from period returns.

Q: Why isn't geometric mean the standard for all fund reporting?

A: Partially because arithmetic mean is simpler to calculate. Partially because it's higher and looks better in marketing. The SEC allows both but doesn't mandate one.

Q: Should I ignore arithmetic mean entirely?

A: No. Arithmetic mean has uses (e.g., estimating forward-looking expected returns, comparing risk-adjusted returns via Sharpe ratio). But for evaluating historical performance and projecting final wealth, geometric mean is correct.

Q: Does my brokerage statement show geometric mean?

A: Most brokerages show time-weighted return, which is similar to geometric mean but adjusted for cash flows. Some show "annualized return," which is also geometric mean. Check your statement's footnotes to confirm the methodology.

Q: How do I calculate my personal return if I've added money to my portfolio?

A: Calculate your internal rate of return (IRR), also called money-weighted return. This accounts for when and how much money you added. Excel has an XIRR function that makes this easy.

Q: Can two funds have the same geometric mean but different volatility?

A: Yes. Fund A might be 7% geometric mean with 5% volatility (expected Sharpe ratio of 0.40), and Fund B might be 7% geometric mean with 12% volatility (expected Sharpe ratio of 0.17). Same return, vastly different risk.

  • Why Average Returns Lie: The broader context of arithmetic mean deception. See Why Average Returns Lie.
  • How Volatility Silently Eats Returns: Volatility drag is the engine driving the gap between arithmetic and geometric returns. See How Volatility Silently Eats Returns.
  • 50% Loss Needs 100% Gain to Recover: The same asymmetry that creates the arithmetic/geometric gap. See 50% Loss Needs 100% Gain to Recover.
  • Sharpe Ratio and Risk-Adjusted Returns: Risk-adjusted return metrics need geometric returns as input.
  • Time-weighted vs. Money-weighted Returns: Understanding when each metric applies.

Summary​

The arithmetic mean is what you see in marketing materials. The geometric mean is what actually happened to your money. The gap between these two—driven purely by volatility—is the most important gap in investment mathematics.

When arithmetic mean vs investor return diverge, the geometric mean (actual investor return) is always lower or equal. This gap grows with volatility, compounding over time into dramatically lower final wealth. A portfolio that advertises "12% average annual return" with high volatility might actually deliver a 9% geometric return, which compounds to 43% less wealth over 30 years.

Understanding this distinction is not academic. It changes which funds you should choose, how you should plan for retirement, and why portfolio smoothness matters as much as portfolio returns. The highest-returning fund is not always the best fund. The highest arithmetic mean return is especially not the best fund. The fund with the highest geometric mean relative to its volatility—its risk-adjusted return—is where your money should go.

Next​

How Volatility Silently Eats Returns →

Authority sources: