What does "the S&P 500 averages 10% a year" actually mean?

Every personal-finance article — including ours — eventually leans on the same statistic: the S&P 500 has averaged about 10% per year over the long run. It's the most-cited number in retail investing. But almost nobody who quotes it explains what it actually means, where it comes from, or what the year-to-year reality looks like.

This piece is the explainer. By the end you'll know exactly what “10% on average” is doing and what it isn't.

What “10% per year” actually refers to

The number is the annualized total return of the S&P 500 index from roughly 1926 through today. Two words doing all the work there:

• Annualized: not the average of each year's return (which would be much messier), but the constant rate that, if applied each year, would have produced the same ending wealth as the actual messy series did.
• Total return: price appreciation PLUS dividends reinvested. Dividends matter — they've historically contributed 30-40% of the index's total return. Just tracking the price index (the “S&P 500” you see quoted on TV) understates the real wealth-building number by a wide margin.

Different start/end dates give slightly different numbers. The ~10% figure is robust across most rolling 30-year windows since 1926. Common variations you'll see cited:

• 1926-2024: ~10.1% annualized
• 1950-2024: ~11.2% annualized
• 1995-2024: ~10.7% annualized
• 2000-2024 (lost decade + recovery): ~7.6% annualized

The point: 10% is a defensible long-run baseline, not a guarantee for any particular 30-year stretch.

What real returns look like year-to-year

Average annual returns hide enormous year-over-year variance. A representative snapshot:

• 2008: -37.0% (financial crisis)
• 2013: +32.4%
• 2018: -4.4%
• 2019: +31.5%
• 2020: +18.4% (despite a 34% mid-year crash)
• 2022: -18.1%
• 2023: +26.3%
• 2024: +25.0%

Even “average” years are rarely +10%. The market tends to be either way up or way down; the smooth 10% is what emerges over decades, not what any single year delivers. This matters for psychological reasons — if you're going to invest based on the 10% number, you need to be prepared for the years that look nothing like 10%.

The “decade of lost returns” possibility

The 2000s offer a cautionary tale. The S&P 500 returned roughly -1% per year from 2000 to 2009 — a full decade of losses, after two crashes (dot-com and financial crisis). If you invested everything in early 2000 and needed it in 2010, you had less than you started with.

This is part of why long horizons matter so much. The decade after, 2010-2019, returned 13.6% annualized — almost entirely offsetting the bad decade. Over the full 20 years, the annualized return was about 6.1%. Below the long-term average but recognizably positive.

The lesson: 10% is the very-long-run average. Plan for 7% in any given 20-year stretch as a more conservative assumption, and you'll rarely be unpleasantly surprised.

Why we (and most personal-finance writing) use 10%

Two practical reasons we and Cost Me use the 10% number despite its imperfections:

1. It's the most widely-recognized baseline.Using a different rate would either be lower (and people would dismiss the math as not aspirational enough) or higher (and we'd be cherry-picking).
2. The directional lesson is robust. A $200 jacket is $3,967 in 30 years at 10%. At a more conservative 7%, it's $1,624. Either number is large enough to change how you think about the purchase. The opportunity-cost framing works at any reasonable rate.

What 10% is NOT

A few things 10% explicitly is not, that get conflated:

Not a guarantee

Past performance doesn't predict future returns. The future could look like 1990-2024 (great), 2000-2010 (terrible), or something different entirely. Anyone telling you otherwise is selling something.

Not adjusted for inflation

10% is the nominal return — the number you'd see in your brokerage account. Inflation has averaged about 3% per year, so the real return (what your money is worth in purchasing power) is closer to 7%.

For long-term wealth-building math, you can use either: just be consistent. If you're comparing today's purchase to future-you's wealth, the nominal 10% is conventional. If you want to think in today's buying power, use 7% real.

Not the same as an “investment”

The 10% number describes a hypothetical investor who bought the index and held forever, reinvesting all dividends. Real investors who try to time the market, switch in and out of funds, or pick stocks individually historically underperform the index by 1-3 percentage points per year, on average. The biggest risk to your 10% isn't the market — it's you.

Not a substitute for diversification

The S&P 500 is concentrated in US large-cap stocks. A properly diversified portfolio includes international stocks, small caps, and bonds. The overall return on a diversified portfolio is lower than 10% (maybe 8-9%) but with meaningfully less volatility. For Cost Me's opportunity-cost framing, the S&P 500 number is the conventional baseline; for an actual investment portfolio, diversification is the standard advice.

How to internalize the volatility

Most personal-finance writing treats “the market goes up 10% a year” as if it actually does. It doesn't. It goes up 25%, then down 18%, then up 8%, then down 4%, then up 15%. The 10% is an emergent property of that messy sequence.

The practical implication: if you're going to invest, you need to ignore years. The market will crash; you will see your balance fall 30% at some point. The right response is to do nothing. The wrong response is to sell at the bottom and miss the recovery, which is the modal way real investors underperform the index.

The 10% number assumes you stay invested through the bad years. That's the only part of the assumption that's up to you.

The honest version

Combining all of the above, the honest claim is:

Over the last century, broadly diversified US stocks have returned ~10% per year nominal (~7% real) on average, with massive year-to-year variance. Most rolling 30-year periods come in close to that average. Future returns are not guaranteed; the closest thing to a guarantee is that you'll need to ride out painful stretches to capture them.

That's less catchy than “10% a year” but it's the version that survives scrutiny.

The takeaway

10% is a useful number, not a sacred one. Use it for long-horizon opportunity-cost framing (where the directional lesson dominates). Don't use it as a prediction for any specific year, or as a substitute for understanding what investing actually feels like in the short term.

And don't let the imperfection of the number be an excuse to ignore the framework. A $4,000 future cost on a $200 jacket is wrong by 30% — and still big enough to change your behavior. That's good enough.