S&P 500 Average Annual Return and Historical Performance

Search Fintrev

Search our guides, reviews and glossary.

The S&P 500 average annual return is the yearly growth rate the index has produced over a chosen period. It is usually expressed as a compound annual growth rate (CAGR) and usually includes reinvested dividends. Measured from 1957, when the index took its 500-stock form, through 2026, the S&P 500 compounded at about 10.51% a year in nominal terms and about 6.64% a year after inflation. Measured from 1928 (using the index’s predecessor) through the third quarter of 2026, the figures are about 10.09% nominal and 6.81% real.

That is the basis for the common “10% a year” rule of thumb. The index rarely returns close to 10% in any single year, though. Some years post large gains, others post double-digit losses, and 10% is roughly what those results average out to over many decades. The method behind the figure matters as much as the number itself: which start date, which kind of average, whether dividends are included and whether inflation is removed.

What the S&P 500 average annual return measures

The S&P 500 is a market capitalization-weighted index of 500 leading U.S. publicly traded companies, maintained by S&P Dow Jones Indices, a division of S&P Global. Its roots go back to the 1920s, when it was a composite index of 90 stocks. It expanded to its present size and name in 1957.

An average annual return for the S&P 500 condenses a long, uneven record into one rate per year. It answers a specific question: if the index had grown at a constant rate every year between two dates, what rate would produce the same ending value? Strictly, that describes the compound annual growth rate, and it is the meaning behind most quoted long-run figures.

Because the S&P 500 is an index and not a security, nobody earns its return exactly. Investors earn the return of an index fund or ETF that tracks it, minus costs. Still, the index’s historical return is the standard benchmark for U.S. large-company stocks and the most common shorthand for “the stock market’s” long-term return.

How the average annual return is calculated

Three choices determine the figure you see: how the average is computed, whether dividends are counted, and whether inflation is removed.

Arithmetic mean vs. CAGR

The arithmetic mean adds each year’s return and divides by the number of years. The compound annual growth rate links the starting and ending values:

CAGR = (Ending value ÷ Starting value)^(1 ÷ number of years) − 1

The two can differ substantially when returns swing. Consider a hypothetical five-year sequence: +20%, −10%, +15%, −5%, +10%.

• Arithmetic mean: (20 − 10 + 15 − 5 + 10) ÷ 5 = 6.0% per year
• Growth of $1: 1.20 × 0.90 × 1.15 × 0.95 × 1.10 = $1.298
• CAGR: 1.298^(1/5) − 1 = about 5.35% per year

The arithmetic mean overstates what an investor experienced, because a loss takes a bigger gain to recover. A 50% gain followed by a 50% loss has an arithmetic average of 0%, but $100 becomes $150 and then $75, a 25% loss. The more volatile the returns, the wider the gap between the two averages. When a long-run S&P 500 figure is labeled “annualized,” it refers to the compound rate.

The 10.51% figure since 1957 is consistent with compounding. A $100 investment in 1957 grew to more than $98,000 by May 2026, and $100 × 1.1051^69 is roughly $98,500.

Price return vs. total return

The S&P 500 is published in two main versions:

• Price return: tracks only changes in the index level. Dividends paid by the member companies are ignored.
• Total return: assumes every dividend is reinvested back into the index.

Long-run “average return” figures, including the ones above, generally refer to total return. Over many decades, dividends make up a meaningful part of the result. As a hypothetical illustration, if the index level rises 8% in a year and its constituents pay a 2% dividend yield that is reinvested, the total return is roughly 10%, while the price return is 8%. Compounded across decades, that difference grows into a very large gap in ending wealth. When comparing figures from different sources, check which version each uses.

Nominal vs. real return

A nominal return is measured in dollars as they existed at each point in time. A real return removes inflation and measures growth in purchasing power:

Real return = (1 + nominal return) ÷ (1 + inflation rate) − 1

Applying this to the 1957 to 2026 averages, 1.1051 ÷ 1.0664 is about 1.036. That implies average inflation of roughly 3.6% a year over the period. In purchasing-power terms, the $100 invested in 1957 that grew to more than $98,000 nominally would be worth about $8,400 in 2026 dollars. The investment still multiplied about 84 times in real terms, but inflation accounted for most of the nominal growth.

Worked example: what the averages mean in dollars

This hypothetical example shows how the nominal and real averages translate into ending values. It assumes a $10,000 lump sum held for 30 years, a constant annual return, reinvested dividends, and no taxes or fund costs.

The 10% line is in future dollars. The 7% line approximates what that money would be worth in today’s purchasing power. Use the real figure when you think in today’s dollars, such as estimating whether savings will cover future living costs. Use the nominal figure when comparing against other nominal figures, such as a fund’s reported return or an interest rate. Mixing the two, for example by projecting with a nominal return while assuming today’s prices, overstates future purchasing power.

The rule of 72 gives a quick feel for compounding. Divide 72 by the annual rate to estimate the years needed to double. At 10%, money doubles roughly every 7.2 years. At 6.6% real, purchasing power doubles roughly every 11 years. Actual annual returns rise above and fall below the constant rate, sometimes by wide margins.

How returns have varied across history

The long-run average blends very different periods:

• Postwar expansion (1957 to 1969): Following World War II, the U.S. economy expanded strongly, and the index climbed with rising industrial output and a growing middle class.
• Stagflation (1970 to 1981): High inflation combined with weak growth weighed on stocks. Nominal returns were poor, and real returns were worse because inflation was eroding purchasing power at the same time.
• Internet boom and bust (1990 to 2002): Enthusiasm for internet companies pushed the index to record highs in the late 1990s. The dot-com crash then produced a steep decline in the early 2000s.
• Financial crisis (2007 to 2009): The index fell nearly 57% from October 2007 to March 2009, the deepest decline of the Great Recession era.
• Long recovery (2009 to 2020): The market entered the longest bull run in its history, rising about 330% over roughly 10 years.
• Pandemic and after (2020s): The COVID-19 shock caused a sharp but brief drop in 2020, followed by a fast recovery and record highs in 2021. The index lost more than 18% in 2022, then recovered in 2023 and set new all-time highs in 2024 and 2025 before renewed volatility tied to recession fears and tariffs.

Every major decline so far has eventually been followed by a recovery, but the time to recover has varied a lot. An investor who needed money in the middle of a long drawdown would have experienced something very different from the average.

Why your entry and exit dates matter

The historical average describes the index between two fixed dates. Your result depends on your own dates. Someone who bought an S&P 500 fund a decade ago and held it would have seen substantial gains. Someone who bought near a peak and sold at the lows of 2020 or 2022 would have locked in losses during the same broad period. Nobody can reliably know peaks and lows in advance.

This is why many advisors recommend dollar-cost averaging: investing a fixed amount on a regular schedule regardless of market conditions. It reduces the chance of putting all your money in at a peak. A fixed dollar amount also buys more shares when prices are low. Hypothetically, investing $100 a month when a fund costs $50, then $25, then $50 buys 2 + 4 + 2 = 8 shares for $300, an average cost of $37.50 per share, below the $41.67 average of the three prices. Dollar-cost averaging does not guarantee a profit or protect against a falling market. It changes the timing risk, not the market risk.

The same logic applies to any trailing-period figure, such as a 10-, 20- or 30-year average. Moving the start or end date by even a single year can change the result substantially if that year was a crash or a boom. A 20-year window that begins just before the 2008 crisis looks different from one that begins at the 2009 low.

How the index’s composition shapes its returns

Buying the S&P 500 is often described as buying the broad market. That is true in the sense that it holds about 500 companies, but because the index is weighted by market capitalization, the largest companies drive a growing share of its returns.

Concentration in the largest stocks

In 2026, the seven companies known as the “Magnificent Seven” (Alphabet, Apple, Amazon, Meta, Microsoft, Nvidia and Tesla) accounted for more than a third of the index’s total market value. The 10 largest U.S. stocks made up about a third of the index, compared with about 27% at the peak of the dot-com bubble in 2000.

Other measures point the same way. Before the 2020s, the last time the index had a company worth more than 500 times the 75th-percentile stock was 2000, shortly before the dot-com bubble burst. A top stock this much larger than the 75th-percentile stock was last seen in 1932. Concentration has increased by earnings as well as by market value.

Concentration magnifies results in both directions. During the 2022 tech sell-off, the largest companies accounted for a disproportionate share of the index’s losses. A few firms having a bad stretch can now move the whole index noticeably.

How the leading sectors have changed

• 1957: Industrials made up more than 40% of the index.
• 1970s: Energy companies briefly dominated during the oil crisis.
• 1990s: Financial services grew to more than 20%.
• 2020s: Technology reached historic highs, exceeding 25% of market capitalization.

These shifts mean the historical average was earned by a changing mix of businesses, from the industrial and energy companies that dominated in the 1960s to today’s technology leaders. The long-run figure describes U.S. large-company stocks as a group across those transitions, not the prospects of any particular set of companies.

How the figure is used

The S&P 500’s historical return serves three main purposes:

1. Benchmarking. It is the standard yardstick for U.S. large-cap investments. An actively managed fund, a stock-picking strategy or a portfolio of individual shares is commonly judged by whether it beat the S&P 500 total return over the same period, after costs.
2. Planning assumptions. Retirement and savings projections often assume a long-run equity return. Using the real figure (roughly 6.6% to 6.8%) instead of the nominal one keeps the projection in today’s dollars.
3. Setting expectations for index funds. The return of an S&P 500 index fund should come close to the index’s total return minus the fund’s costs. Index fund average returns therefore depend largely on which index the fund tracks and what it charges.

Investing in the index

You cannot buy the S&P 500 directly. Assembling it yourself is impractical: even one share of each of the 10 largest constituents could cost around $3,000 without fractional shares, and you would still need to match the index weights across about 503 stocks. The count exceeds 500 because some companies, such as Alphabet, Meta and Berkshire Hathaway, have more than one share class. Most investors use an ETF or mutual fund that tracks the index, such as the SPDR S&P 500 ETF Trust (SPY).

Why funds tracking the same index differ

Funds tracking the S&P 500 do not all return exactly the same. The differences come from:

• Expense ratios (annual management fees)
• Trading costs incurred when buying and selling holdings
• Cash management, since some funds keep small cash reserves for redemptions, which trail the index when stocks rise
• Securities lending, where some funds earn extra income by lending shares to short sellers

These differences, called tracking error, are usually small, but they compound over long holding periods.

Market-cap weighting vs. price weighting

The S&P 500 weights each company by its total market value (share price times shares outstanding). The Dow Jones Industrial Average is price-weighted, so a $100 stock has twice the influence of a $50 stock regardless of company size. In the S&P 500, a company like Apple or Nvidia carries more weight than a stock with a higher share price but a smaller total value. Because of this difference in method, the two indexes can post noticeably different returns over the same period.

Limitations of the average annual return

• It is backward-looking. A historical average describes what happened, not what will happen. The index’s makeup, valuations and the economy behind it are different now from what they were in 1957.
• It hides volatility. A 10% average includes the more than 18% decline in 2022 and the nearly 57% peak-to-trough fall in 2007 to 2009, and gives no indication of how rough the path was.
• It depends on the window. Different start and end dates produce different averages, as the 1957 and 1928 figures show.
• It is an index, not an account. Fund fees, taxes on dividends and gains, and your own contributions and withdrawals all mean your money-weighted return will differ from the index figure.
• Concentration changes the risk. Today’s index relies more on a handful of companies than in most of its history, so past diversification benefits may not carry over in the same way.

The Bottom Line

Anyone projecting savings or judging a fund against the market needs to know which version of the figure they are using. Use the real return of roughly 6.6% to 6.8% for plans in today's dollars, and the nominal figure for comparisons with fund returns or interest rates. The average hides deep drawdowns and depends on the dates chosen, so actual results in an index fund will differ.