Long-run stock market returns are quoted in many different ways, and the differences are large. "The market returns 10% a year" and "stocks return about 7%" can both be correct — they are measuring different things. This guide breaks down the S&P 500's record from 1871 to mid-2023 into its parts: price growth, dividends and inflation.

Three ways to measure the same market

From January 1871 to June 2023, using Robert Shiller's monthly data:

Measure Annual return $1 in 1871 became
Price index only 4.6% about $979
Total return (dividends reinvested), nominal 9.2% about $641,800
Total return, real (after inflation) 6.9% about $26,200 in 1871 purchasing power
Inflation (consumer prices) 2.1% prices rose about 24.5 times

Two lessons jump out:

  1. Dividends did most of the heavy lifting. Price growth alone turned $1 into under $1,000. Reinvesting dividends turned it into more than $640,000. Over long periods, ignoring dividends understates stock returns enormously.
  2. Inflation takes a large share. Roughly a quarter of the nominal return was simply keeping up with rising prices. What matters for your future spending power is the real return.

How the record looks over different starting points

Period Nominal total return Real total return Inflation
1871 – mid-2023 9.2% 6.9% 2.1%
1926 – mid-2023 10.2% 7.0% 3.0%
1950 – mid-2023 11.3% 7.5% 3.6%
1980 – mid-2023 11.6% 8.1% 3.2%
2000 – mid-2023 6.8% 4.2% 2.6%

The real return has been remarkably stable at around 7% a year over very long periods, even though nominal returns and inflation have varied. But the 2000–2023 row is a reminder that a 23-year stretch can fall well short of the long-run average.

Decade by decade

Decade Nominal total return Real total return Inflation
1900s 9.9% 7.5% 2.3%
1910s 4.7% −2.1% 6.9%
1920s 15.5% 16.9% −1.2%
1930s −0.2% 1.9% −2.1%
1940s 9.2% 3.6% 5.4%
1950s 18.8% 16.2% 2.2%
1960s 7.9% 5.2% 2.6%
1970s 6.3% −1.1% 7.5%
1980s 16.6% 11.0% 5.1%
1990s 18.2% 15.0% 2.9%
2000s −0.6% −3.0% 2.5%
2010s 13.5% 11.6% 1.8%

Decades run from January to January. Some patterns worth noticing:

  • Inflationary decades hurt real returns badly. The 1910s and 1970s both delivered positive nominal returns that turned negative after inflation.
  • Deflation can help. In the 1930s, nominal returns were slightly negative, but falling prices meant investors' purchasing power actually grew a little.
  • Great decades often follow poor ones. The 1950s followed the 1930s–40s, and the 1980s followed the 1970s. Starting valuations matter.

How bad can a long holding period be?

Looking at every rolling period that starts in a different month (real total return, annualised):

Holding period Worst Median Best Share of periods with a real loss
1 year −58.1% 8.6% 151.3% 31.1%
5 years −13.2% a year 7.2% 33.4% 19.0%
10 years −5.9% a year 6.9% 20.0% 11.3%
20 years −0.2% a year 6.7% 13.6% 0.1%
30 years +1.9% a year 6.7% 11.2% 0.0%

The worst 10-year period in this data started in March 1999: an investor lost about 5.9% a year in real terms for a decade. Over 20 years, the worst outcome was roughly break-even after inflation; over 30 years, every starting month produced a real gain.

This is the real meaning of "stocks are a long-term investment". Over one year, losing money after inflation happened about three times in ten. Over ten years, it still happened about one time in nine. Only over periods of two decades or more did the risk of a real loss largely disappear — and even then the range of outcomes was wide.

Using this in your own planning

  • Decide whether you are thinking in real or nominal terms, and be consistent. Our calculator works in nominal terms. If you want a projection in today's money, use a return assumption about 2–3 percentage points below your nominal expectation.
  • Do not assume the long-run average for a short horizon. Five- and ten-year outcomes vary widely, which is why the calculator shows a range rather than a single line.
  • Remember fees and taxes. These figures are for the index itself. Fund fees and taxes on dividends reduce what investors actually keep.

Sources and notes

Data: Robert J. Shiller, U.S. Stock Markets 1871–Present, via datasets/s-and-p-500, through June 2023. Monthly average prices; dividends reinvested monthly; inflation from the CPI (spliced with earlier price indexes before 1913, as described by Shiller). Calculations are our own and may differ slightly from figures based on daily or year-end data.

This guide is educational and not investment advice. Past performance does not guarantee future results.