Skip to content
Monte

Jump to anything

Search topics, drills, solvers and tools

Play · 08 of 09

The long run

98 years of real S&P 500 returns. Simulate 2,000 thirty-year futures and see why the average lies.

Every year since 1928, total return

−40%0%+40%1928: 43.8%1929: −8.3%1930: −25.1%1931: −43.8%1932: −8.6%1933: 50.0%1934: −1.2%1935: 46.7%1936: 31.9%1937: −35.3%1938: 29.3%1939: −1.1%1940: −10.7%1941: −12.8%1942: 19.2%1943: 25.1%1944: 19.0%1945: 35.8%1946: −8.4%1947: 5.2%1948: 5.7%1949: 18.3%1950: 30.8%1951: 23.7%1952: 18.1%1953: −1.2%1954: 52.6%1955: 32.6%1956: 7.4%1957: −10.5%1958: 43.7%1959: 12.1%1960: 0.3%1961: 26.6%1962: −8.8%1963: 22.6%1964: 16.4%1965: 12.4%1966: −10.0%1967: 23.8%1968: 10.8%1969: −8.2%1970: 3.6%1971: 14.2%1972: 18.8%1973: −14.3%1974: −25.9%1975: 37.0%1976: 23.8%1977: −7.0%1978: 6.5%1979: 18.5%1980: 31.7%1981: −4.7%1982: 20.4%1983: 22.3%1984: 6.2%1985: 31.2%1986: 18.5%1987: 5.8%1988: 16.5%1989: 31.5%1990: −3.1%1991: 30.2%1992: 7.5%1993: 10.0%1994: 1.3%1995: 37.2%1996: 22.7%1997: 33.1%1998: 28.3%1999: 20.9%2000: −9.0%2001: −11.8%2002: −22.0%2003: 28.4%2004: 10.7%2005: 4.8%2006: 15.6%2007: 5.5%2008: −36.5%2009: 25.9%2010: 14.8%2011: 2.1%2012: 15.9%2013: 32.1%2014: 13.5%2015: 1.4%2016: 11.8%2017: 21.6%2018: −4.2%2019: 31.2%2020: 18.0%2021: 28.5%2022: −18.0%2023: 26.1%2024: 24.9%2025: 17.8%19281950197520002025
Average year
11.9%
What money actually grew at
10.0%
Spread (SD)
19.4%
Down years
27%
Worst · 1931
−43.8%
Best · 1954
52.6%

The average year was 11.9%, but a dollar invested in 1928 grew at 10.0% a year. The gap, 1.8%, is volatility drag: lose 50% then gain 50% and you are down 25%.

Simulate 2,000 futures

Invest $10,000. Each simulated year is a real year from history drawn at random, so every crash and boom stays in the deck.

30
Your futures appear here

Data: Aswath Damodaran, Historical Returns on Stocks, Bonds and Bills (NYU Stern), updated January 5, 2026. Annual total returns, 1928 to 2025. A teaching model, not investment advice.

What’s going on

The arithmetic mean averages the yearly returns. The geometric mean is the rate your money actually compounded at: (Π(1 + rₜ))^(1/n) − 1. The geometric mean is always lower when returns move around, by roughly σ² / 2. For the S&P 500 since 1928 that gap is about 1.8 percentage points a year.

The simulation is a bootstrap Monte Carlo: build each future by drawing real historical years at random, with replacement, then repeat 2,000 times and read off the percentiles. It keeps every real crash and boom, but it assumes years are independent, so it ignores streaks and changing regimes. Treat it as a model, not a forecast.

Data: annual total returns from Aswath Damodaran’s Historical Returns on Stocks, Bonds and Bills at NYU Stern, updated January 5, 2026.

For learning, not investment advice.