Stock market returns have created remarkable long-term wealth, yet the path has never been smooth or predictable. From 1928 through 2025, $100 invested in a large-company U.S. stock portfolio represented by the S&P 500 and its predecessor series grew to about $1.16 million with dividends reinvested. However, that same record included 26 losing calendar years, a worst annual loss of 43.84%, and repeated declines that tested investors’ patience.
Those facts reveal the central bargain of investing. Stocks have historically offered higher average returns than Treasury bills and government bonds. In exchange, investors accepted wider fluctuations, deeper temporary losses, and greater uncertainty. Therefore, a useful study of returns must examine both reward and risk.
This guide explains how to calculate investment returns, why arithmetic and compound averages differ, how economists measure volatility, what the equity risk premium means, and what nearly a century of U.S. data can and cannot tell us. It also updates the concepts shown in the source textbook pages with data through 2025.
Data note: Unless stated otherwise, the historical figures below come from Aswath Damodaran’s January 2026 U.S. returns dataset at NYU Stern. The series covers 1928–2025, and stock returns include both price appreciation and dividends. Damodaran obtains Treasury data from the Federal Reserve and calculates total bond returns using coupon income plus price changes. Download and review the NYU Stern historical returns data.
What Are Stock Market Returns?
A stock market return measures how much the value of a stock investment changes over a given period. Although financial news often focuses on index price movements, an investor’s complete result can include several components.
Price return
Price return measures the change in market price alone:
Price return = (Ending price − Beginning price) ÷ Beginning price
For example, a stock that rises from $100 to $108 produces an 8% price return. Nevertheless, that calculation ignores any cash distributed to shareholders.
Dividend income
Many companies distribute part of their earnings through dividends. Consequently, a comparison that excludes dividends can materially understate the wealth stocks generated over long periods.
The S&P 500 series published on FRED illustrates this distinction. FRED identifies it as a price index and explicitly notes that it does not include dividends. By contrast, the Damodaran series used in this article includes dividends. See the S&P 500 price-index notes at the Federal Reserve Bank of St. Louis.
Total return
Total return combines the price change and cash distributions:
Total return = (Ending price − Beginning price + Cash income) ÷ Beginning price
Suppose an investment begins at $100, ends at $108, and pays a $2 dividend. In that case, the total return equals 10%, not 8%.
| Component | Beginning value | Ending value or cash flow | Contribution |
|---|---|---|---|
| Price change | $100 | $108 | 8% |
| Dividend | — | $2 | 2% |
| Total | $100 | $110 combined value | 10% |
Nominal return versus real return
Nominal return shows the percentage change in dollars. Real return, however, adjusts that result for inflation and therefore better represents the change in purchasing power.
Real return = (1 + Nominal return) ÷ (1 + Inflation rate) − 1
If a portfolio earns 8% while consumer prices rise 3%, its exact real return equals about 4.85%. Simply subtracting inflation gives a close approximation of 5%, but the multiplicative formula gives the correct result.
The Bureau of Labor Statistics calculates its official inflation tool from the Consumer Price Index for All Urban Consumers, or CPI-U, using the U.S. city average for all items, not seasonally adjusted. Explore the BLS CPI program and calculator.
Historical Stock Market Returns From 1928 to 2025
The longest useful record does not promise a future result. Still, it gives investors a disciplined way to understand the range of past outcomes.
Nearly a century in one table
The following calculations use all 98 calendar-year observations in Damodaran’s 1928–2025 dataset. Returns appear in nominal U.S. dollars.
| Asset or measure | Arithmetic average | Compound annual growth rate | Annual volatility | Worst year | Best year | Losing years |
|---|---|---|---|---|---|---|
| S&P 500, dividends included | 11.85% | 10.02% | 19.40% | -43.84% | 52.56% | 26 of 98 |
| U.S. small-cap stocks, bottom decile | 17.78% | 11.98% | 37.95% | -53.94% | 146.60% | 34 of 98 |
| 3-month Treasury bills | 3.41% | 3.37% | 3.04% | 0.03% | 14.04% | 0 of 98 |
| 10-year U.S. Treasury bonds | 4.82% | 4.53% | 7.90% | -17.83% | 32.81% | 20 of 98 |
| Baa corporate bonds | 6.90% | 6.63% | 7.65% | -15.68% | 29.05% | 16 of 98 |
| U.S. residential real estate* | 4.38% | 4.20% | 6.18% | -12.00% | 24.10% | 15 of 98 |
| Gold | 7.36% | 5.61% | 21.51% | -32.60% | 126.55% | 34 of 98 |
*Damodaran’s real-estate series measures home-price appreciation. Therefore, it omits rent, ownership costs, maintenance, taxes, leverage, and transaction expenses. It should not be treated as a complete return on owning rental property.
Several patterns stand out. First, large-company stocks earned more than Treasury bills and bonds over the full period. Second, small-cap stocks recorded the highest arithmetic and compound averages, but they also produced almost twice the volatility of the S&P 500 series. Third, gold was more volatile than large-company stocks even though its compound return was substantially lower.
What happened to $100?
Compounding turns small annual differences into enormous long-term gaps. According to the same dataset, $100 invested at the beginning of 1928 reached the following nominal values by the end of 2025:
| Investment | Ending value of $100 | Approximate multiple |
|---|---|---|
| S&P 500, dividends included | $1,157,009 | 11,570× |
| Baa corporate bonds | $53,952 | 540× |
| Gold | $21,025 | 210× |
| 10-year U.S. Treasury bonds | $7,718 | 77× |
| Residential real-estate price index* | $5,626 | 56× |
| 3-month Treasury bills | $2,578 | 26× |
These dollar values are nominal and do not subtract inflation, taxes, fees, or trading costs. Moreover, the table assumes uninterrupted reinvestment. A real investor could experience different results because of withdrawals, tax treatment, fund expenses, market timing, or behavior.
How often were annual stock returns positive?
The S&P 500 total-return series posted a gain in 72 of the 98 years and a loss in 26. In other words, approximately 73.5% of calendar years were positive. Nevertheless, one-year success was never guaranteed.
| Annual S&P 500 total-return range | Number of years | Share of 98 years |
|---|---|---|
| Below -30% | 3 | 3.1% |
| -30% to below -20% | 3 | 3.1% |
| -20% to below -10% | 6 | 6.1% |
| -10% to below 0% | 14 | 14.3% |
| 0% to below 10% | 15 | 15.3% |
| 10% to below 20% | 21 | 21.4% |
| 20% to below 30% | 18 | 18.4% |
| 30% or more | 18 | 18.4% |
Interestingly, an annual result between 10% and 20% occurred more often than any other interval in this table. Yet the distribution also contained meaningful tails: six years fell below -20%, while 18 years gained at least 30%.
Returns depend on the starting date
Historical averages shift when the measurement window changes. Consequently, no single average describes every investor’s experience.
| Starting year through 2025 | Years | Arithmetic average | Compound return | Volatility |
|---|---|---|---|---|
| 1928 | 98 | 11.85% | 10.02% | 19.40% |
| 1950 | 76 | 12.99% | 11.62% | 16.99% |
| 1975 | 51 | 13.64% | 12.37% | 16.18% |
| 2000 | 26 | 9.57% | 7.99% | 17.79% |
| 2016 | 10 | 15.75% | 14.68% | 15.60% |
The 2000 starting point includes the dot-com crash, the global financial crisis, the 2020 pandemic decline, and the 2022 bear market. Therefore, its compound result trails the full-history figure. By contrast, the 2016–2025 window captures an unusually strong decade and should not automatically become a forecast.
Average Stock Market Returns: Arithmetic Versus Geometric
The phrase “average stock market returns” can hide an important choice. An arithmetic average answers one question, while a geometric average answers another.
Arithmetic average return
Add the annual returns and divide by the number of observations:
Arithmetic average = (R₁ + R₂ + … + Rₙ) ÷ n
Consider returns of +50% and -50%. Their arithmetic average equals 0%:
(+50% − 50%) ÷ 2 = 0%
However, $100 grows to $150 and then falls to $75. Ultimately, the investor lost 25% over the two-year period despite the 0% arithmetic average.
Geometric average return
The geometric average, also called the compound annual growth rate, identifies the constant annual return that connects the beginning value to the ending value:
Geometric average = [(1 + R₁)(1 + R₂)…(1 + Rₙ)]^(1/n) − 1
For the +50% and -50% example, the compound return equals about -13.40% per year. Indeed, compounding -13.40% twice turns $100 into $75.
Why the two averages diverge
Volatility creates a gap between arithmetic and geometric averages. Large losses require disproportionately larger gains to recover.
| Loss | Gain required to break even |
|---|---|
| -10% | 11.11% |
| -20% | 25.00% |
| -30% | 42.86% |
| -40% | 66.67% |
| -50% | 100.00% |
| -75% | 300.00% |
Accordingly, the S&P 500 arithmetic average of 11.85% exceeded its 10.02% compound return. The difference grew much larger for small caps: 17.78% arithmetic versus 11.98% compound. Their extreme variability created a stronger volatility drag.
Which average should investors use?
Use the geometric average to describe actual historical wealth growth across many years. Meanwhile, analysts may use an arithmetic average when estimating the expected return for one independent future period. Even then, estimation error remains substantial.
For a retirement projection, a conservative range often communicates uncertainty more honestly than one fixed return. In addition, projections should account for inflation, fees, taxes, and the order of returns when withdrawals occur.
Stock Market Returns and the Risk Premium
Investors generally demand compensation for bearing uncertainty. The extra historical return on a risky asset above a safer benchmark is called a risk premium, so stock market returns and risk premiums must be interpreted together.
Equity risk premium over Treasury bills
One historical measure subtracts the Treasury-bill return from the stock return each year:
Equity risk premium = Stock market return − Risk-free return
From 1928 through 2025, the arithmetic average S&P 500 return exceeded the average three-month Treasury-bill return by about 8.44 percentage points per year in Damodaran’s dataset. Against 10-year Treasury bond returns, the average difference was approximately 7.03 percentage points.
That premium was not delivered steadily. During many individual years, stocks underperformed bills or bonds. Thus, “premium” describes an average historical reward for risk, not an annual entitlement.
Historical premium versus implied premium
A historical premium looks backward. An implied equity risk premium, on the other hand, uses current market prices and expected future cash flows to infer the return investors demand today.
Damodaran publishes a separate U.S. implied-premium series dating to 1960 and updates its assumptions and outputs. Review the NYU Stern historical implied equity risk premium data.
These measures can differ because markets, interest rates, valuations, and expectations change. Therefore, analysts should label the method, the risk-free benchmark, the averaging approach, and the sample period whenever they report an equity premium.
Risk premiums are uncertain estimates
A sample average contains noise, especially when annual returns vary widely. Adding more historical years increases the sample size, yet very old observations may reflect a different economic and regulatory environment. Conversely, using only recent years makes the estimate more relevant but less statistically stable.
No method eliminates that trade-off. As a result, valuation professionals often compare several estimates instead of treating one historical average as indisputable.
Measuring Stock Market Return Variability
Average stock market returns describe the center of a historical sample. Risk analysis also asks how far individual observations moved around that center.
Frequency distribution
A frequency distribution groups returns into intervals and counts how often each interval occurred. This simple table reveals outcomes that an average hides.
For instance, the 1928–2025 stock series produced both a -43.84% minimum and a +52.56% maximum. Its 11.85% arithmetic average sits between them, yet an investor never received exactly the average in most years.
Variance
Variance measures the average squared distance between each return and the sample mean:
Sample variance = Σ(Rᵢ − R̄)² ÷ (n − 1)
Squaring prevents positive and negative deviations from canceling. Nevertheless, variance uses squared units, so it feels less intuitive than a percentage return.
Standard deviation
Standard deviation equals the square root of variance:
Sample standard deviation = √[Σ(Rᵢ − R̄)² ÷ (n − 1)]
Because standard deviation returns to the original units, investors can read it as a percentage. The 1928–2025 S&P 500 series had annual standard deviation of about 19.40%. Small caps reached approximately 37.95%, while three-month Treasury bills registered only 3.04%.
A worked four-year example
Suppose a portfolio earned 12%, -8%, 20%, and 4% across four years.
| Year | Return | Deviation from 7% mean | Squared deviation |
|---|---|---|---|
| 1 | 12% | 5 points | 25 |
| 2 | -8% | -15 points | 225 |
| 3 | 20% | 13 points | 169 |
| 4 | 4% | -3 points | 9 |
| Total | — | 0 | 428 |
The arithmetic mean equals 7%. Next, dividing 428 by three gives a sample variance of 142.67 percentage-points squared. Finally, taking the square root produces a sample standard deviation of about 11.94%.
Standard deviation is useful but incomplete
Volatility treats upside and downside deviations symmetrically. Most investors, however, do not dislike a surprise gain as much as an equally large surprise loss. Moreover, standard deviation does not directly measure permanent loss, illiquidity, default, or the inability to meet a near-term cash need.
Other useful measures include maximum drawdown, downside deviation, value at risk, expected shortfall, beta, correlation, and the probability of missing a financial goal. Each measure answers a different question.
Are Stock Market Returns Normally Distributed?
Textbook models often introduce a bell-shaped normal distribution because it connects means, standard deviations, and probabilities cleanly.
The 68–95–99.7 rule
In a perfectly normal distribution:
- about 68% of observations fall within one standard deviation of the mean;
- about 95% fall within two standard deviations;
- about 99.7% fall within three standard deviations.
If annual returns had a mean of 11.85% and standard deviation of 19.40%, a simplified normal model would place roughly 68% of outcomes between -7.55% and 31.25%. Similarly, about 95% would fall between -26.95% and 50.65%.
Real markets have tails, skewness, and changing volatility
Actual stock market returns do not follow a perfect normal curve. Crashes and extreme rallies can occur more often than a simple bell curve predicts. In addition, volatility clusters: calm periods can persist, while turbulent episodes can arrive together.
The historical annual sample itself demonstrates the limitation. Its best observation, 52.56%, slightly exceeds the upper two-standard-deviation boundary calculated above. Meanwhile, the worst observation, -43.84%, falls deeply below the corresponding lower boundary.
Therefore, the normal distribution works as an introductory model, not a complete map of market risk. Investors should combine it with scenario analysis and stress tests that deliberately examine extreme outcomes.
Why monthly and daily data look different
Return frequency changes the shape and interpretation of the distribution. Daily stock market returns usually appear small, but rare daily shocks can dominate a year’s result. Annual data smooth many short-term movements, yet a 98-year annual series contains only 98 observations.
Furthermore, returns overlap when analysts calculate rolling multi-year periods. Those overlapping observations are not independent. As a result, an apparently large dataset may contain less unique information than the number of rolling periods suggests.
Stock Market Returns Compared With Other Assets
Cross-asset comparisons help explain why portfolio construction involves more than selecting the highest historical stock market returns.
Large-company U.S. stocks
The S&P 500 represents leading large U.S. companies and covers roughly 80% of available U.S. market capitalization, according to S&P Dow Jones Indices. Read the official S&P 500 overview.
Over its live history since March 1957, S&P Dow Jones Indices reports an annualized price return near 7% and a total return near 10%. Moreover, its 2025 brochure identifies 12 bear markets, an average peak-to-trough decline of about 33%, and an average recovery time near 13 months. See the official S&P 500 historical brochure.
Small-company stocks
Small stocks delivered the highest long-run averages in the table, but they also experienced much wider dispersion. Their worst year lost 53.94%, while their best year gained 146.60%.
Liquidity, business concentration, financing constraints, and economic sensitivity can make smaller companies riskier. Consequently, investors should not interpret the small-cap premium as a guaranteed bonus.
Treasury bills
Three-month Treasury bills showed the smallest variability and no negative nominal calendar-year observations in this dataset. Still, bills can lose purchasing power after inflation. They also carry reinvestment risk because future short-term rates may decline.
Cash-like investments can support emergency reserves and short-horizon spending. However, their lower compound return makes them less effective as the only vehicle for very long-term growth.
Long-term Treasury bonds
Government bonds reduce credit risk, but they still face interest-rate risk. When market yields rise, existing fixed-rate bonds generally lose value because their older coupons become less attractive.
That mechanism explains why the bond return can be negative even when the quoted yield remains positive. Damodaran’s series estimates total return from coupon income and repricing, so it should not be confused with the 10-year yield itself.
Corporate bonds
Corporate bonds add credit and liquidity risks to interest-rate risk. In return, investors generally demand a yield spread above comparable Treasuries.
The SEC notes that bond funds can face credit risk, interest-rate risk, and prepayment risk. Review the SEC’s explanation of bond-fund risks.
Gold
Gold delivered a 5.61% compound return in the 1928–2025 series, yet its annual volatility reached 21.51%. Additionally, gold produced losses in 34 of 98 years.
Unlike a profitable company, gold does not generate earnings or dividends. Its return depends primarily on price changes. Nevertheless, some investors use it as a diversifier or a potential hedge against particular monetary and geopolitical risks.
Residential real estate
The home-price series compounded at 4.20%, but that comparison requires caution. A house can provide rent or housing services, while it also creates expenses such as maintenance, insurance, property tax, and transaction costs. Leverage further magnifies both gains and losses on the owner’s equity.
Therefore, a price index cannot fully represent an individual’s property return. It remains useful for studying home-price appreciation, not for declaring whether stocks or homes universally make the better investment.
How Dividends Shape Stock Market Returns
The contrast between a price index and a total-return index matters more as the horizon lengthens. Therefore, any discussion of stock market returns should state whether it includes dividends.
Reinvested distributions buy more shares
When investors reinvest dividends, each payment purchases additional shares. Those shares may then produce their own future dividends and price gains. Thus, reinvestment creates a compounding loop.
This process does not make dividends free money. A stock’s price generally adjusts when cash leaves the company. Even so, total return correctly counts both changes in share price and cash received.
Compounding is multiplicative
Future value follows this formula when the return remains constant:
Future value = Initial investment × (1 + annual return)^years
At 5%, $10,000 becomes approximately $16,289 after 10 years. At 8%, it grows to about $21,589. Meanwhile, at 10%, it reaches about $25,937. A few percentage points make a large difference because every year’s gain builds on the prior balance.
Fees also compound
Expenses reduce the amount that remains invested. The SEC warns that even small differences in fees can create substantial differences in long-term results. Read the SEC guide to mutual-fund fees and performance.
Suppose two portfolios earn 8% before costs. One costs 0.10% annually, while the other costs 1.50%. Before taxes, $100,000 would grow for 30 years at approximate net rates of 7.90% and 6.50%:
| Annual fee | Assumed net return | Approximate value after 30 years |
|---|---|---|
| 0.10% | 7.90% | $979,000 |
| 1.50% | 6.50% | $661,000 |
| Difference | — | about $318,000 |
The illustration assumes constant returns and a fee deducted annually. Real markets will vary, but the mathematical effect of recurring costs remains.
Stock Market Returns After Inflation and Taxes
Published stock market returns rarely equal the amount an investor can spend.
Inflation reduces purchasing power
A nominal gain can coexist with a real loss. For example, a 4% nominal return during 6% inflation produces an exact real return of about -1.89%.
Consequently, long-term plans should use real returns when the goal involves future purchasing power. Retirement projections expressed only in future dollars can look impressive while hiding how much goods and services those dollars will buy.
Taxes depend on the account and investor
Dividends, interest, and realized capital gains may receive different tax treatment. Additionally, tax-deferred and tax-free accounts can alter the timing or size of the tax burden.
Because rules differ by country, account, income, and holding period, a universal after-tax historical return would mislead readers. Investors should calculate their own situation or consult a qualified tax professional.
Behavior can create a return gap
An index never panics, needs emergency cash, or abandons a plan. People can do all three. Chasing recent winners and selling after declines may cause personal results to trail the funds they own.
Therefore, an appropriate portfolio is not merely the one with the highest theoretical expected return. It is the one an investor can fund, maintain, and use through stressful markets without taking unacceptable risk.
Sequence-of-Returns Risk
The order of gains and losses does not change a lump sum’s ending value when no cash enters or leaves. Once withdrawals begin, however, order matters greatly.
Accumulation without cash flows
Imagine two returns, +20% and -10%. Applied to $100, either order ends at $108:
$100 × 1.20 × 0.90 = $108
Multiplication gives the same result when the factors switch places.
Withdrawals change the equation
Now suppose a retiree withdraws money after the first year. A large early loss forces the person to sell more assets at depressed prices. Consequently, fewer shares remain available for a later recovery.
This sequence risk can matter more than the average return during the first years of retirement. Practical responses may include a cash reserve, flexible withdrawals, broader diversification, or a less volatile near-term spending portfolio.
Diversification and Asset Allocation
Historical stock market returns show why investors seek reward. Meanwhile, historical losses show why they diversify.
Diversification reduces concentration risk
Diversification spreads money among investments so one loss may have less influence on the full portfolio. The SEC summarizes the idea as avoiding dependence on one basket. Read Investor.gov’s definition of diversification.
Owning 20 technology companies does not necessarily create strong diversification because those businesses may react to the same economic forces. By contrast, diversification can operate across companies, sectors, countries, asset classes, maturities, and sources of return.
Asset allocation sets the broad mix
Asset allocation divides a portfolio among categories such as stocks, bonds, and cash. According to the SEC, the appropriate mix depends on risk tolerance and the investment timeframe. See the SEC’s 2026 investor guidance.
An investor saving for a goal next year faces a different problem from someone building retirement wealth for 35 years. Therefore, historical return rankings alone cannot select an allocation.
Correlation matters
A diversified asset contributes value when it behaves differently from the rest of the portfolio, particularly during difficult periods. Correlation measures how two return series move together, from -1 to +1.
Low or negative historical correlation can reduce portfolio volatility. Still, correlations change, and several risky assets may fall together during a crisis. Diversification manages risk; it does not guarantee against loss.
Common Mistakes When Reading Stock Market Returns
Long tables of stock market returns create an appearance of precision. Yet several methodological choices can change the answer.
Mistake 1: Confusing price return with total return
An S&P 500 price chart excludes dividends. Therefore, using it to describe an investor’s complete long-term return understates the result when dividends were available and reinvested.
Mistake 2: Treating an arithmetic average as compound growth
The arithmetic average does not tell you how fast wealth actually compounded. Use CAGR for a multi-year investment journey.
Mistake 3: Ignoring inflation
Nominal dollars from different decades do not have equal purchasing power. Accordingly, real-return analysis becomes essential for long-term goals.
Mistake 4: Comparing unlike datasets
A stock total-return index, a home price-only index, and a bond yield measure different things. Before comparing them, confirm whether each figure includes income, costs, leverage, and price change.
Mistake 5: Assuming the S&P 500 represents every stock
The S&P 500 focuses on large U.S. companies. It does not represent small stocks, private companies, or the full global market. Moreover, its constituents and sector weights evolve over time.
Mistake 6: Ignoring survivorship and selection
The United States became one of the world’s most successful equity markets. Choosing it after observing that success can create selection bias when making global claims. Likewise, modern index histories may rely on backfilled predecessor data before an index’s official launch.
The S&P 500 launched on March 4, 1957, even though researchers often use predecessor series to study earlier periods. Therefore, authors should distinguish live index history from reconstructed historical data.
Mistake 7: Forecasting the next decade from the last one
Recent performance feels vivid, but valuations, interest rates, profits, inflation, and starting conditions change. A strong decade can be followed by a weak one, and vice versa.
Mistake 8: Believing diversification removes market risk
Broad diversification can reduce company-specific risk. Nevertheless, it cannot eliminate economy-wide market declines.
Mistake 9: Forgetting taxes, fees, and trading friction
Gross index performance is not personal net performance. Even modest recurring costs reduce compound wealth over time.
Mistake 10: Treating risk as one number
Standard deviation measures dispersion, not every form of danger. Time horizon, liquidity, drawdown, default, inflation, and behavior can matter just as much.
Practical Ways to Use Historical Stock Market Returns
Historical stock market returns help most when investors use them to test plans instead of predict exact outcomes.
Build a range, not a single forecast
Start with conservative, central, and optimistic return assumptions. Next, calculate results after inflation and fees. Finally, test whether the goal survives weaker returns.
For example, a planner might compare 3%, 5%, and 7% real returns rather than promise one exact number. The appropriate assumptions will depend on the portfolio and horizon.
Stress-test bad starting periods
Model a sharp early decline, a slow recovery, and persistent inflation. If one scenario makes the plan fail immediately, the investor may need more savings, more time, flexible spending, or less near-term risk.
Match assets to the spending date
Money needed soon generally cannot wait decades for a recovery. Conversely, money reserved for a distant goal may need growth to overcome inflation.
A time-segmented plan can connect cash and high-quality short-term bonds to near expenses while assigning diversified equities to longer horizons. Still, no structure removes all uncertainty.
Rebalance deliberately
Rebalancing restores a chosen asset mix after market movements change portfolio weights. In practice, investors may rebalance on a schedule or when allocations cross predetermined bands.
This discipline can reduce unintended risk. However, taxes and trading costs should inform the approach in taxable accounts.
Compare investments on consistent terms
Whenever you evaluate a return claim, ask:
- Does it include dividends or interest?
- Is the number nominal or inflation-adjusted?
- Which average does it use: arithmetic or geometric?
- What dates define the sample?
- Does it include fees and taxes?
- Which index or asset actually produced it?
- How volatile was the path?
- What was the worst drawdown?
Those questions prevent most misleading comparisons.
Stock Market Returns: Facts and Curiosities
Historical stock market returns contain several counterintuitive lessons.
The average year is not a typical promise
The S&P 500 arithmetic average reached 11.85%, but annual outcomes ranged across almost 97 percentage points from worst to best. Accordingly, a forecast near 12% should never imply that most individual years will land near 12%.
Most years were positive, but losses shaped the experience
Only 26 of 98 years were negative. Even so, a few severe declines greatly influenced compound growth and investor behavior.
Small caps had a huge average-to-compound gap
Their arithmetic average exceeded their compound return by 5.80 percentage points. By comparison, the large-stock gap was 1.84 points. This difference offers a vivid example of volatility drag.
Treasury bills avoided nominal annual losses but not real losses
Every calendar-year T-bill observation in the dataset was nominally positive. Nevertheless, inflation sometimes exceeded the bill return, so purchasing power could still fall.
Bonds can lose money when rates rise
Positive coupon income does not guarantee a positive total return. Indeed, 10-year Treasury bonds lost 17.83% in 2022 in this series as yields rose sharply.
Gold’s spectacular best year did not produce the best compound return
Gold’s maximum annual gain reached 126.55%, more than twice the S&P 500 maximum. Yet gold compounded at 5.61%, compared with 10.02% for large-company stocks. A dramatic best year says little about the full journey.
The S&P 500 changes with the economy
The index does not hold a frozen list of companies. S&P Dow Jones Indices updates constituents under its methodology, while market-cap weighting allows successful firms to become more influential. Consequently, the index evolves with corporate America.
Frequently Asked Questions About Stock Market Returns
What is the average annual stock market return?
Using Damodaran’s 1928–2025 U.S. large-stock total-return series, the arithmetic average was 11.85% and the compound annual growth rate was 10.02%. However, the result changes with the index, dates, dividend treatment, inflation adjustment, and fees.
Does the S&P 500 always return 10% a year?
No. The historical compound rate was close to 10% across this particular 98-year sample, but annual returns varied widely. Past performance also does not guarantee future returns.
What was the worst year in the dataset?
The worst S&P 500 and predecessor-series total return was -43.84% in 1931. By contrast, the best was 52.56% in 1954. Because the official S&P 500 launched in 1957, earlier observations use reconstructed predecessor data.
What is a good return on stocks?
A “good” return depends on risk, inflation, fees, taxes, the benchmark, and the time horizon. Beating a safe cash return by two percentage points may be attractive with modest risk, while the same return could disappoint for a concentrated and volatile strategy.
Why do sources report different S&P 500 returns?
One source may report price return, while another reports total return. Furthermore, sources can use different dates, frequencies, data revisions, or annualization methods. Always read the methodology.
What is the difference between return and yield?
Yield measures income relative to price or principal. Total return combines income with the change in market value. Therefore, a bond can have a positive yield but a negative total return when its price falls enough.
Is volatility the same as risk?
No. Volatility measures how widely returns fluctuate. Risk also includes permanent loss, inflation, default, concentration, illiquidity, behavioral mistakes, and failure to meet a goal.
Are stock returns normally distributed?
Not perfectly. A normal model can help explain averages and standard deviations, but real returns show fat tails, skewness, and changing volatility. Stress tests should therefore complement bell-curve assumptions.
Do dividends matter?
Yes. Dividends form part of total return, and reinvesting them can meaningfully increase long-term wealth. A price-only index omits that component.
Should investors use historical averages in retirement calculators?
They can provide a starting point, but one average is not enough. A more robust plan uses a range of real returns, subtracts fees, models withdrawals, and tests unfavorable return sequences.
Final Takeaways on Stock Market Returns
Stock market returns rewarded long-term U.S. investors handsomely across the 1928–2025 historical sample. Large-company stocks compounded near 10.02% a year with dividends, turning a theoretical $100 into roughly $1.16 million before inflation, taxes, and costs.
That reward came with genuine uncertainty. Annual volatility approached 19.40%, more than one-quarter of calendar years produced a loss, and the worst year erased 43.84%. Small stocks offered even higher averages, yet their much greater volatility widened the difference between headline average and compound growth.
Therefore, history supports neither blind optimism nor permanent fear. It supports disciplined expectations. Investors can use long-run evidence to understand compounding, compare consistent measures, estimate a range of outcomes, diversify thoughtfully, control costs, and prepare for periods that look nothing like the average.
Most importantly, historical returns should guide planning rather than masquerade as predictions. A resilient investment strategy connects risk to time horizon, liquidity needs, behavior, and personal goals.
Sources and Methodology
- Aswath Damodaran, NYU Stern, Historical Returns on Stocks, Bonds and Bills. January 2026 spreadsheet edition with annual observations through 2025.
- Aswath Damodaran, NYU Stern, Historical Implied Equity Risk Premiums.
- S&P Dow Jones Indices, S&P 500 official index overview.
- S&P Dow Jones Indices, S&P 500 brochure and live-history statistics. Data cited from the June 30, 2025 edition.
- Federal Reserve Bank of St. Louis, S&P 500 series and price-index methodology.
- U.S. Bureau of Labor Statistics, CPI Inflation Calculator.
- U.S. Securities and Exchange Commission, 2026 investor guidance on asset allocation, diversification, and fees.
- U.S. Securities and Exchange Commission, Diversification glossary.
Disclaimer: This article provides general educational information, not individualized investment, tax, or legal advice. Historical performance does not guarantee future results, and every investment can lose value.
