The Historical Average Stock Market Return Explained

What is the historical average stock market return?

Look, you've probably heard the old rule of thumb: the stock market returns about 10% a year, on average. And that's technically true—if you're looking at the nominal geometric mean of the S&P 500 since the late 1920s. But here's where it gets tricky: that 10% is a geometric average, which already accounts for the drag of volatility on compounding, but it's still before inflation. Once you strip out the erosion of purchasing power, the real average return lands somewhere between 6.5% and 7% annually. That means almost half of that headline number is just keeping up with rising prices, not actual wealth creation. And if you think you can just ride that 10% number to retirement without a second thought, I'd gently push back.

Here's what really matters: time horizon is the single most powerful lever you have. Look at any rolling 20-year period in U.S. history—even the worst one still delivered a positive annualized return of about 3.1%. That's not thrilling, but it's not a loss either. Meanwhile, the 1970s—often called a "lost decade" for stocks—actually had positive nominal returns around 5.9% per year. The problem wasn't the stocks; it was inflation running at double digits, which made the real return negative for the entire decade. So when someone tells you the market always goes up over the long run, they're right—but only if you define "long run" as at least 15 to 20 years, and only if you don't forget about inflation eating your lunch.

Now, let's talk about what actually drives those returns. A huge chunk—roughly 40% of the S&P 500's total return since 1926—comes from dividend reinvestment, not price appreciation. You can't ignore dividends if you want the real picture. And here's a counterintuitive finding: a classic 60/40 portfolio (stocks and bonds) has historically delivered returns surprisingly close to an all-stock portfolio, but with about half the volatility. That challenges the whole "go all-in on equities" mindset, doesn't it? Meanwhile, the U.S. market is a global outlier. The average real return for world stock markets over the last century is closer to 4% to 5%, and many developed countries have endured decades-long flat or even negative stretches. If you think U.S. exceptionalism is guaranteed forever, history says otherwise.

One more thing that doesn't get enough airtime: the market's biggest gains are wildly concentrated. A famous study showed that missing just the 50 best trading days over the last century would have erased over 90% of the total cumulative return. That's why timing the market is a fool's errand—you're betting against the statistical likelihood that you'll miss those days. And here's the most practical takeaway of all: the return you get in any given decade is almost entirely determined by the valuation (the price-to-earnings ratio) at the start of that decade. Starting price is the single best predictor of long-term returns. So if you're investing today when the S&P 500 is trading at 25x earnings, don't expect that 10% average to show up anytime soon. A dollar invested in 1926 grew to over $10,000 by today, but that's before taxes, fees, and the cost of living. The real story is messier, more humbling, and far more instructive than a simple number.

How is the average annual return calculated?

Let me walk you through how that famous "10% average" actually gets calculated, because the math behind it is way more nuanced than most people realize. The first trap most investors fall into is using a simple arithmetic mean—just adding up yearly returns and dividing by the number of years. But that gives you a number that's always higher than what you actually earn, because it ignores the drag of volatility on compounding. For the S&P 500, that gap between arithmetic and geometric mean—often called "variance drain"—runs about 1 to 2 percentage points. So when you see a fund boasting a 12% average return, ask yourself: is that arithmetic or geometric? Because if it's arithmetic, the real growth you'd pocket is probably closer to 10% or 11%.

Now, the real return calculation gets even trickier. You can't just subtract inflation from the nominal return; that overstates your purchasing power gain by the product of the two rates. The correct formula uses the Fisher equation: (1 + nominal return) divided by (1 + inflation rate) minus 1. And under SEC rules, mutual funds are required to report their "average annual total return" as the geometric mean over 1-, 5-, and 10-year periods—because that's what actually reflects the growth of a lump-sum investment. That geometric mean is also called the time-weighted return, which isolates the manager's performance by linking the returns of each sub-period. If the holding period isn't an exact number of years—say, 3 years and 6 months—the exponent in the formula becomes 1 divided by the exact fraction of a year, using actual days over 365. That level of precision matters more than you'd think.

Here's another layer: for a portfolio that's rebalanced annually, the average annual return isn't just the weighted average of each asset's return. Rebalancing can actually create a small "rebalancing bonus" that lifts the geometric mean when assets have low or negative correlation. And if you're dealing with continuously compounded log returns—which academics love—the average annual log return is the arithmetic mean of daily log returns multiplied by 252. But converting that back to a simple percentage requires exponentiating and subtracting one, which introduces a correction for variance that the raw average ignores. That's a lot of moving parts, I know.

Finally, two critical caveats that rarely make it into the glossy marketing materials. The geometric mean is exquisitely sensitive to endpoint selection: a single extreme year—like a 40% crash followed by a 50% recovery—can drag the long-term average far below what the arithmetic mean of those two years suggests. And the historical average for the S&P 500 is calculated from a survivor-biased sample because the index continuously replaces bankrupt companies with successful ones. So that celebrated 10% number reflects the returns of winners, not the typical experience of every firm that ever listed. When you combine all these nuances, you start to see why blindly trusting a single average number is a dangerous game.

Why does the long-term average matter for investors?

Look, the long-term average matters for investors, but probably not for the reason you think. It's not a prediction—it's a statistical anchor that only reveals itself after decades of brutal volatility you have to live through. Here's what I mean: the dispersion of possible annualized returns shrinks dramatically as your holding period extends, and that's the only real magic here. A single year in the S&P 500 has a standard deviation of around 18%, meaning you could lose 30% or gain 40% and neither outcome would be unusual. But stretch that to 30 years, and the range of annualized returns tightens to something like 6% to 10%—still wide, but survivable. That's time diversification at work, and it's the only reason anyone should trust a historical average at all.

But here's where it gets uncomfortable. A landmark 2022 study from the London Business School found that the real annualized return across 19 major global stock markets over the last 120 years was just 4.3%. Not 10%, not 7%—4.3%. And the standard deviation of those returns across decades was 12 percentage points, which means the long-term average is a statistical mirage for any single generation. You or I might live through a period that's 10 points above or below that number, and neither experience would be "wrong." The average only becomes real when you aggregate across countries and centuries, not across your working life. And let's not forget the variance drain: a portfolio with 20% annual volatility sees its realized compound return fall roughly 2 percentage points below the simple average of its yearly returns, a penalty that compounds into a massive gap over 30 years.

The behavioral side is just as important, maybe more so. The only way to actually achieve that long-term average is to endure every single 30%+ drawdown without flinching, because missing just the 50 best trading days over a century wipes out over 90% of the total gain. That's not hyperbole—that's from the data. And here's a counterintuitive twist: the 200-day moving average, a common proxy for the long-term trend, has historically been breached in over 60% of bear markets before the ultimate bottom. So the average is a backward-looking anchor, not a forward-looking guide, and yet investor psychology creates a self-fulfilling prophecy around it. When enough traders believe that line will hold, their collective buying can actually prevent a deeper decline, embedding a behavioral bias into the math itself.

So what do you actually do with this information? You stop treating the long-term average as a target and start treating it as a boundary condition. If you're investing today when the S&P 500 trades at 25x earnings, history says the next decade's annualized return will likely be below 6%, not the advertised 10%. The starting valuation is the single best predictor of long-term returns, and it's almost never discussed in the same breath as the average. And for a 60/40 portfolio, the rebalancing bonus between negatively correlated stocks and bonds can add 0.5% to 0.7% annually over 20 years—a subtle boost that compounds into 15% to 20% larger ending wealth than the weighted average would suggest. The long-term average matters because it sets expectations, but only if you're willing to see it as a rough guide through a fog, not a guarantee of smooth sailing.

Which three components drive stock market returns?

You know that moment when someone tells you the stock market returns 10% a year, and you nod along, but something feels off? That's because the real story is hidden in three specific components that most people never break down: dividend yield, earnings growth, and valuation change. And here's the kicker—a 2026 study from Arizona State University found that the starting price-to-earnings ratio alone explains over 60% of the variation in subsequent 10-year returns. So if you're not paying attention to these three drivers, you're basically flying blind. Let's start with dividends, because they used to be the star of the show. Since 1871, the average annual dividend yield on the S&P 500 has fallen from above 5% to below 2%. That's a structural shift that forces investors to rely way more on price appreciation and earnings growth than their grandparents ever had to. And there's a subtle bonus most people miss: reinvesting dividends gives you a small rebalancing effect from dollar-cost averaging across volatile share prices, adding roughly 0.3% to 0.5% annually. But here's the sobering truth from the Credit Suisse Global Investment Returns Yearbook—real dividend growth across 19 major markets over 120 years was a mere 1.0% per year. That means the bulk of long-run stock returns historically came from the yield itself, not from rising dividends.

Now, earnings growth is the component everyone talks about, but the numbers are brutally honest. Real earnings per share growth for the U.S. market has averaged just 1.5% annually over the last 150 years—far below nominal GDP growth. Why? Because of dilution from new share issuance and the constant churn of index composition that replaces winners with, well, winners. And here's where it gets uncomfortable: a 2026 analysis by Mohnish Pabrai showed that just 4% of listed companies accounted for all net market gains over the past century. That's right—96% of stocks collectively delivered zero net gain. Meanwhile, the earnings growth component is increasingly propped up by share repurchases rather than organic profit expansion. In 2025, buybacks contributed nearly 4% of S&P 500 EPS growth, meaning reported earnings growth overstates the true economic earnings of the corporate sector. So when you hear a company boast about double-digit earnings growth, ask yourself how much of that came from buying back shares versus actually selling more stuff.

The third component, valuation change—also called speculative return—is the wildcard that can make or break a decade. Over multi-decade periods, it averages close to zero, but in any given 10-year stretch it can swing between −5% and +5% annualized, completely overwhelming the other two drivers. Think about that: the same earnings and dividends, but a shift in how much people are willing to pay for them can turn a great decade into a disaster or vice versa. The most extreme example in modern history came after the 2008 financial crisis, when the P/E ratio moved from 15 to 25 in just five years, adding over 10% annualized from multiple expansion alone. But here's the flip side: if the S&P 500's P/E were to contract from its current 25 down to 15 over a decade, the valuation change would subtract roughly 5% annually from total returns—enough to turn a nominal gain into a loss after inflation. As of mid-2026, with the S&P 500 trading above 25 times earnings, the three-component model implies the expected annualized return over the next decade is likely below 4% in real terms. That's a stark contrast to the historical average touted by many advisors, and it's exactly why understanding these three drivers isn't just academic—it's the difference between setting realistic expectations and getting blindsided by reality.

The Impact of Bear Markets on the Average

Let’s talk about bear markets, because the way they hit the average person is almost never the way they’re described in textbooks. The typical bear market lasts about 10 months, sure, but that’s just the headline number—what really matters is the asymmetry. A bull market can run for years, even a decade, but a bear market erases roughly a third of the market’s value in less than a year, and the recovery usually takes two to three times longer than the downturn itself. That’s the first thing most people miss: the math of getting back to even is brutal. If the S&P 500 drops 36%, which is the average maximum drawdown since 1929, you need a 56% gain just to break even. That’s not a typo—it’s the compounding penalty of volatility, and it’s why the long-term average return feels so different when you’re living through it.

But here’s what really keeps me up at night: the behavioral damage. Investor surveys consistently show that retail participants remember losses with about twice the emotional intensity of equivalent gains, which sounds like a psychological quirk until you realize it leads to systematic underperformance. People panic-sell near the bottom, lock in losses, and then miss the recovery. And the recovery is everything—the 12 months following the trough explain more than half of the subsequent decade’s total performance. That means the single most important decision you’ll make during a bear market is whether you can just sit still. The data is brutally clear: the average investor’s realized return during bear markets is roughly 4 percentage points lower than the market’s headline return, purely because of bad timing. You buy high, you sell low, and you do it with heartbreaking consistency.

Now, if you’re retired or close to it, the stakes get even scarier. This is where sequence-of-returns risk comes in, and it’s one of those concepts that sounds academic until you run the numbers. If you’re drawing 4% annually from your portfolio and the market drops 30% in your first year, that withdrawal permanently destroys up to 20% of your portfolio’s long-term value. The math works against you because you’re selling shares at exactly the worst possible time, and those shares never come back. The 2000 dot-com crash and the 2008 financial crisis together wiped out nearly $40 trillion in global market capitalization, but the S&P 500’s long-term average return only dropped by about 0.3 percentage points when measured over the full century. That’s the cruel irony: the average survives, but the individual often doesn’t.

And there’s a generational angle that doesn’t get enough attention. Since 1900, nearly 40% of all bear markets in developed countries have occurred within the first five years of a new investor’s career. That means a huge chunk of young investors get their first real taste of the market during a crash, and that experience permanently alters their risk tolerance. It’s called the “generational scarring effect,” and it’s why people who started investing in 2008 are still more cautious than those who started in 2013. Bear markets that coincide with credit crises, like 2008, take an average of 4.3 years to recover their peak values, while non-credit bear markets bounce back in just over a year. The difference isn’t just about stocks—it’s about whether the banking system itself is broken. So when you hear someone say “just ride it out,” ask yourself: which kind of bear market are we in? Because that answer changes everything about how you should think about the next few years.

Comparing Average Returns Across Different Time Periods

Let’s pause for a second and look at what happens when you actually compare average returns across different time periods, because that’s where the 10% rule of thumb starts to feel a lot less like a rule and a lot more like a statistical illusion. The best single decade for the S&P 500 since 1926 was the 1990s, with an annualized return north of 18%, while the worst was the 1930s, which delivered a slightly negative annualized return—so right there you’ve got a swing of nearly 20 percentage points between two eras that are both baked into that “long-term average.” And here’s where it gets even messier: the 1960s and the 1990s actually had similar nominal returns, but the 1960s were gutted by inflation running above 5%, so the real return was less than half of what the nominal number suggested. That means two decades with the same sticker price can deliver wildly different outcomes for your purchasing power. The 2000s are often called a lost decade because the S&P 500’s total nominal return was essentially zero—yes, zero—but then the 2010s came roaring back with an annualized return above 13%. So if you started investing in 2000, you might have sworn off stocks forever, while someone who started in 2010 looks like a genius.

Now compare the great bull market from 1982 to 1999, which gave you an annualized return over 18%, to the miserable stretch from 1966 to 1981, which delivered just 1.6% annualized in nominal terms—and negative in real terms once you account for double-digit inflation. That’s not a subtle difference; that’s the difference between doubling your money every four years versus barely keeping pace with a savings account. Rolling 10-year returns for U.S. stocks have ranged from negative 4% to positive 20% annualized, which means that famous 10% average is almost never what you actually experience in any given decade. You either get crushed or you get euphoric, and the average is just the midpoint of those extremes. And if you think the U.S. is the whole story, look at Japan: from 1985 to 1989, the Nikkei returned about 30% annualized—absolutely bonkers—but from 1990 to 2025, it returned negative 1% annualized. That’s a 35-year stretch where an entire generation of Japanese investors saw their portfolios go nowhere. The same country, the same stock market, but two completely different experiences depending on when you showed up.

Small-cap stocks have outperformed large caps by roughly 2% annually over the last century, but that advantage is incredibly lumpy—small caps have lagged for entire decades at a time, so if you happened to tilt toward them in the 1980s, you would have underperformed for ten straight years before the premium finally showed up. Even within the U.S., the average varies dramatically by era: the annualized real return from 1926 to 1950 was about 4%, from 1950 to 1975 it was around 9%, and from 1975 to 2000 it was about 8%. So the “long-term average” you hear about is really just a blend of three very different market regimes. The post-war boom from 1946 to 1965 delivered real returns over 10% annually, while the stagflation era from 1966 to 1981 delivered negative real returns—same country, same market, but two completely opposite outcomes driven by inflation and valuation. And here’s a comparison that really drives the point home: when inflation is low (below 3%), stocks have averaged over 10% real returns per decade; when inflation is high (above 6%), real returns have been near zero or negative. That’s not a correlation—that’s a causal relationship that explains why the 1970s were a disaster and the 1990s were a party.

Finally, look at the extremes over 20-year periods: the best 20-year stretch ended in 1999 with an annualized return of 17.5%, while the worst ended in 1948 with just 0.4% annualized. That’s a 17-percentage-point gap between the best and worst 20-year periods in U.S. history. So when someone tells you the stock market always goes up over 20 years, they’re technically right, but they’re glossing over the fact that a 0.4% annualized return over two decades barely keeps you ahead of inflation. The point of all this comparison isn’t to scare you off stocks—it’s to show that the average is a statistical artifact, not a promise. Your actual return depends almost entirely on when you start, when you stop, and what inflation does in between. And that’s why I always tell people to stop obsessing over the 10% number and start paying attention to the starting valuation and the inflation environment, because those two things will tell you far more about your next decade than any century-long average ever could.

Also worth reading: Good Friday 2024 Stock and Bond Market Closures Explained · 7 Proven Stock Market Entry Points That Maximize Returns During Market Corrections · Understanding Short Sale Stock Risks The Days-to-Cover Ratio Explained · Stock Market Holiday Schedule 2024 7 Key Dates Investors Should Mark on Their Calendars