A Century of U.S. Stock Returns: Expect the Unexpected
In honor of America’s 250th birthday, I decided to do a deep dive into U.S. stock returns over the past century. My goal was to answer the following question:
If you had invested in U.S. stocks at a random point in the last 100 years (1926-2025), how would they perform over the next month, the next year, or the next decade? What could you actually expect to happen with your money?
While U.S. stocks have returned around 7% per year (including reinvested dividends and adjusting for inflation) going back to 1871, would a 7% inflation-adjusted return be the right expectation for the next year?
Surprisingly…it wouldn’t be. Why?
7% is too conservative. Yes, you read that correctly.
The expected return for U.S. stocks over a 1-year period is around 9% (with dividends and adjusting for inflation). The chart below plots the distribution of these 1-year returns for the U.S. stock market over the last 100 years:
If you take each return range, multiply by the probability of getting that return, and then sum them up, you’d get your expected return, which is 9.15% in this case.
This expected return is higher than the 7% compound historical growth rate because of volatility on the downside. Negative returns reduce the compound historical growth rate more than they reduce a simple arithmetic average. For example, a 50% gain followed by a 50% loss is a 25% loss overall, but would yield a 0% average return.
But the real takeaway from this chart isn’t the expected return, but how often high positive returns occur. For example, there’s a 47% chance that your 1-year return would exceed 10% and a 26% chance that it would exceed 20%. Of course, there’s also a 17% chance you’d lose 10% (or more), but that seems like a good trade.
What about shorter time periods? Over one month, the expected return of U.S. stocks is just 0.68%. Once again, there is a wide range of outcomes around this result:
Over a 3-month period, the expected return is a bit higher at 2.17%, but the return dispersion remains:
It’s only when we begin to look at longer time periods that the return distribution starts to show more positive skewness (i.e., a fatter right tail than left).
For example, the distribution of 5-year returns for U.S. stocks has far more weight in the right tail than the left one:
In 26% of 5-year periods, U.S. stock returns exceeded 12% on an annualized basis, while in just 10% of all 5-year periods returns were below -4%. This demonstrates how U.S. stocks can trend upward (or downward) for extended periods of time.
But the longer you extend your time horizon, the rarer the negative returns have been historically. Over 10 years, the probability of experiencing a negative annualized return of any kind is just 13%:
And, with a 7.03% annualized return, your money was expected to roughly double every 10 years. So not only are you less likely to lose money over a decade, but it’s reasonable to assume that your investment could double.
You might notice that this 7% annualized return is below the 9.15% 1-year return I cited earlier. This is actually to be expected. The 9.15% is the average of 1-year returns. But as you string returns together across years, the bad years drag the compounded return below the simple average. Remember that a 10% gain followed by a 10% loss isn’t a 0% gain—it’s a 1% loss. Repeat this with U.S. stock returns over time and the expected annualized return drifts down from 9.15% toward the 7% compound average.
This return drift is just as true over 20-year periods as over 10-year periods. While the annualized return declines slightly to 6.93% during 20-year periods (i.e., your money goes up 3.8x across two decades), there were no 20-year periods with a negative return over the last century:
This doesn’t mean that a negative 20-year period can’t occur in the future, only that it was historically unlikely. Going back to 1871, there was only one 20-year period where U.S. stocks didn’t have a positive total real return—June 1901 through May 1921.
And if we look at equity markets outside the U.S., 20-year negative returns are far more common. Japanese stocks peaked in 1989 and didn’t reach new all-time highs until 2024, or 35 years later. Today, Greek stocks are more than 50% below their 1999 peak. I don’t say this to scare you, but to emphasize the exceptional performance that U.S. stocks have had over the last century.
But this performance wasn’t always so exceptional. As you saw above, left tail (or negative) returns do occur. Instead of writing them off as rare, let’s look at what actually happened during these difficult times and what we can learn from them.
What Happened in the Worst of Times?
To understand what happens in the worst of times, I looked at four of the worst moments to invest in U.S. stocks over the last century: September 1929 (The Great Depression), January 1973 (stagflation and bear market), March 2000 (The DotCom Bubble), and October 2007 (The Global Financial Crisis).
If you had invested $1 at these relative peaks, here’s how that investment would have performed over the next 20 years:
As you can see, each of these scenarios started with devastating losses. The 1929 investor saw their dollar drop below $0.25 while the 2007 investor lost half of their money within 18 months. More importantly, these losses persisted for many years. The 1973 investor and the 2000 investor didn’t recover for more than a decade.
The good news is that they did eventually recover. You can see this in the following chart which shows the annualized real returns for the 1-year, 5-year, 10-year, and 20-year timeframes for each of these periods:
After 1 year, every single entry point was deeply negative, with losses ranging from -19% to -38%. We saw negative returns in all of the 5-year periods and most of the 10-year periods as well.
However, by 20 years, each period (with enough history) had positive real returns. Though the 2007 investor is still short of the 20-year mark, I highly doubt the S&P 500 will drop below 1,600 (more than a 75% decline from current levels) within the next 18 months.
This demonstrates the resilience of U.S. stocks in the face of immense economic uncertainty. While I can’t guarantee that U.S. stocks will perform like this in the future, betting against them is usually a bad idea.
Expect the Unexpected
After reviewing both the good and the bad for U.S. stocks over the last 100 years, my main takeaway is simple—expect the unexpected. I mean this both on the downside and the upside. I know a lot about historical U.S. stock performance, yet I still struggle to comprehend just how good the returns can be. This is more of a feeling than anything.
After all, how can stocks keep going up after they’ve gone up for so long? I ask myself this often enough that I get bearish on rare occasions. But my bearishness has been a mistake in hindsight.
Why? Because even after a century of stellar growth, U.S. companies continue to find ways to increase their earnings. And higher earnings are the bedrock of all long-term investment returns. While such progress can seem unsustainable, American companies are inventive. They find ways to get better through technology and other process improvements over time.
This is what I keep coming back to when I look at the last century of U.S. stock returns. They show a range of outcomes that would surprise most people. Outcomes that can feel impossible in the moment, but occur nonetheless.
If you’re wondering what to do with this information, my answer is always the same: Just Keep Buying, be patient, and let your money do the work for you. We’ve had a great run recently, but that doesn’t mean it has to end.
As I’ve said before, “The bull market didn’t start in March 2009. It started in 1776.”
Happy investing, Happy 250th Birthday to America, and thank you for reading!
If you liked this post, consider signing up for my newsletter.
This is post 510. Any code I have related to this post can be found here with the same numbering:
