What history tells us about expected stock returns after a bull run
- Robin Powell
- Jan 3
- 8 min read
Updated: Jan 4
New research analysing 153 years of market data reveals why expected stock returns over the next decade are likely to disappoint, and what disciplined investors should do about it.
I remember the weeks leading up to the millennium. The Y2K bug was all anyone talked about. Planes might fall from the sky. Banking systems could collapse. Power grids might fail at midnight. Some people stockpiled food and withdrew cash. The anxiety was real.
Then January 1st, 2000 arrived. Nothing happened. The lights stayed on. The planes kept flying. The collective relief was palpable.
For many investors, and yes, including me, that relief quickly became complacency. Markets had been on a storming run through the late 1990s. Technology stocks seemed unstoppable. I'd put money into a technology fund, watching it climb and climb. When the millennium bug turned out to be a damp squib, it felt like confirmation that the good times would continue. The threat had passed. Why wouldn't markets keep rising?
That's not what happened. My technology fund fell about 80%. The S&P 500 entered what became known as the "lost decade", delivering annualised returns of -0.95% between 2000 and 2009, according to Dimensional Fund Advisors. An entire decade of nothing. Worse than nothing.
I learned something important. Relief isn't a reason for optimism. And the better markets have performed, the more carefully you should examine what's required for that performance to continue.
"Relief isn't a reason for optimism."
Which brings me to today. Over the last three calendar years, the S&P 500 has delivered total returns of roughly 26% in 2023, 25% in 2024, and around 17% in 2025. Since March 2009, global markets have produced exceptional gains. Performance like that can't continue indefinitely. The question isn't whether returns will moderate. It's by how much.
The three forces that drive stock returns
Stock returns come from only three places.
First, the dividend yield: the cash companies pay you for holding their shares. Second, earnings growth: how much corporate profits increase over time. Third, changes in the price-to-earnings ratio: whether investors become willing to pay more or less for each pound of earnings.
Everything else is noise. Every market movement, every headline, every quarterly surprise flows through these three channels. Think of them as three forces pulling on a rubber band. When all three pull in the same direction, returns stretch impressively. When they don't, the band goes slack. Or snaps back.
Here's what most investors miss: these forces don't like working together.
Javier Estrada, a finance professor at IESE Business School, recently analysed 153 years of US market data, from 1872 to 2024, to understand how these components interact. His findings are uncomfortable for anyone expecting recent performance to continue.
The correlation between earnings growth and P/E changes over ten-year periods is -0.50. That's not slightly negative. It's meaningfully, persistently negative. When profits grow quickly, investors don't also bid up valuations. And when valuations expand, it tends to happen during periods of slower earnings growth.
Look at the pattern across different time horizons. At one year, the correlation is -0.67. At three years, -0.75. At ten years, it's -0.50. Even at 20 years, it remains negative at -0.39. This isn't a quirk of one particular period. It's how markets work.
Why? When earnings grow rapidly, investors notice. They bid up prices in anticipation, leaving less room for further P/E expansion. And when earnings disappoint, P/E ratios sometimes expand anyway, not because investors are optimistic, but because prices fall more slowly than earnings.
You can stretch the rubber band by pulling hard on one end (strong earnings growth) or the other (P/E expansion). But pulling hard on both simultaneously? The tension works against you.
Bullish forecasts implicitly assume all three forces will keep pulling together. Dividends will flow. Earnings will grow. Investors will keep paying premium prices. History says that's not how it works.
Where we stand today, and why 1999 keeps coming up
In September 2025 the Shiller CAPE ratio crossed 40 for only the second time in history. The first was December 1999.
This metric, developed by Nobel laureate Robert Shiller, compares current prices to average inflation-adjusted earnings over the previous ten years. It smooths out short-term fluctuations and gives a cleaner picture of how much investors are paying for underlying profits. The historical median is around 16.
That comparison to 1999 makes people nervous. It should. After the CAPE exceeded 40 back then, the S&P 500 fell 37% over the next two years.
Estrada's framework shows what investors in 1999 would have needed to believe. Given the dividend yield of 1.2% at the time, various combinations of earnings growth and terminal P/E would have produced the returns shown below.
The actual result? Annual returns of approximately -0.7%, according to Estrada. That outcome sits in a zone requiring modest earnings growth and a P/E falling back toward historical norms. Which is what happened. The optimists expecting 8%, 10%, or 12% annual returns were betting on combinations in the upper right of that grid. Combinations that didn't materialise.
But I need to be careful here. Many serious commentators argue today is different, and they have reasonable points.
Current valuations are supported by actual earnings growth, not speculation. Analysts at Goldman Sachs recently noted that the PEG ratio (which measures how much you're paying for each unit of earnings growth) was 3.7x at the dot-com peak versus 1.7x today. Investors in 1999 were paying far more per unit of growth. Today's big tech companies generate enormous cash flows. They're not speculative dot-coms burning through capital with no path to profitability.
The Shiller PE itself has attracted criticism. Some argue it hasn't "worked" as a forecasting tool recently. Markets have remained expensive for over a decade, yet returns have been strong. Perhaps structural changes justify higher multiples.
These are fair points. I'm not dismissing them.
But whether or not we're in a bubble is beside the point. Estrada's analysis isn't about predicting crashes. It's about arithmetic. Whatever you believe about today's valuations, the question remains: what conditions would be required for current expectations to be met?
The maths of expecting more
Let's return to the rubber band. Right now, it's stretched taut.
The S&P 500's P/E ratio stood at 27 in June 2025, according to Estrada. The dividend yield was 1.3%. The historical average P/E is around 16, and dividend yields used to be considerably higher. One of our three return drivers is suppressed, another elevated.
That leaves earnings growth to do the heavy lifting. But remember the negative correlation. When P/E ratios are high, earnings growth and P/E expansion rarely pull together.
Estrada calculated what various combinations of earnings growth and terminal P/E would deliver from June 2025 to June 2035.
If earnings grow at their historical average of 4.2% annually and the P/E reverts to 16, Estrada calculates expected annual returns of 0.4%. Not 4%. 0.4%, before inflation.
What about matching the long-term historical average of around 9%? That requires earnings growth well above historical norms, or the P/E remaining elevated, or both. To reach 9% annual returns with a terminal P/E of 20, you'd need earnings growth of roughly 7% per year. Nearly double the historical rate.
And to match the last decade's returns, around 13%? You'd need earnings growth of 9-10% annually alongside a terminal P/E of 35. Both figures are far above anything history suggests is sustainable.
None of this is prediction. It's arithmetic. Pick your own combination based on whatever you believe. But the maths doesn't change. The more optimistic your expectations, the more you're betting on historically implausible outcomes.
"The maths doesn't change. The more optimistic your expectations, the more you're betting on historically implausible outcomes."
This isn't a call to time the market
I want to be clear about what this analysis does and doesn't tell us.
It doesn't tell us when returns will disappoint. Valuations have poor predictive power over short horizons. The Shiller CAPE has been above its historical average for most of the past decade, yet returns have been exceptional. Anyone who sold in 2015 because valuations looked stretched missed tremendous gains.
Estrada's research tells us about expected returns over the next decade. It says nothing useful about next month, next quarter, or even next year. Trying to time the reversion is a losing game.
But the lost decade taught us something else: diversification worked.
While the S&P 500 delivered -0.95% annualised between 2000 and 2009, other parts of the market did far better. International small cap stocks returned over 190% cumulatively, according to PrairieView Partners. The equal-weighted S&P 500 returned 5.1% annualised while the cap-weighted version lost money, as documented by AMG Wealth.
The problem wasn't equities. The problem was concentration in one expensive corner of the market.
"The problem wasn't equities. The problem was concentration."
You don't need to predict the future to position sensibly for it. When valuations are stretched in one area, returns from that area tend to be muted. But stretched valuations in US large caps don't mean stretched valuations everywhere.
The answer isn't abandoning equities or guessing the top. It's building a portfolio that doesn't depend on any single market defying gravity.
What disciplined investors do instead
Start with your expectations. If you're planning for retirement based on 10% annual returns because that's what the last decade delivered, you may be disappointed. Building your financial plan around more conservative assumptions isn't pessimism. It's prudence.
Next, diversify properly. Not owning different funds that hold the same underlying stocks.
Geographic diversification matters. When US large caps went nowhere for ten years, international markets delivered positive returns.
Size matters too. Small caps outperformed large caps during the 2000s. They've lagged recently, which means they're now cheaper relative to large caps than they've been for years.
Factor diversification helps. Value stocks performed well during the lost decade while growth stocks struggled. The pendulum swings.
Work out what proportion of your portfolio should be in equities versus bonds. This is the most important investment decision you'll make. Base it on your circumstances, time horizon, and capacity for loss. Not on what markets have done lately.
Set up automatic rebalancing. When equities rise and become a larger share of your portfolio than intended, rebalancing forces you to trim. When they fall, it forces you to buy more. A mechanical way to sell high and buy low without timing anything.
Finally, stay the course. The worst response is panic-selling after markets have fallen. If stretched valuations lead to disappointing returns, the investors who suffer most will be those who abandon their strategy at the bottom.
None of this requires predicting the future. It requires respecting what 153 years of history tells us.
Why expected stock returns face a decade of tension
I think back to January 2000, that feeling of relief when the millennium bug didn't bite. How easily relief became complacency.
I'm not saying we're about to repeat 2000. Nobody knows that. What I am saying is that the conditions rhyme. Valuations at levels seen only once before. Investors extrapolating recent returns. A sense that the old rules don't apply.
Maybe they don't. Maybe structural changes really do justify permanently higher valuations. Maybe AI-driven productivity gains will produce earnings growth far beyond historical norms.
But Estrada's 153 years of data tell a consistent story. The three forces that drive returns don't pull together indefinitely. When valuations stretch far above historical norms, something eventually gives.
The disciplined investor doesn't try to predict when the snap will happen. They respect the tension that exists. They calibrate expectations. They diversify beyond the areas that have stretched furthest. They rebalance systematically. And they stay the course.
You can't control what markets deliver over the next decade. But you can control whether your expectations, your diversification, and your discipline are built for whatever comes.
Resources
Estrada, J. (2025). Expected stock returns in bullish times. IESE Business School.
Dimensional Fund Advisors. (2020). A tale of two decades: Lessons for long-term investors.
Goldman Sachs. (2025). Why we are not in a bubble... yet. Goldman Sachs Global Investment Research.
Recently on TEBI
Make 2026 the year you finally sort your life out
If one of your New Year's resolutions is to get your retirement planning sorted, we recommend starting with the award-winning book How to Fund the Life You Want by TEBI founder Robin Powell and Jonathan Hollow. It's published by Bloomsbury and written primarily for UK readers.
Buy the book here on Amazon, or, if you prefer, here on bookshop.org.