You're staring at a spreadsheet. Maybe it's a retirement projection. Maybe it's a DCF model for a stock you're researching. Either way, you need one number: the expected market return.
And that's where it gets messy.
Some people plug in 10% because "that's what the S&P 500 has done historically." Others use 7% because they heard it on a podcast. A few brave souls dig into equity risk premiums and risk-free rates and end up more confused than when they started.
Here's the thing — there's no single "correct" number. But there is a correct way to think about it.
What Is Expected Market Return
Expected market return is exactly what it sounds like: the return investors anticipate earning from the broad stock market over a given period. Usually, "the market" means a broad index like the S&P 500 or a total market fund. Usually, the period is long-term — 10, 20, 30 years It's one of those things that adds up..
But here's where most definitions stop being useful.
Expected return isn't a prediction. That said, it's not a forecast. It's an input — a building block for decisions. The process for arriving at a reasonable estimate? You use it to decide how much to save, whether a stock is cheap, if your retirement plan holds water. The number itself is unknowable. That's learnable.
It's not the same as historical return
This trips up almost everyone. Think about it: the S&P 500 has returned roughly 10% nominal (6. 8% real) since 1926. That's a fact. But using it as your expected return assumes the future will look like the past — same valuation starting point, same productivity growth, same demographic tailwinds, same everything.
It won't.
Valuations matter. Day to day, interest rates matter. Starting P/E matters. Practically speaking, history doesn't care. In practice, if you buy the market at a CAPE of 35, your forward returns will be lower than if you bought at 15. Math does That's the part that actually makes a difference..
It's not the same as required return
Required return is what you need to hit your goals. Expected return is what the market might give you. Also, "I need 12% so I'll assume 12%" is not a strategy. Still, confusing the two is how people take too much risk — or too little. It's wishful thinking Simple as that..
Why It Matters / Why People Care
Every financial model you've ever seen rests on this number.
Retirement calculators. DCF valuations. That's why insurance liability matching. Pension funding ratios. They all take an expected market return as a primary input. Asset allocation studies. Monte Carlo simulations. Garbage in, garbage out.
Get it wrong by 2% over 30 years and you're off by a factor of nearly 2x on terminal wealth. That's not rounding error. That's the difference between retiring at 62 and working until 70.
The valuation anchor problem
Most people anchor to recent returns. After a crash, they plummet. After a bull market, expected return estimates creep up. Both are wrong — but in opposite directions Worth keeping that in mind. And it works..
High recent returns lower future expected returns (valuations expanded). Consider this: low recent returns raise them (valuations compressed). This is mean reversion in action, and it's the single most ignored concept in retail investing And that's really what it comes down to..
The cost of being wrong
If you assume 10% and get 6%, you undersave by ~40% over three decades. If you assume 6% and get 10%, you oversave — which sounds fine until you realize you worked extra years you didn't need to, missed experiences, delayed giving.
Both errors have real costs.
How to Estimate Expected Market Return
There's no perfect method. Here's the thing — smart people disagree. But there are three broad approaches, each with strengths and blind spots. The honest answer? Use all three. Triangulate Simple as that..
1. Historical average — with adjustments
Start here. Think about it: it's the baseline. But don't just take the raw number.
Nominal vs. real. Since 1926, S&P 500 nominal return ≈ 10.2%. Real (inflation-adjusted) ≈ 6.8%. For planning, always work in real terms. Inflation is a separate variable. Mixing them muddies everything.
Time period selection bias. 1926–2023 includes the Great Depression, WWII, the 1970s stagflation, the dot-com bubble, 2008, COVID. That's a feature, not a bug. But if you start in 2009, you get 15%+. If you start in 2000, you get 6%. Cherry-picking start dates is the oldest trick in the book. Don't do it.
Survivorship bias. The U.S. market won the 20th century. Other markets (Japan, Argentina, Russia 1917) didn't. Using U.S. history as the expected return assumes the U.S. keeps winning. Maybe it does. Maybe it doesn't. A global market cap-weighted portfolio is a more honest baseline — historically ~5.5% real Most people skip this — try not to..
Adjust for current valuations. This is where it gets real. The Shiller CAPE (cyclically adjusted P/E) correlates strongly with 10–20 year forward returns. At CAPE 30+, 10-year real returns have historically clustered around 2–4%. At CAPE 15, they've been 7–9%.
As of early 2025, CAPE sits around 34. That said, that doesn't guarantee low returns. But it shifts the distribution. A reasonable adjusted historical estimate: 4–5% real (6–7% nominal).
2. Forward-looking building blocks (the "Graham-Dodd" approach)
Jack Bogle popularized this. Expected return = dividend yield + earnings growth + valuation change Not complicated — just consistent..
Dividend yield. Knowable today. S&P 500 yield ≈ 1.3%. Low by history Small thing, real impact..
Earnings growth. Real earnings per share growth has averaged ~1.5–2% long-term. Productivity + population + pricing power. Could be higher with AI. Could be lower with margins normalizing. Call it 1.5–2.5% real And that's really what it comes down to..
Valuation change. The wildcard. If P/E expands, returns get a boost. If it contracts, a drag. Mean reversion says it should contract from 34. But "should" and "will" are different. Assume zero for a neutral case. Negative for conservative.
Sum: 1.3% + 2% + 0% = 3.3% real. Add 2.5% inflation → ~5.8% nominal.
That's lower than the historical average. Intentionally so — it reflects today's starting point.
3. Equity Risk Premium (ERP) + Risk-Free Rate
This is how institutions do it. Expected market return = risk-free rate + equity risk premium And that's really what it comes down to..
Risk-free rate. Usually 10-year Treasury yield (real or nominal). As of early 2025, 10-year real yield (TIPS) ≈ 2.1%. Nominal ≈ 4.3%.
Equity risk premium. The extra return investors demand for bearing equity risk.
Estimating the equity risk premium (ERP)
The ERP is the extra yield investors demand for shouldering the volatility of equities over the risk‑free rate. Think about it: it isn’t a fixed number; it’s a living estimate that shifts with market sentiment, macro conditions, and forward‑looking expectations. Below are three complementary ways practitioners triangulate a reasonable ERP today.
a. Historical average, adjusted for mean reversion
The classic long‑run U.S. So equity premium hovers around 5–6 % nominal. To avoid overstating future expectations, analysts trim the raw historical figure by the degree of valuation overvaluation observed at the start of the forecast horizon.
- Current CAPE: ≈ 34 (well above the 100‑year mean of ~16.6).
- Historical relationship: Every 5‑point increase in CAPE above the mean has been associated with a 0.8‑percentage‑point reduction in the realized equity premium over the subsequent decade.
Applying that adjustment yields an ERP in the 4.0–4.5 % range for the next 10‑year window.
b. Forward‑looking survey of professional forecasters
The Survey of Professional Forecasters (SPF) and the AAII Asset Allocation Survey ask portfolio managers to project the next‑year equity return and the corresponding risk premium. Think about it: the median response in the latest quarter places the ERP at 3. 8 % nominal, with a 10‑year horizon median of 4.So 2 %. This consensus reflects a more cautious stance than the raw historical average, acknowledging both elevated valuations and the possibility of a slower‑growth macro environment.
c. Implied ERP from options‑derived expectations
A more market‑based approach extracts the ERP from the difference between implied stock returns (derived from options pricing) and the risk‑free rate. On the flip side, using the S&P 500 options‑derived forward return model, the implied ERP as of early 2025 is roughly 4. 1 % nominal. This figure aligns closely with the survey‑based estimate and provides a market‑price validation of the risk premium embedded in current option prices Simple, but easy to overlook..
Putting it together
When we combine the three strands—adjusted historical premium, professional‑forecast consensus, and options‑derived implication—we land in a narrow band:
- Nominal ERP: 3.8 % – 4.3 %
- Real ERP (inflation‑adjusted): ≈ 2.5 % – 2.8 %
Adding the prevailing 10‑year Treasury yield (≈ 4.3 % nominal) gives an expected market return of ≈ 8.1 % – 8.6 % nominal over the next decade, assuming the ERP remains stable. That's why in real terms, that translates to ≈ 3. This leads to 8 % – 4. 3 % after stripping out inflation.
Synthesis: A practical framework for investors
- Start with a baseline real return derived from forward‑looking fundamentals (dividend yield + earnings growth + valuation drift). For 2025, this sits near 3.3 % real.
- Layer on a valuation‑adjusted historical premium to capture the impact of elevated CAPE. This nudges the real return down to ≈ 3.8 %.
- Incorporate the ERP as the risk‑adjusted uplift over the risk‑free rate. With an ERP of ~4 % nominal, the implied market return aligns with the 8 %‑plus nominal figure derived above.
By triangulating these three lenses—fundamental building blocks, historical adjustment, and market‑implied premium—investors can construct a more strong, less biased estimate of future equity performance. The exact number will vary with macro shocks, policy shifts, or sudden valuation swings, but the framework remains resilient.
Conclusion
Projecting future stock‑market returns is as much art as it is science. Relying on a single historical average invites bias; ignoring the present valuation landscape understates risk; and treating the equity risk premium as a static constant disregards market psychology. The most prudent approach blends:
- Fundamental expectations (yield, earnings growth, valuation drift)
- Valuation‑adjusted historical context (CAPE‑driven premium decay)
- Market‑derived risk premium (survey consensus and options‑implied figures)
When these components are weighed together, they produce a forward‑looking return estimate that acknowledges both the upside potential of continued earnings expansion and the downside pressure of elevated valuations. Because of that, for the current environment, a realistic expectation sits in the low‑to‑mid‑single‑digit nominal range, with real returns likely hovering just above 3 %. Investors who internalize this nuanced picture can set more realistic performance goals, allocate capital with appropriate risk buffers, and ultimately stay the course when short‑term market noise tries to obscure the longer‑term narrative Worth keeping that in mind..
