The Equity Risk Premium
The equity risk premium is the additional return investors demand for investing in stocks rather than the safety of Treasury bonds. It is perhaps the most important — and most debated — assumption in valuation. A 1% difference in the equity risk premium can swing a valuation by 10–20% or more. This article unpacks how to estimate it, which methods practitioners use, and why intelligent analysts disagree.
Quick definition
The equity risk premium (ERP) is the difference between the expected return on the stock market and the risk-free rate: ERP = Rm − Rf, where Rm is the expected return on the market and Rf is the risk-free rate. In the CAPM cost of equity formula — Cost of equity = Rf + β(Rm − Rf) — the term (Rm − Rf) is the equity risk premium. It captures the premium investors require for taking on the volatility and uncertainty of stocks.
Key takeaways
- The equity risk premium is unobservable; no one knows the true forward-looking premium, so different methods yield different estimates
- Historical equity premiums (long-term market excess returns) average around 5–7%, but past returns do not guarantee future returns
- Forward-looking estimates using dividend discount models, earnings growth, or market surveys typically yield 3–5%
- Current market valuations embed expectations about returns; high valuations imply low forward-looking premiums
- The premium varies over time with economic conditions, investor sentiment, and risk appetite
- Using a 5.5% or 6.0% ERP is a reasonable middle ground for most DCF models; test sensitivity to ±1% changes
Historical equity premiums
The longest, most rigorous study of historical equity returns is Damodaran's Equity Risk Premium database (available free at stern.nyu.edu/~adamodar). Over the past 100+ years, U.S. stocks have returned roughly 10% per year, while Treasury bonds have returned around 4–5%. The difference, the historical equity risk premium, is therefore roughly 5–6%.
However, history is not prophecy. The 20th century was an unusual period: the U.S. emerged as a global superpower, corporate technology accelerated, and investors were often too risk-averse (the "equity risk premium puzzle"). Going forward, with higher Treasury yields, globalization, and more efficient markets, the premium may be lower.
Breaking down the historical return on stocks into components helps:
Historical stock return ≈ Dividend yield + Growth in dividends (earnings growth)
Historical equity return ≈ Dividend yield + Earnings growth + Multiple expansion
From 1926 to 2025, the average dividend yield was 2–3%, real earnings growth was 2–3%, and multiple expansion contributed another 1–2% (varying greatly by period). This sums to roughly 5–8% total return, or a 4–6% premium over bonds.
The problem is that all three components are mean-reverting. Valuations (multiples) cannot expand forever. Dividend yields were sometimes 6%+ (in value crashes) and sometimes 1% (in bubbles). The historical premium is a useful anchor, but not a reliable forecast.
Forward-looking estimation methods
Dividend Discount Model (DDM). Assuming stocks are worth the present value of all future dividends, you can solve for the implied discount rate that makes the current stock price consistent with expected dividend growth. If the S&P 500 yields 1.5% and long-term dividend growth is expected to be 5%, the implied required return is roughly 6.5%, suggesting an ERP of 6.5% − 4.5% (Rf) = 2.0%. This method is elegant but sensitive to dividend growth assumptions.
Earnings-based approach. Similar to DDM but using total earnings (dividends plus buybacks plus retained earnings). The S&P 500 earnings yield (inverse of P/E) is roughly 5%, and if long-term real earnings growth is 2% plus 2% inflation, the implied required return is 9%, suggesting a 4.5% ERP over a 4.5% risk-free rate. Again, growth assumptions matter enormously.
Supply-side equity risk premium. Damodaran's approach: estimate the forward return on stocks by projecting long-term real earnings growth, expected inflation, and reasonable dividend yields and multiple changes. This yields a range of 3–5% depending on whether you assume multiples expand, contract, or stay flat.
Economist and survey-based estimates. Academic surveys (e.g., CFO surveys by Graham and Harvey) ask finance professionals what premium they expect. Surveys typically yield 4–6%, with a lot of dispersion. The advantage is that these are actual practitioners' views; the disadvantage is that people are often overconfident or influenced by recent markets.
The relationship between valuation and forward premium
A critical insight: high valuations imply low future returns and a low equity risk premium. If the S&P 500 trades at 30× earnings, the earnings yield is only 3.3%. Even with robust growth, the forward-looking required return (and thus the ERP) will be lower than in periods when the market traded at 15× earnings.
This explains why valuations matter to discount rates. In 2024, when valuations have expanded and risk-free rates are high, the forward-looking ERP is lower than in 2009, when valuations were depressed and rates were near zero. Using a 6% ERP in 2024 might be too high; using 4% might be more realistic.
The market implied equity risk premium can be estimated by inverting valuation models. If you assume the market is fairly priced today and back out the ERP, you get the market's embedded expectation. This is often much lower than historical or survey-based estimates, particularly when valuations are lofty.
Variations across industries and countries
The equity risk premium is not uniform. Within the U.S. stock market, different sectors and company types have different risk profiles. Technology stocks, with high growth and low yields, command a lower explicit ERP than mature utility stocks. However, the CAPM adjusts for this through beta, not by using different ERPs.
Internationally, equity risk premiums vary. Emerging markets (higher growth but higher volatility and political risk) typically have higher ERPs — 7–10% — than developed markets. For a U.S.-based DCF of a U.S. company, use the U.S. equity risk premium. For an Indian company, use an emerging-market premium.
Practical recommendations
For a baseline DCF of a typical U.S. publicly traded company, use an equity risk premium of 5.0–5.5%. This sits in the middle of historical, forward-looking, and survey-based estimates and is defensible to most audiences.
If valuations are historically high (Shiller CAPE above 30), consider 4.0–4.5%, reflecting the low forward returns embedded in current prices.
If valuations are historically low (CAPE below 20) or the risk-free rate is unusually low, consider 5.5–6.5%, reflecting higher expected premiums from undervaluation or the need to compensate for low safe rates.
Always test sensitivity: show how your valuation changes if the ERP varies by ±1.0%. For long-duration, stable businesses, this swing can be material.
Common mistakes
Using the same ERP regardless of valuation level. Thoughtful valuations adjust the ERP for the current market cycle. In 2009, a 6% ERP was reasonable (low valuations). In early 2025, with expensive valuations, 4.5% might be more apt.
Confusing historical averages with forward expectations. Just because stocks returned 10% per year in the past does not mean they will going forward. Dividend yields are lower, profit margins are wider (potentially unsustainably so), and valuations are higher. The future ERP is likely in the 3–5% range.
Ignoring the relationship between risk-free rates and ERP. When risk-free rates are very high (say, 5%), investors do not require a 10% return on stocks (5% + 5% ERP) unless equity risk is extraordinarily high. Higher safe rates push down the forward equity risk premium. A 4% ERP in a 5% risk-free-rate environment is defensible.
Using survey data mechanically. CFO surveys show an average expected ERP of 4.5%, but there is huge dispersion — some say 2%, others say 8%. Use the survey as a data point, not a gospel. Understand the distribution.
Failing to document the ERP assumption. Your discount rate depends heavily on the ERP, and the ERP is subjective. Write down which ERP you used, why, and what the implied market expectations are. A reader should be able to understand your reasoning.
FAQ
Q: How does beta interact with the ERP?
A: Beta measures how much a stock moves relative to the market. In the CAPM formula Cost of equity = Rf + β(Rm − Rf), the ERP is multiplied by beta. A stock with beta 1.5 will have a higher cost of equity than a stock with beta 0.8, even if they use the same ERP. The ERP is the market-wide risk premium; beta scales it for the individual stock.
Q: If the ERP is 5% and beta is 2, is the cost of equity 10%?
A: No. The cost of equity is Rf + 2 × 5%. If Rf is 4.5%, then cost of equity = 4.5% + 10% = 14.5%. Beta scales the risk premium; you add it to the risk-free rate.
Q: Should I use a different ERP in my terminal value than in my explicit forecast period? A: Typically, no. The ERP is forward-looking and long-term, so it is the same in both. However, if you assume high growth in the explicit period (10 years) followed by a normalization to steady state, you might argue that the expected return should be lower (closer to growth rate). In practice, most models use a consistent WACC and ERP throughout.
Q: How does the ERP change with recessions and crises? A: During crises (2008, 2020, 2022), equity risk spikes and the ERP widens (investors demand higher returns to hold stocks). After crises, as confidence returns, the ERP compresses. Over very long periods, the ERP mean-reverts. In a DCF, use a normalized, forward-looking ERP, not one distorted by recent crisis or euphoria.
Q: Can the ERP be negative? A: In theory, no. The equity risk premium is the return premium for bearing risk; it should always be positive. However, if you compute the ERP as the market's implied forward return minus a very high risk-free rate, you can get a low number (even near 0%) if valuations are very rich and safe rates are very high. This suggests the market is priced for very low future returns, not a negative premium.
Q: Should I use a different ERP for private companies? A: Yes. Private companies have higher idiosyncratic risk (no liquidity, no diversification benefit). Add a 1–2% private company discount rate premium, or use a higher ERP (6.5–7.5% instead of 5%). This compensates for illiquidity and concentration risk.
Related concepts
- Capital Asset Pricing Model (CAPM) — The framework in which the ERP plays a central role.
- Beta and Systematic Risk — How individual stocks' risk is measured relative to the market.
- Valuation Levels and Forward Returns — Why high valuations imply low forward premiums and vice versa.
- Historical Stock Market Returns — The empirical foundation for ERP estimates.
- Risk Factors Beyond Beta — Size, value, momentum, and other factors that may command additional premiums.
- International and Emerging Market Premiums — How ERP varies across geographies.
Summary
The equity risk premium is the excess return investors demand for holding stocks instead of Treasury bonds. It is unobservable and forward-looking, so estimates vary: historical averages suggest 5–7%, dividend discount models suggest 3–5%, and current market valuations imply 2–4%. A reasonable base case is 5.0–5.5%, adjusted downward if valuations are high and upward if they are low. The ERP, multiplied by beta, determines how much a stock's cost of equity exceeds the risk-free rate. Documenting your ERP assumption and testing sensitivity to it are essential parts of a rigorous DCF.
Next
You now have the risk-free rate and the equity risk premium — the two building blocks of the cost of equity. The third and final block is beta, which measures how much an individual stock moves relative to the market. The next article explains beta and how to estimate it.