Guide

Capital Asset Pricing Model (CAPM) explained

A utility stock and a speculative biotech both trade on the same exchange, but investors demand very different returns from each. The biotech might need to offer 12% expected annual return to attract capital; the utility might clear at 7%. What links risk to that required return? The Capital Asset Pricing Model (CAPM), developed by William Sharpe, John Lintner, and Jan Mossin in the 1960s, is the textbook answer: expected return = risk-free rate + beta × market risk premium. CAPM isolates systematic (market-wide) risk — measured by beta — and argues that diversified investors are not compensated for diversifiable firm-specific shocks. Fund managers use CAPM to set hurdle rates, estimate cost of equity in WACC models, and judge whether a portfolio earned fair compensation for the market exposure it took. This guide explains the CAPM formula and security market line, walks through beta estimation and the equity risk premium, compares CAPM to multi-factor alternatives, works a Harbor Capital equity-sleeve example, lists common pitfalls, and provides an investor checklist alongside modern portfolio theory and Treynor ratio fundamentals.

What CAPM measures

CAPM answers one question: given a stock or portfolio’s sensitivity to the broad market, what return should a rational investor require? It rests on assumptions from Markowitz mean-variance theory: investors hold diversified portfolios, care only about mean return and variance, can borrow and lend at a single risk-free rate, and share homogeneous expectations about future cash flows.

Under those assumptions, only systematic risk commands a premium. Unsystematic risk — earnings surprises, product recalls, management changes — can be diversified away by holding dozens of uncorrelated positions. A well-diversified investor therefore prices assets based on how they co-move with the market portfolio, not on total volatility alone. That co-movement is beta.

Systematic vs total risk

Total return volatility mixes both risk types. A low-beta utility with high idiosyncratic swings might show a large standard deviation but still plot near the security market line — its expected return is modest because market exposure is low. Conversely, a high-beta tech stock with moderate total volatility may require a high expected return because most of its risk is systematic. CAPM deliberately ignores unsystematic variance; metrics like the Sharpe ratio penalize total volatility instead.

The CAPM formula and security market line

The standard CAPM equation for the expected return on asset i is:

E(R_i) = R_f + β_i × [E(R_m) − R_f]

where R_f is the risk-free rate (typically short-term Treasuries), E(R_m) is the expected return on the market portfolio (often proxied by the S&P 500), β_i is the asset’s beta versus that market, and [E(R_m) − R_f] is the equity risk premium (ERP) — the extra return investors demand for holding equities instead of cash.

Reading the security market line (SML)

Plot expected return on the vertical axis and beta on the horizontal axis. CAPM predicts all fairly priced assets lie on a straight line through (0, R_f) with slope equal to the ERP. That line is the security market line (SML). An asset above the SML is underpriced relative to CAPM (positive Jensen’s alpha); below the line is overpriced.

Example with round numbers: if R_f = 4%, ERP = 5%, and β = 1.2, then E(R) = 4% + 1.2 × 5% = 10%. A stock with β = 0.8 would require only 8%. The market portfolio itself has β = 1 by construction, so its expected return equals R_f + ERP.

Estimating beta

Practitioners estimate beta by regressing an asset’s excess returns on the market’s excess returns over a lookback window — commonly 36 to 60 months of monthly data. Choices matter: which index (S&P 500 vs MSCI World), whether to use total vs excess returns, and whether to apply a Blume adjustment (shrinking raw beta toward 1.0 because betas mean-revert). Levered companies have equity beta above asset beta; analysts unlever beta when comparing firms with different capital structures.

Equity risk premium and the risk-free rate

CAPM’s inputs are simple to write but hard to pin down in practice.

Input	Typical proxy	Common pitfall
Risk-free rate (R_f)	3-month T-bill, 10-year Treasury for long horizons	Mixing nominal rates with real cash-flow forecasts
Market return E(R_m)	Historical S&P 500 arithmetic or geometric mean	Survivorship bias in long historical windows
Equity risk premium	Forward ERP surveys (4–6%) or historical spread (~5–7%)	Using peak-era premiums after secular rate shifts
Beta	OLS regression vs broad index, 3–5 year window	Short windows that capture one regime only

Academic debate continues over whether the historical U.S. ERP (~6% arithmetic) is forward-looking or inflated by twentieth-century exceptionalism. Corporate finance teams often adopt a forward-looking ERP from Damodaran-style surveys or implied cost-of-capital models rather than blindly extrapolating past averages. Changing the ERP by one percentage point moves every CAPM expected return by β percentage points — material for hurdle rates and DCF valuations.

CAPM vs multi-factor and fundamental models

CAPM is a single-factor model: market beta is the only priced risk. Empirical research since the 1990s documents anomalies — size, value, momentum, and profitability effects — that CAPM cannot explain. The Fama-French three- and five-factor models add explanatory variables; arbitrage pricing theory (APT) allows multiple unspecified factors.

Model	Best for	Limitation
CAPM	Hurdle rates, quick cost-of-equity estimates, teaching systematic risk	Poor fit for value, small-cap, and momentum tilts
Fama-French multi-factor	Explaining cross-sectional return differences, factor portfolios	More parameters to estimate; factor definitions shift
DCF / fundamentals	Intrinsic value of cash-flow businesses	Input sensitivity; weak for banks and early-stage firms
Build-up / subjective premium	Private companies, illiquid stakes	Ad hoc; hard to backtest

For diversified index investors, CAPM remains the conceptual backbone: you earn the risk-free rate plus beta times the market premium. Active managers tilting toward value or small caps should supplement CAPM with factor investing frameworks rather than interpreting all outperformance as alpha.

Worked example: Harbor Capital equity sleeve

Harbor Capital’s HUC (U.S. Core Equity) sleeve targets long-only large-cap U.S. stocks with a neutral sector weight versus the S&P 500. The investment committee sets a hurdle rate for new positions using CAPM.

Assumptions for Q2 2026 review:

Risk-free rate: 4.2% (1-year Treasury)
Forward equity risk premium: 5.0% (internal survey midpoint)
Benchmark: S&P 500 total return

Candidate A — regulated utility (HUC-EL)
Estimated β = 0.65 (5-year monthly vs S&P 500).
CAPM required return = 4.2% + 0.65 × 5.0% = 7.45%.
Management guides 6–7% EPS growth plus a 3.5% dividend yield. Total expected return ~9.5–10.5% — above the CAPM hurdle, so the name clears for further fundamental work.

Candidate B — high-growth software (HUC-SW)
Estimated β = 1.45.
CAPM required return = 4.2% + 1.45 × 5.0% = 11.45%.
Sell-side consensus implies 13% annualized return over three years. The margin over CAPM is thin; the committee demands a larger moat and margin-of-safety case before overweighting.

Portfolio-level check: HUC’s sleeve beta is 1.02. CAPM fair return = 4.2% + 1.02 × 5.0% = 9.4%. If HUC returned 11.8% last year while the S&P 500 returned 10.1%, raw outperformance is 1.7%. Jensen’s alpha = 11.8% − [4.2% + 1.02 × (10.1% − 4.2%)] = 11.8% − 10.2% = +1.6% — modest positive alpha after adjusting for systematic exposure. The Treynor ratio and information ratio provide complementary views on whether that alpha is repeatable.

When to use CAPM (decision table)

Question	CAPM appropriate?	Alternative
Cost of equity for a large-cap public company	Yes — first-pass benchmark	Add size/value premiums or build-up for private firms
Is my index fund earning fair return for its beta?	Yes — compare to SML	Sharpe ratio if portfolio is not fully diversified
Explaining small-cap value outperformance	No — CAPM underpredicts	Fama-French factors
Crypto token hurdle rate	Caution — beta vs S&P is unstable	Scenario analysis; crypto-specific risk premia
Merger arbitrage or market-neutral fund	No — low beta, idiosyncratic drivers	Absolute-return targets; factor-neutral alpha
DCF discount rate for mature cash-flow business	Yes — CAPM feeds cost of equity in WACC	Cross-check with bond yield plus spread

Common pitfalls

Using total volatility as risk. CAPM prices beta, not standard deviation. High-idiosyncratic-risk stocks can look “cheap” on Sharpe but fair on CAPM.
Stale or short beta windows. A 12-month beta during a rate shock misstates long-run exposure. Prefer 36+ months and sanity-check against sector peers.
Wrong benchmark. A European stock regressed on the S&P 500 produces meaningless beta. Match geography and style.
Ignoring leverage. Equity beta rises with debt. Compare unlevered betas when capital structures differ.
Treating CAPM alpha as skill. Positive Jensen’s alpha over one year may reflect factor tilts CAPM ignores (value, momentum). Test persistence and factor exposure.
Real vs nominal mismatch. Discounting real cash flows with a nominal CAPM rate (or vice versa) biases valuations.
Assuming ERP is constant. Equity risk premia compress in euphoria and widen in crises. Forward-looking estimates beat blind historical means.
Applying CAPM to undiversified holdings. If you own three stocks, you still bear idiosyncratic risk CAPM assumes away. Required return should exceed the CAPM prediction.

Investor checklist

State R_f, ERP, and benchmark explicitly before computing required return.
Estimate beta with at least 36 months of monthly returns vs an appropriate index.
Apply Blume or Bayesian shrinkage when using raw regression beta for forecasting.
Compare expected return to the security market line, not to arbitrary historical averages.
For portfolios, compute weighted average beta before judging alpha.
Supplement CAPM with factor models when evaluating value, small-cap, or momentum tilts.
Cross-check CAPM cost of equity against traded bond yields plus equity risk spread.
Re-estimate inputs after major rate regime changes.
Document whether hurdle rates are nominal or real and match DCF conventions.
Use Treynor or information ratio alongside Jensen’s alpha for performance attribution.

Key takeaways

CAPM links expected return to systematic risk: E(R) = R_f + β × ERP.
Beta measures co-movement with the market; diversifiable risk does not earn a CAPM premium.
The security market line graphs fair expected returns across beta values.
ERP and beta estimation choices dominate CAPM output — document assumptions carefully.
Multi-factor models extend CAPM but the single-factor version remains the baseline for hurdle rates and teaching.