CAPM Calculator (Cost of Equity)
Capital Asset Pricing Model calculator. Computes the required return on equity (cost of equity) from risk-free rate, expected market return and asset beta, with dated, sourced default inputs.
CAPM Calculator — Cost of Equity
Where the defaults come from: the US Treasury's daily par yield curve put the 10-year yield at 5.29% on 30 September 2026, and Aswath Damodaran's implied equity risk premium for 1 October 2026 is 3.70% over that rate (trailing-12-month cash yield; his sheet's other variants run from 3.30% to 5.67%). Rm = 5.29% + 3.70% = 8.99%. Replace all three with your own, dated, in one currency.
How to use the CAPM calculator
Enter the risk-free rate (Rf)
The yield on a "riskless" sovereign bond matching your valuation horizon. For US equity valuation: the 10-year US Treasury yield from the Treasury's daily yield-curve table on the valuation date. For shorter cash flows: 3-month or 2-year T-bill. For non-US valuations: matching-tenor local sovereign yield (UK gilt, German Bund, JGB, MAS T-bill for SGD). The Rf anchors the entire CAPM line — get this wrong and every cost of equity below is wrong.
Enter the expected market return (Rm)
The expected return on a broad market index over your horizon. Two methods: (1) Historical average — the long-run arithmetic or geometric mean of a broad index; Damodaran's historical-returns table (1928 to date) is the standard free source. (2) Forward-looking — the risk-free rate plus an implied equity risk premium backed out of today's index level and expected cash flows, which Damodaran also publishes monthly. Whichever you use, state it: Rm is the input that moves the answer most. Don't use individual-year returns; the noise dwarfs the signal.
Enter the asset's beta (β)
Beta = covariance(asset, market) / variance(market). Sources: Bloomberg / Refinitiv for institutional. Yahoo Finance / Stock Analysis / Stooq for retail — check which window and return frequency each one states. Damodaran's data archive for industry-average betas (preferred for private-company valuation). Levered vs unlevered beta matters: published "raw" betas are levered to the company's current capital structure — unlever and re-lever for like-for-like comparisons.
Read the cost of equity (re)
The headline number is the required return on equity — i.e. what equity investors demand to compensate for the risk of holding this asset. This is the discount rate used for cash flows accruing to equity holders (FCFE-based valuation) and the equity component of WACC for firm-level cash flows (FCFF-based). Higher re = higher discount rate = lower valuation, all else equal.
Feed into WACC and DCF
Cost of equity is one of two ingredients in WACC (Weighted Average Cost of Capital — see WACC Calculator). WACC then becomes the discount rate in DCF valuation. Each tool feeds the next: CAPM → cost of equity → WACC → DCF discount rate. If your DCF is sensitive to small WACC changes, sensitivity tables on the three CAPM inputs reveal how robust your valuation is.
CAPM — the single-factor model behind most cost-of-equity estimates
The Capital Asset Pricing Model, published by William Sharpe in 1964 (and independently by John Lintner and Jan Mossin), prices the trade-off between risk and return for any financial asset. Its central equation — re = Rf + β · (Rm − Rf) — says that an asset's expected return is the risk-free rate plus a risk premium scaled by how much the asset's returns covary with the market. Sharpe shared the 1990 Nobel Prize in Economics for the work. Despite many empirical challenges (the Fama-French three-factor and five-factor models, the Q-factor model, the Stambaugh-Yuan four-factor model — all extend or critique it), CAPM remains the workhorse cost-of-equity model of the corporate-finance curriculum and of most valuation practice.
Why CAPM dominates despite its empirical failures
Three reasons. (1) Simplicity — three inputs, one equation, defensible at a board meeting in 30 seconds. (2) Defensibility — when comp-set analysis or transaction multiples disagree with your DCF, CAPM gives you the language to argue the discount rate. (3) Regulatory acceptance — utility regulators such as the UK's Ofgem and Ofwat and Australia's AER publish CAPM-based "allowed return on equity" determinations, with the beta, risk-free rate and market risk premium argued line by line in the decision documents. The Fama-French models are empirically better at explaining cross-sectional stock returns but rarely show up in regulatory filings or board-pack valuations. CAPM is the language of corporate finance practice — even if academics have moved past it.
A beta is an estimate, not a property of the company. Two data services can show different betas for the same stock on the same day, because each chose its own window, return frequency and index.
Beta — the input that breaks people the most
Beta is the slope of a regression of an asset's returns on market returns. Published "raw" betas have hidden choices baked in: (1) Window and frequency — daily, weekly or monthly returns, over two years or five. Different choices produce different numbers for the same stock. (2) Index — S&P 500 vs Russell 3000 vs MSCI World. Different proxies for "the market" give different betas. (3) Adjusted vs raw — Bloomberg's "adjusted beta" shrinks the estimate toward 1, βadj = 0.33 + 0.67 · βraw, following Blume's (1975) finding that measured betas regress toward the market over time. (4) Levered vs unlevered — levered beta reflects the current debt-equity ratio. To value a target with a different capital structure, you must unlever the comparable's beta (βU = βL / (1 + (1−T)·D/E)) then re-lever to your target structure.
ASEAN-specific considerations
For valuing Singapore-listed companies: use the 10-year Singapore Government Securities (SGS) yield as Rf, the Straits Times Index as the market proxy, and a beta regressed on STI returns. For Malaysia: the 10-year MGS yield and the FBM KLCI. For Indonesia: the 10-year government bond yield and the IDX Composite. Don't apply US inputs to an ASEAN valuation unchanged — the local risk-free rate and a country risk premium shift both ends of the line. Damodaran publishes country risk premiums on his NYU Stern data pages — his table updated 5 January 2026 lists 0% for Singapore, 1.55% for Malaysia and 2.46% for Indonesia — so take the current figure from there rather than from a remembered number.
Rf + β(Rm − Rf): the three inputs, the Blume adjustment and the Fama-French challenge
Published 1964 by William Sharpe in Journal of Finance. Lintner (1965) and Mossin (1966) developed similar models independently. Markowitz's 1952 portfolio theory was the foundation.
1990 Nobel Prize in Economics to Sharpe (jointly with Markowitz + Miller). CAPM made modern finance possible as an academic discipline.
β = 1.0 means market. The index itself has β = 1 by construction; at β = 1 this tool returns exactly Rm, and at β = 0 exactly Rf — the two identities any CAPM result must satisfy.
The market risk premium is Rm − Rf. Damodaran publishes both a historical series (1928 to date) and a monthly implied US ERP; this tool warns when the premium you enter exceeds 15 points.
Bloomberg's "adjusted beta" = 0.33 + 0.67 · βraw, shrinking toward 1 — after Blume (1975), who showed measured betas regress toward the market over time.
Levered vs unlevered beta matters. Published betas reflect current debt-equity. To compare or transfer betas across companies, unlever then re-lever to target structure.
Fama-French (1992) showed CAPM under-explains stock returns — size and value factors matter. Their 3-factor + 5-factor models are academic standards.
Utility regulators such as Ofgem, Ofwat and the AER set "allowed return on equity" with CAPM, publishing the beta and premium they used. Practical staying power.
CAPM is one-period. The original model is single-period — long-horizon DCF applies it tacitly assuming a constant cost of equity, which is a simplification.
Country risk premiums: add a CRP for emerging-market valuations. Damodaran publishes country risk premiums annually on his NYU Stern data pages. This tool computes the single-factor model only; add the CRP yourself.
Frequently asked questions
-
Free sources: Yahoo Finance ("Statistics" → 5-year monthly beta), Stock Analysis, Stooq, WSJ. Free industry betas: Damodaran's data archive (pages.stern.nyu.edu/~adamodar/). Paid: Bloomberg (BETA command), Refinitiv Eikon, FactSet. For private-company valuation, use Damodaran's industry-average unlevered beta, then re-lever to your target capital structure. For valuing comps within a single industry, use a peer-group median unlevered beta and re-lever.
-
There is no fixed number — Rm is the assumption that moves a valuation most, so document how you got it. This page's default is one dated answer: the 10-year Treasury at 5.29% (30 September 2026) plus Damodaran's 3.70% implied premium for 1 October 2026, giving 8.99%. (1) Historical: the long-run mean return of a broad index; Damodaran's historical-returns table gives the S&P 500 series from 1928 with arithmetic and geometric means. (2) Implied: today's risk-free rate plus the implied equity risk premium Damodaran publishes monthly. For an emerging market, add his published country risk premium. Whichever you choose, use the same horizon and currency as your Rf.
-
Three likely causes. (1) Stale beta — your beta may reflect a different growth phase. Check the regression window. (2) Different leverage — comparable's published beta reflects their capital structure; unlever and re-lever to like-for-like. (3) Different Rf or Rm — if comparison is from a year ago, rates have moved. Re-run the comparable with current Rf and Rm. For early-stage companies, CAPM may understate the true cost of equity — the market beta doesn't capture idiosyncratic execution risk, which is why many practitioners add a size premium from a published study rather than relying on beta alone.
-
Levered beta reflects the company's current capital structure (debt + equity). It's what you read from Bloomberg/Yahoo. Use it directly if you're computing the cost of equity for a company at its current capital structure. Unlevered (asset) beta strips out the leverage effect. Use it when (a) valuing a private company by reference to listed comps with different leverage, (b) modelling a target's value at a post-deal capital structure that differs from today's, (c) comparing betas across industries with structurally different leverage profiles. The unlever / re-lever formula: βU = βL / (1 + (1−T) · D/E), then βL,target = βU · (1 + (1−T) · D/E)target.
-
For small-cap and micro-cap valuations, many practitioners extend CAPM with a size premium: re = Rf + β · MRP + SP. Kroll (formerly Duff & Phelps) publishes size-decile premiums annually in its Cost of Capital Navigator; take the current decile figure from there. For private-company illiquidity, a discount for lack of marketability (DLOM) is applied to the valuation rather than to the discount rate. For emerging-market sovereign risk, add a country risk premium per Damodaran's annual estimates. This tool computes pure single-factor CAPM — apply these premiums externally.
-
Three nested layers: (1) CAPM computes cost of equity (re) — the return equity investors demand. (2) WACC blends re with after-tax cost of debt: WACC = (E/V)·re + (D/V)·rd·(1−T), where V = E + D. (3) DCF uses WACC as the discount rate to convert future free cash flows to firm into present value. CAPM is the foundational layer — get it wrong, WACC is wrong, DCF is wrong. Sensitivity analysis on the three CAPM inputs (Rf ±1%, MRP ±1%, β ±0.2) shows how robust your final valuation is to discount-rate uncertainty.
-
Yes — in practice, despite academic critiques. Multi-factor models (Fama-French 3-factor, 5-factor, Carhart 4-factor, q-factor, Stambaugh-Yuan 4-factor) empirically explain cross-sectional stock returns better than CAPM. But they're rarely used in corporate finance because (a) the additional factor premiums are noisy and unstable, (b) board members understand "beta" — they don't understand HML or SMB, (c) regulatory filings, auditor opinions, and fairness opinions universally cite CAPM. For asset-management portfolio construction, multi-factor models dominate. For valuation, M&A advisory, and corporate finance — CAPM still rules.
-
A negative re is theoretically incoherent. It means your inputs imply investors would pay to hold this asset rather than demand a return — only possible if β · (Rm − Rf) is more negative than Rf is positive. Three causes: (a) negative beta (rare — gold, some hedge fund strategies, certain volatility-short products), (b) Rm < Rf (you've made an error — by definition, the market commands a positive expected risk premium over the risk-free rate), or (c) you've used very different time horizons for Rf and Rm. Double-check that all three inputs reflect the same forward horizon and currency.
-
No. The entire CAPM computation — Rf + β · (Rm − Rf), market risk premium decomposition, beta interpretation — runs in JavaScript in your browser. Open DevTools → Network when you click Compute and you'll see zero outbound requests. Safe for confidential M&A valuation work.
-
Sharpe WF. "Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk." Journal of Finance 1964;19(3):425-442. DOI 10.1111/j.1540-6261.1964.tb02865.x. Lintner J. "The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets." Review of Economics and Statistics 1965;47(1):13-37. For textbook treatment: Brealey, Myers and Allen, Principles of Corporate Finance; Damodaran, Investment Valuation. Damodaran's annual "Equity Risk Premiums: Determinants, Estimation and Implications" update is on SSRN and his NYU Stern site.
Method & sources
How it computes
Cost of equity re = Rf + β × (Rm − Rf) (Sharpe 1964). Rm − Rf is the market risk premium; β × (Rm − Rf) is this asset's premium. Computed in percentage points, then shown to 2 dp. Defaults are dated: Rf 5.29% (US Treasury 10-year par yield, 30 Sep 2026) and Rm 8.99% (Rf plus Damodaran's implied premium of 3.70% for 1 Oct 2026). Warns when Rm < Rf, the premium exceeds 15 points, or re < 0. RecaCorpFin.capm.
What this tool implements
- Single-factor CAPM as specified by Sharpe (1964) and Lintner (1965); idiosyncratic risk assumed diversified away and unpriced.
- Definitional identities hold exactly and are tested: β = 0 returns Rf, β = 1 returns Rm.
- Terminology: Rm − Rf is the market (equity) risk premium; β × (Rm − Rf) is the asset's own premium. The page no longer calls the second one "the equity risk premium".
- The default inputs are dated and sourced, not "typical": US Treasury daily par yield curve, 10-year, 30 Sep 2026 = 5.29%; Damodaran's ERPbymonth sheet for 1 Oct 2026 lists a T.Bond rate of 5.29% and an implied premium (trailing-12-month cash yield) of 3.70%, with other variants from 3.30% to 5.67%. The freshness stamp is set to six months because these move.
- ⚠️ THE MODEL IS CONTESTED. Fama and French (1992) found little relation between beta and average US stock returns once size and book-to-market were accounted for. CAPM remains the standard teaching model and the usual input to a WACC.
- ⚠️ Beta is an estimate, not a property: it depends on the window, return frequency and index, and two data services can differ for the same stock on the same day.
Sources
- Sharpe WF. Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk. Journal of Finance 1964;19(3):425-442. doi:10.1111/j.1540-6261.1964.tb02865.x
- Lintner J. The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets. Review of Economics and Statistics 1965;47(1):13-37. doi:10.2307/1924119
- Fama EF, French KR. The Cross-Section of Expected Stock Returns. Journal of Finance 1992;47(2):427-465. doi:10.1111/j.1540-6261.1992.tb04398.x
- Blume ME. Betas and Their Regression Tendencies. Journal of Finance 1975;30(3):785-795. doi:10.1111/j.1540-6261.1975.tb01850.x
- U.S. Department of the Treasury. Daily Treasury Par Yield Curve Rates, 2026 (10-year, 30 Sep 2026: 5.29%): https://home.treasury.gov/resource-center/data-chart-center/inter…
- Damodaran A. Implied equity risk premium by month, ERPbymonth.xlsx (row for 1 Oct 2026): https://pages.stern.nyu.edu/~adamodar/pc/implprem/ERPbymonth.xlsx
- Damodaran A. Country Default Spreads and Risk Premiums (updated 5 Jan 2026): https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/ctry…
- The Nobel Prize. The Sveriges Riksbank Prize in Economic Sciences 1990 — Markowitz, Miller, Sharpe: https://www.nobelprize.org/prizes/economic-sciences/1990/summary/
What can make this go out of date
- Nothing is fetched at runtime. The defaults are a snapshot (Treasury 30 Sep 2026; Damodaran 1 Oct 2026) and are labelled with their dates on the page; replace them with current figures.
Abridged — the full review record for this tool runs longer than the list above.
Pick up where you left off
Stored only in this browser — never sent to our servers.