Module 04 produced the future cash flows. This module produces the rate at which to discount them. The cost of equity via CAPM, the after-tax cost of debt, and the weighted-average that combines them — WACC — is the single most-debated number in valuation. Get it wrong and every downstream answer is wrong.
Foundations Module 03 introduced the time value of money: a dollar today is worth more than a dollar tomorrow because today's dollar can be invested. That intuition was demonstrated with risk-free rates — the kind of rates implied by government bonds, where repayment is essentially certain. But corporate cash flows aren't risk-free. A company that promises to pay you $100 next year might pay $120, or $80, or nothing at all. Risk requires a different rate.
The cost of capital is that different rate. It is the rate of return investors require for putting capital into a particular firm, given the risk of that firm's cash flows. Three observations make this concrete:
The cost of capital is the price of risk-bearing — the compensation investors require for accepting uncertainty about future cash flows. It's the discount rate that converts projected cash flows into present value.
The same number — cost of capital — appears under two different names depending on whose perspective you're taking:
| From the firm's perspective | From the investor's perspective |
|---|---|
| It's the cost of attracting capital — what the firm must pay (in interest, in expected returns to shareholders) to get capital providers to commit. | It's the required return — the minimum compensation an investor demands to accept the risk of putting money into this firm. |
| Cost of debt = interest rate the firm pays on borrowing. | Yield to maturity = return a bondholder earns by holding the debt. |
| Cost of equity = expected return shareholders need to be willing to hold the stock. | Expected return on equity = what shareholders forecast they'll earn. |
The two perspectives are mirror images. In equilibrium, the firm's cost of capital equals the market's required return. When they diverge — when investors demand more than the firm is willing to pay — the firm can't raise capital at acceptable terms, and the investment doesn't happen. The cost of capital is the equilibrium price where supply of and demand for capital meet.
By the end of this module you will be able to compute the weighted-average cost of capital (WACC) for any public company. WACC is the blended cost of all the firm's capital — debt and equity, weighted by how much of each the firm uses. WACC is the discount rate that goes into Module 07's discounted cash flow valuation. The mechanics are not complicated; the judgment calls are where most analysts fall down.
The path through this module:
The interactive WACC calculator at the end lets you flex every input live and watch the answer change. The quiz tests not memorization but applied reasoning — exactly the kind of judgment calls that come up when you're computing a real WACC for a real decision.
Cost of debt is observable — bondholders earn interest at a contractual rate. Cost of equity is harder. Shareholders don't have a contractual return; they have a residual claim on profits and a price that bounces around. So how do you estimate what equity holders require as a return?
The standard answer in modern finance is the Capital Asset Pricing Model (CAPM), developed by Sharpe, Lintner, and Mossin in the 1960s. CAPM is simultaneously the most-used and most-criticized model in finance. Practitioners use it because nothing else is as simple, as theoretically grounded, and as widely accepted. They criticize it because its empirical performance is mixed and its assumptions are heroic. Both reactions are correct.
CAPM says the expected return on an equity investment equals the risk-free rate plus a risk premium scaled by the investment's exposure to market risk:
Where Rf is the risk-free rate (typically a 10-year Treasury yield), β (beta) is the equity's sensitivity to market movements, and (Rm − Rf) is the equity risk premium — the average excess return investors have historically earned by holding stocks rather than risk-free debt.
The three pieces:
The risk-free rate is the return on an investment with no default risk. In practice, this means government bonds in the firm's home currency. For US-domiciled firms, it's the US Treasury yield. For European firms, Bund yields (or French OAT, Italian BTP). For UK firms, Gilts. For Japanese firms, JGB yields.
The choice of maturity matters. The convention is the 10-year rate because:
For US-based valuations in May 2026, the 10-year Treasury yields somewhere in the 4-5% range — that's the typical anchor. Lower in deflationary environments (Japan, parts of Europe in the 2010s); higher in inflationary environments (Türkiye, Argentina). Always use the current rate, not a historical average.
The equity risk premium is the excess return investors earn (or expect to earn) for holding stocks rather than risk-free bonds. Historically, US stocks have outperformed Treasuries by roughly 5-7% annually over very long periods (a hundred years or more). That historical excess return is one estimate of the ERP. Forward-looking estimates derived from current dividend yields and growth expectations give somewhat lower numbers (3-5%). The two camps argue, but for practical purposes, most US-based DCF models use an ERP of 5-6%.
Three things to know about the ERP:
Beta is the third piece — and it's where most of the analyst's effort goes. Beta measures how an equity's returns move relative to the broader market. A beta of 1.0 means the stock moves in lockstep with the market: when the market is up 10%, the stock is up 10% on average. A beta below 1.0 means the stock moves less than the market; above 1.0 means it moves more.
Utilities, consumer staples, established healthcare. Stable cash flows that don't bounce with the economy. Lower required return.
Broad-market firms whose performance tracks the economy. The default for unknown businesses. Average required return.
Tech, semiconductors, autos, banks, airlines. Earnings amplify economic swings. Higher required return.
Consider the Sample Company from Module 04 (a maturing consumer/industrial firm). Suppose its observed beta is 1.1 — slightly more cyclical than the market. With a 10-year Treasury yield of 4.5% and an ERP of 5.5%:
Re = 4.5% + 1.10 × 5.5%This is the rate Sample Company's shareholders require to be willing to hold the stock. Equity capital costs the firm 10.55% per year. That's the number that goes into the WACC calculation in Section 05.
CAPM is a model, not a truth. Three substantial criticisms apply:
Empirical fit is mediocre. Decades of research (Fama-French and others) find that CAPM systematically misprices several categories: small-cap stocks, value stocks, low-volatility stocks. The Fama-French three-factor and five-factor models correct some of these issues by adding extra risk factors. Beta isn't stable. Beta estimates change substantially depending on the time window, the frequency of returns, and the choice of market index. The CAPM assumes investors hold diversified portfolios, can borrow at the risk-free rate, agree on risk and expected returns, and face no taxes or transaction costs. None of these is even approximately true. Despite all this, CAPM remains the practitioner standard because the alternatives have their own problems and CAPM is at least a defensible starting point.
The right response to CAPM's limitations is not to abandon it but to use it carefully — pair it with sensitivity analysis (testing whether the conclusion holds across a range of beta and ERP estimates), with sanity checks against industry comparables, and with explicit acknowledgment of the assumptions. The next section dives into the practical mechanics of estimating beta, where most of the analyst's discretion gets applied.
Beta sounds simple — the slope of the regression line of stock returns on market returns — but the practical estimation involves five judgment calls that can move a single firm's beta estimate from 0.8 to 1.6 depending on choices. The five:
For a public firm with a long trading history, you can estimate beta by direct regression — running the stock's returns against the market's. This is what Bloomberg, FactSet, and Capital IQ produce. For shorter-history firms, private firms, or business segments without their own public stock, you can't run a direct regression. The alternative is the industry beta approach:
This produces a beta estimate that captures business risk (industry-level) without the noise of any single firm's regression. Many practitioners prefer industry beta even for public firms because it's more stable and less subject to estimation noise.
A firm's observed equity beta reflects two sources of risk: the underlying business (operating risk) and how the business is financed (leverage). Leverage amplifies equity returns: a firm with more debt has a more volatile equity return for any given operating performance. Two firms with identical businesses but different capital structures will show different betas — and the difference is purely financial, not operational.
The unlevering / relevering formulas correct for this:
βL = levered (observed) beta. βU = unlevered (asset) beta. T = tax rate. D/E = debt-to-equity ratio (using market values). The formulas assume debt is risk-free; corrections exist for risky debt but are rarely material for investment-grade firms..
The unlevered beta — also called asset beta — captures pure business risk independent of capital structure. Industry comparable firms with different leverage all share roughly similar asset betas, even though their levered betas differ. The relevered beta applies the target firm's specific capital structure to that industry-average asset risk.
Suppose you're estimating beta for a private retailer. You identify a comparable public retailer with observed equity beta of 1.20 and capital structure 30% debt / 70% equity (so D/E = 0.43). The public firm's tax rate is 25%. Your target firm has D/E = 0.25 (less leverage). The procedure:
βU = 1.20 ÷ [1 + (1 − 0.25) × 0.43] = 1.20 ÷ 1.32 = 0.91βL,target = 0.91 × [1 + (1 − 0.25) × 0.25] = 0.91 × 1.19 = 1.08The relevering captures the right idea: the target's lower leverage means its equity is less risky than the comparable's, so its beta is lower. If the comparable's leverage were equal to the target's, the betas would be the same. The mechanics get more elaborate when you average multiple comparables, but the principle is identical.
For private firms or for individual segments of diversified public firms, the industry-beta approach is the only practical option. The procedure: identify pure-play public comparables, unlever each to get asset betas, average the asset betas, then relever to the target's specific capital structure. This is how analysts produce reasonable WACCs for private companies, divisions of conglomerates, or pre-IPO firms.
Choosing comparables is itself a judgment call. A "consumer-products firm" could mean Procter & Gamble (stable, low-beta) or a small DTC startup (volatile, high-beta). The typical rule: pick comparables matched on (a) business model, (b) market segment, (c) firm size, (d) financial profile. The comparable set should ideally be at least 4-6 firms; a single comparable produces too noisy an estimate. The Module 06 work on multiples valuation returns to the comparables-selection problem in greater detail — same skill, different application.
Debt is easier than equity. There's a contractual interest rate, a known maturity, and a stated principal. Most of the work has been done by the bond market — your job is to read the market's pricing of the firm's debt and adjust appropriately.
The first thing to know: the coupon rate printed on a bond is not the cost of debt. The coupon was set when the bond was issued, possibly years ago, when interest rates were different. The cost of debt today is the yield to maturity (YTM) — the return a buyer of the bond would earn if they held it to maturity at today's market price.
Concrete example. A firm issued a 10-year bond five years ago with a 3% coupon. Interest rates have risen since; the bond now trades at 92 cents on the dollar. The 3% coupon is irrelevant — the firm's marginal cost of debt today is the YTM on that traded bond, which would be approximately 5% (the buyer pays 92, gets 100 back at maturity, plus 3% coupons along the way). If the firm wanted to issue new debt today, it would have to offer roughly 5% to attract investors.
The cost of debt is always the marginal cost — what new debt would cost — not the average cost of existing debt. Existing debt was priced under different conditions; new debt prices under current conditions. Use the latter.
| Method | When it works | Limitations |
|---|---|---|
| Direct YTM observation | Firm has publicly traded bonds with sufficient liquidity | Requires the bond to actually trade. Many firms have illiquid bonds where the YTM is unreliable. |
| Credit-rating implied yield | Firm has a credit rating but no traded bond, or bond is illiquid | Use the average yield on the index of bonds with the same rating (e.g., BBB corporate index). Approximate but reasonable. |
| Synthetic credit rating | Firm has neither traded debt nor a credit rating (private firm, smaller public firm) | Estimate creditworthiness from financial ratios (interest coverage, debt/EBITDA), assign a synthetic rating, and use that rating's index yield. |
For most public companies, the second method is the practical default: pull the firm's credit rating from S&P, Moody's, or Fitch, then look up the average yield on bonds with that rating in the firm's currency. Bloomberg, Capital IQ, and the Federal Reserve all publish these indices. A BBB-rated US corporate bond yielded approximately 5.5% in mid-2026; an A-rated about 5.0%; a BB about 7%. The credit rating is a shorthand for the market's assessment of default risk.
The single most important fact about the cost of debt: interest is tax-deductible. When a firm pays $100 of interest, its taxable income falls by $100, which (at a 25% tax rate) saves $25 in taxes. The net cost to the firm is therefore $75 — the after-tax cost. The before-tax interest rate of 5% becomes an after-tax cost of 3.75% at a 25% tax rate.
Where Rd is the pre-tax cost of debt (the YTM or rating-implied yield) and T is the firm's marginal tax rate. This is the cost that goes into WACC — the actual cash cost of borrowing after accounting for the tax shield on interest.
The tax shield on debt is a real economic benefit, and it's why debt looks cheap relative to equity. A firm whose cost of equity is 11% and pre-tax cost of debt is 5% sees an after-tax cost of debt of 3.75% — three times cheaper than equity. This wedge between the two costs is the central force in capital-structure theory (Module 10), and it's also what makes WACC genuinely lower for moderately leveraged firms than for unleveraged ones.
Sample Company has a single bank loan of $200M outstanding, with a 5% interest rate. The firm is privately rated as a BBB-equivalent. Current BBB-corporate yields are 5.5%. The firm's marginal tax rate is 25%.
5.5% × (1 − 0.25) = 4.125%Note that we used 5.5% (the current marginal rate on BBB debt), not 5% (the existing loan's contractual rate). The cost of capital is forward-looking — what would the firm pay to raise additional debt today? If the answer is 5.5%, that's what flows into WACC.
The after-tax adjustment in WACC is the only place the interest tax shield should appear. If you're already discounting unlevered free cash flow (FCFF, from Module 04), the tax shield is captured by using the after-tax cost of debt in WACC. Adjusted Present Value (APV) is a separate methodology that explicitly values the tax shield as a separate cash flow — but if you're using APV you discount with the unlevered cost of capital, not WACC. Mixing the two methods is a common error that leads to double-counting.
The Weighted-Average Cost of Capital combines the cost of equity and the after-tax cost of debt, weighted by the firm's capital structure. WACC is the single discount rate that goes into enterprise-value DCF — it represents the blended required return across all capital providers.
Where E = market value of equity, D = market value of debt, V = E + D (total enterprise value at market), Re = cost of equity (Section 02), Rd = pre-tax cost of debt (Section 04), T = marginal tax rate. The weights E/V and D/V represent the firm's capital structure.
The math is simple. The judgment is in the inputs:
Combining everything from Sections 02-04, plus Sample Company's capital structure (200M debt, 800M market equity):
E = $800M, D = $200M, V = $1,000MWACC = 0.80 × 10.55% + 0.20 × 4.125%= 8.44% + 0.825% = 9.27%This is the rate that goes into Module 07's DCF valuation. Sample Company's projected free cash flows from Module 04 — $50M in Year 1 growing to $125M in Year 8 — get discounted at 9.27% to produce a present value. That present value, plus a terminal value (Module 07), gives enterprise value.
Notice what happens to WACC if Sample Company's capital structure changes. The same firm with 50% debt / 50% equity — instead of 20% / 80% — would have:
WACC = 0.50 × 10.55% + 0.50 × 4.125% = 5.275% + 2.06% = 7.34%
That's nearly 200bp lower. More debt means more weight on the cheaper capital source, which lowers WACC. This is why companies with more debt look more valuable on a DCF basis — the lower discount rate produces a higher present value of future cash flows. Module 10 explores why firms don't just lever up to infinity (the answer involves financial distress, agency costs, and bankruptcy risk).
Modigliani and Miller proved (in their 1958 paper) that, in a world without taxes or distress costs, capital structure doesn't matter. The increased equity beta from higher leverage exactly offsets the lower-cost debt; WACC stays constant regardless of D/E ratio. With taxes, the tax shield on interest tilts the calculation in favor of debt, and WACC falls (modestly) as leverage rises. With distress costs, the calculation tilts back: very high leverage raises the probability of bankruptcy, which is a real cost that offsets the tax benefit. The optimal capital structure is somewhere in between — the topic of Module 10.
For now, the practical takeaway: WACC depends on the capital structure. Use the right structure (current market values, or target if you have a credible target). Don't accept a WACC computed from book-value weights or from a stale historical capital structure.
"V" in WACC is total enterprise capital value at market — equity market cap plus market value of debt. It's not the same as the firm's "enterprise value" used in EV/EBITDA multiples (which subtracts excess cash). The WACC weighting uses the broader notion: total capital deployed in the firm. This is a small but important distinction when you're working through valuations carefully.
The CAPM and WACC frameworks above implicitly assume the firm operates in a single, mature, low-risk economy. For most US, UK, German, or Japanese firms, that assumption is roughly fine. For firms operating in emerging markets or in countries with sovereign or political risk, the standard framework needs adjustments.
The basic problem: a Brazilian firm and a US firm in the same industry, with the same business model, do not have the same cost of capital. The Brazilian firm's cash flows are exposed to currency volatility, political instability, occasional capital controls, and the small but real risk of sovereign default. Investors require additional compensation for those risks. The CAPM and ERP estimates calibrated to US data don't capture this — they need adjustment.
| Approach | Mechanics | When to use |
|---|---|---|
| Country risk premium added to ERP | Add a country risk premium (CRP) to the standard ERP. CRP is typically the sovereign credit-default-swap spread, scaled by the equity-vs-bond volatility ratio (Damodaran's approach). | Most common. Easy to apply. Simple to communicate. |
| Sovereign-spread adjustment to risk-free rate | Use the local-currency government bond yield as Rf, which already reflects sovereign risk. Or use US Treasury + sovereign spread. | When the firm borrows substantially in local currency and FX hedging would be impractical. |
| Ad-hoc cost-of-capital floor | Practitioner shortcut: emerging-market firms get a 12-15% WACC regardless of the formal calculation. Sometimes used in private-equity contexts. | Quick estimates. Not defensible in a formal report; useful for sanity checks. |
The most widely used framework is Aswath Damodaran's. The country risk premium has two components:
Where the sovereign default spread is the country's USD-denominated bond yield minus the US Treasury yield (or equivalently, the sovereign CDS spread), and σequity ÷ σbond is the relative volatility of the country's equity market vs. its bond market — typically 1.3 to 1.5x for emerging markets.
Brazil's sovereign spread in mid-2026 might be roughly 250bp, with an equity-vs-bond volatility ratio of about 1.4. So the country risk premium for Brazil would be approximately 350bp (2.5% × 1.4). For a Brazilian firm, the CAPM cost of equity becomes:
Re = RfUS + β × (ERPUS + CRPBrazil)
or alternatively Re = RfUS + β × ERPUS + λ × CRPBrazil, where λ is a coefficient measuring how much the specific firm is exposed to Brazilian country risk (multinationals with diverse revenue might have λ < 1; pure-play domestic firms have λ ≈ 1). The choice between the two depends on how the analyst views country-risk exposure.
The judgment call is in deciding which category the target firm falls into, and applying the framework accordingly. Section's six country case cards below illustrate how the cost of capital varies meaningfully across major economies — useful background for any cross-border valuation work.
WACC is the most-used and most-misused number in corporate finance. Six mistakes account for the bulk of the trouble:
Computing E/V and D/V from balance-sheet book values rather than market values. Book equity is often dramatically smaller than market equity (especially for tech firms), inflating the debt weight and understating WACC. Fix: always use market capitalization for E.
Using corporate WACC to evaluate a project in a different risk profile. A safe stable utility's WACC misapplied to a high-risk biotech project will systematically over-invest in risky projects. Fix: use a project-specific or divisional WACC reflecting the actual risk being underwritten.
Discounting 30-year cash flows at a rate built from the 10-year Treasury, then adding a "long-term" adjustment. Inconsistent. Fix: match the risk-free rate to the duration of cash flows, or rebuild the WACC at each segment of the projection.
Using last year's D/E ratio when the firm has changed leverage substantially (recent acquisition, recapitalization). The WACC then reflects an outdated reality. Fix: use the target capital structure for forward-looking valuation, or the current market structure for transactional analysis.
Discounting nominal cash flows (which include inflation) at a real discount rate (which excludes it). Or vice versa. The mismatch produces meaningless valuations. Fix: match cash flow basis to discount rate basis. Almost all DCF work uses nominal-on-nominal.
Approving any project whose IRR exceeds WACC. Real businesses face execution risk, optimism bias, and uncertainty that WACC doesn't fully capture. Most firms use a hurdle rate above WACC (often 200-400bp above) for capital-allocation decisions. Fix: add a margin to WACC for project approval thresholds; reserve raw WACC for valuation.
The distinction between WACC (a valuation rate) and the project hurdle rate (a capital-allocation threshold) deserves a clear summary. WACC tells you: "if I deploy capital today and earn this return, I exactly compensate my investors for the risk they're taking." A hurdle rate tells you: "if I deploy capital today, I want a margin above what investors require, because real-world projects underperform plans on average." Both are relevant; they answer different questions.
The Foundations track Module 06 (Capital Budgeting Capstone) used WACC-equivalent rates for NPV decisions — that's the right approach when you're confident in the cash-flow estimates. Real corporate practice often adds 200-400bp on top to account for execution and optimism risk. The judgment is context-specific.
The cost of capital varies meaningfully across countries — not just because beta and ERP differ, but because the entire risk environment differs. Here's how six representative markets compare for a hypothetical mid-cap industrial firm:
10-year Treasury around 4.5%, ERP about 5.5%, mid-cap equity beta ≈ 1.0, BBB credit spread ~150bp. WACC for a representative mid-cap is typically 7-9%. The deepest, most liquid capital markets in the world; the place CAPM was developed and where it works best. The US is the implicit benchmark for cross-border comparisons.
Swiss government bonds traded at negative yields throughout much of the 2010s and into the 2020s. Even in a more normalized environment, Swiss yields run substantially below US Treasuries. WACC for Swiss firms can be 200-300bp below comparable US firms — though the lower-risk/lower-return tradeoff makes the difference somewhat illusory in real terms once Swiss inflation is factored in.
A generation of deflation, near-zero rates, and policy-driven yield-curve control made Japanese cost-of-capital calculations quite distinctive. JGB yields ranged near zero for decades. Equity risk premium debates in Japan have been particularly contentious — historical equity returns are weak, making backward-looking ERP estimates produce uncomfortably low numbers. Many practitioners use forward-looking estimates of 4-6% nonetheless.
Brazil's sovereign-spread component (~250bp at typical levels) plus the equity-bond volatility multiplier produces an emerging-market country risk premium of ~350bp on top of the US ERP. A Brazilian mid-cap industrial might have WACC around 13-15% — substantially above its US counterpart, even though the underlying business may be similar.
Inflation in Türkiye has run between 30-90% in recent years. Local-currency government bond yields exceed 25%. The standard CAPM completely breaks down: at these rates, virtually no domestic project can clear the hurdle. Practitioners typically value Turkish firms in USD terms (translating cash flows) or use real (inflation-adjusted) cash flows discounted at real rates. The choice is consequential and not standardized.
Chinese capital markets have features that complicate standard cost-of-capital calculations: state-owned enterprises with implicit government backstops (effectively reducing default risk for the largest borrowers), capital controls that prevent free arbitrage, and a single-A or BBB sovereign rating that doesn't capture the full geopolitical risk picture. Practitioners often add an opaque "China premium" of 100-300bp to formal calculations.
The pattern across all six: the same business operating in different countries has different costs of capital. The cross-border analyst must choose the right risk-free rate, the right ERP (with country risk premium where appropriate), the right beta, and the right capital structure — all calibrated to the specific market. There's no universal WACC formula that works everywhere; there's a single framework with country-specific inputs.
You now have everything needed to discount cash flows: a projection (Module 04) and a discount rate (this module). Module 06 takes a different approach to valuation — using market multiples (EV/EBITDA, P/E, EV/Revenue) rather than discounting cash flows. Multiples are the practitioner's quick-and-dirty alternative to DCF, with their own strengths and weaknesses. Module 07 will then bring it all together in a complete DCF valuation. Cost of capital is the linchpin connecting Modules 04 and 07.
Set the seven inputs that determine WACC and watch the calculation update live. The Sensitivity panel below shows how WACC flexes when you change each input by ±100bp — useful for identifying which inputs matter most. The default values reflect a typical US mid-cap industrial; modify them to model a specific firm.
The questions test applied judgment — not formulas, but the kind of decisions that arise when you're computing a real WACC for a real situation. Cost of capital is where numerical precision meets practical wisdom.