WACC (Weighted-Average Cost of Capital)

Key takeaways

WACC = (E/V) × Cost of Equity + (D/V) × Cost of Debt × (1 − Tax Rate). Always use market values for the weights, never book.
Cost of equity comes from CAPM: Rf + β × ERP. Beta is regressed over 2–5 years against a broad index.
Typical ranges: 7–9% for large-cap stable, 9–12% for growth, 12–15% for high-risk. If your computed WACC is outside 6–14%, double-check before discounting.
Small WACC errors cascade — a 100 bps change moves DCF enterprise value by ~10–20%. It's the most leveraged input in the model.
For DCF and NPV decisions, every project should compare its IRR to WACC. Above WACC = creates value; below WACC = destroys value.

The formula

WACC = (E / V) × R_e + (D / V) × R_d × (1 − t)

Where:

E — market value of equity (market cap)
D — market value of debt (book is acceptable for non-traded debt)
V — total firm value (E + D)
R_e — cost of equity (from CAPM, see below)
R_d — pre-tax cost of debt
t — marginal tax rate (typically the corporate statutory rate of the country of operation)

Cost of equity (CAPM)

The Capital Asset Pricing Model is the standard framework for cost of equity:

R_e = R_f + β × (R_m − R_f)

R_f — risk-free rate. Use the 10-year government bond yield in the company's reporting currency (10Y Treasury for US, 10Y Bund for Eurozone, etc.).
β — beta. Regression of the stock's returns against a broad market index (S&P 500, MSCI World) over 2–5 years using monthly or weekly returns. Bloomberg's "raw beta" is the standard practitioner reference. Damodaran publishes industry betas free.
R_m − R_f — equity risk premium (ERP). The expected return on the broad market in excess of the risk-free rate. Damodaran's implied ERP (computed monthly from current S&P 500 prices and analyst estimates) is the most defensible reference. Typical: 5.0–6.0% in developed markets.

Levered vs unlevered beta: if you're computing a synthetic beta from peers (because the target is private or has been acquired), unlever each peer's beta by its capital structure, take the median, then re-lever to the target's capital structure. Standard formulas: β_U = β_L / [1 + (1 − t)(D/E)].

Cost of debt

The pre-tax rate at which the company can borrow today — not the historical average rate on outstanding debt. Three approaches in order of preference:

YTM on outstanding bonds. If the company has actively traded bonds, the yield to maturity is the market's view on cost of debt.
Implied from credit ratings. Look up the spread for the company's credit rating (S&P / Moody's / Fitch) over a Treasury benchmark, add to the risk-free rate.
Synthetic credit rating. For private companies, compute interest coverage and leverage ratios, map to an implied rating using Damodaran's tables, then add the spread.

Multiply by (1 − t) to get the after-tax rate. This reflects that interest is tax-deductible — the government effectively subsidises debt.

Why market values, not book

The textbook says market values; many models use book; the gap matters. WACC is the rate the market demands today on the company's capital structure today — and the market values today's equity at the market cap, not the historical book equity from the 10-K.

A company with $1B book equity and $5B market cap has dramatically different WACC weights depending on which you use. Mixing book and market — say, book debt and market equity — produces a WACC that doesn't reflect any actual world. Pick market for both.

Worked example

Apple, illustrative numbers: market cap $3,000B, debt $100B, beta 1.2, risk-free rate 4.25%, ERP 5.5%, pre-tax cost of debt 4.5%, tax rate 21%.

Component	Value
Cost of equity	4.25% + 1.2 × 5.5% = 10.85%
After-tax cost of debt	4.5% × (1 − 0.21) = 3.55%
Equity weight	3,000 / 3,100 = 96.8%
Debt weight	100 / 3,100 = 3.2%
WACC	0.968 × 10.85% + 0.032 × 3.55% = 10.62%

Apple's debt is so small relative to its equity that the WACC is essentially its cost of equity. Most large-cap tech names sit in this range.

Common WACC errors

Mixing book and market values. The single most common silent error. Equity weight should be market cap, not book equity.
Using stale beta. Beta from 5 years ago doesn't reflect today's risk profile, especially for companies post-IPO or post-restructuring.
Wrong risk-free rate currency. If you're valuing a Eurozone company, use a Eurozone risk-free rate, not a 10Y Treasury.
Forgetting the tax shield. After-tax cost of debt is R_d × (1 − t), not R_d. Forgetting overstates WACC.
Net-cash company gets a weird WACC. If cash exceeds debt, the debt weight goes negative. WACC reduces to roughly the cost of equity; that's correct, but check the math.

How Smalt AI builds it

Smalt AI's DCF and project-NPV models include a dedicated WACC sheet that pulls inputs from the assumptions tab, computes cost of equity via CAPM (with documented sources for risk-free rate, beta, and ERP), looks up after-tax cost of debt, computes market-value weights, and outputs the blended rate. Sensitivity tables on WACC × terminal growth and WACC × exit multiple are standard. The computed WACC is range-checked: outside 6–14%, the model flags it for review before delivery.