Estimating the cost of capital requires translating abstract theory into defensible numbers. McKinsey's Chapter 11 and Damodaran's empirical work reveal that practitioners systematically get WACC wrong — using the wrong beta estimation method, anchoring on stale ERP data, ignoring country risk, and confusing book value with market value weights. This lesson covers each component of WACC estimation in the detail that real valuation work demands.
| Company | Raw β | D/E | Tax Rate | Unlevered βU = βL / [1+(1−t)×D/E] |
|---|---|---|---|---|
| Peer A | 1.45 | 0.60× | 25% | 1.000 |
| Peer B | 1.22 | 0.40× | 25% | 0.938 |
| Peer C | 1.68 | 0.80× | 25% | 1.050 |
| Peer D | 1.10 | 0.30× | 25% | 0.898 |
| Peer E | 1.85 | 1.00× | 25% | 1.057 |
| Median βU | Use median to reduce outlier influence | 1.000 | ||
| Relevered βL (target D/E=30/70) | βL = βU × [1 + (1−t) × D/E] = 1.000 × [1 + 0.75 × 0.43] | 1.323 | ||
Beta is the slope of the regression of a stock's excess returns against the market's excess returns. In theory, it is simple. In practice, three choices — the return interval, the index, and the lookback period — each introduce material estimation error. A company's historical beta is also contaminated by its historical capital structure; when capital structure changes (e.g., in an LBO, post-IPO, or strategic pivot), historical beta becomes an unreliable guide to forward-looking systematic risk. McKinsey recommends using industry median unlevered betas, then relevering to the target capital structure.
| Method | Mechanics | When to Use | Key Limitation |
|---|---|---|---|
| OLS Regression (raw beta) | Regress 60 months of weekly stock returns vs. S&P 500; slope = raw beta | Listed companies with 5+ years of stable capital structure | Noisy — single stock has high standard error; affected by corporate events |
| Adjusted (Blume) Beta | Adjusted β = 0.67 × Raw β + 0.33 × 1.0 | Default for listed companies — corrects mean-reversion tendency of raw beta | Arbitrary correction factor; doesn't adjust for capital structure |
| Industry Comparable Unlevering/Relevering | Step 1: Collect peer betas. Step 2: Unlever each: βU = βL / [1+(1−t)×D/E]. Step 3: Median βU. Step 4: Relever at target structure | Private companies; companies changing capital structure; any situation where historical capital structure ≠ target | Requires clean peer group; assumes comparable operating risk |
| Damodaran's Industry Beta Database | Pre-computed median unlevered betas by sector, updated annually | Sanity check; industries with few comparables; emerging markets peers | Sector categories can be broad; US-centric by default |
| Bottom-Up Beta | Weight segment betas by revenue/EBIT contribution; unlever each segment; relever total | Multi-segment conglomerates; companies entering new businesses | Requires judgment on segment weights and comparable selection per segment |
Using the raw historical beta directly — without unlevering and relevering — when the company's capital structure has changed. Example: a company that was 70% debt-financed historically may be modeling a target of 30% debt. Its historical levered beta of 2.1 reflects the old leverage, not the future structure. The correct process: (1) unlever the historical beta using the historical D/E ratio and tax rate; (2) relever using the target D/E ratio. βU = 2.1 / [1 + (1−0.25) × (70/30)] = 2.1 / 2.75 = 0.76. At 30% target debt: βL = 0.76 × [1 + (1−0.25) × (30/70)] = 0.76 × 1.32 = 1.00. The forward cost of equity is dramatically lower — missing this step overstates WACC for de-levering companies.
WACC weights equity and debt by their share of total enterprise value — not by book value. This is one of the most common errors in practice. Book value equity reflects historical accounting accumulation; market value equity reflects what investors currently price the equity claim as worth. Using book value weights systematically understates equity's weight (because equity tends to trade above book) and overstates debt's weight, resulting in a lower WACC that overstates enterprise value.
| Item | Book Value Approach | Market Value Approach | Impact |
|---|---|---|---|
| Equity | $200M (book) | $800M (market cap = 4× book) | Market value 4× higher |
| Debt | $300M (book ≈ market) | $300M (market value) | Similar for investment-grade debt |
| Total Capital | $500M | $1,100M | |
| Equity Weight | 40% (200/500) | 72.7% (800/1,100) | +32.7 percentage points |
| Debt Weight | 60% (300/500) | 27.3% (300/1,100) | −32.7 percentage points |
| WACC (Ke=10%, Kd=5%, t=25%) | 40%×10% + 60%×5%×(1−25%) = 6.25% | 72.7%×10% + 27.3%×5%×(1−25%) = 8.29% | +2.04% — book value approach dramatically underestimates WACC |
Using market value equity in WACC creates an apparent circularity: WACC depends on equity value, but equity value depends on WACC (via the DCF). In theory, this requires iteration; in practice, there are two clean solutions. First, use a target capital structure based on the company's debt policy and peer group — this breaks the circularity because the target structure doesn't depend on the current equity price. Second, if modeling explicitly: set initial WACC → compute equity value → update weights → recompute WACC → iterate until convergence (usually 3-4 iterations). McKinsey consistently recommends the target capital structure approach as simpler and more conceptually defensible.
Key Takeaways
A company has a raw regression beta of 1.8. Its historical capital structure was 60% debt, 40% equity, with a 25% tax rate. The target capital structure for valuation purposes is 30% debt, 70% equity. What is the correctly relevered beta for the WACC calculation?