Why "bottom-up" beta beats regressing your own stock
The textbook way to get a beta is to regress a company's own stock returns against the market. In practice that estimate is often noisy — a couple of years of data, a few unusual quarters, and the number moves around more than the company's actual business risk did. It's also useless for a private company or a new division with no trading history at all.
The fix: take the betas of several comparable public companies, strip out each one's own leverage (unlever), average the results — business risk without financing risk — and then relever that average at your own capital structure. The betas of several peers, averaged, are far more stable than any single noisy regression, and the technique works even when your own company has never traded a day in its life.
Why the interest coverage ratio stands in for a credit rating
Cost of debt should reflect default risk, which shows up in real bond markets as a credit spread over the risk-free rate. Most private companies and smaller public ones don't have a rated bond to read that spread off directly — but the interest coverage ratio (EBIT ÷ interest expense) correlates closely with actual credit ratings, because it's measuring the same underlying thing: how comfortably a company's operating profit covers its debt payments. Mapping coverage to an approximate rating, then a rating to a typical spread, gets you a reasonable cost of debt without needing an actual bond outstanding.
The honest limitations
Every number here is an estimate built from other estimates — the equity risk premium in particular is genuinely debated among practitioners, and reasonable people use figures a percentage point or more apart. The synthetic rating spreads are illustrative, not a live market feed, and real spreads widen and narrow with credit-market conditions. Small changes in WACC compound into large changes in a DCF's output, so treat the number this tool gives you as the center of a reasonable range, not a precise fact — and stress-test your valuation across a spread of plausible WACCs using the sensitivity table in our DCF model.