Company Valuation

DCF Methodology — Detailed Reference

Expands on the summary in SKILL.md. Use this when building the DCF build table or when the user asks for industry-specific treatment.

When DCF Is Appropriate

Good fit:

• Mature companies with predictable cash flows

• Companies whose revenue and margin trajectory can be estimated within a reasonable confidence band

• Strategic valuations requiring intrinsic value assessment

• Cross-checking a relative valuation

Poor fit:

• Pre-revenue / early-stage (no cash flow history)

• Banks, insurance (use DDM or excess return model)

• REITs (use NAV)

• Highly cyclical businesses without a clear cycle baseline — use mid-cycle earnings instead

Projection Model (5-Year Explicit Forecast)

Revenue projection

1. Compute historical 3–5 year CAGR.

2. Pull analyst consensus from yfinance.Ticker.revenue_estimate.

3. Consider industry growth, competitive position, and company guidance.

4. Project revenue for Y1–Y5, fading linearly toward terminal growth rate.

Revenue_t = Revenue_{t-1} × (1 + g_t)

EBIT and Free Cash Flow Build

Revenue
- COGS                          → historical gross margin trend
= Gross Profit
- SG&A                          → historical SG&A % of revenue
- R&D                           → historical R&D % of revenue
- Other OpEx
= EBIT (Operating Income)

FCFF = EBIT × (1 − Tax Rate)
     + Depreciation & Amortization
     + Stock-Based Compensation    ← only if treating SBC as non-cash
     − Capital Expenditures
     − Change in Net Working Capital

Assumption checklist (state explicitly)

WACC Calculation

WACC = (E/V) × Ke + (D/V) × Kd × (1 − Tax Rate)

Cost of Equity (CAPM)

Ke = Risk-Free Rate + Beta × Equity Risk Premium + Size Premium (if applicable)

Cost of Debt

• Preferred: interest expense / total debt from financials.

• Fallback: credit rating spread over risk-free rate.

• Investment-grade: 4–6%. High-yield: 7–10%.

Capital structure

Use market values:

• E = market cap

• D = total debt (balance sheet)

• V = E + D

Terminal Value

Method 1: Perpetuity Growth (Gordon Growth)

TV = FCFF_5 × (1 + g) / (WACC − g)

• Terminal growth g: 2–3% typical; must not exceed long-term GDP (~2.5% US, ~3–4% EM).

• TV normally represents 60–80% of total EV. Flag if outside that range.

Method 2: Exit Multiple

TV = EBITDA_5 × exit EV/EBITDA multiple

• Use current peer trading multiples as reference.

• Apply discount for growth deceleration by Y5.

• Cross-check against Gordon TV — if they diverge by >30%, reconcile assumptions.

Bridge to Equity Value

PV of FCFF = Σ FCFF_t / (1 + WACC)^t  for t = 1..5
PV of TV   = TV / (1 + WACC)^5

Enterprise Value = PV of FCFF + PV of TV
+ Cash & equivalents
− Total debt
− Minority interest
− Preferred stock
+ Equity investments (if material)
= Equity Value

Implied share price = Equity Value / diluted shares outstanding

Sensitivity & Scenarios

WACC × Terminal Growth matrix

5×5 grid. Vary WACC by ±1% in 0.5% steps and g by 0.5% from 1.5% to 3.5%. Highlight base case.

Scenario analysis

Industry-Specific Guidance

Technology / SaaS

• EV/Revenue often more meaningful than P/E if not yet profitable.

• Key metrics: ARR growth, net dollar retention (NRR), Rule of 40 (growth% + FCF margin ≥ 40).

• CapEx light (3–8% rev); R&D heavy (15–30%).

• SBC material — decide cash vs non-cash upfront and disclose.

• Terminal growth: 3–4% for category leaders, 2–3% others.

Retail / E-commerce

• Revenue = same-store sales growth + new store openings (physical) OR GMV growth (digital).

• Working capital matters: inventory turns, payables.

• Split CapEx: maintenance (existing) vs growth (new stores/fulfillment).

• Normalize for one-time charges (store closures, write-downs).

Financial Services (Banks / Insurance)

• Standard DCF is wrong. Use DDM or excess return model.

• If forced: project NII, provisions, non-interest income separately.

• Discount rate = cost of equity only (debt is operational).

Healthcare / Pharma

• Separate existing portfolio from pipeline.

• Key risk: patent cliffs, FDA approval probability.

• R&D: 15–25% of revenue.

• Biotech: risk-adjust pipeline NPV by phase success probability.

Energy (Oil & Gas)

• Revenue tied to commodity prices — use strip pricing or scenarios.

• High CapEx; distinguish development vs exploration.

• Depletion accounting differs from standard D&A.

• Terminal value very sensitive to long-term price deck.

Manufacturing / Industrial

• Cyclical — use mid-cycle earnings for normalization.

• CapEx 8–15% of revenue.

• Working capital swings with cycle — use through-cycle averages.

• WACC 8–11% typical.

Consumer Goods (CPG)

• Stable, predictable — good DCF candidates.

• Distinguish organic vs M&A growth.

• Watch gross margin trends, A&P spend, input costs.

• Terminal growth 2–3% (population + inflation).

Telecommunications

• High CapEx (15–25%) for network buildout.

• Recurring revenue, low churn — good for DCF.

• Spectrum costs lumpy.

• WACC 7–9% for large incumbents.

Real Estate / REITs

• Use NAV as primary; DCF supplementary.

• Project NOI instead of FCF.

• Cap rate replaces WACC at property level.

• Distinguish maintenance vs growth CapEx.

Media / Streaming

• Subscriber growth × ARPU drives revenue.

• Content spend dominant cost — capitalize vs expense debate matters.

• Path to profitability > current margin for growth-stage.

• High operating leverage at scale.

Common Pitfalls

• Terminal value dominance: If TV > 80% of EV, model is really a multiple-expansion bet. Disclose.

• Growth > WACC: Breaks Gordon formula. Cap g below WACC.

• Inconsistent tax rates: Historical effective rate may include one-offs. Cross-check with statutory.

• Double-counting SBC: Either subtract SBC from FCFF OR use diluted shares that price it in — not both, and not neither.

• Stale beta: yfinance beta may be 5-year or 3-year. For recent IPOs or post-restructuring businesses, compute fresh.

• Ignoring minority interest / preferred: These are claims on EV ahead of common equity. Always subtract.

• Circular WACC: WACC uses market cap → which is what we're trying to estimate. For IPOs or controversial names, iterate or use target capital structure.