Five multiples. Five different perspectives on the same question: how much is this business worth relative to its financial performance? Damodaran's taxonomy in The Little Book of Valuation goes beyond simply defining each multiple — it specifies which businesses each multiple is appropriate for, what drives the multiple fundamentally, and what companion variable corrects for cross-company differences in growth and risk.
Valuation Multiples Taxonomy — Five Core Multiples
Each multiple is a compressed DCF — know its fundamental drivers before applying it
P/E
Price / Earnings
Companion Variable
PEG (P/E ÷ growth)
Mature profitable companies; stable earnings
Negative earnings; cyclicals; heavily depreciated
P/B
Price / Book Value
Companion Variable
P/B vs. ROE regression
Banks, insurance, asset-heavy businesses
Asset-light (tech, brands have P/B > 10× normally)
EV/EBITDA
Enterprise Value / EBITDA
Companion Variable
EBITDA margin + capex intensity
Cross-capital-structure comparisons; M&A; LBO
Capital-intensive businesses (adds back needed capex)
EV/Sales
Enterprise Value / Revenue
Companion Variable
Operating margin trajectory
High-growth unprofitable companies; early-stage SaaS
Meaningless without margin context; 10× on 5% margin is expensive
EV/FCF
Enterprise Value / Free Cash Flow
Companion Variable
Reinvestment rate (ROIC implied)
Capital-efficient mature compounders; Buffett-style analysis
High-growth with negative FCF; lumpy capex businesses
Industry-Specific Multiples — When Generic Multiples Fail
| Industry | Correct Multiple | Why Generic Fails |
|---|---|---|
| REITs | Price / FFO | Depreciation artificially reduces GAAP net income; RE often appreciates |
| Banks / Insurance | P/Book vs. ROE | EBITDA meaningless; interest is the product, not a financing cost |
| Oil & Gas (E&P) | EV/EBITDAX; EV/Proved Reserves | Earnings volatile with commodities; asset value per barrel matters |
| SaaS / Tech Growth | EV/ARR; EV/Revenue | Negative EBITDA/EPS undefined; recurring revenue quality is the value |
| Airlines | EV/EBITDAR (+ rent) | High lease costs; RASM (revenue per seat-mile) captures unit economics |
Damodaran's Synthesis: Multiples as Screening Tools, Not Valuation Tools
A low multiple flags a candidate for deeper analysis — it does not confirm undervaluation. Always ask: "Why is this cheap?"If the answer is a permanent impairment of the business, the discount is earned. If the answer is temporary sentiment, the discount may be an opportunity. Use the reverse DCF to see what assumptions the current multiple implies — then evaluate whether those assumptions are too pessimistic.
Figure 12.1 — Five core multiples and their companion variables. No multiple is complete without its companion; no multiple produces intrinsic value without fundamental grounding.
Every valuation multiple is a shorthand for a present value calculation. The multiple compresses a full DCF into a single ratio by making implicit assumptions about growth, risk, and payout. The multiple is not arbitrary — it has a specific mathematical relationship to the underlying fundamentals, which determines when it provides accurate information and when it misleads. Damodaran's rule: before using a multiple, know its fundamental drivers.
| Multiple | Formula | Fundamental Drivers | Best For | Fails For |
|---|---|---|---|---|
| P/E | Price / EPS | Growth (g), cost of equity (Ke), payout ratio | Mature, profitable companies; stable earnings; cross-sector comparisons with growth adjustment | Negative earnings; cyclicals at cycle bottom; heavily depreciated businesses; distorted by one-time items |
| P/Book (P/B) | Price / Book Value per Share | ROE vs. cost of equity; long-run growth | Banks and financial institutions; asset-heavy businesses; distress analysis | Asset-light businesses (P/B >10× is normal for brands, software); goodwill-heavy post-acquisition companies |
| EV/EBITDA | Enterprise Value / EBITDA | Operating margin, growth, cost of capital, capex intensity | Cross-company comparisons across capital structures; M&A; LBO analysis | Capital-intensive businesses (adds back capex needed to sustain operations); highly leveraged companies where interest coverage matters |
| EV/Sales | Enterprise Value / Revenue | Operating margin trajectory, growth, competitive dynamics | High-growth unprofitable companies; early-stage; cross-industry comparisons where profitability differs | Meaningless without margin context — 10× EV/Sales on 40% margin is cheap; same multiple on 5% margin is expensive |
| EV/FCF | Enterprise Value / Free Cash Flow to Firm | NOPAT margin, reinvestment rate, ROIC, WACC | Capital-efficient steady-state businesses; mature compounders; Buffett-style analysis | High-growth companies with negative FCF; lumpy capex businesses (FCF distorted year-to-year) |
The most common error in comparable company analysis: concluding that Company A is cheap because it has a lower P/E than Company B, without accounting for the fact that Company A grows at half the rate and has twice the debt. Damodaran's solution: companion variables. Every multiple has a fundamental driver that must be held constant (or adjusted for) before the multiple is meaningful for comparison:
| Multiple | Primary Companion Variable | Secondary Companion Variable | The Correct Question |
|---|---|---|---|
| P/E | EPS growth rate (use PEG = P/E / growth) | Risk (beta, leverage) | 'Is this P/E justified by the company's growth rate and risk profile?' |
| P/Book | ROE (use P/B vs. ROE regression) | Cost of equity | 'Is this P/B justified by the company's return on equity vs. its cost of equity?' |
| EV/EBITDA | EBITDA margin and capex intensity | Growth rate of EBITDA | 'Is this EV/EBITDA justified given margin quality and reinvestment requirements?' |
| EV/Sales | Operating margin and its trajectory | Revenue growth rate | 'Is this EV/Sales cheap given what the eventual operating margin will be?' |
| EV/FCF | Reinvestment rate (ROIC implied by FCF vs. earnings) | FCFF growth rate | 'Does this EV/FCF reflect genuine cash generation or temporarily low capex?' |
Banks are properly valued using P/Book relative to ROE. The Gordon Growth Model gives: P/B = (ROE − g) / (Ke − g). A bank with 15% ROE, 3% growth, and 9% cost of equity should trade at P/B = (15%−3%) / (9%−3%) = 12%/6% = 2.0×. A bank with 8% ROE, 3% growth, and 9% cost of equity should trade at P/B = (8%−3%) / (9%−3%) = 5%/6% = 0.83× (below book!). The first bank is legitimately worth 2× book; the second bank is correctly valued below book because it earns below its cost of equity. Simply comparing both to a 'sector average P/B' without adjusting for ROE differences is meaningless — the difference in P/B is explained entirely by the ROE gap.
For certain industries, generic multiples like P/E and EV/EBITDA are structurally inappropriate — either because earnings and EBITDA are meaningless or because the industry's value driver is something entirely different from earnings or cash flow. Practitioners in these industries have developed sector-specific multiples that capture the actual value driver:
| Industry | Generic Multiple Problem | Industry-Specific Multiple | What It Captures |
|---|---|---|---|
| Real Estate (REITs) | Depreciation artificially reduces GAAP net income — P/E is distorted | Price/FFO (Funds From Operations = Net Income + Depreciation − Gains on Sales) | Real cash flow to REIT investors after real property depreciation is excluded |
| Banks & Insurance | EBITDA meaningless (interest income is the product, not a financing cost) | P/Book (with ROE adjustment); P/Tangible Book; P/Adjusted Earnings | Book value is the relevant capital base; regulatory capital ratios drive risk |
| Oil & Gas (Exploration) | Earnings highly volatile with commodity prices; capex-intensive | EV/EBITDAX (adds back exploration expense); EV/Proved Reserves; EV/Boe/d (barrels of oil equivalent per day) | Asset value per unit of production capacity |
| Early-Stage Technology / SaaS | Negative earnings and EBITDA make P/E and EV/EBITDA undefined or meaningless | EV/ARR (Annual Recurring Revenue); EV/Revenue; LTV/CAC (Customer lifetime value / acquisition cost) | Quality and durability of the recurring revenue stream |
| Telecom / Cable (Mature) | High depreciation and amortization distort earnings; massive infrastructure investment | EV/EBITDA adjusted for capex; EV/OpFCF (Operating free cash flow after capex) | Cash flow after sustaining the capital-intensive network infrastructure |
| Airlines | High operating leverage makes EPS extremely volatile; capital-heavy | EV/EBITDAR (adds back rent/lease expense — EBITDA + Rent); Revenue per Available Seat Mile (RASM) | Economics per unit of capacity, removing lease structure effects |
Damodaran's final message on multiples: they are screening tools, not valuation tools. A screen that identifies companies trading at low EV/EBITDA relative to peer medians does not tell you these companies are undervalued — it tells you they are candidates for deeper analysis. The analyst's job is to then ask: why is this company cheap on the multiple, and is that reason temporary or permanent?
Every low-multiple stock passes through the 'cheap for a reason' test before investment: enumerate all reasons the stock might be cheap. For each reason, assess whether it is (a) a permanent impairment of the business that the multiple correctly reflects, or (b) a temporary, fixable problem that the market is overpricing. If the legitimate discount reasons account for only part of the observed cheapness, and the remaining discount cannot be explained, you have identified a potentially exploitable mispricing. If the discount reasons fully account for the low multiple, the stock is cheap for good reason and not an opportunity.
Key Takeaways
A real estate investment trust (REIT) reports GAAP net income of $50M but Funds From Operations (FFO) of $120M. A peer REIT trades at 18× FFO. What is the peer-implied value of this REIT, and why is P/E inappropriate?