Relative valuation is not a substitute for DCF analysis — it is a cross-check that reveals whether the market is pricing the subject company consistently with comparable businesses. McKinsey's Chapter 13 establishes that multiples are compressed DCFs: every multiple embeds assumptions about growth, ROIC, and the cost of capital. Understanding what a multiple implies — rather than mechanically applying peer medians — is the difference between rigorous and careless relative valuation.
| Company | EV ($M) | LTM Rev | LTM EBITDA | EV/LTM Rev | EV/LTM EBITDA | NTM Rev | EV/NTM Rev | NTM Growth | EBITDA % |
|---|---|---|---|---|---|---|---|---|---|
| Peer A | $4,380 | $620 | $130 | 7.1× | 33.7× | $760 | 5.8× | 23% | 21% |
| Peer B | $2,750 | $390 | $70 | 7.1× | 39.3× | $490 | 5.6× | 26% | 18% |
| Peer C | $6,420 | $840 | $210 | 7.6× | 30.6× | $990 | 6.5× | 18% | 25% |
| Peer D | $1,695 | $210 | $35 | 8.1× | 48.4× | $280 | 6.1× | 33% | 17% |
| Peer E | $8,680 | $1150 | $310 | 7.5× | 28.0× | $1320 | 6.6× | 15% | 27% |
| 25th Pct: 7.1× / 29.8× / 5.7× | Median: 7.5× / 33.7× / 6.1× | 75th Pct: 7.6× / 43.9× / 6.5× | |||||||
The peer group is the foundation of relative valuation. A poorly chosen peer group produces a meaningless range — it is garbage in, garbage out. The correct standard for peer selection is economic similarity, not industry classification. Two companies in the same SIC code can have dramatically different business models, competitive dynamics, and risk profiles. The analyst must define comparability around the factors that actually drive multiples: revenue growth, operating margins, ROIC, capital intensity, and competitive moat.
For a specialized niche business with few direct comparables, the analyst faces a tension: (a) use a tight peer group of 3–4 genuinely comparable companies with limited statistical meaning; or (b) expand to 8–12 companies with weaker comparability but better statistical robustness. McKinsey's guidance: use the tight peer group as the primary reference, use the expanded group as a secondary sanity check, and explicitly footnote which companies are included/excluded and why. Never include a peer just to move the median toward a desired conclusion — peer selection should be defensible independently of the valuation conclusion.
Trading comps reflect the public market's current pricing of comparable businesses. They represent a minority, liquid stake — no control premium, no synergies, pure market pricing. The choice of LTM (last twelve months), NTM (next twelve months), or a specific forward year matters: growth companies should be valued on forward multiples (their current earnings understate earning power), while distressed or turnaround situations may require LTM to avoid anchoring on potentially overstated recovery forecasts.
| Company | Market Cap ($M) | Net Debt ($M) | EV ($M) | LTM Revenue ($M) | LTM EBITDA ($M) | EV/LTM Rev | EV/LTM EBITDA | NTM Revenue ($M) | EV/NTM Rev | Revenue Growth (NTM) | EBITDA Margin |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Peer A | $4,200 | $180 | $4,380 | $620 | $130 | 7.1× | 33.7× | $760 | 5.8× | 23% | 21% |
| Peer B | $2,800 | ($50) | $2,750 | $390 | $70 | 7.1× | 39.3× | $490 | 5.6× | 26% | 18% |
| Peer C | $6,100 | $320 | $6,420 | $840 | $210 | 7.6× | 30.6× | $990 | 6.5× | 18% | 25% |
| Peer D | $1,600 | $95 | $1,695 | $210 | $35 | 8.1× | 48.4× | $280 | 6.1× | 33% | 17% |
| Peer E | $8,900 | ($220) | $8,680 | $1,150 | $310 | 7.5× | 28.0× | $1,320 | 6.6× | 15% | 27% |
| 25th Percentile | — | — | — | — | — | 7.1× | 29.8× | — | 5.7× | 16% | 18% |
| Median | — | — | — | — | — | 7.5× | 33.7× | — | 6.1× | 23% | 21% |
| 75th Percentile | — | — | — | — | — | 7.6× | 43.9× | — | 6.5× | 29% | 26% |
| Subject Company | — | $150M debt | — | $500M | $95M | — apply | — apply | $620M | — apply | 24% | 19% |
LTM multiples reflect what investors are paying for trailing performance. For growing businesses, this produces artificially high multiples because the market prices forward expectations, not trailing results. A company growing 25% per year trading at 8× LTM Revenue is actually trading at 6.4× NTM Revenue (8 / 1.25) — the 'premium' largely disappears when you look forward. When comparing companies at different growth rates, normalizing to a forward year (NTM or Year 2) produces more apples-to-apples comparisons. The rule: use NTM for stable or growing businesses; use LTM when forward projections are unreliable (distressed companies, cyclical trough/peak analysis where you want to normalize).
Transaction comps (precedent transactions) reflect what acquirers have historically paid to acquire control of comparable businesses. They incorporate: (1) a control premium — typically 20%–40% above the pre-announcement trading price; (2) synergy value — acquirer's estimate of cost savings and revenue uplift; (3) deal dynamics — strategic urgency, competitive bidding, and financing conditions. Because of these factors, transaction multiples are almost always higher than trading multiples — using them to value a minority-stake DCF is inappropriate without explicit adjustment.
| Dimension | Trading Comps | Transaction Comps |
|---|---|---|
| What is being priced | Minority, liquid stake in public company | 100% control of a business (or majority stake) |
| Control premium | None — market price reflects minority value | Typically 20%–40% above pre-deal market price |
| Synergies included | No — standalone business value only | Yes — reflects buyer's projected synergies |
| Market conditions | Current market sentiment and rates | Reflects conditions at deal date — may be stale |
| Relevance for DCF cross-check | Primary — directly comparable to minority value DCF | Secondary — must haircut for synergies and control premium if comparing to standalone valuation |
| Relevance for M&A pricing | Secondary — sets the floor (no one buys at minority price) | Primary — reflects what real acquirers have paid |
| Typical EV/EBITDA premium vs. trading | — | Transaction multiples typically 2×–5× higher than trading |
Transaction comps should be used with care for three reasons: (1) Staleness: a deal done in 2021 at peak valuations is not a valid comparable for a 2024 deal — interest rates, multiples, and competitive dynamics have changed. Filter for transactions within the last 3 years, with tighter preference for 12–18 months. (2) Deal structure: earnouts, rollover equity, contingent payments, and minority-vs-majority stakes create comparability issues. Always verify what was actually paid in cash vs. deferred. (3) Target quality: targets in competitive auction processes with multiple bidders transact at higher multiples than negotiated deals — include only transactions where you can verify the process.
Every multiple is a compressed DCF. This is not a theoretical curiosity — it is the key insight that allows analysts to assess whether a multiple is 'fair' or 'demanding' independent of peer comparison. The P/E ratio, derived from the Gordon Growth Model, implies specific assumptions about payout ratio, cost of equity, and growth. The EV/EBITDA multiple implies an assumed tax rate, capex and depreciation intensity, and long-run FCFF conversion rate. An analyst who understands these implied assumptions can assess comparability at the level of underlying economics, not just surface-level financial metrics.
| Multiple | DCF Equivalent | Key Implied Assumptions | Red Flags |
|---|---|---|---|
| P/E | P/E = Payout Ratio / (Ke − g) | Payout ratio, cost of equity, long-run growth rate | Negative earnings; earnings distorted by non-cash charges; leverage differences across peers |
| EV/EBITDA | EV/EBITDA ≈ 1 / (FCFF Yield × EBITDA conversion) | EBITDA-to-FCFF conversion (taxes, capex, NWC), growth, WACC | Very different capex intensities across peers; different D&A policies; lease treatment differences |
| EV/Sales | EV/Sales = EV/EBITDA × EBITDA Margin | Assumed long-run margin, capital efficiency | Most useful for pre-profit companies; assumes converging margins which may not materialize |
| P/B | P/B = (ROE − g) / (Ke − g) | ROE, growth, cost of equity | Book value distorted by intangibles write-downs, goodwill, and historical depreciation conventions |
| EV/NOPAT | EV/NOPAT = (1 − g/ROIC) / (WACC − g) | The KVD formula directly — most fundamental multiple | NOPAT requires reorganization; less available than EBITDA from reported financials |
For a peer group with median EV/EBITDA = 14×, you can back-solve for the implied growth rate using an assumed FCFF conversion (EBITDA → FCFF ≈ 50% after taxes and capex) and WACC = 10%: EV/EBITDA = (FCFF/EBITDA) / (WACC − g) → 14 = 0.50 / (10% − g) → (10% − g) = 0.50/14 = 3.6% → g = 6.4%. The median peer multiple implies 6.4% long-run FCFF growth. If your subject company is growing at 12%, it deserves a premium to the group. If it's growing at 3%, it deserves a discount. Using this framework, you can compute whether applying the median multiple over- or under-values the subject company relative to what its growth profile actually justifies.
Key Takeaways
A peer group has median EV/EBITDA = 12×. Assumed FCFF/EBITDA conversion = 45% (after taxes and net capex). WACC = 10%. What long-run growth rate does the peer group multiple imply? Is this reasonable for a mature consumer staples industry?