Trading Comps: Public Market Multiples
Build defensible peer sets, calculate EV and multiples, apply medians (not skewed means), and challenge comps in a pitch — with practical shortcuts and honest limits.
What are trading comps?
Comparable company analysis values a business relative to similar public companies. The idea is simple: if peers trade at a cluster of multiples and your target looks like those peers, you can apply that cluster to estimate value. Comps are fast, market-based, and universally used in pitches — and they are also noisy, selection-sensitive, and easy to abuse. Treat them as a sanity check and a range, not a truth.
How to calculate a multiple
Most trading comps use enterprise value (EV) in the numerator so you can compare companies with different leverage. EV is not mysterious: Market cap + Debt − Cash = Enterprise Value. Debt includes interest-bearing obligations you would inherit in a deal; cash reduces what you effectively pay because it is available to the buyer after the transaction.
The workhorse multiple is EV / EBITDA. Divide EV by trailing or forward EBITDA (be consistent across peers). That ratio tells you how many years of operating earnings the market is pricing in — before interest, taxes, and non-cash D&A. For context on the denominator, see EBITDA explained.
Worked example (Company X): Market cap $1.0B, debt $200M, cash $50M → EV = $1.0B + $0.2B − $0.05B = $1.15B. EBITDA = $100M → EV/EBITDA = $1.15B / $0.1B = 11.5x. If your target has similar risk and growth to Company X, you might apply ~11.5x to its EBITDA to ballpark what the market would pay — then triangulate with other peers and methods. The multiple is a shortcut; the thinking is in whether 11.5x is actually comparable.
Why this matters: once you trust the peer multiple, you multiply it by your subject's EBITDA to get implied enterprise value — the bridge from "what the market pays for earnings like this" to "what this company might be worth."
Key multiples to know
Pick one primary metric for the story (often EV/EBITDA for mature businesses, EV/Revenue for high growth), then show others as cross-checks. Ranges below are rough historical norms for discussion — always verify against live data.
| Multiple | Best for | Watch out for / limitations | Common EV/EBITDA range by sector (illustrative) |
|---|---|---|---|
| EV/EBITDA | Most industries; capital structure–neutral view of operating earnings. | Heavy adjustments, different accounting for stock comp, one-time items distorting EBITDA. | SaaS: 8–15x; Utilities: 10–14x; varies widely with rates and cycle. |
| P/E | Mature, consistently profitable businesses; quick equity lens. | Leverage, buybacks, and tax distort EPS; less useful vs. EV multiples for M&A framing. | Banks often quoted on P/E or P/TBV rather than EV/EBITDA — sector-specific. |
| EV/Revenue | High-growth or pre-profit companies where EBITDA is not yet meaningful. | Ignores margin quality; two firms at the same multiple can have very different economics. | SaaS growth names: wide bands; use with gross-margin context. |
| EV/FCF | Cash-heavy, mature businesses where owners care about distributable cash. | Lumpy capex, working-cap swings, and one-time cash items make FCF volatile year to year. | Capital-intensive sectors: pair with cycle and maintenance capex discussion. |
Peer selection deep dive
Good comps start with a tight screen: same industry or NAICS/SIC bucket, similar revenue scale (often 0.25x–4x the target), comparable growth and margin profile, and liquid, investable names. If one peer is a pure-play and another is a conglomerate, you are mixing stories — narrow the set or segment the conglomerate if you can.
Red flags — usually exclude or footnote:
- In M&A: Stock trades on deal premium and synergies; multiples are not a clean read of standalone trading.
- Micro-cap outlier: Thin liquidity and jumpy prices; one print can distort the median.
- Different business model: Vertical integration vs. asset-light software, or multi-line conglomerates vs. your pure-play — not "wrong," but not interchangeable without adjustment.
- Distressed peer: Depressed multiple may reflect solvency or covenant risk, not "cheapness." Include only if you explain why the distress is relevant or temporary.
Briefed's AI Suggest can build a first-pass peer set in seconds and surface why each name is in the ballpark — sector, size, and financial similarity. You should still apply the rules above: automation saves time; judgment prevents embarrassing peer lists in front of a client.
The median vs. the mean
Outliers dominate the mean. In comps, one bid-up name or a broken story can drag the average to a number nobody would actually use.
Example: Five peers trade at EV/EBITDA of 9x, 10x, 10x, 11x, 30x (the last is a rumored strategic target trading on premium). Mean = (9+10+10+11+30)/5 = 14x — misleading if the 30x is not comparable. Median = 10x, which better reflects the "typical" peer. In live books you also show quartiles or a trimmed range.
When to exclude vs. keep: Exclude clear non-comparable events (deal rumors you do not believe are sustainable, one-time accounting). Keep borderline names if you disclose the issue and show sensitivity with and without them. Never silently drop peers just to tighten the range — that is how credibility dies in Q&A.
Challenging comps in a pitch
Expect: "Why is this company a peer?" A good answer is factual and comparative — not vibes. Anchor on metrics: "Similar revenue growth (30% vs. 28%), similar operating margins (22% vs. 24%), and similar customer concentration in enterprise," plus one business-model point if needed.
Practical red lines: If revenue growth differs by more than ~10 percentage points or operating margins by more than ~5 points without a clear narrative (mix shift, one-time cost, geography), your peer set looks weak unless you normalize or narrow the cohort. Skeptics will notice before you finish the slide.
Sector-specific multiple ranges (EV/EBITDA, illustrative)
Different sectors embed different growth, margin, and capex expectations — so "fair" multiples differ. Use these as orientation, not quotes.
| Sector | Typical EV/EBITDA band (broad) | Rationale (high level) |
|---|---|---|
| Software / SaaS | 8–18x | High gross margins and recurring revenue command premium multiples; range widens with growth and profitability. |
| Banks | 8–12x (often P/E or tangible book context) | Regulated, balance-sheet-driven; EV/EBITDA is used less uniformly — pair with bank-specific metrics. |
| Utilities | 10–14x | Stable cash flows, rate-regulated returns; lower growth, lower risk premium in calm environments. |
| Healthcare (services / tools) | 12–20x | Mix of growth and defensibility; payer and regulatory risk moves the band. |
| Manufacturing | 6–10x | More cyclicality and capex; margins mean-revert with volume and input costs. |
Where comps sit in the valuation stack
Comps rarely stand alone. They are one band on a football field, alongside DCF, precedents, and market trading ranges. When the DCF and comps disagree, your job is to explain why (growth, margin, risk, cycle) — not to pick the midpoint that flatters management.
Worked example: Shopify and a peer set
Illustration only — round numbers for teaching; pull live prices, shares, debt, cash, and EBITDA before any real pitch.
Suppose you comp Shopify (SHOP) against BigCommerce (BIGC), Wix (WIX), Fastly (FSLY), Twilio (TWLO), and Cloudflare (NET). You pull EV/Revenue and EV/EBITDA on the same basis (e.g., NTM consensus, or LTM — do not mix).
| Peer | EV / Revenue (illustr.) | EV / EBITDA (illustr.) |
|---|---|---|
| BIGC | 1.8x | 14x |
| WIX | 2.2x | 13x |
| FSLY | 2.5x | 22x |
| TWLO | 2.0x | 16x |
| NET | 10.0x | 48x |
Medians: Sort EV/Revenue: 1.8, 2.0, 2.2, 2.5, 10.0 → median 2.2x. Sort EV/EBITDA: 13, 14, 16, 22, 48 → median 16x. NET is an outlier on both; in a live book you might show "full set" vs. "adjusted set" excluding NET if you argue it is not a commerce comp — that debate is exactly what MDs want to see rehearsed.
Apply the median EV/EBITDA to Shopify: assume illustrative LTM EBITDA of $1.0B. Implied EV ≈ 16 × $1.0B = $16B. If the market prices Shopify at $18B EV at your cut date, the peer median implies you are trading at a ~12% premium to the median multiple — either fair if Shopify deserves a quality premium, or a prompt to revisit adjustments, cycle positioning, and whether NET belongs in the set.
Always reconcile: different peer sets and different EBITDA adjustments move the answer more than a slick chart. That skepticism is the point.
Briefed: faster comps, same judgment
Briefed's AI Suggest builds a peer set in about ten seconds and shows why each name is included — but you should own the logic. Pro users can adjust peers manually, benchmark multiples against historical five-year ranges where available, and export the comp table to PowerPoint for decks that still need to survive a live challenge.