valuation
Comparable companies are the peer set whose trading prices supply market evidence for relative valuation. A good set shares economics—not merely a GICS label—and produces multiples you can defend in a committee.
Comparable companies are the listed (sometimes large private) businesses whose observed prices are allowed to speak for the target. IVS's market approach lives or dies on this set. A GICS or NACE code is a search filter, not a conclusion. The question is whether a buyer of the target would, in the next hour, also consider those names as substitutes for capital.
Useful similarity tests: customers and pricing power, gross and EBITDA margin structure, growth and reinvestment, cyclicality, geography of demand, and capital intensity. Size and listing venue matter for liquidity, not as snobbery: a $40bn US compounder is often a bad multiple donor for a €25m Iberian manufacturer.
Once the set exists, you take trading multiples—typically EV/EBITDA and EV/EBIT, plus equity multiples where relevant—then apply a reasoned statistic (median, upper quartile if the target is genuinely higher quality) to the target's driver. Comparable company analysis is the process of doing that work; this page is about the objects in the set.
Precedent transactions are a different evidence class: control, synergies, stale dates. Do not average them into the same median as live trading comps.
Target: invented Harbourline Components (€42m revenue, industrial distribution, DACH+Iberia). Illustration only.
| Candidate | Keep? | Why |
|---|---|---|
| Listed EU industrial distributor, 12% EBITDA margin, 5% growth | Yes | Closest economics |
| Global mega-cap conglomerate with a small distribution division | No | Conglomerate multiple, different cost of capital |
| US specialty chemical producer | No | Different cycle, different gross margin |
| Thinly traded local peer, 2,000 shares a day | Use with caution | Print is not a market |
| High-growth automation software | No | Wrong reinvestment and risk |
Resulting primary set: five names, median EV/EBITDA 8.2x, interquartile 7.4x–9.1x. That range, not a single peer's 11x, is the evidence.
Buyers attack the peer set because it is easier than attacking your DCF. A defensible set shortens debate. A decorative set of famous logos invites a rebuild of enterprise value in the buyer's model using their comps. Sell-side advisors who cannot explain exclusions lose the room.
Public peers can be scarce in niches. Then you widen geography, use sum-of-the-parts on divisions, or lean harder on precedents and DCF—and you disclose the scarcity. Markets also misprice whole sectors together; comps will not save you from a bubble or a panic, they will embed it.
Dazur builds comparable-company sets that a sceptical principal can audit in fifteen minutes: inclusion rules, exclusions with reasons, and a median that is not hostage to one outlier. False peers are how mid-market owners get talked into Nasdaq software multiples. We would rather have five honest names than fifteen logos.