Comparable company analysis — usually shortened to "comps" or "trading comps" — is the valuation method that asks a simple question: what is the market currently paying for businesses similar to the one you're trying to value? Instead of forecasting cash flows and discounting them like a DCF, comps let you value a company by reference to how its public peers are priced right now.
Why bankers use comps alongside a DCF
A DCF is only as good as its assumptions about growth, margins, and the discount rate — and small changes to those assumptions can swing the output wildly. Comps ground a valuation in an observable, market-based reference point. That's why, as covered in Three Valuation Methods, banks rarely rely on a single method: comps, precedent transactions, and a DCF are run side by side, and where they diverge tells you something about the market's expectations versus your own model's assumptions.
What actually goes into a comp set
A comparable company set starts with businesses that share the target's industry, business model, growth profile, margin structure, and size. From each peer you need two things: its Enterprise Value (not just market cap — EV also accounts for debt, cash, and minority interests) and a normalized measure of its earnings, most commonly EBITDA.
Why EBITDA gets "scrubbed" before you use it
Reported EBITDA in a peer's financial statements often includes one-time items — a restructuring charge, a litigation settlement, an asset write-down — that don't reflect the ongoing earnings power of the business. Before computing a multiple, analysts add these back to arrive at an Adjusted (or normalized) EBITDA. Skipping this step, or applying it inconsistently across peers, is one of the most common mistakes in building a comp set, because it makes the resulting multiples impossible to compare on a like-for-like basis.
Reading the EV/EBITDA multiple
Once you have Enterprise Value and Adjusted EBITDA for each peer, the multiple is simply EV divided by Adjusted EBITDA — see What Is a Valuation Multiple? for the basic mechanics of reading "8x EBITDA." A single multiple in isolation doesn't tell you much; what matters is the range across the peer set and whether that range makes sense given differences in growth and margin. A real peer set will show some dispersion — a company with an unusually high or low multiple relative to its peers is either genuinely mispriced, or (more often) there's a company-specific reason, such as pending deal speculation, that's distorting the number.
From multiple to valuation
Once you've settled on a reference multiple (typically the median of a scrubbed peer set, excluding outliers), you apply it to the target's own Adjusted EBITDA to get an implied Enterprise Value — the same EV concept explained in EV-to-Equity Bridge (Intro), which then needs to be bridged down to an equity value.
To see this entire process worked through with real numbers — from raw peer financials, to scrubbed EBITDA, to a defensible implied valuation — walk through Comparable Company Analysis, a full interview-style case with a step-by-step model answer.