Comparable Companies (Comps)
A relative valuation method that values a company using trading multiples of similar public companies, such as EV/EBITDA or P/E. Distinct from precedent transactions, which use multiples paid in past M&A deals.
Comparable Companies (Comps) · the mechanism
1 min read
Build a peer set you can defend name by name, and read the spread rather than just the median.
Where it comes up. A VP hands back your comps page with two names circled and asks why each one is in the set. You need a sentence for both, and a sentence for the obvious name you left out.
Select on business model, not on sector code
What does it sell, to whom, at what growth, with what margin structure and what capital intensity. A GICS code tells you what a company is filed as, not what it is. Six close comparables beat twenty loose ones, and you should be able to say in one sentence why each name is in and why an obvious name is out.
Put everyone on the same clock
Calendarise to a common year end, or use next-twelve-months figures throughout. Then say which convention the page uses, because a calendarised page and an uncalendarised one look identical and only one of them is right.
Compute the multiples consistently
Enterprise value on top of EBITDA, EBIT or revenue. Equity value on top of net income. Build net debt the same way for every name, and use diluted share counts throughout. One inconsistent net debt makes one company an outlier for no reason at all.
Read the spread, not the midpoint
Present the median, the mean and the range. A set trading between 7x and 9x tells you something; a set between 6x and 22x tells you the peer group is wrong or the market is discriminating on something you have not captured. Investigate the outlier before deleting it.
Worked through
Six peers at 7.2x, 8.0x, 8.4x, 9.1x, 9.6x and 21.5x forward EV/EBITDA.
- Median
- 8.75x
- Mean including the outlier
- 10.6x
- Mean excluding it
- 8.5x
The median barely moves and the mean moves by two turns, which is why the median leads a comps page. But the right move is to find out why one name trades at 21.5x before deciding anything. A depressed EBITDA from a one-off, a pending bid, or an uncalendarised fiscal year would all produce that number mechanically.
Check yourselfYour target has no profitable peers. What do you do?
Answer once you have one →
Move up the income statement to a line that exists for everybody: EV/Gross Profit or EV/Revenue, cross-checked against growth so you are not paying the same revenue multiple for a 10% grower and a 40% one. Then supplement with sector-specific measures that price the asset rather than the accounting, such as EV per subscriber or per megawatt, and say plainly that the comps are indicative.
Be able to say this back next week
- Selected on business model rather than sector code
- Put every company on one clock and said which convention the page uses
- Presented the spread and investigated the outlier rather than deleting it
Why Comparable Companies (Comps) matters in interviews
Comps are the valuation method most used in practice and most casually assembled, so interviewers probe them to see whether you can defend a peer set. The real question is never "what is a comp" — it is "why is this company comparable and that one is not".
How it works in practice
You assemble a set of publicly traded peers with similar business models, size, growth, margins and geography, calculate trading multiples for each (usually EV/EBITDA, EV/EBIT, EV/Revenue and P/E), take the median and quartiles, and apply that range to the target's own metrics.
The median matters more than the mean because a single outlier peer distorts an average badly in small sets. Practitioners typically show minimum, 25th percentile, median, 75th percentile and maximum.
Multiples must be calibrated on a like-for-like basis: calendarised to the same fiscal year end, and using forward figures (NTM or next-fiscal-year) rather than trailing where the sector is growing.
What candidates get wrong
- Building a peer set on industry classification alone. A high-growth software company and a mature one in the same SIC code are not comparable.
- Failing to adjust for one-off items in the EBITDA of either the peers or the target.
- Not being able to explain why comps typically produce lower values than precedent transactions — the answer is the control premium embedded in acquisition prices.
Comparable Companies (Comps): frequently asked questions
Why do comparable companies usually value a business lower than precedent transactions?
Because trading comps reflect the price of a minority stake in the public market, while precedent transactions reflect the price paid for control of a whole company. Acquirers pay a control premium — historically often in the 20-30% range — for the ability to direct the business and to capture synergies.
What makes a good comparable company?
Similarity in the drivers of value rather than in industry label: comparable growth rate, margin profile, capital intensity, end-market exposure, geography and scale. A defensible comps set is usually small and tightly argued rather than large and inclusive.
Practise it
A Wordle for finance. Guess a company’s EV/EBITDA band from the business quality, learn what drives it, share your score.
Open Name That Multiple — free, 3 minGo deeper
This term comes up constantly in valuation interviews and on the desk.
DCF interview questions guideRelated Valuation terms
Go further than reading
The written material is free. These are the ways to get it applied to your own work.
CV Review by a Human
Written margin-note feedback on structure, impact bullets and ATS-readability. Reviewed by Suro, not an AI score.
$25 48h turnaround
Cover Letter Review by a Human
Line-by-line review of argument, tailoring and tone, with a rewritten opening as a worked example.
$50 48h turnaround
Inner Circle
Membership: premium tools, role intros, the private community and members-only intel.
$79 /month
Browse the full glossary — 188 finance recruiting and technical terms, in plain English.