How to Build a Trading Comps (Comparable Companies) Analysis
Trading comps value a company by how the market prices similar public companies today. Here is how to pick a peer set, choose the right multiple, and apply it to get an implied value.
What is a trading comps analysis?
A trading comps analysis values a company using the multiples at which similar public companies currently trade. You select a peer set matched on industry, size, growth, and margins, calculate each peer's EV/EBITDA and P/E, take the median across the set, and apply that median to the subject company's own metric to get an implied value.
Key takeaways
- Comps are a relative valuation: they show what the market pays for similar businesses today, not what a business is intrinsically worth.
- EV/EBITDA is the workhorse multiple because both terms sit before financing effects, making it neutral to leverage.
- P/E pairs equity value with net income, so it is only comparable across peers with similar capital structures.
- Use the median, not the mean, so one outlier peer cannot distort the whole set.
- NTM multiples are generally more comparable than LTM, but they depend on the quality of analyst estimates.
- A mispriced sector produces mispriced comps, which is why comps are triangulated against a DCF.
What is a trading comps analysis?
- Trading comparable companies analysis (comps)
- A relative valuation method that prices a company by applying the valuation multiples of similar publicly traded companies to the subject company's own financial metrics.
A trading comparable companies analysis (comps, for short) values a business by reference to how the public market is currently pricing similar businesses. Where a DCF values a company on its own projected cash flows, comps value it relative to peers, converting each peer's market price into a multiple of some financial metric, then applying that multiple to the subject company. It is a relative valuation method: it tells you what the market is paying for similar businesses right now, not what a business is intrinsically worth on its own merits.
Interviewers ask about comps because building a clean set requires real judgment: which companies actually belong in the peer group, which multiple fits the situation, and how to handle a peer set that isn't perfectly comparable, which it never is.
How do you select a comp set?
A good comp set matches the subject company as closely as possible on the factors that actually drive valuation: industry and end markets, business model, size (revenue or market cap), growth rate, margin profile, and geography. Screening usually starts broad (same industry classification) and narrows from there, dropping companies that are too different in size, growth, or profitability to trade at a genuinely comparable multiple.
Which multiple should you use: EV/EBITDA, EV/Revenue, or P/E?
The multiple has to pair a numerator and denominator that reflect the same claim on the business, exactly the logic covered on the enterprise value guide. EV/EBITDA and EV/Revenue are capital-structure-neutral: since both EV and EBITDA (or revenue) sit before financing effects, these multiples let you compare companies with different amounts of debt on equal footing, which is why EV/EBITDA is the most common multiple in comps work. P/E, price divided by earnings per share, pairs equity value with net income, both already affected by capital structure, making it more useful when comparing companies with genuinely similar leverage, or in sectors like financials where an enterprise value framework doesn't apply cleanly.
P / E = Share Price ÷ Earnings per Share
Multiples are usually shown on both an LTM (last twelve months, actual reported results) and NTM (next twelve months, analyst estimates) basis. NTM multiples strip out the effect of a peer having just had an unusually strong or weak trailing year, and are generally considered more forward-looking and comparable, but rely on the accuracy of analyst forecasts.
| EV/EBITDA | P/E | |
|---|---|---|
| Numerator | Enterprise value (the whole business) | Equity value or share price |
| Denominator | EBITDA, before interest and taxes | Net income or EPS, after interest and taxes |
| Neutral to leverage? | Yes | No |
| Handles differing D&A policies? | Yes, D&A is added back | No, depreciation flows through net income |
| Best used for | Most industrials, healthcare, consumer, and TMT comparisons | Banks and insurers, and peers with similar capital structures |
How do you build the comps table and apply it to the subject company?
- For each peer, calculate enterprise value and equity value from public data (share price, shares outstanding, debt, cash).
- Divide by each peer's EBITDA, revenue, and EPS to get its multiples.
- Take the median (not the mean) across the peer set for each multiple, since the median is far less distorted by one or two outlier companies than a simple average.
- Apply the peer median multiple to the subject company's own metric to get an implied enterprise value (or equity value, for P/E), then bridge to a per-share price the same way the DCF guide does.
Worked example: three peers trade at 7.5x, 8.0x, and 9.0x EV/EBITDA (median 8.0x). The subject company has $150mm of EBITDA.
What are the limitations of trading comps?
No two companies are perfectly comparable, so a comps-based value is always an approximation shaped by whichever peers were chosen. The whole peer set can also be mispriced together, during a sector-wide bubble or downturn the comps simply repeat that mispricing rather than correcting for it, which is exactly why comps are typically triangulated against a DCF and, in an M&A context, precedent transactions (past deals for similar companies, which unlike trading comps include the control premium a buyer pays to acquire the whole company).
Try a real question
This is the same question engine and thirty-second clock a real duel uses, filtered to this guide's topic. No account, no signup, nothing saved.
Common trading comps interview questions
How do you select a comp set for a trading comps analysis?
Start with companies in the same industry and end markets, then narrow to those most similar in business model, size, growth rate, margin profile, and geography. The goal is a set of peers similar enough that the market is plausibly pricing them on comparable terms.
Why is EV/EBITDA used more often than P/E in comps work?
EV/EBITDA is capital-structure-neutral: both enterprise value and EBITDA sit before financing effects, so the multiple isn't distorted by how much debt a given peer carries. P/E pairs equity value with net income, both already affected by leverage, so comparing P/E across peers with different capital structures can be misleading.
Why use the median instead of the average (mean) across a comp set?
The median is far more resistant to being skewed by one or two outlier companies, an unusually high- or low-multiple peer, than a simple average, which gives every data point equal weight regardless of how representative it actually is.
What's the difference between LTM and NTM multiples?
LTM (last twelve months) uses each company's actual trailing reported results. NTM (next twelve months) uses analyst forecasts. NTM multiples are generally seen as more comparable across peers since they aren't skewed by one company having just had an unusually strong or weak trailing year, but they depend on the reliability of the underlying estimates.
What are the main limitations of trading comps as a valuation method?
No peer set is perfectly comparable, so the result is always shaped by which companies were chosen. The entire peer group can also be mispriced together during a sector-wide bubble or downturn, in which case comps simply repeat that mispricing. That's why comps are typically checked against a DCF rather than used alone.
How do trading comps differ from precedent transactions?
Trading comps use current public market prices for standalone, minority stakes in similar companies. Precedent transactions use the prices actually paid in past M&A deals for similar companies, which include a control premium, the extra amount a buyer pays to acquire an entire company and its decision-making control, so precedent transaction multiples tend to run higher than trading comp multiples for otherwise similar businesses.