InvestorsFree tool
Comparable Company and Multiple Analysis
A comps table from your whole screen: every multiple on one definition, the range with and without your exclusions, and which ones a filing can settle.
Free to start · Opens in your River workspace, ready to edit and share
How to do a comparable company analysis
A comparable company analysis, often called comps, values a business by how the market prices similar listed companies. You choose peers, express each one's price as a multiple of EBITDA, revenue or earnings, and apply the range to the company you're valuing. A good analysis uses peers with similar growth and risk and builds every multiple the same way. It reports the median and quartiles, with a reason for every company left out.
The hardest part is choosing the peers. Aswath Damodaran, who teaches valuation at NYU Stern, warns that a biased analyst who picks both the multiple and the comparable firms can justify almost any value1. So record every company you dropped and why, not just the ones you kept.
Choose comparable companies you can defend
Good peers share your target's risk and growth prospects, not just its industry label. The IPEV Valuation Guidelines set out best practice for private equity and venture valuations. They say that's more likely when companies are similar in business activities, markets served, size, geography, and applicable tax rate2.
Screen wide on those traits, using industry codes, size and region, and keep the full list the screen returns. Then cut it down, giving every exclusion a reason a reviewer can check against a document.
- Negative EBITDA: the filings show it, and a multiple of a negative number means nothing.
- Mostly services revenue: the revenue breakdown in the annual report settles it.
- Too large or too small: a size rule you wrote down before you saw the multiples.
- "Not comparable": settles nothing. Name the test it failed, or put the company back.
Pick a multiple that fits the business
Match the top of the ratio to the bottom. Enterprise value belongs to lenders and shareholders together, so divide it by a measure that comes before interest, such as EBITDA or revenue. A share price belongs to shareholders alone, so divide it by earnings per share. Damodaran's consistency test: if the numerator is an equity value, the denominator should be one too1.
Three common multiples
| Multiple | Fits when | Watch for |
|---|---|---|
| EV to EBITDA | Peers are profitable but carry different amounts of debt | Add-backs that differ from company to company |
| EV to revenue | The target or its peers don't yet earn steady profits | Profit margins that vary widely across the set |
| Price to earnings | Peers have similar debt and tax rates | One-off gains and charges inside net income |
Next, fix the period: the last 12 months for every company and the target, or next year's forecast for all of them, never a mix. For a startup, test the revenue itself before you multiply it: a pitch deck analysis checks whether a deck's figures hold together.
Build every multiple the same way
Small differences add up across a peer set. Follow these steps for every company, and note where each figure came from.
Fix one price date
Take every share price from the same day, and write the date at the top of the table.
Count diluted shares
Multiply the price by shares outstanding plus the dilution from in-the-money options and convertibles.
Add every other claim
Add debt, preferred stock and noncontrolling interests, then subtract cash. Consolidated EBITDA counts all of a majority-owned subsidiary3, so its outside owners' share belongs in enterprise value too.
Line up the 12 months
Where fiscal years end in different months, rebuild the last 12 months: the latest annual results, plus this year to date, minus the same period last year.
Treat leases alike
IFRS 16 takes most lease costs out of EBITDA, while US GAAP leaves operating lease costs in, as the IASB's own analysis4 explains. Restate one side, and count lease liabilities as debt only where EBITDA excludes their cost.
Put adjusted EBITDA on one definition
Adjusted EBITDA has no standard definition, so each company writes its own. In the US, a public company that publishes one must also show the closest GAAP measure and a quantitative reconciliation of the two5. For EBITDA, that's normally a table that starts at net income and lists each add-back.
Line the reconciliations up side by side and keep the add-backs the set has in common. Apply that list to every peer and to the target. Otherwise you multiply one definition of EBITDA by a multiple built on another. Stock-based compensation is a frequent split: some companies add it back and some don't.
Question add-backs that keep coming back. The SEC staff says a measure that excludes normal, recurring, cash operating expenses6 can be misleading, and it counts an expense that returns even at irregular intervals as recurring. Treat restructuring charges booked year after year as a cost of doing business.
Show both medians
Put the median on each company's own definition beside the median on the shared one. A reviewer will ask about the gap.
Read the range, not one number
Report the median rather than the average. Most multiples can't go below zero but have no ceiling, so averages run high, and Damodaran calls the median much more representative of the typical firm1. Show the 25th and 75th percentiles and the sorted list too, so a reader can see when one company moves the answer.
Now price your own choices. Take the median of every screened company with a multiple, then of the set you kept, and multiply the gap by the target's EBITDA. Loss-makers fall out of both, because their multiples mean nothing, and Damodaran notes that this biases the group's average upward1.
Target: a valve maker with $12.0 million of adjusted EBITDA.
Screen: 12 listed companies. Two have negative EBITDA, so no multiple.
Kept, six companies: 9.8x, 10.4x, 10.8x, 11.2x, 12.0x and 13.5x. Median 11.0x.
Dropped, four with a multiple: 6.9x and 7.6x (mostly services), 8.4x (revenue over 20 times the target's), 7.2x ("not comparable").
All 10 with a multiple: median 10.1x.
- Price the selection. 11.0x on $12.0 million is $132 million. At 10.1x it's about $121 million, so your choices are worth about $11 million.
- Test each reason. Losses, services revenue and size each point to a figure in a filing. "Not comparable" points to nothing.
- Reverse the weakest. Put the "not comparable" company back and the median drops to 10.8x, or about $130 million.
Apply it to the company you're valuing
Place the target within the range, not automatically at the median. Compare it with the peers on growth and risk, which the IPEV guidelines call the two key variables2 behind an earnings multiple. Write down why it sits where it does.
A private company's shares are harder to sell, and the IPEV guidelines ask valuers to weigh that lack of liquidity. Funds also calibrate to the price they paid. In the guidelines' example, buying at 8x EBITDA while peers trade at 10x sets a 20% discount, reviewed at every later valuation. Keep the same peers each time unless new market information becomes available2.
Enterprise value isn't what shareholders receive. To reach equity value, subtract debt and other claims and add cash. Then a cap table and exit waterfall shows how it splits between preferred and common shares.
Before you share the analysis
- Every exclusion has a reason that points to a document
- All multiples use one price date and the same 12 months
- Enterprise value counts debt, preferred stock, noncontrolling interests and cash
- Peers and target use one set of add-backs
- The median, quartiles and sorted list are all shown
- You know which single exclusion moves the range most
When to bring in a professional
Bring in a qualified valuer when the number has legal or tax weight. For private company stock options, US tax rules presume a value is reasonable if it comes from an independent appraisal dated no more than 12 months before the grant7.
The tool at the top of this page does this first pass: it prices your exclusions and restates the add-backs, but leaves the conclusion to you. Once the set is settled, carry it into an investment committee memo.
How it works
Start with your screen: every company it returned, the ones you kept and why you dropped the rest. Add the company you're valuing if you have its numbers. River reads what you send and looks up the latest filings for any listed company with missing figures. Then it builds the comps table in a sheet in your workspace, with a short summary of what your choices did to the range.
Every figure is labeled as coming from a filing or from you, and a price River can't trace is left out of the median rather than guessed. You get the set, the range and the reasoning; the conclusion stays yours.
Send your screen
Paste or attach every company the screen returned, which ones you kept and why you dropped the others.
River checks the set
It finds missing figures in public filings, lines up each company's add-backs and builds every multiple alike.
Get your comps table
A sheet and summary in your workspace: the range with and without your exclusions, and those a reviewer may question.
Edit and defend it
Change anything, compare the set with last quarter's by name, or ask for a memo for your reviewer.
What you get
- Every company your screen returned, with its multiple, its source and why you kept or dropped it
- Each exclusion marked testable against a filing, or asserted with no criterion behind it
- Medians and quartiles on the full screen and on your set, with every price dated
- What your exclusions are worth to the target's value, in multiple points and in currency
- Each peer's adjusted EBITDA restated to the add-backs the set shares, and the target's too
- A short case for each kept company, and the exclusions a reviewer is most likely to challenge
Common questions
What is a comparable company analysis?
A comparable company analysis estimates what a company is worth from the prices investors pay for similar listed businesses. You work out each peer's multiple, such as enterprise value to EBITDA, find the median and quartiles, and apply them to the company you're valuing. It's often called comps or trading comps, and it sits alongside discounted cash flow and precedent transaction analysis.
How many comparable companies should I use?
Use as many as genuinely match the target's business, growth and risk; there is no fixed number. Test the set by dropping each company in turn: if one name moves the median a lot, the set is too thin. In a narrow sector, widen the screen and accept the tradeoff Damodaran describes: a broader industry gives you more firms but a more diverse group1.
What's the difference between trading comps and precedent transactions?
Trading comps use the current share prices of similar listed companies, while precedent transactions use prices paid to buy whole companies, which can include a premium for control. The IPEV guidelines warn that deal data is often weakened by the time since the deal, thin financial information and unreliable pricing2. They leave the choice between the two, or a mix of both, to the valuer's judgment.
Can I value a private company with comparable companies?
Yes, with adjustments. In the US, the section 409A rules for valuing private company stock list the market value of similar companies' stock7 among the factors to weigh. They also name control premiums and discounts for lack of marketability. Pick peers as you would for a listed company, then adjust for size, growth and how hard the shares are to sell.
How do I build a comparable company analysis with River?
Paste or attach your screen in the box at the top of this page: every company it returned, the ones you kept and why you dropped the rest. Add the company you're valuing and pick a multiple if you like. River builds the comps table in a sheet, with a short written summary, in your free River workspace.
Will the tool tell me what the company is worth?
No. It gives you the peer set, the range and the reasoning behind both, and leaves the conclusion to you or a qualified valuer. It doesn't invent market data either: a share price or multiple it can't trace to a filing or to you is marked unsourced and left out of the median.
Is the comparable company analysis tool free?
Yes, it is free to start. Your comparable company analysis opens in a River workspace, where you can edit it, share it or ask River to change it. Paid plans add more AI use and access to the most advanced models.
Sources
Aswath Damodaran, NYU Stern. “Fundamental Principles of Relative Valuation.” pages.stern.nyu.edu (opens in a new tab)
IPEV Board. “International Private Equity and Venture Capital Valuation Guidelines (December 2025).” privateequityvaluation.com (opens in a new tab)
Aswath Damodaran, NYU Stern. “Earnings Multiples.” pages.stern.nyu.edu (opens in a new tab)
IFRS Foundation. “IFRS 16 Leases: Effects Analysis.” ifrs.org (opens in a new tab)
eCFR. “17 CFR 244.100 — General rules regarding disclosure of non-GAAP financial measures.” ecfr.gov (opens in a new tab)
U.S. Securities and Exchange Commission. “Non-GAAP Financial Measures.” sec.gov (opens in a new tab)
eCFR. “26 CFR 1.409A-1 — Definitions and covered plans.” ecfr.gov (opens in a new tab)
Built by
Chandler Supple
Co-founder & CEO, River
Chandler has spent the last six years working in AI, building tools for millions of users and Fortune 100 companies. In 2025, he co-founded River to build a new kind of operating system for AI agents.