Comparable company analysis is a valuation method that estimates what a business is worth by measuring what the market pays for similar businesses, expressed as a multiple of earnings or revenue. It is the core of the market approach, and in the lower middle market it is usually the first number an owner hears and the last one a buyer defends. The arithmetic takes ten minutes; deciding which companies belong in the comparison takes far longer. Owners who want a market-based read before anyone builds a comp set can start with our free business valuation calculator.
What Comparable Company Analysis Measures
The market approach in plain terms
Valuation recognizes three approaches: income, which discounts future cash flow; asset, which totals what the company owns; and market, which looks at what similar businesses sold for. Comps analysis is the market approach. It assumes buyers have already priced businesses like yours and that the relationship transfers with adjustments, which holds where comparables are close and recent and weakens fast as the set loosens.
Enterprise value versus equity value
Nearly every M&A multiple is built on enterprise value, not equity value. Enterprise value is the operating business independent of how it is financed; equity value is what the owner receives after funded debt is retired and excess cash swept. A company selling at 7.0x on $4 million of EBITDA carries a $28 million enterprise value, but with $5 million of term debt the equity proceeds are nearer $23 million.
Building and Screening the Comp Set
Screening criteria that actually matter
A comp set is built by starting broad and cutting hard. The screens that carry the most weight are rarely the industry code.
- Business model, not industry label. A recurring-contract HVAC service company and a new-construction HVAC installer share a code and little else. Screen on how revenue is earned and renewed before you screen on sector.
- Size band. Multiples scale with size, so a set mixing $2 million and $40 million of EBITDA yields an average describing neither. Keep comparables within roughly half to twice the subject’s earnings.
- Recency. Credit conditions move multiples as much as company quality does. Transactions from a different rate environment need exclusion or an explicit haircut, and most practitioners work from a two- to three-year window.
- Buyer type. A strategic acquirer absorbing a competitor and a sponsor buying a first platform price differently. Tagging each transaction by buyer type often explains more of the spread than industry does.
- Customer concentration and management depth. These move lower-middle-market pricing more than almost anything else. A comparable with a diversified customer base is not comparable to one where one account is 40% of revenue.
How many comparables are enough
No threshold makes a comp set valid, but six to twelve close transactions is a practical range. Below five, one outlier dominates the median. Above fifteen, the screens are loose enough to be averaging businesses that price differently. Quality of fit beats sample size, and a defensible analysis names why each deal was kept or dropped.
Normalizing the financials first
A multiple is only as good as the denominator beneath it. Private financials are kept for tax purposes, not sale purposes. Typical normalizations include owner compensation restated to a market salary, discretionary personal expenses, one-time legal costs, related-party rent restated to market, and non-recurring items.
Every add-back will be tested in a quality of earnings review, and an aggressive schedule that collapses in diligence costs more in credibility than it captures.
The Multiples and When Each One Applies
EV/EBITDA
EV/EBITDA is the default above roughly $1 million of adjusted earnings, because EBITDA is capital-structure neutral and approximates cash generated before financing and tax choices distort it. It is the language buyers, lenders, and industry surveys speak. Its weakness is ignoring capital intensity: two businesses at 7.0x differ if one needs $300,000 of annual maintenance capex and the other $30,000.
EV/Revenue
EV/Revenue is used when earnings are negative, unrepresentative, or too volatile to anchor a multiple, and as a cross-check on an EBITDA conclusion. It only means something within a narrow margin band, since a revenue multiple is an EBITDA multiple times margin. If the set runs a 13% margin and yours runs 15%, its median revenue multiple understates value.
EV/SDE and the main street boundary
Seller’s discretionary earnings adds one full owner’s compensation and benefits back to EBITDA. It suits owner-operated businesses where the buyer steps into the job, and it dominates transaction data below roughly $1 million of earnings. SDE multiples are structurally lower on the same company because the denominator is larger, so mixing the two produces a wrong number.
Trading Comps, Precedent Transactions, and Control
What trading comps tell you
Trading comps, or public comps, derive multiples from the current market capitalization and debt of listed companies. Their advantage is data quality: audited financials, live prices, daily refresh. Their disadvantage for a private seller is that listed companies are larger, more diversified, professionally managed, and freely tradeable, and a share price reflects a minority stake, not control.
Why transaction multiples sit higher
Precedent transaction analysis uses completed acquisitions rather than trading prices, and its multiples generally exceed trading comps in the same industry. The reason is the control premium. An acquirer buying the whole company gains the right to set strategy, replace management, redirect cash flow, and capture synergies, and shareholders do not surrender that at the undisturbed price. Published research on public takeovers has long placed typical premiums in a broad band of roughly 25% to 50%.
Size and marketability discounts on public comps
Where public comparables are the only reference, the derived multiple must be discounted first. Two adjustments do most of the work. A size adjustment reflects long-run cost of capital research showing smaller firms carry higher required returns and lower multiples. A marketability discount reflects that a private interest cannot be sold in three days at a quoted price; appraisal literature commonly supports 20% to 35% for controlling interests.
Together these often cut a public trading multiple by a quarter to a third, with the control premium partially offsetting. That is why a carefully built private transaction set beats a discounted public one.
A Worked Comp Set, Start to Finish
The subject company and its normalized earnings
Assume a specialty industrial services business with $28.0 million of trailing revenue and $3.60 million of reported EBITDA. Normalization adds $350,000 of above-market owner compensation, $150,000 of non-recurring litigation expense, and $100,000 of personal vehicle and travel costs, giving adjusted EBITDA of $4.20 million, a 15.0% margin. It carries $3.40 million of funded debt and $600,000 of excess cash.
The transaction set and the derived multiples
Six closed transactions survive screening, shown in millions of dollars.
- Deal A: revenue $20.0, EBITDA $3.00, enterprise value $21.0, so 7.00x EBITDA, 1.05x revenue, 15.0% margin.
- Deal B: revenue $40.0, EBITDA $5.00, enterprise value $32.5, so 6.50x EBITDA, 0.81x revenue, 12.5% margin.
- Deal C: revenue $16.0, EBITDA $2.20, enterprise value $17.6, so 8.00x EBITDA, 1.10x revenue, 13.8% margin; a strategic buyer paid up for the customer list.
- Deal D: revenue $36.0, EBITDA $4.50, enterprise value $27.0, so 6.00x EBITDA, 0.75x revenue, 12.5% margin.
- Deal E: revenue $36.0, EBITDA $5.40, enterprise value $43.2, so 8.00x EBITDA, 1.20x revenue, 15.0% margin, with most revenue under multi-year contract.
- Deal F: revenue $30.0, EBITDA $3.60, enterprise value $23.4, so 6.50x EBITDA, 0.78x revenue, 12.0% margin.
The six EV/EBITDA multiples are 6.00x, 6.50x, 6.50x, 7.00x, 8.00x and 8.00x. They sum to 42.00x, a mean of 7.00x and a median of 6.75x, with the middle four between 6.50x and 8.00x. The revenue multiples average 0.95x against a 0.93x median, and the median comp-set margin is 13.1%.
Selecting a range and bridging to equity
The subject has neither the contracted revenue of Deal E nor the strategic customer list of Deal C, so the two 8.00x observations are ceilings, not midpoints, while its margin beats Deal D. A range of 6.50x to 7.50x is defensible. On $4.20 million of adjusted EBITDA that gives $27.3 million to $31.5 million, midpoint $29.4 million.
The revenue multiple confirms it. Because the subject’s 15.0% margin sits above the comp-set median of 13.1%, the median 0.93x is the wrong reference and the upper-end 1.05x to 1.10x observations are. On $28.0 million of revenue that gives $29.4 million to $30.8 million, inside the EBITDA range. A public cross-check agrees: a peer group at 9.50x, discounted 30% for size and marketability, implies 6.65x, or $27.9 million.
Then bridge to what the owner receives. From the $29.4 million midpoint, less $3.40 million of funded debt and plus $600,000 of excess cash, indicated equity value is $26.6 million before fees and escrow.
Where the Method Runs Out of Road
Private transaction data is thin
Public deals are documented in filings and press releases. Private deals are not. The transaction databases practitioners rely on are compiled largely from voluntary submissions by brokers, intermediaries, and appraisers. That leaves real gaps: submissions skew toward completed processes, reporting lags run six to twelve months, coverage is uneven by industry and region, and some sets mix decades-old deals with current ones.
Broader surveys aggregating sponsor-backed lower-middle-market deals help fill the gap, and their published data consistently shows multiples stepping up with size, with the $10 million to $25 million enterprise value band clearing well below deals above $50 million. Useful anchors, but averages.
A multiple hides the deal structure
A reported multiple rarely tells you how much of the price was cash at closing. Earnouts, seller notes, rollover equity, escrow holdbacks, and working capital true-ups all sit inside the headline number. Two deals at 7.0x differ sharply if one is 90% cash and the other 55% cash with three years of contingency. Comps price the business, not the terms.
Maintaining the Comparable Companies Set Over Time
A comparable company analysis is a snapshot, and snapshots age. The comparable companies that supported a conclusion twelve months ago may since have been acquired, recapitalized, or grown out of the size band that made them relevant. Treating the peer group as a living file rather than a one-time build is what keeps the work usable on the day an owner decides to sell.
How often to refresh the peer group
Refresh the peer group at least twice a year, and again immediately before any live process. Credit conditions and buyer appetite move faster than most owners expect, and a set assembled in a different financing environment describes a different group of buyers than the ones now at the table.
Between refreshes the maintenance work is small. Confirm each comparable is still independent, note any follow-on deal that repriced it, and flag any company whose operating metrics have drifted far enough that it no longer belongs. Fifteen minutes a quarter keeps a set from quietly going stale.
When to drop and replace comparable companies
Drop a comparable when the reason it was included no longer holds. A company that has doubled in size, sold its recurring service line, or been rolled into a larger platform is not the business you originally screened. Replacing it matters more than simply shrinking the set, because a five-name group behaves very differently from a ten-name one.
Keep the discards in the file. An analyst who can say why a name left the peer universe answers the first question a buyer’s diligence team asks, and the written record stops a weak comparable from being quietly reintroduced a year later.
Documenting the valuation methodology
Write down the valuation methodology as you go: screening criteria, inclusion dates, sources, and every one of the adjustments applied to the subject company’s earnings. A comparable company analysis that cannot be reconstructed six months later is an opinion rather than evidence, and in a negotiation that difference decides who gets believed.
The record should also name the valuation method used and why. Comparable company analysis is a relative valuation method, pricing a business against other businesses rather than against its own projected cash flow, and saying so plainly heads off the objection that the market approach was selected because it produced the friendliest answer.
Financial Modeling Behind a Comparable Company Analysis
Most of a comparable company analysis, or CCA, lives in a spreadsheet, and the financial modeling involved is deliberately plain. There is no discounting and no forecast. What the model has to do is hold the inputs for every comparable in one place, on consistent definitions, so the derived valuation multiples are genuinely comparable to one another.
What goes in the company analysis table
The table is one row per comparable and a fixed set of columns: name, date, revenue, reported and adjusted operating earnings, margin, enterprise value, and the derived multiples. Add one column for buyer type and one for a short note on why that deal priced where it did. Those two columns usually explain more of the spread than any of the numbers do.
Below the rows, a summary block carries the mean, median, and quartiles for each multiple, plus the count of names included. Keep the subject company on its own line directly beneath the set, so its operating metrics sit immediately under the peers being used to price it.
Sanity checks an analyst runs on the company analysis
Three checks catch nearly every error. Recompute one multiple by hand and confirm it ties to the sheet. Confirm every enterprise value in the set is built on the same definition of funded debt and excess cash. Confirm the earnings figure in each row is adjusted on the same basis as the subject, since a single unadjusted comparable can move a median by half a turn.
A fourth check is worth the minute it takes. Sort the set by margin and confirm the multiples broadly rise with it. Where they do not, either one comparable is mispriced or something other than profitability is driving the group, and either answer tells you something about the company you are pricing.
Building a peer universe of similar companies
Start wider than the final set. Pull thirty to fifty names from deal databases, sector reports, and trade association lists, then cut on the screens. An analyst who begins narrow tends to confirm a number already in mind; one who begins wide finds similar companies nobody had considered, including a few buyers who later appear in the process.
Where private coverage in a niche is thin, widen the peer universe to similar public companies and discount the result rather than abandoning the exercise. Practitioners who analyze comparable company data across several sources, including deal databases, sponsor surveys, and filings, usually find the same story told three ways, and a conclusion that survives all three is far easier to defend.
Equity Multiples, Operating Multiples, and Checking the Result
Not every multiple in a company valuation measures the same thing. Some are built on enterprise value and describe the operating business; others are built on share price and describe the residual claim left after debt is repaid. Mixing the two is the most common error in an otherwise careful analysis.
How equity multiples differ from operating multiples
Equity multiples divide the value of the shares by a figure that is already after interest and tax. Operating multiples divide enterprise value by a figure measured before financing. The first says what the residual claim is worth to a shareholder; the second says what the business is worth to an acquirer, independent of how the seller chose to fund it.
In a private sale, operating multiples do nearly all the work, because the buyer refinances the balance sheet at closing anyway. Equity multiples matter at two points: when comparable public listings are the only reference available, and at the end, when an agreed enterprise value is bridged to what the seller actually banks.
Checking a derived multiple against the company’s own history
Test the range against the company’s own record before presenting it. If the business raised outside investment, bought out a partner, or turned down an unsolicited offer in recent years, each of those events implies a view of the company’s worth at a point in time.
A conclusion well above or below those reference points is not wrong, but it needs an explanation grounded in what changed. Where the derived range implies a value several turns above where the company was priced three years ago, the analysis should be able to point at the earnings growth, the customer diversification, or the sector rerating that accounts for it.
Key financial ratios that place a company in the range
The key financial ratios that move a multiple within a range are few. Gross and operating margin show whether pricing power is real. Revenue retention and customer concentration show how durable the earnings are. Working capital intensity and maintenance capital spending show how much of the reported profit a buyer actually keeps.
Score the subject against the set on each of those ratios rather than arguing the multiple directly. A buyer who has accepted that a target company retains more revenue and converts more profit to cash than the median comparable has already accepted the premium. Arguing the premium first, and the evidence afterward, rarely works.
Frequently Asked Questions
What is the difference between comparable company analysis and precedent transaction analysis?
The difference between comparable company analysis and precedent transaction analysis is the source of the multiple. Trading comps use current prices of listed peers, reflecting minority stakes. Precedent transactions use completed acquisitions, which convey control and therefore price higher.
Both are built the same way and often presented side by side. For a private lower-middle-market company, precedent transactions are the more relevant reference, since the business is being sold outright.
How many comparable transactions do you need for a credible analysis?
You need enough comparable transactions that no single deal controls the result, which in practice means roughly six to twelve close matches. Fewer than five leaves the median hostage to one outlier; more than fifteen usually means the screens were loosened too far.
Closeness of fit beats count. Four genuinely similar transactions in the same size band and the same two years support a better conclusion than twenty loose ones.
Why can’t I just use the EBITDA multiple from an industry report?
You can’t rely on a published industry multiple because it averages businesses of different sizes, margins, growth rates, and customer concentrations. It shows where the middle of a distribution sits. It does not show where in that distribution your company falls.
Benchmarks are useful for orientation and for testing whether a conclusion is plausible. They are also often quoted without specifying whether they are EBITDA or SDE based, a difference worth turns.
Does a comparable company analysis replace a formal business valuation?
No, a comparable company analysis does not replace a formal valuation. It is one of three approaches a credentialed appraiser weighs, and it yields a market indication of value rather than a certified opinion suitable for tax, estate, litigation, or ESOP purposes.
For a sale process the comps-based range is the practical tool, since it reflects what buyers are paying now. Owners with estate, gifting, or shareholder-dispute needs should also obtain a formal appraisal.
How old can a precedent transaction be before it stops being useful?
A precedent transaction generally stops being useful beyond about three years, and past two years it needs a stated reason for inclusion. Multiples move with credit availability and buyer appetite, so a deal closed under different financing conditions prices a different market.
If a thin sector forces you further back, say so and adjust the multiple rather than presenting it as current evidence. Silently including stale transactions is how comp sets drift.
Working With an M&A Advisor on Comps
What an advisor adds beyond the spreadsheet
The arithmetic is not difficult. The judgment is: which transactions belong, which add-backs survive diligence, whether a comparable was a strategic outlier, and how customer concentration and management depth move the multiple within the range. An advisor working a sector daily carries context no database captures, including deals that did not close.
Using the range to run a process
A comp set produces a range, and its purpose is to inform a process rather than name a number. A competitive process with the right strategic and financial buyers determines where inside that range, or above it, a business clears. The analysis tells you what is reasonable to expect and how to evaluate offers.
If a sale is one to three years out, the most productive use of a comps analysis today is diagnostic: it shows which attributes separate comparables at the top of the range from those at the bottom, and most can be improved before a process begins.
A comp set prices the business against the market, but it only works once the earnings figure underneath it is right. Where a multiple lands on a headline price rather than what the owner banks, the bridge from enterprise value to equity proceeds is the piece to work through, and on smaller operations the earnings base itself is usually seller’s discretionary earnings rather than EBITDA.
