How to select a peer group, spread the comps, calendarize mismatched fiscal years, and derive a credible implied valuation range — with real worked numbers at every step.
Trading comparable companies analysis (trading comps, or simply "comps") values a company by benchmarking it against publicly traded peers. The logic is straightforward: if five similar companies trade at a median EV/EBITDA of 14x, a target company with $800M of EBITDA is worth roughly $11.2B of enterprise value — before any M&A premium.
Comps sit alongside the DCF and precedent transactions as one of the three core valuation methodologies. Comps reflect what the market is willing to pay for a business today, in the current rate environment, with current sentiment. A DCF tells you what a business is worth based on its cash flows; comps tell you what someone can actually sell it for.
In practice, comps are used to:
Comps are a relative valuation. They tell you what the market values similar businesses at — not what the business is intrinsically worth. In a bull market, the entire comp set may be richly valued. In a downturn, every comp may be cheap. Always interpret comps alongside a DCF that anchors to fundamentals.
The quality of a comps analysis is determined almost entirely by the quality of the peer group. A poorly selected universe — companies that are too large, too different, or in the wrong business — will produce a wide, meaningless range. The goal is to find companies that a rational investor would actually consider substitutes for the target.
Sector and Business Model: This is the most important filter. Companies should be in the same end market and have a comparable revenue model (recurring vs. transactional, product vs. service). A pure-play SaaS company is not meaningfully comparable to a hybrid on-premise/software company even if both sell to enterprises.
Size (Revenue and EBITDA Scale): Multiples compress as companies scale. A $50M revenue SaaS startup might trade at 20x revenue; a $2B revenue SaaS company rarely does. As a rule of thumb, keep comparables within roughly 0.5x to 2x the size of the target on revenue and EBITDA. The smaller the peer group the more flexible you need to be, but avoid mixing micro-caps with large-caps.
Geography: U.S.-listed companies trade at different multiples than European or Asian equivalents, partly due to different investor bases, liquidity premiums, and regulatory environments. Keep the peer set in the same primary market unless cross-listing makes them genuinely comparable.
Liquidity: Thinly traded small-caps introduce noise. Focus on companies with sufficient trading volume that public market prices reflect real price discovery. A company trading 50,000 shares a day is not a reliable market signal.
We are valuing Meridian Software, a mid-market B2B SaaS company with $2.1B in revenue, $800M in EBITDA (38% margin), and a December fiscal year-end. The company sells workflow automation to manufacturing and logistics companies. Below is the selected comp universe after screening 40+ candidates.
| Company | Ticker | Revenue ($M) | EBITDA Margin | End Market | FY End | Included |
|---|---|---|---|---|---|---|
| Veridian Systems | VRID | 1,850 | 35% | Industrial SaaS | Dec | Yes |
| Apex Workflow | APWF | 2,400 | 41% | B2B Automation | Jun | Yes — calendarize |
| Cadence Platforms | CDNP | 1,650 | 33% | Supply Chain SaaS | Dec | Yes |
| Solera Enterprise | SLRA | 2,200 | 39% | Logistics Software | Sep | Yes — calendarize |
| Praxis Cloud | PRXC | 1,950 | 37% | Workflow Automation | Dec | Yes |
| Titan SaaS | TSAS | 12,000 | 28% | Enterprise Software | Dec | No — too large |
| Nimbus Tech | NMBT | 320 | 15% | SMB SaaS | Dec | No — too small, different buyer |
Two companies were excluded despite appearing relevant on the surface. Titan SaaS is nearly 6x larger and at a different margin profile — its multiple would compress the comps set unfairly. Nimbus Tech serves small businesses and has a fundamentally different growth and retention profile. Five comps is a reasonable set; anything fewer than three becomes hard to defend.
"Spreading" a comp means pulling the financial data from public filings — 10-K, 10-Q, earnings releases — and computing enterprise value, equity value, and relevant multiples. For each company, you need the income statement (revenue, EBITDA, net income) and the balance sheet (debt, cash) to compute enterprise value from market capitalization.
Enterprise value (EV) represents the total value of the business, as if you were buying it outright — equity plus net debt. Starting from market cap:
Market cap uses the current share price multiplied by diluted shares outstanding (including in-the-money options and convertible securities on a treasury stock method). Never use basic share count — it understates the true cost to acquire the company.
| Company | Revenue | EBITDA | Net Income | Mkt Cap | Debt | Cash | EV | EV/Rev | EV/EBITDA | P/E |
|---|---|---|---|---|---|---|---|---|---|---|
| Veridian Systems | 1,850 | 648 | 312 | 9,210 | 1,200 | 450 | 9,960 | 5.4x | 15.4x | 29.5x |
| Apex Workflow | 2,400 | 984 | 498 | 13,750 | 1,600 | 720 | 14,630 | 6.1x | 14.9x | 27.6x |
| Cadence Platforms | 1,650 | 545 | 241 | 7,480 | 900 | 280 | 8,100 | 4.9x | 14.9x | 31.0x |
| Solera Enterprise | 2,200 | 858 | 410 | 11,200 | 1,400 | 510 | 12,090 | 5.5x | 14.1x | 27.3x |
| Praxis Cloud | 1,950 | 722 | 341 | 10,100 | 1,100 | 390 | 10,810 | 5.5x | 15.0x | 29.6x |
| 25th Percentile | — | — | — | — | — | — | — | 5.1x | 14.1x | 27.4x |
| Median | — | — | — | — | — | — | — | 5.5x | 14.9x | 29.5x |
| 75th Percentile | — | — | — | — | — | — | — | 5.8x | 15.2x | 30.6x |
A few notes on this spread. All figures use LTM (last twelve months) financials — the most recent trailing twelve-month period, spliced together from the last annual report and the most recent quarterly filings. LTM is standard; forward estimates (NTM) are also common when you have reliable consensus estimates and want to value on future performance.
P/E is less useful for comps when companies have meaningfully different capital structures — two companies with identical operating performance but different leverage will show very different net income figures. Stick to EV/EBITDA and EV/Revenue as your primary multiples, and use P/E only as a secondary check.
Two of our five comps — Apex Workflow (June FY) and Solera Enterprise (September FY) — have fiscal years that do not end in December. If we use their reported annual numbers directly, we are comparing apples to oranges: Apex's FY2024 covers July 2023 through June 2024, while Meridian's FY2024 covers January through December 2024. They do not overlap the same calendar period.
Calendarization adjusts non-December fiscal year financials to align with a December 31 calendar year, so all companies in the comp set reflect the same time window.
For a June fiscal year company, the December calendar year EBITDA is constructed by taking the last six months of the FY2024 (July–December 2024) and the first six months of FY2025 (January–June 2025):
More generally, for any fiscal year that ends in month M (where M is the number of months into the calendar year):
Apex Workflow has a June 30 fiscal year end. Its reported figures are:
After calendarization, Apex's CY2024 EBITDA is $984M — slightly different from either reported FY number. This is the figure that goes into the comps table. The same process applies to revenue and any other metric you are spreading.
Calendarization matters most for rapidly growing companies, where the gap between FY2024 and FY2025 is large. For a slow-growth industrial business growing 3% per year, the difference between calendarized and as-reported is immaterial. For a SaaS company growing 20%+, skipping calendarization can distort multiples by several turns.
With the comps spread and calendarized, we apply the multiple ranges from the peer group to Meridian's financials. The output is a range of implied enterprise values and, after bridging through the capital structure, an implied equity value per share.
| Percentile | EV/Revenue (LTM) | EV/EBITDA (LTM) | EV/EBITDA (NTM) |
|---|---|---|---|
| 25th Percentile | 5.1x | 14.1x | 12.8x |
| Median | 5.5x | 14.9x | 13.5x |
| 75th Percentile | 5.8x | 15.2x | 13.9x |
| Mean | 5.5x | 14.9x | 13.5x |
Meridian Software: $2,100M LTM Revenue, $800M LTM EBITDA, $870M NTM EBITDA (estimated).
| Multiple Basis | Low (25th %ile) | Mid (Median) | High (75th %ile) |
|---|---|---|---|
| EV/LTM Revenue (×$2,100M) | $10,710 | $11,550 | $12,180 |
| EV/LTM EBITDA (×$800M) | $11,280 | $11,920 | $12,160 |
| EV/NTM EBITDA (×$870M) | $11,136 | $11,745 | $12,093 |
The three methodologies tell a consistent story: Meridian's implied enterprise value falls in the range of roughly $11.0B to $12.2B based on public market multiples. When your three multiple bases converge, you have confidence in the range. When they diverge significantly, investigate why — it is usually a sign that one metric is distorted (e.g., an EBITDA add-back inflating the denominator).
Enterprise value measures the value of the entire business (debt and equity combined). To get to equity value — and equity value per share — we bridge from EV through the capital structure.
Meridian's balance sheet: $1,350M total debt, $0 minority interest, $0 preferred, $510M cash. Net debt = $840M. Diluted shares outstanding: 285M.
| Item | Low Case | Mid Case | High Case |
|---|---|---|---|
| Implied Enterprise Value | $11,280 | $11,920 | $12,160 |
| Less: Total Debt | (1,350) | (1,350) | (1,350) |
| Plus: Cash & Equivalents | 510 | 510 | 510 |
| Implied Equity Value | $10,440 | $11,080 | $11,320 |
| Diluted Shares Outstanding (M) | 285 | 285 | 285 |
| Implied Share Price | $36.63 | $38.88 | $39.72 |
If Meridian's current share price is $35.00, the comps analysis suggests the stock is trading at a modest discount to peers — roughly 5% below the low end of the implied range. That is a meaningful input to an investment thesis, but not conclusive on its own.
If a company trades at a 15–20% premium to the median peer multiple, you need a reason: faster revenue growth, higher free cash flow conversion, a more defensible market position, or a superior management team. If you cannot articulate why the premium is justified, the stock is expensive. Conversely, a persistent 20% discount to peers either signals an undervalued opportunity or a structural problem the market has correctly identified. Comps do not tell you which — that is where your research comes in.
Other signals to read from the spread:
Enterprise value multiples — EV/Revenue, EV/EBITDA — are capital-structure neutral. Two companies with identical operating performance but different leverage will show very different P/E ratios because interest expense flows through net income. EV/EBITDA compares the operating business before financing decisions, making cross-company comparisons more meaningful. You use equity multiples (P/E) primarily for financial institutions, where the capital structure is the business model.
There is no hard rule, but fewer than three makes the analysis hard to defend — any single outlier dominates the median. Five to ten comparable companies is a typical range. For very specialized businesses (defense electronics, specialty chemicals, certain fintech niches), three or four true comparables is often the best you can do, and that is fine as long as you are transparent about it. The quality of the comp matters far more than the quantity — one irrelevant company at 8x drags your median down without adding insight.
LTM (last twelve months) uses the most recent trailing twelve months of actual reported financials — fully observable, no forecasting error. NTM (next twelve months) uses forward consensus estimates, which embed growth expectations. NTM multiples are more useful when a company is in an inflection period — a recent acquisition, a margin improvement initiative, or a product cycle entering a new phase. In a stable, slow-growth business, LTM and NTM multiples will be very close. When they diverge substantially, the market is paying for near-term change, and you need to stress-test whether consensus estimates are achievable.
A September FY company's fiscal year ends September 30. To express it on a December calendar year basis, I need October through December data — the last three months of the calendar year — which falls in the company's next fiscal year. So I take 9/12 of the current FY (covering October of the prior year through September) and 3/12 of the following FY (covering October through December). The formula is: CY Metric = (9/12 × FY_current) + (3/12 × FY_next). Applied to Solera Enterprise with FY2024 EBITDA of $836M and FY2025 EBITDA of $880M: CY2024 = (0.75 × $836M) + (0.25 × $880M) = $627M + $220M = $847M. That $847M is what goes in the comps table, not either reported FY figure.