Comparable Company Analysis: How to Build and Spread a Trading Comps Set
Comparable company analysis (sometimes called trading comps, comps analysis, or simply CCA) is the valuation method that prices a target business by benchmarking its trading multiples against a hand-picked set of publicly listed peers. The output is a per-share or enterprise value range expressed in multiples of EBITDA, revenue, or earnings, and it is the first valuation an analyst at Goldman Sachs, Morgan Stanley, Houlihan Lokey, Lincoln International, or Lazard will build on any sell-side mandate, IPO pitch, or fairness opinion. This guide walks the full 2026 methodology: peer selection, calendarization, capital structure adjustments, non-recurring scrubs, multiple calculation, the comps table itself, and the live spreading workflow that turns a Capital IQ download into a defensible valuation range.
The mechanics matter because regulators, courts, and acquirers all scrutinize the inputs. The Delaware Court of Chancery has rejected comp sets in In re Appraisal of Dell (2016) and DFC Global v. Muirfield (2017) for peer drift and stale data, and the SEC routinely comments on Item 1015 fairness opinion exhibits when comp peers look cherry-picked (see SEC Division of Corporation Finance M&A Interpretations). Get the peer set wrong and the entire valuation is wrong, no matter how clean the spreadsheet looks.
Quick-Reference: Comparable Company Analysis in 90 Seconds
| Item | Detail |
|---|---|
| Also known as | Trading comps, trading comparables, public comps, CCA, comps analysis |
| What it values | Enterprise value and equity value of a private or public target |
| Inputs needed | 5-12 listed peers, last reported financials, consensus forward estimates, capital structure as of pricing date |
| Core output | Multiple-based valuation range (typically EV/EBITDA, EV/Revenue, P/E) |
| Typical use cases | Sell-side pitch, IPO pricing, fairness opinion, board valuation update, 409A, purchase price allocation |
| Time to build (first pass) | 4-8 hours with Capital IQ or FactSet, 12-20 hours from scratch |
| Refresh cadence on live deal | Daily during marketing, weekly off-deal |
| Where it sits in the football field | Public benchmark anchor, typically left of precedent transactions, right of 52-week trading range |
| Common multiples | EV/EBITDA, EV/EBIT, EV/Revenue, P/E, P/B, EV/(EBITDA-Capex), FCF yield |
| Cost of data | FactSet ~$12,000/seat/yr, Capital IQ ~$13,000/seat/yr, Bloomberg Terminal $32,025/yr per Bloomberg’s own price card |
What Comparable Company Analysis Actually Measures
A comparable company analysis answers one question: if the market values these 8 peers at a median 11.2x next-twelve-months EBITDA, what is the defensible multiple to apply to my target? The math is trivial. The judgment is in the peer set, the period (last twelve months versus next twelve months), the EBITDA definition (reported, adjusted, run-rate), and the haircut or premium for size, growth, and margin differentials.
The method assumes three things that you have to defend in any committee meeting:
- Comparability: Peers operate in the same end markets, with similar capital intensity, growth, and margin profiles.
- Market efficiency: The peers trade at fair value, or at least at a value the bidder will accept as a benchmark.
- Translatability: Public-market multiples translate to a private control transaction with an adjustment for control premium and liquidity discount.
That third point is where the analysis breaks if you forget it. The historical control premium reported by FactSet MergerStat ran around 30.4% in 2024 (mean) and 27.9% (median) versus the one-day pre-announcement spot price; the 5-year average sits near 33%. The illiquidity discount on private equity stakes typically runs 15 to 35%, per repeated empirical work in the Stout Restricted Stock Study updated annually since 2005.
Trading Comps Versus Precedent Transactions Versus DCF
Analysts almost never present comparable company analysis on its own. It anchors the football field next to two other methods, each measuring something different:
| Method | What it captures | What it misses | Typical output range vs the others |
|---|---|---|---|
| Trading comps (CCA) | Current public-market sentiment on peer set | Control premium, synergies, deal-specific terms | Lowest of the three on average |
| Precedent transactions | Realized multiples in actual M&A deals, includes control premium | Deal vintages may be stale, synergies may be buyer-specific | Highest of the three on average |
| Discounted cash flow (DCF) | Intrinsic value of projected free cash flow | Highly sensitive to terminal value and discount rate assumptions | Wide range, often used as the swing factor |
For the DCF leg, see our walkthrough at discounted cash flow business valuation and the 2026-specific update at DCF valuation for a business sale in 2026. For the LBO floor leg that PE bidders run alongside, see our LBO model step-by-step guide and the paper LBO drill at paper LBO example walkthrough.
The empirical pattern, documented by Damodaran’s annual NYU Stern valuation data set, is that trading comps and precedent transactions tend to bracket the DCF, with precedents 20-35% above trading comps because of the control premium and synergy capture priced into actual deals.
Step 1: Build the Peer Set (The Single Most Important Decision)
Eight to twelve peers is the working range. Fewer than five and a single outlier swings the median; more than fifteen and the set drifts away from genuine comparability. Here is the screen that survives committee scrutiny:
- Industry classification: Start with GICS sub-industry code (8-digit) per the MSCI GICS standard, then layer in SIC 4-digit, NAICS 6-digit, and SEC SIC code from EDGAR. Cross-reference at least two classification schemes.
- Business mix: Pull the 10-K segment disclosure under ASC 280 from SEC EDGAR. Require at least 60% revenue overlap by segment. A diversified industrial whose target segment is 22% of revenue is not a comp, it is a noise generator.
- Size band: Constrain to roughly 0.3x to 3x the target’s revenue or EV. Mega-cap multiples do not translate to micro-cap targets and vice versa.
- Geographic mix: Match home-market revenue share within 15 percentage points where possible. A US-listed company with 80% European revenue is a European business with a US listing.
- Profitability profile: Match EBITDA margin band (within 500 basis points typically) and revenue growth band (within 500 basis points for stable industries, wider for hyper-growth).
- Trading liquidity: Require average daily traded value (ADTV) above $5 million per Refinitiv or Bloomberg. Thinly traded names print noisy multiples.
- Listing status: Exclude names in bankruptcy, going-private process, or those subject to a binding acquisition agreement. Bidder-locked names trade to deal value, not fundamentals.
Wachtell Lipton’s standard fairness opinion peer-set memo, summarized in their M&A practice publications, applies essentially this seven-filter test. So does the Skadden M&A practice in their fairness opinion engagement letters.
The “Tier 1, Tier 2, Tier 3” Bucketing
Senior bankers split the final set into three buckets:
- Tier 1 (3-5 names): Direct competitors with near-identical business mix. These drive the recommended valuation range.
- Tier 2 (3-5 names): Adjacent companies with strong overlap but a meaningful difference (e.g., a heavier services mix, a different geographic skew).
- Tier 3 (2-4 names): Broader sector benchmarks for context but not the recommendation.
Houlihan Lokey’s published valuation primer, available on their insights portal, presents virtually every fairness opinion this way. Lincoln International’s quarterly private company multiples report uses an analogous tiering for the public proxy set.
Step 2: Calendarize the Financials
Peers have different fiscal year ends. Calendarization standardizes them to a common period (almost always calendar-year LTM and CY+1 NTM) so that the multiples are apples to apples.
The mechanics for a peer with a June fiscal year end being calendarized to a December 31, 2025 LTM date:
| Step | Action | Source |
|---|---|---|
| 1 | Pull FY June 2025 reported financials | 10-K filed in August/September |
| 2 | Subtract the H1 FY June 2025 reported financials (July to December 2024) | 10-Q from H1 of the fiscal year |
| 3 | Add the H1 FY June 2026 reported financials (July to December 2025) | 10-Q from the most recent quarter |
| 4 | Result is calendar LTM ending December 31, 2025 | Cross-check against any preannouncement |
For forward-looking multiples (NTM, NFY+1, NFY+2), pull consensus estimates from a vendor that calendarizes for you, typically Visible Alpha, S&P Capital IQ, or FactSet StreetAccount. Always disclose the consensus date and the contributor count. A 3-contributor consensus is not a consensus, it is a coin flip.
Step 3: Clean the Numerator (Enterprise Value)
The enterprise value calculation is where most junior bankers introduce errors. The textbook formula is:
EV = Equity Market Cap + Total Debt + Preferred Equity + Non-Controlling Interest – Cash and Short-Term Investments
That formula is necessary but not sufficient. The professional version layers in seven adjustments:
| Adjustment | Add or subtract | Authority |
|---|---|---|
| Operating lease liabilities (ASC 842) | Add | FASB ASC 842, effective FY 2019 public, FY 2022 private |
| Pension underfunding (PBO less plan assets) | Add | ASC 715 disclosure in 10-K footnote |
| OPEB underfunding | Add | Same 10-K footnote |
| Convertible debt above par (if treasury method) | Add the in-the-money portion as equity | FASB ASC 260 EPS treatment |
| Restricted cash | Do not net against debt | SEC Staff Accounting Bulletin and consensus practice |
| Cash held abroad with tax leakage | Discount for repatriation tax under TCJA | IRC Section 965 transition tax history, current GILTI under IRC Section 951A |
| Equity investments at FMV | Subtract from EV | ASC 321 disclosure |
Skipping the operating lease adjustment is the single most common error. For a retailer or restaurant chain, operating lease liabilities can be 30-60% of EV. Forgetting them inflates EBITDA-based multiples by an equivalent amount and produces a comp set that looks artificially cheap.
Step 4: Clean the Denominator (Adjusted EBITDA, Adjusted EPS)
Reported EBITDA is rarely the EBITDA you put in a comp. The job is to scrub non-recurring, non-cash, and non-operating items to get to a clean run-rate number. Practitioner consensus, codified in the SEC Regulation G non-GAAP reconciliation requirement and Item 10(e) of Regulation S-K, supports the following standard adjustments:
- Stock-based compensation (SBC): The contentious one. Buy-side analysts increasingly include SBC as a real expense per Buffett’s 2002 shareholder letter (“if options aren’t a form of compensation, what are they?”), but sell-side practice typically adds it back. Houlihan Lokey discloses both in their fairness opinions. State your treatment explicitly.
- Restructuring and severance charges (non-recurring component)
- Impairments of goodwill, intangibles, or fixed assets under ASC 350 and ASC 360
- Litigation reserves and settlements (one-off)
- M&A transaction costs (banker fees, legal, diligence)
- Integration costs in the post-close year
- Gain or loss on sale of business or asset disposals
- Mark-to-market on derivatives (if not in operating segment)
- FX remeasurement gains and losses
- COVID-era one-offs (now largely retired but still appear in 2020-2022 LTM windows)
For tech and high-growth comps where SBC is 8-20% of revenue, the SBC treatment alone can move the comp set median EV/EBITDA multiple by 3-5 turns. Bessemer’s State of the Cloud 2024 report flagged exactly this divergence between sell-side and buy-side cloud comp tables.
Step 5: Calculate the Multiples
The standard multiple stack for a comparable company analysis, in roughly the order you would present them:
| Multiple | Formula | Best for | Watchout |
|---|---|---|---|
| EV/Revenue (LTM, NTM) | EV / Revenue | High-growth, pre-profit, early SaaS | Ignores profitability entirely |
| EV/Gross Profit | EV / Gross Profit | SaaS where COGS classification is consistent | Less common but useful for variable-margin businesses |
| EV/EBITDA (LTM, NTM) | EV / EBITDA | Mature businesses, LBO transactions, most industries | Ignores capex intensity |
| EV/(EBITDA-Capex) | EV / (EBITDA – Capex) | Capital-intensive businesses (telecom, industrials, real estate) | Capex can be lumpy, normalize with 3-yr average |
| EV/EBIT | EV / EBIT | Businesses where D&A is economically meaningful | Distorted by purchase accounting intangible amortization |
| P/E (LTM, NTM) | Price / EPS | Banks, insurers, retail investor benchmark | Capital-structure dependent, useless for cross-gearing comparison |
| FCF Yield | Free Cash Flow / Equity Market Cap | Cash-generative mature businesses | Capex normalization required |
| PEG Ratio | P/E divided by EPS Growth Rate | Growth-adjusted P/E, popularized by Peter Lynch | Breaks down at sub-5% growth |
For PE-backed deals and credit underwriting, EV/EBITDA dominates. The Refinitiv LPC middle-market loan pricing reports have shown median sponsored EV/EBITDA in the 6.5-7.5x range for senior debt and 11-13x for total purchase price across 2022-2025. For public equity research, NTM P/E remains the lead multiple, with NTM EV/EBITDA second.
Step 6: Build the Comps Table (The Output Document)
The comps table is the deliverable. A clean institutional comps table has roughly 50 columns and 8-15 rows. Here is the standard column architecture from left to right:
| Block | Columns |
|---|---|
| Identifier | Company name, ticker, exchange, GICS sub-industry |
| Market data (as of pricing date) | Share price, 52-week high/low, equity market cap, EV, ADTV |
| Capital structure | Total debt, cash, net debt, preferred, NCI, lease liabilities |
| Size | LTM revenue, LTM EBITDA, employees |
| Profitability | LTM gross margin, EBITDA margin, EBIT margin, net margin, ROIC |
| Growth | Revenue growth LTM, NTM, NFY+1, NFY+2; EBITDA growth same |
| Gearing | Net debt / EBITDA, EBITDA / interest |
| Trading multiples (LTM) | EV/Revenue, EV/EBITDA, EV/EBIT, P/E |
| Trading multiples (NTM) | EV/Revenue, EV/EBITDA, EV/EBIT, P/E |
| Trading multiples (NFY+1) | Same four |
| Summary statistics | High, low, mean, median, 25th percentile, 75th percentile |
The summary statistics row at the bottom (the “stats block”) is what feeds the valuation range. Sell-side analysts and M&A advisors typically anchor on the median and 25th-75th percentile to set a recommended range, then triangulate against the target’s growth and margin profile.
Anyone running this workflow for a fairness opinion or sell-side process needs to coordinate closely with the M&A advisor heading the engagement, and ideally with a former sell-side analyst or private equity analyst who has built comp sets in the relevant sector.
Step 7: Spreading the Comps (The Live Workflow)
“Spreading” is the verb bankers use for the calendarization, EV calculation, EBITDA scrubbing, and multiple math executed across all peers in parallel. Wall Street Oasis and Mergers & Inquisitions both publish detailed templates (see Wall Street Oasis IB forum and Mergers & Inquisitions comparable company analysis primer), and the workflow is functionally identical across firms.
A clean spreading workflow on a 10-peer set:
- Hour 0-1: Build the peer screen, finalize the 10 names, pull Capital IQ tearsheets.
- Hour 1-3: Pull last 10-K and last four 10-Qs for each peer. Pull consensus estimates from Visible Alpha.
- Hour 3-6: Calendarize. Most institutional templates have a calendarization tab that does the H1/H2 math automatically once you tag fiscal year ends.
- Hour 6-8: Spread the EBITDA scrub. Pull the non-GAAP reconciliation from each 10-K/10-Q. Categorize every add-back as recurring or non-recurring per firm policy.
- Hour 8-10: Build the EV bridge for each peer. Pull lease liabilities, pension, OPEB, NCI from the balance sheet footnotes.
- Hour 10-12: Calculate multiples. Cross-check against Bloomberg or Capital IQ pre-built multiples and reconcile any difference greater than 2%.
- Hour 12-14: Build the summary stats and the valuation range output. Stress test by dropping the high and low peers.
- Hour 14-16: Pretty up the formatting, add data-as-of footnotes, freeze the file, send to associate for review.
On a live deal, the comp set refreshes daily during marketing because peer share prices move. The refresh is typically automated via a Capital IQ or FactSet data plug-in that pulls fresh share prices, market caps, and consensus estimates each morning before the deal team’s 7am check.
Step 8: Worked Example, $400M Revenue Specialty Industrial
Assume a sell-side mandate for a privately held specialty industrial manufacturer with $400M LTM revenue, $72M LTM adjusted EBITDA (18.0% margin), 6.5% revenue growth, and $35M of net debt. The screen produces the following peer set:
| Peer | LTM Revenue | LTM EBITDA | EBITDA Margin | Revenue Growth | EV/LTM EBITDA | EV/NTM EBITDA |
|---|---|---|---|---|---|---|
| Peer A | $385M | $78M | 20.3% | 7.2% | 11.8x | 10.9x |
| Peer B | $410M | $70M | 17.1% | 5.4% | 10.2x | 9.6x |
| Peer C | $520M | $98M | 18.8% | 8.1% | 12.4x | 11.3x |
| Peer D | $340M | $58M | 17.1% | 4.9% | 9.7x | 9.1x |
| Peer E | $465M | $84M | 18.1% | 6.8% | 11.1x | 10.4x |
| Peer F | $280M | $48M | 17.1% | 9.2% | 12.9x | 11.6x |
| Peer G | $610M | $112M | 18.4% | 5.8% | 10.6x | 10.0x |
| Peer H | $355M | $66M | 18.6% | 7.5% | 11.5x | 10.7x |
| Median | $397M | $74M | 18.2% | 6.8% | 11.3x | 10.5x |
| 25th-75th | – | – | 17.6%-18.7% | 5.7%-7.5% | 10.5x-11.7x | 9.8x-11.0x |
Applying the 25th-75th percentile EV/LTM EBITDA range to the target’s $72M EBITDA produces an EV range of $756M to $842M, with a midpoint at $814M. After subtracting $35M of net debt, the implied equity range is $721M to $807M.
This is the pre-control-premium range. For a fairness opinion on a sale process, the precedent transactions analysis would push the high end 20-35% higher to reflect the control premium realized in actual deals. For a 409A valuation of a still-private company, no control premium is applied and a discount for lack of marketability is added per AICPA Valuation Guide for Privately Held Company Equity Securities Issued as Compensation.
Step 9: Pitfalls That Cost Bankers Their Bonuses
The most expensive errors in comparable company analysis, ranked by frequency in my experience and per the standard mistakes catalogued in CFA Institute’s Equity Valuation: Science, Art, or Craft:
- Stale share prices. Pulling prices “as of last Friday” on a Wednesday pitch is a fireable offense at most bulge brackets. Use T-1 close at minimum.
- Forgetting operating leases. Post-ASC 842, the right-of-use asset and lease liability both belong on the balance sheet. Older comp templates often ignore them and produce stale numbers.
- Mixing reported and adjusted EBITDA across peers. Each peer adjusts differently. Pick one definition and apply it consistently.
- Stock-based compensation inconsistency. Either always add it back or never. Mixing is the surest way to a wrong median.
- Calendarization errors. Adding two half-years from different fiscal years without verifying segment continuity (e.g., a divested business in one period but not the other).
- Outlier peers driving the median. Always show high, low, mean, and median. A single peer at 25x EBITDA in an otherwise 11x sector should be flagged and explained, not silently included.
- Cherry-picked peer set to fit a desired valuation. This is the issue Delaware courts have called out repeatedly. In re Appraisal of Dell (2016) and DFC Global v. Muirfield (2017) both featured peer set criticism. The defensible approach is to publish the screen criteria first and let the names fall out of the screen.
- Ignoring trading liquidity. A peer with $200,000 ADTV trades on noise. Its multiples are not market clearing.
- Same-day printing of multiples and market data from different sources. A Bloomberg market cap as of 4:00 PM and a Capital IQ market cap as of 6:00 PM the same day can differ by a basis point on a quiet day, or by 80 basis points on a volatile day. Use one source for everything.
- Counting cash held for restricted purposes. Cash collateralizing a litigation bond or held in a foundation account is not available to bondholders or equityholders. Do not net it.
Step 10: How Public Comps Translate to Private Targets
The mechanical hand-off from public-market multiples to a private control transaction requires three adjustments, in this order:
| Adjustment | Direction | Typical magnitude | Source |
|---|---|---|---|
| Control premium | Add | 25-35% to equity value | FactSet MergerStat annual control premium study; Kroll Cost of Capital Navigator |
| Illiquidity / DLOM | Subtract | 15-35% for unregistered private interest | Stout Restricted Stock Study |
| Size discount (small-cap premium) | Subtract from multiple if target is materially smaller | 1-3 turns of EBITDA for businesses below $50M EV | Kroll/Duff & Phelps Size Premium data, GE Capital middle market multiples reports |
For a strategic sale process, the control premium nets against the size discount and typically the comp-implied valuation lands within 10-15% of the precedent-transaction-implied valuation. For a financial buyer running an LBO, the entry multiple is bounded above by the maximum debt the lenders will support, currently roughly 6-7x EBITDA of senior debt per the Refinitiv LPC data, plus equity check sizing. See our LBO model from scratch walkthrough for the full bidder-side math, plus the deeper business valuation formula methods and math piece for the full Gordon-Growth and venture-capital-method context, and how to determine the value of a business for the owner-facing summary.
Step 11: Sector-Specific Multiple Conventions
Not every sector trades on EV/EBITDA. The sector-standard multiple is what your buyer or investor is actually quoting, and missing it is a tell that the banker has not run a deal in the space:
| Sector | Lead multiple | Secondary |
|---|---|---|
| Software / SaaS | EV / NTM Revenue (or EV / ARR) | EV / NTM Gross Profit, Rule of 40 scoring |
| Banks / depositories | P / Tangible Book Value | P/E NTM |
| Insurance carriers | P / Book Value | P/E NTM, ROE-implied multiple |
| REITs | P / FFO (funds from operations) | NAV premium/discount per Green Street Advisors NAV methodology |
| Oil & gas E&P | EV / EBITDAX, EV / Proved Reserves (boe) | EV / Production (boe/day) |
| Pharma / biotech (late-stage) | EV / NTM Revenue, EV / Peak Sales | Risk-adjusted NPV per molecule |
| Asset managers | EV / AUM, P/E | EV / Management Fee Revenue |
| Hospitals / health systems | EV / EBITDA | EV / Bed, EV / Admission |
| Retail / restaurants | EV / Sales, EV / EBITDA (post-IFRS 16 / ASC 842 lease-adjusted) | EV / Store Box |
| Mining / metals | EV / Reserves, EV / Resources | P / NAV |
| Telecom / cable | EV / EBITDA, EV / Subscriber | EV / Tower (for tower cos) |
| Industrials (capital-intensive) | EV / EBITDA, EV / (EBITDA – Capex) | P/E |
For SaaS specifically, the Bessemer State of the Cloud 2024 report tracked EV/NTM revenue medians across the Cloud 100 dropping from a peak of 34x in late 2021 to roughly 7.5x by year-end 2024, with the highest-growth quartile at 13-15x. The compression is the single largest multiple reset in the public software market since the 2000-2002 dot-com unwind.
Step 12: Comparable Company Analysis in Legal and Regulatory Settings
Comparable company analysis carries legal weight in five recurring settings, each with its own admissibility standard:
- Delaware appraisal proceedings (DGCL Section 262): The Court of Chancery weighs CCA alongside DCF and precedent transactions. The 2017 Delaware Supreme Court ruling in DFC Global v. Muirfield Value Partners emphasized deal price as a strong indicator when the sale process was well-run, demoting CCA when peer drift was material. See the opinion at Justia DFC Global v. Muirfield.
- Fairness opinions under SEC Item 1015: A board ordering a fairness opinion must disclose the methodology including the comp set and the bases for inclusion and exclusion. See SEC Regulation S-K Item 1015 guidance.
- Tax-court valuation disputes: The IRS Engineering Program and the Tax Court routinely consider CCA in gift, estate, and Section 2701-2704 disputes. Revenue Ruling 59-60, still binding for closely-held business valuation, expressly lists “the market price of stocks of corporations engaged in the same or a similar line of business” as one of eight factors. See Rev. Rul. 59-60.
- 409A valuations of private company equity: The AICPA Practice Aid (2013 revision) endorses CCA as a Market Approach method alongside Income (DCF) and Asset approaches. Carta, Pulley, and Eqvista all use CCA as a primary input in their automated 409A models per their published methodology documents.
- Purchase price allocation under ASC 805: Acquirers running PPA on closed deals use CCA-derived multiples to corroborate the income approach valuations of acquired intangible assets and reporting units for the goodwill impairment test under ASC 350.
For the deal-document side that consumes the CCA output, see what is a stock purchase agreement, material adverse effect clause, installment sale versus cash sale, QSBS Section 1202 small business stock, golden parachute 280G, and founder shares.
Step 13: Tools, Data Vendors, and Templates
The institutional tooling for comparable company analysis is concentrated among five vendors plus Excel:
| Vendor | Annual price (per seat) | Best feature | Weakness |
|---|---|---|---|
| FactSet | ~$12,000 | Pre-built peer screen and spreading templates | Steep learning curve |
| S&P Capital IQ | ~$13,000 | Best fundamental coverage of small-caps and private companies | Add-back data not always reconciled |
| Bloomberg Terminal | ~$32,025 per Bloomberg’s price card | Real-time market data and the EQS screener | Most expensive, sometimes redundant with FactSet for valuation work |
| Refinitiv Workspace (LSEG) | ~$22,000 | I/B/E/S consensus and deal data | Migration from Eikon caused 2023-2024 stability issues |
| Visible Alpha | ~$15,000 | Granular contributor-by-contributor consensus model lines | Coverage limited to large- and mid-caps |
| Excel (with manual SEC EDGAR pulls) | $0 incremental | Free, total control, defensible | 20+ hours per refresh |
Boutique advisors and emerging managers often start with Excel plus free SEC EDGAR pulls before committing to a paid platform. For PE firms, Pitchbook (~$25,000+/yr) and Preqin (~$15,000+/yr) are typically additive to one of the above for private market deal data.
Step 14: Interview and Modeling Test Questions on Comparable Company Analysis
Investment banking and private equity interviews regularly include CCA mechanics. The most common ten questions, sourced from Mergers & Inquisitions, Wall Street Oasis, and the published interview guides from Vault and Wetfeet, plus the Street of Walls modeling primer:
- Walk me through how you would build a comp set from scratch.
- Why might you use EV/EBITDA instead of P/E?
- How do you treat stock-based compensation in adjusted EBITDA?
- How do operating leases under ASC 842 affect enterprise value?
- How do you calendarize a peer with a fiscal year ending in March?
- What is a control premium and where does it come from?
- How do you handle a peer that recently announced a divestiture mid-period?
- Walk me through every line of the EV bridge for a typical industrial company.
- Why is EV/Revenue used for high-growth SaaS instead of EV/EBITDA?
- Your comp set median is 11x EBITDA but your target’s growth and margin are below the median peer. What multiple do you actually apply?
The answer to question 10 is the entire point of the exercise. The multiple you apply is not the median. It is the median adjusted for the target’s relative growth, margin, scale, and risk profile, justified explicitly to the deal team and the client.
Step 14a: International CCA, Cross-Border Peer Sets, and IFRS Reconciliation
A US-only peer set is rarely enough for a global business. When a target operates across multiple regions, the peer set has to span home-market and foreign listings, and the financials need to be reconciled from local GAAP or IFRS to a common framework before multiples are calculated. The big three issues:
- IFRS 16 versus ASC 842 lease accounting: Both standards capitalize operating leases, but IFRS 16 treats the entire P&L charge as a combination of depreciation (in EBIT) and interest (below EBIT), while ASC 842 keeps the full lease expense in operating costs for operating leases. The result: IFRS reporters show structurally higher EBITDA than US GAAP reporters even for identical economics. The IFRS 16 standard text and the ASC 842 codification diverge here. Adjust either by stripping IFRS 16 lease interest and depreciation back into operating cost, or by converting ASC 842 operating lease expense to a finance-lease presentation.
- R&D capitalization: IAS 38 permits capitalization of development costs that meet six criteria; US GAAP under ASC 730 mandates expensing of nearly all R&D. Software development costs follow ASC 985-20 (external use) or ASC 350-40 (internal use), with their own capitalization rules. For a tech peer set spanning US and European listings, restate R&D treatment to a common basis or the EBITDA margins are not comparable. The IAS 38 standard documents the IFRS treatment.
- FX translation for multiples: Use period-average rates for income statement items (EBITDA, revenue) and period-end rates for balance sheet items (debt, cash). Mixing the two distorts EV/EBITDA. The Bloomberg-published FX rate page and the European Central Bank’s euro reference exchange rates are the typical reference sources.
For cross-border deals, the country risk premium adjustment also matters at the multiple-application stage. Damodaran’s annual country risk premium update at the NYU Stern country risk premium data provides the standard reference, used in every Houlihan Lokey, Lazard, and Kroll fairness opinion that touches an emerging-markets target.
Step 14b: ESG Adjustments and the 2026 Multiple Premium-Discount Debate
Since 2021, sell-side research from Morgan Stanley, BofA, and Bernstein has documented an emerging “green premium” or “brown discount” in trading multiples for energy-intensive sectors. The MSCI ESG Ratings methodology and the S&P Global Corporate Sustainability Assessment both feed into the data vendors most analysts consult, and a growing minority of fairness opinions now disclose ESG-adjusted comp sets as a sensitivity. The empirical gap, as quantified in a 2024 Bank for International Settlements working paper (see BIS Working Papers index), shows oil and gas E&P trading at 30-50% discounts to historical EV/EBITDA averages, while best-in-class renewables developers carry premiums of 40-80% versus integrated utility peers.
For a CCA on an industrials, utilities, or energy target, the question of whether to include or exclude the ESG laggards in the peer set is now a methodology disclosure point. Two defensible approaches: (1) include all peers regardless of ESG profile and show the high-low range, or (2) split the peer set into “transition-ready” and “legacy” tiers and apply the relevant tier multiple to the target. Either is defensible; mixing them silently is not.
Step 15: Refresh Cadence and Versioning
On a live sell-side process, the comp file lives in a controlled folder with version history. The institutional pattern:
- Daily refresh of share prices, market caps, and EV during marketing (3-6 weeks)
- Weekly refresh of full multiples (the calendarization and EBITDA scrub do not change daily)
- Monthly refresh of consensus estimates as analysts update models around earnings
- Full re-screen of the peer set quarterly off-deal, or any time there is a transformational event (M&A in the peer set, IPO, take-private, bankruptcy filing)
- Hard version snapshot at every committee meeting and every materially changed valuation presentation, with version number and as-of-date stamped on every page
The Lazard, Houlihan Lokey, and Lincoln International published procedures for fairness opinion preparation, summarized in their public engagement materials, all require an as-of-date footnote on every page of the comps tab and a contemporaneous email confirming the data source used.
Step 16: Terminology Notes and Common Aliases
Practitioners use several different names for the same workflow, and a junior analyst should be ready for any of them in a senior banker’s question. The aliases:
- Comparable company analysis (CCA) is the formal academic and textbook name, used in Damodaran’s Investment Valuation and the CFA curriculum.
- Comparable companies analysis (plural “companies”) is the same thing, common in older textbooks and in UK practice.
- Trading comps and trading comparables are the bulge-bracket shorthand and what most practitioners on a live deal will actually say in conversation.
- Public comps emphasizes the public-market input set, often used in contrast with “private comps” which is shorthand for precedent transactions sourced from private deal databases.
- Comps analysis and comps table refer narrowly to the workflow and the deliverable, respectively.
- Spreading comps refers specifically to the calendarization and EBITDA-scrub workflow that produces the inputs to the comps table.
All six terms point at the same methodology. The CFA Institute’s Equity Valuation: Applications and Processes refresher reading uses “comparable company analysis” as the lead term and lists the aliases.
TLDR and Takeaways
- Comparable company analysis (CCA, trading comps, comps analysis) values a target by benchmarking it against 8-12 listed peers using EV/EBITDA, EV/Revenue, P/E, and sector-specific multiples.
- The peer set is the single most important decision in the entire workflow. Use a seven-filter screen: GICS, business mix, size, geography, profitability, liquidity, listing status. Bucket into Tier 1, Tier 2, Tier 3.
- Calendarize to a common period (LTM, NTM, NFY+1). Pull H1/H2 quarters from 10-Qs to bridge fiscal year ends.
- Clean the numerator: equity market cap + total debt + preferred + NCI + operating leases (ASC 842) + pension/OPEB underfunding – cash and short-term investments.
- Clean the denominator: scrub for non-recurring restructuring, impairments, M&A costs, litigation, FX, and decide a consistent SBC treatment.
- Build the 50-column comps table with summary stats. The 25th-75th percentile range feeds the recommended valuation range.
- Spreading takes 14-16 hours on a 10-peer set from scratch, then refreshes daily during a live deal.
- Translate public-market multiples to a private control transaction with a 25-35% control premium, a 15-35% illiquidity discount, and a sector- and size-specific size discount.
- Delaware Chancery, the IRS, the SEC, and the AICPA all rely on CCA as one of three to four primary valuation methods. Cherry-picked peer sets get rejected.
- Sector matters: SaaS trades on EV/Revenue, banks on P/Tangible Book, REITs on P/FFO, E&P on EV/EBITDAX. Use the multiple your buyer actually quotes.
A defensible comparable company analysis is the table stakes valuation deliverable for any sell-side, fairness opinion, IPO, 409A, or fund-level mark. Build the peer screen first and let the names fall out, scrub the EBITDA and EV bridges consistently across all peers, present the summary stats, and apply the multiple to the target after honestly assessing where the target sits on growth, margin, scale, and risk relative to the median peer. That is the workflow that holds up in front of a board, a regulator, and the Court of Chancery.