Public Comps: How to Build a Public Company Trading Comps Set

Public comps, short for public company comparable trading multiples, is the valuation method that benchmarks a target business against publicly traded peers using market-derived multiples such as EV/Revenue, EV/EBITDA, EV/EBIT, and P/E. It is the first valuation any sell-side analyst pulls up when a banker asks “where does this trade?” because it answers a different question than a discounted cash flow: not what the business is worth in theory, but what the market is pricing right now for businesses that look like it. The Goldman Sachs M&A practice, the Morgan Stanley equity research desks, and every bulge-bracket pitch book published between 2020 and 2026 lead with a public comps table on page 4 or 5 for exactly that reason (Goldman Sachs Insights).

This guide walks through how to build a defensible public comps set the way a first-year IB analyst at Houlihan Lokey or Lazard is taught to build one: peer screening, capital structure normalization, calendarization, multiple calculation, scrubbing for one-time items, and presenting a tight median/mean/quartile table that a managing director will not red-pen at 11pm the night before a fairness opinion goes out.

Public Comps in 60 Seconds: Quick-Reference Table

Element	What It Is	Typical Source	Key Watch-Out
Peer set size	6-12 pure-play companies	Capital IQ, FactSet, Bloomberg	Too few = noisy median, too many = diluted comparability
Equity value (market cap)	Shares outstanding x current price	10-Q cover page + closing price	Use diluted shares via treasury stock method
Enterprise value (EV)	Equity + debt + preferred + minority interest – cash	Most recent 10-Q balance sheet	Subtract only excess cash if operating cash is material
EV/Revenue	EV / LTM or NTM revenue	Filings + sell-side consensus	Pre-profit SaaS uses this; mature industrials rarely
EV/EBITDA	EV / LTM or NTM EBITDA	Calculate EBITDA, do not trust GAAP	Add back stock-based comp only when peers do
P/E	Share price / diluted EPS	Income statement	Distorted by capital structure, tax rate, one-timers
Calendarization	Align all peers to same fiscal year-end	Manual stub-period math	Required when one peer has Jan FYE and another Dec FYE
Output benchmark	Median + mean + 25th/75th percentile	Excel PERCENTILE.INC	Median is the headline, mean shows skew

The full mechanics for each row are in the sections below. If you want to see how this same logic gets applied to a take-private LBO, the LBO model from scratch guide picks up where this one ends.

Why Public Comps Exist: The Market-Based Valuation Anchor

The three valuation methods every banker is trained on are intrinsic (DCF), market (public comps + precedents), and transaction (LBO). A DCF tells you what a business is worth based on a forecast of free cash flows and a discount rate; public comps tell you what the market is willing to pay today for businesses that share the same growth, margin, and risk profile (CFA Institute Curriculum, Equity Valuation).

The reason public comps anchor the football-field valuation chart that every banker shows their client is simple: a DCF is a hostage to its assumptions. Change the perpetuity growth rate from 2.5% to 3.0% and the implied equity value swings 15-25% in a typical industrial model. Public comps, by contrast, are observable. Apple closed at $234.41 on a given day, that is its price, no assumption required. The discipline of public comps is therefore not in the calculation but in the peer selection and the scrubbing.

A Morgan Stanley research note dated March 2025 on US software valuations noted that the median EV/NTM Revenue multiple for the 25-company SaaS coverage universe had compressed from 12.5x in late 2021 to 6.8x by Q1 2025, even as median revenue growth held steady at 18% (Morgan Stanley Insights). That kind of multiple compression is invisible in a DCF unless you are willing to mark your discount rate to market every quarter. Public comps make it impossible to ignore.

Step 1: Define the Peer Set

Peer selection is where most junior analyst comps sets fail their first review. The default temptation is to type the target’s industry code into Capital IQ and dump the top 25 results into the table. A managing director will throw that table back in five seconds. The correct screen is multidimensional.

The Five-Filter Peer Screen

Industry classification (GICS or NAICS): Start with the target’s six-digit GICS sub-industry code. The S&P Global GICS taxonomy, jointly maintained by S&P and MSCI, is the institutional standard (S&P Dow Jones GICS).
Business model: Recurring vs transactional, B2B vs B2C, asset-heavy vs asset-light. A pure-play SaaS peer is not a comp for a hybrid hardware-plus-services business even if both sit in the same GICS code.
Size band: Target revenue +/- 50% is a reasonable starting band; widen to +/- 75% if the peer set comes back below 6. A $200M revenue target should not have $20B revenue Microsoft in its peer set even if the GICS code matches.
Geographic mix: A US-only operator should not be benchmarked against a peer with 60% emerging-markets revenue, since FX and country risk premiums distort the multiple.
Growth and margin profile: A 30% grower at 25% EBITDA margins is not a comp for a 5% grower at 25% EBITDA margins. Filter the long list down to peers within +/- 500 bps of revenue growth and +/- 500 bps of EBITDA margin.

A typical screen takes the 25-company industry universe down to 6-10 names. If you end up with 3, the target is genuinely unique and you should disclose that in the footnotes and supplement with precedent transactions. The sell-side analyst career guide covers how junior bankers are evaluated on the rigor of this screen during their performance reviews.

Step 2: Pull the Financials From the Right Source

Three data pulls drive every comps table: balance sheet (for the EV bridge), income statement (for the denominator), and shares outstanding (for the market cap). The source hierarchy is filings first, data provider second.

10-K and 10-Q: The SEC EDGAR system is the canonical source. A 10-Q filed within the last 45 days gives you the most recent balance sheet for the EV bridge (SEC EDGAR Company Search).
Capital IQ / FactSet / Bloomberg: Use these for the standardized financials and consensus estimates. The standardization saves time but check the footnotes, providers often differ on whether stock-based comp is added back to EBITDA.
Earnings release + supplemental: Many companies report non-GAAP adjusted EBITDA, organic revenue growth, and pro forma share counts in the earnings press release that are not in the 10-Q. The supplemental deck for software companies is where you find committed annual recurring revenue (cARR) and net revenue retention (NRR).
Conference call transcript: The Q&A section often clarifies one-time items, segment splits, and forward-looking guidance that did not make it into the prepared remarks. Seeking Alpha and the company IR site both post transcripts within 24 hours.

For UK and EU peers, the equivalent of EDGAR is Companies House and the ESMA filings portal. Annual reports filed under IFRS need a separate set of adjustments for capitalized R&D and IFRS 16 lease accounting before they are comparable to US GAAP filings (IFRS Standards List).

Step 3: Calculate Equity Value and Enterprise Value

Equity Value (Market Capitalization)

Equity value is diluted shares outstanding times current share price. Diluted is the key word: basic share count ignores options, RSUs, convertible notes, and warrants that are in the money. The treasury stock method (TSM) is the GAAP-mandated approach for options and RSUs; the if-converted method handles convertible bonds (ASC 260, Earnings Per Share).

Worked example: a company with 100M basic shares, 8M options struck at $20 with current price $50, has TSM-diluted shares of 100M + 8M – (8M x $20 / $50) = 100M + 8M – 3.2M = 104.8M. At $50 share price, equity value = $5,240M. This same logic feeds into the LBO sources-and-uses table covered in the LBO model step-by-step guide.

Enterprise Value (EV)

EV = Equity Value + Total Debt + Preferred Stock + Minority Interest – Cash and Equivalents. The Damodaran framework that every banker learns in training adds a few refinements: include capitalized operating leases (post-ASC 842 these sit on the balance sheet anyway), subtract only excess cash if the business needs working cash to operate, and include the unfunded portion of pension obligations as a debt-like item (Aswath Damodaran NYU Stern).

EV Bridge Line Item	Source on 10-Q	Treatment
Diluted equity value	Cover page + closing price	Add
Short-term debt	Current liabilities	Add at book value
Long-term debt	Long-term liabilities	Add at book value (use market value if bonds trade)
Operating lease liability	Current + long-term lease liability	Add (post-ASC 842)
Preferred stock	Equity section	Add at liquidation value
Non-controlling interest	Equity section	Add at book value
Cash and equivalents	Current assets	Subtract
Short-term investments	Current assets	Subtract if marketable
Unfunded pension	Pension footnote	Add (debt-like)

Step 4: Calculate the Numerator Multiples (EV/Revenue, EV/EBITDA, EV/EBIT, P/E)

Once EV and equity value are nailed, the multiples are arithmetic. The art is in choosing which multiple to feature.

EV/Revenue

EV / LTM Revenue and EV / NTM Revenue. Used when the business is unprofitable (early-stage SaaS, biotech, hyper-growth) or when EBITDA is too volatile to be meaningful. Bessemer’s State of the Cloud 2025 report tracked median EV/NTM Revenue for public cloud companies at 6.2x with a top-quartile of 11.4x (Bessemer Venture Partners State of the Cloud).

EV/EBITDA

The workhorse multiple for mature businesses. EV / LTM EBITDA is the headline; EV / NTM EBITDA gets the most attention from buy-side investors because it reflects forward earnings power. The Lincoln International Private Market Index publishes quarterly multiples that benchmark middle-market private deals against public peers; the Q1 2025 reading showed public EV/EBITDA at 10.8x median versus private at 7.5x (Lincoln Private Market Index).

EV/EBIT

Sometimes called EV/Operating Income. Used in capital-intensive industries (cement, steel, telecom infrastructure) where depreciation policy varies enough that EBITDA becomes misleading. A 10x EV/EBITDA business with $50M of D&A is a 16.7x EV/EBIT business if the D&A is real maintenance capex.

P/E (Price to Earnings)

Equity multiple, not an enterprise multiple. Share price / diluted EPS. The single most cited multiple in equity research, but the least useful for M&A because it folds capital structure into the denominator. A highly levered company will have a low P/E for the same EV/EBITDA, which makes apples-to-apples cross-peer comparison hard. Useful for sanity-checking the equity research target prices that consensus expects (S&P Dow Jones Indices).

PEG (P/E to Growth)

P/E divided by the projected EPS growth rate. A PEG of 1.0 is considered fair value in the Lynch framework; below 1.0 is cheap, above is expensive. Useful as a tiebreaker between two peers with similar P/E but different growth (SEC Investor.gov).

Step 5: Calendarize

Public companies have different fiscal year-ends. Apple closes its fiscal year in late September; Microsoft in late June; Costco in late August; most US industrials in late December. If you pull “LTM revenue” from Capital IQ for a peer set spanning these year-ends without calendarizing, you are comparing different time periods. In a rapidly growing or rapidly declining market, that mis-comparison is enough to move the median multiple by 10-20%.

The Calendarization Formula

To calendarize a peer with a non-December fiscal year-end to a December year-end, you stub-period the financials:

Calendar-year revenue = Most recent annual revenue + (Most recent stub period revenue) – (Prior-year same stub period revenue)

Worked example: Apple’s FY2024 ended September 28, 2024. To calendarize to December 31, 2024:

FY2024 revenue (Oct 2023 – Sep 2024): $391.0B
Add: Q1 FY2025 revenue (Oct-Dec 2024): $124.3B
Subtract: Q1 FY2024 revenue (Oct-Dec 2023): $119.6B
Calendar 2024 revenue: $391.0B + $124.3B – $119.6B = $395.7B

Apple actual figures from the 10-K filed November 2024 and 10-Q filed January 2025 (Apple SEC Filings). This same logic applies to EBITDA, EBIT, and EPS calendarization, applied line by line.

Step 6: Scrub for One-Time Items and Non-Recurring Adjustments

Reported GAAP EBITDA is almost never the “real” EBITDA a buyer would underwrite. The scrub list is long and industry-specific, but the common adjustments are:

Restructuring charges: Severance, facility closures, asset impairments. Add back if non-recurring; leave in if the company restructures every year.
Stock-based compensation (SBC): The controversial one. GAAP excludes SBC from EBITDA. Many SaaS companies present “adjusted EBITDA” that adds back SBC, inflating margins meaningfully. The Warren Buffett school (and most credit investors) treats SBC as a real cost. Document your treatment and apply it consistently across all peers.
Litigation settlements: Add back if one-time; many financial companies have recurring litigation accruals that are part of doing business.
M&A transaction costs: Add back legal, banker, and integration fees. These are deal-specific.
Goodwill impairments: Non-cash; always added back in EBITDA but should be discussed in the comp footnote because it signals overpaying for past M&A.
FX gains and losses: Add back if they sit above the EBITDA line under the company’s reporting convention.
Divested business results: Subtract the contribution of any business unit sold during the LTM period.

The AICPA Audit and Accounting Guide on Revenue Recognition and the SEC’s Compliance and Disclosure Interpretations on non-GAAP measures both provide the regulatory boundary on what can and cannot be presented as “adjusted EBITDA” in a public filing (SEC Non-GAAP C&DIs).

Step 7: Build the Output Table

The final comps table that goes in a pitch book or fairness opinion has a standardized format. Banker convention is rows = companies, columns = multiples + operating metrics. The bottom rows show high, mean, median, low, and the target company itself (boxed and bolded).

Company	Ticker	Market Cap	EV	LTM Rev	EV/Rev	EV/EBITDA	EV/EBIT	P/E	Rev Growth	EBITDA Margin
Peer A	PXA	$4,250M	$4,800M	$1,200M	4.0x	14.5x	17.2x	22.5x	14%	27.6%
Peer B	PXB	$2,100M	$2,400M	$650M	3.7x	13.8x	16.1x	21.0x	11%	26.7%
Peer C	PXC	$8,900M	$10,200M	$2,800M	3.6x	12.1x	15.0x	19.4x	9%	30.1%
Peer D	PXD	$1,500M	$1,650M	$420M	3.9x	15.2x	18.0x	23.1x	16%	25.8%
Peer E	PXE	$6,200M	$7,100M	$1,950M	3.6x	13.0x	15.8x	20.2x	10%	27.9%
Peer F	PXF	$3,300M	$3,750M	$985M	3.8x	14.0x	16.5x	21.7x	12%	27.4%
High					4.0x	15.2x	18.0x	23.1x	16%	30.1%
75th pct					3.9x	14.4x	17.0x	22.3x	14%	27.9%
Mean					3.8x	13.8x	16.4x	21.3x	12%	27.6%
Median					3.8x	13.9x	16.3x	21.3x	11.5%	27.5%
25th pct					3.7x	13.2x	15.9x	20.4x	10%	26.9%
Low					3.6x	12.1x	15.0x	19.4x	9%	25.8%
Target	TGT	$2,750M	$3,100M	$780M	4.0x	14.2x	16.8x	22.0x	13%	28.0%

The Excel functions to populate the summary stats are PERCENTILE.INC(range, 0.25), MEDIAN(range), AVERAGE(range), MAX(range), MIN(range). Most bulge-bracket templates also include a Z-score column flagging any peer that is more than 2 standard deviations from the mean as a potential outlier to be excluded.

Steps 8 and 9: Choose LTM vs NTM and Apply to the Target

Step 8: LTM vs NTM

Two flavors of every multiple: trailing (LTM, last twelve months) and forward (NTM, next twelve months, or FY+1 and FY+2). LTM is fact-based, NTM is consensus-based. The choice depends on the deal context.

Context	LTM or NTM	Why
Sell-side pitch to client	NTM	Forward multiples typically lower, sets up the “your business deserves more” narrative
Buy-side fairness opinion	LTM + NTM	Show both, board needs the full picture
Cyclical business at trough	NTM and FY+2	LTM understates earnings power at trough
Cyclical business at peak	LTM only, or normalized	NTM consensus often anchors to peak; LTM more conservative
Hyper-growth (SaaS)	NTM Revenue + NTM ARR	LTM is stale for a 50%+ grower
Distressed / restructuring	LTM, then forward case	Consensus often hasn’t reset; use management’s forward case separately

Step 9: Apply to the Target (Imply Equity Value)

Once you have median EV/LTM EBITDA and EV/NTM EBITDA from the peer set, multiply by the target’s LTM EBITDA and NTM EBITDA to imply a target EV range. Then bridge back to equity value by subtracting net debt.

Worked example: target LTM EBITDA = $55M, NTM EBITDA = $62M. Peer median EV/LTM EBITDA = 13.9x, EV/NTM EBITDA = 12.5x. Target net debt = $80M.

Implied EV (LTM): $55M x 13.9 = $764.5M
Implied EV (NTM): $62M x 12.5 = $775.0M
Take the lower-quartile to upper-quartile range for the football field: $720M to $815M
Bridge to equity: $720M – $80M = $640M; $815M – $80M = $735M
Per-share, divide by diluted shares. If target has 25M diluted shares, range = $25.60 – $29.40 per share.

This implied equity range gets layered into the football-field chart alongside DCF, precedent transactions, and (if relevant) LBO floor analysis. The business valuation formula, methods, and math guide walks through how the four methods get reconciled on a single page.

Step 10: Trading Comps vs Transaction Comps

Public comps measure what the market pays for a public company on a normal trading day. Transaction comps (also called precedent transactions or M&A comps) measure what an acquirer paid in a control transaction, which by definition includes a control premium. The two should never be conflated.

Dimension	Public Trading Comps	Precedent Transaction Comps
Reflects control premium	No (minority stake)	Yes (typically 20-40% above trading)
Data freshness	Real-time market	Stale (months to years old)
Strategic synergies	Not included	Included in buyer’s price
Sample size	6-12 active peers	5-15 deals over 2-5 year window
Use case	Where does it trade today	What would a buyer pay

The Wachtell, Lipton, Rosen and Katz takeover compendium discusses how trading comps establish the floor for board negotiations and transaction comps establish the strategic-buyer ceiling (Wachtell, Lipton, Rosen and Katz). The control premium gap between the two is one of the most-debated topics in any fairness opinion, with empirical studies from the Deloitte Recap database showing a 28% median control premium across US public-target deals over the 2019-2024 window (Deloitte Recap M&A Database).

How Multiples Differ by Sector

There is no universal “right” EV/EBITDA. Sector economics drive the multiple. The McKinsey Valuation: Measuring and Managing the Value of Companies textbook, now in its 7th edition, devotes three chapters to how sector multiples should be interpreted (McKinsey Strategy and Corporate Finance).

Sector	Headline Multiple	Typical Range (2024-2026)	Why That Multiple
SaaS / Cloud Software	EV/NTM Revenue	4x-12x	Pre-profit or rule-of-40 driven
Mature Software	EV/EBITDA	14x-22x	High margin, sticky revenue
Consumer Staples	EV/EBITDA + P/E	11x-16x EBITDA, 18x-24x P/E	Defensive, low growth, dividend-heavy
Industrials	EV/EBITDA	8x-13x	Cyclical, capital-intensive
REITs	P/AFFO or P/FFO	15x-22x AFFO	GAAP earnings distort due to depreciation
Banks	P/Tangible Book Value + P/E	1.2x-2.0x TBV, 10x-14x P/E	Balance sheet driven, regulatory capital
Insurance	P/Book Value + P/E	1.0x-1.8x BV, 9x-14x P/E	Float economics, reserve adequacy
Biotech (pre-commercial)	Risk-adjusted DCF, not comps	N/A	No revenue, binary outcomes
Media and Entertainment	EV/EBITDA	9x-14x	Cord-cutting transition, mixed growth
Healthcare Services	EV/EBITDA	10x-15x	Reimbursement risk + roll-up dynamics
Energy (E&P)	EV/EBITDAX	3x-6x	Commodity-price exposed
Energy (Midstream)	EV/EBITDA + Distribution Yield	9x-12x	Fee-based, MLP structure

Sector-Specific Operating KPIs You Will Be Asked About

Beyond the standard EV/EBITDA, every sector has a banker-favorite operating KPI that the comp table needs to surface. Skipping these will get the table sent back.

SaaS: ARR growth, net revenue retention (NRR), CAC payback, rule of 40. Bessemer’s State of the Cloud 2025 report defined top-quartile cloud companies as 30%+ growth with NRR above 115%.
Marketplace / Consumer Internet: GMV growth, take rate, MAU growth, ARPU. Pulled from supplemental disclosures (Uber, Airbnb, DoorDash all disclose these quarterly).
Banks: Net interest margin (NIM), efficiency ratio, return on tangible common equity (ROTCE), CET1 ratio. The Basel III framework defines these metrics (Bank for International Settlements Basel III).
Insurance: Combined ratio, loss ratio, expense ratio, return on equity. The NAIC publishes industry benchmarks quarterly (NAIC).
REITs: Funds from operations (FFO), adjusted FFO (AFFO), implied cap rate, same-store NOI growth. NAREIT’s industry tracker publishes the standard definitions (Nareit).
Healthcare Services: Same-store census, EBITDA per location, payor mix (Medicare/Medicaid/Commercial), case-mix index for hospitals.
Industrials: Book-to-bill ratio, backlog, organic revenue growth, free cash flow conversion.
Energy: Production volume (boe/day), reserve life index, finding-and-development cost per boe, hedge book.

Common Comps Mistakes That Get Junior Bankers in Trouble

Every analyst program at Goldman, Morgan Stanley, JP Morgan, Lazard, Houlihan Lokey, and Lincoln teaches the same comps mistakes to avoid. The list has not changed in 20 years (Lazard Insights).

Pulling Capital IQ at face value: The standardized financials are a starting point, not a finished product. Always tie back to the 10-Q.
Forgetting operating leases post-ASC 842: Pre-2019 comp tables ignore lease debt. Post-2019 they should not.
Mixing GAAP and adjusted EBITDA across peers: If you add back SBC for peer A, do it for peer F.
Not calendarizing: Especially fatal in a moving market.
Including a peer that just announced a deal: The stock price has the bid premium baked in; the comp is no longer “trading” comp.
Letting outliers drive the median: A 25x EBITDA peer in a 13x median set should be flagged and discussed, not silently included.
Using stale forward estimates: Consensus refreshes after every earnings call. A pre-earnings NTM multiple may not match a post-earnings one.
Not adjusting for IPO float dynamics: A company that IPO’d six months ago may have a low float and an inflated trading multiple. Use 30-day VWAP rather than closing price for IPO peers.
Treating equity multiples as enterprise multiples: P/E and PEG are equity multiples and cannot be mixed in a table that otherwise uses EV-based metrics without flagging the inconsistency.
Missing dual-class share structures: Companies like Alphabet, Meta, and Snap have multiple share classes with different voting rights but the same economic interest; sum all classes for diluted equity value.

When Public Comps Do Not Work (and What Replaces Them)

Public comps fail when (a) there are no good public peers, (b) the target is structurally different from any public peer, or (c) the market is dislocated (a 2008 or 2020 type event).

Private SaaS targeting $5M ARR: No public peer is that small; use SaaS Capital Index private benchmarks or recent venture financing comps.
Vertical-specific industrial business: If you find only 1-2 public peers, supplement with European or Asian peers (FX-adjusted) and lean harder on precedent transactions.
Restructuring or distressed: Public multiples are stale; lean on liquidation value, sum-of-the-parts, or recovery analysis.
Disrupted incumbent (Kodak in 2010, Blockbuster in 2008): Historical multiples are no guide; use forward DCF with terminal value haircut.
Pre-revenue biotech or deep tech: Risk-adjusted DCF or real options analysis replaces comps entirely.

In each of these cases, the football-field chart will show a wider valuation range because the methodologies that work disagree more. That is honest valuation. Forcing public comps into a deal where they do not work is the most common cause of valuation overrun reported in Tax Court 280G excise tax cases involving golden parachute payments (see also the golden parachute 280G guide).

Who Uses Public Comps and How

How Public Comps Feed Into the Sell-Side and Buy-Side Process

Public comps are not a one-time pull. They get updated continuously through a transaction process and at multiple decision gates.

Process Stage	Who Updates Comps	Frequency	Purpose
Pre-mandate pitch	Sell-side analyst	Once	Establish valuation range to win the mandate
Kickoff after mandate	Associate + analyst	Weekly	Refresh for IM and management presentation
Buyer outreach	Associate	Weekly	Update bid range guidance for each round
First-round bids	VP + associate	Continuous	Reconcile bids to comp-derived expected range
Final bids	MD + VP	Daily	Negotiate final price against current trading
Fairness opinion	FO committee	Once at signing	Defensible record for board minutes
Closing	Junior banker	Once	Final comp snapshot for the bible

The work is most intense in the first-round and final-round bid stages because the trading multiples can move 5-10% in a quarter and a stale comp can result in advising a board to accept too low a price. The Delaware Chancery Court has been clear in cases like the 2024 Mindbody appraisal decision that boards relying on stale or methodologically weak comps face heightened fiduciary scrutiny (Delaware Court of Chancery).

How Public Comps Differ From DCF and Precedent Transactions

Method	What It Tells You	Strength	Weakness
Public Comps	What the market pays today for similar businesses	Market-based, observable, current	No control premium, peer set quality drives result
Precedent Transactions	What acquirers have paid in control deals	Includes control premium, real cash paid	Stale, deal-specific synergies, often limited disclosure
Discounted Cash Flow	Intrinsic value based on future cash flows	Forward-looking, fundamental	Assumption-sensitive, no market check
LBO Analysis	What a financial sponsor could pay to hit IRR target	Sets a floor in competitive auctions	Only relevant when sponsor is a likely buyer

The four methods are complements, not substitutes. The discounted cash flow business valuation and DCF valuation business sale 2026 guides go deep on the DCF side; the paper LBO example walkthrough covers the LBO floor; and the how to determine the value of a business guide is the executive-level synthesis.

How a Buy-Side Analyst Uses Public Comps Differently

Sell-side bankers use comps to pitch a value; buy-side analysts at private equity firms and hedge funds use comps to test whether the market is mispricing a stock. The mechanics are identical but the conclusion is opposite. Where a banker says “the median is 13.9x, your business deserves 15x,” a hedge fund analyst says “my target trades at 11x, the median is 13.9x, there is 25% upside if the gap closes.”

The private equity analyst career guide covers how PE associates use trading comps as the entry-point screen for take-private opportunities. Apollo, KKR, Blackstone, Bain Capital, and TPG all run quantitative screens against the Russell 3000 looking for companies trading at 30%+ discounts to their sector median EV/EBITDA, then have associates dig into the bottom-quartile names to find the ones where the discount is unjustified (and therefore an LBO opportunity). The Pitchbook 2025 PE outlook noted that take-private deals as a share of US PE deal volume rose from 18% in 2023 to 26% in 2024, with valuation-arbitrage being the most-cited driver (PitchBook Reports).

How M&A Advisors Position Public Comps in a Pitch

An M&A advisor pitching a private business owner uses public comps to anchor the valuation conversation around objective data the owner can verify on Yahoo Finance. The structure of the pitch is consistent across boutique and bulge-bracket firms:

Identify 6-10 publicly traded comparables.
Show their current trading multiples (EV/Rev, EV/EBITDA, NTM Rev growth, NTM EBITDA margin).
Apply a private-company discount of 15-30% (also called the marketability discount) to derive a private trading range.
Add a strategic premium of 10-25% for the synergies a buyer would realize.
Present the range as the expected outcome from a competitive auction process.

The marketability discount has been the subject of dozens of Tax Court opinions; the Mandelbaum v. Commissioner (T.C. Memo 1995-255) decision laid out the 10-factor framework still cited today by valuation firms in family limited partnership and gift-tax cases (US Tax Court).

TLDR: Public Comps Takeaways

Public comps benchmark a target against 6-12 publicly traded peers using market multiples (EV/Revenue, EV/EBITDA, EV/EBIT, P/E).
Peer selection is the most important step: filter by GICS code, business model, size band, geography, growth, and margin to get to a tight set.
Always calculate enterprise value as Equity + Debt + Preferred + Minority Interest + Operating Leases – Cash, with pension and other debt-like items added consistently.
Calendarize peers to a common fiscal year-end so you are not comparing different time periods.
Scrub one-time items: restructuring, SBC (document your treatment), litigation, M&A costs, FX gains, goodwill impairments, divested businesses.
Output table: 6-12 peers as rows, multiples and operating metrics as columns, summary stats (high, 75th, mean, median, 25th, low) at the bottom, target row boxed.
LTM = fact-based, NTM = consensus-based. Use both for fairness opinions; pick the one that supports your narrative for sell-side pitches.
Public trading comps measure minority-interest pricing; transaction comps add the 20-40% control premium.
Multiples are sector-specific: SaaS uses EV/Revenue, REITs use P/AFFO, banks use P/Tangible Book, biotech does not use comps at all.
Public comps fail for unique businesses, distressed situations, or dislocated markets; supplement with precedent transactions, DCF, and LBO floor analysis.
The football-field chart that synthesizes public comps with DCF, precedents, and LBO is the deliverable that goes in every fairness opinion and pitch book.

Public comps are the market-based discipline that keeps every other valuation method honest. Master the peer screen, the EV bridge, the calendarization, and the scrub, and you can build a defensible trading multiples table for any business in any sector.

Public Comps: 2026 Trading Comparables Methodology for M&A Bankers