How to Choose Comps for Stock Valuation
Choosing the right comparable companies is part science, part judgment—and the discipline you bring to this choice directly impacts the accuracy of your valuation. Too many investors grab the first five peers listed in a database and calculate an average multiple without asking whether those peers are truly comparable. A company might share an industry classification with a competitor yet differ wildly in growth, profitability, capital intensity, and risk. Those differences matter enormously. This chapter teaches you how to identify true comps—companies that reflect the fundamental economic drivers of your target stock—and how to use their multiples defensibly.
Quick Definition
Comparable company analysis values a target company by finding similar businesses trading in the market and applying their valuation multiples to the target's financial metrics. If five peers trade at an average price-to-earnings (P/E) ratio of 18x, and your target earns $2 per share, comparable company analysis suggests a fair value of $36 per share. The challenge lies in determining which companies are truly comparable and which multiples are most meaningful.
Key Takeaways
- Comparability rests on similar business fundamentals: growth rates, margins, capital intensity, and return on invested capital (ROIC)—not just industry classification.
- The "comps universe" should include 5-15 peer companies; too few introduces individual company bias, too many dilutes signal and includes poor matches.
- Primary multiples (EV/EBITDA, P/E, P/S) have different strengths; the choice depends on the target company's profitability, accounting quality, and stage of development.
- Adjustments for differences in growth, profitability, and leverage make comparisons more defensible but also more subjective.
- Market-wide multiple expansions and contractions can mislead; comparing your target's multiple to its own history and peer range is more robust than to an absolute standard.
- Sector-specific metrics (P/FCF for utilities, EV/subscribers for telecom, EV/barrels for energy) often reveal value more clearly than generic multiples.
The Foundation: What Makes Companies Truly Comparable?
The biggest mistake in comp analysis is conflating industry classification with economic similarity. Two companies in the "Software" sector might have radically different business models, capital requirements, and risk profiles.
Compare Microsoft to Zoom:
Microsoft: Diversified software and cloud services, 15+ billion in revenue, 30% operating margins, global enterprise customer base, high switching costs, growing 10-12% annually.
Zoom: Single-product communication platform, 4+ billion in revenue, 15% operating margins, high customer churn, newer competitive threats, growing 20%+ annually.
Both are software companies. Both trade on the Nasdaq. But they should not be valued identically. Microsoft trades at 25x earnings; Zoom trades at 15x earnings. The multiple difference reflects genuine economic differences.
True comparability requires alignment on:
1. Growth Rate
A company growing at 25% annually should trade at higher multiples than one growing at 5%. This is the single most important driver of valuation multiples. If your target grows at 15% and your comp universe averages 8% growth, adjusting the comps' multiples downward is essential.
To normalize for growth, many analysts use the PEG ratio (P/E divided by growth rate). A 20x P/E on a 20% growth stock (PEG = 1.0) may be cheaper than 18x P/E on a 10% growth stock (PEG = 1.8).
2. Profitability and Margins
High-margin businesses justify higher multiples. A software company with 40% EBITDA margins commands different pricing than a consulting firm with 15% margins, even at the same growth rate. Compare:
- Operating margin (operating income / revenue)
- EBITDA margin (EBITDA / revenue)
- Net profit margin (net income / revenue)
If a comp has superior margins due to scale or moat, and your target has similar fundamentals, using that comp's multiple may overvalue your target.
3. Capital Intensity
Businesses requiring heavy ongoing capital investment generate less free cash flow for equivalent earnings. A software company (capital-light) with $1B in earnings generates far more free cash flow than a manufacturer (capital-heavy) with the same earnings. Investors should pay more for capital-light businesses.
Use EV/FCF instead of P/E, or adjust comps for capital intensity differences.
4. Return on Invested Capital (ROIC)
ROIC reveals how efficiently a company deploys capital. A company with 25% ROIC (earning $0.25 for every dollar invested) deserves higher multiples than one with 10% ROIC. When selecting comps, ensure their ROIC aligns with the target's.
5. Risk and Stability
Higher-risk companies (startups, cyclicals, those with concentrated customers or products) warrant lower multiples. A utility with stable cash flows trades at lower multiples than a biotech with blockbuster drug prospects—but the biotech's valuation is riskier.
Stability metrics include:
- Revenue volatility (how much does revenue swing annually?)
- Customer concentration (what % of revenue comes from top 10 customers?)
- Competitive position (is there a durable moat?)
- Leverage (high debt increases financial risk)
Building Your Comps Universe
Step 1: Start with Industry and Direct Competitors
Begin with companies in the same industry classification. For a cloud software company, look at other enterprise SaaS providers. For a bank, look at regional and national banks of similar scale. Industry databases (Bloomberg, CapitalIQ, Yahoo Finance) allow you to filter by sector and industry.
Typical comps universes include:
- Direct competitors (companies selling similar products to similar customers)
- Broader sector players (larger or smaller peers serving the same market)
- Adjacent players (companies with related business models or customer bases)
For most analyses, 8-12 comps is ideal. Fewer than 5 introduces individual company bias; more than 15 introduces too much noise and includes poor matches.
Step 2: Eliminate Mismatches
Remove companies that fail the comparability test:
- Outlier growth: If one comp is growing 50% while others grow 10%, exclude it unless your target is also growing 50%.
- Vastly different scale: If your target has $500M in revenue and one comp has $50B, the scale difference affects margins, leverage, and multiples.
- Different business models: A software company that licenses software shouldn't be compared to one that sells consulting services built on the same platform.
- Different capital structure: If one comp is highly leveraged and others are not, EV/EBITDA may be more appropriate than P/E (since leverage affects net income differently).
- Different profitability: If one comp is unprofitable while others are highly profitable, exclude it unless valuing an unprofitable target.
Step 3: Check for Seasonality and Cyclicality
Cycle matters. If you're valuing a retailer, are you using comps in peak holiday season or off-season? If you're valuing a cyclical (energy, housing), use through-the-cycle multiples or normalizing adjustments.
For cyclicals, use:
- Normalized earnings: Average earnings over a full cycle, not trough or peak
- EV/EBITDA: Often less distorted than P/E for cyclicals
- Adjusted metrics: Some analysts adjust out one-time items or use run-rate metrics
Step 4: Organize Multiples by Financial Metric
Create a table:
| Company | Growth % | EBITDA Margin | EV/EBITDA | EV/Revenue | P/E | ROIC % |
|---|---|---|---|---|---|---|
| Target | 15% | 25% | ? | 8.5x | 20x | 18% |
| Comp A | 18% | 26% | 14.2x | 8.8x | 22x | 19% |
| Comp B | 12% | 23% | 12.5x | 7.2x | 18x | 16% |
| Comp C | 20% | 28% | 15.8x | 9.1x | 24x | 21% |
| Comp D | 14% | 24% | 13.1x | 7.9x | 19x | 17% |
| Average | 15.8% | 25.3% | 13.9x | 8.3x | 20.8x | 18.3% |
This reveals where your target fits relative to peers. If your target's growth (15%) is at the average but margins (25%) are at the low end, using the median multiple might overvalue it.
Flowchart
Which Multiples to Use and When
EV/EBITDA
- Best for: Cyclicals, companies with high debt, cross-border comparisons, capital-intensive businesses
- Why: EBITDA is less distorted by capital structure, depreciation, and tax differences
- Weakness: Ignores reinvestment needs and doesn't reflect free cash flow
Price-to-Earnings (P/E)
- Best for: Profitable, stable, low-debt companies; narrow comparisons within a sector
- Why: Most intuitive; available for all profitable companies; reflects bottom-line profitability
- Weakness: Distorted by capital structure and accounting choices; excludes unprofitable companies
EV/Revenue (Price-to-Sales)
- Best for: Loss-making companies, distressed businesses, comparing across accounting regimes
- Why: Works even if company is unprofitable; hard to manipulate through accounting
- Weakness: Ignores profitability; high-margin and low-margin businesses may trade at different P/S ratios
Price-to-Free Cash Flow (P/FCF)
- Best for: Capital-intensive businesses, comparing true economic returns
- Why: FCF is what actually flows to shareholders; most theoretically sound
- Weakness: More volatile; hard to calculate accurately; requires detailed capex forecasting
Sector-Specific Multiples
- Telecom: EV/subscribers (customers), EV/minutes (usage)
- Utilities: P/E, EV/EBITDA
- Energy: EV/barrels per day, EV/reserves
- Retail: EV/square footage, P/comp sales growth
- Banks: P/B (price-to-book), P/E
Real-World Examples
Valuing Nvidia: Navigating the AI Boom
In 2023, as Nvidia's AI chips became essential for large language model training, the company's valuation multiple expanded dramatically. Choosing comps required care. Nvidia had:
- 125% growth rate in 2023
- 50%+ gross margins
- Pure-play exposure to AI infrastructure
Comparing Nvidia to traditional chip manufacturers (Intel, AMD) with 10-15% growth and 40% margins would massively undervalue it. Comparing it to software companies (Microsoft at 10% growth) was closer but still not perfect. Sophisticated investors created a smaller universe of 3-4 comps with 50%+ growth and high margins (like Adobe at 10-15% growth), then adjusted upward for Nvidia's superior growth and margins.
Result: Nvidia traded at 50-60x forward earnings, far above the 20-25x average for the broader market—but justified given the comp analysis showing its economic superiority.
Valuing a Regional Bank Acquisition
When JPMorgan Chase acquired First Republic Bank in 2023, valuers had to identify comps quickly. Key considerations:
- Size: First Republic was mid-sized ($200B in assets); comps included similar-sized banks (Comerica, Zions), not all banks.
- Business model: First Republic served high-net-worth clients; comparable banks included Umpqua and others focused on affluent clients.
- Capital adequacy: First Republic faced capital pressure; comparing it to well-capitalized peers would overvalue it.
- Market conditions: The deal occurred during a banking crisis, when credit comps traded at depressed multiples; using median 2022 multiples would have been misleading.
Adjustments were essential. The acquiring bank paid a 25% premium to fair value based on comparable banks—partially because First Republic's deposit base was valuable even if the balance sheet was troubled.
Valuing a Loss-Making SaaS Company
A 5-year-old SaaS company generates $50M in revenue but is still unprofitable, burning $10M annually. P/E and P/FCF multiples don't exist. Valuation requires:
- Use EV/Revenue: Look at SaaS comps trading at 5-8x revenue (if growing 30%+)
- Adjust for growth: If target grows 40% and comps grow 25%, apply a 10-15% premium
- Adjust for profitability path: If target is approaching breakeven and comps are already profitable, narrow the discount
- Scenario weight: Use probability weighting—60% chance the target reaches profitability, 40% chance it runs out of capital
Result: Apply 6x revenue multiple (lower than 8x due to burn and profitability gap) = $300M valuation.
Common Mistakes
1. Selecting Comps Based on Size Alone
The largest companies in an industry are not always the best comps. A startup with $100M revenue and 50% growth has more in common economically with other growth-stage companies than with the $50B market leader.
2. Using Outdated Multiples
Multiples change with economic conditions, interest rates, and sector sentiment. Peer multiples from six months ago may be obsolete. Update multiples monthly, especially in volatile sectors.
3. Ignoring Leverage Differences
A highly leveraged comp has lower net income and higher P/E than an unlevered peer with the same EBITDA. Use EV/EBITDA to normalize, or adjust P/E comparisons for leverage.
4. Cherry-Picking Comps
Selecting comps that produce a valuation outcome you want is confirmation bias dressed up as analysis. Establish the universe first, then apply disciplined exclusion criteria. If you're tempted to exclude a comp because it doesn't fit your thesis, you've probably excluded it for the right reasons—but document them.
5. Trusting Average Multiples Without Investigating Outliers
If 12 comps trade at 16x P/E and one trades at 8x, don't just average to 15x. Why is that one comp cheaper? Is it:
- Slower growth (exclude or adjust)?
- Higher risk (include but note the risk premium)?
- Misvalued (include and note the opportunity)?
Outliers tell you something; investigate before averaging them away.
6. Forgetting to Adjust for Differences
The biggest edge comes from disciplined adjustments. If your target has superior growth, apply a growth premium. If it has inferior margins, apply a margin discount. Comps analysis is not just averaging—it's understanding why multiples differ and calibrating accordingly.
FAQ
Q: How many comps should I use?
Typically 8-12. Fewer than 5 introduces single-company bias. More than 15 dilutes signal and often includes poor matches. Use a median rather than mean to reduce outlier influence.
Q: Should I use trading comps or transaction comps (M&A)?
Both. Trading comps (public company multiples) reflect current market sentiment. Transaction comps (acquisition prices) include deal premiums and can reflect true intrinsic value better. Use transaction comps for mature companies, trading comps for growth companies.
Q: My target is unique with no perfect comps. What do I do?
Build a narrower universe of partial matches, apply larger adjustments, and back up conclusions with DCF. For truly unique companies (early-stage tech, novel business models), comp analysis should supplement, not replace, DCF.
Q: Should I weight comps differently?
Yes. If one comp is much closer to your target in size, growth, and profitability, give it higher weight (say 30% instead of 8%). But document your weighting rationale.
Q: What if trading multiples are historically high or low?
Acknowledge it and compare your target's multiple to both peer multiples and its own history. If the entire sector trades at 25x (elevated), but your target at 18x (below peer average), it's relatively cheap even if absolutely elevated.
Q: Can I use comps from other countries?
Yes, if risk profiles are similar. European software companies are reasonable comps for U.S. software companies. But adjust for currency, regulatory, and economic differences. Cross-border comps require larger adjustments.
Related Concepts
- Valuation Multiples Explained — Understand the mechanics of P/E, EV/EBITDA, and other commonly used multiples.
- Building Your DCF Model — Learn how to validate comps analysis conclusions against a bottom-up discounted cash flow valuation.
- Adjusting Multiples for Growth and Risk — Master the art of normalizing multiples when comps don't perfectly match your target.
- Detecting Valuation Traps — Recognize when comp-based valuations mislead and what patterns signal overvaluation.
Summary
Choosing comparable companies is the foundation of multiple-based valuation. True comparability rests on alignment in growth, profitability, capital intensity, and risk—not industry classification alone. Building a disciplined comps universe (8-12 companies), calculating relevant multiples, and understanding why multiples differ are skills that compound over time. The greatest value often comes not from mechanically averaging peer multiples but from understanding the economic drivers of those multiples and adjusting them defensibly when your target differs from peers. Combined with DCF analysis and historical context, comp-based valuation provides a powerful cross-check on intrinsic value and alerts you to relative mispricings across the market.
Next
Continue to Which Multiples for Which Sector? to learn the sector-specific metrics and adjustments that unlock valuation clarity in industries from technology to energy to healthcare.