Selecting Comparable Companies#

You've mastered P/E and EV/EBITDA. You understand when to use each multiple. But here's the uncomfortable truth: the most important part of relative valuation isn't the math—it's choosing the right companies to compare.

Why Peer Selection Matters So Much#

A Simple Example#

You're valuing CloudHR, a cloud-based HR software company. Consider three possible peer sets:

Same company. Same metric. 2x difference in valuation based purely on peer selection.

This isn't manipulation—each peer set has logic. But the choice dramatically affects the conclusion.

The Peer Selection Framework#

Step 1: Define the Target's Business Model#

Before searching for peers, deeply understand your target:

Step 2: Identify Initial Universe#

Start broad using industry classifications:

For CloudHR, initial universe might include 30+ HR tech and B2B SaaS companies.

Step 3: Apply Quantitative Screens#

Narrow the universe using objective criteria:

Step 4: Apply Qualitative Filters#

Numbers don't tell the whole story:

Step 5: Final Selection (4-8 companies)#

Your final peer set should be:

• Large enough for statistical reliability (minimum 4)
• Small enough that you know each company well (maximum 8-10)
• Defensible with clear logic for inclusion/exclusion

The Final CloudHR Peer Set#

After applying filters:

Final set: Paylocity, Paycom, Ceridian, TriNet (4 companies)

Common Peer Selection Mistakes#

1. Using Generic Industry Codes#

GICS code "Software" includes everything from Microsoft to a 50-person startup. Too broad to be useful.

Fix: Narrow by subsector, size, and business model.

2. Cherry-Picking to Support a View#

Want to show a stock is cheap? Select higher-multiple peers. Want to show it's expensive? Select lower-multiple peers.

Fix: Establish selection criteria before looking at valuations. Document the methodology. Would you defend these choices to a skeptic?

3. Ignoring Growth Differences#

Two companies at the same multiple but different growth rates are not equivalent.

Fix: Always include growth in your comparison. Use PEG or EV/EBITDA vs. EBITDA growth scatter plots.

4. Selecting Too Few Peers#

With only 2-3 peers, outliers dominate. One company's unusual situation skews the average.

Fix: Aim for 5-7 peers. If fewer exist, acknowledge the limitation.

5. Including Acquirable/Distressed Names#

A company being acquired trades at the deal price, not fundamental value. A distressed company trades at a discount for non-comparable reasons.

Fix: Exclude companies with pending M&A, restructuring, or severe distress.

Handling Imperfect Comparables#

No peer set is perfect. Here's how to address differences:

Method 1: Adjust for Growth#

If peers grow faster on average, your target may deserve a discount:

Target Growth: 18%
Peer Average Growth: 25%
Growth Gap: 7pp

If 1pp of growth = 0.5x EBITDA multiple:
Adjustment: -3.5x from peer average

Method 2: Create Premium/Discount Framework#

If peer average is 10x EV/EBITDA, target might warrant 6.5-7.5x.

Method 3: Regression Analysis#

For sophisticated analysis, regress multiples against key drivers:

EV/Revenue = a + b(Growth) + c(Gross Margin)

Then predict what multiple your target "should" have based on its growth and margin.

Presenting Comparable Analysis#

The "Football Field" Chart#

Show valuation range from multiple methods:

                    $5    $10    $15    $20    $25
                     |      |      |      |      |
Peer EV/EBITDA     |=====[=======]=====|
Peer EV/Revenue              |=====[====|
DCF Analysis            |========[====|
52-Week Range           |===========[===|
Analyst Targets              |==[====|

Current Price: $14

This visualization shows where different methods converge (likely fair value) and diverge (uncertainty).

The Summary Table#