Benchmarking Companies Using Peer Distribution Analysis

Hello! In the previous lesson, you calculated growth rates, margins, and market shares with careful attention to periods and denominators. Those figures become more useful when you can answer a second question: How unusual is this company’s result among genuinely comparable businesses?

This lesson builds a compact peer benchmark using the median, quartiles, dispersion, and outlier checks. You will also see how to turn the numbers into a qualified finding for your first company brief. Allow about 40–45 minutes, including the resources and a small Excel schedule. All figures in our worked example are invented.

A benchmark begins with the peer group

Imagine that your subject company has an 18% operating margin. Is that strong? Comparing it with one competitor at 12% suggests yes, but that competitor might be unusually weak. A peer distribution gives you a more defensible reference point.

First decide which firms belong in it. Look for broadly comparable products, customers, geographies, and business models. Then check differences in scale, growth, and financial risk. “Same industry” is a starting point, not proof of comparability: two firms in the same sector may earn money in substantially different ways.

Watch two short passages from Kenji Explains’ Comparable Company Analysis Excel Walkthrough. Although the video ultimately builds a valuation, focus here on peer selection and summary statistics, not the final valuation calculation.

For your current research brief, benchmark an operating metric—such as revenue growth or operating margin—before worrying about trading multiples. Keep each column like-for-like: the same metric definition, a comparable reporting period, and a consistent consolidated or standalone basis. Record material differences you cannot resolve. Do not insert a zero for an undisclosed metric; zero is a result, whereas missing means you do not know the result.

Your subject company should be clearly separated from the comparison set in the first-pass benchmark. Otherwise, its own performance helps determine the standard against which you judge it. There can be reasons to include a publicly traded target in a market-based valuation set, but that choice must be explicit. The spreadsheet image below illustrates the layout—not a peer group or set of summary formulas you should adopt without checking.

Read the middle and the spread, not just the average

Suppose you identify six sufficiently comparable companies and calculate their operating margins on a consistent basis. In ascending order, they are 8%, 10%, 12%, 14%, 16%, and 38%. Your subject company’s margin is 18%; it is not one of the six peers.

The median splits the ordered peer observations into a lower and an upper half. With six observations, it is the average of the middle two:

The mean uses every observation:

One 38% observation pulls the mean well above the median. Neither statistic is automatically “correct”: the difference tells you to inspect the distribution.

Quartiles add information about its middle. The first quartile, $Q_1$, is the 25th percentile; the third quartile, $Q_3$, is the 75th percentile. Using Excel’s inclusive quartile method for this example gives:

For a small set, quartiles can fall between observed values. Their exact values also depend on the quartile convention, so label your method rather than treating the output as a natural boundary. Our subject company’s 18% margin is 5 percentage points above the peer median and above $Q_3$. That is more informative than saying it “beats peers”: it describes where it sits, without claiming why.

Dispersion describes how far apart the observations are. Two useful, differently sensitive measures are:

The IQR describes the width of the middle half; the full range responds strongly to extreme observations. Here, the narrow 5-point IQR alongside a 30-point full range is a reason to inspect the high-margin firm. It is not yet a reason to delete it. With only six peers, each firm has considerable influence, and these sample quartiles should not be presented as precise estimates of an entire industry.

Use an outlier rule to start an investigation

A common screening rule marks observations below $Q_1-1.5\times IQR$ or above $Q_3+1.5\times IQR$. For our margins, the lower fence is 3% and the upper fence is 23%. The 38% peer therefore receives an outlier flag. Our subject company’s 18% is above $Q_3$ but below the upper fence: above the peer middle and flagged as an outlier are different descriptions.

Steven Bradburn’s How To Create A Box Plot In Excel (Including Outliers) makes the quartiles and fences visible. Watch for the distinction between a point the rule flags and an observation an analyst can explain.

An outlier might reflect a data-entry error, a one-time gain, a different business mix, or a genuinely exceptional operating model. Inspect its filing and metric definition before deciding what to do. If you remove it, state why and show how the result changes. In this example, excluding the 38% firm leaves five margins—8%, 10%, 12%, 14%, and 16%—with a 12% median. The subject company remains above the peer median, but that robustness check does not make the exclusion automatically justified.

Aswath Damodaran’s Relative Valuation is useful here because it treats a benchmark as a description of a chosen distribution, not a verdict. Read the specified slides with your operating-margin example in mind; the slides often use valuation multiples, but their cautions about definitions, outliers, and comparability carry over.

A statistical flag answers “Which observation needs review?” It does not answer “Which observation is wrong?” That distinction matters in an interview: you should be able to explain both the rule and the business reason for keeping, qualifying, or excluding a peer.

Build a benchmark you can audit in Excel

Create a small sheet headed “Illustrative operating-margin benchmark.” Put Peer in A1 and Operating margin in B1. Enter six peer labels in A2:A7 and the values 8%, 10%, 12%, 14%, 16%, and 38% in B2:B7. Put Subject company in A8 and 18% in B8. The subject company must stay outside the peer ranges in the formulas below.

In D2:D11, add the labels shown; enter the corresponding formulas in column E:

Format B2:B8 and E3:E10 as percentages; format E11 as a number, not a percentage. Its result should be 5: the margin gap is 5 percentage points. Check that E2 is 6, E4 is 13%, E6 is 5% (meaning a 5-percentage-point width), and E8 is 23%.

To flag peers for review, put Outlier review in C1 and enter =IF(OR(B2<$E$7,B2>$E$8),"Review","") in C2. Copy it through C7. Only the 38% peer should show Review. The anchored fence references stay fixed as the formula moves down.

For a real case, replace the invented inputs with source-linked figures and add columns for period, metric definition, source, and comparability note. If a firm’s margin uses a materially different profit definition, do not silently blend it into the same range. Keep a count of valid observations: a benchmark based on four disclosed margins is not a benchmark based on all eight firms you initially considered.

Turn the table into a defensible finding

A useful sentence for a research brief connects position, spread, and uncertainty:

“The subject company’s latest operating margin was 18%, versus a 13% median across six selected peers. It exceeded the peers’ inclusive-method upper quartile of 15.5%. One peer reported 38%, above the IQR screening fence of 23%; its business mix and margin definition require review before treating it as directly comparable.”

That sentence reports calculations without asserting that the subject is better managed, that the 38% figure is wrong, or that the market should value either company in a particular way. Before publication, replace “selected peers” with a concise selection rule, specify the financial period, and cite the underlying figures. If peer reporting dates or business mixes differ materially, qualify the comparison.

This benchmark also guides your next research step. A company above the peer median could have stronger pricing, lower costs, a different product mix, or a temporary benefit. The peer table identifies the difference; filings and operating evidence help you investigate its cause.

Key takeaways

A credible peer benchmark requires comparable inputs and an explicit peer set before it requires formulas. The median describes the middle; quartiles locate a company within the group; the IQR and full range show different aspects of dispersion. An outlier rule flags observations for review, while a documented business judgment determines how they are treated.

Your Excel schedule now gives your first company brief a stronger comparison than a single hand-picked rival. Next, you will decide which chart communicates that comparison accurately, with appropriate scales, labels, and sources.