Sector comparisons: why identical multiples are not equal

A multiple compresses a business model

Two companies trading at the same earnings multiple can represent very different economic claims. One may have stable recurring demand and modest reinvestment needs; another may require large periodic investment and face volatile input costs. The multiple does not erase those differences. Its usefulness comes from connecting a market value with an economic denominator that is comparable across the selected businesses.

Start with a map of customers, revenue recognition, cost structure, capital intensity and financing. A bank's funding is part of its operating model, so subtracting deposits as though they were ordinary industrial debt can make an enterprise-value comparison meaningless. A property company may be better understood through asset values, rental economics and carefully defined cash measures. A software business may require attention to retention, acquisition costs and employee dilution.

Control the denominator

Accounting earnings can be affected by different depreciation schedules, lease treatment, acquisition amortization and tax positions. An EBITDA comparison removes some charges but does not remove the underlying need to replace assets or pay obligations. A capital-intensive business can appear inexpensive on EBITDA while producing little cash after necessary investment. Conversely, a current earnings loss can reflect spending that supports future revenue, though that spending is not guaranteed to succeed.

Check whether periods match. A forward estimate for one company and a trailing result for another can produce a spurious ranking. Fiscal years, seasonality, currencies and exceptional items also matter. Segment reporting can reveal that two companies assigned the same industry label have different economic mixes. A conglomerate with a fast-growing service division is not necessarily comparable to a pure manufacturer merely because both sell related products.

Distinguish cyclicality from structural quality

Cyclical earnings often look strongest when conditions are least sustainable. A low trailing multiple can therefore signal peak profitability rather than a bargain. Estimate a range of mid-cycle prices, volumes and margins, and show how the conclusion changes. Avoid replacing observed results with an optimistic normalized number that cannot be tied to capacity, costs and competitive behavior.

Structural advantages should also be tested, not asserted. Strong gross margins may reflect differentiated products, but they can also invite competition. High returns on invested capital can arise partly because past spending was expensed rather than capitalized. A comparison becomes more informative when it traces the source and durability of returns, the reinvestment opportunity and the probability that competition erodes them.

Build a comparison that teaches something

Select a small peer group with a written inclusion rule. Record what differs as carefully as what matches. Use several measures rather than creating a league table from one ratio: growth, operating margins, cash conversion, balance-sheet resilience and valuation can reveal different dimensions. Do not average ratios blindly when denominators are negative or close to zero, because those averages can become economically meaningless.

Finish with an explanation of the premium or discount that would be justified under explicit assumptions. A higher-quality company can still be overpriced, and a lower-quality business can already reflect substantial pessimism. The goal is not to declare the cheapest ticker the winner. It is to determine which expectations are embedded in the comparison and which evidence would change the ranking. That framework remains useful even when the correct conclusion is that the available peers are too different for a clean relative valuation.

A peer median is a description, not a fair-value law. If an entire sector shares an optimistic assumption about commodity prices or discount rates, relative valuation may identify the least expensive member of an expensive group. Include at least one absolute cash-flow or balance-sheet cross-check so that a common market assumption does not pass unnoticed through every comparison.

Write down why each peer belongs

A useful peer group begins with the economic question. If the question concerns replacement spending, group businesses with similar asset requirements. If it concerns customer retention, compare similar selling models and contract structures. The exchange's industry label may be convenient, but it does not guarantee that two companies face the same customer behavior or reinvestment needs.

Consider two fictional industrial companies. One sells equipment once and then competes for the next order; the other earns a substantial share of profit servicing an installed base. Their reported revenue growth could match while the durability of future cash flows differs. Combining them in one peer median without describing the service mix hides precisely the difference the comparison should investigate.

Write exclusions as well as inclusions. A company may be too small, too leveraged, or too dependent on a different region to provide a clean comparison. Excluding it solely because its multiple undermines the preferred conclusion is a different matter. A short rule written before examining valuations makes that selective reasoning easier to spot. Keep an imperfect peer in view when it provides a useful stress case, but label the reason it is imperfect.

Give the same sales shock two cost structures

Suppose two fictional businesses each make $100 million revenue and $20 million operating profit. Business A has $40 million variable costs and $40 million fixed costs. Business B has $70 million variable costs and $10 million fixed costs. In a simplified 10% sales decline, with variable costs falling proportionately and fixed costs unchanged, A's profit becomes $14 million while B's becomes $17 million. The initial margins were identical; the earnings sensitivity was not.

The exercise does not establish that B is the better investment. A may have advantages in an expansion, a stronger customer franchise or a much lower valuation. It shows why a peer multiple needs a view of how profits change when conditions change. Treating both companies' last year's earnings as equally durable would conceal an important difference.

Real cost behavior is less tidy. Staff, distribution and capacity spending can change in steps, and management can respond by cutting investment or changing prices. Those responses have consequences for later performance. Use the simple model to identify which assumptions deserve research, then add evidence about contractual flexibility, utilization and restructuring costs instead of multiplying every cost line by the same sales percentage.