Portfolio Lab

Valuation multiples: valuing companies with comparable firms

In depth11 min read · updated 18 November 2026 · 7 sources

Valuation multiples (relative valuation) estimate a company's value from the prices of similar firms: if comparable companies trade at a median of 10.8 times their EBITDA, that is a reference point for your own. It is faster than a DCF and anchored in the market, but it depends entirely on the choice of peer group and multiple. Here you work through the valuation of the fictitious Sample Corp. against nine comparable firms and learn which multiple suits which purpose and what research knows about accuracy.

What a multiple is

A multiple relates a price to a company metric. The best known is the P/E ratio: market capitalization divided by net income. Valuing companies with multiples assumes that similar firms cost about the same per dollar of earnings, revenue or book value. In practice the method is also called trading comparables or peer group analysis; comparing with prices paid in acquisitions is called transaction comparables.

What matters is that numerator and denominator match. Equity multiples such as P/E or price-to-book (P/B) relate market capitalization to figures that belong only to shareholders. Enterprise multiples such as EV/EBITDA, EV/EBIT or EV/sales take enterprise value (EV), i.e. market capitalization plus net debt, and divide it by figures before interest that belong to all providers of capital. Dividing market capitalization by EBITDA compares highly leveraged and debt-free firms on the wrong basis.

The formula: from multiple to value per share

usually the median of the comparable firms, such as EV/EBITDA
the corresponding figure of the company being valued for the same period, such as EBITDA of the last fiscal year
financial debt minus cash, plus leases, pensions and noncontrolling interests if they are included in the peers' EV

With the P/E ratio there is no bridge: it values equity directly. It is important that the same definitions apply to your own company and the peer group: the same period, the same treatment of leases and the same adjustments for one-off items.

Example: Sample Corp. against nine comparable firms

The fictitious Sample Corp. from the article DCF valuation had revenue of $1,000 million, EBITDA of $190 million, EBIT of $150 million and net income of $105 million in its last fiscal year. Net debt is $300 million, and there are 50 million shares at $30 each. The nine comparable firms A to I come from the same industry and size class; their multiples are fictitious too:

FirmEV/salesEV/EBITDAEV/EBITP/E
A0.97.910.813.5
B1.18.611.915.8
C1.39.413.117.2
D1.510.214.218.6
E1.610.814.919.5
F1.811.615.820.8
G2.012.417.022.4
H2.313.218.424.1
I (fast-growing)4.121.529.041.0
Median1.610.814.919.5
25% and 75%1.3 and 2.09.4 and 12.413.1 and 17.017.2 and 22.4
Sample Corp. today1.89.512.014.3

The other multiples paint a different picture. At the peer median, Sample Corp. would be worth $26.00 per share by EV/sales, $38.70 by EV/EBIT and $40.95 by P/E. The sales multiple undervalues it because, with an EBITDA margin of 19%, it earns more per dollar of revenue than a typical member of the group (1.6 / 10.8 ≈ 15%). The P/E gives the highest value, but the P/E also depends on leverage and interest, which differ between the firms. A DCF with the assumptions from the article DCF valuation gave $30.75, i.e. less than all median values except sales.

Value per share of Sample Corp. by peer multiple

Multiples of the nine fictitious comparable firms from the table, applied to Sample Corp.'s revenue, EBITDA, EBIT and net income; EV values minus $300 million of net debt, divided by 50 million shares.

Which multiple for what

MultipleSuitsWeaknesses
EV/salesyoung or temporarily unprofitable firms, comparisons across marginsignores differences in margins; research finds it the least accurate multiple
EV/EBITDAcapital-intensive industries and firms with different leverageoverlooks capex and depreciation; lease accounting under IFRS 16 raises EBITDA
EV/EBITfirms with similar depreciation policies and steady capexamortization from acquisitions and accounting choices distort it
P/Emature, profitable firms with similar financing; banks and insurersdepends on interest, taxes and one-off items; useless with losses
P/Bbanks, insurers, holding companiesintangible assets are often missing from the balance sheet

Forecast figures (forward, usually for the next twelve months) beat past figures (trailing) because the price values the future. Past figures, on the other hand, are reliable and audited. What matters is that the same period applies to your own company and the peer group.

Choosing the right peer group

The peer group should resemble the company being valued in what drives multiples: growth, return on invested capital and risk. In practice you start with the industry and narrow it down by size, region and business model. Size matters because small firms usually have higher costs of capital and trade less liquidly; Banz (1981) found that small companies earned higher returns than the CAPM would predict. Conversely, higher expected returns mean lower prices per dollar of earnings. These size classes by market capitalization are common:

Size classMarket capitalization in US dollars
Micro capunder 300 million
Small cap300 million to 2 billion
Mid cap2 to 10 billion
Large cap10 to 200 billion
Mega capover 200 billion

The thresholds are a widespread convention, not a standard; index providers draw them differently. A smaller but well-matched group is often better than a large, mixed one. Outliers like firm I, which grows fast, only belong in the group if the company being valued grows similarly.

Median, mean or harmonic mean

How you summarize the peer multiples changes the result noticeably. The mean of the nine EV/EBITDA values is 11.73 because firm I pulls it up; without it, it would be 10.51. The median of 10.8 is robust to such outliers. The harmonic mean (the count divided by the sum of the reciprocals) is 10.86 and gives high multiples less weight.

SummaryEV/EBITDAValue per share
Median10.8$35.04
Harmonic mean10.86$35.25
Mean11.73$38.59
Mean without firm I10.51$33.95

Because a multiple is a price per unit, many consider the harmonic mean the appropriate average: if all firms have the same enterprise value, it equals exactly the total value of the group divided by its total metric. The median and the harmonic mean are therefore often close together, with the simple mean above them.

Advanced: deriving multiples from a DCF

A multiple is a shortened DCF. If a company grows permanently at g and earns the return RONIC on new capital, the value driver formula from Koller et al. (2020) gives a justified multiple:

tax rate
long-term growth
return on newly invested capital
weighted average cost of capital

With 25% tax, 2% growth, 12% RONIC and an 8.5% WACC, the result is 0.75 × (1 − 0.02 / 0.12) / (0.085 − 0.02) = 9.6. At 3% growth and 20% RONIC it is 0.75 × (1 − 0.03 / 0.20) / (0.085 − 0.03) = 11.6. If the peer median is well above that, like the EV/EBIT of 14.9 in the example, the market is pricing in more growth, higher returns or lower costs of capital. This lets you check whether a multiple fits the company's characteristics.

  • Justified multiples: Bhojraj and Lee (2002) estimate a justified multiple for each firm from growth, profitability, risk and industry and choose as peers the firms with the closest value instead of only those from the same industry.
  • Regression instead of averages: According to Liu et al. (2002), regressing the price on the metric with an intercept offers some improvement over the simple ratio, especially for multiples that otherwise perform poorly.
  • Measurement date: Peer multiples reflect the share prices of one day. In turbulent markets, an average over a few weeks keeps a single price jump from driving the result.

What the research shows

  • Earnings before sales: Liu et al. (2002) compared a long list of multiples. Multiples based on forecast earnings explained stock prices best; for about half of the firms, the error was within 15% of the share price. They were followed by historical earnings, then cash flow and book value, and finally sales. This ranking held in almost all industries.
  • Assets, EBITDA and size: Lie and Lie (2002) found that the ratio of market value to book value of assets often estimates value more precisely than sales and earnings multiples, forecast earnings better than trailing ones, and EBITDA better than EBIT. How well a multiple performed depended strongly on size, profitability and the share of intangible value.
  • IPOs: According to Kim and Ritter (1999), the P/E, price-to-book and price-to-sales ratios of comparable firms predict the value of IPO candidates only modestly without further adjustment, because the ratios of young firms within an industry vary widely. Forecast earnings gave much more accurate values than trailing earnings.
  • Peer selection: Bhojraj and Lee (2002) showed that comparable firms chosen via justified multiples predicted future EV/sales and price-to-book ratios much better than groups chosen by other methods.
  • Multiples and DCF: In 51 highly leveraged transactions, DCF values were on average within 10% of market values and at least as close as valuations based on comparable companies and transactions (Kaplan and Ruback 1995).

Limits and common mistakes

  • Relative, not absolute: Multiples tell you whether a firm is expensive compared with its peer group, not whether the whole group is too expensive. In a bubble, every firm looks fairly valued when all of them are expensive; the Shiller CAPE measures the overall market.
  • Different accounting: IFRS and US GAAP, leases, capitalized development costs and amortization from acquisitions shift EBITDA and EBIT. Only compare adjusted figures with the same definitions.
  • Negative denominators: Losses or negative EBITDA produce meaningless multiples. Leave such firms out or switch to sales.
  • Cyclical peaks: At the top of the cycle, multiples on past earnings look cheap, at the bottom expensive. Use normalized earnings over a cycle.
  • Size and liquidity: A mega-cap peer group for a small firm usually overstates its value; conglomerates with many divisions rarely fit pure specialists.
  • Mean instead of median: A single outlier can shift the mean considerably, like firm I in the example.

This article explains a method; it is not investment advice and not a recommendation of individual stocks. Sample Corp., the comparable firms and all figures in the example are fictitious.

How Portfolio Lab calculates

The multiples calculator applies this scheme to real peer groups: it searches all companies listed in the US by region, size, sector and industry and shows EV/sales, EV/EBITDA, EV/EBIT, P/E and P/B with median and quartiles, plus the distribution and the median for each size class. The figures come from the companies' mandatory SEC filings and are refreshed every night; Europe follows. The default is Sample Corp. from this article in the group Industrials, Machinery, Mid. Because the group contains real companies, its median differs from the fictitious example above (methodology). Related analyses in the app:

  • DCF as a cross-check: The DCF calculator values the same company from its free cash flows; if the two values differ widely, it is worth looking for the reason.
  • Market valuation: The Valuation page shows the Shiller CAPE of the S&P 500 since 1881, i.e. the multiple of the overall market on ten-year real earnings (methodology).
  • Value and size as factors: The Factors page measures how strongly a fund or portfolio tilts toward cheaply valued (value) and small companies (size), following Fama and French (methodology).

Frequently asked questions

What is valuation with multiples in simple terms?

You value a company through the prices of similar firms: if comparable companies trade at a median of 10.8 times their EBITDA, Sample Corp. with $190 million of EBITDA gets an enterprise value of $2,052 million.

Which multiple is the best?

Studies put multiples based on forecast earnings first and the sales multiple last. Which one fits, however, depends on the company: EV/EBITDA with different leverage, P/E or P/B for banks, EV/sales for firms without profits.

What is a good EV/EBITDA?

There is no generally good value. The multiple depends on growth, return on invested capital, cost of capital and industry. What is meaningful is the comparison with similar firms and with the company's own history.

How do you choose the peer group?

Same industry, similar size, similar growth and similar profitability. Better a few well-matched firms than many poorly matched ones; check outliers and use the median instead of the mean.

Why use enterprise value instead of market capitalization?

Because EBITDA and EBIT belong to all providers of capital. Enterprise value therefore includes net debt alongside market capitalization; this makes firms with different financing comparable.

Multiples or DCF?

They complement each other. Multiples show how the market values similar firms today; the DCF shows which assumptions are behind a value. If the two differ widely, it is worth looking for the reason.

Literature

  1. Banz, Rolf W. (1981): The relationship between return and market value of common stocks. Journal of Financial Economics 9 (1), pp. 3–18. doi:10.1016/0304-405x(81)90018-0
  2. Bhojraj, Sanjeev; Lee, Charles M. C. (2002): Who Is My Peer? A Valuation-Based Approach to the Selection of Comparable Firms. Journal of Accounting Research 40 (2), pp. 407–439. doi:10.1111/1475-679X.00054
  3. Kaplan, Steven N.; Ruback, Richard S. (1995): The Valuation of Cash Flow Forecasts: An Empirical Analysis. The Journal of Finance 50 (4), pp. 1059–1093. doi:10.1111/j.1540-6261.1995.tb04050.x
  4. Kim, Moonchul; Ritter, Jay R. (1999): Valuing IPOs. Journal of Financial Economics 53 (3), pp. 409–437. doi:10.1016/S0304-405X(99)00027-6
  5. Koller, Tim; Goedhart, Marc; Wessels, David (2020): Valuation: Measuring and Managing the Value of Companies. Wiley, 7th edition (McKinsey & Company).
  6. Lie, Erik; Lie, Heidi J. (2002): Multiples Used to Estimate Corporate Value. Financial Analysts Journal 58 (2), pp. 44–54. doi:10.2469/faj.v58.n2.2522
  7. Liu, Jing; Nissim, Doron; Thomas, Jacob (2002): Equity Valuation Using Multiples. Journal of Accounting Research 40 (1), pp. 135–172. doi:10.1111/1475-679X.00042

Full bibliography

This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.