Fundamentals · Valuation
EV/EBITDA vs. P/E: two valuation multiples that disagree
Compare EV/EBITDA and P/E on seven large stocks, see why Microsoft and Cisco share a P/E of 28 but not an EV/EBITDA, and learn which multiple fits which company.
On this page
The P/E ratio compares a company's market value with its net income, the profit left for shareholders after interest and tax. EV/EBITDA compares its enterprise value, which adds debt and subtracts cash, with EBITDA, the profit before interest, tax, depreciation and amortization. Both answer "how many years of profit am I paying for?", but they count different profit for a different owner.
Most of the time they rank companies the same way. Sometimes they do not. On November 17, 2026, Microsoft and Cisco had almost the same P/E, 28.1 and 28.3. On EV/EBITDA, Microsoft was 19.2 and Cisco 21.7. Walmart's P/E of 46.4 made it the most expensive stock on the list by far; its EV/EBITDA of 23.9 was still the highest, but only just ahead of Cisco. This page explains why the two multiples part ways, works them on seven companies, and shows which to use for which kind of business.
The two multiples compared
| P/E | EV/EBITDA | |
|---|---|---|
| Numerator | Market cap (or price per share) | Enterprise value |
| Denominator | Net income (or EPS) | EBITDA |
| Whose value and profit? | Shareholders' | All capital providers' |
| Affected by debt? | Yes | Designed not to be |
| Affected by tax rate and depreciation? | Yes | No |
| Works with a net loss? | No | Often (EBITDA can be positive) |
| Works for banks? | Yes | No (debt is their raw material) |
| Main blind spot | Financing and accounting choices | Capital spending, which EBITDA ignores |
The two are paired on purpose: market value belongs to shareholders, and so does net income; enterprise value belongs to lenders and shareholders together, and so does EBITDA. The full lessons are Price-to-earnings ratio (P/E): formula, examples and limits and EV/EBITDA: the capital-structure-neutral valuation multiple.
Seven companies, both multiples
Prices are the closes of November 17, 2026. Earnings, D&A, debt and cash are from each company's latest Form 10-K. P/E here is market cap ÷ net income, which matches price ÷ diluted EPS to within a few tenths.
| Company | P/E | EV/EBITDA | Net debt ÷ EBITDA |
|---|---|---|---|
| Microsoft | 28.1 | 19.2 | −0.19 |
| Cisco | 28.3 | 21.7 | 0.76 |
| Walmart | 46.4 | 23.9 | 0.77 |
| Procter & Gamble | 24.9 | 18.5 | 1.06 |
| Home Depot | 23.9 | 16.1 | 2.23 |
| Verizon | 11.3 | 7.0 | 2.92 |
| AT&T | 7.1 | 6.1 | 2.62 |
Why the two disagree
Depreciation. EBITDA adds it back; net income charges it. A company with large D&A relative to profit, such as Walmart or Microsoft (whose D&A was $38.5 billion), has an EV/EBITDA far below its P/E. Cisco's D&A, $2.5 billion, is small relative to its profit, so its two multiples sit closer together. That is most of the Microsoft-Cisco split: the same P/E, but Microsoft carries more depreciation from its data-center build-out, which EV/EBITDA does not charge.
Debt. P/E sees debt only through interest; EV adds the debt itself. Microsoft holds more cash than debt, which lowers its EV; Verizon and AT&T carry over $100 billion of net debt each, which raises theirs. On P/E the telecoms look very cheap (11.3 and 7.1). On EV/EBITDA they still look cheap (7.0 and 6.1), but by less, because their lenders' claim is counted.
Tax and one-off items. Net income moves with tax rates and with gains or charges below operating income; EBITDA does not. AT&T's low P/E of 7.1 partly reflects items below the operating line in 2025, which is why its P/E and EV/EBITDA are unusually close.
The blind spot of each
P/E punishes leverage and accounting choices that do not change the business. Two identical companies, one financed with debt, get different P/Es. It also fails for loss-makers.
EV/EBITDA treats capital spending as free. Microsoft spent $115.9 billion on property and equipment in fiscal 2026, almost three times its D&A; those assets will depreciate in coming years, and EBITDA ignores them until then. For capital-hungry companies, EV/EBIT, which charges depreciation, or price to free cash flow is safer (see EBITDA vs. operating income: what adding back D&A changes).
The bridge from P/E to EV/EBITDA, for one company
Both multiples can be rebuilt from the same filing, which shows exactly what separates them. Home Depot, fiscal year ended February 1, 2026, at the November 17, 2026 close:
The numerator grew by 16% and the denominator by 72%, so the multiple fell by about a third. For almost every profitable company the denominator grows more than the numerator, which is why EV/EBITDA sits below P/E. How much below depends on the company: debt widens the top, while tax, interest and depreciation widen the bottom. Two companies with the same P/E can therefore have very different EV/EBITDA ratios, and the difference tells you which of those four items is large.
Which to use when
| Situation | Better multiple | Why |
|---|---|---|
| Mature, lightly indebted, steady earners | P/E | Simple, widely quoted, close to EV-based measures |
| Peers with very different debt | EV/EBITDA | Neutral to financing |
| Companies in different tax regimes or with different depreciation policies | EV/EBITDA | Removes both |
| Companies with temporary net losses | EV/EBITDA (or EV/sales) | P/E has no meaning |
| Banks and insurers | P/E (and P/B) | Debt is their product; EV does not apply |
| Heavy capital spenders | EV/EBIT or price to free cash flow | EBITDA ignores reinvestment |
| Takeover pricing | EV/EBITDA | A buyer pays for the debt too |
A practical workflow: start with P/E because it is everywhere, check EV/EBITDA when debt or depreciation differs a lot between the companies you compare, and if the two rankings disagree, work out which of the four reasons above explains it. The disagreement is often the most useful thing either multiple tells you.
How Finabase does it
A stock's Valuation tab charts ten fiscal years of P/E, EV/EBITDA and related ratios side by side, so you can see whether a company is expensive against its own history on both, for example Walmart. The Comparison tab lines up P/E for several stocks, and the Financials tab lists the operating income, EBITDA, debt and cash used to rebuild EV/EBITDA. The multiples on this page were computed by Finabase from each company's 10-K and the November 17, 2026 closes; the app's figures use the latest twelve months and can differ.
Frequently asked questions
Why is EV/EBITDA usually lower than P/E?
Because EBITDA is larger than net income (it is before interest, tax, depreciation and amortization), while enterprise value is usually only slightly larger than market cap. A bigger denominator gives a smaller multiple.
What is a good EV/EBITDA compared with P/E?
There is no fixed conversion. Compare each multiple with the company's own history and with close peers in the same industry. On this page's seven companies, EV/EBITDA ran from 6.1 to 23.9 and P/E from 7.1 to 46.4 on the same day, which shows how much sectors differ.
Should I use EV/EBITDA or P/E for tech stocks?
Both, with care. For cash-rich, lightly indebted software companies they tell a similar story. For companies spending heavily on data centers, EV/EBITDA flatters, so add price to free cash flow.
Can a stock look cheap on P/E and expensive on EV/EBITDA?
Yes, when it carries a lot of debt (which raises EV but only lowers net income through interest) or when net income is temporarily boosted by a one-off gain or a low tax rate.
Do I use trailing or forward numbers?
Either, as long as both companies are measured the same way. Trailing multiples use the last reported year or twelve months and are facts; forward multiples use analysts' estimates for the next year and are opinions. Mixing a trailing P/E for one company with a forward P/E for another makes the comparison meaningless.
Summary
P/E divides market value by net income; EV/EBITDA divides enterprise value by EBITDA. The first is simple and suits steady, lightly indebted companies and banks; the second removes the effects of debt, tax and depreciation and suits comparisons across different capital structures. On November 17, 2026, Microsoft and Cisco shared a P/E of about 28 but not an EV/EBITDA (19.2 and 21.7), and Walmart's P/E of 46.4 halved to 23.9 on EV/EBITDA. Use both, and when they disagree, find out whether depreciation, debt or a one-off item is the reason.
Check your understanding
4 questions on this lesson. Each answer comes with the reason, and nothing is scored.
Microsoft and Cisco had almost the same P/E (about 28), but Microsoft’s EV/EBITDA was 19.2 and Cisco’s 21.7. What explains most of the gap?
- Microsoft’s larger depreciation, which EBITDA adds back, plus its net cash
- Cisco’s higher share price
- Different accounting standards
Show the answer
A. Microsoft’s larger depreciation, which EBITDA adds back, plus its net cash EBITDA adds back D&A, so a company with heavy depreciation has a much bigger EBITDA than net income. Microsoft’s net cash also lowers its EV.
Which multiple is designed to be neutral to how a company is financed?
- P/E
- EV/EBITDA
- Dividend yield
Show the answer
B. EV/EBITDA EV includes debt and EBITDA is measured before interest, so moving between debt and equity financing changes neither much. P/E moves with interest and the size of the equity.
You are comparing two banks. Which multiple should you start with?
- EV/EBITDA
- EV/sales
- P/E (with price to book)
Show the answer
C. P/E (with price to book) For banks, borrowing is the business, so enterprise value and EBITDA have no clear meaning. Bank analysts use P/E and P/B.
A company spends three times its depreciation on new equipment every year. Which multiple is most likely to make it look cheaper than it is?
- EV/EBITDA, because EBITDA ignores capital spending
- P/E, because net income includes interest
- Price to free cash flow
Show the answer
A. EV/EBITDA, because EBITDA ignores capital spending EBITDA leaves out both depreciation and the cash spent on new assets. Price to free cash flow charges the spending; P/E at least charges depreciation.
Sources and further reading
- Microsoft Corporation — Form 10-K for the fiscal year ended June 30, 2026 (SEC EDGAR)
- Cisco Systems — Form 10-K for the fiscal year ended July 25, 2026 (SEC EDGAR)
- Walmart Inc. — Form 10-K for the fiscal year ended January 31, 2026 (SEC EDGAR)
- Procter & Gamble — Form 10-K for the fiscal year ended June 30, 2026 (SEC EDGAR)
- The Home Depot — Form 10-K for the fiscal year ended February 1, 2026 (SEC EDGAR)
- Verizon Communications — Form 10-K for the fiscal year ended December 31, 2025 (SEC EDGAR)
- AT&T Inc. — Form 10-K for the fiscal year ended December 31, 2025 (SEC EDGAR)
- Finabase chart data — daily closes of November 17, 2026; earnings, D&A, debt and cash from the SEC EDGAR XBRL company facts of each 10-K
