EV/EBITDA — The multiple that sees through debt
Worked example: Anheuser-Busch InBev — 2016
In November 2006, a private equity consortium paid $44 billion to take Clear Channel Communications private. The deal valued the company at roughly 13 times EBITDA — a multiple that looked reasonable to the banks and sponsors involved. Two years later, the renamed iHeartMedia was struggling under $20 billion in debt. Its P/E ratio was useless as a valuation tool because there were almost no earnings. Its price-to-book was distorted by acquisitions. But its EV/EBITDA told the story clearly from the beginning: the enterprise had been purchased at a price that left no margin for error, and the debt load made any operating stumble catastrophic. iHeartMedia filed for bankruptcy in 2018. The enterprise value multiple had not been wrong. The assumptions beneath it had been.
The concept in 60 seconds
EV/EBITDA divides a company's enterprise value by its earnings before interest, taxes, depreciation, and amortization. Both components matter.
Enterprise value is the full price tag of the business — market capitalization plus net debt (total debt minus cash). If you wanted to buy the entire company outright, enterprise value is approximately what you would pay: the equity at market price, plus the obligation to assume its debt, minus the cash you would inherit. EV captures the whole capital structure, not just the equity layer.
EBITDA is a proxy for operating cash generation before financing choices. By stripping out interest (debt cost), taxes (jurisdiction choice), depreciation, and amortization (historical capital allocation), it enables comparison across companies with different capital structures, debt levels, and accounting treatments for fixed assets.
Divided together, EV/EBITDA answers: what multiple of operating cash generation is the market charging for this entire business?
Common ranges: capital-light, high-growth businesses trade at 20–40 times EBITDA. Mature industrials and consumer staples trade at 8–14 times. Capital-intensive cyclicals — energy, mining, basic materials — often trade at 4–8 times at mid-cycle. Private equity historically pays 8–12 times EBITDA for platform acquisitions, with leverage added on top.
The inverse of EV/EBITDA — EBITDA yield on enterprise value — is useful for the same reason earnings yield is useful for P/E: it can be compared to borrowing costs and bond yields to quickly assess whether a deal makes financial sense.
Mental model
Think of EV/EBITDA as a whole-business price tag expressed in years of operating cash flow.
An EV/EBITDA of 10 means you are paying 10 years of operating cash flow for the entire business at current run-rate. Whether that is reasonable depends entirely on how confident you are that the operating cash flow will persist, grow, or decline.
The key advantage over P/E is capital structure neutrality. Two identical businesses — same revenue, same margins, same growth — will have very different P/E ratios if one carries debt and the other does not. The leveraged business pays more interest, which reduces net income, which compresses the P/E denominator. EV/EBITDA sees through that: it captures the same operating economics for both. This is why acquirers, private equity firms, and analysts comparing companies across industries almost always start with EV/EBITDA rather than P/E.
The key limitation is the EBITDA figure itself. Depreciation is stripped out — but depreciation approximates real economic wear on assets that eventually must be replaced. A capital-intensive manufacturer with $100m in EBITDA and $60m in annual maintenance capex is not in the same position as a software company with $100m in EBITDA and $5m in maintenance capex. EV/EBITDA treats them identically. That is why EV/EBITDA works best for capital-light businesses and can badly mislead for capital-heavy ones.
The simplest discipline: always calculate EV/EBITDA alongside EV/EBIT (which includes depreciation) and EV/FCF (which captures actual cash generation). The spread between these three figures reveals how capital-intensive the business really is.
Worked example: Anheuser-Busch InBev, 2016
When AB InBev completed its $107 billion acquisition of SABMiller in October 2016, it created the largest beer company in the world. The combined entity held roughly 28% of global beer volumes. For a deal of that size and complexity, P/E was close to useless: SABMiller's net income was distorted by acquisition costs, currency effects, and restructuring charges. EV/EBITDA was the primary lens for the deal.
SABMiller's last reported EBITDA before the deal was approximately $6.7 billion. The enterprise value paid (equity price plus net debt assumed) was approximately $107 billion. That implied an acquisition EV/EBITDA of roughly 16 times — at the high end of historical brewing sector ranges of 11–14 times, but justified by the synergy thesis.
AB InBev's management projected $1.4 billion in annual cost synergies within three years: procurement savings, distribution network consolidation, and overhead elimination. If those synergies materialized, the effective acquisition multiple fell to approximately 13 times — within normal sector parameters.
By 2019, the synergies had largely been achieved, but a second problem had emerged: the combined group had taken on $108 billion in debt to fund the acquisition. Debt service consumed a large portion of operating cash flow. The stock fell from a peak of $130 in 2016 to under $60 by 2018 — not because EBITDA had collapsed, but because leverage magnified every revenue miss and investors re-rated the equity risk premium upward.
The lesson is not that EV/EBITDA was wrong. The deal was priced at a defensible multiple of operating cash generation. The lesson is that EBITDA is not cash flow available to equity holders when there is significant debt. EV/EBITDA evaluates the business; capital structure determines what the equity is worth once debt obligations are met.
| Year | Enterprise Value | EBITDA | EV/EBITDA | Net Debt | Context |
|---|---|---|---|---|---|
| 2015 (SABMiller standalone) | ~$80bn | $6.7bn | ~12x | $7bn | Pre-deal |
| 2016 (acquisition price) | ~$107bn | $6.7bn | ~16x | $108bn | Acquisition premium |
| 2018 (combined, synergies) | ~$180bn | $14.2bn | ~13x | $102bn | Synergies delivered; debt pressure |
| 2019 (equity re-rating) | ~$140bn | $13.8bn | ~10x | $95bn | Multiple compression on leverage risk |
Historical pattern
2004–2007: The LBO era. Private equity drove acquisition multiples for mid-market businesses from 7–8 times EBITDA toward 10–12 times, enabled by cheap debt. At peak leverage, sponsors were paying 11–13 times EBITDA for consumer and industrial businesses, layering 6–7 times debt on top. When credit markets froze in 2008, the leveraged loan market collapsed and many of these businesses — valued at defensible EBITDA multiples but carrying unsustainable debt — entered restructuring. The multiple had been reasonable; the capital structure had not.
2012–2019: The multiple expansion decade. As the 10-year Treasury fell from 3% to 1.5%, EV/EBITDA multiples across public markets expanded steadily. The S&P 500 median EV/EBITDA rose from roughly 8 times in 2012 to nearly 14 times by 2019 — the same multiple compression story as P/E, but visible across capital structures. Acquirers who paid 12 times in 2015 found their acquisitions had re-rated to 16 times by 2019 with no fundamental change to the business.
2020–2021: Peak multiples. Near-zero interest rates pushed EV/EBITDA multiples to extremes. Software-as-a-service businesses traded at 40–80 times EBITDA. The average S&P 500 EV/EBITDA reached 17 times in 2021. Strategically, this made debt-funded M&A attractive: if you could borrow at 3% to buy a business earning an EBITDA yield of 6%, the arbitrage was compelling. That arbitrage closed sharply in 2022.
2022–2023: The rate reset. Rising interest rates compressed EV/EBITDA multiples across every sector. Software multiples fell 50–70% in many cases. The S&P 500 median EV/EBITDA dropped from 17 times toward 12 times. Highly leveraged companies were hit doubly: their EBITDA multiples contracted and their debt service costs rose simultaneously. The 2022 selloff demonstrated again that EV/EBITDA is a cleaner signal of fundamental value than P/E — but it is not insulated from rate environment effects, because the multiple itself is anchored to the cost of capital.
Decision framework
Step 1 — Build the enterprise value correctly. Market cap is easy to find; enterprise value requires adding total debt (including leases under IFRS 16) and subtracting cash. Do not use market cap as a proxy for enterprise value when the company carries material debt or cash. A business with $5bn in market cap and $8bn in net debt has an enterprise value of $13bn — very different conclusions follow from using the wrong number.
Step 2 — Normalize EBITDA. Reported EBITDA is the starting point, not the ending point. Strip out one-time items: litigation settlements, restructuring charges, gains on asset sales, stock compensation (which is a real cost), and any accounting adjustments. For cyclical businesses, use a 3–5 year average EBITDA to avoid buying at a peak multiple on peak earnings. For acquisition targets, adjust for known synergies — but be conservative: synergy estimates have a consistent history of optimism.
Step 3 — Calculate EV/EBIT and EV/FCF alongside EV/EBITDA. The gap between these three figures tells you how capital-intensive the business is. A software company might have EV/EBITDA of 20 times, EV/EBIT of 22 times, and EV/FCF of 21 times — the three numbers are nearly identical because capex is minimal. A toll road might have EV/EBITDA of 12 times, EV/EBIT of 25 times, and EV/FCF of 30 times — depreciation and capex are real and large. If you only look at EV/EBITDA for the toll road, you severely underestimate the true cost of ownership.
Step 4 — Compare to sector medians and the company's own history. EV/EBITDA varies enormously by sector — comparing a retail business at 6 times to a software business at 25 times tells you almost nothing about relative value. Compare within sector and against the company's own 5–10 year history. A business trading at a 30% discount to its own 5-year median EV/EBITDA, with no fundamental deterioration, is making a different argument than one at a 40% premium.
Step 5 — Stress-test for debt. Calculate what happens to equity value under scenarios where EBITDA falls 20% and 40%. With significant leverage, a 20% decline in EBITDA can wipe out a disproportionate share of equity value — not because the business is destroyed, but because fixed debt obligations consume a larger percentage of reduced operating cash flow. The equity cushion below EV determines how much protection there is.
Common mistakes
Using EBITDA as a proxy for free cash flow. EBITDA ignores capital expenditure, working capital changes, and taxes paid. For capital-intensive businesses — manufacturers, utilities, telecoms, airlines — the gap between EBITDA and actual free cash flow can be enormous. An airline trading at 5 times EBITDA sounds cheap; when maintenance capex, working capital, and debt service are incorporated, the true FCF yield is often thin or negative. Warren Buffett has described EBITDA as "earnings before bad stuff" — the stripping out of depreciation in particular bothers him, because depreciation is the accounting recognition of real capital consumption.
Ignoring the debt load when comparing EV/EBITDA across companies. Two companies at 10 times EV/EBITDA are not equivalent if one has net cash and the other carries 5 times EBITDA in debt. The unlevered business's equity holders receive the full benefit of EBITDA improvement; the leveraged business's equity holders receive residual value after debt service. When comparing on EV/EBITDA alone, it is easy to miss that leverage is making the equity riskier, not cheaper.
Applying sector multiples without understanding why they are what they are. "Software trades at 20 times EBITDA" is a description, not a justification. The reason software trades at a premium is capital-light economics, high recurring revenue, strong retention, and scalable margins. A software company with high churn, low gross margins, and declining revenue does not deserve a 20 times multiple because it has the word "software" in its filing. Always ask why the sector trades where it does, then ask whether the specific business deserves to be priced at, above, or below that level.
Treating EV/EBITDA as a standalone verdict. Like P/E, EV/EBITDA is a starting point. A low multiple can reflect genuine cheapness — or a secularly declining business, a balance sheet crisis, or a pending earnings collapse. A high multiple can reflect deserved premium for quality — or speculative excess with no margin of safety. The multiple opens the conversation; the analysis of earnings quality, competitive moat, management track record, and balance sheet health concludes it.
How VI Stack uses this
EV/EBITDA is the second of five valuation lenses in Gate 3 — The Forensics. It is used alongside P/E, not instead of it.
In the Gate 3 framework, EV/EBITDA is calculated on three bases: trailing reported, normalized (3-year average), and adjusted (stripping one-time items and stock compensation). The divergence between these three figures flags EBITDA quality. A company whose normalized EBITDA is consistently below reported EBITDA is using one-time items to flatter the multiple.
EV/EBIT and EV/FCF are always calculated alongside EV/EBITDA. The capex intensity ratio (capex as a percentage of EBITDA) is pulled directly from the financial statements. High capex intensity means EV/EBITDA is an unreliable standalone metric and the analysis shifts weight toward EV/FCF.
In Gate 4 — The Pitch — any EV/EBITDA multiple used in the target price must be justified with reference to: the business's earnings quality, its position in the capital cycle, the current cost of capital environment, and its historical trading range. A multiple pulled from "sector average" without justification is not accepted.
What's next
val-03 — Free Cash Flow builds on this module by moving from operating earnings proxies to actual cash generated for owners. FCF yield, owner earnings, and the gap between GAAP earnings and real cash are the subject — and they resolve several of the limitations that EV/EBITDA leaves open.
vistack.io