Key takeaways
- EV/EBITDA = (Market Cap + Net Debt) / EBITDA. Capital-structure-neutral, unlike P/E.
- Typical ranges: industrials 6–10×, consumer 10–15×, software 15–30×.
- Use trailing twelve months (LTM) for current valuation; forward (NTM, FY+1) for forecast valuation. Always be explicit which one you mean.
- The multiple compresses for low-growth and high-leverage businesses; expands for growth and margin scalability.
- EV/EBITDA hides capex intensity. Compare with EV/(EBITDA − CapEx) when companies have very different capital requirements.
What it actually measures
Enterprise value is the value of the entire business — what an acquirer would pay to take it private and pay off the lenders. EBITDA is the operating profit before financing, taxes, and non-cash charges. Their ratio answers: "how many years of operating profit am I paying for the whole business?"
EV / EBITDA = Enterprise Value / EBITDA
The capital-structure-neutrality is the key advantage over P/E. A leveraged company has lower net income (more interest), so a lower P/E — but the underlying business may be just as good. EV/EBITDA strips that out: same business, same multiple, regardless of how it's financed.
LTM vs NTM — be explicit
Two timing conventions, and they can imply very different valuations:
- LTM (Last Twelve Months / Trailing) — actual reported EBITDA from the most recent four quarters. Concrete, audited, but backwards-looking.
- NTM (Next Twelve Months / Forward) — forecast EBITDA from analyst consensus or company guidance. Forward-looking, but estimate-dependent.
- FY+1, FY+2 — fiscal year forward multiples. Common in IB pitches.
For growing companies, LTM EV/EBITDA looks higher than NTM because the denominator is smaller. State explicitly which one you're using; mixing them across companies in a comps table produces nonsense conclusions.
What drives the multiple
Higher EV/EBITDA reflects the market's view that:
- Growth — EBITDA will be higher in future periods, so today's denominator understates the value.
- Margin scalability — incremental revenue earns higher EBITDA margin (operating leverage).
- Capital efficiency — the business doesn't need much capex to grow (high ROIC).
- Quality — recurring revenue, customer stickiness, defensible moats.
- Low risk — predictable cash flows, low cyclicality.
Lower multiples reflect the inverse: low growth, capex-heavy, cyclical, or low-quality. The right multiple isn't a number — it's a function of the business profile, and comps tables reveal whether the market views your target similarly to its peer set.
Worked example
A target reports LTM EBITDA of $100M. Comparable peers trade at:
| Peer | Market Cap ($M) | Net Debt ($M) | EV ($M) | LTM EBITDA ($M) | EV/EBITDA |
|---|---|---|---|---|---|
| A | 2,000 | 500 | 2,500 | 200 | 12.5× |
| B | 1,500 | 300 | 1,800 | 150 | 12.0× |
| C | 3,000 | 1,000 | 4,000 | 320 | 12.5× |
| D | 1,200 | (200) | 1,000 | 80 | 12.5× |
| Median | 12.5× |
Applying the median 12.5× to the target's $100M EBITDA: EV = 12.5 × $100M = $1,250M. Subtract net debt of $200M: equity value $1,050M. Across 50M diluted shares: implied price $21 per share.
The arithmetic is the easy part. The judgment is in selecting the peer set. A peer set should match on industry, scale, growth profile, geography, and capital structure. A bad peer set produces a defensible-looking multiple that's actually irrelevant.
EV/EBITDA vs other multiples
| Multiple | Best for | Drawback |
|---|---|---|
| EV/EBITDA | Most M&A, PE, capital-structure-neutral comparison | Ignores capex intensity, cash taxes |
| EV/EBIT | Capital-intensive industries (penalises higher D&A) | Sensitive to depreciation accounting choices |
| EV/Revenue | Pre-EBITDA companies (early-stage SaaS) | Says nothing about profitability |
| P/E | Mature companies, especially financials | Capital-structure-dependent; tax-jurisdiction-dependent |
| P/B | Banks and asset-heavy financials | Useless for asset-light businesses |
| EV/(EBITDA − CapEx) | Capital-intensive cross-industry comparisons | Less standardised; harder to source consistently |
Common EV/EBITDA errors
- Wrong EV calculation. EV must include preferred stock and minority interest, and subtract cash. Many practitioners use Market Cap + Total Debt only — that's wrong.
- Mixing LTM and NTM across the comps table. Pick one convention and apply consistently.
- Bad peer selection. A peer set should match on industry, scale, growth, geography, capital structure. Picking peers by industry alone produces noisy comps.
- Median of a small set. Three peers and one outlier produces a misleading median. Use mean, median, and quartiles; show the full set.
- Ignoring growth. A 12× multiple on a 5%-growth business and a 12× multiple on a 25%-growth business mean very different things. Compute PEG-style adjustments (multiple/growth rate) for cross-checks.
How Smalt AI builds it
Smalt AI's comps tab pulls market cap, net debt, and LTM/NTM EBITDA for the peer set, computes EV consistently (including preferred and minority interest where present), and outputs the median, mean, and quartile multiples. Selection logic is documented (industry, size band, geography, growth band) so the peer set is defensible. The implied valuation range is shown as a fan: median, 25th percentile, 75th percentile.
Further reading
- Rosenbaum & Pearl — Investment Banking, comps construction chapter.
- Damodaran — Investment Valuation, multiples chapter.