EV/EBITDA by Industry — 2026 Sector Benchmarks

EV/EBITDA strips out two of the biggest distortions in P/E analysis: capital structure and tax rates. A highly leveraged company looks expensive on P/E because interest charges eat into net income — EV/EBITDA eliminates that by using Enterprise Value (equity plus net debt) in the numerator. This makes it far more useful for capex-heavy sectors like energy, utilities, and industrials where debt is a structural part of the business model, not a warning sign.

Why Tech trades at 18x and Energy at 5x. The spread is not random — it reflects how much of each sector's EBITDA survives as free cash flow. A software company might convert 80–90% of its EBITDA to free cash flow because it needs almost no capital to grow. An oil producer might convert only 30–40% after maintenance capex and wellfield depletion. A dollar of software EBITDA is simply worth more in cash terms, and the market prices it accordingly.

The capex haircut matters most at sector boundaries. EBITDA is earnings before depreciation — but depreciation approximates the real cost of maintaining assets. For capital-light businesses (software, asset managers, marketplaces), D&A is minor and EBITDA closely tracks cash generation. For capital-intensive businesses (telecom, airlines, mining), D&A understates true economic wear and the EBITDA multiple inflates the apparent value. Always check capex as a percentage of EBITDA before trusting a sector multiple at face value.

Beware the cyclicality trap. Energy and materials often show their lowest EV/EBITDA at the peak of the commodity cycle, right before earnings collapse. A 5x multiple that looks cheap may reflect peak-cycle EBITDA that will normalize down 50% — implying the stock actually trades at a normalized 10x. The sector averages above are through-cycle medians, not point-in-time reads. Compare a company's current EBITDA against its own 5-year average before concluding it's cheap.

The only useful comparison is within sector. Never compare a retailer at 11x to a SaaS company at 22x and conclude the retailer is cheap. The multiples embed completely different assumptions about growth, cash conversion, and cyclicality. The question worth asking is whether a specific company trades at a premium or discount to its own sector median — and whether the deviation is justified by better margins, faster growth, or higher-quality EBITDA. For a deeper dive on the metric itself, see our EV/EBITDA guide.

For banks and insurance companies, EV/EBITDA is not just imprecise — it is conceptually broken. Debt is an operating input for financial firms, not capital structure. A bank borrows to lend; its liabilities are the raw material of its business. Enterprise Value (which adds debt to market cap) therefore double-counts the core activity. The resulting multiple is not comparable to non-financial sectors and should be ignored entirely. Use Price-to-Book for banks — it anchors valuation to tangible asset quality and equity return, which is what actually drives bank earnings. Insurance companies add another complication: float (policyholder premiums held before claims) distorts the balance sheet further.

REITs are a more nuanced case. EV/EBITDA technically exists for a REIT, but accounting depreciation on real estate substantially understates the asset's economic durability — properties don't actually depreciate the way GAAP suggests. This means EBITDA is inflated relative to true cash generation. The industry standard is Funds From Operations (FFO), which adds back real-estate depreciation, or Adjusted FFO (AFFO), which also backs out maintenance capex and straight-line rent adjustments. EV/EBITDA can serve as a rough cross-check for REIT comparisons but should never be the primary valuation metric for this sector. For more on what EBITDA captures and where it breaks down, see our EBITDA explained for investors guide.