EBITDA Explained (and Why It Can Mislead)

EBITDA — Earnings Before Interest, Taxes, Depreciation, and Amortization — is the most widely quoted profitability metric in private equity, M&A, leveraged finance, and corporate development. Almost every deal that closes is sized in EBITDA multiples. Almost every credit decision references EBITDA. Almost every analyst includes EBITDA prominently in their model.

It is also, according to a vocal minority of serious investors (Warren Buffett, Charlie Munger, Howard Marks), a misleading metric that has done substantial damage to the discipline of finance. The argument is not that EBITDA is meaningless — it has uses — but that it is so frequently substituted for more meaningful measures that an entire generation of finance professionals reasons through it without realizing what it's hiding.

This article covers what EBITDA is, why it became popular, the cases where it's useful, why Buffett hates it (and he's mostly right), "adjusted EBITDA" and the games played there, real cases of EBITDA misuse, and what to use instead.

What EBITDA is

EBITDA is operating income with depreciation and amortization added back.

$\text{EBITDA}=\text{EBIT}+\text{D\&A}$

or equivalently:

$\text{EBITDA}=\text{Net Income}+\text{Interest}+\text{Taxes}+\text{D\&A}$

The intuition: EBITDA strips out items that vary by capital structure (interest), tax jurisdiction (taxes), and historical investment choices (D&A). What's left is a measure of operating profitability that is, in theory, comparable across companies with different financing structures, geographic footprints, and asset bases.

This is the case for EBITDA: it normalizes for things you don't necessarily care about when comparing two businesses' operational performance.

Why EBITDA became popular

Three factors drove EBITDA's rise to dominance in finance.

1. Private equity and leveraged buyouts.

In an LBO, the acquirer takes on substantial debt. Interest expense is large. Tax shields from interest payments are valuable. The pre-deal income statement reflects the seller's capital structure, not the buyer's. PE firms wanted a metric that abstracted from financing — something that measured the underlying operating cash generation.

EBITDA serves that purpose. From the EBITDA number, you can build out a new capital structure, model new interest expense, and arrive at an equity return projection. This is the foundation of LBO modeling.

2. Comparability across companies.

If you want to compare a US company (with high tax rates) to a similar Irish-domiciled company (with low taxes), EBITDA removes the tax distortion. If you want to compare a debt-free company to a highly levered one, EBITDA removes the interest distortion. These are legitimate comparability concerns.

3. Approximation to cash flow.

EBITDA approximates the operating cash generation of a business before financing decisions. For businesses with stable D&A and CapEx (think mature consumer staples), EBITDA tracks roughly to operating cash flow.

For these reasons, EBITDA became the default operating metric in deal contexts. EV/EBITDA multiples became the standard valuation shorthand. "What's the company's EBITDA?" replaced "what does the company earn?"

When EBITDA is actually useful

EBITDA has legitimate uses. The objection isn't that EBITDA is always wrong — it's that EBITDA is frequently used outside its valid range.

Use 1 — Comparing operations of similar-risk businesses across capital structures.

If you want to know which of two businesses is operationally better, ignoring financing, EBITDA can help. The caveat: only if their CapEx requirements and D&A are similar.

Use 2 — Modeling LBOs.

In LBO modeling, you start with EBITDA, build the new debt service, and project equity returns. EBITDA is the natural starting point because it's pre-financing.

Use 3 — Credit analysis (with caution).

EBITDA is the conventional measure of debt capacity. Debt covenants are often expressed as "net debt to EBITDA" or "EBITDA to interest." These are useful — but with the recognition that EBITDA overstates the cash actually available for debt service.

Use 4 — Comparing companies in industries with similar CapEx intensity.

For two SaaS companies with similar CapEx, EBITDA is a reasonable proxy for cash generation. The same is largely true for franchise-model businesses, asset-light services, and similar capital-light industries.

Why Buffett hates EBITDA

Buffett and Munger have argued for decades that EBITDA is a misleading metric in most uses. The argument has three components.

Component 1 — D&A is a real cost.

The "DA" in EBITDA represents the periodic allocation of past CapEx. Adding D&A back implicitly treats it as if it weren't a cost. But the underlying CapEx was a real cash outflow. For most businesses, sustained CapEx (at least at the maintenance level) will recur. Adding D&A back gives you a number that isn't sustainable cash generation.

Buffett's frequent point: in a business that requires $100MofannualmaintenanceCapEx,the"true"annualcostincludesthat$100M. Reporting EBITDA without subtracting maintenance CapEx — and adding D&A back — produces a number that overstates the cash truly available to capital providers.

Component 2 — Interest is a real cost.

The "I" in EBITDA also gets added back, on the theory that interest depends on capital structure choices and you want to abstract from those. But interest is, in fact, a cost that must be paid. A company with high interest expense has less free cash than one with low interest expense, regardless of EBITDA.

Component 3 — EBITDA encourages overpayment in deals.

When deals are priced at EV/EBITDA multiples, the buyer is implicitly paying based on a metric that overstates cash generation. The historical record shows that high EV/EBITDA multiples are correlated with poor post-deal returns. Buffett's argument: EBITDA-based pricing systematically inflates deal valuations.

Charlie Munger put it more bluntly: "Every time you hear EBITDA, just substitute 'bullshit earnings.'"

Adjusted EBITDA: the bigger problem

If standard EBITDA has problems, "adjusted EBITDA" — the metric companies disclose in their earnings releases — has much bigger ones.

Adjusted EBITDA starts with EBITDA and then adds back additional items that management argues are non-recurring or non-cash. Common additions:

• Stock-based compensation — argued as "non-cash," even though it dilutes equity holders.
• Restructuring charges — argued as "one-time," even when they happen every year.
• Acquisition-related costs — argued as "non-recurring," for serial acquirers who acquire every year.
• Integration costs — same.
• Legal settlements — argued as "one-time," even when the company has a pattern of legal issues.
• Goodwill impairments — argued as "non-cash," which is true, but the impairment reflects real value destruction from past acquisitions.

Each adjustment has some defensible logic. The problem is the cumulative effect. By the time a company's "adjusted EBITDA" is calculated, the metric bears little resemblance to actual cash flow.

A 2018 SEC study found that companies' reported "adjusted" metrics differed from GAAP earnings by more than 30% in many cases, and the adjustments almost always made the picture look better. The bias is systematic.

Real cases of EBITDA misuse

WeWork (pre-IPO 2019) famously introduced "Community Adjusted EBITDA," which added back not just typical items but also marketing, general and administrative expenses, and development costs. The metric was so detached from cash reality that the SEC required disclosure adjustments, and the IPO ultimately failed.

Theranos never disclosed audited financials, but reported EBITDA-style metrics in private fundraising that bore no relationship to its actual (negative) cash generation. The metric was used to support valuations that were entirely fictional.

Various LBO failures. Many private equity acquisitions that ultimately failed financially were priced at high EBITDA multiples that didn't survive contact with actual cash flows. The pattern: EBITDA looked attractive; cash flow after debt service didn't.

These aren't anti-EBITDA arguments per se. They're arguments against using EBITDA as a substitute for cash flow analysis.

What to use instead

The disciplined alternative to EBITDA:

For operational comparison: EBITDA minus maintenance CapEx (sometimes called EBITDA-CapEx or FCF Margin). This captures the genuine cash generation of the operating business.

For deal valuation: free cash flow. Compute it explicitly. Don't shortcut to EBITDA multiples.

For credit analysis: cash flow available for debt service (CFADS), which starts with operating cash flow and subtracts taxes, working capital changes, and maintenance CapEx. CFADS is what bondholders care about.

For business performance over time: track multiple metrics. EBITDA, operating cash flow, free cash flow, and FCF margin together give a fuller picture than any single number.