In the 1980s, leveraged buyout firms had a problem. They wanted to buy companies using massive amounts of debt, then point at how 'profitable' those companies were to justify the deals. But after subtracting interest payments on all that debt, plus depreciation, plus taxes, the reported profits often looked thin or nonexistent. So they invented a different number — one that conveniently ignored all of those things.
That number was EBITDA: Earnings Before Interest, Taxes, Depreciation, and Amortization. It made highly-leveraged companies look much more profitable than their net income suggested. It also became, over the next forty years, one of the most quoted metrics in all of finance — and one of the most criticized.
What it actually measures
EBITDA is a company's earnings before you subtract four specific things: Interest, Taxes, Depreciation, and Amortization. The idea is to measure 'core operating profitability' — what the business itself generates, before financing decisions (interest), government claims (taxes), and accounting estimates of asset wear-and-tear (D&A).
The formula:
EBITDA = Net Income + Interest + Taxes + Depreciation + Amortization
Or equivalently, working from the top of the income statement: EBITDA = Revenue − Cost of Goods Sold − Operating Expenses + Depreciation + Amortization. (Operating expenses already include D&A, which is why we add them back.)
The motivation makes some sense. If you're comparing two companies in the same industry but one was funded with debt and the other with equity, their interest expenses will be wildly different even if the underlying businesses are equally good. Stripping out interest puts them on a level field. Same logic for taxes (jurisdictions vary) and depreciation (accounting policies vary).
Why analysts use it anyway
Despite the criticism, EBITDA has stuck around for forty years because it solves real problems:
- Comparing across capital structures. A company with $5B of debt and a company with $0 debt can be compared on EBITDA in a way they can't on net income.
- Comparing across tax regimes. Multinationals with operations in dozens of countries face different tax rates everywhere. EBITDA strips that out.
- Approximating cash flow. Depreciation and amortization are non-cash expenses — accounting entries, not actual money leaving the business. Adding them back gets you closer to the cash the business actually generates.
- Valuation multiples. The EV/EBITDA ratio is the standard valuation metric in private equity and M&A. It's used because it's comparable across companies regardless of capital structure.
The Munger problem
Charlie Munger famously called EBITDA 'bullshit earnings.' Warren Buffett has piled on, saying any management team that prefers to talk about EBITDA is asking him to ignore the real costs of running their business.
Their objection is specific and devastating. Depreciation isn't a fake expense. If a manufacturer buys $100 million of machinery that wears out over 10 years, that's $10 million of real cost per year. The cash already left the building when the machinery was bought, but the economic reality is that the business consumed $10M of useful asset life every year. Adding depreciation back to 'earnings' pretends that consumption never happened.
The same logic applies to amortization of acquired intangibles. If a company paid $1B for another company's customer relationships and is amortizing them over 10 years, that's a real cost. The $1B is gone; the customers are wearing out. Pretending otherwise inflates earnings by exactly the amount you're pretending.
EBITDA works as an intermediate analytical tool. It fails as a substitute for actual profitability. The sin isn't using it — the sin is using it as if it represented economic reality.
How to use it without getting fooled
A few rules:
- Never use EBITDA in isolation. Always check it against net income, free cash flow, and capex. If a company has growing EBITDA but flat free cash flow, that's a red flag.
- Watch the 'capex / D&A' ratio. If a company is spending much more on capex than it's recording in depreciation, the depreciation number is understating real economic costs. The EBITDA flatters the business more than it should.
- Be suspicious of 'adjusted EBITDA.' Companies sometimes add back stock-based compensation, restructuring charges, 'one-time' costs that recur every year, etc. The more adjustments, the less the number means.
- Use EV/EBITDA for comparisons, not absolute levels. Saying 'EBITDA is $500M' tells you nothing without context. Saying 'EV/EBITDA is 8x vs industry average of 12x' tells you something useful.
The Stock Price Decoder shows EBITDA alongside net income, revenue, and ROIC for any US public company — so you can see whether a company's 'great EBITDA' is actually translating into anything that matters.