Articles

Every quarter, companies that lost money on a GAAP basis issue press releases leading with EBITDA. Every leveraged buyout is priced at a multiple of EBITDA. Every loan covenant is written around it. The metric is everywhere, and the reason is not that it is the most accurate measure of profit. The reason is that it is the most flattering one that a sophisticated audience will still accept without laughing. Understanding the EBITDA definition means understanding both why the metric is genuinely useful and exactly what it is designed to obscure.

What EBITDA Is and Where It Came From

EBITDA is earnings before interest, taxes, depreciation, and amortisation. The formula: start with net income and add back interest expense, tax, depreciation on physical assets, and amortisation on intangibles such as patents, software licences, and customer lists from acquisitions. Alternatively, start with operating income and add back depreciation and amortisation only.

John Malone at TCI cable built the metric into standard practice in the 1970s and 1980s. Cable networks required massive upfront capital investment that created depreciation charges large enough to wipe out reported profit for years. Malone’s argument was that those charges were accounting artefacts, not measures of business health, and that operating earnings should be assessed before them. The leveraged buyout industry of the 1980s adopted the same logic because LBOs are financed with debt whose interest expense would otherwise make every acquisition look like a money-loser from day one. EBITDA let deal-makers present a purchase price as a multiple of operating earnings without the financing structure contaminating the comparison.

The result is a number that describes what the business earns from its core operations before anyone gets paid for lending it money, before governments take their cut, and before accountants reduce profit to reflect assets wearing out.

Why Companies Report It

The honest answer is that EBITDA makes the numbers look better, and it does so in ways that are technically defensible. A capital-intensive company with heavy debt, a recent acquisition, or assets that depreciate quickly will show a far larger EBITDA than net income. Presenting EBITDA is not fraud. It is a choice about which lens to use, and companies choose the lens that shows them in the best light.

Three legitimate uses exist alongside the flattering ones. First, EV/EBITDA is the standard acquisition pricing metric because it strips out the target company’s existing capital structure. A buyer replacing the target’s debt with its own financing needs to see operating earnings before interest to compare deals on equal footing. Second, net debt divided by EBITDA is the standard leverage ratio in loan covenants. Lenders set ceilings because EBITDA is the closest thing to a cash generation metric that does not require forecasting capital spending. Third, comparing operating margins across companies in the same sector is cleaner with EBITDA than net income when the companies have different tax rates or financing structures.

Telecoms, utilities, cable operators, and private-equity-owned businesses are routinely priced on EV/EBITDA for exactly this reason. These are sectors where the gap between EBITDA and net income is structural and persistent, not temporary.

What It Hides

Warren Buffett has criticised EBITDA for decades, and the critique is precise: depreciation is not a fiction. It represents real capital assets that will eventually need replacing. A logistics company depreciating $60 million of trucks per year will spend approximately $60 million replacing those trucks over time. Adding depreciation back to earnings and calling the result profit is telling investors they can ignore a cost that will arrive on schedule.

The metric ignores capital expenditure entirely. A company spending $200 million per year on maintenance capex to keep its assets functioning has $200 million less cash than its EBITDA implies, every single year. The standard check is to compare EBITDA with free cash flow after capital expenditure. A persistent and widening gap between those two numbers means the business is consuming cash at a rate EBITDA does not show.

The fraud vector is more serious. WorldCom booked approximately $3.8 billion of ordinary network maintenance costs as capital expenditure, keeping those costs out of operating expenses and therefore out of EBITDA, until the accounting fraud collapsed in 2002. The mechanism worked because EBITDA ignores capex: moving costs from the income statement to the balance sheet as capital spending inflates EBITDA directly.

WeWork in 2018 presented bond investors with a “community-adjusted EBITDA” that excluded building costs and growth costs, converting a 2017 net loss of more than $900 million into a stated profit. That version was aggressive enough to attract attention, but the underlying problem, that companies define their own adjusted EBITDA with no standardised rules, exists across markets.

Neither GAAP nor IFRS defines EBITDA. The SEC requires US-listed companies to reconcile any non-GAAP metric to the nearest GAAP equivalent in public filings. Reading that reconciliation, specifically looking at what was added back and why, is the only way to assess whether a company’s EBITDA is conservative or engineered.

The Right Way to Use It

EBITDA is most useful when used as a starting point rather than an endpoint. Calculate it, then subtract maintenance capital expenditure to get a figure closer to genuine cash generation. Compare it to free cash flow over several years: a company whose free cash flow consistently runs 80% to 90% of EBITDA is generating what the metric implies, while one whose free cash flow is 30% of EBITDA is hiding significant costs in the gap.

For loss-making companies, heavily indebted businesses, or companies carrying large acquisition-related amortisation charges, EV/EBITDA often works where the P/E breaks down entirely. A negative net income produces a meaningless P/E. EBITDA can still be positive and the multiple meaningful. That is a genuine advantage, as long as the capital expenditure reality is checked separately.

For any adjusted EBITDA figure, list the adjustments and ask whether each one is genuinely non-recurring. Stock-based compensation is routinely excluded from adjusted EBITDA as a non-cash charge, but it dilutes shareholders every year and is a real cost of retaining employees. Excluding it consistently is not adjusting for a one-time event; it is permanently removing a recurring expense from the profit figure.

Conclusion

Companies report EBITDA because it is the largest defensible profit figure available and because the financial system, from loan covenants to acquisition pricing to analyst models, is built around it. It is genuinely useful for comparing operating performance across different capital structures and for industries where depreciation is a structural feature rather than an annual surprise. What it hides is capital expenditure, the real cost of maintaining the asset base that generates those earnings. Take EBITDA, subtract maintenance capex, read the reconciliation to net income, and you have a metric worth using. Take it at face value and you have exactly what management wanted you to see.