The EBITDA Trap: Why the Market's Favorite Metric Measures Less Cash Than It Promises

Anyone who knows me knows I'm no fan of EBITDA. I've lived with it for more than twenty years across negotiating tables, and I've reached a conclusion that sounds like a contradiction: in most cases, EBITDA does more harm than good. It became a buzzword in board meetings, in fund pitches, in bar conversations between business owners. And like every trend that catches on, it started being repeated without many people stopping to ask whether it makes sense.

Some people 'teach' that EBITDA is the indicator of a business's cash-generating capacity. This is where I first raise my hand. It's nothing of the sort. EBITDA measures operating income before a few accounting adjustments, which can be useful for some very specific analysis, but it's terrible at precisely what people sell it as doing.

Charlie Munger used to say that every time you see the word 'EBITDA,' you should mentally replace it with 'bullshit earnings.' The phrase is crude on purpose. And it is the thesis of this piece: EBITDA doesn't measure what people claim it measures, and when it becomes the only yardstick of valuation in a sale process, the damage can be significant.

Where the fashion came from

EBITDA (earnings before interest, taxes, depreciation and amortization) caught on with the market in the 1980s and 1990s, largely thanks to the leveraged deals of that era, when the buyer needed a number that showed gross operating generation to size up how much debt the business could carry.

In practice, it starts from the company's operating income and adds back the depreciation of assets. The idea is to strip out depreciation, which shows up depressing operating income, on the argument that it has no cash effect: since it isn't a disbursement, the amount supposedly 'stays' in the cash box. It also excludes income tax, because that is contaminated by financial and non-operating results and would distort comparisons between companies.

So far, so good. There is a logic to it. By neutralizing capital structure (interest), tax policy (income taxes) and non-cash items (depreciation and amortization), EBITDA makes it easier to compare companies in the same sector that carry different debt loads and tax regimes. It is a standardization tool. The problem starts when a standardization tool is sold as a measure of cash.

The flaw with no fix: pretending depreciation isn't money

The central flaw of EBITDA, and to me it is beyond repair, is treating depreciation as if it were an accounting fiction with no consequence in your pocket. There is a handful of businesses where that premise really holds: certain technology and service sectors, delivered with low use of equipment and fixed assets. But it is a short list.

For the overwhelming majority of the economy's sectors, depreciation is a cost or an expense by nature. It isn't hard to see. Today's depreciation corresponds to a spend that, if the company didn't incur today, it incurred yesterday and will incur again tomorrow. The assets that keep the operation running grow old, become useless, and have to be replaced. What depreciation signals, with all the imprecision of accounting allocations, is the amount the company spent or will have to spend to replace what it uses to function.

Anyone who runs a company's cash feels this firsthand. Just open the bank statement: nearly every month there is an outflow to buy a machine, a vehicle, a piece of equipment, a system, something that replaces what wore out. How, then, can anyone argue that we should ignore depreciation to gauge a business's cash generation? That is ignoring one of the very accounts that hurts the most.

Warren Buffett treats depreciation as the worst kind of expense there is, because it is a reverse flow: you lay out the money before the revenue arrives, and only recognize the expense little by little, much later. In Berkshire Hathaway's 2000 annual letter, he summed up his skepticism with an image that became famous, asking whether management thinks the tooth fairy pays for capital expenditures. The jab lands because EBITDA, by erasing depreciation, creates the illusion that replacing assets is free. BH letter, 2000

Perhaps the best Brazilian illustration of this point is Localiza. The fleet-rental business carries enormous depreciation, because a car is an asset that runs, loses value and has to be replaced all the time. Looking at the company's EBITDA and concluding that all of it is available cash would be a gross mistake: much of it 'vanishes' into renewing the fleet, month after month. Depreciation there is not an accountant's abstraction. It is the real cost of keeping the business standing. Sweeping it under the rug does not make the money appear.

My practical advice is simple: prefer EBIT, the good old operating income. Imperfect as it is, it keeps depreciation in the account and gets far closer to the operation's cash-generating potential. Less glamour in the acronym, more truth in the number.

In M&A, the damage is bigger

I find it curious that, even though it does not point to real cash generation, EBITDA still reigns as the valuation reference in mergers and acquisitions. The buyer shows up, applies a multiple to EBITDA and says the company is 'worth x times EBITDA.' It is a dangerous path.

I acknowledge the argument in its favor. Since companies in the same sector tend to have similar risk premiums and market challenges, their discounted free cash flow projections converge toward similar EBITDA multiple ranges, which gives the market a comparison shortcut. That is true. But a comparison shortcut is not a measure of value, and confusing the two is where the problem lives.

The first kind of damage comes from the freedom the market has taken to stretch what it calls 'non-recurring' revenues or expenses. It became a game. Spent on a consulting project? Non-recurring. Took a loss with a client? An isolated event. Each of these adjustments inflates EBITDA and, multiplied by six, seven, eight times, turns into a small fortune in the price. The second is the gray zone of buying replacement assets: is that an investment to be capitalized or an expense that reduces earnings? Depending on the answer, the same business changes value. The third, and perhaps the most treacherous, is the metric's indifference to changes in working capital and to financing interest, which often devour the cash at breakfast.

Oi is a Brazilian example that needs no introduction. For a long time it displayed a telecom operator's EBITDA, robust on paper, while its cash was eroded by heavy network investment and a mountain of debt interest. EBITDA saw none of it, because seeing it is not EBITDA's job. The outcome was one of the largest judicial reorganizations in the country's history. Anyone who had looked only at EBITDA, which the year before (2015) came to R$7.2 billion, would have seen a profitable company; anyone looking at cash flow after capex and interest would have seen a very different one. No wonder that, in the same period, total net debt reached R$38.1 billion. Oi's reports

Abroad, the examples are even more blatant. The HP and Autonomy case is a business-school classic. HP bought the British firm Autonomy for around US$ 11.1 billion in 2011 and, a little over a year later, recognized an US$ 8.8 billion writedown, attributing much of it to accounting irregularities that had allegedly dressed up the target's results. A deal that seemed to fit a comfortable multiple evaporated once the numbers behind the metric proved fragile. HP financial report

But my favorite is WeWork. On the eve of its 2019 IPO attempt, the company used a creation called 'Community Adjusted EBITDA.' With justifications as elegant as they were devious, that version excluded marketing expenses from the calculation (supposedly 'investments'), the pay of part of its back-office staff (supposedly 'temporary' growth costs, therefore 'non-recurring'), and even property rent costs (again, 'investments'). The maneuver made the company look far more profitable and sustainable than it was, propping up a valuation that was pegged at as much as US$ 47 billion. This time the market did not fall asleep at the wheel: once investors caught the trick, the value collapsed, the IPO was shelved and the company slid into crisis. An EBITDA so adjusted that it needed a name of its own should have set off the red light long before. WeWork filing, SEC

'But everyone uses the EBITDA multiple'

I do, yes, I know I do. And I am not preaching that owners throw EBITDA in the trash. I am preaching moderation. The EBITDA multiple is a first-conversation tool, a quick way to place a company relative to its peers. As the starting point of a negotiation, it has its uses. As a verdict of value, no.

The difference between the seasoned advisor and the improvised one usually lies right there. One treats the EBITDA multiple as the first line of an analysis that still has to pass through discounted cash flow, quality of earnings, working capital needs, an investment plan. The other stops at the multiple, and that is the one who can leave value on the table without noticing. A quick aside: here I am talking about EBITDA as a measure of performance and of cash. The specific confusion between the EBITDA multiple and the fair price of a Brazilian mid-sized company I have already covered in another piece, and I won't repeat the argument here.

What to do on Monday morning

If you own a company and want to truly understand its health, start by swapping EBITDA for EBIT on your dashboard. That alone puts depreciation back in the account and brings the number closer to the cash that actually remains. Then look at three things EBITDA insists on ignoring: how much you spend to replace the assets that keep the operation running (take the chance to compare it against book depreciation), how much of your cash is locked in working capital, and how much your debt interest eats every month. On that last point, working capital, I have written a separate guide.

And if you are on the side of whoever is selling or buying a company, keep one practical rule: be wary of heavily adjusted EBITDA. The more gymnastics it takes to reach the number, the more fake it tends to be. Ask to see EBITDA without the 'non-recurring' adjustments and compare. The distance between the two numbers is, often, the distance between the company they are selling you and the company that exists. A good buyside advisor does this reconstruction work before and after due diligence, and that is where many prices agreed in the euphoria quietly begin to move.

EBITDA is an acronym that promises cash generation and delivers standardization. They are different things, and the owner who confuses the two pays the difference in the price. In the end, every metric is a lens: EBITDA magnifies what matters to whoever wants to sell high and blurs what hurts in the pocket of whoever runs the business. It is up to you to decide whether you will see the company through the seller's lens or the owner's.

This content is informational and general in nature and does not constitute a recommendation or specific advice for any transaction.