The Real Cash a Business Makes

Ask most people how much a company "made" last year, and they will read out the net profit - the big bold number at the bottom of the accounts. (Net profit is what is left after a company subtracts all its costs from its sales.) It feels like the real answer. An accountant signed it, the newspaper printed it. But if you want to know how much money the owner could actually take out and spend, this number is often wrong - sometimes very wrong.

Warren Buffett's answer was to build a different number. He called it owner earnings. Here is the idea in one line: owner earnings is the cash a business truly throws off each year that the owner can put in their pocket without weakening the business. Not the accounting profit. Not cash that just passed through. The cash you could really take out and still wake up next morning owning the same strong business you had before.

Why is this different from the profit number at all? Because the accounts follow rules - sensible rules - that were never made to answer "how much can I safely take out?" Two rules matter here. First, profit is charged a cost called depreciation - a slow write-down for machines bought years ago. That cash already left long ago, so counting it again can make profit look smaller than the real cash. Second, and more dangerous, profit is not charged for the fresh cash a business must spend every single year, forever, just to keep its existing machines, trucks, and shops working. So profit can make the cash look bigger than what the owner can safely take. A business can show a fat profit and hand its owner almost nothing, because every rupee of that profit was swallowed at once by the machines it must keep replacing. This study is about reading past the bold number to the real cash underneath.

Buffett explained all this plainly in his 1986 letter to shareholders, and the idea has not needed changing since. He said owner earnings is the reported profit, plus depreciation and other costs where no cash actually left this year, minus the money a business must spend each year on machines and buildings just to keep its position and keep making the same amount. Read that last part slowly: just to keep - not to grow. The spending that only keeps you standing where you already are.

He was just as sharp about a popular number that quietly skips all of this: EBITDA. That is a long word that means "profit before subtracting interest, tax, and depreciation" - in plain terms, profit before you take out the cost of your machines wearing down. Buffett's complaint, in his own blunt words, was that showing a business on EBITDA pretends the money spent on machines is not real - as if, he said, the tooth fairy pays for it. For a business that must keep replacing costly machines, EBITDA is not the cash the owner can keep. It is the cash before the single biggest, most unavoidable bill. The whole point of owner earnings is to put that bill back where honest reading needs it.

The sum looks like easy arithmetic but is mostly judgement. Start at the bottom - the reported profit - and adjust it.

Reported net profit - the accountant's answer. This is where you start.

Add back the non-cash costs - depreciation and its cousins. These made the profit look smaller, but no cash actually left the building this year; the cash left years ago, when the machine was bought. Adding them back moves you closer to the real cash the business made.

Subtract the keep-it-running spending on machines - and this is the whole game. Every year the business must spend real cash to replace worn-out machines, redo tired shops, and keep the trucks on the road. This spending does not show up as a cost in the year it happens; it only trickles into the profit slowly, as future depreciation. Owner earnings puts the actual yearly amount back in, right now, because that cash truly is not available for the owner to take.

The trap in that last step is that the total spending on machines mixes two very different things: keep-it-running spending (staying level) and growth spending (getting bigger). Only the keep-it-running part should be subtracted, because growth spending is a choice - an owner could stop growing and pocket that cash instead. Telling the two apart is the judgement the whole number rests on, and the accounts will not do it for you. A rough, honest first check: compare a few years of total machine-spending against depreciation. If a business spends far more on machines than the depreciation it books, it is usually growing, and some of that is a choice. If its machine-spending is about equal to depreciation, it is mostly just maintaining itself.

The reward for this work is that owner earnings, not the reported profit, is the number you should use to judge what a business is worth. Buffett's point is that a business is worth the stream of owner earnings it will produce, and that two businesses with the very same reported profit can be worth completely different amounts if one keeps its cash and the other must feed it all back into machines.

Take two made-up Indian businesses that both report the same net profit of ₹100 crore. illustrative One is a light services firm - think a tuition company, few machines. The other runs heavy plant that wears out fast. Watch the bold number stay equal while the cash the owner can actually pocket splits apart.

Read the two right-hand columns as a warning. On EBITDA, Ironworks looks better - ₹160 crore against ₹115 crore - exactly because EBITDA adds back the heavy depreciation and then conveniently forgets that this depreciation is a preview of cash the business must spend again, and more, to replace those machines. Put the machine-spending back in, and the order flips: the light services firm hands its owner ₹95 crore of truly free cash, the heavy-machine firm only ₹75 crore, and the gap only widens over time as Ironworks' machines keep coming due. Same headline profit. Very different businesses to own.

The keep-it-running number is a guess, and guesses can be nudged. Because nobody tells you this number, you must estimate it - and an optimist estimates it low, making owner earnings look generous. An owner keen to show off big free cash can quietly under-spend on upkeep for a few years, letting shops get tired and machines get old; cash looks wonderful right up until all the delayed spending arrives at once. So reading owner earnings means also reading whether the business is truly staying fresh or quietly running itself down.

Growth spending hides inside upkeep, and the other way round. A company growing fast will show machine-spending far above depreciation, and it is tempting to call all the extra "growth" you can ignore. But some of what looks like growth is really higher upkeep on a bigger set of machines, and some businesses must keep spending just to stay relevant. Split it wrong and the whole number is wrong.

Cash stuck in the business, and one-off years. Buffett's version also notices cash trapped in a fast-growing business - money tied up in more stock on the shelf and more bills customers have not yet paid. And a single unusually good or bad year twists everything, which is why owner earnings should be read as an average over several years, never off one year alone.

It is a rough figure, and false precision is a trap. The honesty of this number is in its range, not its decimal point. Anyone quoting owner earnings to the exact rupee has mistaken a judgement for a measurement. The discipline is to be roughly right about the cash, not exactly right about the accounting.

Owner earnings is one of the most portable ideas Buffett gave - it works on any business in any country, because every business, everywhere, must one day replace its machines. What does not carry over cleanly is how easy the guess is. In markets that share a lot of detail, you can sometimes find management's own split between keep-it-running and growth spending. In many Indian filings you cannot, and the numbers by segment may be rough or missing. So the Indian reader gets the idea whole but must do more of the estimating themselves - reading several years of machine-spending against depreciation, listening on company calls for what management says its normal upkeep spend is, and treating a business that will not discuss this difference as one to read more carefully, not less.

• Start from net profit, not EBITDA. If someone sells you a business on EBITDA, ask at once what it spends every year just to stay level - and be careful if there is no clear answer.
• Add back the non-cash costs, then hunt for the real yearly keep-it-running spend.
• Compare machine-spending to depreciation over five to ten years. Spending well above depreciation means growth (some of it a choice); spending about equal to depreciation means the business is mostly just maintaining itself.
• Watch for under-spending on upkeep. Free cash that suddenly looks wonderful may be a business putting off spending it will owe later.
• Read it as an average, over several years, and quote it as a range, never an exact figure.
• Judge value on owner earnings, not reported profit - two equal profits can hide very unequal cash.