EBITDA Margin Explained: What It Hides

Two companies closed the twelve months to December 2025 with almost the same EBITDA margin. Colgate-Palmolive India earned ₹1,858 crore of operating profit on ₹5,903 crore of sales — a 31.5% margin. PVR Inox earned ₹1,914 crore on ₹6,422 crore — 29.8% (Inve data, TTM to Q3 FY26). Read only the EBITDA line and you'd call them cousins. They are not cousins. Colgate kept ₹1,328 crore as net profit; PVR Inox kept ₹22 crore. One business converted 22.5% of its sales into owner profit, the other 0.3% — and the gap is almost entirely the two costs EBITDA is built to ignore (Inve data, TTM to Q3 FY26). Illustration of a metric, not a view on either stock.

Nobody lied. EBITDA margin simply does what it is designed to do — it shows profitability before the three costs that decide whether a business actually generates cash: depreciation, interest, and tax. Across 15,726 management commitments tracked on Inve over 1,547 listed Indian companies, the average kept rate is about 55% (Inve data, as of 2026-06-12), and margin-expansion guidance is one of the more commonly dropped classes of commitment. "EBITDA margin will improve to 18% by FY27" is easy to say, easy to applaud, and easy to quietly stop mentioning. This article is about why the metric flatters, what it hides, and how to read it without being fooled — using one real, awkward, well-documented case rather than a worked-up hypothetical.

What does EBITDA margin actually measure?

EBITDA is Earnings Before Interest, Tax, Depreciation, and Amortisation. EBITDA margin expresses it as a share of revenue.

The idea has a legitimate use, and it's worth stating fairly before we attack it. By stripping out financing decisions (interest), tax regimes, and the bookkeeping of past capital spending (depreciation and amortisation), EBITDA margin tries to isolate the cash-generating power of the core operations — letting you compare two firms in the same business without their debt loads and tax situations getting in the way. A leveraged-buyout analyst comparing two cement plants genuinely wants that view.

The problem is what gets stripped out. Depreciation is not a fiction — it is the annual cost of plant, machinery, fit-outs and equipment wearing out and needing replacement. Interest is not optional — it is real money leaving the building every quarter. By parking these "below the line," EBITDA margin lets an asset-heavy, debt-funded business present the same clean face as a capital-light one — which is exactly why it can mask the warning signs of a debt-trap stock. Colgate and PVR Inox are not the same business, and the owner's cash says so even when the EBITDA line won't.

Why is EBITDA margin so flattering for asset-heavy companies?

Think of EBITDA as the rent you collect on a flat before you've paid the home loan EMI, the society maintenance, and the cost of repainting every few years. It looks like a handsome yield — right up until you remember the flat came with a loan and the paint peels. EBITDA is the rent; depreciation and interest are the EMI and the repainting. Quote only the rent and any landlord sounds rich.

PVR Inox is the multiplex version of that flat. It runs cinemas it largely leases, fits them out at heavy cost, and writes those fit-outs and lease rights down year after year. In the twelve months to December 2025 it carried ₹1,258 crore of depreciation and ₹756 crore of interest (Inve data, TTM to Q3 FY26) — together ₹2,014 crore, more than its entire ₹1,914 crore of EBITDA. That is the whole story in one comparison: the cost of owning and financing the assets slightly exceeded the operating profit those assets produced. After both, profit before tax was ₹23 crore on ₹6,422 crore of sales (Inve data, TTM to Q3 FY26). The 29.8% EBITDA margin was real. It was also almost entirely consumed by the cost of being in that business.

Charlie Munger put the objection bluntly: every time you see the word EBITDA, he said, substitute "bullshit earnings," because depreciation is a real expense — the machine genuinely is wearing out. That is harsh, and it is unfair to capital-light businesses where the gap is small. But for a leased, fitted-out, depreciation-heavy model, the arithmetic is on his side, and PVR Inox's own statement is the proof.

What does EBITDA margin hide, line by line?

Here are the two companies side by side, the same twelve-month window for both, walked down from EBITDA to what the owner actually keeps. Every figure is from Inve's parsed financials.

Real companies, illustrating a metric — not a recommendation on either. TTM = the four quarters Jan–Dec 2025 (Q4 FY25 through Q3 FY26).

Source: Inve data (financialreport and balancesheet), TTM to Q3 FY26; fixed assets as at 2025-09-30. PVR Inox fixed assets include right-of-use lease assets under Ind AS 116.

Both wave a roughly 30% EBITDA margin in the headline. By the time you reach the bottom line, Colgate keeps 22.5 paise of every revenue rupee and PVR Inox keeps a third of one paisa. Look at the fixed-asset row to see why: ₹742 crore of assets behind Colgate's margin, ₹13,633 crore behind PVR Inox's (Inve data, 2025-09-30). The same operating profitability, sitting on eighteen times the asset base — and that base has to be depreciated and financed every single year, whether the films work or not.

This is the single most useful habit: never read an EBITDA margin without asking what does it cost this company to keep its EBITDA from shrinking? For a toothpaste brand, very little. For a leased multiplex, very nearly all of it.

When a company quotes you two EBITDA margins for the same quarter

Here is the part that should make you slow down. On its Q3 FY26 call, PVR Inox did not report a 30% EBITDA margin at all — management reported 18%. Both numbers are "EBITDA margin." Both describe the same three months. They differ by roughly 1,200 basis points.

The reconciliation is in one line of the transcript. "The following numbers were calculated after adjusting for the impact of IndAS 116 on lease accounting" (PVR Inox Q3 FY26 concall). Ind AS 116 forces a company to move lease rentals out of operating costs and re-cast them as depreciation and interest — which mechanically inflates reported EBITDA, because rent that used to sit above the EBITDA line now sits below it. The ~30% figure in our table is the as-reported, Ind-AS-116-inclusive number (the one a screener shows you). The 18% is management's own number after stripping the lease accounting back out (Inve data, quarterly_kpi, Q3 FY26: "18% EBITDA margin at 28.5% occupancy").

Notice which way each party leans. The statutory accounts carry the higher figure; management, talking to analysts, reaches for the lower, "cleaner" one — because at 18% it can make a fairer point about the underlying business. That's the honest use of an adjustment. But it leaves the retail reader with two defensible EBITDA margins for one quarter and no warning that they aren't comparable across companies unless you know which lease convention each used. If a single metric can legitimately read as 18% or 30% depending on a footnote, treating it as the measure of profitability is a category error.

How do you read EBITDA margin without being fooled?

A few disciplines turn the metric from a trap into a tool.

Pair it with the cash-flow line. A rising EBITDA margin alongside a falling free-cash-flow margin (free cash flow ÷ revenue) is the classic divergence — reported operating profitability improving while the cash the owner can take out is not, usually because capex, interest, or working capital is eating the difference. When the two move in opposite directions for several quarters, trust the cash line. We cover this gap in detail in our piece on quality of earnings.

Watch the gap between EBITDA margin and net margin. It is the combined weight of depreciation, interest, and tax. For Colgate the gap is 9 points (31.5% to 22.5%); for PVR Inox it is the near-entirety of the margin (29.8% to 0.3%) (Inve data, TTM to Q3 FY26). A small gap is a capital-light, low-debt business; a chasm is an asset-heavy or leveraged one where the headline barely reaches the owner. (For how the margins above EBITDA behave, see gross vs operating margin.)

Distinguish maintenance capex from growth capex. Not all spending is equal. Building new screens is investment; stopping existing ones from falling apart is just the cost of staying open. EBITDA ignores both, so a business whose capex is mostly maintenance has a far lower true margin than its EBITDA implies.

Treat "adjusted EBITDA" with extra suspicion — and ask which way the adjustment leans. Used properly, adjusting strips out genuine non-operating income and truly one-off items. PVR Inox flagged a ₹44.6 crore one-time labour-code provision in Q3 and placed it below EBITDA as an exceptional item (PVR Inox Q3 FY26 concall) — that is the legitimate version. The abuse is the opposite: recurring costs dressed up as one-offs. Restructuring charges that appear in four consecutive annual reports are not one-offs; they are a cost the company would prefer you ignored.

Where management guidance meets the EBITDA margin

Now the part that matters for a long-term owner, because EBITDA margin and the return it eventually produces are easy to talk about and hard to be pinned to.

On the same Q3 FY26 call, an analyst pushed the CFO on returns, not margins. "What is the ROCE that you said you achieved in nine months?" The reply: "I think we will be in single digit, in the high single-digit number adjusted for goodwill, if you look at the overall capital employed." The analyst answered with a fact from his own screen: "So the screener that I used, we were at 18% in FY'19." The CFO: "Correct. So we are improving... the trajectory is upwards" (PVR Inox Q3 FY26 concall). Sit with that exchange. The same business that the EBITDA line presents at ~30% earns a high-single-digit return on capital today, against roughly 18% before the pandemic — and the gap is precisely the depreciation and interest on all those leased, fitted-out screens that EBITDA waves away.

Management's forward framing was a soft target on the same call: return on capital "will definitely be in double digits very soon" (PVR Inox Q3 FY26 concall). That is exactly the kind of guidance worth marking and revisiting, because it has no quarter attached and no number beyond "double digits." Inve's Promise Tracker records that commitment with its birth quarter and original quote, alongside the rest of the company's guidance history. It also shows what already went quiet: PVR Inox guided a film-hire cost of 44.5%–46% in Q4 FY25, then stopped reporting against it on the next two calls — guidance that simply went silent (Inve data, Promise Tracker, PVRINOX). Illustration of how guidance is tracked, not a view on the stock. A margin screen would never surface that disappearance; you have to track the commitment itself.

The useful move is to read the two together: the EBITDA margin tells you what the income statement wants you to see, and the delivery record tells you whether the return story is one management keeps returning to with numbers, or one they raise once and let fade.

Where this lens could be wrong

The strongest case against dismissing PVR Inox on its bottom line goes like this, and it deserves an honest hearing. The lease-heavy model is a deliberate, capital-light pivot: leasing screens rather than owning them keeps cash capex low and lets the chain expand on landlords' balance sheets, and Ind AS 116 mechanically converts that rent into depreciation and interest — so the very costs crushing net profit are partly an accounting artefact of a sensible strategy, not pure value destruction. Management's point that it now earns the same 18% underlying EBITDA margin at 28.5% occupancy that it once needed 32% to reach (PVR Inox Q3 FY26 concall) is a genuine efficiency gain, not spin. And a cinema chain at the bottom of a weak-content cycle, carrying a one-time ₹44.6 crore provision, is showing you a trough year — judge it across a cycle, not at its low.

All true. None of it changes the reading; it sharpens it. The pivot is real, and that is exactly why you watch the return on capital climb back toward its FY19 level rather than trusting the 30% EBITDA line today. The accounting is an artefact — but the lease liabilities and the ₹756 crore of interest are cash, artefact or not. The point was never "asset-heavy is bad." It is that EBITDA margin cannot tell you which of these two readings is true, and only the cash and the return on capital can.

How to use EBITDA margin in your own screening

A repeatable sequence keeps the metric honest:

1. Compute EBITDA margin and net profit margin side by side. A wide gap tells you how much depreciation, interest, and tax are taking. The wider it is, the less the EBITDA margin is worth alone.
2. Layer in free-cash-flow margin. The real test. If EBITDA margin is rising but FCF margin is flat or falling, ask where the cash is going — capex, interest, or working capital.
3. Sector-adjust. A 22% EBITDA margin is unremarkable for software and excellent for a commodity producer. Compare within the sector, never across.
4. Check the capex and lease intensity. Capex and lease liabilities against five years of EBITDA tell you how much the business must reinvest and finance just to sustain itself.
5. Cross-reference the guidance. If management has guided to a margin or return target, check whether they keep reporting against it — or whether the Promise Tracker shows the commitment going quiet. The concall summaries on Concall AI surface the guidance per quarter, so you can see whether a target is being raised, maintained, or quietly revised away.

Frequently asked questions

EBITDA margin is not a bad number. It is a useful one, pointed at the wrong question by people who would rather you not look further down the statement. The honest reader treats it as the first line of an inquiry, not the last. Colgate and PVR Inox closed the year on the same EBITDA margin; one owner kept seventy times as much of each rupee as the other (Inve data, TTM to Q3 FY26). So the owner's question is not "how high is the EBITDA margin?" It is: what must I believe about this company's depreciation, its interest, and its return on capital five years out for that margin to ever reach me? When the margin rises and the cash doesn't, the margin is telling you what management hopes — the cash is telling you what's true.