EBITDA — Earnings Before Interest, Taxes, Depreciation, and Amortisation — is the most cited profitability metric in business finance. Analysts use it. Bankers use it to size loans. Founders use it to pitch valuations. Private equity uses it as the basis for LBO models. The metric is on every management presentation, every analyst note, and every deal term sheet.
It is also one of the most misleading numbers in finance — not because it is calculated incorrectly, but because it is used as a proxy for cash generation when it explicitly excludes several of the largest cash outflows a business makes. Warren Buffett called it "a fiction." Charlie Munger was less diplomatic. Understanding why — and what to use instead — separates clear financial thinking from institutional habit.
The four adjustments, and why each is questioned.
Interest. EBITDA adds back interest expense, making it a pre-debt-service profit measure. This is useful for comparing businesses with different capital structures (a leveraged buyout vs a debt-free startup). It is deeply misleading as a measure of absolute profitability because a business with ₹200 Cr of debt at 12% interest has a real, recurring cash obligation of ₹24 Cr per year. That cash goes to lenders, not to shareholders or reinvestment. EBITDA makes it disappear.
Taxes. Tax is a cash payment. The actual amount paid depends on the tax structure — and businesses with tax shields, deferred tax assets, and tax havens pay very different effective taxes on identical pre-tax earnings. Adding it back creates a pretax comparison, which is sometimes useful and often obscures how much cash actually leaves the business annually.
Depreciation. This is the central problem. Depreciation is the annual non-cash charge that allocates the cost of an asset over its useful life. A ₹100 Cr factory depreciated over 20 years creates ₹5 Cr of annual depreciation — a charge on the P&L with no cash outflow in the current period. Adding it back to profit creates EBITDA. The problem: the factory will need to be maintained, renovated, and eventually replaced. The cash outflow is real — it just occurs at a different time than the depreciation charge. For any capital-intensive business, depreciation is a deferral of capex recognition, not an absence of capex.
Amortisation. Similar to depreciation but applied to intangible assets. A business that acquires customers or technology through an acquisition takes an amortisation charge on the acquired intangibles. This charge represents the declining value of what was purchased. Adding it back pretends the acquisition was free.
EBITDA is a useful approximation for cash generation in asset-light businesses. It is a dangerous fiction for asset-heavy ones.
Why maintenance capex changes everything.
The most important adjustment EBITDA ignores is maintenance capex — the capital expenditure required to keep the existing asset base operational at current capacity. A steel plant, a cement factory, a power generation facility, a fleet of commercial vehicles: all require ongoing investment just to maintain current output. That investment is not optional. It is the cost of staying in business.
A ₹500 Cr EBITDA business with ₹300 Cr of annual maintenance capex generates ₹200 Cr of real, distributable cash. A ₹500 Cr EBITDA business in a capital-light sector (software, insurance, wealth management) with ₹20 Cr of capex generates ₹480 Cr. Both have identical EBITDA. Their actual cash generation is entirely different.
The EBITDA-based valuation multiple — "this business trades at 15x EBITDA" — is applied uniformly across both, producing a nonsensical comparison. The capital-light business at 15x EBITDA is priced at 15.6x FCF. The capital-heavy business at the same multiple is priced at 37.5x FCF. These are completely different financial propositions.
| Business | EBITDA | Maint. Capex | FCF | EV @ 15x EBITDA | FCF multiple |
|---|---|---|---|---|---|
| Capital-light SaaS | ₹100 Cr | ₹5 Cr | ₹95 Cr | ₹1,500 Cr | 15.8x FCF |
| FMCG company | ₹100 Cr | ₹25 Cr | ₹75 Cr | ₹1,500 Cr | 20x FCF |
| Steel manufacturer | ₹100 Cr | ₹70 Cr | ₹30 Cr | ₹1,500 Cr | 50x FCF |
| Airline (pre-Covid) | ₹100 Cr | ₹90 Cr | ₹10 Cr | ₹1,500 Cr | 150x FCF |
The hierarchy of cash-based metrics.
Operating Cash Flow (OCF) is the first upgrade from EBITDA. It captures the actual cash generated by operations — after working capital changes, before capex. It is harder to manipulate than EBITDA because cash either moved or it did not. OCF is available in every company's financial statements and should be the default check on reported profitability.
Free Cash Flow (FCF) = OCF minus total capex. This is the cash available after maintaining and growing the business. The problem with FCF is that it does not separate growth capex (optional, discretionary, funded by investor capital) from maintenance capex (mandatory to survive). For a growing company, FCF will systematically understate the cash available to mature investors, because growth capex is included in the deduction.
Owner Earnings — Buffett's preferred metric — is net income plus depreciation/amortisation minus average maintenance capex. It strips out the non-cash fiction of depreciation and replaces it with the real cost of staying in business. It requires a judgment about maintenance versus growth capex, which is why it is less standardised. But that judgment is exactly the exercise that forces clear thinking about what a business actually generates for its owners.
The cases where the metric earns its keep.
EBITDA is not useless. It has two legitimate applications. First, as a debt capacity proxy: banks use EBITDA to size loans because interest and debt service are paid before tax, depreciation is non-cash, and the bank wants to know whether the business generates enough operating cash to service the debt. The DSCR (Debt Service Coverage Ratio) is EBITDA divided by debt service obligations — a reasonable operational measure that does not require capex assumptions.
Second, as a cross-company operating comparison tool for businesses at different stages of their capital structure or tax optimisation. Comparing the operating profitability of a highly leveraged private-equity-owned business to an unlevered listed company requires stripping out the financing differences — EBITDA does this cleanly.
The discipline is to use EBITDA only where it is valid — operating comparisons and debt sizing — and to use FCF or Owner Earnings where capital intensity matters. Most investors use EBITDA everywhere, for every business, as a valuation anchor. That is the error. Not the metric. The application.
Four questions before accepting an EBITDA number.
The EBITDA reality check
- What is the maintenance capex? Ask management directly. If they cannot answer, or if the answer is "all capex is growth capex," apply a benchmark from industry peers. For heavy manufacturing, assume maintenance capex of at least 80% of depreciation. For asset-light businesses, 20–30% is reasonable.
- What is the OCF to EBITDA ratio? For a capital-light business, it should be 75–95%. For a working-capital-intensive business, 50–70%. Below 50% consistently means either working capital is absorbing profitability or accounting is being aggressive on revenue recognition.
- Are there add-backs in the reported EBITDA? Adjusted EBITDA that adds back stock-based compensation, restructuring charges, or "one-time" items deserves scrutiny. If one-time items recur annually, they are not one-time. Stock compensation is a real cost — the dilution is paid by existing shareholders even if no cash leaves the company.
- What is the working capital trend? Rising receivables and inventory as a percentage of revenue reduce OCF relative to EBITDA. A business reporting EBITDA growth while OCF is flat or declining is funding its growth through balance sheet expansion, not earnings quality improvement. That is a short-term source of apparent profitability that cannot persist indefinitely.
EBITDA is a tool. Like all tools, it works when used for the right job. The right job is not "what is this business worth." It is "how much operating cash did this business generate before we have a conversation about capital structure and capex requirements?" Start there — and then do the work that comes after.