IndiGo is buying long-haul capacity from a position with no margin cushion

As reported, EBITDAR margin fell from 26.3% to 17.8%, a drop of 8.5 points. Excluding forex on dollar lease liabilities it rose, 26.3% to 27.3%. The gap between the two is about nine and a half points of revenue and it is entirely non-operating.

Both are published here because publishing either alone is an error this project has already committed once. An earlier version of the site claimed the operating margin had halved, from a convention the company does not publish; that claim was withdrawn and the retraction is still in the methodology. Quoting only the flattering ex-forex figure would be the same mistake pointing the other way.

financials.margin_ladder() returns four rows, both bases across both years, so neither figure can be read out of the module without the other. A test asserts it.

Three bases, because the middle one is where the year hides. All-in unit cost rose 0.34. Ex-fuel it rose 0.52, which looks worse. Ex-fuel and ex-forex it rose 0.11, which is what the airline actually did to its own cost base. The 0.41 in between is the rupee.

Two different quantities in this project both equal 3.00: FY2025 CASK ex-fuel, and FY2026 CASK ex-fuel ex-forex. Confusing them collapses the bridge, which is why the exhibit names the basis on every bar rather than relying on the reader to track it.

These are the only two yields this project has verified against a primary source, and they are what the contested-pool band is built from: the floor prices the leaked traffic at IndiGo’s own realisation, the ceiling at Emirates’.

Emirates publishes passenger yield directly, at 38.1 fils per revenue passenger kilometre, converted at a dated reference rate. IndiGo’s is from its own results release. No proxy, no blend.

The gap bounds the prize; it does not measure a connect premium. Emirates carries substantial premium cabins where IndiGo is all-economy, and yield per kilometre normally falls with stage length, so some of the difference is cabin mix and network shape rather than pricing power. Air India, the third carrier that matters here, publishes nothing.

The commitment is roughly a year of earnings expressed as annual revenue potential, which is the honest order of magnitude for a decision of this size. It is scale, not a financing plan: how the aircraft are funded changes who owns the metal, not where it should fly.

No aircraft price appears anywhere in this exhibit, and a test fails if one ever does. A list price is not a transaction price and transaction prices are commercially confidential, so scale is expressed in capacity and in revenue at realised unit revenue instead. The seat count is read off the same gated order-book table the capacity sizing leg uses, so the two cannot diverge.

The utilisation basis is the owned fleet, 10.06 hours a day. An active-fleet basis would be higher and would raise this figure by roughly a third. The conservative basis is used and named, because the widely quoted 13 hours a day does not survive arithmetic: it requires 100 of 441 aircraft grounded.

The second both-ends cross-check in this project, after DGCA against Eurostat. It matters because the utilisation figure it underwrites is what converts an order book into capacity, and capacity is what the whole absorption argument runs on.

Run on the financial year, April to March, not the calendar year, because DGCA publishes monthly and IndiGo reports to March. Comparing two different twelve-month windows would have produced a spurious gap.

The published headline is 0.31%, on scheduled services alone. That residual turned out not to be measurement error but non-scheduled international flying sitting outside the filter. Both bases are reported rather than the flattering one alone.