Variant Perception

Where We Disagree With the Market

The market is still pricing IOC's LPG under-recovery as a permanent P&L drag when the Centre has now codified it into a sovereign-guaranteed, calendarized receivable — and that single reclassification dissolves most of the structural bear case. Consensus has dispersed into a ₹150–₹195 range (a 30% spread on a Nifty-50 PSU) without resolving the central question: is IOC a refiner mispriced at trough multiples, or a regulated utility mispriced at fair? Our read is that the dispersion is itself the finding — both sides are anchoring on the wrong variable. The real disagreement is mechanism, not magnitude. The cleanest way this gets resolved is the Q4 FY26 print on ~21–28 May 2026 followed by Panipat/Gujarat commissioning in June, because either landing reframes the LPG accrual treatment and the BPCL ROCE-gap story in the same window.

Variant Perception Scorecard

The variant strength score reflects three real disagreements with consensus, each tied to a dated, observable signal. Consensus clarity is moderate because the analyst range is unusually wide (₹150–₹195) but the implied price (₹142) clearly skews to the bearish end of that range. Evidence strength is high because the LPG framework has now been demonstrated twice (₹22,000 Cr FY23 + ₹30,000 Cr Aug-2025, plus the ₹14,486 Cr currently in monthly tranches), and the Q3 FY26 concall cancellation is a documented BSE filing — these are not narrative claims. The four-month resolution horizon is exact: Q4 FY26 print, three brownfield commissionings, and the FY26 Annual Report sign-off all land between mid-May and late August 2026.

Consensus Map

The consensus map shows where the market has internally settled and where it is genuinely split. The two issues consensus is most certain about (LPG as discretionary, BPCL as the right peer) are exactly the two we disagree with. The dispersion in target prices comes mostly from disagreement on margin sustainability and brownfield execution — but those are already arithmetic debates, not framework debates. Where the framework itself is wrong is in how analysts handle LPG.

The Disagreement Ledger

Disagreement 1 — LPG as a calendarized receivable. Consensus modelers strip LPG comp out as non-operating Other Income because the historical pattern was a single FY23 ₹22,000 Cr bailout. Our evidence is that the Centre has now done this twice (₹22k + ₹30k = ₹52k Cr cumulative), with the second tranche specifically calendarized as ₹1,207 Cr × 12 monthly payments running Nov-2025 through Oct-2026 — that is not a discretionary one-off, that is a mechanism. If we are right, IOC's ROCE on the cash-economic basis is meaningfully higher than reported, the BPCL gap is mechanically narrower than it appears, and the structural-cap reading (the Bear's central claim) loses its anchor. The market would have to concede that IOC is not a regulated utility absorbing permanent losses but a sovereign-cash-managed cyclical refiner where one liability moves to working capital. The cleanest disconfirming signal is a tranche pause inside FY26 — if any of the remaining six monthly payments doesn't land on time, the framework is discretionary, not codified.

Disagreement 2 — The cancelled Q3 concall. This is the variant view we hold with the lowest conviction but the highest information value because it is genuinely under-priced. On 5-Feb-2026 IOC reported a 4× PAT beat and cancelled the analyst call two hours after the print. The stock barely moved. Consensus would say either (a) PSU bureaucracy, or (b) coincidence with board reconstitution. Our read is that a confident management with a clean beat hosts the call — the cancellation reveals a question management did not want answered live, and the most plausible candidate is integrated margin sustainability (HDFC's FY26-28 +49/30/20% EPS revision rests entirely on the MoPNG channel check that the call would have tested). If we are right, the HDFC framework is too aggressive and fair value is closer to ₹150-160 than ₹190. The cleanest disconfirming signal is the Q4 FY26 board meeting on ~21-28 May 2026 — if management hosts the call AND defends Q3 margin sustainability with an unchanged forward posture, the variant is wrong and the cancellation was procedural.

Disagreement 3 — Stale through-cycle EPS math. Consensus benchmarks current 5.6× P/E against the 12-year median of ~9× and concludes IOC is "cheap." But the EPS the market is multiplying has changed character: the asset base has roughly tripled, capacity moves 80.75 → 98.4 MMTPA by FY27, and TTM PAT of ₹36,869 Cr is on the new base, not the old. At 5.6× × ₹26 the stock is fair, not cheap. The implication is that re-rating requires a real catalyst — either policy unlock (genuine LPG mechanism that gets repriced) or earnings step beyond TTM (brownfield commissioning earning brownfield economics) — not just multiple reversion. The market would have to concede that 5.6× is the regime, not the trough. The disconfirming signal is the Q1 FY27 print: throughput rises mechanically with Panipat live; if EPS rises with it (run-rate above ₹26 annualized), through-cycle has stepped up. If EPS stalls, today's price was already fair.

Evidence That Changes the Odds

The cleanest evidence row in the table is #1 — the calendarized LPG tranches. Two retroactive bailouts plus an explicit monthly schedule is no longer a discretionary policy stance; it is a mechanism. The HDFC channel check (#3) is the variant case in plain sight — if the MoPNG hands-off-margins read is right, every consensus model is too low. Evidence row #2 is the asymmetric one: the Q3 concall cancellation is either a procedural footnote or the most important under-priced signal of the cycle, and the Q4 print resolves the read in 17 days.

How This Gets Resolved

The eight signals collapse into three event windows: the Q4 FY26 print (~21-28 May 2026, four signals fire here), the brownfield commissioning window (June-August 2026, two signals), and the FY26 Annual Report sign-off (late August 2026, two signals). All three windows close within 120 days of today. This is the unusual case where a variant view is fully resolvable inside one fiscal half — the holding cost of patience is bounded.

What Would Make Us Wrong

The strongest case against our LPG-mechanism reading is also the simplest: the Centre has done this twice over three years, and "twice" is not yet a regime — it's a precedent that could break the moment a different government takes a different view. The same political machinery that codified ₹14,486 Cr in 12 monthly tranches can pause a tranche, change the accrual treatment, or convert the mechanism into a non-cash equity infusion at any point. If a single Nov-2025-to-Oct-2026 tranche fails to land on schedule, our calendarized-receivable reading collapses into the consensus framework that LPG is discretionary. We would have to concede that the Bear's structural-ROCE-cap reading is right, and that 1.0× book is fair, not floor. Watching this is mechanical: every quarterly Other Income line and every monthly cash-flow disclosure resolves it.

The strongest case against our Q3-concall-cancellation reading is that we are over-reading a procedural event. Maharatna PSUs cancel calls for many reasons that have nothing to do with sustainability — board reconstitution overlap, ministerial scheduling, regulatory restraint windows around budget season. If the Q4 FY26 print lands clean and management hosts a normal call defending Q3 margins, our "un-priced negative signal" reading was wrong, and the cancellation was procedural. We accept this as a near-coin-flip variant — held with the lowest conviction of the three precisely because the disconfirming signal is so cheap and so close.

The strongest case against our stale-through-cycle-EPS reading is that the asset base expansion has been visible for years and the market does aggregate-correct multiples in real time. If the brownfield wave delivers and run-rate EPS sustains above ₹26 with multiple stable, our "fair, not cheap" framing is moot — the cheapness was real but expressed through a 2-year lag. Either way, the Q1 FY27 print resolves it. We would also have to concede that the bull target (₹190+ in 12-18 months) was the right read all along, and that the multiple-band anchor was legitimately stale rather than legitimately structural.

The most important fact that breaks our entire variant frame in one shot is a single Cabinet decision that institutionalises the LPG framework into a formal pricing-pass-through formula AND confirms hands-off marketing margins on a recorded MoPNG statement. That is HDFC's variant case as the new baseline — and at that point our "calendarized receivable" reading is not variant anymore, it is consensus, and the price target moves to ₹190+ before we can act on it. The honest answer is that we would rather be early-and-wrong on the framework than late-and-right on the price.

The first thing to watch is whether IOC hosts the Q4 FY26 conference call on ~21-28 May 2026 and defends the Q3 integrated margin live — that single binary event tests the un-priced concall-cancellation signal, the LPG accrual treatment, and the through-cycle-EPS sustainability in one room.