Hindalco Q2 Record: The Non-LME Half of EBITDA Was Flat as a Share, and Copper, Currency and Insurance Lifted It

Hindalco's consolidated revenue from operations for July-September 2026 was ₹91,708 crore against ₹66,058 crore a year earlier (up 39%, the company's figure), EBITDA ₹13,757 crore (up 42%) and profit after tax ₹7,533 crore against ₹4,741 crore (up 59%). The company says "the non-LME-linked businesses contributed almost 50%" of earnings. On the company's own segment table that share is 49.5% of business-segment EBITDA, against 50.3% a year ago (our arithmetic): the record was broad, not a shift in mix. What lifted the non-LME group quarter on quarter was copper, and the copper line carries by-product and "one-time" gains the company does not size. The auditors' conclusion on both statements is unmodified.

The idea in one sentence

A label such as "non-LME-linked" tells a reader what an earnings line is priced off, not what it is made of; in this quarter the group labelled that way matched the rest of the company in growth; its year-on-year gain was helped by rupee translation and an insurance recovery at Novelis, and its quarter-on-quarter lift came almost entirely from copper, which carries by-product and one-time gains the company does not size and a rebound after a planned shutdown in Q1, so the quarter is a weaker test of a more stable earnings base than the headline suggests (our reading).

The fact that breaks the assumption

The press release says: "Together, the non-LME-linked businesses contributed almost 50% to the company's earnings reflecting sustainable growth despite LME volatility and global macroeconomic headwinds." The presentation (slide 17) defines the group: "Non-LME linked business defined as Novelis, Copper and Aluminium downstream whose earnings have low correlation with LME." Its EBITDA was ₹6,860 crore, an all-time high, up 50%.

Now put it next to the segment table ([RS], segment results; [IP] annexure). Business-segment EBITDA was ₹13,869 crore against ₹9,104 crore, up 52% (company figure). Novelis ₹4,858 crore, Copper ₹1,673 crore and Aluminium Downstream ₹329 crore make ₹6,860 crore, which is 49.5% of ₹13,869 crore. A year earlier the same three were ₹3,685 crore, ₹634 crore and ₹261 crore, ₹4,580 crore of ₹9,104 crore, or 50.3% (our arithmetic). The LME-linked Aluminium Upstream segment, EBITDA ₹7,009 crore against ₹4,524 crore, supplied ₹2,485 crore of the ₹4,765 crore year-on-year rise in segment EBITDA, 52% of it (our arithmetic). The non-LME group is as large a share as before, not a larger one, and half of this quarter's growth still came from the line priced off the London Metal Exchange. It is sequentially a different story: in Q1 FY27 the group was ₹6,091 crore of ₹13,480 crore, 45.2% (our arithmetic), because the upstream EBITDA fell ₹381 crore quarter on quarter to ₹7,009 crore while the group rose ₹769 crore, which is the figure to follow.

One concrete thing, followed: Copper's ₹1,673 crore

Of the ₹769 crore quarter-on-quarter rise in the non-LME group, ₹755 crore is Copper (₹918 crore to ₹1,673 crore); Novelis was ₹16 crore lower (₹4,874 crore to ₹4,858 crore) and Aluminium Downstream ₹31 crore higher (our arithmetic from the segment table). So the "up 13% sequentially" the deck prints (slide 17, "despite 10% decline in LME prices") is almost entirely a copper story.

What the filings say about copper. Metal sales were 116 kt against 113 kt, up 2%, and continuous cast rod sales were 96 kt against 97 kt ([IP] media release; slide 25). Revenue rose 47% to ₹21,374 crore "on account of high metal prices and better by-product realisations". EBITDA rose from ₹634 crore to ₹1,673 crore. That is a rise of ₹1,039 crore on roughly flat volume, and EBITDA per tonne went from $641 to $1,510 (slide 18). The press release gives the reasons for the record: "strong operational performance, and healthy domestic demand across segments despite elevated LME prices and higher realisation of by-products like sulphuric acid and one-time gains". Slide 25 adds, beside the per-tonne chart: "↑ 63% QoQ on account of higher LME and by-product realization, includes one-time gain". And the quarter-on-quarter rise of 82% in rupee EBITDA is put down by the deck to "normalization post planned smelter shutdown in Q1", with metal shipments up 10% on the quarter (slide 25).

So the company's own text lists three things inside the copper number besides operations: "higher LME", by-product realisations, and a one-time gain. It sizes none of them. Copper EBITDA on revenue was 7.8% against 4.4% a year ago (our arithmetic). Our reading: a business described as having "low correlation with LME" earned its record in a quarter when the deck itself credits "higher LME" for part of the per-tonne gain. The relevant question is not whether copper is good, but how much of ₹1,673 crore a reader may treat as repeatable. The filings cannot answer it.

The mechanism: what "non-LME" does and does not remove

Our reading of the structure: a downstream converter earns a conversion margin on top of a metal price, and a metal price that passes through to customers moves revenue (consolidated revenue is up 39% here; Novelis revenue is up 25% in US dollars, "driven by higher metal prices", per the release) far more than EBITDA. That is the sense in which Novelis, Copper and Aluminium Downstream are less LME-sensitive than a smelter. It does not make their earnings independent of anything else. Three other drivers sit in this quarter's numbers, and none of them is operating volume or margin in the ordinary sense.

Currency. The deck says (slide 16): "Novelis EBITDA benefited from favorable INR/USD exchange rate movements. As a result, EBITDA grew by 32% in INR terms versus 20% growth in USD terms". Novelis' adjusted EBITDA was $508 million against $422 million (the company says up 20%) on shipments of 919 kt against 941 kt, down 2%. Had the rupee EBITDA grown only at the dollar rate, the segment would be about ₹4,436 crore, and about ₹422 crore of the ₹1,173 crore rise is therefore translation (our arithmetic, approximate: the rupee segment figure and the US GAAP dollar figure are not on identical bases).

Insurance. Slide 22 says Novelis' adjusted EBITDA is "up 20% YoY, incl. $39 Mn in Oswego insurance recoveries and benefitted by better scrap spreads and higher recycling". Note 6 of [RS] confirms business interruption recoveries of ₹370 crore (US$39 million) booked in other income for the quarter, and a separate net recovery of ₹284 crore (US$29 million) in exceptional items. Excluding the $39 million, EBITDA would be $469 million, 11% above $422 million (our arithmetic), if the prior-year figure held no such recovery, which the deck does not say. The "Novelis recovery" the company names as the first driver of the record shows in the sequence too: EBITDA per tonne was $563 in Q1 and $553 in Q2, and rupee EBITDA was flat quarter on quarter (₹4,874 crore to ₹4,858 crore). Note 6 of [RS] puts business interruption recoveries at ₹819 crore (US$86 million) for the six months against ₹370 crore (US$39 million) for the quarter, so about ₹449 crore (US$47 million) fell in the June quarter (our arithmetic): the insurance recovery was smaller in September than in June, and Novelis' rupee EBITDA fell slightly on the quarter.

Tax. Slide 16 says: "This increase in PAT is aided by HIL's transition to the new tax regime effective 1st April 2026. Consequently, Effective Tax Rate reduced to 25% in Q2 FY27 versus 29% in Q2 FY26." The India-business statement in [IP] gives tax of ₹1,969 crore on profit before tax of ₹7,821 crore, 25.2% (our arithmetic), against ₹1,257 crore on ₹4,317 crore, 29.1%. Standalone, tax expense was ₹1,734 crore (current ₹1,708 crore plus deferred ₹26 crore) on profit before tax of ₹6,866 crore, 25.3%, against 30.3% (our arithmetic). At last year's 29.1% rate, India-business profit would have been about ₹308 crore lower, and consolidated profit after tax would have grown about 52% rather than 59% (our arithmetic, holding everything else). The consolidated effective rate was 28.9% against 27.5% (our arithmetic), because, our reading, the Novelis side carries its own rate. The company presents the lower rate as a regime transition effective 1 April 2026, so our reading is that it should recur in later quarters; it is a tax effect, not an operating gain.

Taken together, our reading is that the non-LME group's record is a measure of three things at once: genuine operating improvement (Novelis' structural cost cuts, which the deck puts at "Over $240 million in run-rate cost savings though Q2 FY27"; Aluminium Downstream's 119 kt, up 6%, and EBITDA per tonne up 9% to $289), a commodity-adjacent windfall in copper, and translation and insurance effects at Novelis. The filings do not let a reader weigh them.

Where this breaks: the bear case, and the case against it

The strongest argument against our thesis is in the company's own sequence. Non-LME EBITDA has risen three quarters running (₹3,930 crore, ₹5,364 crore, ₹6,091 crore, ₹6,860 crore on slide 17; the Q2 FY26 figure was ₹4,580 crore, so Q3 FY26 was a fall, our arithmetic), and the last step came with LME falling. The deck's chart (slide 12) gives the aluminium LME at $3,576 per tonne in Q1 FY27 and $3,227 in Q2, a fall of 9.8% (our arithmetic), and $3,071 for Q3 "till date". If the group had the correlation of a smelter, its EBITDA should have dropped in Q2; the segment that dropped was Aluminium Upstream (down 5%). Net of the copper by-product and "one-time gain", the group may still have risen; the filings do not size them to test this.

Our share arithmetic is a segment-level view and the segments are not the whole group. The company's consolidated EBITDA of ₹13,757 crore differs from segment EBITDA of ₹13,869 crore by unallocable items and eliminations ([RS] segment note), and consolidated EBITDA fell 8% quarter on quarter from ₹14,989 crore (the Q1 figure carried an unallocable income of ₹1,992 crore). A different denominator would give a different share. We use the company's own ₹6,860 crore and its segment total, and say so.

One quarter does not make a pattern. The 50.3%-to-49.5% comparison is a single year-on-year pair; the quarter-on-quarter jump from 45.2% to 49.5% shows the share is volatile, not trending. We have no longer series for the group.

The upstream side is the profit engine, and it is where the cost-advantage moat would show. Aluminium Upstream EBITDA margin was 52% against 45% ([IP] media release and slide 23), on EBITDA per tonne of $2,125 against $1,521. The company says it was "driven by favourable macros, higher shipments and stronger operational performance". The filings give no cost-per-tonne, so they cannot separate a lower-cost position from a higher price. The deck itself says the per-tonne rise came "owing to favorable macros partially offset by increase in input costs" (slide 23). Our reading is that a smelter whose margin moves from 45% to 52% on favourable macros is showing the price, and that this quarter shows nothing about the cost curve.

Cash does not match the earnings headline. The deck's slide 20 (a bridge from 31 March to 30 September 2026) reports "Healthy Cash flow from operations at ₹23,319 crore (↑ 57% YoY)" and "Net Cash inflow of ₹10,231 crore", and net debt to EBITDA of 1.80x against 1.23x a year ago. The statutory consolidated cash flow statement in [RS] shows net cash generated from operating activities for the six months of ₹2,825 crore against ₹8,016 crore, down 65% (our arithmetic), after an inventory build of ₹8,279 crore and a trade payables reduction of ₹4,634 crore, with payments for property, plant and equipment and intangibles of ₹19,344 crore against ₹10,685 crore (up 81%, our arithmetic). The deck's net debt bridge goes from ₹64,841 crore at 31 March 2026 to ₹80,435 crore at 30 September 2026, up ₹15,594 crore (our arithmetic), with capex of ₹20,719 crore. The company ties the higher working capital to "rate impact of higher metal prices". The two cash numbers use different definitions that the filings do not reconcile.

Review, not audit. The limited review conclusion is unmodified; a review is "substantially less in scope than an audit", and the subsidiaries covered by other auditors' reports carry total assets of ₹262,204 crore and quarterly revenue of ₹57,915 crore (paragraph 7), which is most of the group (our reading).

Why it costs the reader something

A reader who takes "almost 50% non-LME" as a floor under earnings is assuming the ₹6,860 crore repeats. This quarter it contains unsized copper by-product and one-time gains (copper added ₹755 crore of the ₹769 crore sequential rise, which the deck also ties to normalisation after a planned smelter shutdown), about ₹422 crore of rupee translation at Novelis (our arithmetic), and $39 million of insurance recoveries. The record is real, and the review report does not qualify it. But the quantity that matters for a moat reader, the part of earnings that does not move with metal prices, rates or one-offs, cannot be read from these two documents. Capital that is sized to that number is sized to something the company has not measured. The cash side points the same way: operating cash flow was lower and net debt higher, while the deck puts capex of ₹20,719 crore "towards strategic growth projects"; Satish Pai, Managing Director, says Bay Minette, copper recycling, aluminium battery foil, Aditya FRP and Inner Grooved Tubes "will further accelerate the share of our non-LME-linked businesses" and that the aluminium refinery and smelters, copper smelters and coal mine projects "remain on track". The deck's own guidance is a test to watch, not a result: "Long-term $600+/ton EBITDA guidance intact" against $553 this quarter at Novelis.

The concrete thing, again

Copper's ₹1,673 crore on 116 kt is the record the release leads with, and the line the sequential non-LME growth hangs on. The company's own slide says it "includes one-time gain" and credits "higher LME and by-product realization". Until a document gives the size of the one-time gain and the sulphuric acid effect, the right reading of Hindalco's quarter is that the non-LME half held its place in a good year for metals, and that the proof it is more stable than the LME half has not arrived.

Sources

• [RS] Hindalco Industries, outcome of the board meeting, results and review reports, 6 Nov 2026 (NSE)
• [IP] Hindalco Industries, media release and Q2 FY27 earnings presentation, 6 Nov 2026 (NSE)
• [AR] Hindalco Industries, audio recording notice for the Q2 FY27 earnings call, 6 Nov 2026 (NSE)
• Earlier Moat & Margin page on the company: Hindalco Industries Ltd., first-pass moat teardown

Educational research, not investment advice. Every figure above is drawn from Hindalco Industries' own filings of 6 November 2026 (results and review reports, media release and earnings presentation, and the audio-recording notice; no call transcript was read), read from the primary documents; derived figures are our arithmetic, labelled as such. MoatSCORE: 5.74 (read live on the Moat Screener on 8 Nov 2026, where it is labelled first-pass). It is provisional. No buy/sell recommendations, no price targets, no share prices. Moat & Margin is not a SEBI-registered Research Analyst or Investment Adviser.