How to Analyse a Metals & Mining Stock (Non-Ferrous)

In the year the zinc price spent mostly drifting sideways, Hindustan Zinc still pushed its cost of production to a five-year low of $959 a tonne for FY26 — below its own guidance of $1,000 — and the Q4 exit was $903 (Hindustan Zinc Q4 FY26 concall). That single number is why the company swung from roughly ₹4,000 crore of net debt in Q1 to a net cash position of ₹5,594 crore by year-end (Inve data, Q1 and Q4 FY26), and earned an ROCE management put at 58% (Hindustan Zinc Q4 FY25 concall). (Illustration of how to read the numbers, not a view on the stock.)

Notice what did the work. Not the metal price — that the company does not control. The cost of getting a tonne out of the ground and through the smelter. That is the whole secret of a metals and mining business: when you sell into a price set by a global auction in London you cannot influence, the only lever you own is being the lowest-cost producer left standing when the price falls. Everyone makes money at the top of the cycle. The question is who is still making it at the bottom.

This is how to analyse a metals and mining stock the way a cyclical analyst does — focused on non-ferrous (aluminium, zinc, copper) and mining (coal, iron ore), where the economics differ from a steel mill in ways the headline P/E will never tell you. For the ferrous spread business, read its sibling: how to analyse a steel company. Here we cover the metals whose price is quoted on the LME, and the miners whose product is the rock itself. On the Indian market that universe runs from diversified producers like Vedanta to iron-ore miners like Lloyds Metals & Energy.

A boundary first: you will not forecast the LME aluminium price, the silver price, or the China coal-import number, and neither can management. What you can read is whether a producer's cost position, its degree of integration, and its balance sheet are built to compound through a cycle it cannot time.

What actually drives the economics of a metals producer?

Picture two miners digging the same ore body. One owns its power plant, its coal mine, and its ore at the pit; the other buys electricity off the grid, coal on the market, and concentrate from a third party. When the metal price is high, both look like geniuses. When it halves, the first one is still profitable and the second is begging its bankers for covenant relief. The metal price is the same for both. The cost curve is everything.

Three forces fall out of that, and they govern the sector.

You are a price-taker on revenue and a price-maker only on cost. Realisation is the LME (or global benchmark) price minus or plus a small regional premium — handed to you by the market. So the entire competitive game is fought on the cost side: cheaper power, captive raw material, richer ore. A non-ferrous producer's moat is a position on the global cost curve, not a brand or a customer.

Integration is the cost moat. Aluminium is congealed electricity — power is roughly a third of the cost — so a captive coal mine and captive power plant are the difference between a survivor and a casualty. For a zinc miner the equivalent is owning a high-grade ore body and an integrated smelter. The more of the chain you own, the flatter your cost stays when the market panics.

Debt decides who survives the down-leg. Mines and smelters are built with borrowed money, and the interest bill keeps arriving after the metal price has gone. A producer's leverage must be judged against trough earnings, not the comfortable peak, because the same EBITDA that clears the debt in a good year evaporates in a bad one.

Hold those three — cost position, integration, balance sheet — and the metrics below stop being a list and become one story.

The metrics that matter — and where they hide

Here is the part that catches anyone used to reading a P&L: the numbers that decide a metals investment are almost never on the income statement. Cost of production per tonne, EBITDA per tonne, LME realisation, captive-power share, ore grade, stripping ratio — these live in the investor presentation and get quoted on the concall, expressed per tonne or per ounce. The income statement gives you sales and operating profit; it will not tell you whether you produced more or got a better price, and those two mean entirely different things. You have to go and get them.

Cost of production per tonne

The single most important number in the sector, and the one beginners skip straight past to "profit." It is what it costs to produce one tonne of finished metal, all-in, usually quoted ex-royalty in dollars so it can be benchmarked against a global cost curve. It matters because revenue is handed to you by the market — so your rank on the cost curve is your competitive position. Where to find it: the deck and the call, never the P&L. Hindustan Zinc tracked its zinc cost of production down from $1,052 in Q4 FY25 to a full-year FY26 figure of $959 a tonne, a five-year low and below its $1,000 guidance, with a Q4 exit of $903 (Hindustan Zinc Q4 FY25 and Q4 FY26 concalls). "Good" is not a fixed level — it is being in the bottom quartile of the global curve and holding there as the cycle turns, because that is who stays profitable at the trough.

LME / benchmark realisation

The price actually received per tonne of metal — for non-ferrous, anchored to the London Metal Exchange (LME) plus a regional premium; for a miner, a benchmark grade price. It matters because it is the revenue you do not control, and it swings violently. Where to find it: the deck, often as the LME price the company realised plus its hedge book. Hindustan Zinc reported a zinc LME of $2,764/tonne in Q1 FY26 climbing to $3,350 by Q3, and — the more interesting line — silver surging from ~$37/oz to an all-time-high ~$93/oz spot over the same window (Hindustan Zinc Q1 and Q3 FY26 concalls). Because realisation is exogenous, what you watch is the hedge book: a producer that has locked in 87 KT of zinc at $2,872 (Hindustan Zinc Q2 FY26 concall) has chosen certainty over the upside, and you want to know that before you assume the spot price flows to the bottom line.

EBITDA per tonne

Cost of production tells you the input side; EBITDA per tonne tells you what each tonne actually earns after costs — the cleanest single measure of unit economics, stripped of volume and accounting noise. It is to a metals producer what NIM is to a lender. Where to find it: the deck, quoted explicitly, and here the vertical-integration story jumps off the page. Hindalco's upstream India aluminium segment — fed by captive coal and power — earned an EBITDA per tonne of $1,572 with 45% margins in Q3 FY26 (Hindalco Q3 FY26 concall). Its downstream Novelis business, which buys aluminium and rolls it, earned roughly $495 a tonne the same quarter against a long-standing $600 target (Hindalco Q3 FY26 concall). Same group, two businesses, a 3x gap in unit economics — that gap is the value of owning the upstream cost moat. Read it over six to eight quarters; one quarter only tells you where the cycle is.

Vertical integration: captive ore, coal, and power

Not a single line but a set of them — captive power share, captive coal tonnage, captive ore share — and collectively the most durable cost advantage in the sector. It matters because power and raw material are the bulk of the cost, and owning them flattens your cost curve through the cycle. Where to find it: the strategy section of the deck and the call. Hindalco told analysts that when all three of its captive mines run fully they deliver "around 20 million tons of coal" and "about a 30% reduction in the cost level" (Hindalco Q1 FY26 concall). That is the number that explains why its India cost of production barely moves while a non-integrated smelter's lurches with the spot coal price. When a producer's captive share is falling over time, its cost moat is quietly eroding — flag it.

Ore grade and reserves (for the miners)

For a pure miner — iron ore, zinc concentrate, coal — the quality of the rock is the business. Ore grade is the metal content per tonne mined; reserves are how many years of it remain. It matters because a richer grade means less rock moved per tonne of metal, which means lower cost; a depleting reserve is a melting ice cube. Where to find it: the deck and the call. Hindustan Zinc's ore grade ran around 7.4–7.5% in FY25–26 (Hindustan Zinc concalls), and management was explicit that lifting silver grade toward 650 ppm is what would roughly double silver output to 1,300 tonnes under its 2x plan (Hindustan Zinc Q3 FY26 concall). For a miner, falling grade with rising production is the classic warning that the cheap ore is running out.

Stripping ratio and evacuation (for the coal/ore miners)

The stripping ratio is how many units of waste rock (overburden) must be removed to extract one unit of mineral; evacuation is the logistics — rakes, conveyors, ports — that get the mineral to the customer. Both matter because they are the real cost and capacity constraints of a mine, and both hide off the income statement. Where to find it: the operational section of the deck. Coal India ran a stripping ratio of 2.67 in FY25 and removed 1,404 million cubic metres of overburden in nine months — a 22% jump (Coal India Q4 FY25 and Q3 FY24 concalls) — which is the leading indicator of future production, because you must strip before you mine. On the cost side, its open-cast coal costs roughly ₹1,000/tonne against ₹2,100–2,200 underground (Coal India Q2 FY24 concall), and it is pouring ₹24,700 crore over five years into First Mile Connectivity to fix evacuation (Coal India Q3 FY24 concall). A miner that grows production while overburden removal stalls is borrowing from next year.

Net debt (and net debt to EBITDA)

The survival metric. Absolute net debt is the burden; net debt to EBITDA is how many years of cash flow would clear it — and the trap is that EBITDA is itself cyclical, so a ratio that looks safe at the peak can double at the trough without a rupee of new borrowing. Where to find it: the deck states it directly; the ratio is quoted on the call. The contrast is the lesson. Hindustan Zinc moved to net cash of ₹5,594 crore by Q4 FY26 (Hindustan Zinc Q4 FY26 concall). Hindalco's India operations also sat in net cash of ₹18,657 crore (Hindalco Q1 FY26 concall) — but its consolidated net debt, carrying the leveraged Novelis acquisition, was around ₹34,257 crore in Q1 FY26 (Hindalco Q1 FY26 concall), running at a consolidated net-debt-to-EBITDA of 1.73x at end-December 2025, with management committing to hold that ratio "at 2 or below" (Hindalco Q3 FY26 concall). Judge the ratio against trough EBITDA, and look at where in the group the debt actually sits.

Value-added / downstream mix

The share of output that is specialised, higher-margin product — precious metals for a zinc miner, rolled and recycled products for an aluminium group — rather than raw commodity. It matters because it dampens the cycle. Where to find it: the deck, as a percentage. Hindustan Zinc said precious metals (mainly silver) now contribute 44% of profits (Hindustan Zinc Q3 FY26 concall) — so a "zinc" company is increasingly a silver company, and you would misread it watching only the zinc price. A rising value-added share is one of the few structural improvements a price-taker can engineer for itself.

How do you value a deep cyclical like a metals producer?

This is where most investors lose money, and the error is mechanical: they value a metals stock on P/E.

A P/E divides price by current earnings. For a business whose earnings swing several-fold across a cycle, that denominator is meaningless. At the bottom, earnings collapse, the P/E balloons, and the stock looks "expensive" exactly when it is cheapest. At the top, a fat metal price inflates earnings, the P/E looks low, and the stock looks "cheap" exactly when it is most dangerous. A deep cyclical's P/E is highest when you should be most interested and lowest when you should be most wary. It is a contrarian indicator used backwards by almost everyone.

So use two lenses instead.

EV/EBITDA strips out capital structure and accounting noise and compares the whole enterprise against operating cash generation — the cleaner cyclical multiple. But it must be read against normalised, mid-cycle EBITDA, never the peak. The discipline is to ask what this producer earns per tonne in an average year across the cycle, multiply by sustainable volume, and value that. Hindustan Zinc's quarterly EBITDA climbing from ₹3,816 crore in Q1 FY26 to ₹7,666 crore in Q4 (Inve data) is not a run-rate to capitalise — it tracks the zinc-and-silver price up-leg, and a mid-cycle number sits well below it.

Price to book (P/B) anchors on the replacement value of the assets — the mines, the smelters, the power plants — which do not evaporate when the metal price does. For a capital-heavy cyclical, P/B paired with through-cycle ROCE is often the most honest gauge. The question is whether the normalised return on capital justifies the premium to book. A producer earning a mediocre through-cycle ROCE while priced at a rich multiple of book is pricing in an up-cycle that has to actually arrive — and stay.

The owner's frame: don't ask "is this cheap on this year's earnings?" Ask "what does a tonne of this producer's metal earn in a normal year, how many tonnes can it sell, and what am I paying for that mid-cycle stream?"

A worked case: Hindustan Zinc and the cost that beat its own guidance

The cleanest way to feel why cost, not price, is the metric is a producer that drove cost down while the metal price was unremarkable — and to read its guidance against what actually happened. Hindustan Zinc set a FY26 zinc cost-of-production guidance of $1,025–1,050 a tonne; it delivered $959, a five-year low (Hindustan Zinc Q4 FY26 concall). On Inve's record that commitment is marked achieved — and management did not stop there, guiding FY27 cost at "$975 to $1,000 per ton, reflecting prevailing global uncertainties" (Hindustan Zinc Q4 FY26 concall). When a price-taker keeps beating its own cost guide, that is the rarest thing in a commodity business: a genuine, repeatable edge. (Illustration, not a view on the stock; figures as reported by the company.)

But read the same record for what did not hold. The FY26 silver-production guidance of 700–710 tonnes was missed; the FY26 refined-metal guidance of 1.1 million tonnes (±10 KT) was missed; and the marquee 2x capacity expansion — taking smelter capacity from 1.2 to 1.45 million tonnes "by FY27–28" — together with the 70% renewable-energy target, sits delayed (Inve data, guidance record). The cost guidance was beaten while the volume and expansion guidance slipped, all in the same company, across a few quarters.

The point is not that management misled anyone — on the record this is a low-cost operator delivering its headline cost number. The point is the texture: the cost guidance was kept, the production targets were missed, and the big expansion was pushed out. You only see that pattern by tracking each commitment against the quarter it was made — which is the entire job of Promise Tracker, and the kind of thing nobody reconstructs by re-reading four transcripts and four investor decks by hand.

Hindalco's record shows the same discipline of reading commitment-by-commitment. Its captive-coal strategy is real and its India business sits in net cash — but a target of 600 Kt of India downstream capacity "by June of calendar year '25" was simply not mentioned again on later calls (Inve data, guidance record: ghosted), and the Chakla coal mine's production start guided for early 2026 slipped to delayed. A captive mine that keeps slipping is a cost moat that hasn't been built yet.

Red flags specific to a metals and mining company

• Profit rising while cost rises faster than price. If the metal price is up but cost of production is climbing quicker, the producer is busier and poorer. Always read cost per tonne next to realisation, never realisation alone.
• Falling captive share. A producer mining less of its own ore, coal, or power over time is losing the only moat a price-taker has. Track captive-power and captive-coal percentages, not just the cost headline.
• Grade decline with rising output. A miner lifting production while ore grade falls is moving more rock for the same metal — costs are about to rise, and the cheap reserve is running down.
• A new smelter or mine commissioned into a weak price. A new asset is a fixed cost. If it ramps as the cycle turns down, it converts operating leverage into a drag — high depreciation and interest with no spread to cover them.
• Net debt judged against peak EBITDA. A 2x net-debt-to-EBITDA ratio at the top can become 4x at the bottom when EBITDA halves, with no new borrowing. Stress it against trough earnings — and check where in the group the debt sits.
• One-time income dressing up the profit. A fat reported profit can be a non-operating windfall — strip to operating EBITDA per tonne before believing any earnings number.
• Expansion guidance that keeps slipping. A capacity target pushed out call after call (or quietly dropped) is the tell that the growth thesis is further away than the deck implies.

Frequently asked questions

A repeatable workflow

1. Rank the cost. Cost of production per tonne, benchmarked to the global cost curve and tracked as a trend — the low-cost producer survives the trough. Read it next to realisation, never alone.
2. Read EBITDA per tonne over time. Six to eight quarters from the deck; one quarter is the cycle, the trend is the company — and the upstream-vs-downstream gap is the integration story.
3. Check integration. Captive power, captive coal, captive ore share — rising is a deepening moat, falling is an eroding one.
4. For miners, check the rock. Ore grade, reserves, stripping ratio, evacuation — the constraints that set future cost and volume.
5. Stress the balance sheet. Net debt against trough EBITDA, and look at where in the group the leverage sits.
6. Value on the cycle, not the quarter. EV/EBITDA and P/B on mid-cycle numbers — and treat a cheap-looking P/E as a red flag.
7. Audit the guidance. Check the cost, volume, and expansion commitments against what actually happened next.

Inve's KPI Screener lines up cost of production, EBITDA per tonne, net debt and volume across metals and mining companies — value, trend and a data-confidence flag per number — so the per-tonne hunt takes minutes, not an afternoon of PDF-mining. For the ferrous spread business read in a different shape, see how to analyse a steel company; for a spread business with no tonnes at all, how to analyse an NBFC.

Where this lens can be wrong. The strongest case against everything above is that cost position, integration and a fortress balance sheet are necessary but not sufficient — the metal price is set by global supply and demand no analyst can forecast, and at a deep enough trough even the lowest-cost producer earns little. You can read cost of production, captive share, grade and net debt perfectly and still be blindsided by a China stimulus reversal, an LME crash, a power-tariff change, or a royalty hike that resets the economics overnight. Reading the operating numbers tells you whether a producer is built to survive and gain share through a cycle. It does not tell you when the cycle turns, and a great low-cost mine earns a poor return for as long as the metal price stays depressed. The honest claim is narrower than it looks: this analysis lowers your odds of owning a fragile, high-cost, over-levered producer into a downturn, and raises your odds of owning a durable one into the recovery. It cannot time the recovery.

The owner's question to sit with before buying any metals or mining stock: across a full cycle — not this quarter's metal price — where does this producer sit on the global cost curve, how much of its own ore, coal and power does it own, and is the balance sheet strong enough that it is still mining, still investing, and taking share from weaker rivals when the price finally turns? If the answer leans on this year's high metal price holding forever, you have read the peak, not the business.

Related reading

• How to Analyse a Bank Stock (NIM, CASA, GNPA, ROA)
• How to Analyse a Building Products Company (Pipes, Tiles)
• How to Analyse a Cables and Wires Stock