Is Copper in a Supercycle? | Subtext by Zerodha | Markets by Zerodha

ELI5 / TLDR

Copper prices roughly doubled — from around $6,000 a tonne to a peak near $13,000–14,000 — and the question is whether this is a genuine long-running boom (a “supercycle”) or just another spike that crashes 50% like the last four did. The guest, Amit Jeswani of Fintech Capital, argues this cycle is real and probably lasts another three to five years, because demand is being pulled by AI data centres, power grids, and electric vehicles, while new supply takes 5–15 years to dig out of the ground. The most interesting wrinkle: when copper booms, the miner pockets almost all the upside, the smelter in the middle gets squeezed (sometimes to zero profit), and everyone downstream just passes the cost along. India, embarrassingly, used to export copper and now imports it — a self-inflicted wound from a single plant shutting in 2018.

The Full Story

Four booms, four busts, and the question of “this time”

Copper has a reputation as the metal with a PhD in economics — it shows up in nearly everything industrial, so its price reads the global economy’s pulse. The interviewer opens skeptical: copper has rallied four times this century and every single rally was followed by a roughly 50% drawdown. Why believe this one ends differently?

Jeswani walks through the four. The 2000s rally was pure China — urbanization, grid buildout, a property boom — copper doubled from ~$3,000 to ~$6,500 before the 2008 crisis knocked it down. Then came the 2009–2013 “QE supercycle,” central banks pumping money, followed by a wave of new mines that flooded the market and forced a correction. Third was the post-COVID spike, killed off by China’s property slowdown (the Evergrande collapse — he calls it “Granada”) and tightening liquidity. Each time the pattern held: demand shock up, supply catches up, price down.

The new cycle, he says, is different in its driver. It is not one country urbanizing — it is structural electrification everywhere at once.

“We have this cycle which is basically led by AI which is led by power transmission and distribution which is happening because of electrification of vehicles.”

His honest caveat: he does not know when it stops, and another 50% drawdown is always possible. The point is only that, for now, the tailwinds are blowing the right way.

Why supply can’t just show up

Here is the mechanism that makes a supercycle plausible. Imagine mining as panning for gold, except the gold gets scarcer the deeper you dig. Ore “grade” is the percentage of actual copper in the rock. A 1.5% grade means a tonne of ore yields 15 kg of copper; the best surface deposits have largely been scraped, and the global average grade has slid toward 0.6%. Lower grade means you crush three times the rock, burn three times the energy, and manage three times the waste for the same copper. Think of it like squeezing oranges that keep getting drier — same juice, far more squeezing.

And a brand-new mine is not a factory you switch on. It takes 5 to 15 years from discovery to first metal, after you survive permitting, financing, and — increasingly — geopolitical risk, with countries now weaponizing raw-material exports. So even when the price screams “dig faster,” the supply response is glacial. Copper, in his phrase, is “oscillating between surplus and scarcity,” and if AI demand stays on a straight upward line, genuine shortage is on the table.

The escape valve: substitution

High prices, Jeswani notes, are information — they tell every engineer in the world “find something cheaper.” History obliges: the 1970s oil shocks made economies vastly more energy-efficient; fibre optics displaced copper telecom cable; carmakers re-engineered catalytic converters to use less platinum when platinum spiked. For copper, the obvious substitute is aluminium — cheaper, lighter, already used in some transmission lines and power applications. But aluminium is less conductive, so it can’t go everywhere copper goes, especially in dense EV and data-centre wiring. For now there is no full replacement, which is precisely why copper’s prized combination of conductivity, durability, and recyclability keeps it irreplaceable.

Who actually makes the money

This was the sharpest part. When copper doubles, the value chain does not share the spoils evenly.

The miner wins biggest. Their costs — labour, energy, sustaining capex — are largely fixed and rise slowly, so when the copper price doubles, the extra revenue drops almost straight to operating profit. A miner producing at $4,500–$6,000 sees margins explode when the metal hits $13,000. Classic operating leverage.

The smelter — the middleman who turns raw concentrate into refined copper — gets squeezed, which is counterintuitive. Smelters earn “TC/RCs” (treatment and refining charges), essentially a toll for processing, not a cut of the copper price. When ore is scarce, miners hold the bargaining power and beat those tolls down. At the December peak, some Chinese smelters reportedly agreed to process at zero dollars a tonne just to keep their feedstock and their factories running. Think of a toll booth that has to start waving cars through for free because there are barely any cars and it cannot afford to sit idle.

The downstream — wires, cables, electrical equipment — mostly just passes the metal cost through, like Asian Paints passing crude-oil swings to its dealers. They don’t profit from the price rise itself; they profit from volume (demand is strong) and from working-capital muscle. A spiking copper price means you need far more cash tied up in inventory, which quietly crushes small unorganized players who can’t fund it — explaining a steady shift toward listed, branded, well-hedged firms.

The mining majors won’t merge

Jeswani notes the repeatedly-floated mergers — Glencore-Rio Tinto, Rio-BHP — that keep collapsing on valuation gaps and the sheer legal-political tangle of assets spread across many jurisdictions. His read: the copper story is real, but you can’t fully trust the big miners to convert it into clean shareholder returns, given environmental litigation (Glencore especially) and governance friction.

India’s own-goal

The most pointed thread is India. The country exported copper until 2018, when Vedanta’s Sterlite plant in Tuticorin — supplying ~90% of domestic demand, employing ~20,000 — was shut after worker agitation, a police firing that killed protesters, and a Supreme Court order. A Green Tribunal review years earlier had found no environmental case, yet the plant remains closed and analysts now assign it zero value on Vedanta’s balance sheet. India became a net importer overnight.

The response is the “Atmanirbhar” push: Adani (a mine in Kutch), Hindalco, JSW and the Jindals are all entering copper. But none of this is near-term — Adani’s plant arrives around 2029, full capacity by 2031. These are 20-to-30-year strategic bets on India’s electrified future, not bets on today’s prices. The clinching number: India’s copper consumption is about 0.5 kg per capita against China’s ~9 kg. That gap is the runway.

How to read the tape

For tracking the market, Jeswani names three indicators: LME inventories (London Metal Exchange stock levels, published weekly on Thursdays — falling stocks signal tightness), the Yangshan premium (what Chinese buyers pay above benchmark for imported copper — rising = strong physical demand and stockpiling; China consumes ~65% of the world’s copper), and mine-disruption headlines from Chile and Peru, where strikes and unseasonal rains routinely knock out supply and spike prices.

Key Takeaways

Copper has had four boom-bust cycles this century, each followed by a ~50% (post-COVID: ~30%) drawdown; the current cycle is driven by AI data centres, grids, and EVs rather than a single country’s urbanization.
Ore grade — the copper percentage in mined rock — has fallen from ~1.5% toward 0.6% globally; lower grade means far more rock crushed, energy burned, and waste managed per unit of copper.
A new copper mine takes 5–15 years from discovery to production, making supply structurally slow to respond to price.
Copper peaked near $13,000–14,000/tonne, roughly double the ~$6,000 base.
Aluminium is the main substitute — cheaper, lighter, but less conductive, so it caps but cannot replace copper demand.
In a price boom, miners capture most of the gain (fixed costs + operating leverage); smelters get squeezed because their TC/RC processing fees collapse when ore is scarce.
At December’s peak, some Chinese smelters reportedly accepted $0/tonne treatment charges just to keep feedstock and factories running.
Downstream cable/wire/equipment makers pass copper cost through (like paints passing on crude); they win on volume and working-capital strength, not on the price rise itself. Rising prices push out unorganized players who can’t fund inventory.
India exported copper until 2018; Vedanta’s Sterlite Tuticorin plant (~90% of domestic supply, ~20,000 jobs) shut after worker protests, a fatal police firing, and a Supreme Court order — making India a net importer.
Analysts assign zero value to the shuttered Sterlite plant on Vedanta’s books.
Adani’s Kutch copper plant is expected around 2029, full capacity ~2031; Hindalco, JSW, and the Jindals are also entering. These are 20–30 year strategic bets, not near-term plays.
India consumes ~0.5 kg copper per capita vs China’s ~9 kg — the demand gap is the long-term runway.
EU’s CBAM (carbon border tax) phase one began Jan 2026 without copper, but the EU has proposed extending it to ~180 downstream products from 2028, including copper wires and cables — a coming cost for Indian exporters.
Three tracking metrics: LME inventories (weekly Thursday data), the Yangshan premium (Chinese import-demand gauge; China consumes ~65% of global copper), and mine-disruption headlines from Chile/Peru.

Claude’s Take

This is a solid, honest interview rather than a hype reel — which is to its credit. Jeswani repeatedly refuses to promise the supercycle thesis, hedges that another 50% crash is possible, and openly says he wouldn’t play the “pure copper beta” (Hindustan Copper) because its valuation is “out of whack.” That self-skepticism is the tell of someone actually thinking rather than selling.

The genuinely useful, durable content is the value-chain breakdown — miner wins, smelter gets squeezed, downstream passes through — and the TC/RC mechanism, which is the kind of plumbing detail most copper commentary glosses over. The zero-dollar Chinese smelting anecdote is a vivid, checkable signal of how lopsided the concentrate market has become.

Weaknesses: it leans heavily on one analyst’s framing with no opposing voice, and the supercycle case rests almost entirely on “if AI demand stays linear” — a load-bearing assumption that is doing a lot of work and could just as easily wobble. The recurring India-grievance framing (reopen Sterlite!) is reasonable but one-sided; the plant closed after people were killed in a police firing, and “if there are environmental concerns they should be resolved” rather waves past a serious story. The transcript is also a bit meandering, with the interviewer doing a lot of “I read somewhere” thinking-aloud.

A 7. Clear, mostly rigorous, genuinely educational on market mechanics, let down slightly by single-source framing and an AI-demand assumption it can’t really defend.