There is a small company in Visakhapatnam that makes only one thing, a white powder called aluminium fluoride. I did not expect a flux chemical (white powder) to be the most interesting thing I read this year.

Here is why. Somewhere between 14 and 20 Kg of it is consumed for every tonne of aluminium produced. No aluminium fluoride, no aluminium. And this year aluminium is the one metal everyone suddenly cannot get enough of. Even the beer and cola makers have been running around for cans.

The company makes this powder out of someone else’s waste. Let’s understand basics first.

Why the world ran short of aluminium

Start with China, because with aluminium you always start with China. It makes roughly 60% of the world’s primary aluminium. Towards the end of 2024 it cancelled the 13% export tax rebate on the metal and capped its own production at around 45 million tonnes a year. The logic was simple. Why ship cheap raw metal abroad when you can keep it at home and turn it into EVs, solar frames and other things that earn more?

That alone would have left the world a little short. Then West Asia caught fire. Gulf smelters in places like Qatar and Bahrain had to shut, because smelting runs on enormous amounts of electricity and the war disrupted their energy. On top of that, the bauxite and alumina they import, and the metal they export, both move through the Strait of Hormuz. Around 1.6 million tonnes, close to 9% of global smelting capacity, got hit.

Take that much supply out of a market that was already tight and the price does what you expect. Aluminium peaked near $3,800 a tonne in the middle of this year. It has cooled since, and on 16 October it was trading around $3,000, the lowest since February, but still about 8% higher than a year ago.

India felt it in a strange way. We are the world’s second-largest producer of primary aluminium, so in theory we should be fine. But making a can needs a very thin, food-grade rolled sheet, and we do not make enough of that. So we import it, much of it from the Gulf. Hence the missing cans.

India makes a lot of it, and uses very little

India has roughly 4.2 million tonnes of primary smelting capacity, almost all of it sitting with three groups. Hindalco, Vedanta (including BALCO) and NALCO. Out of about 74 million tonnes produced globally last year, that is not a small number.

But look at consumption. An average Indian uses around 3 Kg of aluminium a year. The global average is around 11 Kg. China is near 25 Kg. Nearly half of what we do use goes into one thing, electricity. Transmission conductors, cables, transformers. Add 500 GW of non-fossil power targets, a grid that has to carry it, EVs that need to be light, and the gap between 3 Kg and 11 Kg starts to look less like a statistic and more like a runway.

The producers are behaving like they believe it. Hindalco is spending about 21,000 Crore to add 3.6 lakh tonnes of smelting at its Aditya complex in Sambalpur, taking it towards 0.72 million tonnes. Vedanta keeps talking about a 3 million tonne footprint. NALCO has lined up roughly 30,000 Crore over five years for a new smelter, with groundwork targeted around mid-2027. And the Government has started saying 37 million tonnes of capacity by 2047, which is aspirational, but tells you the direction.

Here is the bit that matters for the rest of this piece. Every tonne of new smelting capacity does not just need bauxite and power. It needs a steady drip of consumables, every single day, for as long as it runs. One of them is aluminium fluoride. Every 1 lakh tonnes of new smelting needs somewhere between 1,400 and 2,000 tonnes of it a year, forever.

How aluminium actually gets made

You dig up bauxite. You refine it into alumina, which is just aluminium oxide, a white powder. Then you send the alumina to a smelter, and this is where it gets interesting.

Alumina does not want to give up its oxygen. To pull the metal out, smelters dissolve it in a molten salt bath and pass a massive electric current through it, in long rows of cells called pots. Hence the word potline. Molten aluminium collects at the bottom and gets tapped out. This process has not fundamentally changed in over a century.

The problem is that alumina on its own melts above 2,000C. No one can run an economical plant at that temperature. So the bath is mostly cryolite, with aluminium fluoride added on top. That brings the operating temperature down to under 1,000C, and keeps the bath chemistry and conductivity where it needs to be. Less heat, less power, more metal.

Aluminium fluoride is not part of the final metal. It is a consumable. It gets used up, lost to fumes and absorbed into the pot lining, so smelters keep topping it up, continuously.

And this is where the moat comes from. A potline is never supposed to stop. Shut it down and restarting can take weeks and a lot of money. So if one bad consignment of aluminium fluoride messes up the bath, that is a disaster. Which is why smelters take ages to approve a new supplier, and once they do, they rarely switch. Slow to win, slow to lose.

Two ways to make the same powder

There are broadly two routes to aluminium fluoride, and which one you are on decides almost everything about the business.

The first is the fluorspar route, and it is how most of the world does it. You mine fluorspar, a calcium fluoride mineral, react acid-grade fluorspar with sulphuric acid to get hydrofluoric acid (HF), and then react that HF with alumina hydrate. You need roughly 1.5 tonnes of fluorspar for every tonne of aluminium fluoride. Clean, scalable, well understood.

The catch is the rock. China mined about 6 million of the world’s roughly 9.5 million tonnes of fluorspar in 2024. India does not have enough of its own, so fluorspar-route makers here import it, along with the currency risk and the China risk that come with it. There is also a quieter problem. HF is the starting point for almost every fluorine chemical, refrigerants, pharma intermediates, battery chemicals. If you own an HF plant, you have many more profitable things to do with it than sell bulk flux to a smelter. In India, Tanfac Industries (jointly controlled by Anupam Rasayan and TIDCO since 2022) is on this route, and its recent growth story is mostly about HF and refrigerant gas.

The second is the FSA route. When a fertiliser plant makes phosphoric acid, it treats phosphate rock with sulphuric acid. The fluorine sitting in that rock escapes as gas, gets scrubbed with water, and comes out as fluosilicic acid, FSA. FSA is toxic and annoying to dispose of. For the fertiliser plant it is a waste problem. For an aluminium fluoride maker, it is the fluorine you would otherwise have had to mine. React FSA with alumina hydrate and you get aluminium fluoride, plus silica as a byproduct.

Read that again. No mined fluorspar, no China, and the key input is something the supplier needs to get rid of. That is a real cost advantage.

But nothing is free. You can only make as much as the fertiliser plants near you throw off. You cannot just build a bigger plant, you need someone to build a bigger phosphoric acid plant next door first. Globally, FSA has historically been only a small slice of the fluorine that goes into aluminium fluoride, for exactly this reason. Cheap but hard to scale, versus scalable but import-dependent. That is the whole trade-off.

The Company

Alufluoride Ltd has been doing exactly this since 1984, from a single plant at Mindi in Visakhapatnam. It buys FSA from IFFCO’s Paradeep complex under a long-term contract (originally 7,000 tonnes a year, now over 17,500, with room to go to 19,000), tops up from Paradeep Phosphates, and since April 2026 also draws from Coromandel’s new phosphoric acid plant at Kakinada. It turns that into aluminium fluoride and sells it to Hindalco, Vedanta and NALCO. It is the only aluminium fluoride maker in Andhra Pradesh, and one of fewer than 5 in the country.

One product. Three customers who matter. No exports. A captive solar plant on the side to keep power costs down. That is the whole company, and I mean that as a compliment.

The capacity has been built in steps. 4,000 tonnes a year at commissioning, then 8,000, 12,000, 18,000, and finally 24,000 tonnes a year, completed in June 2026. In FY26 it produced 17,787 tonnes against 18,000 of capacity. It was basically running flat out, which is why it needed the expansion.

The FY26 numbers (year ended March 2026) were the best in its history. Revenue of 208.8 Crore, up 10.7%. EBITDA of 46.4 Crore, margin up to 22.1% from 20.2%. Net profit of 24.3 Crore, up 32.7%. EPS of 31 rupees, dividend raised to 4 rupees. Return on net worth around 20%, with low debt. On 18,866 tonnes sold, that is a realisation of roughly 1.1 lakh rupees a tonne.

Now put that against demand. India’s 4.2 million tonnes of smelting, at 14 to 20 Kg per tonne, needs something like 60,000 to 85,000 tonnes of aluminium fluoride a year. Alufluoride at full 24,000 tonnes would be roughly a third of that. The rest comes from Tanfac and from imports, mostly Indonesia, China and the Gulf. So it is not fighting for share in a crowded market. It is substituting imports in a market that is about to get bigger.

The capex, and where the real cost sits

This is the part that took me longer to understand.

You would assume FSA is the big raw material cost. It is not. In FY26 the company consumed 124 Crore of raw materials. FSA was about 30 Crore of that. Alumina hydrate, the other half of the reaction, was about 68 Crore, more than half the bill. FSA is cheap. Its problem is availability, not price. Alumina hydrate is the one that actually moves the margin.

Which is why the backward integration plan makes sense. Since FY25 the company has been working on a greenfield alumina hydrate plant, both to secure its own supply and as a separate business. Shareholders approved the change in the objects clause in April 2025. To be fair about where it stands, though, the March 2026 balance sheet shows only about 11 lakh rupees spent on it so far. It is still on paper. The money in capital work-in-progress went into the expansion itself, a new calciner (about 12.3 Crore) and a new centrifuge (about 9.1 Crore).

Around that, a few smaller pieces. Captive solar went from 3 MW to 4.1 MW in May 2026. A second FSA source at Kakinada came online. IFFCO has indicated more FSA once it revamps its fluorine recovery. And management is looking at turning the silica byproduct into something sellable, plus fluorine derivatives and a possible second aluminium fluoride site, all at the assessment stage.

The simple arithmetic is this. 24,000 tonnes at FY26’s roughly 1.1 lakh a tonne would be around 265 Crore of revenue, against 209 Crore last year. That is not guidance, the company has not given any. It is just what full capacity looks like if the FSA keeps flowing. And that “if” is the whole next section.

The catch

The easy line to write is that Alufluoride does not depend on imported raw material. That is only half true. It does not depend on mined fluorspar. It depends on someone else’s waste. And the fertiliser plants making that waste run on imported rock phosphate, sulphur and ammonia, with sulphur and ammonia coming largely through the Strait of Hormuz.

So when West Asia flared, the same war that took Gulf aluminium off the market also throttled Alufluoride’s FSA. The company told the exchanges on 23 May 2026 that FSA supply had dropped and operations were being scaled back. Q1 FY27 (April to June 2026) revenue fell to about 26.6 Crore, down 39% from a year earlier, and profit to about 2.75 Crore. Even Q4 FY26 was soft, because two FSA suppliers shut early for maintenance. The supplier said the problem was resolved on 15 July 2026, and the plant is back to normal. But the lesson stands. This is a business one step removed from the import problem, not free of it.

A few other things I would keep an eye on.

Customer concentration. The largest customer took 69% of aluminium fluoride revenue in FY25, up from 52% the year before. That is what happens when smelting is in three hands, but it is still one relationship carrying most of the company.

The import ceiling. Domestic makers cannot price far above landed imports, or smelters simply buy from Indonesia. The aluminium industry has lobbied for years to cut the duty on aluminium fluoride from 7.5% to 2.5%. If that ever goes through, it is good news for Hindalco and Vedanta, and a lower ceiling for Alufluoride.

The product itself. Alufluoride is India’s largest maker of low bulk density aluminium fluoride, which is what the FSA route naturally gives you. Much of the world’s capacity makes the high density grade through the fluorspar route, and the two are not perfect substitutes. It is qualified at all three Indian majors today, so this is not a problem now. It is a question for the new potlines being built.

One more data point, make of it what you will. Between late May and the end of June 2026, exactly while the plant was running below capacity, two people classified as immediate relatives of insiders bought roughly 90 lakh rupees worth of shares from the market, at prices between about 440 and 500 rupees. The stock has since run to around 670 rupees (9 October 2026), a market cap of roughly 530 Crore, or a little over 20 times FY26 earnings.

End note

The can shortage and the aluminium fluoride story are the same story, read from two ends. Aluminium got scarce because the world’s cheapest supply decided to keep its metal and the Gulf’s supply got caught in a war. India’s answer is to build more smelters. Every one of those smelters will need a white powder, 14 to 20 Kg at a time, every day, from a supplier it trusts and will not easily replace.

Alufluoride sits right in that pipe, on a route that skips the fluorspar problem entirely, and it has just finished building the capacity to sell a lot more of it. But its own raw material is tied to the same West Asia that caused the shortage in the first place. Demand is not the question here. Supply of waste is.

Can a business chained to a fertiliser plant’s effluent grow as fast as the smelters it feeds? I have not decided yet.

Disclaimer: Not a recommendation or investment advice. I work in equity research and this is my personal reading of public filings. Everything above is drawn from the company’s own disclosures and public records, and the interpretation is mine. Verify it yourself before putting a single rupee to work. The only person responsible for your money is you.