Why I Keep Coming Back to Tokyo and Seoul for Our Tungsten and Moly

A buyer in Osaka once asked me a question that stuck. "What's your tungsten trioxide percentage, and can you hold it batch to batch?" Not price. Not volume. Consistency. That single question told me more about the Japanese market than any trade report I'd read.

Because here's the thing about Japan and South Korea — they don't chase the cheapest concentrate. They chase the one they can plan around. And that suits us more than the buyers who only want to talk about tonnage.

Let me back up.

We hold 16 concessions across Gilgit-Baltistan, and among them are scheelite-bearing skarn zones and molybdenite occurrences tied to the Karakoram batholith intrusives. Tungsten and molybdenum both sit on the US, EU and UK critical-minerals lists. Both are heavily concentrated in Chinese supply. And both Japan and Korea have spent the last decade trying to reduce that single-country dependence — quietly, without press releases.

The concentrate quality question is the whole game

Tungsten off-take isn't like selling gold. Nobody pays you for raw rock. They pay for WO3 content in a clean concentrate, and the penalties for impurities can eat your margin alive.

Japanese processors — think the ammonium paratungstate (APT) converters feeding the carbide tool industry — want scheelite concentrate at 65% WO3 or better, with phosphorus, arsenic and tin held down hard. Get the arsenic wrong and you're looking at deductions that make the shipment barely worth loading.

Our scheelite from the skarn contacts assays well on the tungsten side. What I got wrong at first was assuming grade alone would sell it. It doesn't. The deleterious element profile matters just as much, and that means beneficiation on our end — flotation and magnetic separation to strip the molybdenum and sulphides before we ship. We're building that testwork now with an independent lab rather than pretending we can skip it.

Korea's a slightly different animal. The Korean tungsten story runs through Sangdong, the mine Almonty has been bringing back into production. So Korean buyers already understand tungsten geology and pricing at a level most markets don't. When I talk to a Korea molybdenum or tungsten buyer, I don't have to explain what APT is or why the London metal exchange doesn't set the price. That saves everyone weeks.

Honestly, dealing with a buyer who already knows the commodity is worth a small discount on price. The relationship costs less to run.

Molybdenum, steel, and why Korea keeps calling

South Korea's steel sector is enormous. POSCO alone is one of the biggest steelmakers on the planet, and molybdenum goes into the high-strength, corrosion-resistant alloys they produce — pipeline steel, pressure vessels, the stuff that has to survive North Sea conditions or a chemical plant.

Molybdenum demand tracks steel and it tracks energy infrastructure. Both of those are growing in Korea and across the wider region. Japan's the same story on the aerospace and specialty-alloy side.

Our molybdenite shows up as disseminated mineralisation and in quartz veining within the intrusive suites. Roasting molybdenite to technical-grade molybdenum oxide is a step we'd rather see done closer to the buyer, at least early on — the roasting infrastructure and the sulphur-dioxide handling aren't something you stand up overnight in Gilgit-Baltistan. So the realistic first move is shipping clean molybdenite concentrate and letting the Korean or Japanese converter handle the oxide stage.

That's not a weakness. It's just honest about where the value chain sits today.

The logistics reality nobody sugarcoats

Here's where I have to be straight with you, because the geology is only half the deal.

Gilgit-Baltistan is landlocked-adjacent — we're a long way from a port. Concentrate moves by road down the Karakoram Highway to the Punjab, then on to Karachi or Port Qasim for containerised export. That's roughly 1,800 kilometres of trucking before anything touches saltwater. Winter closes sections of the KKH. Landslides happen. I won't pretend otherwise.

But — and this is the part that changes the maths — tungsten and molybdenum concentrates are high-value, low-volume. A single container can hold hundreds of thousands of dollars of WO3. The freight cost as a percentage of cargo value is small compared to, say, shipping bauxite or granite. That's exactly why these two metals are the right ones to build an export relationship on first. The economics survive the trucking.

For a Japan tungsten import buyer, the sea leg from Karachi to Yokohama or Kobe is well-trodden. Same for Busan. No exotic routing. The friction is all on the Pakistan side of the port, and that's manageable with the right freight forwarder who's actually moved mineral concentrate before, not just textiles.

What a real JV or off-take looks like from here

I'm not interested in a one-container trial that goes nowhere. Those waste everyone's time.

What makes sense is a staged off-take — a bulk sample first, properly assayed by a lab both sides trust (SGS or Intertek, umpire samples split three ways), then a defined tonnage commitment tied to agreed WO3 and MoS2 specifications with a transparent penalty schedule. The kind of contract structure a Korean or Japanese trading house already uses every day.

And on the bigger concessions, I'd rather bring in a JV partner who puts capital into the beneficiation plant on the ground. Because whoever controls the concentrate quality controls the price. I'd rather share that upside with a serious Asian off-take partner than ship dirty concentrate and eat penalties forever.

One thing I've learned talking to buyers in both countries — they move slowly, then they commit for years. That patience used to frustrate me. Now I think it's the healthiest way to build this. A Chinese buyer might sign faster, but a Japanese converter who's satisfied with your batch consistency will still be buying in a decade.

So if you're on the procurement side for a carbide manufacturer in Japan, or a specialty-steel buyer in Korea, and you're mapping supply outside the usual single source — what specification would you need to see before you'd take a bulk sample?