Here Comes the Scandium Supercycle (Koala)

Does the West actually value having a resilient robust supply chain with a hundred percent integrity that isn't at the mercy of China, which I think it's fair to say is the consensus number one geopolitical rival of the United States? Kuala, I can remember when we first met in person. We had a pre-rebus conversation about super cycles as they apply in commodity markets. We talked about the I&R super cycle, the potential lithium super cycle and its parallels. I never would have thought we'd be here three years later talking about a potential scandium super cycle. A very niche commodity is they can get like 75 times per annum of production, but here we are. The world needs it. And I just want to say this coming. Look, the biggest consumer globally of scandium today has put out a 2030 production target that says we're going to need incremental scandium equivalent to 150 percent of the current global annual supply. I know I've been doing metals and mining for 15 years or so and that sounds to me from the producer or developer side like the dream setup. 15 years or so, but you visited Ningan in 2014, which was a hell, I mean, I'm a futurist, but I'm sure you didn't see a scandium's relevance in 2014. I saw the potential for it. I think Bloom was actually still, it was actually a thing back then. And I think also the question was where did Bloom's technology fit in the stack, but the other big appeal outside of Solid Oxide Puel cells, which obviously given what's happened with the AI wave is the topical point today, but the potential for this in scandium aluminum alloys. I was always been something pondered. I mean, the Soviet Union put scandium aluminum alloys in the big fighter jets because they produce scandium from an underground uranium mining Ukraine. So it's been known that there are applications for this very niche metal, but it's always been a case of, well, where's the supply? I'm not going to adopt it in a car or a plane unless I know the supply is there. So it's always had this, what I call a failure to launch, a failure to get the adoption. And yet here we are today, and we've had this magical situation where we're building out gigawatts of compute at, call it $50 billion a gigawatt. That needs power. You can't get a connection to the grid or an interconnect or a gas turbine. So all of a sudden, the balloon box, which is a Solid Oxide Puel cell, it basically takes natural gas directly and converts electricity on site to very simplify it. The balloon box has gone from a cute niche power solution that had to compete on marginal economics with things like the electric grid and other solutions. To now, it's considered a go-to solution for the likes of Oracle and Brookfield who are trying to build out compute and data centers as fast as they can. So you've had this ironic demand pull from the downstream to the midstream. And now the question is, Bloom is out there saying, we're going to go from a gigawatt to five gigawatts. Well, if you're going to 5X your production, you need to 5X your supply chain. And yet, and so here we are. I mean, how long have you actually followed Bloom for CoreWild? It's fascinating because one of the points that you tease out in your writing is that this is a $70 billion company acting like a $3 billion company. And for the longest time, they were a single gigabillion dollar company. But there's a lot of people now pointing the finger at like management saying, hey, you might be lying about having no integration into the Chinese supply chain. And all these other things, you don't put out any sort of guidance this that in the other. You can't have invested in Scandium in 2013, 2014 without having known what Bloom is and what Bloom's potential could be one day. I don't think anyone in their wildest dreams thought something like this would happen. I mean, we're we've gone from in at least in the United States, black electric electricity demand for like 20 plus years to now, we're talking about actually having electric demand growth. And you can tell I'm not utilities trained, so I'm not calling it load or any of the fancy terms that I sometimes spout off when we're talking about basic things like mining. So what me through the the emergence of Bloom and like that, like what exactly the driver behind why Scandium is the answer to all of the world's like infrastructure growth problems here and why Bloom's kind of, you know, this one company's demand to be concentrated in China is so fragile. I think what Bloom's blog post, which they put out, basically the same time Hunter Brook put out their report last week. I mean, this is kind of the first time Bloom has publicly spoken about their Scandium requirements. There was a tweet, I wish I'd saved it down, but someone joked this is the first time Scandium has been in a bloom SEC filing. But basically, and look, when you have strategic things like this, you don't want to flag your vulnerabilities. I totally understand that point, but basically you need a ceramic in these SOSCs. And it's basically pure zirconium oxide or whatever. And this Scandium, a very small amount of it is basically doped into it to optimize the performance. If you talk to crossroads and you who did a great job looking at all of Bloom's patents, they had estimates of it was like six or seven kilograms per 100 kilowatt bloom box. My estimate would be that they've thrifted it down to probably closer to 30 tons per gigawatt, which would be three kilograms per 100 kilowatts. That's a pretty substantial change. Let's come down decently. Absolutely, but I think we know this is commodity investors, innovation and thrifting and optimization is part of the game. I mean, bear in mind, when we had the rare earth mania in 2010, 2011, and Syrium went to $150 a pound. And Molly Corp actually asked for Japan to pay that number. All of a sudden, the glass manufacturers found a way to not use Syrium. I mean, cliche, but cure for high prices, high prices, careful low prices, low prices. People will substitute at a certain point. And to lead to that point, you don't necessarily have to use Scandium in these SOFCs. You could use like E-Trium, but you might not get the same performance. It's a trade-off. And also bear in mind, E-Trium has been very politicized in geopolitics by China and the US in the past year. Scandium, it's been in the headlines, but you haven't really seen, I would say, Scandium become a geopolitical football the way E-Trium and some other rare earths have. I mean, it hasn't needed to, right? It could very well because we're just on a one-year pause for the big US China debate over rare earths and all the rest of it. So three months out, what do you reckon? Do you reckon it becomes a bigger talking point as people now start talking to that Scandium in the media? I think it's a broader critical minerals question. You have China saying things like, "Oh, we don't want tons to export tons to and that goes into defense applications." I think, but this is a whole critical mineral. This is a broader, this is a Scandium, a tons to an aware. This is a broader critical minerals thematic question. Do, does the West actually value having a resilient, robust supply chain with a hundred percent integrity that isn't at the mercy of China, which I think it's fair to say is the consensus number one geopolitical rival of the United States and the Western world at this point in time? Do you want to have that dependency? And look, they're dependent on us for things, but I think that's a big broader question right now. Because look, what happened to rare earths after 2012, China resumed exports and flooded the market. I think that's a question everyone who's looking at critical minerals in the West right now is saying, whether it's to Washington DC or others, which is, we aren't, are we going to get washed out if China exports again or we are a priority? That's a, that's a core question on top of all the things we talk about when we talk about different tungsten projects, different rare earth projects. It's, is that global supply chain coming through China a thing of a past? I mean, I think, I think today, point, it's worth for people that haven't read your paper or haven't followed the story too closely, just giving the quick recap of what Western projects look like at this stage. Sure. I mean, right now the majority of the supply chain in Scandium is out of titanium waste streams in China. There is a seven and a half ton per year Scandium circuit on the Technico nickel mine owned by Coral Bay, which is a Sunitomo metal mining subsidiary in the Philippines. And there is a three ton titanium waste Scandium recovery circuit at Surreal Tracy for Rio Tinto and Quebec, which off of a Canada growth fund investment of $25 million Canadian will go from three tons to nine tons. Beyond those supply sources, the future really comes down to you have the pretty well publicized Niochor on IOBM titanium Scandium underground project in Nebraska. And then though the real potential future source is there's a belt in New South Wales that runs from in the north the Ningen deposit, which is two hours drive west of Dubbo to the Sireston now known as Sunrise for Oendale formerly known as Burra Flemington is down there as well. But what I call would call the New South Wales five field Scandium belt, which is selection of laterites with a limeliant top satellite on the bottom that have concentrations of Scandium ranging from 250 300 ppm up to and some of the highest grades portions of Sunrise 7 or 800 ppm. And it's really a phenomenal geologic anomaly that could be the source of another 150 200 tons of Scandium on a primary basis. And the other options for the West are can we extract it from titanium waste streams, which Bloom kind of wants the market to look to. I would prefer to Bloom or every people in the market as I did in my piece. Why does the West not do this already? And the conclusion I can only come to is that at the Chinese price for Scandium, it doesn't make sense for the Western titanium producers to put those circuits on. And then of course it's a question of do the nickel cobalt laterites in say Indonesia want to put on Scandium recovery circuits because there is some Scandium in those. It's very small, be a very small percentage of revenue and earnings. But then again, most of the Indonesian nickel mines are controlled by Chinese shareholders. So that doesn't really I think necessarily solve the geopolitical question. Bloom will try and reassure the market that these titanium waste feeds will be able to secure Scandium to a price of about a thousand bucks per kilogram and to the future. You think that is incredibly insufficient. Oh, incredibly like a low relative for what the likely incentive price is to actually do it up these ways of the splotchains, right? They claim they claim line of sight to be able to supply enough Scandium to their business to produce 25 gigawatts. I note in there they did not say at what price. And frankly, I think that's where those who know commodities or sit between what I'll call the genitals and the commodities know is what the price is everything. I mean, there are shale oil basins like the Bakken in North Dakota that exists one because of shale revolution but have shale technology development. But really because oil went over a hundred bucks in the US for a few years. So all the sudden North Dakota merited drilling. As the title says, there is plenty of Scandium there for them. It's just not at the price they're used to paying and at a $3 billion company saying, okay, Sunrise will cost 120, Ningen 120, Burrow 120, okay, 360 million of cat-backs. Oh, and we're going to have to pay 1500 or 2000 or 2500 kilos instead of 800 or 1000. We're talking about quantums of dollars both on the call up front capital. If theoretically they were doing this all in house, that's real percentages of what their market cap was only three years ago. And it's do they actually want to increase the price of a bloom box by two or three percent with that have changed pre AI the amount of bloom boxes they could sell given they were not a base load go to solution. They were a marginal solution. Now it's completely different. And this is easily solvable for them, but that conversation has to happen within that organization. Are they actually confident they can deliver 120 tons of incremental Scandium in 2030 without the price going up. And also can they do that without being reliant on quote unquote Asia? I mean you say 25 gigawatts there and in your number factory in some thrifting already you're talking about 30 tons of gigawatts on a market that this year was what 60 tons if you say 60 consumed 80 produced. And I think I'm going to just take a shot in the dark here. I think we probably know who bought those who holds that inventory would be my conservative guess. Yeah, but we're talking about a market in its entirety growing many many multiples to get to that number that they say they have line on side line of side on right. Yeah, I mean you've been in the market a bit longer than us, but is there anything comparable that you think back on any market commodities? I think it's I think all of us were too young to be in the trenches during the China super cycle, which was the everything cycle, but I would say lithium over the last 12 years. We've gone from a market that was like what 0.5 T LCE and now we're talking about one and a half else ton million tons LCE we're talking about three million tons like we're talking like but it's it's it's the lithium growth it's the growth of the lithium market once EVs really started to get cool in the mid 20 tons just great divide the factory got built divide by a thousand everywhere you know it's it's the same dollar price per kilograms set of a ton of C6 the lithium numbers were already so small for everyone to wrap their head around. Yeah, so I mean let's just run it back except we're doing titanium waste streams all over the world and we're doing H-PAL and new self whales which I mean I don't need to say anaconda nickel but we all should acknowledge that these are okay these are very small autoclaves this is like a quarry with like what a nickel coming would call a pilot plan autoclave but yeah I mean nothing's perfect in this world but this is what you dream of in a commodity you dream of demand being pulled in an almost price bloom is price is price sensitive due to historical organizational behavior and survival instinct but I think if we talk to an Oracle a Brookfield a hyper scalar let's use my 30 tons of giga lot 1000 a kilogram that's a million a ton that's 30 million of scandium and a giga lot of compute metaphorically let's take that up to 3000 a kilo which would unlock the new self whales belt it would also unlock titanium waste streams and we'll probably get the nickel cobalt guys to at least think about putting on circuits so look I'm not nearly I'm not that emotional guy so we probably know the right number starts with the tube not three but come on come on million becomes give me a bit of freedom put a put a five there when one day Travis one day but 30 million becomes 90 million 60 million of incremental cost on a 50 billion dollar capital bill for a giga lot of compute okay does that change whether or not you're going to build the compute no it's just not it's it's 0.1 percent like the price of steel moving would better more yeah the some of the best businesses are when you're a small but important part of a large a large inexpensive crisis or China or the China super cycle we need an electric grid copper went from what a dollar a pound to four dollars a pound in the 2000s the two companies come out with the two investors come out with these short reports on on bloom one of them you know lists a set of 10 questions what did you actually make of them stepping through and answering in that post because on the on the surface if you just take what you've just said it doesn't seem that earth shattering to go to some of the companies we've spoken about and start to talk about ceilings laws off takes these these types of things like why do you think they haven't gone about answering it that way do you think that just highlights too clearly the bottleneck I don't think you want. I don't think this is something you want to get deal with in public. It's been pulled into the public sphere. But I think it comes back to a cultural point. They don't want. I mean, what if Beijing wakes up and says, "We are in an AI race, like the space race in the Cold War with the US and these blue boxes are critical?" Okay, LB Group, you can't export anymore to bloomers, affiliates, whether that's Thailand, Korea, Japan, in route to the US. That. I just don't think that you want to have your. No business wants its vulnerabilities out there for public discussion. But at the same time, culturally, this is eminently fixable. So it's really, I think the market is going to be asking. I think Bloom's earnings in two weeks would be very fascinating because the question will be, "Show us a pathway to deliver the scandium for five gigawatts." And then, does that reflect a change in what you're willing to pet? I would not surprise me, by the way, if Brookfield Oracle and some of these hyperscalers weren't aware of this issue or this critical mineral pinch point. After all, I don't think people really were thinking about where the race came from when they started using these magical permanent magnets once upon a time. I mean, our whole generation is still pretty wired to this globalized, just in time supply chain philosophy that it'll be there when we need it. You mentioned the Pentagon and the implied pricing that they bought kind of flashes out at a number with the six at the beginning of it. But thinking about all the stuff that I have read about this administration and their enthusiasm to win the AI award. Do you think there's any. What do you think is the probability that they work with Bloom to not kind of blot the market to get something here? Given that's pretty important for them to win this, to have energy that's not insanely expensive so they can remain competitive with China. I think if I was advising Bloom or I was sitting in Bloom's Board of Directors, right now I'd be saying we are going straight to TCN where we're figuring this out because this is a question of national security and geopolitical importance. Now that it's out there and Hunter Brook went on CNBC, the cat's out of the back now. I think you have to go figure out because I don't see a world where Lockheed has done an offtake with sunrise for 25% of their initial production. You see the sunrise valuation. I don't think Lockheed is going to be paying 1,500 to Kilo. I also don't think for a second Lockheed isn't going to insist on first nation, most favorite nation status that says you can't sell to someone on more favorable terms than you sell to us. And what world, so either Bloom will not be buying from sunrise or Bloom will be buying at the Lockheed price. It's very difficult in commodities to create differentiated prices for basically the same products. Can you talk me through Reo Tanto's history here, JC? Ding, ding, ding. I think the funny thing, so Blue, so Rio under Yacob started to explore one lithium but also critical minerals and started with the three-ton plant at Cerelle with an attitude towards look with Alcan, they understand the aluminum angle, the aluminum alloys. And then they looked to expand to 12 tons after getting the three tons but they weren't seeing and they openly spoke about the Settuskandium symposium. They paused building that 9 ton expansion because they weren't seeing demand at a price that justified the returns on investment. The fact that Canada growth came in in late 2025 with 25 million Canadian and an off-take agreement, that's what unlocks the expansion from three tons to nine tons. I mean, at the end of the day, we are a mining industry. We need to make a return on capital invested. The tech bros don't understand, we don't get valued on TAM. We don't get to brag about when we bought our Tesla stock or which round of SpaceX we got access to privately. Our stocks and our companies are not the digital mail, Burkin bag equivalent that we flexed to each other about. The mining industry is. You brag about when you sold the mining instead of when you bought. It's about making money. We actually look at the income statement. We look at the IRR's unspended money. We are not an optimistic bunch. I mean, valuations in the rare earth space just to quickly put the point in there have been pretty wild the last kind of couple years. If you're talking about net profit and you look at something like Lannus or maybe N Pay Materials, then there's a few ways to cut it. Completely. I think the generalist money coming in, if we. I thought about this a bunch and the way I would distill it down is our rare earth and critical minerals, not commodities anymore, are they defense inputs? To be compared to a defense contractor, a defense prime where you almost have a protected market because no, you're not buying a fighter jet made by Russia or China for obvious reasons to be totally extreme. You're also not going to use a magnet made in China anymore. I mean, that's kind of, I think, the question that if we are decoupling from critical minerals dependency in China, then you have to assume that the critical minerals western industry is going to be protected and also, therefore, be more like not a commodity but more like a defense contractor. Which we'll find out in due time whether that turns out to be the correct thesis or not. I remember that I'm trying to think of some of the. I remember when RIO bought that Platinum project for like $15 million just thinking, "What is this?" This is so irrelevant to RIO. I remember another analog that someone taught us once about. RIO with their aluminium business. They've forever had the ability or the possibility to add a gallium circuit and they never did because it's just this irrelevant rounding error. But I would have bought them so much geopolitical heft had they have done that and they didn't. I just find it so interesting why they saw the light with Scandium early but they just were absent with gallium. I think that there's. Besides the solid oxide fuel cells, which to be frank, pre-AI, the opportunity was in the aluminium. Whether that was radiators, because the real power of it with aluminium is you can kind of weld aluminium Scandium alloys together and the weld is as strong as the rest of the sheet. So if you think about the wing of an airplane, all the rivets, because normally they rivet together pieces of aluminium. Yeah, it allows for incredible light weighting but also it allows. Aluminum handles feet a lot better when it has Scandium in it even if it's 0.1%. But it's been getting. Airbus built a motorbike out of Scandium, their proprietary Scalmeloi, which is also used by F1 teams in some parts. So the question is you're dealing with, okay, how do you get from that to. Okay, and this is with hindsight, I look back on the last 12 years, where was maybe I way too early and didn't appreciate? Is if you're going to get this into cars or planes, one, there's the certification timeline, but there's also the question of, people want to see, oh, there's 5, 10 tons and it's in this car already. Oh, we can play around with that because there's already a supply of it. And so it's easier to get these bigger 10 ton, 20 ton demand growth drivers if there's already a base of 5 or 10 to work off of. And that's why I look, I think this has always had a. Freelinus said this, I've said it. It's Scandium said a chicken and egg problem for so long. This AI demand shock is kind of the. I won't say who said this to me, but it was a person with one of those projects besides Ningen in New South Wales. All that's missing is a killer application that just needs this. And it turns out these data sensors need solid oxide fuel cells because you can't get gas turbines, you can't get a grid hookup. And solid oxide fuel cells need Scandium or E3M. And here we are. and it doesn't really matter to the end consumer what the price of the scandium is. It's just, you know, as all things with cycles, it takes a little time for these things, for these recognitions to kind of flow through. So I know in the piece I was like, you know, Tesla won public in 2010-2011, Lithium started to get crazy in 2015-2016. China joined the WTO in 2001, and when did the Kamyi Supercycle kickoff, O304? Like it just, it, it, chat GPT comes out November 2022. Last year we started talking about Claude and Anthropic, and Bloom's stock started running the last two years. Sunrise started running last year. You kind of see that it takes a little while for the, oh, well, we can always get the power hook up for the DASR, oh, we can't get the power hook up. Oh, well, let's go with a bloom box. Oh, let's get the bloom box. Okay, well, we need a lot of these bloom boxes. You can produce all these bloom boxes, right? Yes. Okay, we now, we now need to go upstream and get backfill our supply chain. For a business that's 5x, what it was a year ago. That's, you see, this is how Supercycles happen. Now, that's why I think the whole thing is Hunter Brook and Crossroads are so right to point out this bottleneck for Bloom's technology. But it's so solvable. The way we solve all Supercycles have demand, it's, you just pay more for the building blocks because the economics pass through entirely through the value chain. The miners actually get a chance to make some money given we chose this stupid industry that just gives us PTSD nine years out of 10. And everyone kind of guessed what they want. It's just going to cost a little more than everyone thought it was going to. What do you think of the other weaknesses in, in like the thesis? I mean, there's all these other points that people who are focused on different parts of the chain would, would look into. But is there any data out here that keeps you up at night? On the specific detail of the demand poll for Bloom is what's giving Scandium this window. Is there a technology that can replace Scandium definitively at the same efficacy in, in these Bloom boxes and in other, there are sold oxypucils that don't use Scandium. So there's always that substitution and thrifting risk. I think other question is do you see a slowdown in this AI CapEx where instead of needing five gigawatts in 2030 they end up saying, we only need three and it's not because of raw material. Inability to get it, it's our customers are slowing our demand, our slowing on their spend. It's all the standard I would say risks. But if you want, if you want material in 2030, there's only so many fully permitted and shovel ready projects in the world. But you also then need to think about for everything else you need. People need time to actually go get approvals and actually build these things. You can't just turn these things on overnight. Mines are not like social media apps that just the next time someone downloads it, it's zero marginal cost. That's what the tech, the tech gazillionaires are starting to realize. The marginal cost of incremental supply is not zero. The way it isn't software. How will we win with Moningen Space? Well, I know maybe you put Scandium aluminum in the rocket, makes the lifting cost cheaper and we'll get a three-time platinum media and crash the platinum market. It's been a wicked to talk through this one with you. I think you guys are out saying you've got a ding-ding-ding on your side, your long Scandium. So just. Yeah, of course. But hey, I think it comes down to Bloom waking up and realizing they just have to pay up and their customers can easily afford the price push. This isn't an issue for them. Miners don't build for zero return on spec. Because if you build a mine into a market with dominant customers, you are not going to get the price you need. Twiggy does not build fordiscue if the price of iron ore paid by China is equal to what his cash cost would have been when he made the decision to build. Simple as that. There's a reason I or went from 23 bucks a ton to over 100 in the China super cycle. I have a 200 if you want to get really excited. But look, but what happened? You got a four-to-scue out of it because you could take something that otherwise you thought would be left in the ground because they needed to balance the market. They needed to satisfy the demand. I mean, the unique piece with iron ore, against pretty much every other commodity example, is just the scale that four-to-scue was able to do it at. It's just kind of staggering. There's just no thrifting or getting around steel as necessary in infrastructure in the build-out. China had. Just divide by a million. You don't need to be in a high-norbed billion. Just take one billion. I've done the job. But does Twiggy get the money if people can't look at what's happening in iron ore in the iron ore price and say, "Oh, this actually makes sense to build because you can produce it for less." China is so desperate for this material, even at a discount to re-on-bhp. It works. Yeah, the comparable example comes right through. These quilets are always awesome to chat, looking forward to having another deep dive with you in the not-too-distant future on a wider variety of things. But I appreciate you putting scandium on our radar well and truly, only eight, nine months after SRL has had its staggering run. I was trying to have a relaxing summer. I think August and September are going to be very busy and exciting times. And then this came along. So, yeah, I'm sure we'll be chatting real soon. Absolutely. Cheers, Kuala. Been too long since we have had the Kuala on. And as we hinted out at the end there, going to have them on again shortly to talk about everything going on in our awesome world with mining. Massive thank you to our fantastic partners, Sandvik, Ground Support, Focus the platform by market tech, Interlinks and Exceed Capital. Now remember, I'm an idiot, Jay Gays an idiot. If you thought any of this was anything other than entertainment, you're an idiot and you need to read out a disclaimer.