Rare Earth Elements - The Physical AI Conundrum

Podcast Transcript

Ed Coyne: Hello, and welcome to Sprott Radio. I'm your host, Ed Coyne, Senior Managing Partner at Sprott. I'm excited today to have a new guest on our show, Ryan Castilloux, Managing Director at Adamas Intel, a market research and advisory firm specializing in critical minerals and materials. Our topic for today: rare earths. Ryan, thank you for joining me today on Sprott Radio.

Ryan Castilloux: Thanks for having me, Ed. Great to be here.

Ed Coyne: Ryan, as a new guest to the show, I thought it'd be helpful to get a bit of background on you and some of the work you're doing at Adamas, and then we'll go into our topic of rare earths today.

Ryan Castilloux: My background is in geology, which feels like a lifetime ago, where I worked in nickel, copper, cobalt exploration and mining in the Sudbury region, up at the Raglan camp in northern Quebec, with Falconbridge, Glencore and Xstrata. 2009 came around, and I found myself working on the exploration end of things. With the global economic crisis of that time, it was one of the first areas to be hit. I took off to Europe, completed an MBA program, and upon graduating, I was in the Netherlands. If you've been, you'll know there's not a rock to be found in the country. It's essentially table-flat.

I was keen to work in something mining and mineral-related. That's where my passion is. I found myself working for a technology scouting company, advising on the latest and greatest technologies for green buildings and smart cities. Think energy-efficient air conditioners, light bulbs and building management systems. I noticed a common concern amongst the clients I was speaking with about rare-earth elements.

This was just after China cut off supplies to Japan, leading to price spikes by orders of magnitude in 2011, which really piqued my interest. It ultimately led to the founding of Adamas Intelligence some 14 years ago now, where we've been covering the global rare earth market and industry, from exploration to mining, processing, magnet making, end use, policy, prices and so on.

Ed Coyne: The term rare earths is something that's obviously hit the media more recently for various reasons. You've been at it now for 14 years with your company; you've obviously been ahead of the curve on that. I think for a lot of our listeners out there today, it'd be helpful just to get a general 101 on rare earths. I know there's heavy and light, that kind of thing. Maybe just give us a general lay of the land of what rare earths are, and then we'll go a little deeper into it after that.

Ryan Castilloux: If you can take yourself back to chemistry 101 and recall that periodic table of elements, the rare earth elements include what are referred to as the lanthanide series. It's the elements spanning from lanthanum all the way through to yttrium, as well as scandium, 17 elements that I like to think of as the superheroes of the periodic table. They're subcategorized as light or heavy, as you mentioned, based on their electron configurations.

What's important to note about those two sub-categories is that the heavy rare earths are indeed rare. They justify the terminology, whereas some light rare earths, such as cerium, are more abundant in the Earth's crust than copper. In terms of uses, rare earths are literally all around us, whether you realize it or not. They're in the magnets that make your phone vibrate. They're in the speaker that you may be listening through. They're in wind turbines that you might see on the horizon.

They're in electric vehicle traction motors, micro motors and sensors across all types of vehicles, as well as catalytic converters, the catalyst used to crack fuel. I could go on and on. They are truly all around us, are often used in relatively small amounts in applications, and have proven very challenging to duplicate or replace with alternative materials.

Ed Coyne: Is there any technology out there at all that could duplicate what rare earths are currently doing or what magnets are currently doing as it relates to operating in systems and motors and so forth?

Ryan Castilloux: Yes. In most cases, there are alternatives you could use, but they come with trade-offs, whether it's performance, cost, or the physical size of the devices we interact with. Were it not for neodymium-iron-boron magnets, which are the main type of rare earth magnet in many of those applications I just described, you would probably need help carrying your laptop into the office, or your cell phone would be in a backpack. These rare earth magnets have enabled the miniaturization of the gadgets and electronics that pervade modern society and enable the conception of the next generation.

Having a small computer allowed for a small phone and a smart watch, and the story continues. In the electric vehicle market, our data shows that around 93% of all electric vehicles produced to date use a rare-earth permanent magnet motor to drive the vehicle forward. Alternative motor types include induction motors and electrically excited synchronous motors that don't use rare earth permanent magnets. Those motors tend to be less efficient. They use more of the battery's power per unit of driving distance.

There's a trade-off: either the car is less competitive than alternatives on the lot, or the automakers have to beef up the battery size to compensate. Long story short, there are alternatives, but they come with a trade-off, and nobody is rushing to embrace them if they can get their hands on the rare earths they need.

Ed Coyne: It sounds like, to your point, it has to do with scale, size and weight. I think about the old war movies with the phone operator guy with the backpack. Let's talk about the sustainability, not from an environmental standpoint, but from a functionality standpoint. Do these magnets then have to be replaced? They break down. Do they break and stop working? How sustainable are these magnets once they go into a motor? How dependable are they?

Ryan Castilloux: It's a great question. The answer is in the materials' names. They are permanent magnets, which means they are permanently magnetized. There's a big, important asterisk that I think should go after that. It is possible to demagnetize a permanent magnet by exposing it to temperatures above its threshold, which allows the grains to disorient, or by exposing it to an opposing magnetic field.

That means that when you're designing an electric vehicle motor, and you're choosing the right grade of magnets to use, you need to understand the operating conditions and what kind of peak conditions that motor might experience. There's always a risk of demagnetizing the magnets, rendering the vehicle basically useless until it's fixed.

A takeaway there is that magnets are not a commodity. There are over 100 different grades of sintered neodymium-iron-boron magnets commercially available, and depending on the application and the conditions the magnet will experience, this determines which grade you need, its composition, and which rare earths it contains.

Ed Coyne: If installed and used correctly, it is a permanent solution to that mechanical item that it is being put into.

Ryan Castilloux: Absolutely, yes. The vehicle itself, if we were talking about that scenario again, would break down long before the magnet ever would. There's still an opportunity to recycle that magnet, but you wouldn't do so in its physical form. It would have to be ground up and reprocessed into a new magnet for a new application, possibly a motor again.

Ed Coyne: They can be recycled then, but you have to bring them back to their original form and rebuild them into their new form. Is that how it has to happen, effectively?

Ryan Castilloux: Yes. Today, that's how it has to be, because there's a lack of standardization on the end-user side. Even within a company, between vehicle models, the motor, the magnet or the dimensions might change. They do have to come back in. Today, there are two conventional approaches. One is a long-loop process in which the magnets return further upstream into the leaching stage. You would need to separate the rare earths into oxides, reduce them back into metals, produce the magnetic alloys and then produce the magnets.

The other one being developed is a short-loop process where you get the magnets out of their host, their end-of-life device, using hydrogen to decrepitate and break them down into powder. They do not oxidize in that state, so they return to a metal input further upstream. One has a lower environmental footprint, but it's maybe more technically challenging.

It's exciting to see that space unfold, but ultimately, for any recycler in the rare-earth space, technology is just one frontier they need to conquer. The other is securing access to feedstock, because it's not readily available in a bin at the scrap yard. You have to find this stuff and try to secure predictable flows of that feedstock so you can grow your business.

Ed Coyne: That leads us then nicely into the supply-demand dynamics. Let’s start with the overall size of that market. How much is trading not from an investment standpoint but from a pure feedstock standpoint? Every year, how much is coming online? Where are we in that ecosystem right now, with the supply-demand dynamics? Clearly, we've all read about the demand side. What does the supply side look like?

Ryan Castilloux: Globally, the consumption of rare earths last year amounted to US$10 billion. It's amazing because it feeds into and is critical to trillions of dollars' worth of downstream industries, including defense, consumer electronics, robotics, automotive and many others. Despite that seemingly low value, they have the potential to stall trillion-dollar industries downstream. I think we've really seen that come into play over the past 12 months or so.

In terms of the market dynamics and the supply demand balance, I think there's an important piece of the puzzle to come to terms with up front, and it's that while there are hundreds of different end uses of rare earths, some I described up front, hundreds of end uses that fall into one of eight end use categories, including catalysts, ceramics, battery alloys, polishing powders, metallurgy, alloys, phosphors and permanent magnets.

Permanent magnets alone are responsible for approximately half of global demand by volume, but over 95% of global consumption by value. Permanent magnets are the bread and butter of the rare earth industry. Therefore, the four core rare earth elements used in those magnets are the economic focus of the entire industry. That's neodymium and praseodymium on the light rare earth side of the spectrum, and dysprosium and terbium on the heavy rare earth side of the spectrum. Those are what we call the magnet rare earth oxides, or the MagREOs.

The supply-demand balance for those elements is generally very tight. It is the demand that suppliers are chasing. They're dictating, strategizing and planning their future production to chase demand for neodymium (Nd) and praseodymium (Pr), on the light rare earth side of things. If they're a heavy rare earth producer, they're ramping up production to chase the demand for dysprosium (Dy) and terbium (Tb). Those elements remain closely balanced on the supply-demand side, but as a consequence of chasing the fastest-growing elements in the basket, many of the other rare earths are being sacrificially overproduced.

Elements like cerium, lanthanum, yttrium and numerous others are being pulled out of the ground just to get the neighboring neodymium, praseodymium, dysprosium and terbium. That leads us to what is referred to in the industry as the balance problem, or the basket problem, where some elements, like magnet rare earths, need to cover the cost of production and losses on the others, which are chronically overproduced.

Ed Coyne: Who is the biggest player in rare earths today, and how do they become the biggest player? Talk about that, and then I want to go backward and talk about the U.S. and where we sit in that ecosystem right now.

Ryan Castilloux: China is, by far, the biggest player in the industry today. It dominates every step of the value chain from mining to refining into separated oxides, to production of metals and alloys, which are the inputs for magnetic alloys, which themselves are the cookie dough from which the final magnets are machined. China dominates that entire value chain, and over the past five to seven years or so, it has increasingly dominated the downstream industries that feed off of it as well. It leveraged an initial upstream monopoly to move downstream and today owns it all.

That said, there's been a major turning point over the past 12 months, when China decided to restrict exports of several magnet rare earths and defense-critical rare earths to the rest of the world, and continues to do so today. That has galvanized governments resolve to fix this problem like never before. This $10 billion industry is paralyzing their trillion-dollar downstream industries and potentially their ability to defend themselves. Ultimately, it has driven a wave of investment and support into the rare earth industry, more so over the past 12 months than in the past 12 years.

I think it can be argued that we're witnessing a rare earth renaissance in the West, led by the U.S. and Australia. That is moving the needle as we've never seen. However, it's a multi-year challenge that is ahead. I think the U.S. can make great strides toward resolving this by 2030, but it will need ongoing investment and support in the years to follow, as demand will continue to grow. I think the pandemic was a stark reminder for many that it's dangerous to have all your eggs in one basket.

Then, just years later, this is a reminder for the entire rare earth industry and the technology industry that relies on it that we've done it again, and now we're really seeing a lot of action and investment, which is hugely encouraging, but the hard work is not over just yet.

Ed Coyne: I read somewhere that about 20 years ago or so, the U.S. was the China of rare earths. For whatever reason, cost, material, environment, whatever it was, we took a step back. We're clearly trying to get back into that chair again. Do you see, or are you concerned at all, that the U.S. is taking a very active approach, whether it's the government investing in some of these companies, creating effectively price floors to encourage more capital to go into this space? Is there any risk of the U.S. nationalizing the rare earth ecosystem?

Ryan Castilloux: It's a tough and contentious question to answer, depending on one's view on government intervention in free markets. I think the evidence I would point to is just looking back over the past decade or so. In 2011, when China first weaponized rare earths and halted exports to Japan, prices rose thousands of percent, and the risk became crystal clear. For years after, the investor decks and government language around the issue consistently highlighted that this risk was looming and that we needed to address it, but the free market was not moving fast enough.

I think there was too much risk in looking into the rare earth industry compared to the battery materials party kicking off alongside it. Ultimately, it dissuaded investors who instead chose to invest in a lithium or nickel company. These companies just couldn't really get themselves to stack bricks or dig holes.

Some level of intervention was necessary. We could probably argue about different ways that intervention could go, whether price floors or equity investments are needed. I think what was needed was an intervention. That's my careful stance on that. I suppose there's always a risk of nationalization, but I don't think it's the intent at this point.

I feel a more venture-capital approach is being taken, where they're looking at who can bring us scale in the near term and who can bring us vertical integration. They're placing bets on those players, and started with the one that was furthest along, MP Materials, the lowest risk, and are identifying others that I think check those boxes at the same time.

Ed Coyne: Didn't the U.S. make a meaningful financial commitment to MP Materials? Am I correct in saying that?

Ryan Castilloux: Yes, they did. It's a public-private partnership. Part of it is an equity investment, and the other part is a price protection agreement. Rather than a direct price floor, they're providing a top-up for prices below a certain threshold, in this case, $110 per kilogram of NdPr. Today, the price is above that, so it's already turned into a net positive. There is some participation in it. The same goes for some of the landmark investments that followed. One in Vulcan Elements and another in USA Rare Earth.

Ed Coyne: I think that's probably a good sign because how do you draw long-term capital if you don't know if the price is going to be half of what it is today? That's very risky. On the flip side, it's so important to our national security. In the early days of our current war, I read that every day we were at war was consuming months, if not a full year's worth, of supply and demand for rare earths. That's a little shocking to read. From a growth standpoint, China is obviously the biggest player. Is it fair to say that the growth opportunities are probably in other places besides China, given that they have so much of it right now? Where will the growth come from?

Ryan Castilloux: Over the next ten years, we expect magnet demand growth to be the fastest globally, with the U.S. as the fastest-growing market. It's a function of a few things. Its base today is much smaller than that of China's, so we can naturally expect a faster rate of growth. It's more importantly a function of where we see the growth coming from: robotics and the low-altitude economy, encompassing everything from drones to electric vertical takeoff and landing (eVTOL) aircraft or hybrid variants of that.

All of those industries are very magnet-intensive. They're all industries where the U.S. is actively trying to lead. I think there's recognition in the U.S. that they've fallen dramatically behind China on electric vehicles, and it's not necessarily a race worth leaning into so much. They see robotics and eVTOL as other industries where they can really lean into and compete, so that's exciting. A humanoid robot, for example, can have magnetic mass that is multiple times that of an electric vehicle. As units scale for advanced robotics, the demand is astronomical.

Ed Coyne: It goes back to something we talked about earlier: the replacement or alternatives. The size and weight components are clear when you're talking about robotics and drones; it almost seems irreplaceable to me, because yes, you could do it, but the weight and size won't allow it. You don't have any alternative to using rare earths for these technologies. Is that too far of a stretch, or is that pretty fair to say?

Ryan Castilloux: That's very fair to say. I also think it speaks to a change in price sensitivity on the demand side going forward. If we look back 10 years ago, a lot of rare earth magnet demand was driven by price-sensitive applications, consumer gadgets and electronics, speakers where they could switch to a ferrite magnet, and maybe the sound quality goes down a bit, but it's not entirely mission-critical, and it's key to keep the price of that thing cheap.

As we move forward, robotics, drones, aircraft and hybrid or electric vehicles benefit immensely from the high efficiency of permanent magnet motors, which enable them to use smaller batteries or perform better on the factory floor. It's just a level of benefit to using rare-earth magnets in a lot of these future applications that will allow the price to rise significantly before it would ever make sense to switch to an alternative that is going to use more battery capacity or is going to do less hours of manufacturing work, and so on.

Ed Coyne: As you've watched this market evolve, what are some of the things that maybe have you paused with concern or are worried about as you're watching this market unfold? What has you concerned right now?

Ryan Castilloux: One is bottlenecks that I see as likely, as the alternative mine-to-magnet supply chains continue to be built up. Another is something I alluded to earlier: the post-2030 horizon. Right now, there's the right level of support for the alternative industry, and I think we could argue about the mechanics. We were talking earlier about the contentious ways it could be approached. My concern is beyond 2030, having the same level of support be maintained. With respect to bottlenecks, there are many steps from mine through to the production of a finished magnet, which then goes to a motor and into an electric vehicle.

The first is the separation of rare earths into oxides. The next is the production of metals, and then the rest is magnet-related. Today, both ends of that value chain are well supported. We have a lot of magnet factory capacity under construction or announced. We have a modest amount of mine production coming up, which is now being separated into magnet rare earth oxides that can feed into that downstream. Two bottlenecks are in the midstream. One is the reduction of oxides into metals that are the critical inputs for the cookie dough that magnets are made from.

Another is the production of heavy rare earths specifically. I mentioned earlier that there are over 100 grades of neodymium-iron-boron magnets commercially available today. Each of those has a different composition, which dictates its maximum operating temperature and strength. For a lot of these high-growth applications that we have talked about, robotics, electric vehicles, drones, eVTOLs, they require magnets that can operate at temperatures from 150°C to 180°C to 220°C and beyond.

For a magnet to preserve its magnetic strength and properties at that temperature requires the addition of dysprosium and terbium, those heavy magnet rare earths. In the Earth's crust, overall, the vast majority of rare earths we find in typical rocks or in average rare earth deposits are light rare earths. We have a pipeline with a modest amount of Nd and Pr, neodymium, praseodymium, to feed the magnet factories of tomorrow.

What we really need is also a certain amount of dysprosium and terbium to ensure that they cannot just produce magnets, but they can produce the specific magnets that the demand side is calling for, and that is those high-temperature performance magnets that have heavy rare earths in them. Just to summarize, the bottlenecks are around metals and heavy rare earth supply.

Ed Coyne: You mentioned 2030 a couple of times as we've been talking. What's special about 2030? Is that just a supply-demand clock that's ticking, or what's up with 2030?

Ryan Castilloux: It seems to be the tail end of a lot of announcements for building up magnet factory capacity. Companies that have been coming into the market over the past 12 months are stating, "We're going to build the factory that's going to be commissioned by 2028 or so and ramp up to X tons by 2030." We have good visibility out to 2030, where the market's going. There's potential for the U.S. to be self-sufficient, producing as much magnet tonnage as the U.S. market needs.

That's just on a depth basis. It doesn't acknowledge the full breadth of demand in the U.S. across all these different grades and compositions, which is not satisfied by tonnage alone. It requires assembly lines that are tooled up for those specific grades, form factors and coatings. That's a bigger challenge that won't be solved by that time.

Ed Coyne: If you listen to the Elons of the world, we're going to have robots in every one of our houses the way we have laptops in our houses today in the next decade or sooner. It doesn't seem like the demand's going away any time soon, if ever. It just seems like the supply needs to somehow catch up.

Ryan Castilloux: Yes, absolutely. I think back over the past decade and a half, recall when EVs seemed far off, and look at where we are now. There are so many other uses that continue to creep into the market because rare earths are the superheroes of the periodic table. They possess remarkable properties that make them the best material for an application. That doesn't just relate to magnets only. It's true with phosphors. It's true with a lot of catalysts, a lot of ceramics and optics, and so on. I'm confident that, as we look ahead to robotics and eVTOL, something else will undoubtedly come along and be just as impressive.

Ed Coyne: It's amazing. I remember my first job out of college, where I was marveling at the fax machine. I don't know if that speaks to how simple I might be or just how quickly technology has advanced over time, but I used to get the nightly closing prices for all of our funds by fax, and then I'd have to read them all over the phone. It's only about 30-something years ago. It wasn't that long ago.

Ryan Castilloux: Yes, and I'm sure it seemed stressful at the time, just like life does today. Somebody will look back in 50 years from now and tell us we had it easy.

Ed Coyne: I just prayed every night after the market closed that the thing didn't jam up because I didn't know how to fix it. I'd like to end on a positive too. What are some of the things that really have you pumped up about this space? You've been at it for a long time now. Supply and demand are picking up, so that's got to be exciting. What else is going on that has you enthusiastic about the rare earth ecosystem?

Ryan Castilloux: Historically, one of the biggest challenges for this industry has been end users' focus on securing the lowest-priced rare earths available. Dollars per kilogram ruled. It didn't matter where they came from, how they were produced or what risks were associated with them. A lot of that was driven by the fact that a procurement manager at a big auto original equipment manufacturer (OEM) or a tier one supplier operated in a silo, with their own bottom line to worry about, and it didn't matter what the effect was on the system. What's really encouraging now is that we've seen a big change in mentality, where procurement is viewed at the system level.

It's not just about the magnets in your EV motor. It's what are the implications on battery cost and system cost as a whole? The recognition among end users that they may need to pay more to get the supplies they need, to be produced under best-in-class conditions and transparent conditions, without the environmental impacts that come from Southern China or Myanmar, and that while they might pay more, they'll have more transparency into supply and more predictability on what the input price will be. One downside of just getting the cheapest material you could, and historically getting it from China, was that those prices were super volatile.

You might be happy with the price at the start of the year, but your magnet price doubles by mid-year, and that, all of a sudden, starts to chisel into what are really thin vehicle margins, making planning a challenge. There's a more mature approach to rare earth procurement now, and we see the majors willing to get involved with companies early on.

We should look at the GM and MP Materials relationship, which was established well before the magnet factory was built. In short, there are more legs under the momentum that we have today than ever before, with real fundamentals underpinning it. It's an exciting time to look ahead at how robotics and advanced air mobility overlap with defense tech, and how AI is about to become physical. That's all enabled by motors, magnets and rare earths.

Ed Coyne: I love that term: AI is about to become physical. That's very true. I use it daily now for research purposes. It's amazing how quickly things are evolving. Ryan, I can't think of every question to ask, so was there anything else that maybe you wanted to deliver today before this call that you said, "Hey, I really can't wait to talk about this."

Ryan Castilloux: I would just go back to robotics again, not to beat the topic to death, but in recent weeks, we continue to see developments that make us think that's being pulled forward even faster than we were anticipating. We're not quite as bullish as Elon Musk. I don't think there'll be 10 billion robots on Earth by 2040, because there aren't enough materials to make that happen.

We are very optimistic about the space and the numbers that we're seeing companies ramping up towards, and the deployment commitments for robots hitting factory floors are super exciting. If not in the next couple of years, I think by the end of this decade, we'll see tens of thousands rolled out in industrial settings, and eventually in home and military settings as well, which is a topic for another story.

Ed Coyne: Yes, it is. That's probably another whole hour's worth of podcast. Ryan, I really appreciate you carving out the time today to talk to us. Really interesting stuff here. Thank you for your time.

Ryan Castilloux: Thank you, Ed. It's a pleasure.

Ed Coyne: Once again, I'm Ed Coyne, and thank you for listening to Sprott Radio.