Welcome to the First Edition of The Weekly Briefing and thank you for joining us!
Who we are:
Digital Materials is a research, advisory, and investment firm hyper focused on breakthroughs in the world of basic materials and commodities. We believe that we are on the cusp of a massive transformation in the way objects are made and what they are made of. We intend to be a close to materials revolution as possible!
What is the Weekly Briefing:
Our weekly briefing, published on Thursdays, is our chance to think out aloud about a company, concept, or trend we find interesting and important as part of our overall research effort. This is a raw feed meant to provoke discussion and ideation; we come with strong beliefs, loosely held, so we ask for your feedback!
Become at client:
If you interested in finding out more about what’s coming next to disrupt agriculture, forestry, chemicals, and metals, consider becoming a client with DM+.
In addition to the weekly briefing, DM+ Clients receive monthly reports that dive deeply into the opportunities that we see in one of our Mega Themes.
As we grow, our DM+ offering will expand, offering clients enhanced information flow and access.
Of course, if you think we can help you directly with strategy, investment or advisory work please get in touch.
We are both active in the market raising special purpose vehicles for opportunities we think are amazing and available for individual research projects.
You can reach us at bill.birmingham@digital-materials.co
This Week: Mining’s quantum leap.
TLDR: Kobold Metals brings mining into the information age with AI.
Tags: Metals, Company, Copper, Cobalt, AI, Critical Minerals, Africa, Geopolitics, Cold War Redux, US IRS, AGOA, Green Premium, Crypto/Blockchain,
What is it?
This week the Wall Street Journal revisited AI-mining unicorn KoBold Metals in a long-form tech piece highlighting how the startup is changing the way the mining industry does exploration and production (E&P) for critical minerals.
KoBold’s data collection methods run the spectrum from physically deployed, cutting edge muon detectors to finding and digitizing old geologist’s paper notes buried in dusty mining bureaus. In between are a host of other novel data collection and analysis technologies that disrupt a very staid and conservative exploration playbook. The net result is a growing geological database primed for the application of artificial intelligence with the goal of identifying previously unknown critical mineral deposits.
Kobold has ~60 projects under exploration, the most promising being a Tier 1 deposit named Mingomba in Zambia capable of producing ~300k/t of copper per year. In a world where, at minimum, the electrification of transportation plus rapid data center growth will require between, on average, three new copper mines of this size per annum, the need for a step-change in how we source minerals is overdue.
Why it’s interesting?
KoBold’s ability to successfully deliver on its pipeline of production would put it ahead of global mining majors, implying at least 40x upside from its recent Series C valuation to match, say, Southern Copper’s market cap (SCCO) of $80b; though if our thoughts on TerraShed and the CRO model of mining are correct (see below) we suspect it will be much higher. The more important question will be how long it will take and can KoBold navigate the risks in between.
KoBold’s database, named TerraShed, is estimated to contain only 3% of exploitable global geological data. Clearly the age of intuition driven, wildcat exploration is over. The next era is more about “drilling for information” than the ones just drilling for ore.
The fact that KoBold’s flagship vein in Zambia sits underneath a depleted surface mine, in the heart of the copper belt, undiscovered by the previous owner, says a lot about the new type techniques that are needed to find the next set of resources. The long lament that mining needed to digitize is now finally changing. The algorithm replaces the heuristic.
Along these lines, what is the right model for KoBold? Traditional exploration budgets among the majors for copper alone has ballooned to ~$3b in 2023 with only around a 1% hit rate to show for it. KoBold has brought money-ball style analysis to mining spending only $60m per year on sixty projects across four continents and it has already discovered a viable Tier 1 mine.
AI in mining feels a lot to us like Contract Research Organizations (CROs) in Pharmaceuticals. Similarly to mines, drugs have only a 12% hit rate and take around 12 years to go from discovery to commercial sales while facing a host of risks along the way.
CROs have reorganized the pharma value chain by bifurcating R&D from production and the result has been ~11% CAGR for the past decade. We would expect TerraShed and a wave of AI competitors to reinvent mining finance, perhaps in the CRO model, hopefully using decentralized finance. Again, the entities who own the data, rather than the mine, will be where the ultimate value accrues in the supply chain. Digital Materials certainly hopes to be involved.
In the meantime, the majors seem to be moving in the opposite direction. While AI is a staple topic on earnings calls it usually refers to mining site process automation and lowering production costs. A worthy goal but also in line with the old adage that incumbents still get disrupted despite having every opportunity in the world to co-opt the disruption. As far as we can tell, most majors are increasing supply through acquisition. If there are any real in-house efforts that can rival KoBold please let us know.
Implications for the Copper Market
What does AI ultimately mean for the supply curve of copper? We won’t speculate here on the near term moves in copper supply, demand, and price (we will if you ask us!); Until we get more detail from the company, we would rather highlight the longer term cost implications of AI in mining E&P.
First, the application of technology to natural resources is always deflationary in the long run. This is good news no matter what your concern (economic or political).
In this case, there is the first order price impact from simply increasing supply into a market that is expected to go into deficit by 2026 and remain tight for the indefinite future. We will dig into the trend of “electrify everything” in a future mega report but for now, the current supply outlook alone is enough to justify rationing to at least 2050. There hasn’t been an Escondida size find in a decade nor will there be using traditional methods.
By contrast, KoBold is showing not only that there is new copper to be found but that exploration can be done faster, cheaper, and most importantly at higher grade ores (3.5% at Mingomba v. .39% average in the US). More copper is coming.
Of course, there is a catch. In this case, it is the fact that the Mingomba vein is a mile underground and will require expensive shaft mining versus and overall higher production costs than is currently seen across the bulk of the cost curve, which is open pit. We don’t see the KoBold’s edge in the production space such that it impacts these costs yet. Further, the new AI driven supply will more than likely all be deep underground.
As the ten major mines that make up the bulk of global production deplete, the push of cheaper exploration versus the pull of more expensive production will an important topic. Again, we look to the CRO model and more specialization on both sides.
As a reality check on the deflation argument, the long dated tail of the CME futures curve (Jul 2029) is pricing copper at around $4.27/lb, which we consider to be the “market’s” current estimate for long term breakeven production costs. Compare that to the C1 (direct mine cost) + sustaining capex of the very flat copper supply curve at ~$2.30/lb and the market is signaling that serious inflation is still the expectation. While the market surely knows about KoBold it’s not convinced. Not to mention that Mingomba is not scheduled to produce until 2030 at the earliest. To us, the long term break even for copper is at least 20% lower.
Africa: Things can’t fall apart
Last but not least, we are very intrigued by the geopolitical shifts that the KoBold effort in Zambia portends - which is a word that does not always mean negative outcomes! What is clear is that the US is waking up to the reality that Africa is the front line of the Cold War Redux.
China, through Belt & Road, and Russia, through the Wagner Group, have been playing the realpolitik long game in securing critical mineral capacity in Africa. The US, on the other hand, has lost close to a decade of opportunity in a region that holds critical minerals in abundance. While the Zambia maintains friendly relations with the United States, and the Lobito Corridor (rail from Zambia through DRC to the Angola coast) project is an excellent start, there is still the outstanding issue of what US policy is towards African countries in general.
We were surprised to hear KoBold’s Director of Government Affairs, Jennifer Fendrick, refer to KoBold as the American “beachhead” in Africa. This represents a small, but exciting, shift in thinking away from the previous mindset and policy preference that we can either recycle, friend-shore, or domestically supply critical minerals. It’s not going to happen. We need to embrace African supply ethically but effectively, recognizing that the rules of engagement have changed when we frame critical minerals as national security priority and Africa as a key part of our national interests.
With that in mind though, how far will the US government now go to protect the interests of US driven mining efforts of the long haul? For example, Zambia currently holds a 20% stake in Mingomba. President Hakainde has persuasively stated he thinks Zambia, a poor country that has benefitted little from decades of copper extraction, deserves 30% We don’t think this puts this project at risk as the US and Zambia currerntly have cordial relations. We do think it speaks to age old risk in mining of nationalization.
How, given the urgency of the critical minerals goals now, compared to the long dated time frame for mining, does the US sets up for dealing with successive administrations from a security guarantee perspective? We won’t suggest NATO because Africa is now a young continent that associates colonialism with the alliance, but a similar first principles effort needs to start now.
When you go next door to the Democratic Republic of Congo, which holds 76% of the current global supply of cobalt, almost exclusively financed by the Chinese, business done by US companies operating alone, without a creative alliance, then KoBold’s odds of success diminish.
One simple step in the right direction is to renew and modify the African Growth and Opportunity Act of 2000. (AGOA). Currently, the US Inflation Reduction Act (IRA) requires EV’s claiming the Section 30D tax credit (which, at $7,500 is all EVs) use critical minerals in their batteries that are at least 50% sourced from the US or countries with reciprocal free trade agreements with the US. By 2027, this amount moves up to 80% If the US wants the fleet to be electric, it needs to stop exclusively incentivizing domestic processing and start crafting a foreign policy that deals with upstream supply.
As we’ve mentioned, conditions for supply are not perfect for IRA requirements to be met by US sources and its allies. The US needs Africa and needs to embrace Africa. Re-affirming AGOA as a reciprocal trade agreement both keeps the IRA on track and sends a signal to African countries that they are indeed strategic partners. From there, much more robust specific trade agreements are possible. If this sounds like fantastical thinking then you are right. Yet we are optimistic but also realistic on all fronts. We quote Thomas Sowell here: “there are no solutions, there are only tradeoffs.” For KoBold and the Quantum Leap in Mining to succeed, the US needs to start making trades.
Putting real world assets “on-chain”
There is a lot more to say about critical minerals policy and copper, but we would be remiss not end here on the opportunities for the materials themselves and the workflows associated to become digital. While we will be talking a lot more about reserves as intangibles and the outright ownership of commodities through decentralized blockchains (“real world assets on chain” being the term) the near term opportunity for KoBold is to put post-trade execution on a decentralized platform. We’ve seen massive success in agriculture markets with Covantis; the value of similar efficiencies in metals would be tremendous.
Talk to us about how this can happen!
THIS INFORMATION SHOULD NOT BE RELIED UPON AS INVESTMENT ADVICE, OR A RECOMMENDATION REGARDING ANY PRODUCTS, STRATEGIES, OR ANY INVESTMENT IN PARTICULAR. THIS INFORMATION IS STRICTLY FOR ILLUSTRATIVE, EDUCATIONAL, OR INFORMATIONAL PURPOSES AND ISSUBJECT TO CHANGE. CAREFULLY CONSIDER THE RISK FACTORS, INVESTMENT OBJECTIVES, FEES, EXPENSES, AND OTHER INFORMATION ASSOCIATED WITH ANY INVESTMENTS BEFORE MAKING ANY INVESTMENT DECISION.