What We’re Reading (Week Ending 29 June 2025) - 29 Jun 2025
Reading helps us learn about the world and it is a really important aspect of investing. The late Charlie Munger even went so far as to say that “I don’t think you can get to be a really good investor over a broad range without doing a massive amount of reading.” We (the co-founders of Compounder Fund) read widely across a range of topics, including investing, business, technology, and the world in general. We want to regularly share the best articles we’ve come across recently. Here they are (for the week ending 29 June 2025):
1. China’s rare earth choke hold – Amber Zhang
Rare earths comprise a group of 17 elements, typically categorized into light, medium, and heavy groups. These materials are indispensable for making high-performance magnets used in both civilian and military technologies. Among them, medium and heavy rare earths — critical for aerospace, defense, and other cutting-edge sectors — are particularly scarce and difficult to source.
Don’t be fooled by their size. Rare earth magnets are no larger than a stick of chewing gum, yet pack magnetic force 15 times stronger than traditional iron magnets. Heat-resistant and cost-efficient, they are essential components in electric motors — not only in EVs and hybrid vehicles, but also in robots, drones, offshore wind turbines, missiles, and fighter jets…
…According to the International Energy Agency, China accounted for over 60% of global rare earth mining output in 2023 — and an even more dominant 92% of the world’s refining capacity. According to the International Energy Agency, China accounted for over 60% of global rare earth mining output in 2023 — and controlled a staggering 92% of global refining capacity…
…Between 2020 and 2023, 70% of the rare earth compounds and metals used in the U.S. were imported from China, according to the U.S. Geological Survey…
…Ford recently halted production for a week at its Chicago plant due to rare earth shortages, affecting its Explorer SUV line. In early June, the Motor & Equipment Manufacturers Association (MEMA), along with General Motors, Toyota, Volkswagen, Hyundai, and other major automakers, issued a joint letter warning that without a stable supply of rare earth magnets, production of essential components could come to a standstill…
…The U.S. once boasted the world’s largest rare earth magnet industry. Its Mountain Pass mine in California had supplied most of the global market since 1965. But in 1998, the mine was shut down following a pipeline leak that released trace heavy metals and radioactive materials into the Mojave Desert. Chinese firms made three separate attempts to acquire the mine — all blocked by U.S. authorities.
Alarmed by Japan’s supply crisis, the Obama administration supported Hitachi Metals’ investment in a rare earth magnet plant in North Carolina, operational from 2011 to 2013. But the costs were prohibitively high compared to China’s vertically integrated, state-backed operations in cities like Ganzhou. U.S. buyers, ultimately unwilling to pay a “made-in-America premium,” continued sourcing from Chinese suppliers. In 2020, Hitachi shut down the facility and mothballed its equipment…
…Back in 2010, Mountain Pass — the U.S.’s only remaining rare earth mine — received over $1 billion in Pentagon funding just to stay afloat. But lacking commercial competitiveness, it shut down again the following year. In 2017, MP Materials acquired the site, restarted mining operations, and began exporting raw ore to China for processing. The company now plans to begin producing rare earth magnets at a new facility in Texas by the end of this year. Still, even at full capacity, its annual output would match just a single day of production in China…
…Domestically, the Round Top project in Texas has emerged as a cornerstone of America’s rare earth strategy. Operated by U.S. Rare Earths Inc., the site holds estimated reserves of 130,000 metric tons across 16 different elements and aims to supply 20% of U.S. rare earth demand by 2027. The company is also building a $100 million magnet manufacturing facility in Oklahoma, which is expected to process up to 2,000 metric tons of rare earth materials annually.
Meanwhile, the U.S. Department of Energy has launched the ReElement initiative, allocating $50 million to recover up to 90% of rare earth elements from electric vehicle batteries by 2025. But these recycling systems have yet to achieve commercial scale and remain economically marginal.
The National Defense Authorization Act for fiscal year 2025 earmarks $1.2 billion for strategic stockpiling and $350 million for domestic development. These funds are being channeled into American firms like MP Materials, aimed at accelerating the construction of a domestic rare earth processing infrastructure…
…According to the Center for Strategic and International Studies (CSIS), the Pentagon has invested over $439 million since 2020 to develop a rare earth industrial base — but most U.S. production remains in its early stages.
RAND Corporation estimates that it would take at least 10 years and $10–15 billion in investment to establish a fully independent domestic rare earth supply chain, factoring in infrastructure, permitting, environmental compliance, and workforce training.
2. The Great Decoupling (or Why Your Clicks Are Down and Impressions Up) – Ryan Law
Impressions are increasing because AI Overviews now give companies two chances to log an impression for a given keyword: once as a “traditional” blue link in the search results, and again as a citation in an AI Overview…
…At the same time, clicks are decreasing because AI Overviews are increasing zero-click searches. Searchers can get all the information they need to resolve their query without leaving the search results page.
When we studied this at scale across 300,000 keywords, we found that the presence of an AI Overview correlates with a 34.5% reduction in clickthrough rate…
…While our clicks are tanking because of AI search, recent data from Patrick Stox shows that—at least on the Ahrefs website—visits from AI search convert 23x better than visits from traditional search.
The way content marketing functions is very different, but guess what? There are more potential customers in the world, more demand for products and services. That is the real determinant of growth, not clicks to a blog. We’ll find different ways to reach those people.
3. A Cheeky Pint with Meta CFO Susan Li (Transcript Here) – John Collison and Susan Li
Susan: When I think about it, I go back to when I was IC4 and I joined in 2008, I’m building these first revenue models. I’d gone from banking – which is super organized, super structured, they don’t even need to know your name, they just train you to immediately figure out how to find the backup to everything, so that two years later someone else can do this and so on and so forth – to there was no infrastructure. So I’m hunting down the exact engineer who has built some ad server so that he can tell me what the parameters mean. And of course, the next time he changes them, he’s not gonna tell me, and I have to go find him again, and he’s like, “Oh, she’s coming. Don’t look her way.” A few months in, I got a meeting invite for power users of SQL and I thought, “My gosh. I’d been getting a good amount of feedback about how things could be better, and here was finally this moment of recognition that – I didn’t even know how to write queries in SQL when I started.” I show up to this meeting and there are five other people and the meeting organizer tells us that we have been called because we are the five users of SQL who consume too much power. And we have just been churning with our massive joint tables through the…
John: I love that you were all called to the principal’s office.
Susan: Basically, yes. But I often think back to this because this was a data analyst who didn’t know any of us that well, but had just generated his reports of who’s using the most infrastructure and looked at the top people on the list and thought, “Okay, this person in finance, it doesn’t make sense why she’s the third highest person on the list,” and called us in and then taught us to write better queries. No one I think specifically told him to do that. I think it’s a little awkward when you call people in to do this, but he did it because it would make us all better at our jobs…
…Susan: So, there’s this very measurable part of the company and we generally try to trade those things off against each other when we’re evaluating things within that bucket and we generally try to fund the things that are positive ROI. I’m usually the person who’s trying to make sure we understand, for every individual experiment, the expected return is something, but that’s where we are on the curve today, but what about 50 experiments later? Does the curve still have the same slope?
Then there’s a set of things which we constrain more in terms of, there’s some envelope of investment that we’re willing to make that’s not in this really ROI-driven bucket. It is very difficult to pencil out what the annual revenue forecast for Reality Labs is gonna look like over the next 20 years. For bets like that, we invert the problem. But when we talk about the return on the investment, the question that we pose, as a finance organization, to Mark – and make sure that Mark and the board understand – is what does this have to be worth to pencil out at the end? Does that pass the sanity check, the intuition, about what the size of these markets can be based on maybe some comparisons to markets that exist today, but of course in another 10, 20 years, you expect that the world will look different and maybe those markets should be bigger or smaller for whatever reason. That’s the guide, which is, for this thing to succeed at the rate at which we’re investing, it needs to be worth this at the end and does that make sense?…
…Susan: I am not a tech visionary. There are many things I’m good at, but envisioning the future of the world and what I want it to be like is not one of them. I’m a very happy beneficiary of the technology built by the world around me.
But Mark very much has a vision for what he wants that world to be. And for him, I think the strategic imperative is that we have to be building these next states of the world for us to again, be a good business, but also just be a compelling company that builds technology and puts it out in the world and builds incredible experiences for people.
I remind people in the finance organization all the time, we are very good at skeptically evaluating each bet. But the point is not that we have to look at every bet and be like, “This bet is going to work.” The point is there is a portfolio of bets, and some of them are going to pay off massively beyond, in fact, what the case on paper looks like when you make the bet. Many of them are going to not work out, but the ones that pay off are gonna more than justify the overall investment strategy or the overall roadmap that you’re building toward. If we just allowed ourselves to nix everything that the paper-case didn’t seem high-confidence, then we would never make a lot of the important bets that have been really important over the history of the company…
…Susan: That is the question that I assume all of my counterparts at these companies and I are all thinking about. For us, there are the drivers of the way we’re investing in capex today. Of course, we have, first of all, just a massively-scaled consumer business and core AI infrastructure that powers all the ranking and recommendations work and so on and so forth. That’s always been a reasonably big number for us, but also because it was getting more mature that we were driving to be more efficient over time. Then now you have, among many of our peers and ourselves, this big investment to train what we all aspire to be, frontier models. If you use those models to build great and scaled consumer experiences, then how much inference compute you’re gonna need on top of that? If compute required continues to scale up in this way forever, then you’re gonna run into some true problems of physics. But hopefully, there will be different kinds of research innovations along the way that will unlock things like being able to distribute the training so you don’t need one extremely large cluster somewhere and that will help with a lot of the energy and other challenges. So there’s some question about what that looks like over time.
Then there’s this question about, “Great, you can build all this capacity, and what do you do with them if it turns out you don’t need as much compute for either training or inference as you thought?” I think a lot of us have different backup use cases. So, up to some point, we would use a lot of compute very happily still, in the core business and what we expect the core business to be, three years from today. But frankly, we’d use more compute in the core business. Now, that doesn’t scale forever. So the real question is what happens in like two years if you’ve built so much compute that you cannot envision a reasonable ROI on the backup use case if what you’re building doesn’t come to fruition. That’s something I think we’re all gonna learn in the next few years…
…Susan: As part of not wanting to miss the boat, we built out enough capacity for Reels but also for future things. We found that we were in fact able to put that capacity towards very good use – exactly as you said. So I do think an interesting question in the future will be allocating compute as a resource, It’s a muscle we’ve built later as a company, because we had gotten very good at allocating headcount as a resource, and headcount’s really easy to account for because you have org charts, you know exactly this person reports to this person, to this person, this person is incontrovertibly working on Facebook Marketplace, for example. GPUs don’t have that property. In fact, you often want to build out your infrastructure for it to be very fungible. Because you need to divert capacity to where – suddenly something has happened in India and you want a lot of compute to be available to be used there. So it’s not like this GPU is labeled for Facebook Marketplace, and this is labeled for – it’s actually quite a bit more difficult to account for where the capacity is being used at any given point in time. That means it’s harder to manage, and it’s harder to create the incentives around are you using GPUs efficiently?
John: You allow people to trade between people and GPUs, right?
Susan: In the budgeting process, we have allowed people to trade. Not too surprisingly, even though you’ll find that groups are often asking for compute, when that particular trade is on offer, people almost never trade for compute for exactly the reason I described, which is that if they get allocated 100 new headcount, there is no chance that 26 of those headcount will accidentally be working for something else.
4. My Trip to Washington to Get in Sync with Republican and Democratic Leaders on the Budget and Debt Situation – Ray Dalio
Everyone I spoke with on both sides agreed that:
- We are likely to have a big debt-economic crisis if we don’t get the budget deficit down to 3 percent of GDP, so 3 percent should be an agreed-on goal,
- Getting the deficit to 3 percent will require both spending cuts and tax revenue increases because if they come from just spending cuts or just tax revenue increases alone, the cuts or increases would be too big and shocking.
- It’s not possible for politicians to say these things publicly even though they believe them because they would be thrown out of office…
…So, our biggest problem is that our country’s political representatives can’t even say, let alone do, what they need to do to fix our debt issues because their constituents would throw them out of office if they did that. Such is the condition of our political decision-making system.
We discussed my idea of a “3 percent 3-part solution,” which would be to cut the budget deficit to 3 percent of GDP through a mix of spending cuts, tax revenue increases, and interest rate cuts. For example, cutting spending by 4 percent, increasing tax revenue by 4 percent, and lowering the real interest rate by 1%** so that the adjustments wouldn’t be unbearably large to achieve that 3% deficit goal. The leaders I spoke with said that they’d love to do this or something like it — in fact, they thought it would be wonderful if the “meme” of reducing the deficit in this way took hold in the electorate and there was public pressure to get it done.
As for where things are likely to go, there won’t be big enough changes to the current proposed budget to change the overall picture this tax year.
5. The Speed of Patience – Paul Higgins
To understand how patient preparation creates decisive speed, I’ll show you three different maps of the same territory I’ve found practical.
- Pace layers reveal where to be patient and where to be urgent, showing how businesses operate across multiple timescales simultaneously, from seasonal fashion to generational culture.
- S-curves illuminate when those layers will hit their inflection points, helping you recognize which growth curve you’re actually betting on.
- Trust as a leading indicator – what emerges from ongoing interactions across and between layers (employees, communities, customers and processes), the invisible asset that compounds for decades…
…I like Stewart Brand’s pace layering framework for understanding how businesses operate across time. It reveals why this matters so profoundly. In most complex systems, different elements change at different speeds. Fashion moves seasonally, commerce shifts yearly, infrastructure evolves over decades, governance changes generationally, and culture moves so slowly it appears frozen in time…
…Layers don’t exist separately, they form a single, interconnected living system which is sometimes hard to see. We tend to see layers as independent parts to optimize separately, but in living systems, layers are how the whole organism breathes – each rhythm nested within another, each movement part of a larger dance. The fast movements at the surface and slow currents in the depths aren’t separate phenomena but the system’s way of being alive at every scale simultaneously. Speed doesn’t come from stability – they arise together from the coherence of the whole system…
…Apple master this temporal arbitrage. New iPhone colors arrive every season to satisfy the fashion layer, while annual product cycles drive the commerce layer with reliable predictability. But the iOS ecosystem, which represents their true competitive moat, took twenty years to build in the infrastructure layer, creating switching costs and network effects that compound with each passing year. Their App Store governance evolves with glacial deliberation, each change carefully considered for its long-term implications, while their design philosophy – the cultural layer that infuses everything they create – hasn’t fundamentally changed since Jobs articulated it decades ago. You just have to look at their cumulative cash reserves to see whether they have the capacity to keep it up or not.
Competitors try to destroy Apple’s fashion layer moat and assume that’s the game being played. They miss the insight that Apple’s speed in the fashion layer comes from stability in the infrastructure layer, that the layers aren’t independent but deeply interdependent, with the slow layers enabling the fast ones to move with confidence and clarity…
… In business, you’re never riding just one S-curve. You’re managing a portfolio of them, each operating at different speeds across different layers of your organization. Your product adoption might be hitting exponential growth (measured in months) while your infrastructure build-out is still in early grind (measured in years) and your culture formation hasn’t even begun its curve (measured in decades)…
…Netflix understood this with brutal clarity. In 2010, they were shipping 2 million DVDs daily – a massive operation at the peak of its S-curve. But Reed Hastings saw streaming was at the bottom of its S-curve, barely functional, with terrible selection and constant buffering. While Blockbuster optimized their mature retail model, Netflix deliberately cannibalized their profitable DVD business to ride the next wave. They moved $200 million from DVD operations into streaming content when streaming represented less than 20% of revenue. Today Netflix is worth $240 billion; Blockbuster is a cautionary tale…
…Kerry Group’s transformation from Ireland’s smallest dairy cooperative to a €6.3 billion ingredients empire illustrates how patience creates opportunities invisible to those focused on shorter horizons. Every dairy producer faced the same challenge with whey, the protein-rich liquid left over from cheese-making that represented both a disposal cost and a compliance headache. While the entire industry treated this as expensive waste, Kerry’s leadership recognized something profound: they were looking at two different S-curves operating on completely different timescales.
The dairy business that consumed everyone’s attention was approaching the top of its S-curve, with margins thinning and consolidation inevitable, while the ingredients business hadn’t even begun its exponential climb. For fifteen years, Kerry invested in extraction technology and scientific capabilities while competitors focused on optimizing dairy margins. By the time health consciousness and specialized nutrition exploded into mainstream consciousness, Kerry had spent two decades perfecting protein extraction, understanding molecular structures, and building relationships with food manufacturers who needed exactly these capabilities…
…Warren Buffett’s 2008 moves exemplified how trust operates across all three maps simultaneously. While others mocked Berkshire’s growing cash pile – $40 billion sitting “idle” – he was building in the infrastructure layer (pace layers), preparing for the inevitable down-cycle in financial services’ S-curve, and accumulating trust with every patient year. That cash pile represented more than financial capacity; it was trust crystallized into capital. Every year Buffett didn’t chase returns, every quarter he resisted leverage, every deal he walked away from, he was depositing into an invisible trust account. When 2008 hit, that patient accumulation enabled lightning-fast execution: $8 billion deployed to Goldman Sachs with one phone call. The $7.7 billion total return exceeded Coca-Cola’s entire 20-year dividend stream to Berkshire. Trust had compressed decades into days.
Disclaimer: None of the information or analysis presented is intended to form the basis for any offer or recommendation. We currently have a vested interest in Alphabet (the company behind AI Overviews), Apple, Meta Platforms, and Netflix. Holdings are subject to change at any time.
