Picks and shovels and copper ⛏️
TLDR: Humanity must mine more copper in the next 20 years than we have in human history to meet demand.
The first model I built in banking was a copper mine model. Actually, it was two copper mines and two models but I don't like to brag. I spent a whole weekend in the office learning about sulphides, oxides, grades, recovery, crushing, flotation and much more. You take real-world engineering and finance bro it into a spreadsheet to spit out a valuation. God's work - discounted cash flow.
I stuck around the metals and mineral resources world because it was real (plus I liked the team at my bank). The stuff came from the ground. Someone dug it up, worked on it and it had a value. It was a fascinating world of large open pit or deep underground mines (sometimes both). Turquoise tailings ponds that looked like fresh mountain lakes. A landscape of junior exploration companies (think pre-seed - pre-product, pre-revenue, dirt and a pitch deck) all the way to the big guys like Anglo America, BHP, Rio Tinto, Vale, Glencore, Freeport (the FAANGs).
China drove the 2000s commodities super cycle with their insatiable hunger to devour anything that came out of the ground. The buying, building and boom of China's metal industry made them a voracious consumer but also a major producer of metals like copper and aluminium (and lots of others). By the time I joined banking, the super cycle had become a slow unicycle and as always, I had nailed my timing.
Over the past few years the move towards decarbonisation and electrification is driving yet another boom in the metals space. The miners have dropped their dirt-laden overalls and adopted a 'transition metals' or 'future enabling metals' suit. Your EV cars need copper, the batteries need copper, the wires need copper and so. In fact, on average, an EV car needs about ~83kg of copper which is 4x that of an ICE car.
"Screw Tesla. I care about AI and software". Those data centers in operation and the new ones being built to provide capacity for compute contain a lot of copper and other metals. So whilst I knew that we needed a lot more copper, I didn't know that we needed A LOT. In fact, it blew my mind when I read that humanity must mine more copper in the next 20 years than we have in human history to meet demand! And you stop and think for a second, "wait, what??".
Like all things that collect data, as mines and miners do and lots of it, you can optimise and improve with tech. Fractional improvements in recovery or processing driven improvements in the millions of dollars. The miners built their own tech teams and most failed. Seeing start-ups build in this space brings a smile to my face :) because let's face it (and no offence), I trust founders a heck of a lot more to innovate than the innovation teams of corporates in the industry. They will build and they will build with forward deployed engineers at assets implementing tech that does difficult things for difficult assets.
Source and stuff: Mining.com