25 June 2023 |

A unicorn goes to mines


Climate tech has a new unicorn: KoBold Metals surpassed a $1B valuation after raising nearly $200M from T Rowe Price and other a-list investors, like Breakthrough Energy Ventures. 

On the ~surface~, this sounds like a straightforward story. KoBold Metals is a Bay Area-based tech company that builds AI and ML technology to aid in the discovery and exploration of critical minerals and metals. As written in this article about the raise:

KoBold analyzes vast amounts of geological data to pinpoint potential deposits of metals crucial for battery production, such as lithium, nickel, cobalt, and copper. Leveraging AI means the company can reduce its exploration time and costs.

We’ve also written plenty about the extent to which the global supply chain of metals and minerals will need to grow to accommodate the twin trends of electrification and energy transition. Whether it’s rare earth metals like neodymium or more ‘common metals’ like copper needed for countless types of electrical equipment and infrastructure, demand for these metals will expand significantly in coming decades:

Thanks to Alex Grant for turning me onto these graphs from the DOE

If the story ended there, I could tie it off with a neat bow, speaking to how much sense it makes that a now-unicorn tech company that accelerates discovery, exploration, and development of mines for some of the above-pictured critical inputs has investors burnishing bridles to ride it.

We’re not on Sand Hill Road anymore

There’s more to the story, though. Last year, KoBold decided to invest $150M to buy a controlling stake in undeveloped copper reserves in Zambia. The $200M they just raised will likely be used to develop those reserves. 

That’s a head-turning decision for a software company to make. Why take a stake in an actual mine? In a business that is the antithesis of developing software? 

I don’t know that much about developing copper reserves. Still, I can tell you it will tie up a lot of capital. I can tell you it is very risky. I can tell you it will take a lot of time and requires very different operational experience than running a software company does.

Plus, we’re talking about Zambia here. Which I also know little about. But I know it’s about as far away from KoBold Metals’ headquarters in Berkeley, California, as you can get.

A processing plant at the Kansanshi Zambia copper mine (not KoBold Metals’ mine) (Shutterstock)

There are plenty of ways this move could pay off. The Mingomba project, in which KoBold invested, has some of the world’s highest-grade undeveloped large copper deposits. KoBold’s technology should help to find optimal spots to drill and mine. They also noted they would use their tech stack to process actual drilling data.

I could also make the case that what KoBold Metals is after here is not a wholesale vertical integration from reserve discovery to reserve development. They may see this Zambian mine as the perfect proof-point for their tech to show how effectively it can help other developers identify undervalued reserves and profit on existing holdings. That’d make it a lot easier to sell software.

And they’ll have help with the drilling and the mining; the mine is a joint venture between KoBold and EMR Capital (private equity) and Zambia’s state-backed miner ZCCM-IH. Kurt House, the CEO of KoBold Metals, has publicly stated his company does not intend to be a mine operator

Still, regardless of how effective KoBold’s tech is in identifying ideal areas in which to drill and mine copper, they’re taking on a lot of risk by investing in this mine and by working with the mining company. Their stake could turn from pay dirt to dust for a whole host of reasons that someone with more experience covering mining could expound on at length. If that sounds like you, I hope you’ll respond to this to outline the bear case in greater ~depth~ for us! 

The net-net

The hardware vs. software divide is an oft-discussed divide in climate tech. Many venture investors prefer software-only businesses or businesses with a supporting software component alongside hardware. 

You thus don’t see that many software-first companies willingly invest full bore on hardware. Let alone, in KoBold’s case, invest full bore on some of the ‘hardest’ hardware applications, at least as measured by the risk of failure, CAPEX, and operational intensity. 

All that said, when I step out of my armchair analysis mode, I should thank KoBold Metals for taking the risk here. The world may need twice as much copper by 2040 to support all of the other climate and energy technologies we normally write about. 

I just as often discuss the need for more climate tech hardware businesses. So if a Silicon Valley-based company wants to get its hands dirty in Africa, and will get good reps at testing their tech in the real world in the process, maybe I should just close by saying, “Heck yeah.”