If elecricity is cheap, arc replaces met coal. For most of the world that's not true right now. It could be true someday, but with AI demand, maybe a long time from now
not really on topic, but related to electricity demand: we know that energy/ai is a big topic right now.
on the one hand, we have three mile island re-opening a nuke to sell exclusively to microsoft. and there are discussions of siting data centers in e.g. the permian to use cheap nat gas and local water for cooling. [e.g. see the write up of landbridge at the specialsituationinvesting substack]
but otoh we have many discussions of ai/big data not producing any value except in niche specific applications.
is ai a bubble, comparable to the dot com bubble of the late '90s? the internet went on to be a huge economic component, but not before the nasdaq managed to drop 80% could ai be on a similar trajectory, crashing in the near to intermediate term?
if so, what does it imply about electricity build-out/demand over the near to intermediate term?
We almost certainly will have a bubble in AI stocks where the price crashes down from a peak. Some of the best tech experts have said that they don't know if that is at NVIDIA $3T or $5T.
But at the same time, AI is already replacing thousands of call center jobs just at the beginning, so we probably have a full ten to twenty years of jobs getting outsourced to compute. The effect on the economy will be huge. Historians will say it was fast, but living through it will still feel slow.
Electricity buildout could also get overdone, especially if AI can code itself to use less electricity, which I have heard is some low hanging fruit, as machine learning is good at writing software code. If we overbuild electricity, like everything else, low marginal cost producers will survive, unless regulations and tax credits upset the pricing. I think this would benefit nuclear now that the carbon cult is on board, and would eventually pose a challenge to natural gas but probably not for a decade, and who knows if by then the carbon dioxide hysteria is even still a thing. Five percent of global deserts have greened due to all this wonderful carbon dioxide that humanity is putting back into the atmosphere. Without leftist narrative control, it is unclear if the institutional control over carbon tax credits and pricing will be a problem when we have an electricity capacity glut.
Is TMI exclusive to MSFDT? Or is that just an accounting trick (with data centers all over the east)? I think this deal is very different from the Talen/AWS deal where the data center is off-grid.
I only bought METC. I know Ben at The ROI Channel loves royalty streams. I haven't quite jumped on that bandwagon, but I would put it in an income account.
Lovely piece as always prof, coal and royalties are both near to my heart.
My uncle Rick is avoiding small cap coal names due to their inability to access finance/capital should something unfortunate befall them. Does that concern you if you're taking a stake in the operating entity v the royalty co?
That's a good risk to worry about. They have almost no debt, and the rare earth asset to borrow against, so if credit was cut off from coal, they could try and be creative on the critical minerals side. In theory the rare earth royalties and intellectual property held in METCB would be safer from a total loss on the coal side.
very interesting write up, thank you. a couple of questions:
1. if the rare earth play comes through, does metcb participate somehow?
2. what's your best guess on when/whether arc furnaces or hydrogen displace demand for met coal?
Yes, METCB has rare earth royalties.
If elecricity is cheap, arc replaces met coal. For most of the world that's not true right now. It could be true someday, but with AI demand, maybe a long time from now
not really on topic, but related to electricity demand: we know that energy/ai is a big topic right now.
on the one hand, we have three mile island re-opening a nuke to sell exclusively to microsoft. and there are discussions of siting data centers in e.g. the permian to use cheap nat gas and local water for cooling. [e.g. see the write up of landbridge at the specialsituationinvesting substack]
but otoh we have many discussions of ai/big data not producing any value except in niche specific applications.
is ai a bubble, comparable to the dot com bubble of the late '90s? the internet went on to be a huge economic component, but not before the nasdaq managed to drop 80% could ai be on a similar trajectory, crashing in the near to intermediate term?
if so, what does it imply about electricity build-out/demand over the near to intermediate term?
We almost certainly will have a bubble in AI stocks where the price crashes down from a peak. Some of the best tech experts have said that they don't know if that is at NVIDIA $3T or $5T.
But at the same time, AI is already replacing thousands of call center jobs just at the beginning, so we probably have a full ten to twenty years of jobs getting outsourced to compute. The effect on the economy will be huge. Historians will say it was fast, but living through it will still feel slow.
Electricity buildout could also get overdone, especially if AI can code itself to use less electricity, which I have heard is some low hanging fruit, as machine learning is good at writing software code. If we overbuild electricity, like everything else, low marginal cost producers will survive, unless regulations and tax credits upset the pricing. I think this would benefit nuclear now that the carbon cult is on board, and would eventually pose a challenge to natural gas but probably not for a decade, and who knows if by then the carbon dioxide hysteria is even still a thing. Five percent of global deserts have greened due to all this wonderful carbon dioxide that humanity is putting back into the atmosphere. Without leftist narrative control, it is unclear if the institutional control over carbon tax credits and pricing will be a problem when we have an electricity capacity glut.
Is TMI exclusive to MSFDT? Or is that just an accounting trick (with data centers all over the east)? I think this deal is very different from the Talen/AWS deal where the data center is off-grid.
i don't know but my assumption is that it's an accounting trick since there was no mention of co-locating a data center.
Arc furnaces require scrap steel, i.e., an OECD economy.
not sure of the economics, but found this:
Direct Reduced Iron (DRI): EAFs can use DRI, which is produced from iron ore using natural gas. The DRI is then melted in the EAF to produce steel.
Did you buy both METC and METCB or just METC?
I only bought METC. I know Ben at The ROI Channel loves royalty streams. I haven't quite jumped on that bandwagon, but I would put it in an income account.
Ok. Thanks. I was looking at METC and wondering where to raise the money to buy. Cloned your idea of trimming my other coal stocks.
Anyone else reading the 2 letters to shareholders on Ramaco’s site about their potential REE mine?
Lovely piece as always prof, coal and royalties are both near to my heart.
My uncle Rick is avoiding small cap coal names due to their inability to access finance/capital should something unfortunate befall them. Does that concern you if you're taking a stake in the operating entity v the royalty co?
That's a good risk to worry about. They have almost no debt, and the rare earth asset to borrow against, so if credit was cut off from coal, they could try and be creative on the critical minerals side. In theory the rare earth royalties and intellectual property held in METCB would be safer from a total loss on the coal side.
Most of METC’s production increase is low-vol. Does this bother you?
It's my understanding that low-vol is the expensive stuff. So that would make it a good thing.
Yes, your understanding is correct!