At his level of capabilities, tokenisation infrastructure (once fully operational and reliable) enables financialisation of absolutely everything, while regulation on tokenisation will inevitably be late by years and decades, due to inner complexities of infrastructure. If a banker can collateralise (via tokens) **everything** and make \~50x in credit funny money, while having little or no regulations (like crypto in early days), that is a business that gets all his attention. Just random examples. A gold mining company (a thing for the next 10 years) needs one or ten billion dollars to build a mine. What do they do today? Sell a stream or royalty, sell equity and warrants, or mostly they get a credit line. Slow, complex and holds them back. Now Mr Fink's minions can make a deal, tokenise the whole contained gold in the deposit (after the usual bankable feasibility study), and with a little bit of banking magic they can sell it today, get the money tomorrow, fund the company's mine build, and get some 0.01% commissions at every step. But 0.01% of a 2Moz gold deposit will be worth $1M soon (almost there), so any further trading in these tokens between holders can bring $1M every day, or every hour, whatever the tading volume might be. Then there are copper mines that are costly to build ($10\~20b), so a major pain to get funding for. And again, Mr Fink's magical tokeniser machine will turn it into tokens available for trading, and backed by real copper in the ground. This machinery unlocks illiquid high-quality assets with guaranteed demand, and once operational and reliable, it does it repetitively like a factory robot makes cars. Mr Fink can replace most of the private credit community that has been funding such mining projects, should he decides so. Next obvious market is real estate: fund the development by tokenising unbuilt buildings, sell it in open market to anyone interested, get his 0.01% on each transaction. In this one, Mr Fink will eat commercial bank's food, all of it, anywhere on the planet, not being limited to one country or something. Another thing is tokenised metals, all of them, from gold to minor metals that don't even have futures. Some XYZ Co may be sitting on several tons of iridium, and the only buyer is a refining company that will not be named here. A call to Mr Fink's minion, and bam! - iridium stash is tokenised, now freely traded by anyone interested, Mr Fink gets his 0.01% from each trade. In terms of financialisation power, this is like a klingon battleship vs blonde in bathtub - who wins is not a probability. Mr Fink wants to win, so he wants his klingon battleship ready asap, and not become the blonde in bathtub when someone else does it first.

---

Few questions. 1) why aren’t you using a CAPM for the cost of equity as it seems like some factors such as idiosyncratic risk and size premia are missing? Also the cost of debt to get the WACC. The 7.4% seems a bit low compared to the high on! growth / high assumed base case returns. 2)40% base case growth for on! products seems like an extreme bull case, no? 20% seems like a bull case even, given pouches were a big trend from ‘22–24. My co workers/ friends were big into Zyns during that timeframe but I’ve since seen them quit. 3)Younger Americans (18-30) would be more likely to convert to Juuls/nicotine pods than cigs in my opinion. I don’t see many people smoking cigs routinely, but I do commonly see people sneaking in a shameful Juul hit in public places. 4)Total debt is high, but coverage ratios are pretty goodish. What are the current maturities? Also what % of par are their bonds trading at? (You said this is a long term play but still a good indicator to see how much lower it could go). 5)Revenues have been declining since 2020 but net of excise tax they have barely increased? Was this due to a specific legislation? The current admin is beneficial to the tobacco industry but is a 14% excise tax a realistic projection? I do like the high div as that will incentivize long plays and gives the company an extra lever to pull incase on! doesn’t grow and the situation becomes distressed. This is well put together and very thorough. I really enjoyed reading this. Thank you.

---

What's the general thoughts on Fink around these parts? Buddy of mine just had dinner with him, it was a design meeting as he is an architect on the team designing Finks new vacation home in CO. Surprised to hear he was a super down to earth guy and generally nice during the meeting. Curious about others opinions on the guy.

---

stop buying shit co

---

Is NBIS a scam co

---

My job has a direct contract with Microsoft so we can’t use anything but co pilot. But yeah, it’s trash LMAO

---

Co worker here, none of your business

---

I’m concerned about my co worker. She doesn’t talk anymore, no smiles, or anything to anyone, even our boss. She’s now missed three company outings including the Christmas party. You may not like it, but that puts a stain on you from management perspective Is she trying to get fired?

---

The “compare launch MSRP vs launch rental rate” lens is useful, but it’s only part of the story. I'm looking at today’s economics. I.e. what people actually paid for H100/A100 hardware vs what they can rent it out for now . See my other comments where I did the math using today's rates. If older GPUs naturally drift cheaper as H200/Blackwell arrive that’s fine ONLY if the owner has largely paid down the original capex. The problem is a lot of this gear was bought in the last 1-3 years at peak pricing during the VC spending spree via these providers. When rental rates get pushed down toward or below full cycle breakeven while the loans are still outstanding, the age doesn’t save you. it just means you’re servicing yesterday’s gear / capex with today’s compressed cash. In that sense falling H100/A100 rental prices are a red flag. They show current supply/demand and ROI pressure on recent purchases, not just a normal generational price. And it’s not like most of this gear was bought for cash. Look at the public disclosures from the AI infra players we CAN see. CoreWeave has raised well over $10B in secured debt facilities for GPUs and data centers, and Oracle AI build out is explicitly tied to a debt load north of $100B. That’s before even getting into Nvidia‑linked financing and co‑signed deals. This isn’t a clean, pay as you go upgrade cycle. It's a lot of leverage that only works if rental economics hold up...which right now they don't seem to be at all. It looks more like scalp demand padding Nvidia’s books than durable, self funding infrastructure...and that can only run so long before the financing and ROI math collide and investors get wise to all of this.

---

At least magnum ice cream co is green

---