Elite technical talent voluntarily leaving high-status, high-paying jobs to work on something that still looks weird, unproven, and financially irrational.
The leading indicator is missionaries accepting lower status and higher risk; the lagging indicator is mercenaries arriving for higher status and lower perceived risk.
@stats_feed 2026 is expensive operationally, but not remotely comparable to Qatar on infrastructure because North America already has stadiums, airports, hotels, roads, and media infrastructure
History is full of examples where distributors traded at higher multiples than suppliers because they owned the customer relationship. The supplier may control production, but the distributor can control demand, pricing, and access. By Chanos's logic, cloud providers should never have become enormously valuable because they were "middlemen" buying servers from hardware vendors. Yet much of the value accrued to cloud platforms because they packaged infrastructure into a more usable service.
Also, growth rates matter. The market doesn't assign multiples based solely on where a company sits in the ecosystem. It assigns multiples based on expected future cash flows. If a neocloud is growing 200%+ while the supplier is growing 30–50%, a premium multiple is not irrational.
META owns some of the most valuable attention assets on earth, prints cash from ads, and is now layering AI across the entire ecosystem. If AI meaningfully improves monetization + WhatsApp/business messaging scales, the market is probably still underestimating earnings power. Reality Labs is basically a free call option.
@MarketMatrixs Great list. Wouldn’t touch $IREN with a 10ft pole tho - dilution is real and if you want exposure to the neocloud space, you’re better off allocating to $NBIS
Don’t see this being a material revenue driver for $RACE - $640k price tag is too high compared to their legacy vehicles.
Market won’t blink at EV performance given what we’ve come to expect from Tesla/BYD. This will be a limited run of vehicles where they end up partnering with a contract manufacturer or just licensing the brand.
Interesting article explaining how AI agents are becoming active players in prediction markets, executing trades using strategies that were previously available only to HF quant guys.
It begs the question - if AI agents dominate liquidity, do prediction markets still reflect human consensus or just algorithm efficiency?
https://t.co/DKtWTQbnZq
Interesting chart from Apollo.
Direct lending market is roughly $2 trillion, or about 3% of total debt outstanding for US households and businesses. Mortgages accounted for about 60% of the same at the peak of the housing bubble in 2006.
The LP perspective is different from the market commentary. The people writing the doom-and-gloom pieces (and even some GPs publicly talking their book about discipline) aren’t the same as the allocators actually deploying capital. LPs are looking at 150-250bps of illiquidity premium over BSL, floating rate exposure, senior secured positioning, and historical loss rates that still look favorable versus public credit.
@Restructuring__ Very few jobs will be done entirely AI. In this case, you’ll have 80% of customer service handled by AI, w/ difficult requests escalated to a human.
Jobs exposed to AI are repetitive, back-office tasks. Clients wont care who prepped their K-1 forms, so long as they get one.
@Restructuring__ Banks are still competing with investors on the broad majority of sponsor-backed deals. Private credit is still dominant, but it’s worth noting the field is full of players.
You don’t measure Citadel’s performance to the S&P because hedge funds are not correlated to the weighted index.
Most institutional LPs (insurance cos, public pensions, etc.) need 5-6% annual returns to cover their liabilities. They’re not paying Citadel to outperform Big Tech, they’re paying them to hedge their portfolios.
@BenHorney@FOS Not sure why DK would launch a separate app rather than adding prediction markers to their existing one. CAC is the largest expense for betting companies.