Rate moves challenging IPO liquidity could hit neocloud compute resellers harder than today's market implies. Neoclouds are running out of ammunition, and AI labs are running out of time and capital. If labs fail while hyperscalers have excess compute, today's shortage could become a glut. These shifts would not happen overnight. x.com  x.com  x.com 

Record-high positioning in hyperscalers and the Magnificent Seven could bring short-term weakness if the market trades against crowded holdings. The medium-term hyperscaler thesis still looks attractive, but the near-term positioning risk tempers it. x.com 

Hyperscalers' accelerating growth risks raising peak-growth concerns, making Alphabet the first to short. SouthernValue95 says hyperscalers are growing at their fastest rate since Covid while Anthropic is decelerating sharply. x.com 

Neocloud selection is crucial because debt-heavy buildouts, a few customers able to build their own capacity and fragile financing could leave most entrants working for lenders or unable to finish. A handful may become massive winners while many never bring capacity online. Survivors could benefit from years of demand outstripping supply, contracts with deep-pocketed customers and customers inherited from failed rivals. x.com 

Hyperscalers look positioned to capture value from other industries in the agentic era. Amazon, Microsoft, Meta and Alphabet are all well positioned and proactive, with a good chance they all extract value from other markets, so no loser among them is clear yet. Businesses relying on convenience or brand, or charging high fees or interest such as banks and card issuers, must trade margin for volume. Alphabet has the largest agentic distribution, from Search and Android to Gemini. x.com 

Alphabet's lagging AI lab undermines the case for owning it for ANT exposure. The semiconductor-business framing does not make up for that lag, and hyperscalers are no longer needed for this exposure. x.com 

Hyperscalers' forward P/E has contracted 20% even as estimates ramped, making it rare to see things get much worse. Rising GPU rents and new lows in availability make the AI fears a fade; hyperscalers should lead. The next two weeks will be gut-wrenching amid the OpenAI raise, Anthropic S-1, Fed, midterms and mutual-fund selling into year-end, but that is the time to buy. x.com 

Slower neoclouds such as CoreWeave face an uphill battle against SpaceX's faster data-center buildout. Demand may be effectively unlimited, but falling a year or two behind creates a $5-10B cash cost disadvantage they may not survive. Some could still pull off a miracle; SpaceX itself competes as a neocloud and has no reason to buy firms that cannot build data centers. x.com  x.com  x.com 

Microsoft, Amazon, Alphabet and Meta could be the big short if their AI capex is no longer funded from free cash flow but from dilution and bond and note sales. x.com 

Contracts, operating rental revenue and access to capital should let five neoclouds survive consolidation as debt tightens. IREN combines customer contracts with $14B of cash, committed GPU financing and prepayments; Nebius has over $40B of commitments and a new $3.5B Citi facility. Cipher has begun HPC rent and funded Stingray; TeraWulf has a long Anthropic lease and cash; Applied Digital has $36B of contracted base-term leases and 250MW operational. Musk warns many other neoclouds may fail. … Show more