Feels like the current narrative is pretty simple
AI capex bad. RAM/HBM/power/data centers too expensive. No clean ROI on AI capex. Hyperscalers are just burning cash. Why bet on them when you can bet pick & shovel companies with high margins. I get the bear case. It’s not dumb.
If memory prices keep ripping, every AI build gets more expensive. Cloud margins can get hit. Devices can get more expensive. Eventually normal people get pissed because their Xbox & iPads now cost more.
But I’m wondering if this is becoming groupthink now and what can shift the narrative on hyperscalers.
Earnings are just round the corner. We will get numbers on q2 and forward guidance. If the capex guidance stays steady from increasing, that minor change alone could reverse the trade where we see hyperscalers go up and memory, chip & storage stocks go down. Seeing clear ROI in financials will also help.
One another blind spot for memory bulls: if RAM/HBM prices get too stupid, DOJ/FTC could start sniffing around. Not because high prices are illegal, but because DRAM has had price-fixing history before. If Apple/Microsoft/Dell/etc. start blaming memory costs for popular consumer devices inflation and consumers get mad, politicians may ask questions.
Betting on hyperscalers is the clear contrarian trade right now. Obviously being contrarian doesn’t always mean more upside. Anyone buying MSFT/GOOG/AMZN/META/ORCL here? What will change the narrative on hyperscalers?