Member-written theses with a clear stance. Publish yours and build a track record.
Bears argue the DRAM cycle is topping. I disagree. Bit demand from AI training clusters is structurally different from the PC/mobile cycles of the past. Inventory days at the top three suppliers are still below the five-year average. That said, valuation already embeds a lot. I'd rather add on pullbacks than chase strength here. Watch capex discipline — the moment someone breaks ranks and floods capacity, the thesis changes.