The swing variables are (1) AI
capex persistence vs the company’s aggressive multi-year build, (2) next-node and advanced-packaging ramp quality, and (3) geopolitics/export-control policy that can impose a sudden risk
discount or constrain customers/tools. The bull case works if demand stays ahead of qualified supply; the bear case is mainly a load-factor/
multiple shock, not tech obsolescence.