Resolved calls: 2026 Q4

1 resolved. HIT 0. PARTIAL 1. MISS 0.

ticker	type	claim	verdict	resolved
	Structural shift	Through 2026-09-30, the binding bottleneck on AI compute deployment shifts away from GPU compute itself (NVDA/AMD/AVGO supply) toward four named adjacent constraints (memory: HBM3E/HBM4 capacity; networking: optical interconnect / scale-up fabric; power: chip-to-grid + behind-the-meter gas + rack power; cooling: liquid-cooling adoption), and equity outperformance follows, with these four pure-play constraint stacks meaningfully outperforming pure-GPU-exposure on a 6-month basis.	PARTIAL	2026-10-01