$SNDK

Does he own it?

Multiple explicit first-person ownership signals: a Feb 7 portfolio breakdown lists Sandisk at 5% (within a 35% memory bucket); an Apr 30 earnings post states plainly that he owns Sandisk; and he repeatedly refers to it among names he personally holds and to its returns being underperformers "for me." Not just coverage.

His thesis

Sandisk is a core beneficiary of a structural NAND/memory "supercycle": AI-driven demand vastly outstrips supply, producers are taking multi-year prepayments and repeatedly hiking NAND prices well above analyst estimates, with no relief expected until 2028. He frames Sandisk as a pricing-power bottleneck over NAND output (via Kioxia), argues forward P/E stays cheap even after huge rallies, and reads blowout earnings and short-seller capitulation as confirmation. He treats it as the template "supercycle" winner to benchmark other plays against.

How his view evolved

Early on it is mainly a reference point for the memory thesis and a yardstick for smaller plays. As earnings repeatedly blow out and the stock rallies hard, his conviction hardens; he dismisses doomposters and shorts. Later it shifts to a benchmark/"meme stock" he holds while he rotates focus to photonics names as "the next Sandisk."

Key points

• Holds it; disclosed at roughly 5% of book within a 35% memory allocation
• Thesis: structural AI NAND shortage, repeated price hikes, multi-year demand, cheap forward earnings
• Used as the recurring 'supercycle' benchmark for other longs (photonics, etc.)

Derived from a holistic read of his public posts — paraphrased, never quoted. See his actual posts about $SNDK on X ↗