AMKR
watching
Conviction over time
Sentiment weighted by conviction, oldest to newest — above the line is bullish
Reasoning trail
Every post mentioning AMKR, newest first — the thesis as it was built and revised
-
29 Jul 26
Says he's interested in Amkor but only knows the basics and needs to research it further before forming a view.
-
28 Jul 26
Says he didn't have time to research Amkor today and asks if the earnings were really that bad.
-
28 Apr 26
Flags Amkor as starting to look interesting, in reply to a post about impulsively selling on account snapshots.
-
11 Apr 26
Agrees with a follower that $AMKR and $DRAM look very interesting.
-
10 Apr 26
With his portfolio 20% in cash, lists a set of stocks he plans to research over the weekend before deciding whether to deploy capital.
-
05 Apr 26
Thesis that companies solving AI infrastructure bottlenecks (memory/HBM, packaging, servers, networking, photonics, power/thermal) have pricing power to pass on war-driven energy inflation, making them attractive during market corrections; lists names across each category.
-
01 Apr 26
Suggests additional names to add to a supply-chain map: SK Hynix for memory, Vertiv for thermal management, Amkor for packaging, and Aehr Test Systems for testing.
-
14 Mar 26
Thesis post: major tech layoffs (AMZN, META, ORCL, XYZ, CRM, ASML) are being redirected into AI R&D and capex, not signs of weak fundamentals. He is bullish on companies that control the emerging AI infrastructure bottlenecks — memory/HBM, packaging/OSAT, server/rack integration, networking silicon, photonics/optical, and power/thermal/grid — naming SK Hynix, Samsung, MU, ASE Technology, AMKR, JCET, SMCI, DELL, HPE, Foxconn, AVGO, MRVL, CSCO, ANET, Ayar Labs, ALAB, CRDO, COHR, LITE, VRT, MOD, NVT, SU.PA, IREN, NBIS, CIFR.
-
08 Mar 26
Lists names he's researching more closely in the AI/data center supply chain space: Marvell, SK Hynix, Vertiv, Amkor, Applied Optoelectronics, Lumentum.
-
04 Mar 26
Shares a full article breaking down the AI chip supply chain into 4 phases and 13 layers, listing companies at each stage from raw materials to hyperscalers.