Reading List
Notes
Discovery candidate YSS. Score 75/100. Catalyst 6.00 pts, source 2.00 pts, risk 0.00 pts. Stock reacted 6.8879% since catalyst, but extracted value is still 10.7391% of market cap. Matched: million, acquisition.
Notes
Discovery candidate YSS. Score 75/100. Catalyst 4.00 pts, source 2.00 pts, risk 0.00 pts. Stock reacted 6.8879% since catalyst, but extracted value is still 10.7391% of market cap. Matched: acquisition.
Notes
Discovery candidate YSS. Score 75/100. Catalyst 4.00 pts, source 2.00 pts, risk 0.00 pts. Stock reacted 6.8879% since catalyst, but extracted value is still 10.7391% of market cap. Matched: acquisition.
Notes
Discovery candidate BW. Score 77/100. Catalyst 0.00 pts, source 2.00 pts, risk 5.00 pts. Lag still present: since the catalyst, price is 36.2712% over 100 days, near normal range 5.5583%/day. Risks: lawsuit.
Notes
Discovery candidate BW. Score 77/100. Catalyst 11.00 pts, source 2.00 pts, risk 0.00 pts. Lag still present: since the catalyst, price is 36.2712% over 100 days, near normal range 5.5583%/day. Matched: billion, approval.
Notes
Discovery candidate NET. Score 48/100. Catalyst 5.00 pts, source 2.00 pts, risk 0.00 pts. Catalyst appears repriced: price is 4.8773% since the catalyst over 17 days. Matched: zero trust.
Notes
Discovery candidate NET. Score 48/100. Catalyst 5.00 pts, source 2.00 pts, risk 0.00 pts. Catalyst appears repriced: price is 4.8773% since the catalyst over 17 days. Matched: zero trust.
Notes
Discovery candidate BAH. Score 57/100. Catalyst 5.00 pts, source 2.00 pts, risk 0.00 pts. Lag still present: since the catalyst, price is 2.8968% over 25 days, near normal range 1.5929%/day. Matched: zero trust.
Notes
Discovery candidate BAH. Score 57/100. Catalyst 4.00 pts, source 2.00 pts, risk 0.00 pts. Lag still present: since the catalyst, price is 2.8968% over 25 days, near normal range 1.5929%/day. Matched: acquires.
Notes
Discovery candidate IONQ. Score 75/100. Catalyst 5.00 pts, source 2.00 pts, risk 0.00 pts. Catalyst appears repriced: price is 86.4198% since the catalyst over 109 days. Matched: zero trust.
Research Docs
No documents yet.
Thesis
The Next Semis — Investment Thesis
Semiconductors were the defining infrastructure bet of the AI wave: whoever controlled compute controlled the bottleneck. That trade is maturing. The question now is: what is the next layer of physical infrastructure that AI and the broader digital economy will hit at scale?
The Pattern
Every technology supercycle runs through a constraint. In the 1990s it was networking equipment (Cisco). In the 2000s it was storage and servers. In the 2010s it was cloud hyperscalers. In the 2020s, chips.
The next constraint will be something that today looks adjacent and boring — the way ASML looked adjacent and boring before EUV became the only path to advanced nodes.
Candidates Under Surveillance
Semiconductor equipment — EUV and High-NA EUV are decade-long moats. ASML is not the trade; it is the template. Who is ASML for the next generation of fab constraints?
Power infrastructure — AI datacenters are approaching 1 GW per campus. The grid, transformers, and cooling systems are the new bottleneck. This is not a software problem.
Photonics and optical interconnects — As GPU clusters scale beyond a single rack, copper hits latency and power limits. Silicon photonics is 5–10 years from being the default interconnect inside datacenters.
Advanced packaging — CoWoS and HBM are already constrained. The companies that own advanced packaging capacity — TSMC, ASE, Amkor — are infrastructure, not just contract manufacturers.
Investment Criteria
A position graduates from watchlist to portfolio when:
- The physical constraint is independently verifiable (capacity data, lead times, capex announcements)
- The moat is structural, not cyclical — pricing power that survives a down-cycle
- Management has demonstrated capital discipline in at least one prior capex cycle
- Entry price implies a reasonable margin of safety on through-cycle earnings
This document is updated as conditions change. It is a working hypothesis, not a conclusion.