Agentic Winners 40
A bet on who captures AI's value — when it is genuinely unclear who wins. No clean public pure-plays exist; every name is a hybrid — a legacy business being disrupted, with agentic upside. Far less certain than the OPEX layer, so near-term performance may be soft by design. 40 names across 4 tiers, clean equal-weight. Performance plays no role — ABR is the gate.
Base 30 Dec 2025 = 1,000. Reconstituted from the AW25 momentum sensor into a clean equal-weight tradable index; pre-launch is a backtest of the current 40 constituents. Updated daily after global close.
Megacap Gateway
6The platforms agentic adoption flows through — Copilot, Gemini, AWS, on-device. They benefit largely regardless of who wins; gateway optionality, not pure-play conviction. NVDA is cross-index (Rubin); AAPL is a latent/laggard gateway, not a winner yet.
Control Plane
7The enterprise operating layer — ERP, CRM, HCM, workflow, orchestration. The Cannibalize-or-Die cohort: they must eat their own legacy revenue to win the agentic pivot. Mostly punished now; ABR splits the genuinely-pivoting from the legacy drag.
Application Leaders
16Vertical + horizontal software whose business outcome improves with AI — or gets eaten by it. Disrupted-may-emerge; the contrarian core. ABR separates the pivoting (Adobe, Intuit) from the roadkill (Chegg, Shutterstock — dropped as Terminal).
Endpoints
11The consumer + transaction frontier — commerce, payments, content, social. Where agentic action lands: shopping agents and payment agents need trusted endpoints to act through. The most consumer-weighted tier.
Two ways to read the book
Beyond the 4 tiers, AW40 cuts on segment and region. Enterprise (B2B): 27 · Consumer: 13. The consumer/transaction tier (Endpoints) is where agentic action lands — shopping agents and payment agents need trusted endpoints to act through.
Methodology — ABR is the gate, not performance
Winners holds names that may emerge but are punished today on AI-disruption fears — the bet is forward and contrarian, so trailing return is the wrong screen. The gate is ABR — the Agent Beneficiary Ratio, Closelook's 5-dimension AI-disruption scorer. Natural Position (benefits regardless) and Cannibalize-or-Die (incumbents genuinely pivoting) get in; Terminal names (agents replace them — e.g. Chegg, Shutterstock) are screened out. Every constituent carries its archetype so you see the legacy-vs-agentic transition, not a victory lap.
Clean equal-weight (1/N). ABR shown per name but informational — not a weighting input. Part of the three-index ecosystem: Rubin builds the factory · Agentic Infrastructure operates it · Agentic Winners use it. Read the thesis →
Closelook C+ Portfolio
$25 / month or $250 / year — founding rate
A research diary, not investment advice.