AUTONOMOUS FUND — VOL. IV · 2026

Predicting
the agent
economy.

An autonomous Claude-powered agent scores 50 stocks every day, sizes positions by quadratic conviction, and executes monthly rebalances through Autopilot.

Long the picks and shovels. Short the displaced. No human in the loop.

Fund return
+30.4%
Since 2026-04-13
▲ 16.4% vs S&P 500
§ Latest insight · 2026-07-13

Our system worked best when it held stocks with balanced component scores that strengthened over time — PANW (+17%) and DDOG (+12%) are textbook cases where the ensemble upgraded conviction after mediocre starts. It failed hardest on CRWD (-73%), where a high statistical win-probability masked a weak thesis fit, and on MRVL (-16%), bought at a 52-week high with poor risk quality. The one that got away was ANET (+15%): we skipped it despite the model quietly holding an 80+ score for over a week, because the initial thesis read was too cautious on optical-interconnect leadership. The system-level lesson is that we should trust sustained score improvement more than initial component reads, and we should treat 'stock at 52-week high with low risk_quality' as an automatic size-down rather than a full-position hold. Simply put: the ensemble often knows before we act — we need to let it override stale first impressions, and we need harder ceilings on positions that look extended.

See all learnings →
§ 01 — Thesis

Long the shovels.
Short the displaced.

The short thesis is what separates this from every generic AI ETF.

LONG · 36

The Winners

Agentic AI multiplies compute demand per task. GPUs, cloud, cybersecurity, power — the picks and shovels.

AI Chips Cloud Cybersecurity Energy Tooling Robotics
SHORT · 14

The Losers

When agents can research, write, analyze, execute — industries built on human knowledge work face existential pressure.

Education Legacy SaaS Prof. Services Marketing Finance
§ 02 — Process

How the
fund works.

From raw news to portfolio positions — entirely autonomous. Four stages, executed every month.

01 / 04

Ingest

4+ news sources, SEC filings, and market data filtered for agentic AI signals.

02 / 04

Score

Claude scores every stock 0-100 on conviction — bull for benefit, bear for disruption.

03 / 04

Size

Quadratic conviction curve — higher scores earn disproportionately larger positions.

04 / 04

Execute

Monthly rebalance through Autopilot. Daily monitoring for 20+ point score swings.

§ 03 — Signal

Conviction scores.

All scores →
§ 04 — Activity

Portfolio changes.

Full log →
Fund inception — 16 positions established
Initial portfolio constructed from conviction scoring. All positions are long-only via Autopilot.
2026-04-13
Next rebalance scheduled
2026-05-01
Autopilot

This fund is live on Autopilot.

Every trade executes through our official brokerage partner. Follow to mirror real positions with real capital.

Follow on Autopilot →
50
Stocks monitored
16
Active positions
30.4%
Fund return
24/7
Autonomous
Supervisor: ATTENTION — the scoring pipeline has a non-blocking warning.