Reference Portfolio: Skin in the Game, Not Benchmark Tracking
Closelook operates ten reference portfolios — not benchmarks to track, but live positions with real capital behind them. Five are active-discretionary conviction baskets (AI Build-Out, Hypergrowth, Global Tech 50, Global ETFs, Derivatives), two are passive-systematic sleeves of the Rubin index (equal-weight, momentum) and three are active-systematic books — Rubin sector rotation, the Jensen NVIDIA-stakes cohort and the T1 Trillion-Dollar Club. The distinction matters: a benchmark is a target to match or beat. A reference portfolio is a demonstration of conviction — proof that the author acts on their own research. Every buy, sell, roll, and adjustment is published with reasoning. Every loss is documented. This is the "skin in the game" principle applied to investment research.
Why Reference, Not Benchmark
Most investment newsletters publish "model portfolios" that exist only on paper. There's no capital behind them, no emotional weight to the positions, no real consequence to being wrong. Closelook's portfolios are different: they represent actual allocation decisions backed by real capital.
This changes the quality of research. When you have money at risk, you question your own thesis harder. You size positions more carefully. You respect stops. The published portfolio is not a showcase of ideas — it's a record of decision-making under uncertainty.
The Five Active Discretionary Portfolios
AI Build-Out: Concentrated AI infrastructure positions. High conviction, high volatility. 15-25 positions.
Hypergrowth: Emerging AI infrastructure leaders with asymmetric upside. Higher risk.
Global Tech 50: 45-55 global technology leaders. The non-US opportunity at the widest discount in a decade.
Global ETFs: Macro allocation across geographies, commodities, and crypto. The diversification layer.
Derivatives: Pure options book — cash-secured puts, covered calls, directional bets. 80-95% cash at all times. The income and hedging layer.
The Two Passive Systematic Portfolios
Rules-based sleeves of the Rubin Build-Out 100 universe with no discretion — the contrast against the conviction baskets above is part of the research lens: same universe, different rules, observable performance gap.
Rubin Equal Weight: Quarterly rebalance to equal weights. Pure breadth — every name carries the same risk regardless of market cap.
Rubin Momentum: Equal-weight at the 2025-12-30 inception, then left to drift — no scheduled rebalance, so winners compound a heavier weight. First manual rebalance 2026-06-30.
The Three Active Systematic Portfolios
Rule-based construction driven by an active signal rather than a fixed schedule.
Rubin Sector Rotation: Event-driven — repositions when the Rubin sector composition flips between Memory, Power, Substrates, and Packaging dominance. Trigger-based, not calendar-based.
Jensen — NVIDIA Strategic Authority: Adds a public company the day NVIDIA discloses an equity stake or admits it to a named NVIDIA architecture layer; holds until NVIDIA's 13F shows the position gone. The signal is NVIDIA's own capital allocation.
T1 — Trillion Dollar Club: Equal-weight every stock that has ever closed at or above a $1 trillion market cap, added on its crossing date, base the last trading day of 2025. The signal is a one-trillion-dollar membership test.