Live AI Model Portfolio · Fully Automated · No Emotion

Can AI
beat S&P 500?

We're finding out in real time — every decision logged, every position transparent.

investors visited this page
🏆
⚡ Alex Fund
since inception
VS
📈 S&P 500
same period
— Alpha
since inception
Active Positions
Rebalances
Open P&L
Closed P&L
See which positions Alex holds, would open, or would close — right now
NAV vs S&P 500 — Indexed to 100 at Inception
Alex Fund
S&P 500
Loading performance chart…
Performance
Fund Overview
Loading…
About
What Is Alex Fund?

ALEX FUND is a simulated rule-based portfolio generated by the Alex Finance AI engine. It selects high-conviction mispriced assets from the Asymmetry Scanner, applies region and risk caps, and rebalances automatically after every scan.

It is not a signal list. It is a persistent, auto-managed portfolio with entry and exit rules, benchmark comparison, risk controls, and a fully transparent decision log. Every position has a thesis. Every exit has a reason.

Educational and analytical model portfolio. Not investment advice.
Open Platform → View Methodology
Process
How It Works
01 · SCAN
🔍
Universe
807 assets scanned daily across AXI, EMS, CBS and regime fit.
02 · SELECT
Rank
Alex ranks candidates by structural asymmetry, mean reversion, and business quality.
03 · ALLOCATE
⚖️
Size
Equal-weighted with region and concentration controls. Max 15 positions.
04 · EXIT
🚪
Discipline
Closed when score deteriorates, profit target hit, or a stronger candidate displaces the position.
Why Alex Fund Exists
Most AI systems try to predict.
Alex tries to survive.
The objective is not maximum return. It is finding structurally mispriced assets while controlling fragility across market regimes — then getting out of the way when the edge disappears.
Asymmetry Engine
Every position is scored on structural dislocation from fair value — not momentum or narrative. If the upside isn't meaningfully larger than the downside, Alex doesn't enter.
AXI · EMS · CBS · APS
🌐
Regime Engine
The model reads the macro environment — Risk On, Risk Off, Stagflation, Recovery — and adjusts position sizing and sector exposure accordingly. Same asset, different regime, different decision.
4 Macro Regimes
🧠
Honesty Architecture
Every entry, exit and hold is logged with a reason. The backtest carries a confidence score — not a cherry-picked CAGR. If the model is uncertain, it says so.
Confidence Score · Full Audit Log
This is not a "ChatGPT portfolio." It is a rules-based engine with explicit logic, regime awareness, and the discipline to close a position when the thesis breaks — regardless of what the market is doing.
Live Portfolio
Current Positions
What the AI holds right now — and why every position was selected.
Loading positions…
Signal Engine
🎯 Next to Enter
Quality companies at meaningful discounts from the 52-week high — ranked by QaV Score (Quality × Value). Entry triggers when fundamentals hold and price stabilises near the floor.
⚠️ How entry actually works: These are QaV quality candidates (Source 5 of the engine). The fund's live scan (Source 1) runs first and may add a completely different ticker — as happened when Atlassian entered while other names were shown here. A candidate shown below only enters if the live scan produces no higher-priority pick on rebalance day.
Scanning entry signals…
Portfolio Architecture
📊 Structural Gaps
Factor and sector gaps identified by the portfolio diagnostics engine — candidates that would improve diversification and Sharpe ratio if added.
Loading watchlist…
Risk Analysis
🔬 Portfolio Stress Test
Beta-adjusted scenario simulation across bear, base, bull, and tail-risk regimes — plus structural analysis of concentration, correlation, and fragility signals across the fund's positions.
Running simulation…
Allocation
Portfolio Composition
Loading…
Transparency
Decision Log
Loading…
Trade Journal
Transactions
Every automated fund decision — chronological, with reason and result
Loading…
Learning Engine · v1
Reflection Report
What works, what doesn't — automated analysis of every closed position
Loading…
Learning Engine · v2
🧠 What Alex Learned
Pattern-based insights from every closed position — with proposed rule changes. Rules don't change automatically. These are suggestions for manual review.
Analysing closed positions…
Under the Hood
Risk & Methodology
The hard rules the AI follows. Constraints, not suggestions.
Risk Framework
Region Caps
US 70% · Europe 50% · Asia 35% · China 20% · Crypto 27%
Score Floor
Close when ALEX Score < 0.55 (absolute deterioration)
Deterioration Rule
Close if score falls >20% below entry score
Drawdown Exit
Standard: −25% hard stop · Extended: −32% when Alex score ≥ 0.65 (thesis intact — wider discount, not broken thesis)
Profit Take
Low-upside positions (implied upside <30% at entry) close at price target · High-vol / no-target positions close at +50%
Hedge Sleeve
5% GLD constant · 10% GLD + 8% SH in RISK_OFF
Entry Filter
Combined = Score×75% + Seasonal×25% ≥ 0.68 · Relaxed to ≥ 0.64 for seasonal divergence picks & crypto slot
Crypto Slot
1 of 15 positions always reserved for best crypto (SOL / BTC / ETH / AVAX) · Rotation on 1st of month with +0.05 buffer
Benchmark
S&P 500 (SPY) · alpha tracked continuously
Alex Fund is designed to survive regime changes, not simply chase high scores.
Scoring Methodology
AXI — Asymmetry Index (40%)
Measures structural upside potential relative to downside risk. Primary driver of position selection.
CBS — Business Strength Score (30%)
Quality filter. Screens out value traps and broken business models. Prevents high-asymmetry traps.
EMS — Mean Reversion Score (20%)
Quantifies dislocation from fair value. High EMS signals structural over-punishment — the core thesis.
Seasonality Overlay
10Y monthly win-rate history. Blocks entry in seasonal headwind months. Three divergence types: Type A (catch-up — strong month, stock lagging), Type B (reversal — bearish month overrealized, strong month ahead), Type C (oversold — MTD ≥15pp below seasonal avg, very strong next month). Classification range: HEADWIND → STRONG TAILWIND.
Macro Regime Overlay
5 regime states tracked with 8 FRED indicators. Modulates hedge allocation and entry aggressiveness.
ALEX = 0.40 × AXI + 0.10 × APS + 0.30 × CBS + 0.20 × EMS
Why Alex Buys
Position Thesis Videos
Every position has a thesis. Every thesis has data. Watch the reasoning behind each entry.
📺 View all thesis videos →
Historical Validation
Walk-Forward Backtest · 2017–2026
113 monthly periods · top 15 holdings · 0.25% transaction cost · equal-weight monthly rebalance
Loading backtest data…
Ask Alex AI
Click any position ticker to ask why it was selected — or ask me anything about the fund's logic, risk, or strategy.
Company Report ↗ Open in new tab