↻ Refresh

Filters: composite ≥60 · setup ∈ {PULLBACK_50DMA, BELOW_200DMA, BETWEEN_MAS, BREAKOUT_52W} · R:R ≥1.5 · not in portfolio · setup backtested R:R ≥1.10×.

↻

+ ADD

No tickers yet. Type above and press + ADD.
Saves automatically — Firebase cloud sync available in ⚙ settings.

☁ Firebase sync active · Your watchlist saves to cloud automatically

↻

COMPARE →

↻ REFRESH

↻ REFRESH

FILTERS: ⭐ ONLY MY HOLDINGS Sort: Trade Date (newest) Sort: Trade Date (oldest) Sort: Disclosed (newest) Sort: Amount (largest) Sort: Amount (smallest) Sort: Ticker (A-Z) Sort: Politician (A-Z) Sort: Disclosure Delay (longest) 👑 PERFORMANCE LEADERBOARD ↻

↻ Refresh

Sourced from short_lockups.json. Maintained by Claude (manual updates from feed emails + periodic refresh). Data feeds: MarketBeat, Briefing.com, IPOScoop, SEC EDGAR S-1/S-4.

⚠️ Educational/research use only. Verify options availability + chain liquidity in your broker before trading. Not financial advice.

PREDICT →

TIMEFRAME: 1D 1H 15M 5M cone forecast scales to selected bar size

❓ HELP — How This Dashboard Works A walkthrough of every tab, every signal, and how to actually use it.

🎯 What This Is

A custom-built swing-trading dashboard that scans ~850 stocks every night, scores them across 10 factors (technical, fundamental, ML, sentiment, options flow, sector themes, etc.), and surfaces the best setups for 5-60 day holds. It's calibrated for a family $100k → $200-300k 5-year challenge — meaning capital preservation matters as much as returns.

It is not a day-trading platform. The system generates ideas; you execute on your broker.

📋 Every Tab, Explained

📖 Reading Guides

Crash Probability Levels (RISK tab + RISK BOX)

🟢 LOW (<8%): Normal market noise. Trade as usual.
🟡 MODERATE (8-18%): Raise stops. No new buys.
🟠 ELEVATED (18-32%): Trim winners. Consider hedge.
🔴 HIGH (32%+): Defensive posture. Cut exposure.

Setup Types (SETUPS + PREDICT)

🎯 PULLBACK_50DMA — stock testing 50-day MA support.
🔄 BELOW_200DMA — oversold quality stock below its long-term trend.
🛡️ BETWEEN_MAS — between the 50 and 200 SMA, testing 200 support.
📈 BREAKOUT_52W — at a 52-week high.
⏳ EXTENDED_50DMA / 🚨 VERY_EXTENDED — stretched above the 50-DMA (momentum).
🚫 PASS_BEARISH — composite below 45. Model says no.
Which setups have edge changes with the market — every card on SETUPS shows its current 2-year backtest (re-run weekly), and the tab only lists setups whose live edge clears the bar. No hardcoded favorites.

Where the Money Is Going (Pulse bottom)

Two ranked lists: 💰 MONEY FLOWING IN — sectors beating SPY over the last 5 days (buy setups in these first) · 💸 MONEY FLOWING OUT — sectors lagging SPY (avoid new buys here). Number = percentage points ahead/behind SPY; ↑/↓ = flow accelerating or fading. Everything in between is noise and hidden.

🌅 Suggested Daily Routine

6:30 AM (pre-market): Open RISK tab → check 30-day crash probability. Sets the day's risk posture.
9:00 AM (pre-open): Open SETUPS tab → review today's actionable trades. Click any name → opens in PREDICT for full analysis (cone, entry zone, stop, target, R:R).
During market hours: Glance at the PULSE heatmap (live prices) for unusual moves in your names. This is an end-of-day swing system — no need to watch it all day.
4:00 PM (close): Open MORNING tab → review portfolio P&L + near-stop alerts.
Weekly: Open RISK → realized leaderboard → see how flagged picks have performed since first flagged.

⚙️ Behind the Scenes

Every night, a Windows scheduled task scans 850 stocks. The scan computes a 10-factor composite score per ticker (technical, fundamental, macro, sentiment, ML, options flow, insider, political, peer momentum, theme score). A LightGBM machine learning model adds expected return for the 60-day horizon. The system writes JSON files to a Cloudflare Pages-hosted dashboard.

Every morning, Claude reviews the overnight scan, adds web research context, and writes the daily brief. Every hour during the day, Claude wakes up to refresh the macro market take.

Every Monday, the entry-plan classifier is re-backtested against 2 years of data so we know which setup types actually work. The cone calibration is also re-tested to keep the displayed probabilities honest.

Backtested portfolio replay (Jan-Apr 2026, 3-month window): +8.56% return, Sharpe 2.11, 61 trades. The model's edge is real, but past performance ≠ future results.

⚠️ Educational/research use only. Not financial advice. Past model performance does not guarantee future results. Always do your own research and consult a qualified financial advisor before making investment decisions.

↻ Refresh

Live SPY/QQQ/VIX strip refreshes every time you open this tab. Crash probabilities are calibrated to 10 years of SPY history and rebuild nightly (pre-market) — they lag intraday moves by design.

↻ NOW

📖 HOW TO READ THIS

The three big numbers below are crash probability scores (0-100) — NOT raw percentages of SPY dropping. They measure how much the current macro backdrop resembles historical pre-crash conditions across 8 leading indicators (VIX, positioning, yield curve, credit spreads, breadth).

• 0-34 LOW: Calm conditions. Trade your edge normally, full position sizing on conviction setups.
• 35-49 MILD: Watch daily. Keep standard stops, reduce size on marginal setups.
• 50-64 ELEVATED: Reduce new entries, tighten trail stops, raise cash to 25-30%.
• 65+ HIGH: Cut position sizes 50%, avoid new entries, consider SQQQ/VXX hedges.

Three horizons show WHEN the risk is most pressing. A 1-month score of 70 with 1-week score of 40 means "calm now, storm forming". The backtest table at the bottom shows how accurate these signals have been historically.

↻ REFRESH