MARKET INTELLIGENCE SYSTEM
MULTI-FACTOR · TECHNICAL · FUNDAMENTAL · SECTOR · SOCIAL · ODDS · OPTIONS · BUY-POINT · DISTRIBUTION DAYS · HISTORICAL ANALOG
initialising… auto 10m
Worker URL ● checking… · key no key — using Yahoo relays (often slow/fail) get free key →
Settings backup
TradingView Premium does not provide a data API for this page. A free Twelve Data key (800 calls/day) loads stocks reliably. Sign up at twelvedata.com → copy key → paste above → Save → Refresh.
Market Posture distribution days + O'Neil follow-through · S&P · Nasdaq · small caps · BTC · ETH · SOL
  reading the tape…
Alerts buy: Stage 2 breakout · sell: trend-exit breakdown (tuned) · watchlist
sell exit: entry mode:
Phone alerts: ntfy topic or webhook URL phone (SMS webhook only) free push: install the ntfy app, subscribe to a private topic, paste its URL
Circuit breakers: max open · daily loss $ · stale > d
Alert Track Record $1,000 paper trade per BUY→SELL · accuracy & P&L · PAPER
SymbolStatusEntryExit / nowReturnP&L ($1k)MFE / MAEReason
no tracked trades yet — a $1,000 paper position opens on each BUY alert and closes on the matching SELL alert
Journal Insights — how to improve patterns in your tracked trades · LEARNING LOOP
Signal Log every buy/sell signal captured & graded by what price did next · SAMPLE
DateSymbolSignalPriceAgree/CompModeFwd 21dVerdict
no signals yet — every alert is captured here automatically, independent of paper round trips
Captures every signal — including sell-only events with no matching buy — so nothing falls through. Click score past signals after ~2+ weeks to grade older ones: a BUY is "correct" if price rose over the next ~21 sessions, a SELL "correct" if price fell (you avoided a drop). Export CSV regularly — this is your growing sample, and it survives across file versions. Not investment advice.
Near-Miss Log strong candidates (≥70, 5–6/7) that DIDN'T fire — did they run anyway? · GATE TEST
DateSymbolScoreMissingPriceFwd 1wkRan?
no near-misses yet — captured automatically when a ≥70 composite / 5–6-of-7 name doesn't trigger a buy
This measures the cost of the entry gate: strong names one or two conditions short of a buy, checked a week later. A high "ran +10%" rate suggests the gate is too strict; a low rate means it's filtering well. Honest caveat: a near-miss that ran isn't proof you should\u2019ve bought — loosening the gate to catch it would also admit more losers. Judge the group, not the one that got away. Export the CSV to keep the sample across versions. Not investment advice.
Suggested Changes algorithmic · re-derived every refresh
deriving…
Composite Weights drag to re-weight · totals normalised · LIVE
weights sum to 100 · applied
Industry Rotation — Acceleration ranked by improving momentum (1-mo → 1-wk → 3-day) · leaders flagged near buy points
Industry1-mo1-wk3-dayAccelExt vs 50dLeaders (● = near buy)
press "rank industries" — scores each group's momentum, its extension above the 50-day (in ATR units), and surfaces each group's leading stocks (highlighted when near a buy point)
Watchlist — Multi-Factor click a row for detail · composite 0–100 · TV = open chart on TradingView
research / add any ticker:
#SymbolRatingCompositeBuy pointRSIStage TechFundSectorSocialOptionsOdds3-mo
scoring watchlist…
Risk & Position Sizing survival first · size the trade from your stop · portfolio risk cap
Live Screener rolling scan · ~150 stocks & ETFs · setup score COMPUTED
80
warming up…
#SymbolRatingSectorScoreBuy point200-dayRS vs S&P1-mo
warming up the scanner…
Closest Historical Analog nearest-neighbour regime match on S&P 500 · BACKTEST
Best match
scanning history…
What happened next
Trend Template Backtest Minervini TT entry · 2-year lookback · ON DEMAND
SymbolSignals1-mo win%1-mo avg3-mo win%3-mo avgIn-trend 1-mo win%
press "run backtest" — fetches 2y of daily bars per symbol and evaluates the Trend Template historically
Per-Stock Backtest & Tune buy = Stage 2 entry · sell = best exit MA · 2-year round trips · ON DEMAND
SymbolBest exitIS tradesIS win%IS expOOS tradesOOS win%OOS exp
press "backtest & tune" — for each watchlist name it round-trips the Stage-2 buy against three exit rules (21/50/200-day) over 2 years, picks the best, and wires that exit into the sell alert
Sector-Tuned Exit pools all ~150 screener names by sector · walk-forward · ON DEMAND
SectorBest exitIS tradesIS win%IS expOOS tradesOOS win%OOS exp
press "backtest & tune by sector" — pools all ~150 screener names' round trips by sector, picks the best exit on in-sample data, and validates it out-of-sample
Strategy Metrics & Monte Carlo risk-adjusted quality across all trades · sequencing-luck simulation · ON DEMAND
ExitTradesWin%PayoffExpectancyProfit factorMax L-streakSharpe*Max DD ($)RecoveryTotal ($/1k)
press "run" — pools every trade across the ~150-name universe for each exit rule and computes full risk-adjusted metrics, then bootstraps 5,000 resampled sequences to show how much the result depends on luck
Monte Carlo exit: slippage %/round-trip ($0 commission)
Sector-strength impact does entering in a strong sector actually help? · same pooled trades, split by sector RS at entry
ExitBucketTradesWin%PayoffExpectancyProfit factorSharpeTotal $
runs automatically after "run" above — splits the pooled trades into those entered when the stock's sector was outperforming SPY (strong) vs underperforming (weak), point-in-time
QQQ Timing vs Buy & Hold — 10-year test does trend-timing the index beat holding it? · ON DEMAND
StrategyTotal returnCAGRMax drawdownTime in marketTrades
press "run" — applies each moving-average trend rule (in when price > MA, in cash when below) to ~10 years of QQQ, with round-trip friction, and compares to simply holding QQQ
Not investment advice. This is an educational, algorithmic tool. Scores, suggestions, buy-point reads and historical analogs are heuristics computed from delayed public data, not recommendations. Live feeds are best-effort through free public relays and frequently fall back to modelled or static values — always confirm against your broker and against TradingView charts before acting. Past analogs and backtests do not predict the future. Use TradingView Premium for precise volume-profile, multi-timeframe, and real-time confirmation.