Demo data
X

Find the market's
strongest setups.

StockPilot analyzes thousands of market signals and organizes them into actionable research for your chosen time horizon. No hype. Just data-grounded setups.

0 stocks scored 4 strategy horizons 1,284 tracked predictions
Choose your strategySwing horizon (2-4 weeks). Rankings weight 20-day momentum, relative strength vs. SPY, and volume acceleration most heavily.
Stocks scored
0
across 11 sectors
Bullish setups
0
outlook: bullish
Avg top-25 score
0
current strategy
SPY 5D trend
+1.8%
market bias: long

⚡ Top ranked setups

RankStockAI ScoreTrendOutlookRisk

Stock Explorer

Full universe with all four horizon scores side by side. Filter by sector, search by name, or set a minimum score floor.

🛰 Universe

StockSectorPriceChgDaySwing6-MoLongOutlook

AI Rankings

Per-horizon ranking with the key driver behind each score. Click any row to see why the model ranked it there.

🏆 Leaderboard

RankStockScoreOutlookKey driver

My Watchlist

Track the tickers you care about. Outlook and scores update with the model.

⭐ Watchlist

Your watchlist is empty. Add a ticker above to start tracking.
MODEL TRACK RECORD

Proof over promises.

StockPilot ranks stocks using systematic models. This page shows how those historical rankings performed when tested against forward market returns, SPY, and the broader stock universe.

— backtest periods 4 models 5D · 10D · 20D horizons SPY benchmark No lookahead
Top result
—
—
highest average forward return
Win rate—
vs SPY—
Swing average return
—
Average forward return of the top-ranked Swing selections over 20 trading days.
Swing win rate
—
Percentage of tested selections that finished positive.
Excess vs SPY
—
Average Swing return above SPY over the same forward period.
Excess vs universe
—
Average Swing return above the tested stock universe.

Performance across every model

See the complete 5D, 10D, and 20D backtest results at a glance.

DATABASE RESULTS · UPDATED WEEKLY
Day
Short-term ranking model
5D——
10D——
20D——
Swing
2–4 week research horizon
5D——
10D——
20D——
6-Month
Longer-horizon positioning model
5D——
10D——
20D——
Long Term
Highest-horizon ranking model
5D——
10D——
20D——

📊 Full backtest results

Every available model, horizon, benchmark return, and excess-return measurement.
ModelHorizonPeriodsWin rateAvg returnSPYExcess vs SPYUniverseExcess vs universe
Loading live backtest results…

Built to be tested, not trusted blindly.

Each historical test date uses the scores available on that date, selects the highest-ranked stocks, and measures what happened afterward. Future prices are not used to make the historical selection.

✓Historical scores determine the selections.
✓Forward 5D, 10D, and 20D returns are measured afterward.
✓SPY is measured over the same forward period.
✓The average-stock universe is also measured.
✓Results are stored in PostgreSQL and refreshed weekly.

One page. The whole proof.

This page reads directly from StockPilot's backtest API. When the weekly Railway backtest runs, the database updates and these results update with it.

Historical backtest performance is not a guarantee of future results. The 6-Month and Long Term models currently use the available 20-trading-day forward horizon in this backtest, so those figures should not be presented as six-month or one-year realized returns.

🕐 Individual prediction log

Live prediction logging is not enabled yet.
DateTickerHorizonScoreOutlookEntryStatus

Settings

Customize your default view and alert preferences. Connect your own yfinance server for live prices.

60
Run a small yfinance server at home and paste its URL here. When reachable, the app fetches live prices from it; otherwise demo data is used. See Settings docs for the server script.

Disclosure. StockPilot AI provides AI-powered market research and decision support, not personalized investment advice or money management. All scores are model-generated from technical indicators (momentum, trend, relative strength vs SPY, volume, volatility, setup) and may be wrong. Equities carry risk of loss. Past model performance is not indicative of future results. Review appropriate risk disclosures with counsel before acting on any signal.