Bullish Momentum
Cup & Handle Breakout
SPY4H
- Win rate
- 81%
- Avg move
- +1.9%
- Max DD
- -3.2%
Strategy Lab
Explore Live Strategies
Pick a setup, inspect the evidence, and refine it in your own strategy builder.
Setups with current momentum and fresh confirmation.
Continuation
Bull Flag Momentum
QQQ1D
- Win rate
- 77%
- Avg move
- +1.7%
- Max DD
- -3.5%
Mean Reversion
RSI Oversold Bounce
DIA1D
- Win rate
- 72%
- Avg move
- +1.2%
- Max DD
- -2.1%
Research Pipeline
Configure Signal Model and Portfolio Engine stages, then launch runs from a focused research setup.
Configure the pipeline visually, then refine domain controls below.
Configure
Signal Model
Next-5-bar Breakout
Predicts breakout probability over the configured lookahead window.
Decision cutoff: p_hat >= 0.50 · Feed: channel asc 1D/50
Configure
Portfolio Engine
Portfolio Engine
TP/SL search and holdout simulation.
1D execution · 5 max bars · 3 constraints
Run control
ReadySelected mode: Signal Model -> Portfolio Engine
Ready to submit.
Research Configuration
Tune the next run by domain instead of scrolling through a single mixed form.
Create new run
Highest-leverage iteration areas are the data window and timeframe, the Signal Model feature set and thresholds, and the Portfolio Engine feasibility and cost assumptions.
Data
Define the research window, feed attachment, and feature resolution.
Stage A data selectors
Attach at most one labelled feed before Signal Model (Stage A) training.
Attached feed: ascending channel (50d, valid).
Channel feed selected: use a selector-safe window to avoid fold starvation (train 12m / holdout 3m).
Objective
Pick the ranking profile used to score and compare candidates.
Gates
Control robustness gates and portfolio feasibility filters.
Features
Configure Signal Model search budget, labels, thresholds, and feature set.
Predictor features
7 selected
Gamma / ATR RatioATR Slope (14)ATR (14)Close Position (20)MACD Delta MARSI Delta MAWavelet Curvature
Portfolio Engine (Stage B)
Configure execution timeframe, TP/SL search, and the shared execution policy used by every A+B run.
Max holding bars
5
5 trading days on 1D candles.
Stage B contains everything needed for a complete A+B run. Choose 1D, 4H, or 1H here to control how execution is optimized.
Run Launch
Mode: Signal Model -> Portfolio Engine
Canvas + Backtest
Live symbol for canvas:SPY
SPY · 1D | SPY
CheckDetected 2026-06-03
Replay from publish date
Use Open Backtest inside the chart to validate setup.
Indicators
Keep evidence clear with overlays and momentum context.
EMA 24RSI 14Stoch 14/3/3
Backtest
Adjust inputs. These also feed backtesting.
1D2026-05-04 → 2026-06-03$100,000
Strategy
Adjust strategy assumptions and execution inputs.
Date range
Set the historical window used for validation.
Execution assumptions
Composite rulesAND
Rules apply to entry; exit occurs when rules flip.
Indicators
EMA
RSI
Stoch
No run yet
Submit backtesting to see OHLC candles, insights, and overlays.
Trade segments
Ready to run with current inputs
Broker connection
Account requiredChecking broker access...
Loading your account access before we enable broker connections on this page.