Volatility · Intermediate · Daily
Intraday ORB
How it works, and when it doesn't.
Strategies 1-10 have correlations ranging from -0.3 to 0.7. The ensemble achieves ~30% lower variance than the average individual strategy while maintaining similar expected returns.
Gap 0.91% but RF signal conflicting (46%). Cash.
Why it works
Combining signals from strategies 1-10 via Sharpe-weighted voting reduces individual strategy noise. When multiple uncorrelated strategies agree, the ensemble signal is more reliable than any single strategy. Disagreement is informative — persistent disagreement signals Watch.
The signal logic
- Collect signals from strategies 1-10
- Weight each vote by strategy's rolling 63-day Sharpe ratio
- Long: weighted majority (>60%) signal Long
- Short: weighted majority (>60%) signal Short
- Watch: 40-60% split (no clear consensus)
When it fails
In rapid reversals, the ensemble is slower to react than individual strategies. Consensus formation takes time.
- Best market
- Markets where multiple strategies align — broad trend confirmation
- Worst market
- Regime transition periods where strategies systematically disagree
Backtest
| Sharpe ratio | 0.12 |
| Annual return | 0.7% |
| Worst drawdown | -30.0% |
| Trades in the backtest | 138 |
Backtest on daily bars, positions taken the bar after the signal (no look-ahead).
Intraday ORB, live.
Re-scored after every close.
The platform runs this and 32 other strategies against the live tape each day — with the evidence behind every number.