How the U.S. Stock Model Works

An ensemble of CatBoost, GRU, and Transformer networks estimates next-session upside for U.S. equities.

Data Pipeline

Ensemble Architecture

Recommendation Flow

  1. Aggregate features and load the latest model artefacts.
  2. Score every eligible ticker with each model and compute the blended score.
  3. Filter for liquidity and volatility constraints, then surface the top 10 names along with projected upside.
  4. Track next-session highs to evaluate realized returns and recalibrate the ensemble.

Risk Considerations

The model is purely statistical and does not guarantee profits. Liquidity conditions, slippage, and unexpected news can materially impact results. Use these signals as one input within a broader investment process.