Explore our latest posts on machine learning, market dynamics, strategy architecture and design
Many features are redundant or noisy. High dimensionality = harder to generalize.
Part 2: Infrastructure, Governance & Roadmap. What it takes to implement AI in asset management.
We are selective with deep learning. Expensive to train, easy to overfit, harder to debug.
If you discover a signal today, how long will it work?
Part 1: Use Cases & Value. Real-world use cases: idea generation, regime analysis, risk management.
Volatility ≠ returns: heavy tails, clustering, mean reversion. Dedicated volatility models are essential.
Explicitly model temporal dependence with transparent structure.
Compare GARCH, HAR, and ML models for volatility forecasting.
Financial machine learning rarely fails because the model is 'bad'. It fails because the market regime changed.
Gradient boosting dominates tabular ML. Learn how XGBoost and LightGBM deliver strong performance.
Most ML performance in finance doesn't come from the model — it comes from the features.
If you use financial data and your model does not use a rolling window, the backtest is wrong.
Sharpe ratio is dangerously incomplete for ML strategies.
Tree-based ensembles capture nonlinearities and interactions in market data.
Linear and regularized regressions still do serious work in finance.
Episodes where statistical properties are stable enough: high vol vs low vol, risk-on vs risk-off.
An alpha factory needs predictions for every asset at multiple horizons from multiple models.
Most backtests report a single Sharpe. But ML models fail by regime.
Loading...