Intraday Strategy Using GARCH Volatility Modeling

A quantitative research implementation fitting rolling-window Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models on daily log returns to forecast 1-day-ahead volatility and guide systematic strategy execution.

Research & Methodology

• Volatility Forecasting: Employs GARCH(1,1) to estimate conditional variance, filtering out trades during regimes where forecast volatility exceeds risk thresholds.
• Signal Synergy: Combines the GARCH regime filter with momentum and mean-reversion triggers (RSI and Bollinger Bands).
• Performance Evaluation: Jupyter notebook featuring comprehensive backtest metrics including cumulative returns, maximum drawdown curves, and Sharpe ratio sensitivity visualizations.