This project investigates the effectiveness of dynamic volatility models for delta-hedging European call options within a Black-Scholes framework. It moves beyond the standard model's assumption of constant volatility by implementing and backtesting three different volatility forecasting methods: a 30-day Historical Volatility (HV), a symmetric GARCH(1,1) model, and an asymmetric GJR-GARCH(1,1) model.

The backtest, run on five major tech stocks (AAPL, GOOGL, MSFT, NVDA, TSLA), reveals that "no one size fits all". The findings demonstrate that a successful hedging strategy requires a volatility model that is correctly specified for the unique (and non-uniform) volatility structure of the underlying asset.