Financial modeling project for SPY (S&P 500 ETF) volatility estimation and prediction using classical and rough volatility models. Developed during CentraleSupélec academic program (June 2024).
This project implements advanced volatility modeling techniques to predict financial market volatility and develop trading strategies using SPY time series data (2000-2013).
Part I: Volatility Estimation
- Close-to-close volatility estimators
- Range-based estimators (Parkinson, Garman-Klass, Rogers-Satchell, Yang-Zhang)
- Statistical comparison of estimator performance
Part II: Rough Volatility Modeling
- Rough Fractional Stochastic Volatility (RFSV) implementation
- Fractional Ornstein-Uhlenbeck process for log-volatility
- Scaling properties and monofractality analysis
Part III: Trading Strategy
- Volatility-based trading signals
- Performance metrics (Sharpe Ratio, Drawdown)
- Backtesting on SPY and VXX data
- Python with SciPy for numerical analysis
- Time series modeling (ARCH/GARCH)
- Yahoo Finance data processing
- Statistical testing frameworks
Complete specifications are available in Volatility_prediction_Project-3.pdf.
Quantitative finance project combining mathematical modeling with practical trading applications.