Work

Selected projects and case studies showcasing AI, ML, and data engineering work.

Value-at-Risk

PythonRGARCHHistorical SimulationQuantitative Finance

This study conducts a comprehensive comparison of eleven Value-at-Risk (VaR) models to evaluate their effectiveness in estimating market risk for 473 S&P 500 stocks from 1962–2023, encompassing over 3.2 million data points. The models tested include Rolling Normal (three windows), Historical Simulation (four windows), Normal GARCH (three windows), and an Optimized Student-t approach.

Read case study →