Course: RSM432: Risk Management for Financial Managers
In this assignment, my partner and I analyzed market risk (specifically, downside risk) in a global equity portfolio using real historical data. The assignment covered a range of techniques, from basic VaR calculations to more advanced methods like EWMA, GARCH(1,1), and copula-based models. We built these models from scratch in Excel and ran backtests to see how the models held up. One thing I took away from this was just how easily Value-at-Risk can break down in the tails, especially when markets move together. Modeling joint loss probabilities and stress-testing the portfolio over different holding periods gave me a much clearer picture of how risk actually behaves in practice. It was a technical assignment, but honestly one of the most rewarding ones I’ve done so far.
Key Concepts:
Excel VBA, Solver, and dynamic modelling
MLE estimation of volatility parameters (λ, α, β, ω)
EWMA and GARCH(1,1) volatility forecasting
Value-at-Risk (VaR) and Expected Shortfall (ES) calculations
Backtesting of 1-day VaR violations vs theoretical expectations
Copula functions to model tail dependence and joint probability of loss
Stress testing over 5-day, 21-day, and multi-month holding periods
Power law distribution fitting using MLE for fat-tailed asset return models