Thesis written for the BSc Mathematics with Statistics degree at the University of Bristol.
The thesis examines the statistical techniques used to forecast next day Value-at-Risk (VaR) and Expected Shortfall (ES), under different market conditions. The statistical techniques examined involve parametric distributions and time series methods for variance forecasting. These are tested in various international equity markets under stressed conditions (higher volatility) and normal conditions (lower volatility) and are evaluated using hypothesis testing methods on the backtested estimates.
For more information please review "Risk Theory thesis presentation.pdf" or "Risk Theory Thesis.pdf".