EVT, EWMA (GARCH), and MC (note that this is only for parametric VaR and not applicable for historical VaR)
Use the results of the Var previously calculated (will be sent once project accepted); and continue the following:
1. Add EVT to all VaR estimates (in doing so you would need Nu/N where Nu is obtained by counting your histogram using the critical value i.e., 1.645σ • Estimate the two parameters (ξ and β) in EVT ( need long history of data, say 5 years)
2. Simulate a one-day distribution using the simulated factors (assuming Gaussian) and coefficients from the model (computing means and var-cov’s of the selected factors)
3. Simulate two days of prices using EWMA
The Geske model
Use the two-period Geske model (equation 14.4) and:
1. Collect one year daily stock prices for 10 companies; Compute daily returns, and then mean and standard deviation.
2. Collect (1) market cap, (2) current and (3) long-term liabilities of each firm.2
3. Run the model to solve for asset value and asset volatility (for each of the 10 firms).
4. Compare to the book value of each firm. 5. Report the one and two year default probabilities of each firm.