- What is the risk associated with estimating losses in long tails, i.e. for very low (or high, depending on how we view the profit and loss distribution, i.e. Profit & Loss distribution) alpha levels in VaR and CVaR? Are VaR and CVaR the right measures to determine the likelihood of extreme events?
- What are the benefits and risks of adopting a parametric distribution for estimating tail losses?
- Does the fact that the historically observed correlation of rates of return in normal market conditions between two assets oscillated around zero give us confidence that by simulating VaR and CVaR from such a two-element portfolio we can assume a correlation of zero? Does the fact that the correlation has been stable over the last few decades mean that it will be stable in the future?
- The model presented assumes that the distribution of return rates is a normal distribution. Would you have more reservations about VaR (1%), to VaR (5%), to CVaR (5%) and CVaR (1%)? Explain why
Where did you get these questions?
And why, as a Level I candidate, are you asking them in the Level III forum?
