- Both VARs, when used for regulatory capital measurement, need to be validated against actual loss experience
- Both are built on data (market prices for Market VaR and operational loss data for Operational VaR) that is readily available
- Both are modeled based on a normal distribution
- Extreme Value Theory can be used to model extreme losses at the tail of the distribution for both Operational and Market VaR.
- I and IV
- I, II and III
- I, II and IV
- II, III and IV
I and IV are correct comparisons.
II is not a correct comparison. While market risk data is readily available, operational losses (especially extreme operational losses) data are relatively sparse and pose significant difficulty for operational VaR modeling.
III is not a correct comparison. Other statistical distributions also are in use for modeling VaR. E.g. an Operational VaR can be derived from convolution of a frequency distribution (e.g. Poisson distribution) and a severity distribution (e.g. lognormal distribution).