Question 1

Which of the following is true in relation to the application of Extreme Value Theory when applied to

operational risk measurement?

I. EVT focuses on extreme losses that are generally not covered by standard distribution assumptions

II. EVT considers the distribution of losses in the tails

III. The Peaks-over-thresholds (POT) and the generalized Pareto distributions are used to model extreme value

distributions

IV. EVT is concerned with average losses beyond a given level of confidence

Options :

A : I and IV

B : II and III

C : I, II and III

D : I, II and IV

Answer: C

Question 2

Under the KMV Moody's approach to calculating expectingdefault frequencies (EDF), firms' default on obligations is likely when:

Options :

A : expected asset values one year hence are below total liabilities

B : asset values reach a level below short term debt

C : asset values reach a level below totalliabilities

D : asset values reach a level between short term debt and total liabilities

Answer: D

Question 3

The VaR of a portfolio at the 99% confidence level is $250,000 when mean return is assumed to be zero. If the assumption of zero returns is changed to an assumption of returns of $10,000, what is the revised VaR?

Options :

A : 260000

B : 240000

C : 273260

D : 226740

Answer: B

Question 4

Which of the following decisions need to be made as part of laying down a system for calculating VaR:

I. The confidence level and horizon

II. Whether portfolio valuation is based upon a delta-gamma approximation or a full revaluation

III. Whether the VaR is to be disclosed in the quarterly financial statements

IV. Whether a 10 day VaR will be calculated based on 10-day return periods, or for 1-day and scaled to 10

days

Options :

A : I and III

B : II and IV

C : I, II and IV

D : All of the above

Answer: C

Question 5

Which of the following is true for the actuarial approach to credit risk modeling (CreditRisk+):

Options :

A : Default correlations between obligors are accounted for using a multivariate normal model

B : The number ofdefaults is modeled using a binomial distribution where the number of defaults are considered discrete events

C : The approach considers only default risk, and ignores the risk to portfolio value from credit downgrades

D : The approach is based upon historical rating transition matrices

Answer: C