Allright, who is up for collecting Q&A for Level 2? I make a start:
(X) Liquidity Duration: shares owned / shares traded onaverage = 500,000 / 200,000 = 2.5
(X) Matching given securities with key rate duration exposure table: the 2Y instrument has exposures up to the 2Y KR, the 30Y instrument was the (only) one that had expsoure to the 30 KR and the remaining MBS (?) had expsoure to a couple of KRs but none to the 30Y KR
(x) Basel & Revision of the market risk requirements: the revision introduced stressed VAR.
(X) Basel III & Capital requirements: it was asked for the wrong option which was that a Tier 2 capital of 4% has to be maintained at all times
(X) Expected loss for a bond & CSD position with default probability for both: in case the bond
and the CDS seller default the expected loss is just EL = PD x LGD = 0.02 x 0.9 = 0.018 = 1.8%. In all other cases either the bond doesn't default or the CDS seller doesn't deault and makes the CDS buyer whole, i.e. there is no loss.
(X) Counterparty exposure with master netting agreement: there was a long position worht +$25mn, a long position worth -$35mn and a short position worth +$35mn.
(X) Difference between normal and lognormal VAR: mu = 10%, sigma = 40%, alpha = 5% and face value = 1mn GBP.
(X) Internal vs External Data in modeling operational losses: was about the bias that internal and external data tend to exhibit. not sure about this one.
(X) Most common operational loss types: cannot remember the asnwer options
(X) UL, EL and correlation: UL(P) > UL(1) + UL(2) where rho=0.02
(X) VAR backtesting & acceptable exceedances: a plot showed 4 exeedances over the last 250 days and 2-tailed tests were used. At which level of confidence would the correct model hypothesis be rejected. Given were 99%, 95%, 90% and 86% and hence 1.25, 6.25, 12.5 and 17.5 acceptable exceedances. Only at the 99% confidence intervall we would rejecte the hypothesis of a correct model
(X) Expected loss calcualtion at the 96.%: Given were the values at the 96.5%, 97%, 98% and the 99% confidence level.
(X) Interest rate swap & counterparty risk for the fixe-paying party: rising interest rate
(X) Total return swap & default: with a recovery rate of 30% the CLN buyer pays (ignoring accrued interest) 70% x face value
(X) Basel & Diversification: Which approach does explictly recognize diversification? Internal models approach (IMA)
(X) Liquidity Duration: shares owned / shares traded onaverage = 500,000 / 200,000 = 2.5
(X) Matching given securities with key rate duration exposure table: the 2Y instrument has exposures up to the 2Y KR, the 30Y instrument was the (only) one that had expsoure to the 30 KR and the remaining MBS (?) had expsoure to a couple of KRs but none to the 30Y KR
(x) Basel & Revision of the market risk requirements: the revision introduced stressed VAR.
(X) Basel III & Capital requirements: it was asked for the wrong option which was that a Tier 2 capital of 4% has to be maintained at all times
(X) Expected loss for a bond & CSD position with default probability for both: in case the bond
and the CDS seller default the expected loss is just EL = PD x LGD = 0.02 x 0.9 = 0.018 = 1.8%. In all other cases either the bond doesn't default or the CDS seller doesn't deault and makes the CDS buyer whole, i.e. there is no loss.
(X) Counterparty exposure with master netting agreement: there was a long position worht +$25mn, a long position worth -$35mn and a short position worth +$35mn.
(X) Difference between normal and lognormal VAR: mu = 10%, sigma = 40%, alpha = 5% and face value = 1mn GBP.
(X) Internal vs External Data in modeling operational losses: was about the bias that internal and external data tend to exhibit. not sure about this one.
(X) Most common operational loss types: cannot remember the asnwer options
(X) UL, EL and correlation: UL(P) > UL(1) + UL(2) where rho=0.02
(X) VAR backtesting & acceptable exceedances: a plot showed 4 exeedances over the last 250 days and 2-tailed tests were used. At which level of confidence would the correct model hypothesis be rejected. Given were 99%, 95%, 90% and 86% and hence 1.25, 6.25, 12.5 and 17.5 acceptable exceedances. Only at the 99% confidence intervall we would rejecte the hypothesis of a correct model
(X) Expected loss calcualtion at the 96.%: Given were the values at the 96.5%, 97%, 98% and the 99% confidence level.
(X) Interest rate swap & counterparty risk for the fixe-paying party: rising interest rate
(X) Total return swap & default: with a recovery rate of 30% the CLN buyer pays (ignoring accrued interest) 70% x face value
(X) Basel & Diversification: Which approach does explictly recognize diversification? Internal models approach (IMA)
