Nov 2013 FRM Part 2 Review

[abh2013]

So, the Nov 13 FRM part 2 is over across the globe! Please share your experience. In my opinion, it was very theoretical test which tested on very fine details in theory . What do you think?

 

[Hope]

I think so, very theoretical test. Actually i prefer calculation like part 1 because my memories is not good.
Btw please share what do you remember in part 2?

 

[MJ2013]

Mostly theory around 80 %
1 historical shortfall var for different confidence levels given
2.solvency II ..correlation among risk factors...not in basel
3 4 ques caselet on basel III ..trick was not to include tier iii capital for calculations
4.many questions on CDS
5.Economic capital ..which would increase EC
6.Asset allocation problem
7.netting factor calculation
8.which was a confirmed fact about BMIS..didnt charge commission/management..performance fees
9.two ques from flash crash..one on results of the study
10.relation between financial and sovereign crisis
11.two ques caselet on risk management framework..

 

[Jugoncete]

Very theoretical exam:
My annotations( I missed the bassel III tip (dropping Tier III, result was 8,4%):

12-Which of the following options would retrieve the largest LVar(constat spread) to Var ratio: smallest confidence level and holding period
13 - Regression: Beta when positive returns and negative. Asimmetry.
14. Impact on ES on economic capital
15. Which of the following would increase ARAROC the most: scenarios based on changes on operating cost, revenus, were given(guess it's Q number 5 above).
16: Liquitidy cost calculation: Assuming non-constant spread.
17: A set of options contracts were given plus a forward contract: One of the options deep in the money the other out of the money, derive the VaR, based on delta?.
18: Senior tranche: Is the one that has the highest duration of the tranches.
19: IO(negative duration).Not 100% sure about this one.
20: ZC Bond: yield=8% RR=0% Riskfree rate=3%. Derive the PD. Two correct options were given, 5% and 4,63%. I chose the latter(more accurate in my opinion).

 

[maheshs]

My additions:
18. Calculation of PD for an individual bond given a basket
19. Prepayment principal calculation for a MBS bond
20. Calculation of debt using merton model
21. Computing strike of barrier option given spot, barrier rate, RF, and other inputs
22. Calculating rate using simulated approach using model 1 (no drift) and dw was provided
23. Calculation of Var and LogNormal VaR and their difference
24. Calculating Implied PD where Recovery was 0

 

[MJ2013]

Jugoncete..sry your version of q5 is correct.answer was B
Q 20 ans was 4.63

 

[Jugoncete]

Regarding question 20:

You can obtain it two ways: PD=spread/ LGD --> 5% or (1+Rf)=PD*RR(1+r+s)+(1-PD)*(1+r+s)-->4,63%

Question 21, I came across the solution this way: All digital options given summed and european call and a european put. After that use put call parity to obtain the strike, solution was X=40

 

[Hope]

25. Surplus risk? Initial asset 100, liability 90, return asset 6% liability 7%. Volatility asset 10%, vol liability 5% correlation 0.8.

26. Distribution operational risk about severity.

27. Risk neutral in short term rate.

28. About MBS principal only

29. External credit enhancement

 

[abh2013]

Overall it was highly theoretical test in which engineers like me don't perform well (who love to play with numbers). Fingers crossed though . Any guesses how much % score could be the passing score?

 

[Hope]

30. 95% CVAR? ---> 110-101 = 9 (dont remember exactly)
31. Largest Vega? at the money with long maturity or short maturity (dont remember exactly)

 

[wenqc]

Most, if not all, questions require understanding of concepts, rather than memorizing calculation formulas. I was worried they might test the GPD/EVT formula; but thank goodness they didn't. The coverage was pretty well given the huge amount of materials in the L2 curriculum. I suspect they might move to a 3-level exam format soon given the explosion of risk-related knowledge in recent years reflected in the jammed L2 reading materials. Hope we all pass this time; so that we don't have to deal with FRM level 3.

 

[wenqc]

One question is straight from the practice set (normal vs. lognormal VAR)

 

[SoBeFla]

Very helpful...thx everyone. Another q I recall - two counterparties to an OTC swap have a CVA agreement. Both C/Ps have their ratings downgraded. We were asked to pick which C/P would be able to charge more/pay more following the downgrades. I think the bank/dealer rather than the end user can request more CVA compensation b/c their rating is still higher by a notch. Any other ideas! Thanks!

 

[Jugoncete]

@SoBeFla My choice was different, as you said both counterparties had their credit spreads increased in the market. However one of them had only a 30bp increase vs 130 bp. Therefore the first would ask for a CVA reduction.

 

[GiriFRM]

what was the answer for the last question on EVT.... was it b or c ?

 

[GiriFRM]

surplus var answer.... is it 11.34 or by considering the duration parameter ...is it different ???

 

[abh2013]

I thnk surplus question anwswer was some 1.7%. Sometng like 11%-9% .......though dont remember exactly ...

 

[abh2013]

@SoBeFla My choice was different, as you said both counterparties had their credit spreads increased in the market. However one of them had only a 30bp increase vs 130 bp. Therefore the first would ask for a CVA reduction.

I used the same reasoning too i. e. compared delta chages for both the parties

 

[Hope]

I also got 11 something for var surplus.

Btw Do you still remember the question related Binomial tree (Risk Neutral)?

 

[SoBeFla]

I used the same reasoning too i. e. compared delta chages for both the parties

Yeah, i considered that and you may well both be right. I didn't go that route because I recalled reading that the higher rated C/P usually charges the lower rated party. Also, as matter of practice, dealer/banks typically charge and there is not much of a precedent for an end user to charge CVA to a dealer. Tricky question for sure.