FRM Level 2 , Nov 2012 : Post what you remember here

[Jot]

here a few more questions I remember and the answers I picked

- How to best mitigate systemic risk - Answ. Insure senior tranches of securitizations
- Modified duration of a 15y semi-annual compounding zero bond - Answ. 14.5
- Impact of new products - Answ. new products can alter fundamentals and standards of a market or similar
- How to replicate debt under Merton Model - Answ. short put plus risk free bond
- Question on non-parametric VaR methods - Answ. either bootstrapped simulation can yield outliers that distort Var
or bootstrapping can lead to higher Var estimates
- Calculate Liquidity Cost with random spread approach, a so called volatility spread factor of 3 was given - Had never heard of a volatility spread factor in this context before and didnt get to any of the numbers given and finally picked 1.5m or 1.7m
- What is not a building block regarding economic capital discussions –
interest rate risk in banking book, counterparty risk, dependence in credit risk, ? --> Answ. interest rate risk in banking book
- What kind of debt does a typical Private Equity LBO deal use? - Answ. mostly senior bank loans
- One on Prepayments and Strips – What will decrease the avg life of a PO? Answ. higher prepayment speed
- Question on securitization - Answ. subordination of equity tranche is considered part of internal credit enhancement

 

[fnctkaa]

here a few more questions I remember and the answers I picked

- Calculate Liquidity Cost with random spread approach, a so called volatility spread factor of 3 was given - Had never heard of a volatility spread factor in this context before and didnt get to any of the numbers given and finally picked 1.5m or 1.7m

I also don't know what vol spread factor is, and this was one of the questions that I can't figure out the numbers.

- What is not a building block regarding economic capital discussions –
interest rate risk in banking book, counterparty risk, dependence in credit risk, ? --> Answ. interest rate risk in banking book

I got a different answer but I forget what I picked
Interest rate risk in banking book is difficult to estimate due to its maturity nature ???

 

[shady007]

here a few more questions I remember and the answers I picked


- Impact of new products - Answ. new products can alter fundamentals and standards of a market or similar.

I dont remember choosing this. Can u list other options in case u remember...

- How to replicate debt under Merton Model - Answ. short put plus risk free bond
This is correct.


- Calculate Liquidity Cost with random spread approach, a so called volatility spread factor of 3 was given - Had never heard of a volatility spread factor in this context before and didnt get to any of the numbers given and finally picked 1.5m or 1.7m

In case of exogenous approach, LC = P*(mean spread + K* volatility spread)/2 where K is volatility spread factor....

- What is not a building block regarding economic capital discussions –
interest rate risk in banking book, counterparty risk, dependence in credit risk, ? --> Answ. interest rate risk in banking book

Can u explain the rationale for the same. I think I chose liquidity risk.

- What kind of debt does a typical Private Equity LBO deal use? - Answ. mostly senior bank loans

Schweser says "Different levels of debt include senior debt, provided by banks and secured by firm's assets, aiind also subordinated debt, which is unsecured. This type of high-yield debt is raised in the high-yield capital markets"

- Question on securitization - Answ. subordination of equity tranche is considered part of internal credit enhancement

Can u list other options....is it just the overcollateralization question we were already discussing....

 

[klartxt]

Not sure why the volatility spread factor confuses some - in many practice questions for LVaR, for the LC (= liquidity cost = spread) part of the LVaR there was either a factor given, usually 3 (pretty much equivalent to a 99.9% conf level, i think), or you were asked to use a confidence level which in case of the spread is a two-tailed z-value for the significance level.
The VaR part can therefore use a different factor or confidence level than the spread part which is added to arrive at LVaR (=VaR + LC).

Regarding the question about what is not part of the building block of economic capital ... Was it economic capital or regulatory capital? That will be important in answering the question. Not sure about the answer on that one, but my thought was that mkt risk, credit risk and operational risk would have to be included - and liquidity risk is "the ultimate" (as Schweser calls it) part of operational risk. Therefore I did not choose liquidity risk to be the exception. Counterparty risk = CVA risk, I think (credit value adjustment?), and, if I recall correctly, is part of regulatory capital in Basel III - however, many banks do not want to price CVA risk for ECONOMIC capital because they fear losing business... So I would narrow the answer down to either dependency modeling of credit risk or counterparty risk

 

[klartxt]

klartxt, after going through the answers the first time I also thought that none of the answers can be correct. So what I finally did is to adjust the given past returns of the market fund (7% I think it was) with the betas of stock A and B (0.8 and 0.9 I believe) in order to get some forward looking/expected returns for the market fund. Then I re-calculated the IR with the adjusted market fund returns and this time the answer involving the IR was correct. Not sure if this is the right approach but it was the best I could come up with, everything else did not make any sense.

Jot, thanks for replying. Interesting take on that question. sounds plausible. So now we have 3 equally correct choices... You recall the betas correctly, just the other way around i think, 0.9 for asset A, 0.8 for asset B. Instead of using the benchmarks past return to calculate the excess return, I thought of using the expected return of A, which was 10%, and dividing it by its beta, so 10%/0.9, to arrive at 11% for the benchmark (which per definition has a beta of 1)... But I stopped there - which probably was a mistake since using the same method for asset B also roughly (rounded down) comes out to that return for the benchmark: 9%/0.8 = 11%. This gives an "excess" return of -1% for asset A and -2% for asset B, so asset A still has the "higher" excess return. So maybe choice a) (choosing asset A for its higher excess return) makes no sense because the excess return is negative?
Following the logic of recomputing the info ratio:
Asset B had the lower volatility - and the lower excess return of -2% - ... not sure if i could recalculate the Info Ratio now, as calculating the difference in volatility in the denominator involves the covariances, doesn't it (i don't think it's just one std deviation minus the other, is it?). Anyway, just thinking of it: wasn't the information ratio given? Why recalculate?

 

[deaf rodent]

Q9 was straight forward right? I looked at the table and took the values corresponding to 99% CL i.e 3*Var(t-1)+3*SVar(ave)...funny I remember 275 for some reason? I hope i didnt mistake 3*35=95 instead of 105

 

[klartxt]

i thought they asked for general market risk charge - that did include the SVaR, yes? Assuming so, the answer would be max(VaR of t-1; 3 times average VaR over last 60 days) + SVaR whereby the SVaR is calculated the same way here. There can be a plus-factor which is added to 3 in the case of VaR depending on the number of exceptions (no plus factor for SVaR) for a maximum of 3+1, but there was no plus factor given in that question if I recall correctly.
So, in short: max(VaR t-1; average VaR over last 60 days times 3) + max(SVaR t-1; average SVaR over last 60 days times 3).
The VaR values above are 10 day 99%.

I think I answered that one wrong because I thought the question asked for the GENERAL market risk charge would only be the first part, and the SVaR would apply to the SPECIFIC market risk charge. Adding both together would come to the TOTAL market risk charge. So - not sure actually whether i got that one right or wrong...?

 

[varun34by02]

There were two questions, one about the Singaporean client with its USD revenues and the other about Nasdaq. Any thoughts?

Imad - For SGP client - I think it was abt wrong way risk and what action to take ...

 

[varun34by02]

Duration of IO strip - Was it -20? (For PO they gave it as 25 if i remember correctly)

Level2WaitingForResults - Qstn was - 5 yrs eff durn for MBS and PO had durn of 20 years and the qstn asked IO .... If i wrote the qstn correct, is the probable answer IO = -15 (5=20+IO) ?

 

[deaf rodent]

Imad - For SGP client - I think it was abt wrong way risk and what action to take ...

There was only one sttement abt wrong way risk in the available choices? I chose the the answer relating to wrong way risk here too.

 

[deaf rodent]

yeah..i also remember this one...-15 looked correct. I remeber that IO price relationship is positive to increasing interest rates

 

[Jot]

tracking errors for stocks A and B were given, so you could re-calculate IR with the adjusted excess returns for stocks A and B

 

[Jot]

klartxt, thanks. If you think of the volatility spread factor as a z-value for the spread it makes perfect sense to me. Didnt think of it this way in the exam, too bad

 

[Jot]

Regarding the question about what is not part of the building block of economic capital ... Was it economic capital or regulatory capital? That will be important in answering the question. Not sure about the answer on that one, but my thought was that mkt risk, credit risk and operational risk would have to be included - and liquidity risk is "the ultimate" (as Schweser calls it) part of operational risk. Therefore I did not choose liquidity risk to be the exception. Counterparty risk = CVA risk, I think (credit value adjustment?), and, if I recall correctly, is part of regulatory capital in Basel III - however, many banks do not want to price CVA risk for ECONOMIC capital because they fear losing business... So I would narrow the answer down to either dependency modeling of credit risk or counterparty risk[/quote]

pretty sure it said economic capital. I was guessing on this one

 

[Jot]

Imad - For SGP client - I think it was abt wrong way risk and what action to take ...

chose the answer that said 10m threshold for this client is too low and should be revised

 

[Jot]

Level2WaitingForResults - Qstn was - 5 yrs eff durn for MBS and PO had durn of 20 years and the qstn asked IO .... If i wrote the qstn correct, is the probable answer IO = -15 (5=20+IO) ?

yes, sum of durations for IO and PO must equal that of MBS

 

[Jot]

shady, the question on securitization I listed above was different from the O/C question. On the O/C question I picked the excess spread answer.

 

[David Harper CFA FRM]

Thank for this great feedback thread on P2, I've collected some very useful information, we really appreciate the efforts to share

 

[nageshtheone]

Thank for this great feedback thread on P2, I've collected some very useful information, we really appreciate the efforts to share

Thank you sir.
You know, sometimes I feel that the manner in which the syllabus is designed, and also the exam, is done on purpose. It is not only a test of concepts, but also a test of one's ability to apply them under tremendous pressure. The philosophy seems to discourage the candidate as much as possible and then test his/her mettle. I hope that I pass the Nov ' 12 P2.
If possible, please do give us a probable passing score range ( out of 80). Just an approxiamte idea.

 

[varun34by02]

I think VaR is the answer here. Let's not overanalyze the question. It clearly says a coherent risk measure and easy to understand. I read in schweser that ES is not easy to understand, but VaR is. And also, even if VaR is only a coherent risk measure for normal dist, it is still a coherent risk measure. That's why the answer is VaR.

But did the question say 'NORMAL DIST' .. I doubt that.. I feel this question can be reported to GARP ... Could we have David's comment here.