Exam Feedback FRM Part 2 (May 2015) Exam Feedback

[JDGutzmann]

Hello,
firstly, I found the exam to be fair both in terms of time to be used and difficulty of the questions, although there were again many "most appropriate" / "best practice"-questions. I find these highly irritating.

Here are a few thoughts on what has been mentioned above:

• For the OpRisk SA >> BIA question I realized you had to ignore the last column (year 2011), since both approaches only employ three years of data history. --> plus 37M
• QQ-Plot: I found it to have lighter tails and a POSITIVE mean (shifted upwards).
• I also chose the CDS-correlation graph that goes to zero, because I wouldn't pay for a protection that becomes worthless the moment the underlying defaults.

What I didn't get:

• Did anyone have any idea on the "long correlation" position? Also, there were some questions regarding the proper use of distributions that had me stumped.
• I am not quite sure what GARP's position is regarding clearinghouses.

Best luck to everyone!

Johannes

 

[vamsib555]

But isnt it the same for the spread? I would pay nothing for a cds that has no value for me?

CDS Value means value of the derivative while spread is the premium that is being charged from buyer. Why would any investor/buyer buy a cds when correlation is enough to tackle the issue at hand ? hence when correlatoion -> 1, then cds spread -> 0

 

[vamsib555]

i was stumbled.. whether to consider BB- as investment grade or non investment grade when answering the risky loans question (X,Y,Z).. i used the formula Expected loss = (1-RR) * PD* Exposure and arranged the loans with respect to their Expected loss values.. did any one solve that question?

 

[robin3301]

i was stumbled.. whether to consider BB- as investment grade or non investment grade when answering the risky loans question (X,Y,Z).. i used the formula Expected loss = (1-RR) * PD* Exposure and arranged the loans with respect to their Expected loss values.. did any one solve that question?

non investment grade

 

[berimbolo]

Another few I can remember
- Use of lognormal wrong for bondprices
- Had lighter tails for the QQ plot
- Convexity adjustment downwards
- CCP question about a fradulous trade
- Net stable funding ratio (had not enough funding)
- zero correlation doesn't imply independence
- LIBOR swap valuation, didn't know how to do that
- CLN question
- A question about a CDS for a BBB rated company with a AA rated company or something

 

[vamsib555]

Another few I can remember
- Use of lognormal wrong for bondprices
- Had lighter tails for the QQ plot
- Convexity adjustment downwards
- CCP question about a fradulous trade
- Net stable funding ratio (had not enough funding)
- zero correlation doesn't imply independence
- LIBOR swap valuation, didn't know how to do that
- CLN question
- A question about a CDS for a BBB rated company with a AA rated company or something

At the initiation of swap. fixed leg is equal to floating leg and net value of swap is zero. Question asked was to calculate the floating leg. So, it will be

= 6 (fixed swap payment)/ (1+ 1 year OIS rate) + 6+100 (fixed swap payment + notional)/(1+ 2 year OIS rate)^2 . answer i remember is vaguely 100.95 or 4th option. Not sure

 

[robin3301]

Another few I can remember
- Use of lognormal wrong for bondprices
- Had lighter tails for the QQ plot
- Convexity adjustment downwards
- CCP question about a fradulous trade
- Net stable funding ratio (had not enough funding)
- zero correlation doesn't imply independence
- LIBOR swap valuation, didn't know how to do that
- CLN question
- A question about a CDS for a BBB rated company with a AA rated company or something

convexity adjustment increase bond price

 

[Becky]

There was one question with the Jensen's and convexity. I had some doubt between two answers because of the concexity part in the explanation.
An other one related to scoring model. Which can help during the process: SVM, merton, Area under the Curve or ROC.
Even for the smirk and the 4 graphs I had some doubt between two.

 

[vamsib555]

Can anybody tell me.. is there any sectional cutoff for FRM part 2 ? i.e. getting 4 in any section leads to failure ?

 

[ami44]

- LIBOR swap valuation, didn't know how to do that

I felt something was missing from the IRS question. The 2 year swap payed 6% fix and vas valued at par. I.e. Was worth zero. The question was the value of the floating leg with ois rates 5% for first year and 5.5% for 2 years. I think you need the current libor fixing and the forward rate from first to second year to solve this. From the information, that the swap is valued at par you can get one unknown, but you are still missing the second. Or am I missing something here.

I solved it by assuming that the current fixing and the forward rate are both 6%. Dont know if result is correct, i think other solutions are possible.

 

[Goldfish3]

Hello All,

To me the exam was a hit or miss. about 30 questions were very straght forward which i think everyone would get right.so im guessing about 25 to 30 correct questions would be the cutoff for the 4th quartile. I struggled with majority of the remaining questions by not being able to choose between 2 choices (as i was able to eliminate 2 in most questions).

Overall dont think it was too hard but i wasnt prepared enough For sure. Struggled with the simplest questions (or at least which seemed simple) e.g. where 100m bugget needs to be assigned to 3 uncorrelated managers with 10% volatility. Kept trying till the very last second couldnt get the answer. Can someone please scratch this itch for me?

Best of luck.

 

[robin3301]

There was one question with the Jensen's and convexity. I had some doubt between two answers because of the concexity part in the explanation.
An other one related to scoring model. Which can help during the process: SVM, merton, Area under the Curve or ROC.
Even for the smirk and the 4 graphs I had some doubt between two.

1. convexity adjustment increases bond price
2. I choose SVM because it mentions lending money to small business

 

[ami44]

Overall dont think it was too hard but i wasnt prepared enough For sure. Struggled with the simplest questions (or at least which seemed simple) e.g. where 100m bugget needs to be assigned to 3 uncorrelated managers with 10% volatility. Kept trying till the very last second couldnt get the answer. Can someone please scratch this itch for me?

Best of luck.

Im not 100% sure, but i thought the whole budget was 100 Mio. And from the answers only one was below that. The others were 140 Mio. and above. So i choose the one below the total budget, because of no correlation, the seperate VaRs must be lower than portfolio VaR.

 

[berimbolo]

There were also two questions about illiquid bonds (one with a hedge fund?). Options were autocorrelation, high beta,..
Can't remember it very well.
With the budgetting I had C, pretty sure about that one

 

[robin3301]

There were also two questions about illiquid bonds (one with a hedge fund?). Options were autocorrelation, high beta,..
Can't remember it very well.
With the budgetting I had C, pretty sure about that one

i choose autocorrelation

 

[BTO]

Hello All,

To me the exam was a hit or miss. about 30 questions were very straght forward which i think everyone would get right.so im guessing about 25 to 30 correct questions would be the cutoff for the 4th quartile. I struggled with majority of the remaining questions by not being able to choose between 2 choices (as i was able to eliminate 2 in most questions).

Overall dont think it was too hard but i wasnt prepared enough For sure. Struggled with the simplest questions (or at least which seemed simple) e.g. where 100m bugget needs to be assigned to 3 uncorrelated managers with 10% volatility. Kept trying till the very last second couldnt get the answer. Can someone please scratch this itch for me?

Best of luck.

Since stdev is 10% each and they are uncorrelated. This implies portfolio stdev of 5.72% or thereabout. VaR = 100 = p * 5.72% * z. Solving for p = ~ 750. Since equally weighted divide by 3 ~ 248. Could be wrong but that was my solution.

 

[robin3301]

Since stdev is 10% each and they are uncorrelated. This implies portfolio stdev of 5.72% or thereabout. VaR = 100 = p * 5.72% * z. Solving for p = ~ 750. Since equally weighted divide by 3 ~ 248. Could be wrong but that was my solution.

my calculation:
sqrt(3 x VaR^2) = 100
x=58

 

[ami44]

Th

Can anybody tell me.. is there any sectional cutoff for FRM part 2 ? i.e. getting 4 in any section leads to failure ?

your pass/fail only depends on your total score. No sectional cutoff.

 

[ami44]

At the initiation of swap. fixed leg is equal to floating leg and net value of swap is zero. Question asked was to calculate the floating leg. So, it will be

= 6 (fixed swap payment)/ (1+ 1 year OIS rate) + 6+100 (fixed swap payment + notional)/(1+ 2 year OIS rate)^2 . answer i remember is vaguely 100.95 or 4th option. Not sure

I gave the same answer, but i'm not so sure, that it is correct. Essentially you valued the fixed leg with ois discounting. Its not clear to me, why that should be the same as the floating leg. They are valued at par with libor discounting, that might be different under ois discounting, right?

 

[Abhishek...]

On the lighter side ....If GARP ask 2-3 simple questions like:
What is the full form of GARP?
What is the difference of questions asked in Part 1 and Part 2?
It will lighten the candidate's stress or what?