Exam FeedbackFRM Part 2 (November 2014) Exam Feedback

Roshan Ramdas

Active Member
Hi guys we have discussed/posted 60 questions out of 80 that perhaps appeared on exam...just keep grinding..we need another 10 questions to see how we all have performed in 70 questions that appeared on exam..
There was one question on factors that need to be considered while looking at quantile standard errors - I selected asymmetric confidence intervals
Another on facts pertaining to bootstrap, age weighted & volatility weighted HS - I selected age weighting being more responsive to losses,....data cluster.

HT_William

New Member
They give the formula :

z = (x - pt) / SQRT(p*(1-p)T)

so x = z * SQRT(p*(1-p)T) + pT

x = 1,96 * SQRT(0,99*0,01*2500) + 2500*1%

x = 1,96 * 4,97 + 25 = 35

The trick was 99% VaR one tailed but 95% confident two tailed so z=1,96...

I first made the computation with 2,33 and 1,65 but luckily no answer matched.

I understand, but my struggle was that the exact figure computed should be 34.75....
Therefore if X=35, the hypothesis should be rejected....

monumonumental

Member
There was one question on factors that need to be considered while looking at quantile standard errors - I selected asymmetric confidence intervals
Another on facts pertaining to bootstrap, age weighted & volatility weighted HS - I selected age weighting being more responsive to losses,....data cluster.
even i selected asymetric confidence interval..

monumonumental

Member
There was one question on factors that need to be considered while looking at quantile standard errors - I selected asymmetric confidence intervals
Another on facts pertaining to bootstrap, age weighted & volatility weighted HS - I selected age weighting being more responsive to losses,....data cluster.
I think i selected bootsrap may cause greater volatality compared to HS....I don't think your answer is correct as in age wt HS the probablity of impact of historical losses are reduced...i hope you read the question carefully..question was talking about historical losses..

Last edited:

monumonumental

Member
u agree with whose answer? me or rams?

Hoang

New Member
I agree with Bootstrap too !

Tanya Vasileva

New Member
Subscriber
Hi Tanya - Are you sure that there was mention about the haircut already being applied to the level 2A ??
I did perform the 40% check on the Level 2A assets but forgot to account for the haircut and landed up using the Level 1+Level 2A in the numerator. My answer should still be correct if the Level 2A figure already contained the haircut.

frmexam

New Member
about var/Svar someone replied answer as 320, how did you get to that? I know how its calculated as per Basel not sure prior to basel. So i picked difference 500-XXX as 350 as a guess

Prince31

New Member
I marked 25 for the question. My reasons for marking 25 were as follows
35-25/4.975=2.01 which is greater than 1.96. Hence the hypothesis that the true p=.01 is rejected.

They give the formula :

z = (x - pt) / SQRT(p*(1-p)T)

so x = z * SQRT(p*(1-p)T) + pT

x = 1,96 * SQRT(0,99*0,01*2500) + 2500*1%

x = 1,96 * 4,97 + 25 = 35

The trick was 99% VaR one tailed but 95% confident two tailed so z=1,96...

I first made the computation with 2,33 and 1,65 but luckily no answer matched.

Tanya Vasileva

New Member
Subscriber
The reason why I selected consumer credit was because there was a mention of the usage of statistical tools and (I think) scoring methodologies and this matches the profile of individual (consumer) credit. I may be wrong here,....but there is a write up in paragraph 2 page 21 (GARP's credit risk book) that leads me to believe that that this was the answer.

Scoring models are relevant to consumers, so it is the correct answer here

Roshan Ramdas

Active Member
about var/Svar someone replied answer as 320, how did you get to that? I know how its calculated as per Basel not sure prior to basel. So i picked difference 500-XXX as 350 as a guess

The Basel 2 market risk capital charge did not have an SVAR component

So the difference is -> Market Risk Capital Charge (with SVAR) - Market Risk Capital Charge (without SVAR)

monumonumental

Member
about var/Svar someone replied answer as 320, how did you get to that? I know how its calculated as per Basel not sure prior to basel. So i picked difference 500-XXX as 350 as a guess
I was not sure about it as well...but i think historically Basel only use to consider Var and later it started considering both Var and Svar....So the difference between two methods will be the value of Svar ie max of (3*100,320)=320

monumonumental

Member
I think going by the general discussion in this thread i have got at least 45 correct on 60 questions being posted here....i think i have some chances of passing now..

frmexam

New Member
Roshan - Can you correct what I am missing here, my understanding is
MRC diff = Basel VAR - Prior to Basel Var
Basel MRC = Max(var, M*Varavg) + Max (svar, M*svaravg) = 500
As per your definition then Prior to Basel Var is 150 (3*50k)
then the answer should be 500-150 = 350

monumonumental

Member
I think both of us are
Roshan - Can you correct what I am missing here, my understanding is
MRC diff = Basel VAR - Prior to Basel Var
Basel MRC = Max(var, M*Varavg) + Max (svar, M*svaravg) = 500
As per your definition then Prior to Basel Var is 150 (3*50k)
then the answer should be 500-150 = 350
correct concept wise...but one of us have surely messed with calculations..

Nicole Seaman

Director of FRM Operations
Staff member
Subscriber
Hi Nicole/David,

Based on your experience, do you always get this kind of "very Difficult" feedback from students or are we a "bunch of ill prepared losers"
Hello @sudssuds,

There are no "ill prepared losers" here There has always been mixed feedback, but many past candidates have stated that the exam was very difficult. Here is a link to the feedback thread that we posted for the FRM Part 2 May 2014 exam: https://forum.bionicturtle.com/threads/frm-part-2-may-2014-exam-feedback.7834/. I think you will find that there are many of the same comments.

Roshan Ramdas

Active Member
Roshan - Can you correct what I am missing here, my understanding is
MRC diff = Basel VAR - Prior to Basel Var
Basel MRC = Max(var, M*Varavg) + Max (svar, M*svaravg) = 500
As per your definition then Prior to Basel Var is 150 (3*50k)
then the answer should be 500-150 = 350
Hello,
The method you have followed is perfect. I don't remember the exact numerical answer.
Thank you

krenate

Member
it was one question regarding convertible arbitrage... I do not remember the answers though, however, maybe someone does?
due to 50 out 80...I am certain that it is a strong pass....

monumonumental

Member
it was one question regarding convertible arbitrage... I do not remember the answers though, however, maybe someone does?
due to 50 out 80...I am certain that it is a strong pass....
So if we get 50/80 we pass??

Replies
37
Views
6K
Replies
26
Views
2K
Replies
40
Views
10K
Replies
23
Views
6K
Replies
275
Views
52K