Exam Feedback May 2019 Part 2 Exam Feedback

[nikic]

I remember a) 60000, b) 1000000; and no more, I remember selecting 60000, and put in the formula 10 in the first part, and -5 in the second part. But of course not sure on that.

Mv

Was this 240000 / 4?

I did the same but it was a blind guess really.

 

[AK88]

(1) On selecting the manager with the best performance, options were alpha, mogdiliani, information ratio. etc
I choose " Mogdiliani " . Anyone ? many of my friends choose " IR "
(2) Answer = "declaring cash dividend will reduce Tier 1 capital" ( what about the other choices for this question? cannot remember clearly what the questions is..)
(3) New CDO, assesses data quality, which is the biggest issue? -> Answer = different business units have different formats of risk data (option D)
I think I select something like " build a centralized data warehouse to store data ?" anyone with the same answer ?
(4) Regresssion hedge - obtain beta value from the table -> Answer : I pick " 711" but I think I'm wrong. the correct answer should be 677 ?
(5) Collateral – threshold, remargin period etc -> Answer = high threshold ( what's the question about ? anyone remember whats the other answers for this question? )

Thank you.

I also chose IR for selecting the best manager;
For sharp ratio optimization I selected do nothing, as return/sigma was the highest already;
For #5 I vaguely recall selecting something about threshold as well

 

[nikic]

Anyone else who's sat for Part II before, what's your take on the May 2019 exam difficulty with prior sittings? On par, trickier etc?

 

[bhawsinghka.ritwik]

There was this question that I remember. A table was provided highlighting the VaR at different Confidence Levels and we had to calculate ES. Now, when we're calculating this, we simply take the average of all values given, except the first one (95% VaR).
I was getting two answers - one using 10 values and one using 9 values.
Can anyone tell me which one is correct?

 

[bhawsinghka.ritwik]

But this question has appeared exactly the same in the GARP samples and the answer was 10.0% there?

I honestly thought this was a free and easy mark...

Memory is hazy now....but wasn’t the question worded EXACTLY as above? Or was it worded differently?

This is how I calculated -

A. PD in Year 1 = 1-e^(-0.12) = 11.30%
B. Cumulative PD in Year 2 = 1-e^(-2x0.12) = 21.33%
Hence, PD in Year 2 = A - B = 21.33 - 11.30 = 10.03%

Now, Conditional PD = 10.03/.887 = 11.3%

This is the answer for all years.

 

[kumaranimesh09]

I also chose IR for selecting the best manager;
For sharp ratio optimization I selected do nothing, as return/sigma was the highest already;
For #5 I vaguely recall selecting something about threshold as well

I chose Treynor ratio for selecting best manager, because the question mentioned that it was a well diversified portfolio. This means that unsystematic risk should not be considered for selecting best manager, which left us only the Treynor ratio as the answer.

 

[nikic]

I chose Treynor ratio for selecting best manager, because the question mentioned that it was a well diversified portfolio. This means that unsystematic risk should not be considered for selecting best manager, which left us only the Treynor ratio as the answer.

Was this the question where the other options were alpha (vs benchmark), alpha (vs peers) and mogdiliani?

 

[kumaranimesh09]

For the BCVA question, was it required to use the discount factor in the calculation? Or was it a simple application of EE*LGD*PD-NEE*LGD*PD?

 

[nikic]

There was also a question where they gave a table with the number of surviving companies over 3 or 4 years.

Then they asked to calculate the probability of default within Year 4 (or Year 3, I can’t recall the number of years shown in the table).

Was this a marginal PD or conditional PD?

 

[nikic]

There was this question that I remember. A table was provided highlighting the VaR at different Confidence Levels and we had to calculate ES. Now, when we're calculating this, we simply take the average of all values given, except the first one (95% VaR).
I was getting two answers - one using 10 values and one using 9 values.
Can anyone tell me which one is correct?

It should be average of 9 values - you exclude the 95.0th percentile.

 

[AK88]

For the BCVA question, was it required to use the discount factor in the calculation? Or was it a simple application of EE*LGD*PD-NEE*LGD*PD?

I did apply this, but not sure if it is correct.

 

[amit.m.sharma]

There was also a question where they gave a table with the number of surviving companies over 3 or 4 years.

Then they asked to calculate the probability of default within Year 4 (or Year 3, I can’t recall the number of years shown in the table).

Was this a marginal PD or conditional PD?

I was hard pressed for time when I got to this question. Just went with my intuition, PD = number of defaults in year 4/ number of companies surviving at the end of year 3

 

[nikic]

I was hard pressed for time when I got to this question. Just went with my intuition, PD = number of defaults in year 4/ number of companies surviving at the end of year 3

That's precisely what I did, but it'd only be the correct answer if it was Conditional Probability of Default. If it was Marginal Probability of Default, then we'd need to take Probability Default in 4 years - Probability Default in 3 years.

@David Harper CFA FRM if a question doesn't specify and merely asks "calculate the probability of default in the 4th year", which do we use - conditional or marginal?

 

[amit.m.sharma]

That's precisely what I did, but it'd only be the correct answer if it was Conditional Probability of Default. If it was Marginal Probability of Default, then we'd need to take Probability Default in 4 years - Probability Default in 3 years.

@David Harper CFA FRM
if a question doesn't specify and merely asks "calculate the probability of default in the 4th year", which do we use - conditional or marginal?

To add to @nikic, the numbers as I recall were similar to the following:
Say there are 400 companies at time t=0 and the number of companies surviving at the end of each year is,
T=1 , surviving = 392
T=2 , surviving = 383
T=3 , surviving = 376
T=4 , surviving = 372
Then what is the default probability during year 4?

 

[Anoop Kumar]

To add to @nikic, the numbers as I recall were similar to the following:
Say there are 400 companies at time t=0 and the number of companies surviving at the end of each year is,
T=1 , surviving = 392
T=2 , surviving = 383
T=3 , surviving = 376
T=4 , surviving = 372
Then what is the default probability during year 4?

the asnswer to this would it be 28/400 - 24/400 =1% ??

 

[Anoop Kumar]

I felt that the paper was tricky for at least 20-35 questions i spent close to 200hrs on material n revision @nikic & other members how much time did u spent on material n revision. Based on historical passing percentages and compared to CFA in terms of level of difficulty the MPS would not exceed 50 (Pessimistic range 44 - 50 ,Optimistic range 42 - 46). if it goes beyond 50 then it would suggest that more than 50% of exam takers felt the paper was easier as compared to us and better prepared than us. All the best to everyone for the results.

 

[nikic]

I felt that the paper was tricky for at least 20-35 questions i spent close to 200hrs on material n revision @nikic & other members how much time did u spent on material n revision. Based on historical passing percentages and compared to CFA in terms of level of difficulty the MPS would not exceed 50 (Pessimistic range 44 - 50 ,Optimistic range 42 - 46). if it goes beyond 50 then it would suggest that more than 50% of exam takers felt the paper was easier as compared to us and better prepared than us. All the best to everyone for the results.

I spent probably between 140 and 180 hours on revision. But I’ve got Risk experience and familiarity with the ops risk component (such as the different loss categories and Basel 3) due to my work experience. Still I’m not sure if reading more would’ve helped as I’m not the memorising type. But I do think I paid too little attention on credit risk as it’s the one area I neither have personal interest nor working experience in. If I fail it’d be due to a 4 in credit risk, that I am sure of.

On the other point in your post, you say MPS would not exceed 50. Is this 50/80 questions or 50% (i.e. 40/80). If it’s the latter what makes you think the passing score is so low?

 

[Anoop Kumar]

I spent probably between 140 and 180 hours on revision. But I’ve got Risk experience and familiarity with the ops risk component (such as the different loss categories and Basel 3) due to my work experience. Still I’m not sure if reading more would’ve helped as I’m not the memorising type. But I do think I paid too little attention on credit risk as it’s the one area I neither have personal interest nor working experience in. If I fail it’d be due to a 4 in credit risk, that I am sure of.

On the other point in your post, you say MPS would not exceed 50. Is this 50/80 questions or 50% (i.e. 40/80). If it’s the latter what makes you think the passing score is so low?

I spent probably between 140 and 180 hours on revision only on revision thats very huge. waht about reading it for the first time ? and i said pass rate 50% and not pass perecentage marks and 50 is 50/80

 

[nikic]

Sorry

I spent probably between 140 and 180 hours on revision only on revision thats very huge. waht about reading it for the first time ? and i said pass rate 50% and not pass perecentage marks and 50 is 50/80

Sorry 140 to 180 hours in total including the first read, revision and questions.

 

[y88888888]

Anyone remember a question to compare VaR between cashflow mapping and duration mapping, I marked cashflow because I think cashflow VaR is unsmoothed.