Exam Feedback November 2018 Part 2 Exam Feedback

Narchie02

New Member
Am I the only one who had suggested 'pool insurance' for securitization CE? Waterfall and OC will be rendered useless - even for senior tranches - if loss defaults and LGD is expected to spike tremendously...
 
Last edited:

raghav159

New Member
1. for the edpm the option mentioned something else instead of plain vanilla edpm being lower in subscription database. I think there was something wrong with another option - damage to physical assets (it only features in severity and not frequency but the option mentioned frequency also.)
2. For unsmoothing the option that said increases observed risk and decrease returns or any other option that had anything to do with returns was wrong to my understanding because unsmoothing wont affect the returns. Still cant remember what the other options were.
3. This question on AML cost me 300 seconds or so -
  • socially backward people should be treated as higher risk [out of bounds]
  • In the algorithm for filtering AML transactions good credit the filter should also be applied on good credit individuals
4. Another strange question on validation of models. Here there were 3 options that were structurally right/sound but there was something to do with the scope of what comes under the ambit of model validation
  • chose utlizing knowledge of developers
  • something to do with independently getting the rates from some backoffice
5. There were two questions on dollar. The easier one was probably the CIP one. the tougher one dealt with -
If dollar strength increases, the economic activity in USA increases & banks will reduce borrowing,
-- Here there was some other option I am forgetting. Both the options had one part that was convincing but the part followed by the conjunction "and" was odd.

6. Merton & KMV -
  • Merton doesn’t account for changes in asset volatility - I didn't mark this because my understanding is that in the 1st step to merton we input equity price & volatility basis which we get a combo of outputs on asset value and volatility so asset volatility changes can (probably) be reflected (in the D2D formula the updated asset vol. can serve as a dynamic input)
  • For calculating PD KMV takes into account historical equity price only - I chose. This was such an appealing choice but probably the "only" was little odd here. Still not sure though.
7. Screening, stratification, LP and QP? It definitely required an open textbook to answer this as 2 choices were really close
The one on screening said - it can be used to initially create a portfolio (which was definitely right) and then screening can also be used to rebalance the portfolio. - I chose this
LP doesn't take into account transaction costs explicitly. (The word explicitly made me spend another 100 seconds or so on this question because I know that LP optimises alpha - TC but doesn't factor the active risk {unlike QP}, but the word explicitly made it sound fishy. Still don't know whether LP optimises TC implicitly (through controlling for turnover of stocks like stratification & screening) or explicitly like QP.

On that note, there were atleast 15 questions which would not have a conclusive answer even if you had 10 people doing a group discussion on those questions. Take for example the following question

8. You invest in Class A bonds of company. Which of the following credit enhancement will be the best for you :
  • Cashflow Waterfall
  • Overcollat
  • Step-up margin
  • Some other option (which was a cred. enhancement.

I again wasted time thinking between waterfall & overcollat. I had read somewhere (in the assigned reading) that overcollat is only meant for equity tranche although the numericals suggested otherwise. I had almost selected waterfall (after unneccesarily wasting so much time on this supposedly simple question all because of that piece of text, then I reread the question and saw the word credit enhancement mentioned & stayed away from that option.
I am sure people would have selected overcollat & called it a simple question but in fact it is not.

There were many such questions honestly and I am not sure whether the pattern is the same every year, but to all the folks who thought the paper was too demanding on some areas at times, I am sure it definitely was - there were questions where you just had to put the value of lamda to get the cpd and then there were questions like the aforementioned ones.

On a different note, If GARP does ever publish these (retired) questions alongwith answers, I would be as curious then (though with grey hair) as I am today to find out what the answers to them actually were. (Only if I would still manage to remember what "I" had answered).
 
1. for the edpm the option mentioned something else instead of plain vanilla edpm being lower in subscription database. I think there was something wrong with another option - damage to physical assets (it only features in severity and not frequency but the option mentioned frequency also.)
2. For unsmoothing the option that said increases observed risk and decrease returns or any other option that had anything to do with returns was wrong to my understanding because unsmoothing wont affect the returns. Still cant remember what the other options were.
3. This question on AML cost me 300 seconds or so -
  • socially backward people should be treated as higher risk [out of bounds]
  • In the algorithm for filtering AML transactions good credit the filter should also be applied on good credit individuals
4. Another strange question on validation of models. Here there were 3 options that were structurally right/sound but there was something to do with the scope of what comes under the ambit of model validation
  • chose utlizing knowledge of developers
  • something to do with independently getting the rates from some backoffice
5. There were two questions on dollar. The easier one was probably the CIP one. the tougher one dealt with -
If dollar strength increases, the economic activity in USA increases & banks will reduce borrowing,
-- Here there was some other option I am forgetting. Both the options had one part that was convincing but the part followed by the conjunction "and" was odd.

6. Merton & KMV -
  • Merton doesn’t account for changes in asset volatility - I didn't mark this because my understanding is that in the 1st step to merton we input equity price & volatility basis which we get a combo of outputs on asset value and volatility so asset volatility changes can (probably) be reflected (in the D2D formula the updated asset vol. can serve as a dynamic input)
  • For calculating PD KMV takes into account historical equity price only - I chose. This was such an appealing choice but probably the "only" was little odd here. Still not sure though.
7. Screening, stratification, LP and QP? It definitely required an open textbook to answer this as 2 choices were really close
The one on screening said - it can be used to initially create a portfolio (which was definitely right) and then screening can also be used to rebalance the portfolio. - I chose this
LP doesn't take into account transaction costs explicitly. (The word explicitly made me spend another 100 seconds or so on this question because I know that LP optimises alpha - TC but doesn't factor the active risk {unlike QP}, but the word explicitly made it sound fishy. Still don't know whether LP optimises TC implicitly (through controlling for turnover of stocks like stratification & screening) or explicitly like QP.

On that note, there were atleast 15 questions which would not have a conclusive answer even if you had 10 people doing a group discussion on those questions. Take for example the following question

8. You invest in Class A bonds of company. Which of the following credit enhancement will be the best for you :
  • Cashflow Waterfall
  • Overcollat
  • Step-up margin
  • Some other option (which was a cred. enhancement.

I again wasted time thinking between waterfall & overcollat. I had read somewhere (in the assigned reading) that overcollat is only meant for equity tranche although the numericals suggested otherwise. I had almost selected waterfall (after unneccesarily wasting so much time on this supposedly simple question all because of that piece of text, then I reread the question and saw the word credit enhancement mentioned & stayed away from that option.
I am sure people would have selected overcollat & called it a simple question but in fact it is not.

There were many such questions honestly and I am not sure whether the pattern is the same every year, but to all the folks who thought the paper was too demanding on some areas at times, I am sure it definitely was - there were questions where you just had to put the value of lamda to get the cpd and then there were questions like the aforementioned ones.

On a different note, If GARP does ever publish these (retired) questions alongwith answers, I would be as curious then (though with grey hair) as I am today to find out what the answers to them actually were. (Only if I would still manage to remember what "I" had answered).

KMV takes current equity prices into account.
 

jkowalonek

New Member
I had OC as well -- if the payments completely default besides the Overcollateralization, I would think the senior tranche would be paid first out of the OC cushion.

Overcollateralization is achieved when the value of assets in the pool is greater than the amount of the asset-backed security (ABS). So, even if some of the payments from the underlying loans default or are late, principal and interest payments on the ABS can still be made from the excess collateral.

Read more: Overcollateralization (OC) https://www.investopedia.com/terms/o/overcollateralization.asp#ixzz5XKHNkCJa
Follow us: Investopedia on Facebook
 

cfa2015

New Member
FRM folks do not know how to write good questions. They need to spend some time with CFA or CPA examiners. A concept can be tested with a concise question. You do not need to make a question long, verbose and meander around the core issue. According to me the questions were tough and obscure and; there was little resemblance to the mock. Comparing my experience with the CFA and CPA certifications FRM requires a lot of improvement on both the syllabus development and examinations.
 

ympetit

New Member
FRM folks do not know how to write good questions. They need to spend some time with CFA or CPA examiners. A concept can be tested with a concise question. You do not need to make a question long, verbose and meander around the core issue. According to me the questions were tough and obscure and; there was little resemblance to the mock. Comparing my experience with the CFA and CPA certifications FRM requires a lot of improvement on both the syllabus development and examinations.

I completely agree. There were many questions that were lengthy and did not resemble the mock exams available in BT.
 

QuantFFM

Member
Hi folks,

I think it was a difficult exam. A lot of questions were really long. Furthermore sometimes the the answer options were in my opinion not well designed.

Regarding the RAROC question I think I choosed 4,75%? I set the formula for RAROC equal to 9,75% and solved it for r.

Maybe wrong..:-(
 

jkowalonek

New Member
What did people put for the question about when an Outsourcing Vendor, uses a Third party to outsource? I'm pretty sure I put you needed to make sure the third party would be held accountable?
 

jkowalonek

New Member
Pool insurance makes sense theoritically, until we saw how the monoliners handled the situation during the financial crisis was.
I keep flip flopping to, but if they don't quantify the size of the default (pretty sure it was going to just be a big more right?), then we wouldn't know that it was completely wiped out, if it was completely defaulted then the pool insurance would kick in? But if there is still some good outstanding loans, then the insurance wouldn't kick in? Is pool insurance all or nothing? This question was terrible .... back to everyones point of poorly written questions.
 

fxlprasetyo

New Member
I keep flip flopping to, but if they don't quantify the size of the default (pretty sure it was going to just be a big more right?), then we wouldn't know that it was completely wiped out, if it was completely defaulted then the pool insurance would kick in? But if there is still some good outstanding loans, then the insurance wouldn't kick in? Is pool insurance all or nothing? This question was terrible .... back to everyones point of poorly written questions.
Agreed, there were a few poorly written questionnaires, hopefully they would do the same as what the CFA does, which is to throw away these poorly written ones.
 

max8126

New Member
I also opted for waterfall thinking that OC helps to protect equity tranche in the first place
I'm genuinely curious now that quite a few people mention OC protects equity tranche.

From my limited exposure to the relevant products, OC kicks in when senior tranches's OC ratio (aka principal coverage) falls below a certain threshold, at which point the payout to equity tranche gets diverted to the senior tranche to either purchase more collateral or (partially) pay off the senior tranche, in order to bring the OC ratio back.
 
Top