I think there are 2 options on adverse selection. Based on David’s notes, resolution is due diligence on the Arranger.
Exam Feedback May 2018 Part 2 Exam Feedback
- Thread starter Nicole Seaman
- Start date
-
- Tags
- feedback
terrance00232
New Member
In Taiwan, it's not allow to have any clock or any watch in the room and I remember that proctors shall tell everyone the time that we remain every half hour during the examination.
yes but the answer was not about due diligence. it has something wrong in the sentence.
schumi_frm
New Member
Indeed, I think it was written "due diligence on the asset manager", which is the other way around.
I was under the same impression and this was what happened in London but then how is it possible for other countries to have had a wall clock in the actual exam room?? I think this is not fair game so was wondering if anyone could advise on any formal policy that GARP might have released regarding this. Maybe @David Harper CFA FRM or @Nicole Seaman would be the best persons to assist.
Last edited:
I think correct answer was the one about assets manager asking the arranger to maintain interest in certain portion of portfolio.
drewmanfsu
New Member
I finished in 2 hours and left 6 questions blank on first pass thru. I expected more theory than math, but was disappointed that only 23/80 questions were math related/involved calcs. I also felt frustrated leaving the exam because it seemed as if they tested a lot of what I would consider fringe material. No real regulatory or Basel questions either. I would suggest BT start incorporating similar questions to what is found on GARP practice tests in previous years. (I learned how to compute CVA charges from old garp and not BT which came up huge on exam) Here is the rest of what I remember:
23 calcs total
2 CVA calcs, one with collateral and one without
Ho lee in binomial tree calc
Vasicek calc 2 periods forward
2 CVaR calcs
Stressed loss minus EL
Contribution by asset allocation calc
IFSCL9 rules for writing off Credit EL vs CECL
Workplace ethics/behavior
Regression hedge calc
2x Identifying risks present in a scenario
Stratification vs screens portfolio construction theory
Freq/severity theory for LDA
Data aggregation principles/best practices
Implication of default % on tranches
Frictions in 2008 crisis theory/scenario “which is right”
2x Implied PD w/CDS spreads
TRS
Marginal VaR calc
Component VaR calc
Option VaR calc
LVaR calc
VaR theory about what mapping methodology reduces VaR
2x ES question calcs - one tail quantiles and other in dollars by %
Incremental VaR calc
Convoluted protection/credit risk scenarios w/CDS exposure and identifying wrong way risk
Scenario for Right way risk increasing or decreasing
Ops risk vs credit risk theory
Machine learning best application
23 calcs total
2 CVA calcs, one with collateral and one without
Ho lee in binomial tree calc
Vasicek calc 2 periods forward
2 CVaR calcs
Stressed loss minus EL
Contribution by asset allocation calc
IFSCL9 rules for writing off Credit EL vs CECL
Workplace ethics/behavior
Regression hedge calc
2x Identifying risks present in a scenario
Stratification vs screens portfolio construction theory
Freq/severity theory for LDA
Data aggregation principles/best practices
Implication of default % on tranches
Frictions in 2008 crisis theory/scenario “which is right”
2x Implied PD w/CDS spreads
TRS
Marginal VaR calc
Component VaR calc
Option VaR calc
LVaR calc
VaR theory about what mapping methodology reduces VaR
2x ES question calcs - one tail quantiles and other in dollars by %
Incremental VaR calc
Convoluted protection/credit risk scenarios w/CDS exposure and identifying wrong way risk
Scenario for Right way risk increasing or decreasing
Ops risk vs credit risk theory
Machine learning best application
Hi @drewmanfsu Thank you for very helpful feedback. Question: what do you mean by "old garp" with respect to CVA? (because we had to explain why their first approach was wrong https://forum.bionicturtle.com/threads/garp-2017-p2-76.10344/). Just trying to understand where is the example of a good CVA question .... Generally with respect to their old exams, we've mostly contributed dozens of errors contributed by our members. It's not often that somebody suggests that we should mine older GARP practice exams which are well known to be plagued with problems. (Don't get me wrong the 2018 Practice Exams are much improved but they are the beneficiary of years of edits/corrections). Thanks,
Last edited:
In KL we had a digital projector Showtime the time
drewmanfsu
New Member
Hi David,
Regardless of whether or not there were mistakes on previous GARP practice tests/exams, I simply meant that BT did not have us practice how to calculate CVA charges [(1-RR)(EE)(PD)(DF)] on any of the BT practice tests. and the only place I had seen or practiced the question setup was from previous GARP practice exams (2012-2015, I can’t remember) So regardless of whether their calcs in practice exams were correct, GARP seems to have demonstrated an importance on the concepts —> and maybe something to add to BT going forward. This is the 2nd GARP exam I’ve taken (after FRM 1) where they have re asked questions that look VERY SIMILAR to the practice questions they release over the years
While thinking of it, GARP also asked some convuluted exposure type questions where several counter parties are involved facing each other with different credit exposures, and it was asked to identify the risks and/or how to hedge. I think questions of this nature would be a huge benefit for understanding to future FRM 2 test takers
Regardless of whether or not there were mistakes on previous GARP practice tests/exams, I simply meant that BT did not have us practice how to calculate CVA charges [(1-RR)(EE)(PD)(DF)] on any of the BT practice tests. and the only place I had seen or practiced the question setup was from previous GARP practice exams (2012-2015, I can’t remember) So regardless of whether their calcs in practice exams were correct, GARP seems to have demonstrated an importance on the concepts —> and maybe something to add to BT going forward. This is the 2nd GARP exam I’ve taken (after FRM 1) where they have re asked questions that look VERY SIMILAR to the practice questions they release over the years
While thinking of it, GARP also asked some convuluted exposure type questions where several counter parties are involved facing each other with different credit exposures, and it was asked to identify the risks and/or how to hedge. I think questions of this nature would be a huge benefit for understanding to future FRM 2 test takers
Did not receive the email yet.
Hi @drewmanfsu Oh I see. But we have always incorporated the concepts (if not the mistakes or confusions) embedded in GARP's previous Practice Exams! Importantly, GARP has never published other questions: neither previous actual questions, nor originally written questions ("embedded") associated with the readings. My view is that they have plenty of budget for both: fresh Practice Exams plus some embedded questions. (Each authors own end of chapter questions, as you probably know, vary significantly in quality and/or relevance; e.g., Diebold's EOC questions are high quality but somewhat inaccessible, while on the other hand, Jorion's or Meissner's EOC questions are not hard enough). Hence the significance of their Practice Exams.
The CVA question is included in our interactive quiz, see 713.3 https://forum.bionicturtle.com/thre...les-and-credit-value-adjustment-ccr-cva.10622
... this is pretty much a direct adaptation of the same CVA question found in the practice exam. At the end of the question, I even wrote: "Note: this question is inspired by GARP's 2017 P2 Practice Exam Question #76."
... it appears in our quiz, or topic review, precisely because it is located in the Practice Exam and we deem it might have special significance
... between topic reviews and/or interactive quizzes, we've attempted to include adaptations of almost anything we find in GARP's Practice Exams
I do very much appreciate your feedback because the broader theme is important: as an EPP, we want to write relevant questions, but GARP needs to help us with what I have termed prospective clarity. Unless/until actual exam questions are shared, our realistic roadmap is limited to: learning objectives (study guide), and their practice exams. If they are asking about "some convuluted exposure type questions where several counter parties are involved facing each other with different credit exposures," as an EPP we need help understanding, especially if the questions deviates from a learning objective. There are a lot of sub-issues, which I have already raised with GARP, including but not limited to:
The CVA question is included in our interactive quiz, see 713.3 https://forum.bionicturtle.com/thre...les-and-credit-value-adjustment-ccr-cva.10622
... this is pretty much a direct adaptation of the same CVA question found in the practice exam. At the end of the question, I even wrote: "Note: this question is inspired by GARP's 2017 P2 Practice Exam Question #76."
... it appears in our quiz, or topic review, precisely because it is located in the Practice Exam and we deem it might have special significance
... between topic reviews and/or interactive quizzes, we've attempted to include adaptations of almost anything we find in GARP's Practice Exams
I do very much appreciate your feedback because the broader theme is important: as an EPP, we want to write relevant questions, but GARP needs to help us with what I have termed prospective clarity. Unless/until actual exam questions are shared, our realistic roadmap is limited to: learning objectives (study guide), and their practice exams. If they are asking about "some convuluted exposure type questions where several counter parties are involved facing each other with different credit exposures," as an EPP we need help understanding, especially if the questions deviates from a learning objective. There are a lot of sub-issues, which I have already raised with GARP, including but not limited to:
- The learning objectives need to be precise and, in some cases, should be streamlined. Gregory, for example, who is assigned here (CVA) is just over-assigned: to cover all of the LOs is, by definition, to fail to give candidates strong direction in where they should focus. There are so many LOs in Gregory that the vast majority cannot be tested. I don't blame you for not noticing this particular CVA question! We have many CVA questions, as there are almost ten CVA LOs, and many more CVA-related LOs. Over-assigned.
- Recency is an issue I have raised: if we go back in time, depending, on the topic, we experience author and/or version madness; e.g., Gregory changed his CVA formula between 2nd and 3rd edition. The further back in time we go for a reference, the greater are the odds that a book chapter has been switched or issued a new edition in the meantime. For what it's worth!
Last edited:
Was it the same questions where one of the options was that it doesn't reflect any risky activity the bank might take?
June 28,2018
I think so. However I am not sure about exact wording of the answer possibilities...
The wording is important. In general, jumps suggest an volatility smile; this is hopefully intuitive, a greater probability of big move suggests heavier-than-lognormal tails. However, the confusion arises because a binomial (e.g., bimodal) distribution tends to have lighter tails and therefore suggests a frown; this is per Hulls 20.8, but his example is a bimodal distribution (his bimodal example concerns a "single large jump" which is different than a general jump risk model), not the situation when there are jumps in the underlying asset price (a binomial distribution with p in the neighborhood of 0.50 +/- 0.10 has kurtosis < 3).
In this way, "jump risk" suggests a volatility smile (not frown). See https://forum.bionicturtle.com/threads/vol-frown-jumps-in-underlying-asset.10256/#post-59508 but also Hull's EOC question 20.3:
Bottom line (in my humble opinion): "jumps" is far more likely to suggest smile than frown because for a frown (i) the question would need to be explicit about the bimodial/mixture nature of the distribution and (ii) the smile should be more intuitive, anyway, right? Jump risk suggests heavier tails and therefore smile.
Last edited:
Similar threads
