Exam Feedback November 2015 Part 2 FRM Exam Feedback

[Nicole Seaman]

We hope that everything went well for all of you who took the FRM Part 2 Exam today! We would really like to hear your feedback! How did it go? Did you encounter unexpected questions? Let us know how it went!

 

[Jo_]

I found it in line with expectations - maybe even somewhat easier than expected. Questions were generally "fair", i.e. in line with expectations/AIMS and covering the broad spectrum of the curriculum. Good mix of "theoretical" questions and cases (nice twist to add cases and build multiple questions onto them, well done GARP), as well as a good mix from quantitative and qualitative questions. There were only ~5 questions were i was clueless, of which 2/3 in my perception referred more to P1 than to P2 concepts (FRA, delta on calls/puts), but that may be

So overall, i found it a fair and (dare i say) interesting and enjoyable exam. I got about 55 questions i'm pretty confident off, 20 where i was doubting between 2 answers and 5 complete guesses. I hope that will be enough, let's wait and see.

For future candidates reading this later on (i'm sure more people will complement):
- Strikingly much questions on VaR backtesting
- Quite some questions regarding CVA, both qualitatively and quantitatively
- Multiple qualitative questions on wrong-way risk
- Not so much questions on current topics - I only remember one on HFT and cybersecurity
- Investment risk questions quite focused on VaR under different applications, both quantitatively and qualitative (VaR of portfolio before and after rebalancing, marginal var calculations but also qualitatively)
- Some questions on Basel III but none which required to know the LCR and NSFR ratios, weights, haircuts etc by heart
- On the other hand, Basel questions seemed to be more qualitative, i.e. given this context, which actions are most in line with guidelines
- Q on Jenssens inequality - in line with BT practice Q
- 2 quantitative questions on hazard rate / cumulative PD
- No questions on PD / DD using merton model
- nothing on volatility smiles
- one question on netting factor
- Some qualitative questions on VaR mapping

 

[Sidewinder]

For me it is kind of strange / frustrating (same feeling for Part I in May) to learn about 30 slides of formula and at the end you have got primary quantitative questions about VaR/L-VaR/Fund Surplus... Not a single question (!) about Binomial Trees (as far as I remember) or IR-Models (similar topic for me in Part I with the ARIMA-Model).

 

[afterworkguinness]

A reoccurring them was relationships between variables. Very few direct calculation questions, quite a number of qualitative questions. A few sets of questions pertaining to very lengthy setup information.

- Next to nothing on the Tuckman readings from Market risk; which I spent a good month understanding.

 

[Sargera]

There were a couple of esoteric qualitative questions and some ambiguity (at least from my point of view, maybe it was clear from wherever it appears in the study guide) about the sign convention for delta of a put (referring to the question asking for directional movements of option deltas in the case of the spot falling from (you had to compute it) 24 to 18; clearly the atm call delta decreases here, but technically so does the put's delta, yet almost anyone would say it increases as it's magnitude does).

Apart from that, I thought the exam was rather trivial, at least far easier than my expectations based on the comments from the last exam in June.

 

[HERO]

I don't agree with J0_...the exam was not that easy... it was very difficult. Most of the questions were qualitative. The exam was NOT diversified. Many topics were not tested such as Tuchman (interest rate theories), Correlation Modeling, Merton Models...and many.
In some cases, we got 3 question from 1 page such as VAR Backtesting and VAR Analysis (marginal, component..) I can quant abut 7-8 questions about these two points only.
About 50% of the topics were not tested completely. Here is my performance:

35 questions 100% sure.
25 questions between two choices
20 question random guessing...

I am not sure if this will be enough to clear the exam. lets wait and see.

 

[HERO]

There were a couple of esoteric qualitative questions and some ambiguity (at least from my point of view, maybe it was clear from wherever it appears in the study guide) about the sign convention for delta of a put (referring to the question asking for directional movements of option deltas in the case of the spot falling from (you had to compute it) 24 to 18; clearly the atm call delta decreases here, but technically so does the put's delta, yet almost anyone would say it increases as it's magnitude does).

Apart from that, I thought the exam was rather trivial, at least far easier than my expectations based on the comments from the last exam in June.

I think the answer was both deltas will decrease... the call option will be move from ITM to ATM/OTM and the delta will move to zero.. so decrease. the put option will be more in the money and the delta will decrease (more negative)...

 

[Khiem Tran]

I remember there is no directly calculation questions on term structure models

- Some questions on copula (definition/whats wrong with copula model in crisis)
- Questions on calculation payment on three tranches of CLO given CLO rate, overcollaterization amount, senior/mezzaine tranch interest, loan fixed interest, 100 loans 2 loans default
- Spread based on Jensen inequality
- Some questions related to calculated VaR, ES, one question given individuale VaR MaR beta, calculate differences between diversified and undiversified (guys how did you solve this one?)
- Question on neutralized alpha (Which I was so stupid to to it wrong, neutralized alpha = 0 with benchmark method)
- Quesion on mapping VaR (long 3x6 FRA equal to what, long USD/JPY equal to what)
- 3-4 cases study, 2 which are very long, one related to investment strategy buying index/bond/FRA, 1 realted to China back react to devaluation in CNY and changes in their capital condition (my friends told me they were tired from reading and understanding the case only)
- RAROC given orginal RAROC 14%, the elements, RAROC hurdle 9.5%, what cases make RAROC decrease below hurdlge (increase % operating expenses, expected lost, economic capital require). This one I did not have time to compute RAROC each case, just choose economic capital which have biggest dollar value change. How you guys did it?
- Current readings: cyber security which is best practice (report to all BOD SM...), how to decrease "too-big-to fail" central clearing (Doubt between limit members and let members unwind part or total by themselves, how to do post-trade risk control for high frequency trading
- Compare Var of historical simulation and boostrap ( I chose both are nonparametric)
- One related to EVT (tradeoff between high/low threshold)
- One relared to backtesting reject or accept model 95% confidence level (I chose reject as critical value is lower)
- One related to netting benefit (decrease exposure and collateral required)
- One related to use OIS to calulate LIBOR (I regretted doing this wrong)
- Compute hazard rate based on credit spread contiuously for 2 years, RR 30%
- Theory questions related to subordinated debt
- Compute Credit VaR

I will add more if I remember.
Overall there is balance between theory and computation.
Some questions I can easily omit 2 options and doubt between other two which lots I chose wrong
Mixture of simple and complex questions
Cases are very long and take lots of time which made me lack of time for some last questions

This is second time I took, failed last November. I really dont want to take it again considering time dedication but it is worth for understanding.
Just hope for the best now.

Please add more of your opinions guys.

Thanks
-

 

[nyccc1641]

There were a couple of esoteric qualitative questions and some ambiguity (at least from my point of view, maybe it was clear from wherever it appears in the study guide) about the sign convention for delta of a put (referring to the question asking for directional movements of option deltas in the case of the spot falling from (you had to compute it) 24 to 18; clearly the atm call delta decreases here, but technically so does the put's delta, yet almost anyone would say it increases as it's magnitude does).

Apart from that, I thought the exam was rather trivial, at least far easier than my expectations based on the comments from the last exam in June.

yep, if I remember correctly my answer is the call option should decrease and put option should increase.... just intuitively make sense but not sure if i m missed anything...

 

[asultani]

Questions that I remember;
1. Binomial Trees - Current Spot Rate is 5%. If there is 50% chance that the rate increases to 6% or decreases to 4% + OIS. (No clue.)
2. What is the VAR of portfolio: 15 mm GBP with 0.15 STD DEV and 6% mean at a 99% CI.
3. Portfolio of Asset X and Y - What happens to VAR if we rebalance 800k from Y to X.
4. 3 Questions series on a Chinese Bank - Effects of currency devaluation, purchase of a CDS for protection (Wrong-way risk), forgot the third one.
5. You are looking to reduce the VAR of a portfolio, what would you look at first? (Incremental Var, Component, Marginal). I think the answer was to first look at the component to get a general view of the portfolio and then look at the marginal var.
6. Default rate of 0.07 per year, what is the probability of default within 3 years?
7. Given credit Spread and recovery rate - What is the probability of default?
8. A junior risk analyst is writing a report, what would be the MOST appropriate. The answers included something along the lines "If the assumptions are wrong the model is invalid." but I feel that this answer was too obvious to be an appropriate answer.
9. 2-3 questions on FRAs (Completely forgot everything about FRAs from May P1 to November.)
10. Questions on CDO tranches - What happens when 2 loans default in a portfolio (where does the loss gets allocated?)
11. Delta Put Options? (Absolutely NO clue - Another thing from the May P1 that I completely forgot)
12. Alpha, Beta and Gamma firms dealing in Repos/Collateral - What risk is it facing? (I think its Beta facing Funding Liquidity Risk?)
13. Impact of Collateral with specific changes (The word IMPACT was a bit confusing in the context of the question)
14. Threshold and Minimum Margin transfers.
15. Calculation question on VAR, Spread, Spread volatility, k
16. Information Ratio/Tracking Error/Alpha (May P1.. It was in the notes for P2 as well, but I figured what are the odds of getting something so small tested which has been already covered in P1. Boy was I wrong.. There was a question on each of those and this was only 2-3 pages of T4. (Relatively easy points for anyone who took the time to cover these)

I feel like the amount of material that was dedicated to some parts of the the FRM curriculum and what actually showed up in the questions was completely different.
- For example, there was so much reading material for Tuckman yet only 1 question.
- So much detailed material for Basel yet the material covered in the exam was not that much in-depth (You could basically guess some of them.)
- I felt like there was way more then 25% allocated to Credit Risk.. (So many CVA and CVAR questions - In fact most of the calculations questions were related to Credit Risk)

Overall feeling on the exam;
I felt as it was easier than what I went through with P1 in MAY, but I know it was just a FEELING. Qualitative exams like these are trickier than quantitative ones. At least with P1, you had a feeling right away that your answer was either good or wrong by just comparing your calculated answer to the available choices. Not having this reassurance with P2 makes me a bit nervous about the results. I am not sure, but I would think it would be more difficult passing P2 than P1 since they are fewer people attending the exam and grade averages would be higher since its tied to a smaller group that has passed P1.

 

[babyik]

The paper was definitely not easy though it appeared so in patches because of a few sitters. My analysis is 30% sitters, around 30% difficult and around 40% between 2 choices which has been the usual pattern in FRM Part 2.

List of questions i remember [ some may be a repetition of earlier posts] . Have posted the answers in bracket which i have marked. [ Spent too much time on some questions and had to guess around 10+ questions blindly so some of my answers would soud wierd but will help you guys remember the questions .]

1. Expected shortfall- [ 2.2% ]
2. RAROC- which will decrease to less than 9.5%[ increasing the deposit rates by 2% on 650 million as the loan was fully funded by deposits]
3. OIS- 4/4.5% - LIBOR -5% [ 1100 - guess]
4. A graph between 1-netting /positions on Y and Xaxis [ i chose b which was increasing lognormally and then constant]
5.Surplus - Asset-100 r=10% , Liability=80 - cost of funds -8% [ 3.6%- though i feel ans would be somthing else ]
6.How would the amount be distributed -Senior/Equity/Mezannine - [ i chose the one where equity was wiped off ]
7. Reducing moral hazard [let the originator retain some stake ]
8. Portfolio Var 16 m-27%/8m-20% where 8 m is shifted from A to B [ randomly guessed b]
9. Credit Var - expected 28 defaults with 95% confidence and 1000 bonds with expected loss = 2 % [ 8000]
10.Inflow of CAD/Outflow of Euro
11. Reduce the risk of portfolio by [ by taking MVar]
12. Calculate LVar . P= 1.75 m , s=2.5% , meaan =1% etc [ chose b - exogenous + a number]
13.Frequency-Severity [ Poisson/Lognormal]
14. Calculate daily VaR @ 99% , annual volatility given, mean return =6% etc [ chose b ]
15.PD=7% constant. PD in 3 years ? [ 20%]
16. Call Put Delta [ Call delta decrease , put delta increase ]
17. SVaR - increase in capital requirement [ 330 m ]
18. Mapping of fcators - principal + interest etc [ dont remember ]
19. PCA - 90% / 10 positions/ 70%/7% [ dont remeber ]
20. HFT - justification for reducing latency. [ price may move away ]
21. portfolio manager - + ve intercept and higher slope [ statistically not significant and greater beta]
22. 250 days / 95% VaR - 8.4 exceptions [ Reject since statistically not significant - lies outside ]
23. Risk Budgeting- 800 m- A and B IR was given [ chose b ]
24. POT- threshold [ high enough to be significant and low enough to capture enough data points]
25. Backtesting - 12 execptions in 250 days at 95% VaR .. what can we conclude [ dont remember ]
26. Copula [ VaR ]
27. Jensen inequality [ chose b - i think 86bps]
28. Why FI cannot be valued by BSM [ price change in bond ]
29. Internal Control - [ dont remember]
30. Cybersecurity - [ proactively share the data]
31. Model risk
32. RAAF
33. Alpha, Beta , Gamma [ Beta will face funding risk ]
34. China Bank- 3 questions [ dont remember ]
35. CVA [ dont remember ]
36. Netting [ dont remember ]
37. Securitization [ dont remember ]
38. WWR [ dont remember ]
39. Competitor bank increase deposit rates. what if we do not increase [ LCR - unchanged ]
40. Historical simulation- which is not an assumption? [ Normal distribution ]

Will post more when i remember.

My guess of a safe score would be 50/80 considering the fact that 50% of the test takers pass and achieving 60% which is 48 is competitive. What do you guys think?

 

[asultani]

My guess of a safe score would be 50/80 considering the fact that 50% of the test takers pass and achieving 60% which is 48 is competitive. What do you guys think?

If passing grades are based off the best scores then...

If I look at the May 2015 thread and see the number of people that went through the exam with relative easy, I expect that the passing requirements would be higher.

I feel that the pool of candidates attending the Part II exam are going to get much better grades, thus making it more competitive to make the cut for two reasons:
1) The pool of candidates for P1 was also filled with people that were clueless which I assume would make it easier to pass. (We had someone show up for the exam with no calculator, and a few that left after 15 min.)
2) They were in the top 40% in the P1 Exam - I would assume they are more serious, and it could also be people attempting the exam for a second time.

 

[Sidewinder]

39. Competitor bank increase deposit rates. what if we do not increase [ LCR - unchanged ]

Sure? Net Cash Outflow changes

 

[babyik]

yes sidewinder. logically that seems to be correct as people would withdraw their deposits for better rates and had thought about it. However could not find a suitable option on the exam

 

[babyik]

If passing grades are based off the best scores then...

If I look at the May 2015 thread and see the number of people that went through the exam with relative easy, I expect that the passing requirements would be higher.

I feel that the pool of candidates attending the Part II exam are going to get much better grades, thus making it more competitive to make the cut for two reasons:
1) The pool of candidates for P1 was also filled with people that were clueless which I assume would make it easier to pass. (We had someone show up for the exam with no calculator, and a few that left after 15 min.)
2) They were in the top 40% in the P1 Exam - I would assume they are more serious, and it could also be people attempting the exam for a second time.

@ asultani- Lets hope for the best. We were a small group of around 5 people taking the exam and that 50/80 is based on the small sample. Could be we were a bunch of underprepared lot.

 

[William_Jose]

In the question on delta of call and put, I think both should decrease if we consider on basis of pure number rather than magnitude. The call option delta will obviously decrease and the put option delta will become more negative, so even that should decrease. For RAROC will decrease below 9.5%, I found option (a) operating expenses rise by 13%- my calculations showed none of the other options were giving RAROC<9.5% while this was coming to about 9.125%. Some of the other answers posted here are also not matching with what I selected, but I don't think the paper was too simple so as to easily know who could be correct or wrong.Being mostly qualitative, I found L2 tricky even if not tough.

 

[Sidewinder]

@ asultani- Lets hope for the best. We were a small group of around 5 people taking the exam and that 50/80 is based on the small sample. Could be we were a bunch of underprepared lot.

Yes, I think here in this forum you ve got an "opinion bias"

 

[Jo_]

- RAROC given orginal RAROC 14%, the elements, RAROC hurdle 9.5%, what cases make RAROC decrease below hurdlge (increase % operating expenses, expected lost, economic capital require). This one I did not have time to compute RAROC each case, just choose economic capital which have biggest dollar value change. How you guys did it?
-

I did recalculate for all options just to be sure and since i had the time. However as a shortcut you could see that the change in opex had the highest dollar value in the numerator, so you didn't had to calculate for CoF and EL cause the impact was smaller, that already rules out 2 options. Change in opex was also the correct answer.

Sure? Net Cash Outflow changes

Yes, i thought about that too, and you are correct from an LCR denumerator point of view. However, there was no information on where the bank would get the funds to pay out deposits, i.e. what the impact on HQLA is. Sure you can reasonably assume it's from HQLA with 0% haircut (cash, ECB reserves...) but the question said something like "without further information". Furthermore, from experience, an increase in deposit rates from competitors doesn't necessarily imply deposit outflows from your bank, they can be quite price inelastic. So, "without further information", i also went for the LCR unchanged, by elimination of the other options. Not fully confident on this one though, I guess it can be debated.

 

[Sidewinder]

Yes, i thought about that too, and you are correct from an LCR denumerator point of view. However, there was no information on where the bank would get the funds to pay out deposits, i.e. what the impact on HQLA is. Sure you can reasonably assume it's from HQLA with 0% haircut (cash, ECB reserves...) but the question said something like "without further information". Furthermore, from experience, an increase in deposit rates from competitors doesn't necessarily imply deposit outflows from your bank, they can be quite price inelastic. So, "without further information", i also went for the LCR unchanged, by elimination of the other options. Not fully confident on this one though, I guess it can be debated.

I see your point. But who knows what GARP thinks in the background? Maybe they count both answers as correct.

 

[HERO]

The paper was definitely not easy though it appeared so in patches because of a few sitters. My analysis is 30% sitters, around 30% difficult and around 40% between 2 choices which has been the usual pattern in FRM Part 2.

List of questions i remember [ some may be a repetition of earlier posts] . Have posted the answers in bracket which i have marked. [ Spent too much time on some questions and had to guess around 10+ questions blindly so some of my answers would soud wierd but will help you guys remember the questions .]

1. Expected shortfall- [ 2.2% ]
2. RAROC- which will decrease to less than 9.5%[ increasing the deposit rates by 2% on 650 million as the loan was fully funded by deposits]
3. OIS- 4/4.5% - LIBOR -5% [ 1100 - guess]
4. A graph between 1-netting /positions on Y and Xaxis [ i chose b which was increasing lognormally and then constant]
5.Surplus - Asset-100 r=10% , Liability=80 - cost of funds -8% [ 3.6%- though i feel ans would be somthing else ]
6.How would the amount be distributed -Senior/Equity/Mezannine - [ i chose the one where equity was wiped off ]
7. Reducing moral hazard [let the originator retain some stake ]
8. Portfolio Var 16 m-27%/8m-20% where 8 m is shifted from A to B [ randomly guessed b]
9. Credit Var - expected 28 defaults with 95% confidence and 1000 bonds with expected loss = 2 % [ 8000]
10.Inflow of CAD/Outflow of Euro
11. Reduce the risk of portfolio by [ by taking MVar]
12. Calculate LVar . P= 1.75 m , s=2.5% , meaan =1% etc [ chose b - exogenous + a number]
13.Frequency-Severity [ Poisson/Lognormal]
14. Calculate daily VaR @ 99% , annual volatility given, mean return =6% etc [ chose b ]
15.PD=7% constant. PD in 3 years ? [ 20%]
16. Call Put Delta [ Call delta decrease , put delta increase ]
17. SVaR - increase in capital requirement [ 330 m ]
18. Mapping of fcators - principal + interest etc [ dont remember ]
19. PCA - 90% / 10 positions/ 70%/7% [ dont remeber ]
20. HFT - justification for reducing latency. [ price may move away ]
21. portfolio manager - + ve intercept and higher slope [ statistically not significant and greater beta]
22. 250 days / 95% VaR - 8.4 exceptions [ Reject since statistically not significant - lies outside ]
23. Risk Budgeting- 800 m- A and B IR was given [ chose b ]
24. POT- threshold [ high enough to be significant and low enough to capture enough data points]
25. Backtesting - 12 execptions in 250 days at 95% VaR .. what can we conclude [ dont remember ]
26. Copula [ VaR ]
27. Jensen inequality [ chose b - i think 86bps]
28. Why FI cannot be valued by BSM [ price change in bond ]
29. Internal Control - [ dont remember]
30. Cybersecurity - [ proactively share the data]
31. Model risk
32. RAAF
33. Alpha, Beta , Gamma [ Beta will face funding risk ]
34. China Bank- 3 questions [ dont remember ]
35. CVA [ dont remember ]
36. Netting [ dont remember ]
37. Securitization [ dont remember ]
38. WWR [ dont remember ]
39. Competitor bank increase deposit rates. what if we do not increase [ LCR - unchanged ]
40. Historical simulation- which is not an assumption? [ Normal distribution ]

Will post more when i remember.

My guess of a safe score would be 50/80 considering the fact that 50% of the test takers pass and achieving 60% which is 48 is competitive. What do you guys think?

1. I agree with many of the answers, however, for RAROC, the right answer was increasing operating expenses by 50%... this is about 6.5mn, while deposits will affect by 3.25.
2. You should first look at Marginal VaR - not component VaR.. it is clear written the material. Remember, that to decrease the portfolio VaR you should shift from high MVaR to Low MVaR

I believe, that if you score 50+.. you will pass. Most candidates will get between 45-55. However, I can count 40 question... as easy points... so you need only 10 more.