What's new

Exam Feedback May 2021 Part 2 Exam Feedback

You’re correct
Brain just stuck have no idea how to calculate wcl of that question in the exam

ah
Also there’s one question which they have given the buy cost and sale cost and the risk factor something like that, and asked should the investor/ bank sell the equity/bond or increase the holding of it
Anyone know the answer? I have no idea how to do this as well
was it for MCVAR of around 0.06 , which was between -SC and PC ... i went with last option ...dont do anything ..its optimal
 

giuseppe.dattis

New Member
Dear all, first of all I hope that, apart from different views on the exam you could finally feel relaxed as I'm right now. I'm sure that the toughest part of the exam was the psychological pressure we all faced during the last year with the several postponements of the exam and we should be glad that is finally over. However, I would like to express my opinion too on the exam. It is true and objective that the exam questions were skewed on the more qualitative side and I feel disappointed like you at first but then I realized: "Hell! This is not a memory contest, I don't want to be negatively judged because I cannot learn by heart formulae with 2349498121267 characters! I want to compete on my ability of understanding complex situations and deriving the correct conclusions. A risk manager/me/we don't need to know the Black-Scholes formula hidden in a Python library, I need to know what would happen to my hedging strategy if volatility was to increase. If one of you was to get that fantastic job position in my place, it should have been for merits and not for memory capacity! This was definitely a FAIR EXAM!

That said, below the questions I remember if it could be of any use:

1) Hurdle rate: my answer was 6.5%, definitely easier than RAROC, don't complain please!
2) Asset allocation vs Security selection: security selection was the major component of excess return, no need for formula, just a look at the excess return in fixed income, roughly 6% ( 14-8 ) diff in return times 60% allocation made up 3,6% of a total strategy excess return of 5.1%
3) Default correlation: interesting question, I put CDS spread decrease if default correlation between CRP/ref asset increase. You should pay less ( spread ) because the protection is weak!
4) AML: I put clustering, only thing learnt by heart :D
5) Cybersecurity: I put overlapping functions, in the books it is referenced as "redundancy"
6) ES 97.5%: straightforward, simple average of the last 4 buckets
7) ULC: I found the formula on the internet, put 140K answer B if I'm not wrong. If you want to check Ro = 0.14 UL(a) = 327 UL(b) = 990 UL(ptf) = 1085
8) Climate risk: I put the answer including insurance penetration in equity valuations, haven't checked if right or wrong.
9) SOFR: I put that SOFR is more volatile due to collateral market
10) Time discrete default rate: solved N°issuer(end)/N°issuer(begin) = ( 1 + DEFRATE ) ^ (3 or 4 don't recall)
11) Securitization: predatory lending
12) 3-years ZC price: start rate 2%, risk prem = 10 bps/years VoL = 40 bps/year. Don't recall the exact number
13) GEV 1: don't remember the question
14) GEV 2: don't remember the question
15) 3-years survival prob: easy exp(-Hazardrate*3)
16) Long volatility: buy correlation swap
17) Op Risk losses: losses should be considered gross, no opportunity cost, no enhancement, no provision straight after filling law suite
18) FRTB: don't recall the question but answer was something like 1.5 times something else
19) Liquidation cost: BID ASK spread / 2 * N° shares
20) Information Ratio: data were purchase cost, sale cost, active risk, alpha but definitely double guessed
21) CRO: escalation step in contingent plan
22) Liquidity buffer: the were contingent liquidity, operation liquidity ( maybe 500? ), stressed inflow and outflow 100 and 300 respectively but can't remember the answer. Also here double guessed
23) Hedge Fund perf. 1985 - 1995: I put backfilling bias as the reason for exceptional performance
24) Credit Var: don't recall the numbers but approach was the difference between the price in one year and the stressed one
25) Lower node Model 1: don't remember the values
26) OTM option price: I think I put higher price
27) Var 99% backtest: I put t-statistic ( roughly 2.7 ) higher then p-value ( 2.33 ) hence reject
28) Regression of excess returns: options were regression on benchmark, peers, something else. I've chosen the answer with alpha if I recall correctly
29) Log return VAR: dived the annual mean by 252 and volatility used the square root rule annual -> daily
30) COVID-19: spot oil prices move down the futures curve
31) Repo negotiations: the party most concerned with the shortening/lenghtening of the contract. Don't remember
32) Joint surv. prob: started from the equation of the correlation = (PROBab - PROBa*PROBb)/(PROBa*(1-PROBa)*PROBb*(1-PROBb))^(1/2) but I don't remember the answer

This is everything I remember, hope it might be of any help!
Giuseppe
 

VegaHedge

New Member
Dear all, first of all I hope that, apart from different views on the exam you could finally feel relaxed as I'm right now. I'm sure that the toughest part of the exam was the psychological pressure we all faced during the last year with the several postponements of the exam and we should be glad that is finally over. However, I would like to express my opinion too on the exam. It is true and objective that the exam questions were skewed on the more qualitative side and I feel disappointed like you at first but then I realized: "Hell! This is not a memory contest, I don't want to be negatively judged because I cannot learn by heart formulae with 2349498121267 characters! I want to compete on my ability of understanding complex situations and deriving the correct conclusions. A risk manager/me/we don't need to know the Black-Scholes formula hidden in a Python library, I need to know what would happen to my hedging strategy if volatility was to increase. If one of you was to get that fantastic job position in my place, it should have been for merits and not for memory capacity! This was definitely a FAIR EXAM!

That said, below the questions I remember if it could be of any use:

1) Hurdle rate: my answer was 6.5%, definitely easier than RAROC, don't complain please!
2) Asset allocation vs Security selection: security selection was the major component of excess return, no need for formula, just a look at the excess return in fixed income, roughly 6% ( 14-8 ) diff in return times 60% allocation made up 3,6% of a total strategy excess return of 5.1%
3) Default correlation: interesting question, I put CDS spread decrease if default correlation between CRP/ref asset increase. You should pay less ( spread ) because the protection is weak!
4) AML: I put clustering, only thing learnt by heart :D
5) Cybersecurity: I put overlapping functions, in the books it is referenced as "redundancy"
6) ES 97.5%: straightforward, simple average of the last 4 buckets
7) ULC: I found the formula on the internet, put 140K answer B if I'm not wrong. If you want to check Ro = 0.14 UL(a) = 327 UL(b) = 990 UL(ptf) = 1085
8) Climate risk: I put the answer including insurance penetration in equity valuations, haven't checked if right or wrong.
9) SOFR: I put that SOFR is more volatile due to collateral market
10) Time discrete default rate: solved N°issuer(end)/N°issuer(begin) = ( 1 + DEFRATE ) ^ (3 or 4 don't recall)
11) Securitization: predatory lending
12) 3-years ZC price: start rate 2%, risk prem = 10 bps/years VoL = 40 bps/year. Don't recall the exact number
13) GEV 1: don't remember the question
14) GEV 2: don't remember the question
15) 3-years survival prob: easy exp(-Hazardrate*3)
16) Long volatility: buy correlation swap
17) Op Risk losses: losses should be considered gross, no opportunity cost, no enhancement, no provision straight after filling law suite
18) FRTB: don't recall the question but answer was something like 1.5 times something else
19) Liquidation cost: BID ASK spread / 2 * N° shares
20) Information Ratio: data were purchase cost, sale cost, active risk, alpha but definitely double guessed
21) CRO: escalation step in contingent plan
22) Liquidity buffer: the were contingent liquidity, operation liquidity ( maybe 500? ), stressed inflow and outflow 100 and 300 respectively but can't remember the answer. Also here double guessed
23) Hedge Fund perf. 1985 - 1995: I put backfilling bias as the reason for exceptional performance
24) Credit Var: don't recall the numbers but approach was the difference between the price in one year and the stressed one
25) Lower node Model 1: don't remember the values
26) OTM option price: I think I put higher price
27) Var 99% backtest: I put t-statistic ( roughly 2.7 ) higher then p-value ( 2.33 ) hence reject
28) Regression of excess returns: options were regression on benchmark, peers, something else. I've chosen the answer with alpha if I recall correctly
29) Log return VAR: dived the annual mean by 252 and volatility used the square root rule annual -> daily
30) COVID-19: spot oil prices move down the futures curve
31) Repo negotiations: the party most concerned with the shortening/lenghtening of the contract. Don't remember
32) Joint surv. prob: started from the equation of the correlation = (PROBab - PROBa*PROBb)/(PROBa*(1-PROBa)*PROBb*(1-PROBb))^(1/2) but I don't remember the answer

This is everything I remember, hope it might be of any help!
Giuseppe
Regardin 32) The question was what is the probability that neither of the bonds defaults. This implies, one needs to calculate 1-[- joint default correlation + default(a) + default(b) (in set theory, it is lke saying not (A or B)...
 
Last edited:
Regardin 32) The question was what is the probability that neither of the bonds defaults. This implies, one needs to calculate 1-[- joint default correlation + default(a) + default(b) (in set theory, it is lke saying not (A or B)...
I too did something like 100-(4+9-3.5)
 

adeex18

New Member
14: I think it was friction between originator and assigner. I chose adverse as originator has more knowledge on the mortgager than the assigner
Dear all, first of all I hope that, apart from different views on the exam you could finally feel relaxed as I'm right now. I'm sure that the toughest part of the exam was the psychological pressure we all faced during the last year with the several postponements of the exam and we should be glad that is finally over. However, I would like to express my opinion too on the exam. It is true and objective that the exam questions were skewed on the more qualitative side and I feel disappointed like you at first but then I realized: "Hell! This is not a memory contest, I don't want to be negatively judged because I cannot learn by heart formulae with 2349498121267 characters! I want to compete on my ability of understanding complex situations and deriving the correct conclusions. A risk manager/me/we don't need to know the Black-Scholes formula hidden in a Python library, I need to know what would happen to my hedging strategy if volatility was to increase. If one of you was to get that fantastic job position in my place, it should have been for merits and not for memory capacity! This was definitely a FAIR EXAM!

That said, below the questions I remember if it could be of any use:

1) Hurdle rate: my answer was 6.5%, definitely easier than RAROC, don't complain please!
2) Asset allocation vs Security selection: security selection was the major component of excess return, no need for formula, just a look at the excess return in fixed income, roughly 6% ( 14-8 ) diff in return times 60% allocation made up 3,6% of a total strategy excess return of 5.1%
3) Default correlation: interesting question, I put CDS spread decrease if default correlation between CRP/ref asset increase. You should pay less ( spread ) because the protection is weak!
4) AML: I put clustering, only thing learnt by heart :D
5) Cybersecurity: I put overlapping functions, in the books it is referenced as "redundancy"
6) ES 97.5%: straightforward, simple average of the last 4 buckets
7) ULC: I found the formula on the internet, put 140K answer B if I'm not wrong. If you want to check Ro = 0.14 UL(a) = 327 UL(b) = 990 UL(ptf) = 1085
8) Climate risk: I put the answer including insurance penetration in equity valuations, haven't checked if right or wrong.
9) SOFR: I put that SOFR is more volatile due to collateral market
10) Time discrete default rate: solved N°issuer(end)/N°issuer(begin) = ( 1 + DEFRATE ) ^ (3 or 4 don't recall)
11) Securitization: predatory lending
12) 3-years ZC price: start rate 2%, risk prem = 10 bps/years VoL = 40 bps/year. Don't recall the exact number
13) GEV 1: don't remember the question
14) GEV 2: don't remember the question
15) 3-years survival prob: easy exp(-Hazardrate*3)
16) Long volatility: buy correlation swap
17) Op Risk losses: losses should be considered gross, no opportunity cost, no enhancement, no provision straight after filling law suite
18) FRTB: don't recall the question but answer was something like 1.5 times something else
19) Liquidation cost: BID ASK spread / 2 * N° shares
20) Information Ratio: data were purchase cost, sale cost, active risk, alpha but definitely double guessed
21) CRO: escalation step in contingent plan
22) Liquidity buffer: the were contingent liquidity, operation liquidity ( maybe 500? ), stressed inflow and outflow 100 and 300 respectively but can't remember the answer. Also here double guessed
23) Hedge Fund perf. 1985 - 1995: I put backfilling bias as the reason for exceptional performance
24) Credit Var: don't recall the numbers but approach was the difference between the price in one year and the stressed one
25) Lower node Model 1: don't remember the values
26) OTM option price: I think I put higher price
27) Var 99% backtest: I put t-statistic ( roughly 2.7 ) higher then p-value ( 2.33 ) hence reject
28) Regression of excess returns: options were regression on benchmark, peers, something else. I've chosen the answer with alpha if I recall correctly
29) Log return VAR: dived the annual mean by 252 and volatility used the square root rule annual -> daily
30) COVID-19: spot oil prices move down the futures curve
31) Repo negotiations: the party most concerned with the shortening/lenghtening of the contract. Don't remember
32) Joint surv. prob: started from the equation of the correlation = (PROBab - PROBa*PROBb)/(PROBa*(1-PROBa)*PROBb*(1-PROBb))^(1/2) but I don't remember the answer

This is everything I remember, hope it might be of any help!
Giuseppe
The SOFR is this the one about disadvantage relative to LIBOR? I believe the answer for that was SOFR does not incorporate bank credit risk
 

badbunny

Member
Subscriber
Regardin 32) The question was what is the probability that neither of the bonds defaults. This implies, one needs to calculate 1-[- joint default correlation + default(a) + default(b) (in set theory, it is lke saying not (A or B)...
Dear all, first of all I hope that, apart from different views on the exam you could finally feel relaxed as I'm right now. I'm sure that the toughest part of the exam was the psychological pressure we all faced during the last year with the several postponements of the exam and we should be glad that is finally over. However, I would like to express my opinion too on the exam. It is true and objective that the exam questions were skewed on the more qualitative side and I feel disappointed like you at first but then I realized: "Hell! This is not a memory contest, I don't want to be negatively judged because I cannot learn by heart formulae with 2349498121267 characters! I want to compete on my ability of understanding complex situations and deriving the correct conclusions. A risk manager/me/we don't need to know the Black-Scholes formula hidden in a Python library, I need to know what would happen to my hedging strategy if volatility was to increase. If one of you was to get that fantastic job position in my place, it should have been for merits and not for memory capacity! This was definitely a FAIR EXAM!

That said, below the questions I remember if it could be of any use:

1) Hurdle rate: my answer was 6.5%, definitely easier than RAROC, don't complain please!
2) Asset allocation vs Security selection: security selection was the major component of excess return, no need for formula, just a look at the excess return in fixed income, roughly 6% ( 14-8 ) diff in return times 60% allocation made up 3,6% of a total strategy excess return of 5.1%
3) Default correlation: interesting question, I put CDS spread decrease if default correlation between CRP/ref asset increase. You should pay less ( spread ) because the protection is weak!
4) AML: I put clustering, only thing learnt by heart :D
5) Cybersecurity: I put overlapping functions, in the books it is referenced as "redundancy"
6) ES 97.5%: straightforward, simple average of the last 4 buckets
7) ULC: I found the formula on the internet, put 140K answer B if I'm not wrong. If you want to check Ro = 0.14 UL(a) = 327 UL(b) = 990 UL(ptf) = 1085
8) Climate risk: I put the answer including insurance penetration in equity valuations, haven't checked if right or wrong.
9) SOFR: I put that SOFR is more volatile due to collateral market
10) Time discrete default rate: solved N°issuer(end)/N°issuer(begin) = ( 1 + DEFRATE ) ^ (3 or 4 don't recall)
11) Securitization: predatory lending
12) 3-years ZC price: start rate 2%, risk prem = 10 bps/years VoL = 40 bps/year. Don't recall the exact number
13) GEV 1: don't remember the question
14) GEV 2: don't remember the question
15) 3-years survival prob: easy exp(-Hazardrate*3)
16) Long volatility: buy correlation swap
17) Op Risk losses: losses should be considered gross, no opportunity cost, no enhancement, no provision straight after filling law suite
18) FRTB: don't recall the question but answer was something like 1.5 times something else
19) Liquidation cost: BID ASK spread / 2 * N° shares
20) Information Ratio: data were purchase cost, sale cost, active risk, alpha but definitely double guessed
21) CRO: escalation step in contingent plan
22) Liquidity buffer: the were contingent liquidity, operation liquidity ( maybe 500? ), stressed inflow and outflow 100 and 300 respectively but can't remember the answer. Also here double guessed
23) Hedge Fund perf. 1985 - 1995: I put backfilling bias as the reason for exceptional performance
24) Credit Var: don't recall the numbers but approach was the difference between the price in one year and the stressed one
25) Lower node Model 1: don't remember the values
26) OTM option price: I think I put higher price
27) Var 99% backtest: I put t-statistic ( roughly 2.7 ) higher then p-value ( 2.33 ) hence reject
28) Regression of excess returns: options were regression on benchmark, peers, something else. I've chosen the answer with alpha if I recall correctly
29) Log return VAR: dived the annual mean by 252 and volatility used the square root rule annual -> daily
30) COVID-19: spot oil prices move down the futures curve
31) Repo negotiations: the party most concerned with the shortening/lenghtening of the contract. Don't remember
32) Joint surv. prob: started from the equation of the correlation = (PROBab - PROBa*PROBb)/(PROBa*(1-PROBa)*PROBb*(1-PROBb))^(1/2) but I don't remember the answer

This is everything I remember, hope it might be of any help!
Giuseppe

Correct, Probability of [1- Probability (A or B)] is the same as Probability of (not A and not B). So either way was correct.

P(A or B) = 1 - P(not A and not B)
 

sajedian

New Member
Subscriber
I am working in Risk Analytics of institutional asset management with exposure to all the types of risks except for ORR. If FRM is meant to test and deliver practical experience, I should say GARP is a huge fauilre in the way they developed exam part 2. They tested something for about 70% of the exam that may be less than 10% of risk professionals are actually working in that area. It was really disappointing...
 

PaulTomlin

New Member
Subscriber
Overall I was disappointed with the huge difference between the practice exam and the actual exam, the latter being way, way harder. If it comes to a retake in December, at least I'll know the level I have to prepare to now. It would have been better to have some idea about this before the exam - not sure if others knew what to expect?

I also thought the inclusion of so many difficult qualitative questions could lead to a level of guessing sufficient to introduce randomness to the exam outcomes.

Hoping for better next time around.
 

Eustice_Langham

Active Member
Overall I was disappointed with the huge difference between the practice exam and the actual exam, the latter being way, way harder. If it comes to a retake in December, at least I'll know the level I have to prepare to now. It would have been better to have some idea about this before the exam - not sure if others knew what to expect?

I also thought the inclusion of so many difficult qualitative questions could lead to a level of guessing sufficient to introduce randomness to the exam outcomes.

Hoping for better next time around.
Thats the issue though...you wont be any better prepared should you need to take the exam in December as the exam will (probably) simply change again.
 
Is there a reason why you think GARP would post the results weeks in advance of their “mid July” estimate as indicated on their website ?

Would be great of course !
When I took part 1 in 2019, they also said mid-July but exam date was May 18th and results came on June 27th. My friend took the October 24th exam in 2020 and got the results on December 4th.

so I’m just being hopeful they continue with the ~6 week time frame I’ve seen.
 

Nicole Seaman

Director of FRM Operations
Staff member
Subscriber
@Nicole Seaman any word from GARP on when exam results will be released? Based on 2019, I would almost expect it tomorrow.
GARP states that they release results around 8 weeks after the exam. Usually, they will tell candidates a tentative date (as @mariomansour stated above). It used to also be on their website under the FRM information menu. If I see a specific date annouced by GARP, I will post it here.
 
Top