Exam Feedback November 2015 Part 2 FRM Exam Feedback

William_Jose

Member
Subscriber
What is moneyless? Stock price - strike price. strike price is constant. In John the graph is captioned "variation of delta with stock price for a call option and a put option on a non dividend paying stock".
No, on that graph, it shows variation of delta with moneyness,ie., how much "in the money" or "out the money" the option is in. So, the more ITM a call option is, the more positive its delta and more ITM a put is, the more negative its delta. Now as underlying value falls, call option becomes more OTM and its delta falls. But at the same time, put option becomes more ITM and so its delta becomes more negative, hence even the put delta reduces.
 

William_Jose

Member
Subscriber
Mathematically speaking under BSM, Call Delta - 1 = Put Delta. e.g. Call Delta = 1 then Put Delta = 0

So they must increase/decrease at the same time
Yes, that is another way to look at this-the difference between the two deltas is constant, so they have to move in same direction.
 

a.lesnar

Member
I recall few question, anyone can help to provide your thought

1. Ops gross loss, use what value for the counts. I choose replacement cost instead of book value
2. 4 million CVA maximum, i chose position 4.. seems to be easy [might be trapped]
3. change position, transfer USD800000 to other asset, i chose VaR decreasing USD2XXXXX
4. Jeasen inequality convexity, i chose 0.86 basis point

Hello
For 1. Ops gross loss, I also chose replacement, but should the insurance coverage benefits be subtracted from the replacement cost itself?

One doubt for the last question: I assumed net benefits of diversifications decrease as N increases, so I chose option D, where the curve was declining going from the left to the right side of the chart. Anyone can remember the exact question?

There was also a question on the bootstrap vs historical VaR methods, I cannot remember it exactly but could it be a coherent solution to indicate that Bootstrapping leads to higher VaR estimations due to data re-sampling?

Thanks
 

RickyZ

New Member
Your opinion on my conclusion is welcome: GARP messed up in May 2015 making P1 exam narrow i.e. hard and compensated for it by making Nov 2015 easier.

May 2015 Part 1 exam had extremely narrow questions - e.g.: a lot of questions on dollar roll and newly added topics (MBS specifically). I was sure I failed coming out of it yet passed 1,1,2,2. Met one candidate that studied 1 week prior to and passed the P1 (frustrating to hear that). Coming out of Part 1 I thought to myself that all the time I spent studying benefited my ability to answer few questions really (I work in risk management, and I didn't know about bionic turtle). In other words: studying did not help too much if you followed the GARP recommended studying plan - do the readings and good luck when we ask you our questions.

Yesterday however, talking to a few other candidates that didn't study too hard, I got the impression that they had a hard time answering (whereas I also feel like the exam, albeit tricky, was fair and one could answer if one had studied, especially in comparison to May15 Part 1). Could GARP be adjusting? Is it possible that they noticed that statistical distribution of grades for May15 (part 1) were not displaying results reflecting level of preparedness (I'm thinking very low mean of 30-35% - not too far from pure guessing @ 25% - and leptokurtic of course with BT candidates being the fat tail).

End result: GARP decided to attempt to make it more normal explaining the easier exam?
 

William_Jose

Member
Subscriber
Hello


One doubt for the last question: I assumed net benefits of diversifications decrease as N increases, so I chose option D, where the curve was declining going from the left to the right side of the chart. Anyone can remember the exact question?


Thanks
There is a formula for netting factor given in BT as well as Schweser. The (1-netting factor) is the benefit and the plot of this wrt n gives an initially increasing function which then flattens. Here the benefit of diversification is always there as n becomes larger and larger but the rate of increase of this benefit reduces (not that the benefit itself reduces). So choice seemed to be (b).
 

William_Jose

Member
Subscriber
Hello
For 1. Ops gross loss, I also chose replacement, but should the insurance coverage benefits be subtracted from the replacement cost itself?

There was also a question on the bootstrap vs historical VaR methods, I cannot remember it exactly but could it be a coherent solution to indicate that Bootstrapping leads to higher VaR estimations due to data re-sampling?

Thanks
Not sure of these answers. I also chose replacement. Bootstrap leads to more accurate VaR estimations due to resampling, but is it necessarily higher?
 

Twol84

New Member
Subscriber
Premiums should not be included -- they are indirect costs.

Diversification benefit is the graph that increases and plateaus.

I think the answer to the bootstrap question is they are not normal (if that is the same question)

Convexity is 0.86.
 

Sidewinder

Member
Subscriber
Your opinion on my conclusion is welcome: GARP messed up in May 2015 making P1 exam narrow i.e. hard and compensated for it by making Nov 2015 easier.

May 2015 Part 1 exam had extremely narrow questions - e.g.: a lot of questions on dollar roll and newly added topics (MBS specifically). I was sure I failed coming out of it yet passed 1,1,2,2. Met one candidate that studied 1 week prior to and passed the P1 (frustrating to hear that). Coming out of Part 1 I thought to myself that all the time I spent studying benefited my ability to answer few questions really (I work in risk management, and I didn't know about bionic turtle). In other words: studying did not help too much if you followed the GARP recommended studying plan - do the readings and good luck when we ask you our questions.

Yesterday however, talking to a few other candidates that didn't study too hard, I got the impression that they had a hard time answering (whereas I also feel like the exam, albeit tricky, was fair and one could answer if one had studied, especially in comparison to May15 Part 1). Could GARP be adjusting? Is it possible that they noticed that statistical distribution of grades for May15 (part 1) were not displaying results reflecting level of preparedness (I'm thinking very low mean of 30-35% - not too far from pure guessing @ 25% - and leptokurtic of course with BT candidates being the fat tail).

End result: GARP decided to attempt to make it more normal explaining the easier exam?

For me it felt like Part I - only with 20 questions less and hence more time to think on the questions...(and I also think that 20-30% of my learning time was wasted (again)...especially regarding the quantitative formulas)
 

Toy

New Member
Subscriber
I have only ever taken one Garp exam before this: Nov 14 Level 1. When it ended, I thought I lacked enough time and I was surprised I passed very well. Three of my results were in the first quartile of performance.

This time around, I read the books and used BT for revision. I felt there was adequate time for the exam, unlike Level 1. I am not sure the balance of quantitave and qualitative questions matters as Garp never set any expectations on that.

At the end of the day, in these exams, what matters most is one's relative performance to the rest of the cohorts. If the exam is easier for everybody, it will mean that the level required to pass will be set higher. On the other hand, if the exams were very diffcult for most people, then they will lower the bar.

I will rather think that exams are set based on the grasp of the concept rather than on level of preparedness. Of course ability to grasp concept is related to level of preparedness.

This is my first attempt of level 2, so I do not know how to compare the exam to previous ones. I do think the exam reflects the syllabus. I do feel better about this than Nov 14 level 1, but I am by no means certain of success.

I have nevertheless benefited from preparing for the exam: most of the stuff from the Gregory readings are related to my day to day job so going through these materials this year has been very useful. In fact now that I am free I will use my free time to go through the Gregory book at a very slow pace.
 

Attachments

  • IMG_20151122_0003_NEW.pdf
    421 KB · Views: 49

a.lesnar

Member
There is a formula for netting factor given in BT as well as Schweser. The (1-netting factor) is the benefit and the plot of this wrt n gives an initially increasing function which then flattens. Here the benefit of diversification is always there as n becomes larger and larger but the rate of increase of this benefit reduces (not that the benefit itself reduces). So choice seemed to be (b).

yes I understand and I agree on the formula, but the variable on the axis was (1-benefits) as expressed in the problem, that is why I chose answer D (decreasing line) instead of B
 

dundaking

New Member
Subscriber
I came to the same conclusion as a.lesner. I actually circled B first because that was the logical answer but then the question threw me off as it also stated that diversification benefits was defined as 1-netting factor and the netting factor tends towards 1 as n increases so I changed my answer to D.... -_-
 

Jason1991

New Member
I came to the same conclusion as a.lesner. I actually circled B first because that was the logical answer but then the question threw me off as it also stated that diversification benefits was defined as 1-netting factor and the netting factor tends towards 1 as n increases so I changed my answer to D.... -_-
what are the variables for both axis? I chose d as well but answer depends on axis variables
 

asultani

New Member
Subscriber
Y was diversification benefits and I think X axis was N?

I answered D as well, to show that the benefits decrease.
Wasn't there something about the granularity of a portfolio having so many positions which which would reduce the benefits of netting?
Wouldn't having hundreds of minimal size positions make netting not as effective as having your entire portfolio invested in 2 assets?
 

Ranjith21

New Member
Hi, could anyone tell me how the answer for convexity using jensens inequality is arrived at. The 1-yr spot rate now is 5% and 1-yr rate one year from now is expected to be either 4% or 6%. I couldnt get either 0.86 or 1.82. Thanks.
 

Tanya Vasileva

New Member
Subscriber
there was also a question about a fund manager with 4 answers - active management, policy mix a
Hi, could anyone tell me how the answer for convexity using jensens inequality is arrived at. The 1-yr spot rate now is 5% and 1-yr rate one year from now is expected to be either 4% or 6%. I couldnt get either 0.86 or 1.82. Thanks.

here is the example from BT notes/R29.P2.T5.Tuckman_v3
upload_2015-11-23_9-54-27.png
 

Tanya Vasileva

New Member
Subscriber
I've just remembered a question about modeling losses in stress testing - the answer was about LGD
we have around 60 questions revealed so far, can we get more ? :) come on
 

Ranjith21

New Member
There was a question about which of the options would reduce the VaR of a short position in 3x9 FRA.. The options were like decrease/increase in correlation between 3-month and 9-month treasury bills, increase in volatility of 3-month bill. I am not sure what the answer is.
 

Ranjith21

New Member
@tanya: could you tell me the exact answer for the jensens inequality question as I am getting only 0.47bps and not any of the answers even going by the formula..
 
Top