What's new

Exam Feedback FRM Part 1 (May 2014) Exam Feedback

Jo_

Member
Subscriber
Also,....there was a question that spoke about identifying an arbitrage scenario given a set of options with different strikes & premiums,....any idea how this needs to be worked out.

This one caught me off-guard too. My first intuition was put-call parity but that didn't lead me anywhere, so i skipped the question in first instance and reviewed with the timei had left at the end. Then i got the following idea: i calculated what should be the price of the middle call (2nd row), based on the law of one price, same as you would do it for bonds. Outcome was that the weights of option A and C were both 50% and that the "model price" of the option B was 8, which is lower compared to the given price of 8-something. So answer for me was to short 2 times call B ("sell the expensive thing", as David would say), long call A, long call C. I must say i was pretty happy to "discover" how to do it. Now just hoping it's correct :)

I cant remember the analytical formula for Modified Duration and computing it using average weighted maturity would take too much time so I just shock the yield by 0.1% using the calculator and formula deltaP/deltaY to compute the Dollar duration. I think the answer was 173.

To each his own time management and preferences in methods of course, but the weighted average maturity is exactly what i did. It was a three year bond with annual payments (so only 3 cashflows), so I actually don't think your method (though also correct) isn't much faster but is more prone to errors. Just my two cents. Don't remember my answer on this, sorry.


I chose the limit order option.
The reason was more to do with the wording - $5 USD or "better",.....which seems to be in sync with the meaning of a limit order - "the order can be executed only at this price or at one more favourable to the investor"

Same answer and reasoning for me.
 
Last edited:

Roshan Ramdas

Active Member
This one caught me off-guard too. My first intuition was put-call parity but that didn't lead me anywhere, so i skipped the question in first instance and reviewed with the timei had left at the end. Then i got the following idea: i calculated what should be the price of the middle call (2nd row), based on the law of one price, same as you would do it for bonds. Outcome was that the weights of option A and C were both 50% and that the "model price" of the option B was 8, which is lower compared to the given price of 8-something. So answer for me was to short 2 times call B ("sell the expensive thing", as David would say), long call A, long call C. I must say i was pretty happy to "discover" how to do it. Now just hoping it's correct :)

Lucky me,...this option that you selected is precisely what I guessed,...I.e., the butterfly spread.

To each his own time management and preferences in methods of course, but the weighted average maturity is exactly what i did. It was a three year bond with annual payments (so only 3 cashflows), so I actually don't think your method (though also correct) isn't much faster but is more prone to errors. Just my two cents. Don't remember my answer on this, sorry.

Agreed,...this is the method that should have been followed -> calculate Macaulay duration -> and then modified duration....I landed up using the yield shock method and did not get the exact answer :(


Same answer and reasoning for me.
 

Roshan Ramdas

Active Member
These 2 questions I was not able to get with even second attempt.. can someone throw some light

and one other question on finding value a random variable 2 standard deviations away.. i get caught in wording of this question
Sorry dont remember exactly questions now...

Was only able to crack into the 2 standard deviation question on a second attempt.
From what I remember,.....what was provided was the value of 2 securities (think it was 40K & 60K) , their individual expected return, volatilities & correlation.
I landed up -
Calculating the portfolio weighted return, using their current price / value as the weight.
Calculating the portfolio volatility, using the standard formula - SQRT(W1^2(Variance A)+W2^2(Variance B)+2*W1*W2*Covariance(A,B))
Multiply the -> standard deviation * 2 (based on the question)
Grow the portfolio value -> Current value 100K (1+Weighted Return)
And finally
New Portfolio Value + standard deviation * 2 //New Value - standard deviation * 2
 

Babak083

New Member
Subscriber
Hi, candidates who seat also for PART I 2013 Nov. exam, how do you think the difference between Nov 2013 and May 2014. I think this exam was more difficult than November exam.
 

Roshan Ramdas

Active Member
How did zou calculate the bond price where 3 bonds were given and the price of the 2nd was the one they asked for_

The quickest way to do it is by using linear interpolation, this method can only be used if the bonds come with the same maturity date & have the same coupon payment dates.
Another alternative is to use the replicating bond portfolio creation method,....wherein you try and ascertain the face amount that needs to be held in bonds 1 & 3 that will generate cash flows which exactly match that of bond 2. You then multiply the face amounts times the respective bond prices & sum them up. This is a slightly longer but more robust method.

Thank you
 

plulutes

New Member
Hi, candidates who seat also for PART I 2013 Nov. exam, how do you think the difference between Nov 2013 and May 2014. I think this exam was more difficult than November exam.

Babak083, I sat for both part I (3,3,2,2) and part II (not scored). I thought Nov.13 part I was very tricky. Was thinking that the questions would come in the form that most exams come in...meaning, I thought it was going to be a straight application of the formulas and concepts. Was not ready for the twists that GARP has in store for us poor, unsuspecting candidates (hats SERIOUSLY off to those who pass Part I without seen it in person first!). With that said, I thought May 14 Part I was 1.5 times trickier! There were a good amount of questions that I felt good about, but there are just some that I don't think you can really prepare for. So while I can rip off every formula and talk about GARCH (1,1) and negative convexity until I'm blue in the face, It does ya no good when you're not sure of what the question is. It's no shame (IMO) to fail at least once. Personally, I was glad to have failed. If I passed (which I must have been extremely close to doing) I would not have had the benefit of really understanding, to a much greater depth, the material as I do now, regardless of passing the exam...I do understand the material so much better, and that's a small victory in its own right. How did you feel about it?


How did zou calculate the bond price where 3 bonds were given and the price of the 2nd was the one they asked for_

rate.1 (x) + rate.3 (1-x) = rate.2 Solve for x, then x(bond1) + (1-x)bond3 = bond2 price. Just a trick Dave taught us through the reading.
 
Last edited:

Alex_1

Active Member
Hi everybody,

first of all thanks for the detailed feedback here, I planned to write some remarks myself but I am honestly still feeling very tired after the exam. I didn't sleep too well before Saturday and I took both parts on the same day which was (in my opinion) very, very exhausting...

Anyway the exam was pretty tricky, I am really not sure what the outcome will be, in the first two hours I was really "in the zone" and I think I have worked through 52 or 54 questions then I took a break to drink and eat something (which lost me some time as I had to give the papers and leave the room - thanks for this restrictive policy! @"§$%) and then my focus decreased gradually...:mad:

I remember some pretty tough questions (which may not be tough when thinking about them in peace, but in the stress of the exam...):
  • 7-4-3 crack spread - I've even explained a similar question here on the forum -> total blackout in the exam. 30 seconds before the end I vaguely remembered how to solve it but the time was up...
  • The one with Jensen's alpha and several returns, deviations --> I took the way some people here recommended but I didn't get any of the answers...
  • The one with the barbell convexity I lost too much time and wasn't very sure of it, although in hindsight it was a doable question
  • The F-stat questions - not so sure, but to me it appeared significant so I ruled out the two answers which stated "fail to reject", but had to guess on the other two, so it was a coinflip...
  • How did you cope with the question on the computational efficiency? Didn't remember a similar one in the PQs or elsewhere.. Was it something from the simulations reading (Pachamanova)? Or just a more advanced question testing the understanding? I tried to apply logic and first thought it should be something with the least time but then this would have been too easy, so I looked at the standard deviations and tried to look for a combination of "least time" + "lowest vola", but still wasn't sure, what was your take on this one?
  • A lot of questions on the new readings, I am very, very glad I read them all, but still I managed to not solve the APT questions (or at least one of them)...
So, I will be very grateful to pass, but not so sure at the moment, my guess is I will be somewhere between 60 and 75% (that's my optimistic guess), or something like 2,2,2,3...
 

Babak083

New Member
Subscriber
Babak083, I sat for both part I (3,3,2,2) and part II (not scored). I thought Nov.13 part I was very tricky. Was thinking that the questions would come in the form that most exams come in...meaning, I thought it was going to be a straight application of the formulas and concepts. Was not ready for the twists that GARP has in store for us poor, unsuspecting candidates (hats SERIOUSLY off to those who pass Part I without seen it in person first!). With that said, I thought May 14 Part I was 1.5 times trickier! There were a good amount of questions that I felt good about, but there are just some that I don't think you can really prepare for. So while I can rip off every formula and talk about GARCH (1,1) and inverted convexity until I'm blue in the face, It does ya no good when you're not sure of what the question is. It's no shame (IMO) to fail at least once. Personally, I was glad to have failed. If I passed (which I must have been extremely close to doing) I would not have had the benefit of really understanding, to a much greater depth, the material as I do now, regardless of passing the exam...I do understand the material so much better, and that's a small victory in its own right. How did you feel about it?

Plulutes, thank you for so comprehensive answer, Actually I agree with you about failing for the first time, so in this time concept was more clear than first time. In this exam first hour I was headache so I just look through questions, so I actually use 3 hours of time and answer 70 of questions, 20 of questions i choose between 2 answer (hope half of them will be right), and rest of questions just use my chance. How do you think if right answers will 65-70, have I any chance to pass?
 

plulutes

New Member

I've tried to figure out how they score based on lots of assumptions, so who really knows, BUT my guess is that about 55%-60% is about the cutoffs. I'm sure there is a lot of differing opinions on this but I can't imagine that the top 5% of scores average more than 80% - 85% and the quartile begin from there...so I bet Im pretty close. With that said, I think that if you feel like you can reasonably review your performance based upon how many you think you answered correctly, the Percentage you widdled down to 50-50%, and the pure guesses in which you'd expect to get 1/4 right, and you come up with 60%+ (with wiggle room), then I think you've got a fair chance of making it. Also, remember that of the 45% that pass Part I, on average, a good percentage of those that pass are Return testers (like us), making the first time success rate probably closer to 30% (my guess again). Good luck man!

That first exam for me I thought I answered correctly not more than 40% and just missed it...so there ya go. ;)
 

Roshan Ramdas

Active Member
Hi everybody,

first of all thanks for the detailed feedback here, I planned to write some remarks myself but I am honestly still feeling very tired after the exam. I didn't sleep too well before Saturday and I took both parts on the same day which was (in my opinion) very, very exhausting...

Anyway the exam was pretty tricky, I am really not sure what the outcome will be, in the first two hours I was really "in the zone" and I think I have worked through 52 or 54 questions then I took a break to drink and eat something (which lost me some time as I had to give the papers and leave the room - thanks for this restrictive policy! @"§$%) and then my focus decreased gradually...:mad:

I remember some pretty tough questions (which may not be tough when thinking about them in peace, but in the stress of the exam...):
  • 7-4-3 crack spread - I've even explained a similar question here on the forum -> total blackout in the exam. 30 seconds before the end I vaguely remembered how to solve it but the time was up...
  • The one with Jensen's alpha and several returns, deviations --> I took the way some people here recommended but I didn't get any of the answers...
  • The one with the barbell convexity I lost too much time and wasn't very sure of it, although in hindsight it was a doable question
  • The F-stat questions - not so sure, but to me it appeared significant so I ruled out the two answers which stated "fail to reject", but had to guess on the other two, so it was a coinflip...
  • How did you cope with the question on the computational efficiency? Didn't remember a similar one in the PQs or elsewhere.. Was it something from the simulations reading (Pachamanova)? Or just a more advanced question testing the understanding? I tried to apply logic and first thought it should be something with the least time but then this would have been too easy, so I looked at the standard deviations and tried to look for a combination of "least time" + "lowest vola", but still wasn't sure, what was your take on this one?
  • A lot of questions on the new readings, I am very, very glad I read them all, but still I managed to not solve the APT questions (or at least one of them)...
So, I will be very grateful to pass, but not so sure at the moment, my guess is I will be somewhere between 60 and 75% (that's my optimistic guess), or something like 2,2,2,3...
Hello,
Had a similar experience with losing sleep,....a real nightmare especially when you need to take a shot at a 4 hour exam.
Managed to work through close to 90 odd questions,.....and from a time management standpoint, this was better than any of my other mocks......,the 10 that were randomly guessed were not topics that I was familiar with.
Of the 90 questions,....I can recollect "at the least" 10 questions where the answers could be wrong.
The Jensen's alpha was one amongst these,....just didn't get the answer & landed up selecting the closest match. The APT was also time intensive & both these questions were supposed to be areas which I felt, I was good at :). These questions were positioned at the start of the paper,....which raised anxiety levels.
There were two questions which were copy pasted from one of GARP's 2011,12,13 or 14 papers,.....thank them for being merciful :)
Hope all goes well !!

Roshan
 

Jo_

Member
Subscriber
  • 7-4-3 crack spread - I've even explained a similar question here on the forum -> total blackout in the exam. 30 seconds before the end I vaguely remembered how to solve it but the time was up..
  • How did you cope with the question on the computational efficiency? Didn't remember a similar one in the PQs or elsewhere.. Was it something from the simulations reading (Pachamanova)? Or just a more advanced question testing the understanding? I tried to apply logic and first thought it should be something with the least time but then this would have been too easy, so I looked at the standard deviations and tried to look for a combination of "least time" + "lowest vola", but still wasn't sure, what was your take on this one?
If i am correct, there actually was an exercise on this in David's set, which is were i based myself on. Basically, if i remember correctly, you need to chose the one for which SE times sqrt(t) is lowest. Stdev, number of scenarios and total time to run scenarios was given. So i calculated SE for each based on Stdev/sqrt(#scenarios run), then multiplies with sqrt(#time needed to run all scenarios) and took the minimum of that. However, i'm not 100% sure if i needed to use sqr(#time all scenarios) or sqrt(#time one scenario).

And I thought of you for the crack spread question indeed :) However, the one you explained was 1 output from 2 inputs, here it was 1 input for 2 outputs. Despite spotting that difference, i didn't find the correct answer either. I came up with 5, whereas the options were -10 and 10.

On the barbell one, i didn't get that one either. I'm getting the feeling there were some questons which many many candidates won't get :)

Also, remember that of the 45% that pass Part I, on average, a good percentage of those that pass are Return testers (like us), making the first time success rate probably closer to 30% (my guess again).

Never really thought of it that way, but that's absolutely correct! Scares me a little bit, but i remain hopeful :)
 

Floris

New Member
Well this one could have gone better..

I was not able to fill in the bubbles on the answer sheet as time was up and I allowed myself 5 mins. to fill those in, what a disastrous mistake that was of me (..and yes I know I am a complete klutz for choosing this strategy). Chances of me passing this exam are diminishingly low now. Otherwise I would say 50-50 chance of passing as 90 percent of the questions were familiar to me and I should be able to find out the right answer.

The major challenge that I find most frustrating is that a have to find out what the problem is that needs to be solved, especially when questions are accompanied with long texts. This takes up much of the valuable time. Also, I find it very strange that you are not given a scrap paper: why would you not allow that when you have to make so many calculations? This adds a pointless difficulty to your problem-solving capabilities from my perspective.

Ow, and one last thing: next time I will skip binomial option valuation until the very last. The answer was $15 as it was an American option by the way to my knowledge.
 

hoang.vu90

New Member
There is also one question about unlevered/level beta where they increase the debt/equity ratio from 1 to 2 and decrease tax rate from 25% to 15%. Original levered beta is 1.75, unlevered beta is 1. So I plug the number in (1+(1-0.15)*2) and got 2.75..which is none of the answer :( anyone solved this question ?
 

Roshan Ramdas

Active Member
Didn't see any issue with the crack spread.....what Alex had explained on the forum was a 3-2-1 crack spread.What was provided was a 7-4-3 -> that would mean -> Buy 7 barrels of A ( cost) + Sell 4 barrels of B (income) + Sell 3 barrels of C (income).
One additional step needed -> the pnl would be for 7 barrels and you would need to divide by 7 to get it down to pnl per barrel.
 

Roshan Ramdas

Active Member
There is also one question about unlevered/level beta where they increase the debt/equity ratio from 1 to 2 and decrease tax rate from 25% to 15%. Original levered beta is 1.75, unlevered beta is 1. So I plug the number in (1+(1-0.15)*2) and got 2.75..which is none of the answer :( anyone solved this question ?
Are you absolutely certain that you didn't get the answer ??.....I remember doing this problem in the exact same manner and getting a match with one of the options ......the calculation above should give 2.7 and not 2.75
 

amresh

Member
Subscriber
Was only able to crack into the 2 standard deviation question on a second attempt.
From what I remember,.....what was provided was the value of 2 securities (think it was 40K & 60K) , their individual expected return, volatilities & correlation.
I landed up -
Calculating the portfolio weighted return, using their current price / value as the weight.
Calculating the portfolio volatility, using the standard formula - SQRT(W1^2(Variance A)+W2^2(Variance B)+2*W1*W2*Covariance(A,B))
Multiply the -> standard deviation * 2 (based on the question)
Grow the portfolio value -> Current value 100K (1+Weighted Return)
And finally
New Portfolio Value + standard deviation * 2 //New Value - standard deviation * 2
.....

I strongly felt that none of the options were correct, as the return on 40 M portfolio was 8 % and the same on 60 M portfolio was 9 % . Which gives average return of 8.6. Hence at-least one of the options had to have mean return off 100*8.6. i.e. the mid point of the +/- Sigma range. I checked for it but did not get it. And then decided to guess "A" and move ahead.
 
Top