FRM MAY PART 1 2013 Feedback

[Milan K]

I think BSM can be used for both Eur and Ame calls, but not Ame Puts. American calls shudnt be exercised early and puts shud be exercised early and because of this, put prices cannot be calculated using BSM.

There was a question asking about the payoff of a fwd contract in CAD: I think the answer was D something which said: 1,000,000 - k*U)(*Y(*)*()....

 

[Backwardation]

I got a 17500 answer too. but for question 41 I think the answer should be long struddle because the question talk about high change in volatility

Maybe we are referring to different questions, but my question was about decreasing volatility.

A Long straddle and a short calendar spread have very similar payoffs. Hence, if they both are listed in a question and you have to Chose one of them, it can be concluded than None of them is correct

 

[nabil1234]

Maybe we are referring to different questions, but my question was about decreasing volatility.

A Long straddle and a short calendar spread have very similar payoffs. Hence, if they both are listed in a question and you have to Chose one of them, it can be concluded than None of them is correct

I remeber that question was about big change in price so there is a high volatility, so the only choice deal with high volatility is long straddle, this is what I studied in Schweser notes book 3, and I am not able to determine if the struddle and calendar have very similar payoffs.

 

[Backwardation]

ok, then we are talking about different questions because my question was talking about "decreasing volatility"

 

[nabil1234]

ok, then we are talking about different questions because my question was talking about "decreasing volatility"

really I think there was a one question, so either me or u misunderstand the question

 

[Milan K]

ok, then we are talking about different questions because my question was talking about "decreasing volatility"

decreasing volatility was for the question where the analyst felt that the hedge ratio was overestimated. Q was what cud b the reason for that. Answer is: decrease of spot price volatility.....

 

[nabil1234]

decreasing volatility was for the question where the analyst felt that the hedge ratio was overestimated. Q was what cud b the reason for that. Answer is: decrease of spot price volatility.....

dear Milan, is that mean there was a two question and my answer "long struddle" was correct?

 

[nabil1234]

what answer of this question : Futures margin question about purchasing 100 tons with initial margin at 7500 and variation margin at 6200 ?
for me the answer was something like $17700

 

[Uchica__Itachi]

Hi All

Just to touch on a few points that have been made. It is possible that I did put in a few questions from FRM part 2 because I took that exam as well.

Question about volatility - A long straddle doesnt have the same payoff as a short calendar spread. Also, I'm 100% certain the question said the trader thought there would be little volatility (this is one of the only questions I had to re-read a few times to make sure I was with it AND its one of the only ones in the whole exam where I thought the answer should have been obvious hence I remember it like it quite well). Having said that, just like in the exam, Long straddle was the very first asnwer I eliminated because that is a bet on high vol. If you thought vol would be low you would short a straddle.

Question about margin - I think the main point here is that the margin account should have been topped up the day before the rise in price which would mean it goes to 18800 (difference in our answers is the difference btwn init n variation margin). Did your calc show that margin acc didnt go below variation the day before final price increase? If it did then maybe you right. My brain was fried during most of this exam so iits possible a few mistakes were made in simple calcs.

Question with 10 & 20 hrs training - I put there was heteroskedasticy because of difference in st dev of distributions.
Question about hybrid approach - I read that hybrid approach is combination of historical sim n exponentially weighted scheme which is similar to Riskmetrics and GARCH so my assumption is that the answer for that question is GARCH. May be wrong.

Question about BSM - The issue about using BSM is one where you compare the American put and call. This was on a non-dividend paying asset so its never optimal to exercise the call but it may be optimal to exercise the put. As a result its less of an issue to use BSM for put. This is one of the few questions where I'm 100% certain of the answer.

Question about overestimated hedge ratio - I also put it was due to decrease in spot vol.

10) for the trader worried about his position and willing to sell off if unemployment rate is higher than expected - I re-re-read the question and the word 'immediately' stood out. So, may be as soon as the numbers come, he would like to liquidate his position immediately and hence would place a market order. - In this case i would have to agree with you! *SIGH* Why do they test us on such specific stuff when our brains are fried!!

Question about Libor - You may be correct on that one. I assumed the crux of the question was more on the fact that he was bent on using LIBOR and wasnt going to consider EIONA for the european stock.

Question about stress test - I think they were saying which one didnt incorporate correlations!

Oh well. I guess we'll have to wait and see how we do. Good luck to you all.

 

[nabil1234]

Hi All

Just to touch on a few points that have been made. It is possible that I did put in a few questions from FRM part 2 because I took that exam as well.

Question about volatility - A long straddle doesnt have the same payoff as a short calendar spread. Also, I'm 100% certain the question said the trader thought there would be little volatility (this is one of the only questions I had to re-read a few times to make sure I was with it AND its one of the only ones in the whole exam where I thought the answer should have been obvious hence I remember it like it quite well). Having said that, just like in the exam, Long straddle was the very first asnwer I eliminated because that is a bet on high vol. If you thought vol would be low you would short a straddle.

Question about margin - I think the main point here is that the margin account should have been topped up the day before the rise in price which would mean it goes to 18800 (difference in our answers is the difference btwn init n variation margin). Did your calc show that margin acc didnt go below variation the day before final price increase? If it did then maybe you right. My brain was fried during most of this exam so iits possible a few mistakes were made in simple calcs.

Question with 10 & 20 hrs training - I put there was heteroskedasticy because of difference in st dev of distributions.
Question about hybrid approach - I read that hybrid approach is combination of historical sim n exponentially weighted scheme which is similar to Riskmetrics and GARCH so my assumption is that the answer for that question is GARCH. May be wrong.

Question about BSM - The issue about using BSM is one where you compare the American put and call. This was on a non-dividend paying asset so its never optimal to exercise the call but it may be optimal to exercise the put. As a result its less of an issue to use BSM for put. This is one of the few questions where I'm 100% certain of the answer.

Question about overestimated hedge ratio - I also put it was due to decrease in spot vol.

10) for the trader worried about his position and willing to sell off if unemployment rate is higher than expected - I re-re-read the question and the word 'immediately' stood out. So, may be as soon as the numbers come, he would like to liquidate his position immediately and hence would place a market order. - In this case i would have to agree with you! *SIGH* Why do they test us on such specific stuff when our brains are fried!!

Question about Libor - You may be correct on that one. I assumed the crux of the question was more on the fact that he was bent on using LIBOR and wasnt going to consider EIONA for the european stock.

Question about stress test - I think they were saying which one didnt incorporate correlations!

Oh well. I guess we'll have to wait and see how we do. Good luck to you all.

about volatility question I understood that the question was talking about high volatility because it was as I remeber about high decreasing in price....Mr Milan above agree with me.

 

[Backwardation]

hi uchica... just a few remarks on your post. But then I will stop discussing the questions.. that's just making me less confident

Hi All

Just to touch on a few points that have been made. It is possible that I did put in a few questions from FRM part 2 because I took that exam as well.

Question about volatility - A long straddle doesnt have the same payoff as a short calendar spread. Also, I'm 100% certain the question said the trader thought there would be little volatility (this is one of the only questions I had to re-read a few times to make sure I was with it AND its one of the only ones in the whole exam where I thought the answer should have been obvious hence I remember it like it quite well). Having said that, just like in the exam, Long straddle was the very first asnwer I eliminated because that is a bet on high vol. If you thought vol would be low you would short a straddle.

If you look at the payoff diagram of a short calendar spread and a Long straddle you will see that there are quite similar, despite the fact that the calendar spread is not symmetric. But in fact both (Long straddle and short calendar) are bets on high vola. So we basically agree on this Point. Like you, I'm also certain that low volatility was expected in this question and therefore the Long Butterfly spread is the only choice.

Question about margin - I think the main point here is that the margin account should have been topped up the day before the rise in price which would mean it goes to 18800 (difference in our answers is the difference btwn init n variation margin). Did your calc show that margin acc didnt go below variation the day before final price increase? If it did then maybe you right. My brain was fried during most of this exam so iits possible a few mistakes were made in simple calcs.

I'm not so sure about this question... possible too that I have made a mistake...


10) for the trader worried about his position and willing to sell off if unemployment rate is higher than expected - I re-re-read the question and the word 'immediately' stood out. So, may be as soon as the numbers come, he would like to liquidate his position immediately and hence would place a market order. - In this case i would have to agree with you! *SIGH* Why do they test us on such specific stuff when our brains are fried!!
this question was not very clear. The market order also made sense to me.


Oh well. I guess we'll have to wait and see how we do. Good luck to you all.

 

[Uchica__Itachi]

hi uchica... just a few remarks on your post. But then I will stop discussing the questions.. that's just making me less confident

Backwardation - I agree about the option one. I was mixing up short calendar spread with long Butterfly profile (remnants of the Saturday's fried-brain-syndrome maybe?).

I'm interested in seeing if other guys can remember a few other questions. Does anyone remember if there was a question on converting an int rate to semi-annual compounding? Also the two Bin tree questions. One where they asked for the prob of dwn move in risk neutral setting and the other, a 2 step calc for price given risk neutral tree.

 

[Spronkworks]

hi uchica... just a few remarks on your post. But then I will stop discussing the questions.. that's just making me less confident

To add to the confusion, a short calendar spread position benefits from a decline in implied volatility...

 

[Spronkworks]

@Uchica_Itachi: Thank you for that summary. First of all, it seems that we had slightly different exams, because some questions you are referring to, were not present in the exam which I did. (I am quite sure, because there are several questions which sound new to me)

Maybe you could argue on the following Points:

Question 4: my result was 17500 and I would wonder if I made a mistake in this straight forward calculation...
Question 30: I think the correct answer would have been the one with the subadditivity because the formula stated VAR(PF) <= VAR (A) + VAR (B). I would assume, that this also holds even if VaR is not aubbadditive because it does not say < but <=, which should also hold for Portfolios which are not subadditive.
Question 41: I think your choice (short calendar spread) is not the correct answer because a calendar spread has a payoff similar to a butterfly spread. Hence, a short calendar spread is a bet on high vola. The correct answer should have been Long butterfly spread, which bets on decreasing vola.
Question 49: What are your arguments for American call? I would argue that an American call can be approximated by the Price of an European call (they have the same bounds). This is not true for an American put, which has other bounds than a European put.

Question 30: For quantile VaR, there can be situations where VaR(PF) > VaR(A) + VaR(B)
Question 41: As I recall the question specified a decline in "volatility", not a small price move. If GARP uses standard market terminology, that means a drop in implied volatility. Of the choices, the only one that was short vega was the calendar spread.

 

[over]

re question about chosing to either creat internal ratings or go with agencies, the logical choices seemed to be either a bond rated investment grade at the beginning of the year is unlikely to default or ratings in the depths of the business cycle have less quality. Still unsure as to the best answer.

 

[Turtle King]

I also find exam hard. Main problem was time.Even I couldn't have a look to some questions. The worst, there wasn't any watch so you could control your time. I've practiced garp mock, schweser and bionic turtle. I'd say David's questions are much better reflection of real exam than schweser. Even there was question about coskewness (question was like this: which value indicates highest risk. I chose highest positive number). About other questions: roughly :
1. Manager implements different risk models and asked to create a model Indian market. Manger is not familiar with the Indian economy. When asked he told that he will create model but can't guarantee time. I think he violated CoC by creating model without knowledge about economy .
2.Two step binomial model with american put. Despite I know how to do it I couldn't attempt, again because of time.
3. First question was about VaR. you have mean return and s.d calculated over 3 year's period. What is annualized volatility over three years. I could't do it. I think it is not something like dividing by 3^0.5.
4. Long put with strike 40 and long call with strike 60. Also payoff table is given for different prices. Question ask which additional option we need to choose so that our payoff becomes the one shown on the table. I remember one line of table: if price is between 40 and 60 then payoff is zero.
5.On which of following Black-Schole can't be implemened: forward, E. call, A.Put, A. call. I chose A. put
6. Same expected return for two stock. Indexed to the same benchmark. Which one is true? I choose lower beta will have higher Treynor but I think it is wrong.
7. Option has positive convexity. which one is true? I choose if volatility is higher then return will be high.
8. Man sell futures and rolls it. and has losses. Which is the cause? Backwardation or Contango?
9. This question was about black approximation. You have inputs and you need to calculated X(1-e^-r(T-t)).
10. Which incorporates correlation: Factor push or Conditional Scenario.
11. Risk of Sinking fund provision to the investor .I choose reinvestment risk.
12. Which one impact interest rate parity: sovern government spread , inflation. I chose inflation.
13. There were lots of VaR question. Typical one was. you have 1 day 95% VaR. need to calculate 10 day 99% VaR.
14. I don't remember Expected Shortfall question. I also read option strategies a lot but there was only one and even I don't know whether I wrote it right. Investor expects volatility could decrease. which strategy to choose ? Straddle, Short butterly, Short calendar, short bull or bear .
15. One portfolio volatility question. I calculated it three times. I think GARP shouldn't test as with digits. I calculated 10.3 then sow there is also 10.2.
16. Jensen alpha. portfolio return. market return, beta are given over 5 years. you need to calculate jensen alpha. I calculated both way: taking average of 5 year and finding CAPM or taking just last year. In both way I couldn't find exact answer.
17. Leveraged beta. sucks.
18. Gamma hedge.
19. You have annual , semi-annual, quarterly, monthly compounding coupon rates . which one offers highest annualized return
20. Monte Carlo with one path

I'd like to know what other guys think.

I beg to disagree on your views in Q1 above... The analyst only wanted time to come up with a model.. i.e to research, study and understand the factors that might have to be coded into the model - which in my view was only a fair requirement to do justice to the task in hand!

 

[noalv4]

Does someone remember the answers to the following qwestions?

1. BLR, is it 17.5 Million?
2. What is part of Risk assesment?
3. The role of a Management?
4. Short FWD EUR/USD of 50 million EUR. gain of 1.584 million.
5. Lower duration with zero coupon?
6. MC simulation, what is the correct answer?
7. Data Quality, somthing with Consistency, currency and more...
8. Probability of A for undergrade. is it 7.5%?
9. Binomial Europian call. is it 0.68 USD?
10. FWD=Future -1/2 convexity with vol of 0.01 t1*t2 act 365. is it 3.03%?
11. Converting VaR from 95% to 99% 10 days. My answer was 12.5. Is it correct?
12. The price of Future contract is 108? (conviniance yield=3.3%, c=5.5% r=don't remember)
13. The bank and the law suit, is the correct answer is the scenarios?
14. CTD, is it BOND B (the lowest loss)?
15. What is the correct answer with the CAPM calculation? mine was 8%. Is it correct?
16. The lowest VaR, is it MC VaR?
17. Another question with data quality... that I don't quite remember... somthing with units... what is the correct answer?
18. Future commodity... doea the correct answer is that the exchange determeine the grade of the commodity at the beggining?
19. What is the Efficiant portfolio? I compered the rates and the SD, and choose the one that seems to have the highest rate with the lowest SD compare to the others. correct?
20. Options with convexity???

Thanks,
Noa

 

[Spronkworks]

I beg to disagree on your views in Q1 above... The analyst only wanted time to come up with a model.. i.e to research, study and understand the factors that might have to be coded into the model - which in my view was only a fair requirement to do justice to the task in hand!

Quoting paragraph 4.5 of the CoC: "GARP Memebers: Shall clearly disclose the relevant limits of their specific knowledge and expertise concerning risk assessment, industry practices, and applicable laws and regulations." Thus there is a positive requirement to disclose the lack of knowledge, which is contrary to your otherwise reasonable argument. The RM can't simply ask for a month without explaining that they need a month to "learn up" on the specific market.

 

[Juan B.]

"3) Regression graph for two assets (both asset vols were supplied) and the question was about the st dev of portfolio I think. Answer: Cant remember exact answer but its one where the the correlation is a negative values close to -1 (this is inferred from the graph of the assets)."

Hi, on this one re portfolio vol, anyone thinking, as per the chart, that the correlation is zero...correlation close to -1 should be quite clear on a scattreplot and the observations were quite disperse with few of them gathered below the x axis...thanks

 

[Uchica__Itachi]

Guys on the GARP COC question the analyst did say he had little knowledge of the markets he was requested to build the model for as well as saying he couldnt promise when he could complete the model. Looks like he did his part and didnt violate COC to me.

Regression graph had a clear pattern which was closish to y=-x pattern so the correlation was close to -1 (not quite -1).

For the VaR question, VaR is not always subadditive which implies that VaR of portfolio could be higher than sum of componet VaRs as Spronkworks pointed out.