Discussion in 'About FRM' started by simhan, May 20, 2011.
How was the exam in your part of the world? what are the questions which you can remember?
I just took the paper in Singapore. I thought it was alright. But i thought there were a couple of mistakes in the paper. (anyone noticed?)
Overall, I think it's mostly quantitative but with quite substantial theory questions (more than expected). Some questions require deeper understanding, not something to rote memorize. I did not study the theory questions and so i mostly answered them with elimination and gut feel. (i saw some older threads on the lack of theory questions so it was a gamble for me) I think the maths part of FRM is easier to tackle given that theory is rather "vast" and "loose" and that was my focus; to really dial the maths. I hope it's worth the gamble. I would say around 30% theory, 70% maths. If its the other way round, i would be in "bleed mode".
As some also mentioned previously, the initial questions were a little trickier for me. Towards the end, it was easier and some questions took only a glance to do.
I completed the paper with around 20 mins left. I didn't bother checking as i thought it may be counterproductive. I had only 1 regret which is not to spend more time on theory topics. As i leave the hall, I managed to overhear some chatter. 1 guy said to his girlfriend,"so difficult, sure fail". I saw a middle aged women looking like she was on the verge of fainting. (maybe she did not choose BT). A scan of the faces, i sensed that most working adults do not look confident at the exit. I was among a small group of students. Although i did not speak to them, I could see that they were more confident. Just my observations.
All the best!
A friend of mine took the exam from India...
He also said the same thing... He got 3/4 questions which had no choices to match with (the choices were wrong)
He also studied the maths part only and he was saying that there were many theory questions from ERM, Op risk and stress testing... However the maths was easy...
He finished 13 mins early...
I have found the exam a bit challenging in terms of concepts being testes (OCL, R2, TSS, probability questions). In particular, the areas I have found not quite straightforward were related to quant’s section (my weakest area). Also I did not take my watch (a must!) with me so I wasted some time along the way (calculations?) and had to accelerate (with 45 minutes to go) with my last 15 -17 questions. Overall impression if compared to the November 2010 exam, which I failed, - is good. Not sure if this due to the fact that this time I was better prepared or GARP really has made some effort in structuring the paper better this time round. Some questions were too easy so I am a bit worried. Some questions I did not get at all (language issues?) and had to guess (4-5 times). But again, there is always be an excuse : "If I only had one more week to revise..."
Good luck to everyone !!!
P.S. my advise for those who take P1 again is to study everything with no exceptions. study to understand every bit of material!! hundreds of questions practiced (i.e. Schweser QBank) may not help if you just don't get the concepts right.. IMHO: Exam was quite different from Schweiser practice papres and QBank question types.. So If I have to retake, I will spend much less time on Schweiser stuff (i.e.QBank) as it is just tool to memorise formula's and nothing more than that...
I agree with Jiew Kwang on that I thought there were few mistakes on the paper. Hence I spent lots of time trying to do those questions over and over again yet not getting any of the answer choices.
Qualitative part was a big step up from sample papers. If not for David's chapter questions-I wouldn't have had a clue. So thanks for that David! Yet, I still struggled with some questions. A lot of those questions were on the same topic (OLS, Stress testing and markets (what to short/long etc.)).
I also thought that exam didn't go across all aims/topics but mainly concentrated on few areas. Hence if you didn't remember a particular formula-you were out for at least two questions. I thought it wasn't all that fair.
So in overall, I though exam was tough enough. I thought I was quite well prepared but was a bit disappointed with my performance today. I'd say for at least 10-15 questions I narrowed answer option to 2-3 choices but then it was a random guess - something I hate to be doing in the exam!
Oh, well, it's done now! I hope everybody did well and will receive good results in July!
P.S. Is there any way exam could start AT LEAST at 9am??? 7.45am is ridiculously early for any exam, especially for 4hrs quantitative.
I agree too with the previous threads!
I lost so much time trying to find the correct answers within wrong choices (especially for the question where we had to calculate the UL)
A lot of questions related to OLS , stress testing, Operational R.
-maybe only 2-3 questions related to GARP code of conduct
- only 2-3 questions related to the cases.
- Many questions were under the format : statement i correct - Statement ii correct - both - neither.
I was expecting to have an exam that will test the knowledges broadly and fairly ... but actually I felt as if it was focusing on some topics.
As David mentioned, the sample apperead to be poorly representattive of the actual diffculty level ... and indeed it is far away from what we discovered today!
I lost so much time trying to find the correct answers within wrong choices (especially for the question where we had to calculate the UL)
- OK, so it wasn't just me then. I got stuck with this question too and thought I had the right answer but it wasn't one of the choices.
apparently for level 2, a lot of questions are under the A or B or both AB or neither as well
I'm kind of scared to say this, but I thought the exam was "ok". Ditto on the UL question because I had that formula down cold and wasted a good 5 min on it. No surprise, but I couldn't figure out the swap question with the series of 6 month libors plus the 1.25%(?). It took me a while to come up with a value, but I forgot to add the 1.25% to the LIBOR! So I ended up guessing. Also, the bond question asking which one had the highest yield based off the different coupon rates and frequencies???
One thing is certain, GARP needs to invest a lot more in proofreaders for these exams. We pay them a fee and they should be able to do this fairly simple task. The cynical side of me wants to think that they put all wrong choices in some of the questions on purpose to trick up people who don't know when to move on to another question. I dunno...
I didn't think the test was that bad, but the time requried to answer the questions on a lot isn't sufficent. It took me about 5 minutes on several questions to read through it and gather the data to answer.
They have to get a clock in there so you can keep better time of how much time you have or taking. That posting a time every 30 minutes just doesn't work. I ended up not answering a good 15 questions. Like the other two, the UL question got me, and when I originally saw the question I thought an easy one. Never could find the answer. The libor floater almost got me, wasted a lot of time on it for I was valuing a swap, and it only asked for the payment. You have to pay very close attention to the questions before you start calculating. I think the problem with the test it's self as others said is that the practice test are not indicitive of the diffifulty of the test. I went through the practice test with ease, however the actual test required a lot more concentration for they trough loops in the question to throw you off.
The test it's self should be reduced to 80 questions or time increased to 5 hours. Plus I forgot to wear a watch with the expectaion there would be a clock in the room, I didn't bother to pick one up on the way to the test.
the result is for July 6th? right?
@Jiew: thanks for conveying your impressions. Given my sense that you genuinely did apply yourself into the material (based on consistent presence in the forum, without a last minute rush), I'll look forward to hearing your result but it sounds like a good feeling is warranted. Thanks for your many forum thoughts!
@step4cfa: thanks for sharing your experience. I think this is very interesting: "my advise for those who take P1 again is to study everything with no exceptions. study to understand every bit of material!! hundreds of questions practiced (i.e. Schweser QBank) may not help if you just don’t get the concepts right.."
... I am going to work hard ahead of the November exam to improve the organization/utility of our practice questions
@Jenny: Thanks for appreciating the chapter questions. (re "a bit disappointed:" i don't think that necessarily is predictive. Total confidence seems to be impossible) ... please share back your result when your hear
@najwa: Thanks for the concrete detail shared. I agree with your expectation that "exam [ought to] test the knowledges broadly and fairly." Based on the ultra-narrowness of the sample papers, this is a principle (concern) I have stressed to GARP. I have offered specific ideas-help, hopefully they will be receptive ... but please share your result when you get it, I look forward to hearing (btw, I appreciated your help in pushing my own understanding of d.f. in OLS!).
@Zack: thanks for your feedback. I 100% agree with your final sentiment (I realize you are only being "cynical" but they would never intentionally give all four incorrect answers; if that did occur, and i have no direct evidence, it would be an unintentional mistake ... but for the fees paid, this should not be allowed to happen on the final exam, IMO, as GARP has the resources to quality check).
Thanks David, my understanding in risk is still far from good. But i feel confident that i have much to learn from you and other forumers. I can't wait to subscribe to FRM part II although i will be attempting CFA level 1 in decemeber. Maybe I'll do it later after breakfast!
@2011 L1 candidates:
I can remember 3 questions that got me: There was the 1) UL , 2) futures margin account and 3) EUR/USD forward questions that I stumbled as possibly containing errors.
UL qn: AE = 20m(calculated from previous question in paper) PD = 0.25% LGD = 60% LGD sd = 25%, so the first thing i did was to find variance of PD which was pq = 0.002494. I plugged into the equation and got UL = 649,335. And all 4 options are 300,000 which was a far cry from my answer. Ended up just shading an answer.
Futures margin account: Forgot the question as there were 5 time steps and I got like 5M+ but answers were 4M+..
EUR/USD short: I believe there was an error in the 1st sentence of this question. Should have been something like USD/EUR..
Questions i remembered:
QN1: Portfolio duration is 7 years. An appropriate hedge is put in place. What is the aggregate gain if a small move in interest rates happen? I chose (D) which was none of the above as I reasoned that a hedge would only offset any losses on the portfolio.
QN2:Expected loss: Total loan Commitment: 30M Drawn: 10M Unused Com: 20M PD: 0.25% LGD:60% UGD:50%. My calculated EL = 30K AnsD)
QN3: Was mentioned above
QN4: Which of the following has the highest claim?
(i) collateral trust bonds
(iii) secured bonds
I chose (iii) secured bonds
QN?: Binomial question on basket of bonds asking for probability of default of at least 1 and 2 bonds defaulting. n=10 p=10% q=90%
My answer was 58%. There were 2 options with 30+% and 1 option with 19%.
QN?: How to test Simultaneously for H0:b1:b2 = 0?
i) Individually test t statistic for b1=0 b2=0
ii) Use F test to test for H0:b1:b2 = 0
I chose (B) use F test only
QN?: (portfolio) Long call options with strike of 60 and long put options with strike of 40(or was it 20?). Which of the following actions would result in (diagram depicting option payoff of unlimited upside and capped downside[horizontal line across])?
Cant recall exact sequence of options. I chose iv)short put option as that will cap downside
QN?:Which of the following will add value? (cant remember actual words used)?
i) reduce bankruptcy cost
ii) reduce financial distress cost
I chose ii) reduce financial distress as i didn't study this topic much. lol.
QN: A particular stock in the index has a function y = Bx + 3.53...(a lot of decimals) then there was a graph of the index that showed the RFR at 2.5% (question also explicitly said RFR = 2.5% with a lot of other information). What is the Jensen's alpha?
Forgot the options but i just took 3.53% - 2.5%
There were alot of regression questions. 1 was to find R^2 given around 8 columns that were similar to what you would normally do in excel. They expected you to know that TSS is total squared difference in Y [Y-mean of Y]^2 and RSS from e^2 column. ANOVA table also appeared and i think doing some excel demonstrations will help.
There were alot of RAPM / CAPM / markowitz portfolio theory questions. ~10 questions on this. These questions gave you a clutter of information which you basically need only maybe 2 data points like Treynor ratio came out. Oh, i remembered a question that asked you to find the CAPM in a form of y = RFR + beta*X + Error term. The RFR i got was 2.5 but the options only had something like 3 for the intercept (RFR) so i chose 3. Hope its a rounding error on their part.
Just thought of 2 question: S0 = $100. expected return = 6%. Volatility = 12%. What is the expected value (mean) of the stock in 2 years? What is the expected value of the variance in 2 years?
options for mean value of stock include:
iii)112.75 (my ans)
I did it with just continuously compounding it.
As for the variance the options ranged from 300+ to 600+. I totally cannot remember how to do this!
Okay, that's all the total recall i have for now. Don't want to come up with more and find out I got them wrong!
p.s anyone from Singapore here? =)
I took Part I in San Francisco this morning and agree with most of the earlier comments. Glad to hear I wasn't the only one who had trouble with that UL question. I honestly have no sense as to whether I passed or failed. My biggest problem was time management. Like some of the other candidates on the forum I spent too much time trying to work out problems for which the correct answer didn't seem to be among the choices (but I won't rule out an error on my part here). Even with a watch, I ran short on time and near the end wound up guessing on around 10 questions for which I had only been able to narrow down the choices to two or three. I was also thrown off by the carbon copy answer sheet. In practice exams my strategy had been to mark all questions with an answer, note on the answer sheet with a small mark the ones I was uncertain about, and then coming back to them and erasing the extra marks. I was not expecting to be unable to erase on the actual exam! Overall, I went in this morning feeling fairly well prepared and was humbled by the difficulty of the actual exam compared with the practice problems I worked. I thought the GARP samples were a cakewalk compared to the actual exam. I also agree with step4cfa that the Schweser questions were of little help in preparing. If I have to retake in November I will spend a lot more time with David's questions. As he's mentioned all along, GARP really tests one's knowledge of the concepts, and with only 4 hours to complete the entire exam there is no time to sit and think things through. Whatever the outcome, this has been a great learning experience and only the first step. A big thanks to David and Suzanne for everything. Please keep up the excellent work!
On the UL question; my thought was that GARP had used Exposure = $10mm. Either from the unadjusted Drawn amount, or "Commitment - (Drawn + UGD * Undrawn)", both giving UL=649,335 / 2
On the variance of a log-normal stock price (average return = 6%, volatility of return = 12%) in 2-years (answers ranged from ~150 to 600); I don't remember seeing variance of log-normal distribution in the readings and certainly have never memorized the formula... In retrospect could have tried something based on LN ~ ln[N(mu,sigma^2)].
There was also a question on which country would have the lowest/highest credit spread based on two categories with which I was unfamiliar; one of which was "Promised Payments", the other I can't recall.
Unlike some posters, I felt the past sample-papers (I reviewed all since 2006, in addition to David's questions) did provide a fair measure of question difficulty in the actual exam. But there is no taking away from the fact that a 4-hour exam requires stamina and that questions which might be relatively easy when fresh become progressively more challenging the further into the exam you get.
On exam technique, the plan I followed was
(1) Start working through David's questions at least 1-month before the exam
(2) Take the preceding 5-days off work to cram/do past papers
(2) Write a formula sheet and memorize it (mine was 9-pages)
(3) Planned on 2-mins per question and skipped any which tool longer than this or was unfamiliar with; this guarantees you'll answer all the 'easier' questions and still have 40-mins to review the difficult ones (about 15 in my case)
I'll let you know how effective it was in July .
Guys, so helpful info..... I'll write FRM this Nov2011 and I thought I should start preparing early.... The friend of mine who wrote the exam yesterday gave me the following questions to cross check... I think it might help other BT members get a feel of the exam...
Q1. See Jiew Kwang's post above. Yes JK, the question did not say anything about the convexity. So my ans. would also be D)
Q3. UL, my friend also got a similar number and me too...
Q4. The other option was mortgage bonds (My ans. would be Secured bond)
Qn. (portfolio) Long call options with strike of 60 and long put options with strike of 20. Which of the following actions would result in the below payoff?
i)long call 40
ii)short call 40
iii)long put 40
iv)short put 40
Ans. is short put 40
Qn. Under risk management irrelevance proposition which of the following will add value?
i) reduce bankruptcy cost
ii) reduce financial distress cost
My ans. would be iv) since risk management irrelevance deals with perfect market. If it is not perfect, the ans. should be iii) both. I guess the ques. is not well presented.
Qn: A plot of Excess stock return and Excess market return was given which had the eqn. y = Bx + 3.53. What is the Jensen's alpha.
Ans. its just the intercept term. 3.53
Another wrong question (I GUESS):
Qn. Stock return=11%, Mkt return=7%, Cov(stock,Mkt)=6%, Volatility(Mkt)=18%. What is the regression eqn. Ans. b1=1.85... and 11%=b0+1.85*7% which yielded b0 -ve.
Qn. 10,000 options with delta 0.6 each strike 100 used to hedge equity portfolio of 100,000. Volatility yearly 30%, 252 trading days. 95% Var of option position. Ans. came around 1,8-- USD. Do not rem the numbers exactly. Some one could help.
Qn. Question involving no arbitrage price range for forwards with convenience yield. Four options were given and only one forward price lied inside the range.
Qn. Another ques. asking to calculate the forward price directly given storage cost and market conveniyence yield.
Qn. One/two questions involving american P-C inequality and European P-C parity.
Example. Two at the money amer. call and amer. put. Which of the following is true. The correct choice was C >= P. Dun rem. the other choices.
Qn. Risk neutral probability of an up move (Given volatility, R and time)
Qn. In ERM framework, which of the following is/are natural hedges.
i) Firm sells both life insurance and annuities
ii) Losses in one trading desk offset by the gain in another.
Options - i) only, ii) only, both, neither.
My guess would be both.
Qn. Distro. of FICO B-score and states.
States-> A B C ...
The number of people falling in each square was given. Find P(score atleast 600/customer is From California). Ans. very simple prob which requires looking up at only one column (California).. The extra info is to make you nervous and go out of time.
Qn. Two questions from Transition matrices.
1) Which inv. grade bond has the highest chance of getting upgraded in one year.
1) Which investment grade bond has the highest chance of getting downgraded in one year.
Qn. I remember a question posted by Saray in one of the threads (Is the VaR calculated using delta normal approach over-estimator or under-estimator of the actual VaR for options)
The original question from yesterday's exam - Two analysts A and B calculates VaR (A by delta normal) and B by monte-carlo simulation with 10,000 trials. Who has reported a higher VaR. Ans. would be A
Qn. Another question directly from the Sample paper. (3 bonds given with one of the price missing) What is the price.
Qn. Purchase Pwer Parity: Which is the reason for exchange rate fluctuations.
i) change in money supply between two countries
ii) GDP of the two
iii) Inflation in the two countries
iv) Forgot. Clueless about the answer.
Qn from APT.
1) Find Rp given b1,b2,Rf,Rx,Ry
2) Beta of Rx and Ry given. Find the beta of Rp.
Qn Oh, I instantly loved this question when i heard it from my friend.
Given a plot of VaRs vs. Confidence level (92%, 94%, 96% and 98%)
Four possible curves were given:
1) Increasing, concave function
2) Increasing, mix of convex and concave
3) increasing, linear
4) increasing convex. (I wish I could post the plot)
Ans. answer is 4). For uniform loss distro (which is never the case), the ans. will be 3). For loss distro with more weight in the tails than the mid region, a concave plot might result (which is even more impossible).
Qn on portfolio of two assets A and B with correlation of 1. What is the possible shape of Return vs. Total risk (sigma). Ans. the straight line.
Will ask my friend if he remembers any more and will keep posted.
Also, on a different note I think it's a good idea to visit the forums regularly for new problems. You are increasing your portfolio of known problems and reducing the idiosyncratic risk !!! (Bad Joke) (IMO, of many other things, the forum is going to play a tremendous role, thanks BT).
Your friend has amazing memory to remember so many questions with such detail. But I think I've learnt a few things so far already just from this thread. I just hope for the best. There were a couple of questions as mentioned de and yourself that; ~2 that somehow I felt were not in the prescribed syllabus. Add the questions with mistakes in them that would be ~6 questions that were ambiguous? (i wonder how will GARP handle this cause its not fair to candidates who knew how to derive the answer!)
I just felt my confidence dip a little seeing some of your answers but i guess its done and i cant do anything about it though i probably could recall a handful of questions which if i post here may be detrimental to my mental state for the next 1 month if discussed. But on hindsight i thought the exams could potentially be more intimidating. And i find consolation from a previous thread discussed by a candidate who thought he would flung the paper but scored 1st quartiles for all sections!
And so i took a leap of faith and purchased BT level 2 earlier this morning. I hope that my next focus will be the level 2 exams and wish everyone here the same!
@de & JK,
The way i'd go about the variance thing is try to find E(St^2) from the characteristic function of LN distro.
Though, it might just be easier to memorize the E(St^2)
Qn. A firm wants to transform current fixed liabilities into floating using a swap deal.
A plot of the spreads of possible counterparties' borrowing rates in fixed and floating markets were given.
Divided in four quadrants - X axis - Spread in fixed rate, Y axis - spread in floating rate.
1) X in (0%-1%) Y in (0%-1%)
2) X in (1%-2%) Y in (0%-1%)
3) X in (0%-1%) Y in (1%-2%)
4) X in (1%-2%) Y in (1%-2%)
In which quadrant will the firm have a comparative advantage.
Ans should be 3) since in floating rate counterparty borrows at > 1% more and in fixed counterparty borrows at < 1% more. Clearly an edge for our firm.
Qn. An analyst is learning MC and starts with simulating standard normal. What is the proportion of outcomes lying between 0 and 1 - 34%
Qn. Which of the following not a source of basis risk:
1) Heating oil to hedge against jet fuel
2) Mkt volatility fluctuation
3) Mismatched expiry
Qn. An analyst while stress testing multiplies the historical return data by 3. What is the weakness in this approach. Forgot the options.
Dun worry. If it makes you feel better I sat for Nov 2010 Part 1 and failed (I used only Kaplan materials which are anyday substandard to BT). And personally I feel I have learnt a lot from BT and I have more to learn. On a different note it feels good to think that I know more now which would not have happened if I passed the exam with a weaker preparation.
And also, GARP will divide your score by the average of top 5% candidates (which I estimate at 90). So, if they require you to score 70% (not very sure about this number) relatively, you need approx 63 correct choices.
So, a few questions can not take anything away from you. I think people who have completed the paper in almost one go has a good chance. There will always be a few mistakes.
It's my gut feeling that everyone using BT has a really bright chance to win. If you compare this base them to people using Kaplan.
Qn on DV01 based hedging
Long X face value with DV01 a
Short Y face value with DV01 b
What is an appropriate hedge. Forgot the choices.
GARP tricked my friend giving the DV01 per 100 face value. He used DV01 of portfolio = Sum of DV01's (as per Tuckman) without scaling them.
Qn on delta-gamma neutral hedge.
Delta neutral position with non zero gamma. Two options - 1) delta=0.25, Gamma=1.5 2) delta=0.25 and Gamma=3
Another trick - My friend started solving the simultaneous eqns (GammaPort+1.5x+3y=0 and 0.25x+0.25y=0) which yielded an answer not among the choices. Then he went back trying each choice. This consumed a lot of time.
Qn Eurodollar futures quoted at 95 for delivery in 6 years. Sigma also given. If the quote of the ED futures is 95 after 6 years the LIBOR for the 3 month period has
Two questions on GARCH and EWMA
EWMA - Given daily volatilities and returns, you were required to find the valatility on the nth day.
GARCH(1,1) - four models were given. Which had the highest(or lowest) mean reversion.
Case studies two - MGRM and Barings
Qn on Op risk:
1) Earnings = A+B*MarketRisk+C*CreditRisk+error. What happens to op risk estimate as R^2 increases or decreases.
R^2 increases implies better fit, lower residual variance, lower estimate of OpRisk.
2) Which one below is true regarding the use of external and internal data for OpRisk models
i) External data must be scaled by the size of the firm / size of external firm (I think this is wrong since there is a power 0.23)
ii) Only external data should be adjusted for inflation (Why not internal also)
iii) Use external data only to model the loss freq. (Clueless, I mean why can't one use it to model severity also)
Qn Difference between Credit rating analyst and equity analyst ... Choices regarding the type of models used (thru the cycle for Credit rating analyst, do not know what equity analysts use - thru the cycle or at the point)
Qn Equivalent of margining in OTC mkt..... 1)netting 2)collateral 3)... 4)... My guess would be between netting and collateral
Qn Wrong way risk is/was found in 1) monoline insurers 2) .... 3) ... 4)...
Separate names with a comma.