Any Feedback on FRM 2009 Exam ?

[David Harper CFA FRM]

Hi Melle,

Thanks for your feedback, it is really appreciated.

In regard to, "In preparing for the exam, I did most of the practice questions on Bionic Turtle and the FRM practice exam. I still found the actual exam questions to be very different."

That's obviously not where we want to be: I want our questions to be a good simulation of the exam (actually, I think perfect would be: a good simulation but maybe a notch more difficult).

This alignment is a real priority for me ... I know we will be working on this and *also* I intend to make it part of my conversation with GARP (b/c some component of it depends on GARP providing prospective clarity) ...historically, i have trusted that a focus on the basics will take care of itself, but lately I do percieve that our customers are hungrier for a "more representitve" practice question experience.

... thanks, David

 

[mdecav]

David,

I've already sent to GARP my issues (1. two typos, 2. box spreads, 3. two-parter where one answer depends on the other) so I know they're aware of it. The Bloomberg question was quite verbose, too.

To reiterate my point - I thought the exam was pretty straight-forward, where a few questions are very tough - these might be the questions many people will focus on when giving their opinions. One of things you have to expect when taking these type of exams is that 10-15% of these questions are going to be hard.

I think I took an absolute random guess on around 8 or so questions out of the 100 (I did Level 1), but I know I'm still borderline. I feel the same way I did when I took CFA L2 in June. Thought I did well enough just to pass and unfortunately got a 'fail-10'.

 

[troubleshooter]

I guess while evaluating how well you did, you have to take it relatively. For those 10-12 % really hard questions, how many of the folk could have gotten them right? So, in average, I think if you got the easy questions right, chances are you will likely pass.

 

[David Harper CFA FRM]

mdecav, thanks for sharing those three items (I'll repeat them in my list, to give further support) ... the FRM post-exam self-evaluation (IMO) isn't directly comparable to CFA (rarely does anyone feel confident; the best almost anyone feels is "borderline" ... those folks have generally passed)...I truly believe, based on what you wrote here, that you should be confident.

David

 

[andiela]

Also L1 from NYC, NewTurtle really has an awesome memory... Many people feel the AM is quite tough but the PM part is better, why I just have the opposite feeling....There are so many calculations which I had to give up and guess them ultimately.....

Can anybody still remember the answers for the :

1) Bloomberg question;

2) The one which require us to obtain the covariance between A and B, given covariance matrix between Global equity index and Global bond index ;

3) T-Distribution Question: Given sample mean, sample size and standar deviation, need to find a confidence interval of 95%. t-distribution is given for df = sample size, df = sample size - 1 and for 2.5% and 5% at the right tail.
Which df should we use? 30 or 29?

BTW, how many questions need we keep right to pass the test? 70 out of 100 for L1?

Thanks !!

 

[benoit16]

I think I took an absolute random guess on around 8 or so questions out of the 100 (I did Level 1), but I know I'm still borderline. I feel the same way I did when I took CFA L2 in June. Thought I did well enough just to pass and unfortunately got a 'fail-10'.​

Hello mdecav,

Having not to to guess on more than 90% of questions, I definitively think that you are going to pass. ;-)

See you,

 

[benoit16]

Good morning,

Would you know how the pass rate is computed:
nbPassed / nbRegistered or nbPassed/nbTakenTheExam

This percentage varies from 37% to 62.7%.
Would you know why the volatility of the pass rate is so high?

Thanks in advance for your answer.

 

[troubleshooter]

@andiela:

1) Bloomberg question - If you are talking about the one that gave a bunch of bonds maturing in succeeding 6 month time frame, there was about 6 bonds but you had to use only first two or three. The Price, time to maturity coupon was given. We needed to find the discount rate. I thought it was bootstrapping calculations. First bond maturing in 6 months gave the discount rate for that period. This discount rate was used in the second bond's first coupon and then come up with the discount rate for 12 months... It thought this was a straightforward question with more information provided than needed as ususal.

2) The one which require us to obtain the covariance between A and B, given covariance matrix between Global equity index and Global bond index ;
This one was a killer. I had about 6-7 mintues to spare in the end and spent all that time trying to work this one out but still unsuccessfully. What I reember from this question is as follows (I have posted this earlier).

"Question on APT but was not able to figure out how it was to be solved. I think the regression is to be done on Global equity index and Global bond index of two markets A and B. Covariance matrix given for Gloabl Equity Index and Global bond index. Also Given are betas of Market A with both indexes and the same for market B. Also given are standard deviations of market A and market B. Need to find correlation coefficient between market A and market B. Very very tricky, had to guess." Still wondering how the heck you would solve this one.

3) T-Distribution Question: Given sample mean, sample size and standar deviation, need to find a confidence interval of 95%. t-distribution is given for df = sample size, df = sample size - 1 and for 2.5% and 5% at the right tail.
Which df should we use? 30 or 29?
Sam[ple size was 30 so df = 29. Since it was confidence interval hence used critical values based on 2.5% at the right and 2.5% at the left tail.

BTW, how many questions need we keep right to pass the test? 70 out of 100 for L1?

70 seems to be a reasonable guess but who knows. I don't think GARP is transparent in deciding pass mark. I think it varies every year.

 

[David Harper CFA FRM]

Hi benoit16,

1. pass rate = X * Top 5% of (nbTakenTheExam); e.g., 0.7 * 5th quantile of (nbTakenTheExam)
which, of course, translates into some ratio of (nbPassed)/(nbTakenTheExam), which is likely to be less than 70% or 75%.

2. Volatility of pass rate is high (or, at least higher than typical standardized exam) for the same reason exam contents are not highly predictable (what I call a relative lack of prospective clarity): because they don't repeat the same questions each year, it's hard to calibrate with consistency. (IMO, this is mostly a good thing, but on the other hand, poses some challenges).

David

 

[mdecav]

I think I took an absolute random guess on around 8 or so questions out of the 100 (I did Level 1), but I know I'm still borderline. I feel the same way I did when I took CFA L2 in June. Thought I did well enough just to pass and unfortunately got a 'fail-10'.​

Hello mdecav,

Having not to to guess on more than 90% of questions, I definitively think that you are going to pass. ;-)

See you,​

Adding to what I said I still had to guess on a number of others. My count was exactly 30 that I wasn't 100% on (17 in the AM and 13 in the PM), which was my goal.

Re: the Bloomberg question. The question asked for a seven month discount factor based on the other bonds. This was one of the other questions where *I think* logic took precendence over actually doing the calculations. From what I remember all of the bonds had coupons close to the $4-5 range, with their prices hovering around par. This, to me, points to interest rates to be in the 4-5% range. The 7-month discount factor choices - annualized - were something like 0.5%, 1%, 4-5%, and 7-8%. I chose the third one.

 

[cine]

Q: Find 2 year forward rate after 3 years. Given were 5 spot rates for the next five years. Ans was coming to be 6%

Regarding Box Spread, Yeah definitely the Schwer just touched upon it but donno whether core readings did any justice. I think they distributed HULL for derivatives and donno if at least that covered it. I ordered core readings but they were impossible to read even after having done courses in Fixed income and derivatives during my course !!! ( by the way what's the answer? I market option 3, just like that !, also found that option 2 can't be the answer !)

It makes sense to say that large losses may occur inspite of having flawless RM in place as David says. Completely agree with it. And I have a premonition that they were testing exactly this point !!!


Cheers.

 

[prabal_sidana]

I think with box spreads you have locked onto a price and therefore there is no possibility of a gain or a loss from that price, irrespective of the direction of share price.

I know in the preceding discussion David has mentioned that the cutoff mark is usually calculated as X*5%ile *0.7x, but does anyone know what the actual cutoff was in the previous years? Was it 65-70% for the full exam? (i.e. between 91-98 out of 140 questions). I personally think that the cutoff this year will be lower than the previous years, something around 60%.

 

[benoit16]

personally think that the cutoff this year will be lower than the previous years, something around 60%.​

Good morning,

Why should it be lower this year?

Thanks in advance for your answer.

 

[prabal_sidana]

Hi,

I took the full test from Bangalore and there were another 250 candidates taking the same exam. Almost all of them had the same thing to say as has already been discussed in this forum - "the first session was very difficult, and the second session was RELATIVELY easy." Many of them said that they would struggle to make a 55% overall. I've had similar discussions with friends taking the exam from other cities in India and they expect a similar score. Hence, I felt the cutoff might drop this year.

Of course, I might be generalising, which is why I thought David might be able to answer this question better considering that he does receive feedback from students taking the FRM each year.

David, do you think that candidates this year have felt that the paper was more difficult than previous years, or is this reflective of normal level of difficulty?

Thanks in advance.

 

[David Harper CFA FRM]

GW -

My impression is: on average, candidate perceptions seem to be about the same; i.e., difficult or very difficult exam. The average candidate almost always feels like maybe they did not pass (such that it is quite common for several candidates to pass, and to be surprised that they passed).

However, so far, at the upper end, only medcav (plus 2-3 others that i infer) are reporting back what I would call "confidence" or a prediction that they will pass (e.g., guessing on only 8 is clearly a strong performance). So I definitely perceive fewer "confident" self-reports. Last year, we at least had several "confident" self-reporters (e.g., there were maybe 10 or so that knew they would pass. But again, there were MANY who did not think they'd pass, but did. That was the typical experience).... so, it will not surprise me if the full exam has a tougher calibration (which some had predicted).

David

 

[mdecav]

I guessed on 30, meaning not 100% sure. I blindly guessed on around 8.

 

[David Harper CFA FRM]

apologies mdecav, I misinterpreted. Okay, then, there is definitely a much thiner layer of "confidence at the top" (nearly non-existent)...David

 

[troubleshooter]

My breakdown would be:
Confident 75
Educated Guess 10
Wild Guess 15

 

[David Harper CFA FRM]

@NewTutle -

That deserves confidence (plus, I think you probably did well on your educated guesses, based on your other posts)

@mdecav -

I saw your post about bionicturtle.com on analystforum. Thank you for sharing feedback there. I saw your comment about slow site, so i just got off the phone with my host liquidweb, and they agreed, the site has slowed due to the (considerable) traffic increase . Customers' site experience is very important to me, and i know response time is part of that, so I ordered a hefty upgrade: will take a few days to migrate to new server, but they are telling me that it should make the site quicker
(if you notice that, I would be very grateful if you could update your feedback to that effect...)

Thanks, David

 

[troubleshooter]

David: I would like to think so but only 5th Jan will tell.