Exam Feedback FRM Part 1 (November 2014) Exam Feedback

[sdonahue]

100) question was about one step bionominal calculation for American put option. Not difficult to calculate but I ran out of time as with the other 10 questions.

Yeah, the one-step binomial as question 100 seemed like a cruel punishment for those with time management issues. Not hard but time consuming lol.

I am not too confident in my performance. I ran out of time and realized that I was not as prepared as I wanted to be. I think I focused too much on memorizing the formulas not upon the application of said formulas.

 

[JDGutzmann]

Hi, in addition to the "...Max/Min of 100,000,000 vs 80,000,000...etc."-question, I found the exam was rather heavy on regression and test statistics.
There was one where you had to calculate portfolio variance without the covariance given, just a plotted graph of dots that would remotely hint at at negative covariance and the STDEV of two variables. No idea, so I just guessed. All in all it was a lot harder than expected, to be honest - especially since it completely left out some areas (unexpected loss) and put strong focus on some selected ones.
Best wishes to everyone! Johannes

 

[kevinyuen]

Overall I thought it was harder than I expected. The test was much more qualitative than I was expecting. There were a lot of questions pulling from obscure details of the readings that you wouldn't expect to be tested. Conversely, it felt like there were not as many questions on the typical bread and butter calculations that get emphasized in the practice tests. And if there were, those questions were familiar questions but with a new twist that you were probably seeing for the first time.

 

[Goldfish3]

I think i messed up most of the simple questions and did most complicated ones right.

The var and es question. No idea. I just chose the arbitrary averages

Bayes question. No idea. Chose B.

The scatterplot im my opinion exhibited 0 correlation. Sd of porfolio would just be without corr term. Answer A

There was another scatter plot which showed visual hetro. Gave t values and r2. But hetro wasnt in the choices so i just chose multicol.

The garch question. Chose the one which has highest long run variance. Turns out, the higher the variance the more mean reverting. So should have chosen lowest variance.

Guessed around 10 questions.

Unsure on the answer of 10-15 others.

Overall dont think the exam was hard for someone who had paid attention to the concepts and application of the material.

Suppose it all now depends on how candidates did overall.

Good luck om the outcome.

 

[LucasC]

I think i messed up most of the simple questions and did most complicated ones right.

The var and es question. No idea. I just chose the arbitrary averages

Bayes question. No idea. Chose B.

The scatterplot im my opinion exhibited 0 correlation. Sd of porfolio would just be without corr term. Answer A

There was another scatter plot which showed visual hetro. Gave t values and r2. But hetro wasnt in the choices so i just chose multicol.

The garch question. Chose the one which has highest long run variance. Turns out, the higher the variance the more mean reverting. So should have chosen lowest variance.

I also chose 750(avg)/2250(sum), but I suppose this is wrong. VaR of a combined position just should not be an average.
I think the scatterplot showed some negative correlation definitely. So I used -1 and -0.5 and picked 1.10%, think it's A too. Why do they give a question that any reasoning still let you arrive at the right answer, since it is the most close?

The t-values showed that the coefficients are not significant. Insignificant coefficients and high R2 is a sign of multicollinearity.

For the GARCH question (is it GARCH or EWMA?), the answer should be the one with the lowest sum of alpha and beta (0.91)

There was 1 question about the impact of changes in interest on price(?): Duration/Key rate/Convexity... I remember choosing Duration, but now as I think about it, is it possible that they gave such a straightforward and easy question? I may have not read the question carefully. Anyone remembers this one?

 

[alinafrmmaster]

[URL='http://bionicturtle.com/images/uploads/WindowsLiveWriterGARCHapproachandExponentialsmoothingEWMA_863image_2.png']

[/URL]

Persistence is a feature embedded in the GARCH model.
Tip: In the above formulas, persistence is = (b + c) or (alpha-1+ beta). Persistence refers to how quickly (or slowly) the variance reverts or “decays” toward its long-run average. High persistence equates to slow decay and slow “regression toward the mean;” low persistence equates to rapid decay and quick “reversion to the mean.”
So shouldnt you have chosen the lowest alpha and beta?

 

[LucasC]

Oh sorry! I picked 0.91, not 0.97

 

[hamu4ok]

Hi, in addition to the "...Max/Min of 100,000,000 vs 80,000,000...etc."-question, I found the exam was rather heavy on regression and test statistics.
There was one where you had to calculate portfolio variance without the covariance given, just a plotted graph of dots that would remotely hint at at negative covariance and the STDEV of two variables. No idea, so I just guessed. All in all it was a lot harder than expected, to be honest - especially since it completely left out some areas (unexpected loss) and put strong focus on some selected ones.
Best wishes to everyone! Johannes

There was a hypothesis test with t-stat and the excerpt of the look up table, but showing one tail when the question actually required a two tailed test.

 

[hamu4ok]

I also chose 750(avg)/2250(sum), but I suppose this is wrong. VaR of a combined position just should not be an average.
?

It required ES and Var after diversification as one postfolio and so I chose low ES due to subaddativity but high VaR.

 

[hamu4ok]

[QUOTE="LucasC, post:

There was 1 question about the impact of changes in interest on price(?): Duration/Key rate/Convexity... I remember choosing Duration, but now as I think about it, is it possible that they gave such a straightforward and easy question? I may have not read the question carefully. Anyone remembers this one?[/QUOTE]
The question was tricky, it mentioned a simple measure in percentage terms but somehow indicating ker rate. (don't remember exact wording). At first I opted for duration but finally chose key rate 01.

 

[berimbolo]

I chose duration, because it didn't mention different time intervals for hedging. I chose the higher VaR with the higher ES, as the VaR is a lower bound for the ES.
Questions I had difficulty with or couldn't spot a clear answer:
- Put option up/down step going to 8 from 5, but no K or So was given.
- The two bond portfolios with two types of bond (one had a higher dependence), it was asked which portfolio has a higher probability of at least one default.
- One question with the historical volatility (I chose something like 9%).
- Indication of omitted variable bias.
- APT assumption.
I found the exam to be pretty hard compared to for example the 2013 practice exam. But maybe that is because I focussed too much on past practice exams and this exam was quite different. Hope the best for all of us.

 

[ami44]

There was one question which i wasted a bit of time on. I think it was either the second or third question where you had to value a fx future. The agreement was to purchase £80,000 at fx rate was GBP/EUR 0.9, converting to Euros that is €88,889. However i was only seeing options including €100,000. Now i'm not sure whether i being stupid or whether they meant to write the rate as 0.8. Anybody else encounter this?

That did cost me some extra time too. Eventually I also decided that it was an error and should have read 0.8 instead of 0.9. Took me awhile though.

For the question with the scatter plot that indicated negative correlation, it was asked about the standard derivation of the sum of two portfolios. According to my (faulty?) calculation answer A and B could have been possible, without knowing the covariance exactly. I guessed A at the end (the lowest value). Was somebody able to solve that question without guessing?

In another question were two Bonds with prices and coupon rate given (I think it was 3% and 9% coupon semiannually) and it was asked about the price of a Bond with, i think, 5%. I tried to solve that by calculation the YTM for the given Bonds but I got two totally different results. Was anybody more successful than me?

 

[LucasC]

The question was tricky, it mentioned a simple measure in percentage terms but somehow indicating ker rate. (don't remember exact wording). At first I opted for duration but finally chose key rate 01.

but key rate 01 is not a measure in percentage terms, it's in monetary terms, like DV'01

 

[LucasC]

I chose duration, because it didn't mention different time intervals for hedging. I chose the higher VaR with the higher ES, as the VaR is a lower bound for the ES.
Questions I had difficulty with or couldn't spot a clear answer:
- Put option up/down step going to 8 from 5, but no K or So was given.
- The two bond portfolios with two types of bond (one had a higher dependence), it was asked which portfolio has a higher probability of at least one default.
- One question with the historical volatility (I chose something like 9%).
- Indication of omitted variable bias.
- APT assumption.
I found the exam to be pretty hard compared to for example the 2013 practice exam. But maybe that is because I focussed too much on past practice exams and this exam was quite different. Hope the best for all of us.

Since the gain from an upward move increases by $3 while the downward move still leaves the option expire worthless, so I took: original call price + probability of an upward move*present value of$3 discounted back 1 period
This question confused me a bit since the conclusion at first seemed a bit counter-intuitive for me: P(X1|X2) > P(Y1|Y2) --> P(X1|X2)*P(X1) > P(Y1|Y2)*P(Y2) since P(X1/2) = P(Y1/2) --> P(X1 & X2) > (P(Y1 & Y2)
P(at least one X) = P(X1) + P(X2) - P(X1 & X2) < P(at least one Y) = P(Y1) + P(Y2) - P(Y1 & Y2)
For the omitted variable question, chose E(error term|Y) not 0
For the APT assumption question I picked idiosyncratic risks could be diversified.

Anyone has different ideas?
Do you remember the question about country risk vs corporate risk ratings? (Some answers are country risk does not affect corporate risk ratings; Indian companies with AA ratings and companies with AA ratings in another country......) What is your answer?

 

[Finaspirant]

There was one question (2 question set) on DV01 where 5 bonds table of a portfolio was given with DV01 of each bond. It was asked to calculate the hedge for this portfolio with a 5 year bond. I went totally blank on this one as was not aware of DV01 formula for portfolio or the corresponding hedge etc..

Also, there was a straightforward question which I did not know why but could not match the answers. It was about given the gold spot rate for a mumbai exchange, a forward rate, storage cost and it was asked to determine the implied lease rate.

 

[hamu4ok]

but key rate 01 is not a measure in percentage terms, it's in monetary terms, like DV'01

Sorry, the answer said just key rate, because if it had said key rate 01, I would have chosen Duration then.

 

[hamu4ok]

Since the gain from an upward move increases by $3 while the downward move still leaves the option expire worthless, so I took: original call price + probability of an upward move*present value of$3 discounted back 1 period
This question confused me a bit since the conclusion at first seemed a bit counter-intuitive for me: P(X1|X2) > P(Y1|Y2) --> P(X1|X2)*P(X1) > P(Y1|Y2)*P(Y2) since P(X1/2) = P(Y1/2) --> P(X1 & X2) > (P(Y1 & Y2)
P(at least one X) = P(X1) + P(X2) - P(X1 & X2) < P(at least one Y) = P(Y1) + P(Y2) - P(Y1 & Y2)
For the omitted variable question, chose E(error term|Y) not 0
For the APT assumption question I picked idiosyncratic risks could be diversified.

Anyone has different ideas?
Do you remember the question about country risk vs corporate risk ratings? (Some answers are country risk does not affect corporate risk ratings; Indian companies with AA ratings and companies with AA ratings in another country......) What is your answer?

Same answers for:
For the omitted variable question, chose E(error term|Y) not 0
For the APT assumption question I picked idiosyncratic risks could be diversified.

 

[berimbolo]

Since the gain from an upward move increases by $3 while the downward move still leaves the option expire worthless, so I took: original call price + probability of an upward move*present value of$3 discounted back 1 period
This question confused me a bit since the conclusion at first seemed a bit counter-intuitive for me: P(X1|X2) > P(Y1|Y2) --> P(X1|X2)*P(X1) > P(Y1|Y2)*P(Y2) since P(X1/2) = P(Y1/2) --> P(X1 & X2) > (P(Y1 & Y2)
P(at least one X) = P(X1) + P(X2) - P(X1 & X2) < P(at least one Y) = P(Y1) + P(Y2) - P(Y1 & Y2)
For the omitted variable question, chose E(error term|Y) not 0
For the APT assumption question I picked idiosyncratic risks could be diversified.

Anyone has different ideas?
Do you remember the question about country risk vs corporate risk ratings? (Some answers are country risk does not affect corporate risk ratings; Indian companies with AA ratings and companies with AA ratings in another country......) What is your answer?

I had the same reasoning for the default question and it cost me a lot of time because it seemed counter intuitive, but I think this solution is correct.
Aah I completely forget about the corporate rating question, that one was also difficult. I chose a different answer but wasn't sure. I think for the APT I had the same. For the option I had something like 1.70?

 

[Goldfish3]

There was one question (2 question set) on DV01 where 5 bonds table of a portfolio was given with DV01 of each bond. It was asked to calculate the hedge for this portfolio with a 5 year bond. I went totally blank on this one as was not aware of DV01 formula for portfolio or the corresponding hedge etc..

Also, there was a straightforward question which I did not know why but could not match the answers. It was about given the gold spot rate for a mumbai exchange, a forward rate, storage cost and it was asked to determine the implied lease rate.

If i remember correctly the DV01 question had 4 bonds. 3 long and one short. The net exposure however was long. I used the net Dv01 to plug in the DV01 hedge formula for the available bond. I got a precise answer which was either B or C i think cant remember. Don't know could be totally wrong.

 

[Goldfish3]

Since the gain from an upward move increases by $3 while the downward move still leaves the option expire worthless, so I took: original call price + probability of an upward move*present value of$3 discounted back 1 period
This question confused me a bit since the conclusion at first seemed a bit counter-intuitive for me: P(X1|X2) > P(Y1|Y2) --> P(X1|X2)*P(X1) > P(Y1|Y2)*P(Y2) since P(X1/2) = P(Y1/2) --> P(X1 & X2) > (P(Y1 & Y2)
P(at least one X) = P(X1) + P(X2) - P(X1 & X2) < P(at least one Y) = P(Y1) + P(Y2) - P(Y1 & Y2)
For the omitted variable question, chose E(error term|Y) not 0
For the APT assumption question I picked idiosyncratic risks could be diversified.

Anyone has different ideas?
Do you remember the question about country risk vs corporate risk ratings? (Some answers are country risk does not affect corporate risk ratings; Indian companies with AA ratings and companies with AA ratings in another country......) What is your answer?

The country v company ratings question was a highly subjective one. I for one did not select independence in both ratings. I.e. an A- rated in Indonesia is not the same thing as an A- rated in say Austria. I know this because I am an asset manager for re/insurance assets. The differences are vast and political integrity of any given country trickles down to corporations in a very obvious manner. If Garp marks this as wrong then they surely don't have the concept in the bag.

Considering they had an entire reading on how rating agencies have proven to be unreliable in many past instances and that ratings made internally should not be ignored, i don't think Garp expects one to depend on rating agencies blindly.