Feedback on part 1 May 2015 exam

[ashwinashu90]

Did anyone have issues with the question of backtesting: VaR exceeding on 5 out of 150 days where we had to find out the probability of it exceeding on 6 out of 250 days? I was not able to match any of the answers. 250C6*(5/150)^6*(29/30)^244 gives around 12% as result which none of the choices had.

 

[Onibiku]

This was for 99% VaR. I thought 5 out of 150 days => 6 out of 250 days means 1 out of next 100 days. So I used 100C1*(1/100)*(99/100)^99 which is roughly 37%.

 

[Sidewinder]

did the same mistake like ashwinashu, there is always a tricky part...

There was another nasty question where they asked to calculative the explained sum of squares with TSS and correlation. So you had to have two things on mind: the basic formula and the connection between R^2 and correlation.

 

[GVD]

Hi,

I used binomial distribution:
250C6*(0.01)^6*(0.99)^244 ...the solution was in there...

I had a problem with the 2 traders and rf = 0.02 (for 1 year) and sigma = 35 % (for 1 year) and delta_t = 1/12 ... (12 step binomial tree) get the probability of a down step

Normally this should be extremely simple question...I found prob = 0.517 although this solution wasn't there...In the end the question must have been more tricky than I thought.

0.517 = 1 - (exp(0.02*(1/12)) - (exp-(0.35*sqrt(1/12))) /(exp(0.35*sqrt(1/12)) - exp(-0.35*sqrt(1/12))))

Any thoughts?

 

[sdonahue]

That is the probability of an upstep, the probability of a downstep is 1 - Probability of Up.

 

[Delo]

Hi,

I used binomial distribution:
250C6*(0.01)^6*(0.99)^244 ...the solution was in there...

That's how I did it too.

I had a problem with the 2 traders and rf = 0.02 (for 1 year) and sigma = 35 % (for 1 year) and delta_t = 1/12 ... (12 step binomial tree) get the probability of a down step
Normally this should be extremely simple question...I found prob = 0.517 although this solution wasn't there...In the end the question must have been more tricky than I thought.
0.517 = 1 - (exp(0.02*(1/12)) - (exp-(0.35*sqrt(1/12))) /(exp(0.35*sqrt(1/12)) - exp(-0.35*sqrt(1/12))))
Any thoughts?

This one really threw me off- it appeared pretty simple in first reading . Wasted a lot of time on this.

 

[ashwinashu90]

What would be the answer to the question on delta changing from 0.5 to 0.65? I got purchase 375 shares. Also, would the answer to the question on credit rating matrix- "at the point approach" be because it had lower diagonal matrix values?
What should be the minimum number of various positions involving call and put required to maintain delta and gamma neutral hedge? I wrote 2.

 

[ashwinashu90]

Ok now I know I am a bit tensed so pardon me for asking for answers to so many questions at a time.

 

[GVD]

In a risk-free world everyone should get the same up and down probability ....but the solution GARP provided was for the same views about 55 % and the other about 45 % (1-45%), the other 2 solutions where each trader dependent, this should have been irrelevant though

This must have been a tricky one!

 

[GVD]

@ashwinashu90
I think I had 375 as well, for internal credit ratings...yes lower-diagonal matrix, yes 2 is minimum number of options.

 

[ashwinashu90]

Thanks for the reply Mr. Genius!

 

[GVD]

That is the probability of an upstep, the probability of a downstep is 1 - Probability of Up.

Yes... I have taken care of that see my expression... "0.517 = 1 - ..." where ... is probability of up-step

 

[ashwinashu90]

I now see that the paper wasn't really tough. It was really very conceptual. If one had strong command over theory and the concepts, it wouldn't be too tough to crack. Atlleast the theory based questions were quite doable.

 

[sdonahue]

Yes... I have taken care of that see my expression... "0.517 = 1 - ..." where ... is probability of up-step

I apologize, I missed that. I guess my brain is still mush lol.

 

[Sidewinder]

I now see that the paper wasn't really tough. It was really very conceptual. If one had strong command over theory and the concepts, it wouldn't be too tough to crack. Atlleast the theory based questions were quite doable.

If you have a strong command over theory and concepts it is always "not tough". But too learn and understand concepts / formula (over 1200 sides) in deepness is obviously very hard and time-consuming (if you have a full-time job inclusive social interactivity).

@ashwinashu90
I think I had 375 as well, for internal credit ratings...yes lower-diagonal matrix, yes 2 is minimum number of options.

Obviously I understood the question false? Gamma-Vega-Neutral needs two Options, Delta-Gamma-Neutral needs Option / Shares, Delta-Gamma-Verga Neutral needs 2 Options / Shares, or am I not right?

 

[Nicole Seaman]

• It was a good balance of quant and qual Qs.
• Our site too had few issues - my booklet had just one seal and same was the case with few others. They changed my booklet. I pray they gave me a right numbered one though. Pencils thing didn't bother me much.
• Writing on the booklet was allowed. It seems to be a site specific issue.
• New topics were tested .. MBS, Diebold (at least 3 4 Qs) but everything was aptly covered in BT's note

Few Questions that I remember. Excuse the typos.
1. Box ljung chi square
2. TBA numerical .. positive carry and proceeds are negative/positive. Might be easy but I was unprepared for dollar roll calculation (frankly didn’t expect)
3. Calc UL give exposure
4. Multifactor beta with gdp and interest rate calc expected returns
5. Fx .. at least 3 questions
6. Calc sopt from swap rates
7. Role of audit committee in risk management - make sure risk mgmt is good
8. Role of CRO
9. Cov stationary series condition
10. 3 regression equation given.. stock vs oil oil vs tbill tbill vs oil.. was there any multicollinearity. Omitted variable
11. Adding another variable in regression increases rsq.. options were does it mean it more significant. .. easy one
12. One drawback of stress testing
13. Gini coeff and human development. .rank the countries
14. MBS calc cpr based on smm
15. Number of positions given asked whther it was calendar / butterfly
16. Easy one on pricing the bond
17. Garch graph was given x axis number od days . Y axis volatility. .. asked whether daily variabce is ... i think it was based on reading the co-ordinates from graph
18. Confidence interval .. it was long question but inc of 100 to 400 conf interval will decrease by half
19. Garch equation given then calc the today's volatility and interval for today index value ..
20. A question of raf basel recommendation
21. Which not a operational risk.. a trader entered wrong vol./ model wrong ..2 other options
22. Easy one on garp code of conduct
23. Which Dist used for loss frequency. .poisson
24. Alied and barings bank .. common trait
25. Stack and roll strategy.. when will you not make money if you are short
26. Delta normal.hedge.. when is it not effective
27. Credit rating matrix was given asked looking at that whether int. Rating used through the cycle or point in time
28. APT question..whether systematic risk in apt is zero or not... cant recall clearly
29. At least 3 questions on conditional / joint probability..miller
30. Regression. .t stat.. std error. When will u reject null
31. Corp bond...which one is most risk..debenture. equipment cert. Subordinate bond
32. Binomial tree. Vol was given. 2 traders has diff expectation by end of year. .6 and 12 percent. . In their models what would be their prob of down movement..
33. Cacl expected shortfall 100 days .. moving 10 days ahead and then recalc expectshorfall. One obs would drop out ..
34. VaR backtesting .. no of exceedances

@td, i think just rename the thread to "FRM Part 1 (May 2015) Exam Feedback"

Hello @ckat

Thank you for suggesting that td rename the post, however, that is the name that Bionic Turtle uses to get our feedback on the exam so that could become confusing if there are two threads with the same name. We try to keep all of the feedback in the same thread to make it easier for others who visit the forum, but there are a few threads that were started before I could post our feedback threads

 

[William_Jose]

• It was a good balance of quant and qual Qs.
• Our site too had few issues - my booklet had just one seal and same was the case with few others. They changed my booklet. I pray they gave me a right numbered one though. Pencils thing didn't bother me much.
• Writing on the booklet was allowed. It seems to be a site specific issue.
• New topics were tested .. MBS, Diebold (at least 3 4 Qs) but everything was aptly covered in BT's note

@td, i think just rename the thread to "FRM Part 1 (May 2015) Exam Feedback"

Hey ckat, could you post some of the answers as well which you remember? Or otherwise just pardon me if that goes against any rule or something that I am unaware of.

 

[riskaverse]

I am confident on about 50, calculated guess on 14-15, after that things got really bad for me. Can I still hope to pass?

 

[Sidewinder]

I am confident on about 50, calculated guess on 14-15, after that things got really bad for me. Can I still hope to pass?

Same for me. Sadly I could write down (out of mind) at least 40 slides with formulas, what a big waste of time

Btw.: These 4 blank spaces for 100 MC-Questions were far too few.

 

[andy88]

yeah same here, got about 80-95% correct on my practice exams, but the actual exam was much more difficult. so actually you fuck up if you just learn the formulas, you have to 1) understand and 2) know all execptions and assumptions

looking forward to the results on 23 june