What's new

Exam Feedback May 2021 Part 1 Exam Feedback

Nicole Seaman

Director of FRM Operations
Staff member
Subscriber
We hope that everyone did well on the FRM Part 1 exams over the past two weeks! :) We would love to hear any feedback that you have about the exam, especially any helpful information that you can provide about the new CBT format. How did the exam go? Did you encounter unexpected questions? How was the experience with all of the COVID guidelines? Thank you in advance for any feedback you can provide!
 

nickcentanni

New Member
Subscriber
I took my exam in Los Angeles with no issues, although my proctor informed me a candidate who came in before me had trouble loading their exam. CBT format was fine - you were able to flag/skip questions if needed. I didn't come close to filling out the full two sheets of blank paper provided. The proctor actually offered me more and I said it wasn't necessary (and for what it's worth, I tended to use a lot of paper when i did practice questions, but I always did my mocks with only two sheets).

This was discussed on other forums with mixed feedback, but at the end of the test, there was an option to click 'Results' and mine showed 70%, but I'm not sure how reliable that number is. We'll have to wait until July for pass/fail results.

As for the exam itself, it was very qualitative. Even a lot of the quant questions were in qualitative form. Definitely not as much calculator work as the Mock exams, but - in my opinion - if you really understand the concepts beyond the formula, you shouldn't have an issue with the exam. Candidates who only prepared by memorizing formulas and counting on a quant-heavy exam might be surprised. Overall I thought it was a challenging exam, but a fair exam. Being well prepared on all topics is key to narrowing down the answers on qual questions. Practicing my time management on my practice exams was incredibly important.

There was a lot of talk on Reddit about the exams being nothing like the GARP Mock, but I tend to disagree. There certainly weren't as many full quant calculation questions, but the question wording was similar and the qualitative questions felt very similar.

As for some specifics on my particular exam, I just sat for the exam on Friday so it's very fresh.

The quant questions I remembered were
  • CAPM / APT
  • Sharpe / Information Ration
  • a few conditional probability questions
  • hazard rates
  • calculating and interpreting a t-stat
  • finding a forward rate given spot rates
  • a butterfly spread payout
  • lower bound of an american call option
  • calculating the notional principal of a swap given a period's payment and the LIBOR / Fixed rate
  • EWMA
  • Long-run mean reverting level of an AR series
  • delta-neutral hedge
  • Expected Future Spot Rate given a currency and the rates of the countries
  • given a lease rate, storage cost, convenience yield and risk free rate and spot price, calculate the cost of carry of a commodity
  • total net payoff from a short stock position (including financing costs)
  • Cheapest to deliver bond (different than a lot of practice questions because they gave you additional information that wasn't needed)
  • Adjusting Beta of a portfolio
  • Poisson
Qualitative topics I remember:
  • A few on ERM
  • Enron / Corporate Governance
  • Volvo Case Study
  • Greeks
  • Binomial Tree vs BSM
  • OLS assumptions
  • BLUE
  • BSM assumptions
  • Country Risk
  • Bootstrapping Methods for Simulations
  • Antithetic Variates
  • CCPs
  • Forwards vs Futures
  • Properties of a lagged variable
  • Spearman's Rank / Kendall's Tau
  • Loss Frequency distribution
  • GARP Code
  • Risk Data Aggregation
I feel good about my preparation. Practice is key to gaining confidence and understanding concepts that prepare you for quant questions in qualitative form. Qualitative topics can't be avoided and is just part of the curriculum. I personally created note cards to try to hit high level points on all the topics. Those topics unfortunately can't be underestimated as much as many of us want a nice math-focused exam.

I hope to get good results in a few weeks and then hopefully it's on to Part 2! Thanks BT for your continued efforts and challenging course material!!
 
Last edited:

nc27

Member
Subscriber
Hi,

I think the above review is on point. I took the exam in Mexico City, I think that globally the test was fair but with some really poorly worded questions which let some space for interpretation (I remember two hypothesis testing questions awfully written). As mentioned above, a bunch of difficult qualitative questions were asked and only a mastery of the curriculum would allow one to get a really high final mark.

As per my individual score I got 69% (a bit disappointing but it could have been worse) , I hope it will be enough to make it through.
 

liyi1989

New Member
I took the exam in Charlotte. At the PSI test center, they only allow/provide to take one pencil and one paper (there are lockers in the lobby. If you need more, you can exchange a new paper with your old one. But since this time there are a lot of qualitative questions, I only use 75% of one side). No watch (needed) since it's just a web app, so it's just a window machine with browser (with clock in the Windows right bottom corner). I don't think there are a lot of candidates in NC as I think the people in the room are taking different exams, like nursing, etc with different length. Candidate can go out to restroom by giving them the ID at the front desk. Overall, the logistic process is very smooth. btw, by clicking sidebar, one can get normal table/index for the questions.

I have worked for ~ 3 years as a data scientist in risk management field. So my main goal to participate in the FRM exam is to learn domain knowledge instead of purely getting a certificate, which means I do not feel annoying when reading "corner cases/topics" in the text book. I almost only read GARP text books (purchased a Sony digital paper), as I think other study materials only summary "key points" which are very exam focus but cut a lot of descriptions on why we need to learn those topics/where does those topics come from.

Overall, I think GARP (text books) does a great job for me to learn the risk management concepts/terminologies. As I mentioned, I participate FRM because I came across various terms "KMV model", "economic capital", "risk aggregation", "market risk", "hedging", etc in my work. And even if I can google those things, I cannot find a good resource for me to learn them systematically (probably, John Hull's risk mgt book, but for most of GARP's text book, it almost directly copy John Hull's book, so the two are almost equivalent), until I paid FRM exam fee and saw GARP's FRM text book. I do admit that for people who wants to rush for passing the exam, reading the text books maybe lengthy. But it is definitely a good resource if you really work in the risk management field. For me, I need to deal with/understand credit risk, market risk, interest risk, insurance risk, risk aggregation, etc, so this exam is definitely a good learning experience for people like me.

This also means that, if you do not have a strong need in risk management (even if there are many risk mgt jobs in the market), like only want to get a certificates in finance to look for a job, some content maybe boring to you. And you may consider getting a (more general) CFA instead of FRM, which is a more specific one in risk mgt.

As for the exam, I think it is very qualitative, which is good. This is not because qualitative is easier, but this means that quantitative methods (statistics, machine learning, quantitative finance, time series analysis) are just TOOLS for risk management (but not the goal). The goal is just risk management (as the exam name suggested), which I think it (GARP's exam) is in the right direction. After all, it is not a final exam from a mathematical finance degree class/fixed income class/Tuckman book's exercises, etc.

With these being said, it is easy to understand why there are no BSM or binomial tree calculation problems. And the quantitative parts are not very complex, and all are served with the same spirit: quantitative methods are just tools. And regarding people saying "you need deep understanding of the materials", after the exam (with a lot of tricky qualitative questions), now I understand the "deep" here really means one not only need to understand the formula/calculations, but also "why" we need those concepts/what is the reason for risk managers need to care about them (but not others).

So, I think the (spirit of the) exam is close to the official practice exam. However, I do admit that, the practice exam almost never changes over years, there are definitely "new (and trickier qualitative) problems" in the exam.

Also, I do admit that the exam problems are not perfect, as
1. some may have multiple correct answers that may require guessing with "exam tricks"
2. some require additional background knowledge
3. some are trivial questions (if I understand correctly)
4. some problem may be wrong

For example,
1. Gaussian copula, they have statements like (i) it transforms into uniform distribution (ii) it assigns relationship with transformed variables.
I believe if you only use the GARP book/John Hull's book, which only introduces Gaussian copula model, you will see (i) is not correct since it transforms into Gaussian variables. However, if you know general copula theory/Skalr theorem, actually, it is (i) the spirit of the copula theory (and Gaussian distribution is just one type of copula built upon the uniform transformation). So, then you may need some "exam trick" to guess what the problem is trying to test you (as we do not have the opportunity to argue with the person who contribute the problem).
2. Some problem ask what events can be assigned with a probability. Even if this is a very simple (i.e. short) question, but, as a math major student, I know the problem is essentially testing probability space for sigma algebra. So it definitely requires people to know (a little bit) measure theory based probability theory (not only the non-math-major probability theory), otherwise one will be confused with this problem (and choose the wrong answer).
3. Similarly, there are definitions about nominal/real interest rate, which I think is out of the scope of GARP text book (but is in the exam). Maybe for the problem contributor, he is a CFA holder or Econ major. But for people like me who have no such background but only reading GARP textbook, I can hardly get those problems right.
4. Some asks a stock can go up/down/stable, what is the probability to go up&down. If I understand it correctly, this is definitely a trivial/stupid question, unless the problem contributor really wants to test us "mutually exclusive events in probability".
5. I met the mortgage default style Bayes problem. I am 99% sure there are no correct answer available in the list. So I just chose a closest one.
 
Last edited:

Eustice_Langham

Active Member
My two cents worth:

1. Was the exam fair? I really dont know if it was. Perhaps, but I think GARP should be testing methods/concepts that are present in the mock exams. I understand that GARP can test whatever they want, but lets be realistic here, the more obscure the testing material the less relevant is the certification. (I have spoken to MD's at bulge bracket banks and they either dont know what the FRM is or see it as not being particularly relevant...true). As an example, how many people need to know about hypothesis testing? Really....
2. Odd that some typical questions were not tested ie Put call parity, or a Binomial Option Model in what I thought and could be excused for thinking was an appropriate format. Can GARP test quantitative material in a qualitative format, yes, but it should be more evenly balanced, not make that the main focus. If GARP want to test these concepts in a qualitative format, fine, but dont ask students to remember the formula's.
3. Students remembering formulas is not such a bad thing, there was a duration/convexity question for instance, was it in INR?
4. From my memory, which doesnt appear to be as extensive as others on this post, too many questions focussed on some single learning objectives ie a disproportionate amount. I go back to mock exams, they need to reflect what is on the exam. I cant see any rationale for differentiating from this. Students are told to practice, practice, practice but if the practice questions are different then why bother.
5. CBT Testing is alot more efficient than paper. Being able to see on the RHS your answered/unanswered question in a single view then easily go to questions saved alot of time. Having multiple dates gives students more choice which is advantageous. The people running the test site were efficient and without any issues. To be able to enter an exam room, do some paperwork then sit down at your desk all within 3 minutes, takes alot of pressure off.
6. I finished with about 7 mins to go, didnt know that you could get your score.....I dont even know if its accurate.
7. Some forums are suggesting that some students/locations had different exams, I dont know if this is the case or not.
8. The entire syllabus needs to change, ditch some obscure topics, incorporate ESG/SCR into the FRM and if need be make it a three level exam.
9. EP like BT are hamstrung by this. EP are a valuable resource to GARP and their thinking needs to be at least listened to.

Im sure that alot will disagree with me. but GARP need to keep this thing relevant. It lost the ERP for this reason, and if I had taken the ERP and passed and payed all that money for it I would be seriously hacked off.

Good Luck to all
 

esteesse

New Member
Subscriber
I sat on ATA test hence it might be little different to you (I had put call parity and BSM in my question sets) but I do agree that there were some concentration of questions for some learning objectives. For example, I saw more than 2 questions on Sharpe and Treynor ratio and around 3 on ARMA. Considering that there are 60 LOBs and 100 questions, the questions could have been more evenly spread. Qualitative questions indeed went in-depth details and options were often 'partially incorrect' hence we need to know every details of the concept (contrast to mock exams)
 

Jagan.Ganti

New Member
Subscriber
I took the exam in Sydney and had a similar feeling about the exam being more qualitative (but no less rigorous). However, I felt I was better prepared thanks to the BT's relentless focus on throwing real life examples and methods.
 

Bluefox21

New Member
Are we allowed discuss specific questions?

I didn't enjoy the exam at all being honest. I thought plenty of questions were badly phrased (especially probability questions).

An example, would you classify Op Risk as a FLOD function? In my mind this sits in between FLOD and SLOD but none of the other answers made sense. I was probably over thinking!

I think 70% would be a definite pass. Sadly I didn't notice option to check results at the end.
 

Nicole Seaman

Director of FRM Operations
Staff member
Subscriber
Are we allowed discuss specific questions?

I didn't enjoy the exam at all being honest. I thought plenty of questions were badly phrased (especially probability questions).

An example, would you classify Op Risk as a FLOD function? In my mind this sits in between FLOD and SLOD but none of the other answers made sense. I was probably over thinking!

I think 70% would be a definite pass. Sadly I didn't notice option to check results at the end.
Hello @Bluefox21

Yes, the exams are over so you can discuss specific questions. That is helpful to members who will be taking the upcoming exams. Although the questions are not the same, seeing the concepts that were tested can be helpful during their studies.
 
I'm a bit disappointed with the exam. It was very qualitative and I feel like I barely had to use my calculator. Taking this into account, half of my studying for the exam was basically a waste of time. Some important concepts weren't asked at all and most of the questions that were asked I wouldn't relate to relevant real-world knowledge.

The fact that I could see my score at the end was interesting, but it just disappointed me further because I know it shouldn't be a passing score.
 

Parsimony

New Member
Subscriber
5. I met the mortgage default style Bayes problem. I am 99% sure there are no correct answer available in the list. So I just chose a closest one.
I also could not find a correct answer to the mortgage default problem. I spent at least 10 minutes at the end of the exam trying to find an answer, and I could not find one. I'm experienced in maths and stats. Did anybody get a solution to this that was in the listed answers?
 

nc27

Member
Subscriber
I also could not find a correct answer to the mortgage default problem. I spent at least 10 minutes at the end of the exam trying to find an answer, and I could not find one. I'm experienced in maths and stats. Did anybody get a solution to this that was in the listed answers?
I do not exactly remember the question but I found an answer that matched, I just follow the classical method (I sketched a decision tree)
 

liyi1989

New Member
I do not exactly remember the question but I found an answer that matched, I just follow the classical method (I sketched a decision tree)
I loosely remember it has 19% and 22% as choices. But somehow I got ~21%, if I remembered correctly.

I use store and recall, so I don't think I have rounding error.
 

Ellie_10

New Member
Subscriber
It was okay for the most part, but still a little frustrating for me. I got a caculation on the Euler Theorem, and a hybird HS VaR question which I thought should be tested in part ii (?).

Also, the time was tight I did not have the chance to review. I got two FULL valuation questions on swap, one for currency swap and one for interest swap, I only got the time to finish one of them..

I got mutiple questions on BSM with dividend and risk parity with dividend.

Some questions on fixed income but none related to term structure which I think is the most difficult part in part I.

Will update more if I recall something important.

Good luck guys.
 

Eustice_Langham

Active Member
I can only recall one irs question which wasn't even a full valuation.....none on BSM at all or put call parity for that matter......I wish someone could explain how it is that GARP can provide different questions I really do....this seems completely insane.
 
Top