CFA level 2 retakers..primary reason for failure in first or previous attempts?


For those out there who failed level 2 (especially those who failed by a close margin)… what was the primary reason you think tipped you towards a band 8 or 9 rather than passing the exam?

My biggest worry (as with everyone) is failing by this small amount. Anything that needs to be done to prevent marginal failure?


I havent taken the test but I know what needs to be done…

STUDY YOUR ASS OFF!!! :slight_smile:

This is a good question. Does somebody with relevant experience have any feedback, please?

Do as many mocks as possible. Practice as much as possible. The volume is large so to get a hang of almost everything, practice will get you there. When you notice that there is a particular question you keep missing, take a break, use video lectures, read that section from the curriculum and do as many questions as available on it. It just might show up.

Do the CFAI EOCs and the online practice tests over and over again. The difficulty of the online practice tests is close to what you will get on the exam day.

Don’t be caught up still reading the curriculum by April. Practice! Practice! Practice!

I was band 9 last year. Basically, I ran out of time to study. About a month out from the exam I realized I was in trouble and simply not where I needed to be in terms of preparation. I knew it was going to be close going in to the exam based on the results from the two Schweser mocks I took (mid 60’s-low 70’s). I do think it could have gone either way but the afternoon session felt very tough and coming out I did not have a good feeling. Its a bit frustrating when you really grind on certain sub-topics and they do not get tested. But, that’s the breaks. It wasn’t one specific topic that tripped me up.

I would echo what others have said. Speed read the curriculum and spend your time doing questions and memorizing formulas and rules. Like I am doing right now! Held-to-maturity! Available for sale!

I passed all three levels on my first attempt, by wide margins. And, I have found the place where this is relevant: being an instructor at a CFA exam prep class at for a university/local society.

So, I had a whole 30 page presentation on this. I will give you some highlights. Schweser and 300hours both have decent lists of what they think cause failures. and

My list includes:

  1. Procrastination/running out of time/not enough hours. We spent a while discussing where to find the time.

  2. Skipping or not studying much on certain topics. Quant and foreign currency accounting seem to be common for this on Level 2.

  3. No study plan. I spent ten minutes on this in class.

  4. Not enough mocks or topic tests/practice problems.

  5. Not paying close attention to what your scores on practice, EOC, or mocks are telling you. I spent about 10 minutes on what to target, what the results mean, and what to do as a result. I have a very particular view on this, and it is not what you generally hear.

  6. Making overly extensive notes. Redoing problems you got right.

  7. Trying to target a score that is barely good enough to pass.

  8. Various problems the day before or day of the exam.

My presentation had about 30 sources of info. Not just my opinion. PM me if you want a copy.

Question on #5 in your list. For L1, I was told 70% on the mocks is good shape for passing. I was averaging 75-83% on the mocks so I felt good going into the test and felt great after the test.

For L2, what is the % on a mock that indicates good shape and great shape? Would it be similar, 70% for likely to pass and 80+ no problem?

True. But is not easy to reach over 70 % in Mocks especially in first round.

I failed in band 10 - I failed Portfolio Management and FRA.

I would say the number one thing that dragged me under was the lack of understanding the actual conceptualized underpinnings of the material rather than just being able to regurgitate a bunch of hieroglyphics.

For example - I could tell you all of the components to any formula on the exam - but understanding the process of how to derive certain figures is where I got lost.

last but not least - MOCK until you drop. I did 10 mocks last year, which I guess that wasn’t enough.

For 125MPH, i would need to do separate calculations for Level 2. However, for Level 1, my calculations were: if scoring at the mean on Schweser mocks you will probably barely fail. At avg + 4% you are about 50/50 chances of passing. At avg + 8% you are doing well, probably about 80% chance of passing. At avg + 12% you are doing very well, probably over 92% chance of passing.

There are tons of explanations, disclaimers, and data sources. One of them is people who score similarly across the topics have more stable overall scores.

Just curious, what was the mean for Schweser mocks for L1? I didnt see that published anywhere.

The mean for each Schweser mock shows up if you do it online. Same kind of thing for QBank. The average scores don’t move much over weeks or months. Common mock average numbers in the mid to high 60s. Since the difficulty varies, the average score varies from mock to mock. Same thing happens with actual exams. That is why it is more important to watch performance vs. mean than absolute scores.

Wow didnt know schweser shows those scores until now. I just looked at the performance trackers :stuck_out_tongue:

I got band 9. Mainly cause I took Ethics lightly and did not practice questions.

So here’s the question…

For the real exam vignettes, will there be questions where if u miss one of them, everything after that is wrong? For example, if they ask to do a calcuation for First year FCFE… then the next question, they ask for the Terminal year FCFE… then the final question is to find the value of the equity of the firm. If you missed the calcuation on the first two, you basically miss the final question too.

Will those be common on the exam?

Or will you be able to miss one and still get all the other ones. Like if they say “Assume for this question only, FCFE year1 is 500”… that would be a way to allow you to miss one question and still get the follow-up questions right?


Each question stands alone.

To me, it is all about percentile rank and average time per problem as recorded on the CFAI practice platform. You obviously need to get a good hit rate, but if you are scoring well on a large enough sample (ideally all ~1000 practice problems, at least for L1) you should be all set assuming you did not take sweet time with doing practice problems and reading curriculum or formulas while doing the practice problems. Furthermore, hit rate and percentile ranks are heavily correlated but since average passing score is not disclosed by CFA, proper metric to look at is percentile rank. If pass rate is 47% for L2, than you need to be above 53rd percentile.

My experience was while doing official practice problems if you do not read into materials and take less than average time than needed on the exam, you will be all set with good percentile rank.

The claim above is for the following reason. My percentile rank prior to the exam day as recorded by CFAI was 75th without resetting/redoing any practice problems while my average time for each practice problems was about 45 seconds on average.

On the actual exam day, I finished both AM and PM sections 90 minutes early which is in line with time average time on those ~1000 problems. My final percentile rank on the actual exam was negligibly below 90th and I suspect difference between the two scores is for the following reasons:

  • Some people reset scores while others do not. This causes some upward bias for those who reset since they know the problem
  • People who visit CFAI practice platform are more likely to pass the test - this causes downward bias on those who utilize the platform

Also note that based on confidence intervals provided by CFAI, my score most likely would be at very roughly speaking +/- 3 percentiles from test day score, meaning that there is wide discrepancy between what was recorded on CFAI practice platform and official score.

I tend to perform well on stress but in my opinion 15 percentiles is way too much of a difference to account for individual performance. Rather, it is likely to be due to two reasons outlined above.

Active learning - practice problems from multiple sources and refer back to curriculum to learn how to answer questions that you missed. Don’t depend on passive learning - watching videos and reading text.

Don’t get bogged down in esoteric questions.

Ensure that you can explain concepts to yourself. Refer to LOS.

Blue box examples for every reading.

Mocks from IFT were directly from curriculum. Repetition is helpful.

Know your equations and be able to diagram and chart payoffs to reason relationships.

Remain focused and plan to peak one week before exam. Don’t give up.

Torsten, you think the range is +/- 3 percentiles? Not 3%? Not a SD of 3%? The latter, would lead to a confidence interval 12% wide on the actual exam.

Not sure if you were able to see the newest report style that kicked off starting from December 2017 but CFAI provides estimate where your overall score would be given that exam re-take was similar to the actual one. The provided up/down score range is rather small (certainly far from above mentioned 15 percentiles) and CFAI does not disclose neither whether the range represents some confidence interval nor the confidence interval . But if the confidence interval was small, then what would be the benefit of such data?