If I recall correctly at the beginning of my preparation for level 2, I read somewhere that level 2 is scored based on how well we do against our fellow exam takers, and not based on raw score? (Whereas in level 1, if you scored 70%+ in each study session, you would almost certainly pass.) Can anyone clarify? I might’ve just been misreading late at night. And this was, as I mentioned above, quite awhile ago that I read this somewhere (might’ve been in a Schweser prep book). It’d be pretty unfortunate if those preparing for the exam were scoring 70%+ on mocks and samples, feeling comfortable, and ending up not passing just because everyone else performed just a bit better.
if you reach the 70% mark and then let up on your studying or think that you have made it and relax, you increase your failing chance imo. you shouldnt aim for any score, just try to score as high as you can. no matter what anyone says, there is NO WAY the CFAI is going to pass 80% of the test takers, even if 80% got over 75%. think about it, if something like that happens, we would never know and all they would do is say that the exam was too easy. its graded on a curve no matter what anyone tells you. study as much as possible. try your best, and when you walk into the test, you better feel confident that you are more prepared that 2 other people in a random sampling of 3 test takers.
What many people believe in is that, the passing grade is determined by 70% of the lowest score of the top 1% performing candidate. e.g. If the lowest score of the top 1% performing candidates is 100/120, the passing grade will be 70/120. In another word, it’s how you perform with respect to the rest of the candidates that matters. You can only be sure you pass if you are among the top 1% performing candidates.
That quite an interesting discussion. I just gave my sample CFA exam and did not do too well. CFAI says, no candidate has ever scored 100% How much score is a good chace? Specially since I need to work on my equity. Gets me in trouble every single time. As if there wasn’t Swaps!
One popular assumption is that the top score on the exam would be about 115/120. I believe someone already calculated that based on this assumption, in order to pass using the evaluating criteria above, you need to consistently get 4 out of 6 questions right per vignette. Nobody knows for sure how the evaluation process works. The above seems to be a good enough rule of thumb to gauge your readiness for D-Day. Some vignettes are pretty difficult to get 5/6 or above (like Ethics). Some vignettes are easier (like Quant, Econ and Corp Fin). You could use the above rule of thumb to allocate time on the preparation of each section. e.g. if you have already managed to consistently get 4/6 on Ethics, it’s probably a good idea to spend the remaining preparation time on other topics (like FI, AI or Derivatives).
I was going to take level 2 next year and was just starting this year in order to make sure I didn’t have to go too crazy next year. But if they automatically fail the bottom 50% of all test takers, does anybody want to pay for me to take the exam this year so I can fail it now and make your grade that much better? I promise to show up for the exam and put down random answers - hey, I haven’t even finished reading all the material yet!
SkipE99 Wrote: ------------------------------------------------------- > > no matter what anyone says, there is NO WAY the > CFAI is going to pass 80% of the test takers, even > if 80% got over 75%. think about it, if something > like that happens, we would never know and all > they would do is say that the exam was too easy. > > If this happens, I will be driving to the CFAI’s headquarters and start breaking every ethics code, line by line.
I am sure someone will pay you to fail the L2 exam this year based on a fees proportional to your weighted contribution to the exam results…
DanLieb Wrote: ------------------------------------------------------- > If this happens, I will be driving to the CFAI’s > headquarters and start breaking every ethics code, > line by line. New Standards of Professional Conduct: VIII. The DanLieb Rule It is considered unethical to use any offices of the CFAI in a manner consistent with a public restroom.
dlpicket Wrote: ------------------------------------------------------- > DanLieb Wrote: > -------------------------------------------------- > ----- > > If this happens, I will be driving to the > CFAI’s > > headquarters and start breaking every ethics > code, > > line by line. > > New Standards of Professional Conduct: > > VIII. The DanLieb Rule > It is considered unethical to use any offices of > the CFAI in a manner consistent with a public > restroom. lol. You guessed it.
eltia Wrote: ------------------------------------------------------- > What many people believe in is that, the passing > grade is determined by 70% of the lowest score of > the top 1% performing candidate. > > e.g. If the lowest score of the top 1% performing > candidates is 100/120, the passing grade will be > 70/120. > > In another word, it’s how you perform with respect > to the rest of the candidates that matters. You > can only be sure you pass if you are among the top > 1% performing candidates. Where does this come from?
Browse around one of the more recent threads and you should find the above assumption and theory. (Pay attention to threads that talk about how to determine the passing grade, or how much study is enough.) As I said, none of us knows for sure. As a rule of thumb, it seems to be a reasonable guess on how the actual evaluation works.
Top 1% all get 100% right Rest get 0% right
I think that’s the idea: they give 100% to the top 1% performing candidate and adjust everybody else accordingly. If this is true, the only way to be sure to pass the exam is to make sure you are in the top 1%. i.e. it’s all relative.
i went to the stalla review a couple weeks ago where they gave a taste of what the mock exam is like. the guy who taught it said he used to work at cfai, and he explained the mps like this: typically, if you plot the scores into a histogram, they wind up looking like two normally distributed samples placed next to each other (i forget the technical term he used). he said they set the mps around the point where the left tail of the higher scores meets the right tail of the lower scores. to me, this make some sort of sense in how they put people into bands, since they don’t say band 10 is the top 10%, but about the top 10%. so it’s possible that they break the bands up into fractions of standard deviations (i’m purely speculating though).
So a curve with two humps? This kind of make sense because we have people who are repeaters and there are people who are first timers. The mean of the humps could correspond to these two types of candidates. Basically, the passing score lies in the valley between the two humps. This would also minimize the population of failing band 10 and passing band 1 as well. Pretty good idea to avoid too many people doing face palms in August. XD