# Mean Reverting Pattern of Level III pass rate

I’ll let it here before the Result Day - so nobody could say I am trying to explain my exam performance with this figures.

So, let’s start from this official document: http://www.cfainstitute.org/programs/cfaprogram/Documents/1963_current_candidate_exam_results.pdf

As you can see, the Level III results (during the last 10 years) perform in a kind of mean-revert pattern. One positive outlyer was the last year (58%). Can we this year the passing rate will be less, even much less vs 2014 (see the pattern of 2006 - 2007 years)?

Yes, I think that is the case.

To begin with, let’s read something extremely interesting: Official CFAI Financial report for 2014.

On the page 16 we’ll read:

The CFA Institute business model continues to be primarily dependent on CFA Program exam fees and related revenue, which constitute more than 80% of total operating revenue (excluding investment income)

Now let’s turn to page 12 and we’ll see that the CFA Program (which counts for the main part of the CFAI earnings) was the only one which suffered from fee decrease.

After that the pass rate grew up and we had the following (http://www.bloomberg.com/news/articles/2015-06-04/cfa-exam-lures-record-crowd-of-candidates-seeking-career-boost):

A total of 159,889 candidates registered to sit for the exam’s three levels, an increase of 7.4 percent from a year earlier, according to a statement Thursday from the Charlottesville, Virginia-based CFA Institute.

So… The pass rate last year grew up, fee-paying candidates have been attracted and this year… well, this year we’ll defenetly see something like 46-48% pass rate.

Just my prediction.

Have a great summer before the result will come.

It would be interesting if we get a table for pass rates indicating how many time one had passed the exam before to pass it.

Just try to ask CFAI. I tried and was not given any figures of dropout rate or the average number of tries.

Ok I will ask, I suspect that L3 has a much higher number of retakers. It is not the same thing if 25% or 75% of the people who pass are retakers.

New Level I candidate demand tends to be the most elastic in response to industry conditions, whereas candidates who have passed the Level I exam, and thus have more effort and resources invested in the program, tend to exhibit less demand elasticity as they move through the remainder of the program.

in other words, if you fail, you’ll come to retake the exam

lol… just like https://youtu.be/YBHQbu5rbdQ

My prediction is in the best they will not give you the number, but will give you a written sales pitch (as it was with my question).Anyway, it’ll be interesting to know if they reply at all.

Regarding your first post: Last year’s pass rate was 54%, the mean of the last decades is 58%. So last year was still below the mean.

Last ten years were pretty much within a band of 10% (except 2007). That’s totally normal since you’ll always have candidates that are prepared a bit better or worse than the year before. Would expect this year’s pass rate between 45%-55%. Don’t think that there is any connection to CFA’s financial situation.

The 58% is the average from 1963… which included 94% passing rate in 1963…

The last 10 years average is 53% (see the last row in the table), with several outlying patterns.

It’s a question of taste… I would not suggest for a standard exam 10% band to be fear.

The rate willbe from 0% to 100%, thats for sure, but I think it will be much less than the 2013 - 2014 years.

We’ll see.

What’s up with the 76% pass rate in 2006? Definitely an outlier.

#TooManyGOODLUCKwishesin2006

I would be surprised if pass rate stays above 50% this year… I would say 46% pass rate for 2015

The pass rate is not mean-reverting (as indicated by the ADF test), but the annual change is a stationary ARIMA (0,0,1) process and the 1-step ahead forecast suggests a 7.5% (relative!) drop in the pass rate over last year. The 95 CI for the pass rate is 37.80% to 62.32% with the point estimate of 49.95%. However, the data is probably unreliable as structural change analysis suggests 3 breaks in 1969, 1994, 2006. You’re welcome.