While I believe that my seminars give the attendees the tools they need to pass the exam, the fact is that good pass rates are the result of the hard work of the participants—especially in the weeks between the end of the Seminar and the exam itself—rather than of the Seminar instructor.
I’ve been teaching actuarial exam-prep seminars since 1994, and over the years have had pass rates as high as 80% and as low as 20%. Only a few participants have reported getting a score of 9 or 10, and I’ve always believed that they would have easily passed without attending my seminar. I doubt that the pool of folks attending seminars contains many of the people who get high passing scores. In fact, the person who invented the seminar business told me once that seminars are really aimed at the candidates who probably would otherwise get a 4 or 5 and just need a boost to get a 6 or above. If indeed the pool of seminar attendees does not contain many of the high-scoring candidates, then I’ve always felt that my seminars would be successful if the pass rates equalled or exceeded the overall worldwide pass rate.
For many years I posted detailed data for all my seminars (how many took each session and how many passed the exam at the following sitting), in comparison with the overall SoA/CAS pass rates, in the hope that other seminars would do the same and that this data would help people in choosing seminars. To my surprise, I was repeatedly criticized (by a competitor in the seminar business) for posting information he considered misleading. As something of a knee-jerk liberal who believes that full disclosure is good for the consumer and the market, I was puzzled by this but decided to remove the detailed results from my website in 2009.
After a little more thought, I decided that it might indeed be misleading to show individual values of the pass rates––after all, individual values of a random variable indicate little about its mean. But sample means are good estimators of the true mean, so I'll give you the following data:
in all my seminars for the various versions of and syllabi for this exam since 2003, during which time the core material on so-called "loss models" has changed only modestly, the percentage of my seminar attendees who passed the exam at the sitting immediately following my seminar has averaged 1.15 times the worldwide pass rate released by the SOA and CAS.