Stuart Klugman, FSA, CERA, PhD is an SOA Staff Fellow. He taught actuarial science for 35 years at two universities and has worked with hundreds of potential actuaries. Stuart has also worked for many years within the SOA’s Education system. He has been involved as author or editor of three books used on SOA exams, has served on the SOA Board as member and as Vice-President, and is a two-time recipient of the SOA Presidential Award.
With a unique perspective on the exams and modules, Stuart will provide responses to questions we have received. If you have a question for Stuart, please send it to email@example.com.
What is known today regarding the Predictive Analytics assessment?
Details are still to be finalized. In particular, a contract is yet to be signed with a vendor to provide computer-based testing. However, we do expect that the assessment will be conducted at a computer-based testing facility using a vendor with sites throughout the world. The computer will provide access to Word, Excel, and RStudio, with a pre-defined set of R packages loaded. Candidates will receive a data set and a business problem and be given about five hours to analyze the data and write a report. Candidates will submit the report in Word along with other relevant files by the end of the session. There will be opportunities to prepare for the assessment through online learning that will include sample projects and solutions. It is expected that the assessment will be administered once in 2018 and two times in 2019. After that, frequency will depend on demand.
The project will be graded using the same processes used for SOA written-answer exams. Graders will use an outline for guidance in assigning points. Papers near the pass mark will be second graded and should there be significant differences, the graders will meet to reconcile them.
It is not clear at this time what sort of break will be available to candidates during the five-hour timeframe. This will depend on the ability of the selected vendor to ensure there is no opportunity to gain outside information during a break.
Candidates should be aware that the Statistics for Risk Modeling exam (SRM) is a formal pre-requisite for taking the Predictive Analytics assessment. Candidates who earn SRM credit through transition rules will be deemed to have met this pre-requisite. However, the assessment will assume full knowledge of the SRM material. Hence, candidates who receive transition credit for SRM will still need to learn any parts of the syllabus they have not covered.
See the ASA Curriculum Changes site to learn more about the Predictive Analytics requirement.
Can you provide information about the grading process for FAP end-of-module exercises?
When the Fundamentals of Actuarial Practice (FAP) modules began, only a selection of exercise submissions were graded. However, now every exercise that is self-assessed as meeting minimum requirements is graded. The exercise is checked against a grading outline and those that come up short are graded as not meeting minimum requirements. It should also be noted that all submissions are checked for plagiarism against other candidate exercises and the model solution. There are also checks to catch those who submit unrelated documents. Disciplinary action is taken against those identified in the plagiarism-checking process.
Exercise submissions are graded approximately seven weeks after submission. If a candidate’s exercise is passed the grade is reflected on the transcript, usually within 24 hours. If the exercise is failed, the candidate is notified via email that a different exercise must be completed for that module. After submission, candidates are expected to compare their work to the model solution to note any areas needing improvement. Individualized feedback is not available for end-of-module exercises.
Do hiring managers view passing exams differently from completing VEE requirements?
Exams passed are viewed as significantly more important than earned VEE credits in the eyes of hiring managers. As designed, VEE credits can ordinarily be earned in a known amount of time and are set at a relatively lower bar. With exams there is the uncertainty due to possible failure. So having exams passed versus attaining VEE credit typically leads to less variance in time to earning a designation. Also, employers believe passing exams is more predictive of future success than is earning VEE credits, due to the difficulty level and the greater quality control surrounding exams. While lack of VEE credits is unlikely to be harmful when applying for a job, it might be viewed negatively if you attended a school with approved VEE courses and either didn’t take them or didn’t earn the requisite grade. An explanation for the failure to earn the VEE credits would likely be needed.
How should I choose my fellowship track?
Before deciding how, it is important to know when. There is no point at which the SOA asks candidates to declare a track. This is done implicitly the first time a candidate takes an exam or module that is unique to a particular track. As an aside, it should be noted that candidates are not allowed to select components from different tracks. All fellowship work must be done within the requirements for a single track. Also, because there are no options to select prior to taking any track-specific component, there are no decisions to make regarding what to do prior to completing all the common requirements.
Nevertheless, at some point in your career a decision needs to be made regarding which track to pursue. Most candidates are already employed at this time and the decision is often a collaborative one between the candidate, the supervisor, and perhaps the director of the actuarial development program. There can be a range of possibilities. Some firms allow little or no discretion while others leave it entirely up to the candidate. Some candidates want the track to align well with the specific activities they are currently or soon plan to do while others use the track to expand their knowledge into areas they aren’t currently engaged with. Finally, note that one of the purposes of the Fundamentals of Actuarial Practice course is to provide grounding in all practice areas. This exposure may help when selecting your track.
Can you explain the role of pilot questions on the computer-based testing (CBT) exams and how they should affect my approach to the exam?
An advantage of CBT exams is the ability to provide instant results. This is made possible by knowing the difficulty level of each candidate’s randomly selected set of questions. To establish the difficulty level, a question does not count the first time it is used. Then the exam statistics are employed to set the difficulty level for future use. To make this work, candidates cannot know which questions are pilot (else they would be skipped). While most every test has pilot questions, the SOA does not make public the number of pilot questions on a given exam. And that number can change from exam to exam and from time to time for a given exam.
A candidate may look at a published 70% historical average pass mark for a 30 question exam and aim to get 21 correct. However, if the exam has four pilot questions, the pass mark is then 70% of 26 or 18 (rounded). A candidate who had the ability to know that exactly 21 questions were answered correctly might still fail if four of them happened to be pilot questions. Trying to guess the number of pilot questions or which ones on a given exam are pilot questions is a fruitless endeavor. I suggest a better exam strategy is to simply maximize the number correctly solved and not worry about pilot questions.
Is there value in pursuing a master’s degree for career changers or those without a job after getting a BA?
There are many schools that offer master’s degrees in actuarial science. Some programs are designed to provide enhanced education for those with a bachelor’s degree in actuarial science. Many others are designed for students who majored in something else, but have a strong mathematics background. Their goal is provide a concentrated dose of actuarial science courses that provide a solid education as well as promote rapid exam progress. These programs also often offer access to a career center, career fairs, and actuarial club resources. There is the possibility of teaching assistantships that may help with the cost of the program.
For those with a bachelor’s degree in actuarial science it may be worthwhile to consider a graduate program in an affiliated discipline, such as statistics, analytics, or finance.
What is the significance of the recent changes to the Code of Conduct for Candidates?
While no one actually asked me this question, I thought it worth bringing up. The major change is the addition of Rule 3: “An Actuarial Candidate shall act with courtesy and professional respect in all interactions with the SOA.” While it is natural to be frustrated and even angry upon learning that an exam or assessment was failed, a threatening or abusive email or call to SOA staff or volunteers is never appropriate. In a few cases, even after warning a candidate that continuing such behavior is unacceptable, the inappropriate behavior continued. This Board-approved change to the Code provides the ability to discipline such candidates. This change is not intended to discourage candidates from contacting the SOA with questions or concerns, but does demand that such contacts be done in a professional and respectful manner.
There has also been a slight change in the definition of a candidate. A person who contacts the SOA with an intent to register for an exam or assessment (even if such registration never takes place) is now considered to be a candidate and is covered by the Code.
Which computer skills are most important for an actuary? Will the R statistical package be incorporated into the SOA examination system?
Having computer skills and the ability to learn new ones are more important than mastering any particular language or software. My first computer languages were BASIC, COBOL, Fortran, and Pascal, none of which are in common use today. For a time, APL was the dominant program language for actuaries. Currently actuaries use a wide variety of tools. It is important to be able to organize large data sets (using, for example, SAS, Access, and SQL), to perform the analysis (for example, SAS, R, and VBA), and perhaps to run whichever actuarial software package is in use at your employer. An aspiring actuary cannot learn all of these, but experience with any of them will make it easier to learn those favored by your eventual employer.
The R statistical language has become very popular. It is free (a bonus for students) and it is open source (many packages have been designed for actuarial applications). The SOA currently requires candidates to learn and use R in its Applications of Statistical Techniques Module, a core component of the the General Insurance Track. An SOA task force is currently looking into expanding the coverage of predictive analytics in the ASA curriculum and it would not be surprising if R is incorporated.
What is the value of the CERA credential?
The CERA credential provides recognition that a candidate has added specific education in enterprise risk management (ERM). While there is basic ERM education in the Fundamentals of Actuarial Practice modules, the CERA components (ERM module and ERM exam) go into greater depth. The curriculum provides rigorous coverage of ERM processes and techniques, emphasizes both qualitative and quantitative skills, and provides education in both actuarial and other approaches to risk. More information is available at http://www.ceranalyst.org/overview.asp.
How do the extensions to the ERM exam work?
When registering for this exam candidates have a choice of extension, one for each track (Individual Life and Annuities, Group and Health, etc.). Extension-based questions comprise 25 percent of the exam. Each extension has associated reading material that relates to ERM within the track. Because some candidates take the ERM exam prior to committing to a track, it is not required that the extension match the track used for fellowship (however, all other fellowship components must be within the selected track).
When discussing multiple-choice exams with others, what is permissible and what is considered cheating?
The multiple-choice questions for the computer-based exams (P, FM, MFE and C) are drawn from a large bank of questions. It is imperative that the questions in the bank remain secret. If individual candidates obtain access to the questions prior to taking the exam they would have a clear advantage. If the questions become publicly available, the question bank would have to be rebuilt at great cost to the SOA and inconvenience to candidates (fewer exams per year and a two-month wait for results). Thus, any activity that would reveal specific questions to others is considered cheating and would result in significant penalties.
While there is no sharp line that can guide you with regard to exam discussions, the following might help. If a fellow candidate asks you about the exam you just took, any answer that reveals one of the questions asked is more than you should be sharing. If you are asking someone else about a question on your exam (perhaps to help you do better on a similar question next time) you should phrase your issue in as generic a manner as possible. It is not a secret that an Exam P question might ask you to calculate the probability that X + Y is less than something given a bivariate density function over an odd-shaped area. So asking how to solve such problems is fine. Presenting the exact problem of this type that was on your exam is going too far.
Does the MLC multiple-choice cutoff for grading depend on the performance of the other candidates?
None of our assessments are graded on a curve. It is possible for every candidate to pass (or in this case have their written answers graded), something that has actually happened. Tentative pass marks are set based on the level of knowledge expected from a qualified candidate. Candidate performance may lead the committee to adjust the tentative pass mark. But this would not be to increase or decrease the number of passing candidates to a desired level but rather because the grading process provided additional understanding with regard to how questions were interpreted.
With regard to the MLC cutoff, the committee has taken a conservative approach, setting the number lower than may be necessary to screen candidates. At the first administration, none of the candidates who hit the cutoff exactly (12 correct out of 20 questions) passed the exam.
More information on the MLC Exam
Is there a time limit for completing the ASA or FSA?
No. As a result, there can be considerable variation in travel time from the first exam to attaining an FSA. We know of one member who earned an FSA in three years and another who took 57 years. The average travel time is about eight years.
How does the choice of major, degree, and level of program affect my chances of gaining employment?
The short answer is that personal accomplishments are more important than any of the above. Important accomplishments include a high GPA, passing actuarial exams, having relevant coursework, being able to demonstrate communication and leadership skills, and prior work experience (which need not be an actuarial internship, though having one is more valuable).
That being said, university actuarial programs, regardless of the type of degree awarded, can make it easier to achieve those objectives. There is a more structured curriculum and exam preparation, student peers to work with, connections to industry and often a career center to facilitate the job search. For career changers, or those who weren’t prepared by their undergraduate experience, there are masters programs that provide a concentrated dose of actuarial exam-related courses and all the services a university program can provide.
One way to locate a university program is to check out the SOA's Universities and Colleges with Actuarial Programs (UCAP) list. UCAP programs along with the SOA’s Centers of Actuarial Excellence can also be found on the SOA Explorer Map. You can access the map by logging in to your My SOA account and find it under communities on the My SOA Services page.
Are there areas of actuarial work that are more math/stat oriented versus business oriented?
I copied this question as it was posed by a candidate. But I think the question raises a false dichotomy. All actuarial work is business oriented. Whether it is coding software, designing hedging strategies, or running generalized linear models, it will be done to solve your employer’s or client’s business problem. Thus an understanding of the business environment is essential for any actuary. A more useful distinction might be math/stat versus finance/investments. While all actuaries need some of both, it would not be unusual for an actuary to go deeper into one or the other. As a general rule (but there are certainly exceptions), actuaries working in short-term insurances (e.g., health and property) tend to focus more on data and modeling and less on finance and investments (because the funds are held for only a short period of time). Actuaries working in long-term insurances (e.g., life, annuities and pensions) focus more on managing assets and liabilities.