How to Pass SOA Exam FAM in 4 Months
SOA Exam FAM (Fundamentals of Actuarial Mathematics) is a 3.5-hour, 34-question multiple-choice exam delivered by computer, and it passed 57.0% at the July 2025 sitting on a 60% pass mark (SOA). It bridges the short-term insurance math from the old STAM and the long-term life contingencies from the old LTAM into a single foundational exam.
FAM is the exam that surprised me the most as a mentor. Candidates who breeze through P and FM often hit a wall on FAM because the material is twice as wide: half the paper is short-term insurance mathematics and half is life contingencies, and almost nobody arrives strong on both. The math isn't harder than P. The breadth is what trips people up.
Here is a 4-month plan and the three pitfalls candidates underestimate.
What FAM Covers
The July 2026 syllabus lists ten topics with weight ranges (SOA). They fall into two halves:
- Short-term half (roughly 35 to 60%): short-term insurance and reinsurance coverages (5 to 10%); severity, frequency and aggregate models (12.5 to 17.5%), including deductibles, policy limits and the (a,b,0) and (a,b,1) frequency classes; parametric estimation by maximum likelihood, including grouped, truncated and censored data (2.5 to 7.5%); introduction to credibility, limited-fluctuation only (2.5 to 5%); pricing and reserving for short-term coverages, meaning expected loss ratio, chain-ladder and Bornhuetter-Ferguson reserving plus ratemaking (10 to 15%); and option pricing fundamentals, meaning binomial trees, Black-Scholes and put-call parity (2.5 to 7.5%)
- Long-term half (roughly 40 to 60%): long-term insurance coverages and retirement programs (2.5 to 5%); mortality models, meaning survival functions, force of mortality and life tables (10 to 15%); present value random variables for insurances and annuities (12.5 to 20%); and premium and policy value calculation (15 to 22.5%)
- What is not on FAM: Buhlmann credibility, Kaplan-Meier and Nelson-Aalen estimators, multiple-state models, and forwards and futures are not on the FAM syllabus; you meet them, if at all, on ALTAM, ASTAM or in older study manuals. Do not spend FAM hours on them.
The weights shift slightly each cycle. Check the current syllabus before exam day.
FreeFellow offers over 1,040 free practice questions for Exam FAM, with detailed solutions, three difficulty levels, and the FAM formula sheet covering severity distribution moments, life contingency commutations, and credibility formulas.
The Format: 34 Multiple-Choice Questions, Two Halves
FAM is entirely multiple-choice: 34 questions in 3.5 hours, five answer choices each, delivered as a computer-based test in the same style as P and FM. There is no written-answer section; the SOA trimmed the paper from 40 questions to 34 starting with the November 2024 sitting, and the written-answer work arrives later at ALTAM and ASTAM. The SOA also provides tables at the exam (abridged inventories of discrete and continuous distributions, plus life and decrement tables), so learn to read them at speed instead of memorizing what they contain.
What catches candidates off guard is the split, not the format. Half the questions come from the short-term side (loss models, credibility, reserving, options) and half from the long-term side (mortality models, insurance and annuity present values, premiums and policy values). A candidate who liked P and skimmed the life-contingencies chapters walks in with 17 questions of unfamiliar territory.
FAM is 34 MCQs, not a hybrid paper. The trap is coverage: drill both halves to the same standard, and practice reading the provided tables under time pressure.
A 4-Month Plan
This plan assumes 18 to 22 hours per week, roughly 3 hours per day, 6 days per week. Over 16 weeks that's 290 to 350 hours, which lines up with what passing candidates typically report.
Months 1 and 2: Build Foundations (130 to 175 hours)
Cover the bedrock topics: severity distributions, frequency distributions, and basic life contingencies. Spend the first two months going deep on the math.
Severity models: know the PDF, CDF, mean, variance, and limited expected value E[X^k] for exponential, Weibull, Pareto, lognormal, and gamma. Practice setting up policy-limit and deductible problems. The exam loves layered coverage problems where you compute the expected payment under multiple deductible-and-limit combinations on the same severity distribution.
Frequency models: Poisson, binomial, negative binomial. Know the conditions under which Poisson approximates binomial, when negative binomial is appropriate (overdispersion in count data), and how compound distributions work (S = X_1 + X_2 + ... + X_N where N is the frequency variable). The (a,b,0) recursive class is testable.
Life contingencies: survival functions, force of mortality, life table mechanics, expectation of life, basic life insurance and annuity present values (whole life, term, endowment, deferred annuity). Memorize the relationship between insurance and annuity functions: A_x = 1 - d * a_x for any whole life on x.
Work 15 to 20 MCQs per day in topic mode. Don't time yourself. Focus on setup and accuracy.
Month 3: Practice Problems Heavy (60 to 80 hours)
Switch to mixed-topic practice. Add limited-fluctuation credibility, parametric estimation by maximum likelihood, and short-term reserving (expected loss ratio, chain-ladder, Bornhuetter-Ferguson) with ratemaking. Add the option pricing content (binomial trees, Black-Scholes, put-call parity).
Aim for 65 percent accuracy across all topics. Every question is multiple-choice, but the ones that separate candidates are multi-step: a policy value that needs a premium first, a Bornhuetter-Ferguson reserve that needs an expected loss ratio first, an MLE on censored data that needs the right likelihood first. Practice writing the setup on scratch paper before touching the calculator; on a five-choice exam the wrong intermediate value usually matches a distractor.
For every multi-step item, follow the same template on your scratch paper:
- Restate the variables you'll use, with units (e.g., "Let X = severity, with X ~ Exponential(theta = 1000)")
- Write the formula or equation you're applying
- Substitute values
- Compute and box the final answer with units
This is the same discipline that carries into the written-answer exams (ALTAM and ASTAM), where the graders reward visible structure. On FAM it protects you from distractors and makes it obvious when a number is off by a factor.
Month 4: Timed Mocks and Targeted Review (60 to 80 hours)
Take a full 3.5-hour timed mock every 7 to 10 days. Review wrong answers and weak topics between mocks. Target 70 percent or higher on your final two mocks before exam day.
FreeFellow's practice exam feature builds timed multiple-choice mocks from the FAM bank and tracks your readiness score over time. Aim for a readiness score of 7 or higher before your exam date.
Between mocks, do focused topic drills on whichever areas your last mock flagged. The point of a mock isn't to confirm you're ready. It's to surface the areas where you're not, while there's still time to fix them.
The Three Pitfalls That Derail Candidates
Pitfall 1: Drilling Only the Half You Like
Candidates coming from P gravitate to the short-term half (loss models feel like more probability) and treat the life-contingencies chapters as reading. Candidates coming from FM do the reverse, since annuity present values feel like more time-value work. Either way, the neglected half is 40 to 60% of the paper, and it shows on exam day.
Fix this by tracking accuracy per half from week 1 and refusing to let either fall behind. FreeFellow's topic analytics split the FAM bank the same way the syllabus does; if your long-term accuracy is 15 points below short-term in month 2, that is where the next 20 hours go.
Pitfall 2: Rusty Credibility and MLE Mechanics
Credibility and parametric estimation look like small slices of the syllabus (2.5 to 5% and 2.5 to 7.5%), and candidates leave them for the last week. FAM credibility is limited-fluctuation (classical) only: the full-credibility standard from the confidence and tolerance parameters, and the partial-credibility factor Z as the square root of the ratio of observed to required exposure. Parametric estimation is maximum likelihood, and the exam likes complete grouped, truncated and censored data because each changes the likelihood you write down.
Most candidates can recite the formulas. Few can set up the likelihood for a right-censored sample or convert a full-credibility standard from claim counts to aggregate losses under time pressure. Drill 20 to 30 problems across the two topics before exam day, and write the setup out each time.
The most common credibility setup the exam tests: you are given a tolerance (say within 5% of the true mean) and a confidence level (say 90%), asked for the number of claims or exposure needed for full credibility, and then for the partial-credibility factor Z when the observed volume falls short. Practice the full chain, including converting the standard between claim counts and aggregate losses, not just the formula recall.
Pitfall 3: Option Pricing Feels Like an Afterthought
The option pricing content (a one- or two-period binomial tree, the Black-Scholes price and delta for a simple European option on a non-dividend asset, and put-call parity) feels like a leftover from FM. Candidates skim it. Then they sit a question that hands them u, d and the risk-free rate and realize they cannot build the risk-neutral probability quickly.
The section is small (2.5 to 7.5%) but it's free points if you've drilled it. Spend 4 to 6 hours on option pricing in month 3. Memorize put-call parity (C - P = S - PV(K) for a non-dividend stock), the risk-neutral probability in a binomial tree, and the Black-Scholes inputs. Drill 10 to 15 problems until the tree and the formula are automatic.
Don't skip topics because they're a small percentage of the syllabus. Every question carries the same weight, so a 2-minute option-pricing item you drilled is worth exactly as much as a 9-minute policy-value item you sweated over.
Practice Volume Benchmarks
Over the 4 months, target the following totals:
- MCQs completed: 800 to 1,200
- Multi-step setups written out on scratch paper: at least 100, so the setup habit is automatic
- Full timed mocks: 4 to 6
- Final readiness score: 7+ on FreeFellow's 0-10 scale
FreeFellow's analytics dashboard shows your accuracy by topic and difficulty. Use it to spot which areas need more drilling.
Exam Day Tips
Pace yourself. 3.5 hours for 34 items is a little over 6 minutes per question. Some take 2, some take 9. If an item is past 10 minutes, pick the best remaining choice, flag it, and move on; the CBT interface lets you return.
Write the setup for every multi-step item. Even on multiple-choice, the setup is what keeps a premium-then-policy-value chain from landing on a distractor. If a final answer matches nothing on the list, the setup is where you find the slip.
Use the calculator wisely. Only the models on the SOA approved list are allowed (the TI BA II Plus and Professional, the TI-30XS and XB MultiView, the TI-30Xa, the TI-30X II, and the discontinued BA-35). Practice the moment and present-value computations on the model you will bring until they are automatic.
Sleep before the exam. Cliche, but candidates underestimate this. The 3.5-hour test format requires sustained focus. Pulling an all-nighter the day before is a strategy that has failed thousands of candidates and worked for none.
I used the BA-II Plus for FAM and every other actuarial exam. Learn the TVM keys, the nCr/nPr functions, and the natural-log/exponential keys cold.
Free Resources Worth Using
- FreeFellow Exam FAM Practice, 1,040+ free questions with solutions, three difficulty levels, full-mock generator, readiness scoring
- FAM formula sheet, severity moments, life contingency commutations, credibility, derivatives put-call parity
- SOA Sample Questions, official samples with worked solutions, written by the same committee that writes the exam
- Klugman, Panjer, Willmot, Loss Models, the canonical reference text for severity and aggregate model topics
- Dickson, Hardy, Waters, Actuarial Mathematics for Life Contingent Risks, canonical reference for life contingencies
A Word on Background and Pacing
If you passed P and FM in the past 12 months, the FAM material won't feel foreign. Severity distributions extend the probability work from P. Time-value calculations for life annuities lean on FM mechanics. Plan for 12 to 14 weeks instead of 16.
If P and FM are more than 18 months in your rearview mirror, plan an extra 2 to 3 weeks for review of probability fundamentals at the start. The exam moves fast and assumes you can integrate density functions, manipulate expected values, and apply time value without reaching for the textbook.
If you're coming to FAM without P or FM (the SOA doesn't require them), you'll need closer to 350 to 400 hours and a 5-month timeline. The math compounds: severity models assume probability fluency, life contingencies assume time-value fluency. Skipping P and FM saves the exam fees but costs study time.
Your Path to Passing FAM
FAM isn't harder than P or FM in raw math difficulty. It's harder because it covers twice the ground and asks you to be equally fluent in loss models and life contingencies. Build the foundations in months 1 and 2, drill volume in month 3, run timed mocks in month 4, and don't underestimate the three pitfalls.
Start your prep today with free Exam FAM practice questions on FreeFellow and the FAM formula sheet.