Sitonce
Country: US
Show exams for United States Hong Kong
Sign in

SOA FAM Master Guide 2026

Updated 11 min read
Key takeaway

SOA FAM is a 210-minute actuarial examination with 34 five-option multiple-choice questions.

  • The October 2026 syllabus spans short-term insurance models, credibility, ratemaking and reserving, options, mortality, life-contingent values and premiums.
  • A few pilot items are unscored, but SOA does not publish their exact count.
On this page16 sections
  1. What FAM covers
  2. Current exam format
  3. How to use the syllabus
  4. Assumed knowledge and registration
  5. Registration and dates
  6. Tables, calculators and test-day rules
  7. Scoring and results
  8. An original worked life-contingency example
  9. Plan preparation around model selection
  10. What follows FAM
  11. How the short-term models connect
  12. Life-contingent values have their own dependency chain
  13. Read the October 2026 syllabus as a range
  14. Use practice as diagnosis
  15. Final readiness across the whole syllabus
  16. A day-by-day test-day plan
Questions
34, including unscored pilot items
Time
210 minutes
Current standard U.S. fee
USD 408
October 2026 syllabus
Effective October 2026

What FAM covers

SOA’s Fundamentals of Actuarial Mathematics examination tests actuarial methods for modeling, short-term ratemaking and reserving, and contingent payment models used in insurance and financial risk. It brings general insurance and life-contingency mathematics into one exam. The current full exam has replaced the transition-era FAM-S and FAM-L routes; candidates who did not earn full credit by July 2024 now take the full FAM.

The October 2026 syllabus is the scope authority for this guide. It contains ten topics, each with a weight range rather than a guaranteed item count. Some questions can address more than one outcome. Use the ranges to plan coverage, not to predict the exact mix on your form.

Current exam format

The exam is three and one-half hours with 34 multiple-choice questions. Every item has five choices, A through E, and one correct answer. A few randomly placed pilot questions are unscored, but the current syllabus does not state how many. All unanswered questions are incorrect, so answer every item.

FAM is administered by computer for most candidates, while the schedule separately lists limited paper-and-pencil administrations. Do not assume every center or window offers both. CBT forms are selected from a question pool and statistically scaled so forms in the same testing period have comparable content and passing criteria.

How to use the syllabus

The ten topics span short-term coverage and reinsurance; severity, frequency and aggregate models; parametric estimation; classical credibility; short-term pricing and reserving; option pricing; long-term products; mortality; present value random variables; and premiums and policy values. Topic 9’s former broad estimation content is no longer current scope; parametric estimation is Topic 3.

The largest published range is Topic 10 at 15-22.5%, followed by Topic 9 at 12.5-20% and Topic 2 at 12.5-17.5%. Do not turn a percentage range into an exact item count from 34 presented questions. Pilot items and overlapping learning outcomes also make exact counting inappropriate.

Assumed knowledge and registration

The syllabus assumes probability, mathematical statistics and financial mathematics as covered in Exam P, VEE Mathematical Statistics and Exam FM. This describes background knowledge. It does not by itself say those credits are formal registration prerequisites. Check the FAM registration page for administrative eligibility and your ASA transcript for whether those components remain in your personal pathway.

The standard U.S. fee for 2026 is $408. Eligible full-time university students may declare status for the $326 FAM student rate. The registration fee is nonrefundable and cannot move to a future period. Prometric appointment changes and travel can add separate costs.

Registration and dates

For CBT, register with SOA first. After SOA processes the registration, the acknowledgment email supplies a candidate number; schedule the appointment separately through Prometric when the number is active. Update your My SOA record and match your first and last names to your identification. The appointment confirmation is a separate record from the exam registration receipt.

The October 23-29, 2026 CBT window is closed for registration. SOA lists February 18-24, July 13-19 and October 22-28, 2027, with deadlines January 19, June 15 and September 21 respectively. The registration page says exam registration closes at midnight U.S. Central Time. A published window does not guarantee a preferred Prometric seat.

Tables, calculators and test-day rules

The October syllabus says official distribution, life and decrement tables are available in the CBT under an Exhibit button and a normal-distribution calculator is included. Personal copies of tables are not allowed. The SOA separately publishes a short list of approved physical calculator models; confirm the current list before test day.

The SOA rescheduling page permits FAM appointment changes within the same window through noon on the fifth business day before the scheduled exam, subject to seat availability and any Prometric fee. The published fee schedule is no charge 31 or more days before, $70 at 5-30 days, and forfeiture inside five days, for a no-show, or arrival more than 30 minutes late.

Scoring and results

SOA uses statistical scaling to keep forms comparable. In the February 2026 CBT results report, the administration’s pass mark was 60.0% average correct. That is a dated result for one sitting, not a permanent cutoff to apply to later forms. Do not estimate a fixed number of misses or translate a practice quiz percentage into an official result.

Starting with February 2026 FAM CBT administrations, candidates receive an unofficial pass/fail email from Prometric within an hour. For October 2026, SOA currently lists candidate-number release on January 8, 2027 and transcripts on January 11, 2027. An unsuccessful CBT candidate receives diagnostic feedback, but SOA cautions that a small number of items can move a topic rating and that feedback is independent of grading.

An original worked life-contingency example

Suppose a life table has lₓ=1,000, lₓ₊₁=980 and lₓ₊₂=950. Then one-year survival from age x is 980/1,000=0.98. Two-year survival is 950/1,000=0.95. Conditional survival through the second year is 950/980≈0.9694. Death during year two as a fraction of the original cohort is 30/1,000=0.03, while the conditional probability of death in that year is 30/980≈0.0306.

The distinction is the conditioning set. A life table counts survivors at each integer age, while a conditional probability starts from those alive at the beginning of the interval. Before calculating, write the starting cohort in the denominator. That habit prevents a common error in survival, mortality and annuity questions. These invented life-table values are examples, not published mortality rates.

Plan preparation around model selection

Start with a diagnostic across all ten topics and assumed knowledge. For each error, decide whether the gap is probability/statistics, selecting the model, writing the cash-flow timeline, applying a formula, or arithmetic. Learn a concept, work an original example without looking, and then test it in a mixed set. Topic-by-topic practice helps first learning; mixed practice tests whether you can recognize which method applies.

Use the SOA syllabus, listed readings, official tables and sample questions. Verify current notation because an older note is not automatically the October 2026 edition. For each calculation, define symbols, state assumptions, show timing and use units. Review every distractor that was tempting, not only the selected answer.

What follows FAM

A FAM pass earns exam credit, not the ASA designation. The current ASA path includes other exams, VEE topics, modules, an assessment and the Associateship Professionalism Course, followed by an approved application for admission. Candidates have flexibility in the order of components, but should use their own SOA transcript to track remaining requirements.

FAM is also not the final word on actuarial practice. Its syllabus checks technical foundations in modeling and insurance mathematics. A candidate should avoid describing one exam as a complete professional qualification or as a guarantee of a particular job. Use the SOA pathway page for current designation rules and keep employment expectations separate from exam facts.

How the short-term models connect

The first five syllabus topics form a chain. Frequency describes how many claims arise; severity describes the size of one claim; their combination creates an aggregate loss distribution. Parameter estimation supplies model inputs, and credibility addresses how much weight to give limited experience. Pricing and reserving then use different views of that uncertainty. A premium is set before or during coverage using a pricing basis, while a reserve estimates obligations attached to business already written. Keeping these jobs separate prevents a common error: treating an observed paid-loss total as if it were automatically a future premium or reserve.

A practical model map starts by writing the random variable and time period. For one policy, let N be claim count and X_i individual severities; aggregate loss S is the sum from i=1 to N of X_i. A reserve may instead use reported development by accident period, while a rate calculation may adjust historical losses for exposure, trend, development and expenses. The syllabus moves among these representations. When a problem changes from per-claim severity to aggregate annual loss, pause and identify whether the asked quantity is a probability, expected value, percentile or rate.

Life-contingent values have their own dependency chain

Topics 8 through 10 build from survival probabilities to cash-flow values. A mortality model gives probabilities of death or survival over an interval. Those probabilities weight benefits that occur only while a policy is in force or upon death. Discounting moves each conditional payment to a common valuation date. Premium principles then compare these benefit values with premium cash flows. In a level-premium example, premiums may stop at death even when the benefit is payable; forgetting that condition changes the premium present value and therefore the answer.

For a discrete term insurance benefit B paid at the end of the year of death, the present value random variable assigns Bv^k when death occurs in year k, with v=(1+i)^-1. Its expectation sums Bv^k times the probability of death in that year. A life annuity instead pays at specified survival times. The same mortality table and interest rate can therefore produce different values because payment timing and contingencies differ. Make a timeline before applying a formula, especially when the statement uses “immediate,” “due,” “temporary” or “deferred.”

Read the October 2026 syllabus as a range

The published topic percentages are ranges, not a promise that each sitting will allocate an exact midpoint to every topic. Their total range permits different combinations across forms while maintaining the blueprint. Use the listed outcomes to decide what knowledge belongs in a review set, and retain the exact October 2026 version when comparing notes. Older FAM material may reflect different topic language or a prior syllabus. In particular, the October 2026 credibility outcome is limited fluctuation credibility; broad credibility texts can include more methods than this exam outcome requires.

A useful tracking sheet has one row per syllabus topic, a concise list of examinable actions, and a confidence column based on closed-book problems. “Know mortality” is too vague. “Derive a one-year death probability from l_x and identify whether an annuity payment is made at the beginning or end of the year” is testable. Mark errors by cause: model selection, algebra, calculator, timing, or interpretation. If errors cluster in one action, review that action rather than rereading an entire chapter.

Use practice as diagnosis

A multiple-choice result alone does not reveal why an answer was wrong. For each missed item, record the model, the quantity requested, the assumption that mattered, and the first line where your work diverged. If you selected an aggregate-loss formula for a per-claim question, the repair is model recognition. If your setup was sound but the discount exponent was off by one, the repair is timeline discipline. This short error record makes later practice more useful because it distinguishes content gaps from execution errors.

When reviewing a correct answer, still compare the distractors. A distractor may encode a plausible mistake, such as using unconditional expected payment where a conditional mean is requested. Write a one-sentence reason each incorrect choice fails. Then solve a nearby variation with a changed deductible, payment date, or survival probability. This checks whether you understood the structure instead of memorizing one numerical result. Keep generated examples clearly separate from official questions; the examples in this guide are original teaching illustrations.

Final readiness across the whole syllabus

A balanced readiness check samples every topic, but it should not give every topic equal time automatically. Use the current published range as one input and your error record as another. A topic with a lower range can still expose a serious gap; a larger range does not justify neglecting basic accuracy elsewhere. Rotate among short-term and long-term problems so that a familiar formula does not conceal weak switching between models. On a timed set, include the official-style reference tables and an approved calculator where the practice environment supports them.

Before test day, confirm the selected administration, booking details, identification match and the current SOA policy pages. The registration fee, scheduling deadline and permitted appointment changes are operational facts that can change between windows. Content readiness and administrative readiness are separate checks. Keep the syllabus version attached to your study notes, and avoid projecting one sitting’s empirical pass mark or pass rate onto a later form. FAM is one examination in a broader actuarial pathway, so plan subsequent requirements from current SOA rules rather than treating this result as a credential by itself.

A day-by-day test-day plan

In the week before the appointment, verify the center address and the exact start time on the Prometric confirmation. Check the permitted identification and the current calculator policy. Reconfirm the syllabus version and avoid trying to learn an entire new technical area the night before. Use short mixed retrieval sessions and correct known errors.

On the day, allow travel time beyond the route estimate and follow the center’s entry instructions. Once seated, read directions about navigation and exhibits. Spend the first moments identifying the interface controls, then begin. Keep a pace that lets you return to flagged questions and leave enough time to answer every presented item, since blanks are incorrect.

Common questions

How long is the SOA FAM exam?

The October 2026 syllabus specifies 3.5 hours, or 210 minutes.

How many questions are on FAM?

There are 34 presented multiple-choice questions, including a few unscored pilots.

Is FAM the same as FAM-L and FAM-S?

No. The transition ended in July 2024; candidates who lack full credit now take FAM.