What should you know before studying for Exam P?
The SOA's study instructions assume calculus, including series, differentiation and integration. Check whether you can use those tools inside a probability problem. These original examples identify a few starting gaps; they are not a pass predictor.
Can you turn a density into a probability?
Suppose X has density 2x for 0<x<1, and zero elsewhere. Find the probability that X exceeds 0.6.
The density integrates to one over its full support. Evaluating the density at 0.6 gives 1.2, which is a height, not a probability. If the integral is the obstacle, review calculus. If the limits are the obstacle, study probability models and support.
Can you calculate an expectation without assuming equal probabilities?
A random variable takes values 0, 1 and 3 with probabilities 0.2, 0.5 and 0.3. Weight each value by its probability.
The ordinary average 4/3 would describe three equally likely values. Here, a frequency check with counts 2, 5 and 3 out of ten gives a total of 14 and the same mean of 1.4.
Can you identify which event is given?
In “given that a policy had a claim, find the probability it belongs to the high-risk group,” the denominator concerns all claims. Follow the worked Bayes example and explain why the claim probability within the high-risk group is a different quantity.
What order should you study the topics in?
Use the following dependency order. The study sequence is our recommendation; the official syllabus determines what can be tested.
- Events, counting and conditioning. Translate “at least,” “exactly,” “given,” and “independent” into events before calculating. Start with conditional probability and the general probability collection.
- One random variable. Move among a mass or density function, a cumulative distribution, and a survival function. Then calculate expectations, variances and quantiles. Compare distribution assumptions rather than memorizing names. Use the univariate problem collection.
- Insurance payments. Write the payment as a function of the loss before taking its expectation. A deductible, payment cap and coinsurance percentage affect different parts of that function. Work the payment-per-loss versus paid-claim example.
- Joint models and combinations. Derive marginals from a discrete joint table, then conditional distributions, covariance and linear combinations. Include order statistics and normal approximations. Continue with multivariate problems.
- Mixed questions. Remove the topic label. Decide which model applies without a section heading giving the method away. Return to a focused collection only after identifying a specific gap.
The November 2026 topic ranges are General Probability 23–30%, Univariate Random Variables 44–50%, and Multivariate Random Variables 23–30%. They are not exact question counts or a reason to skip a smaller topic. See the official learning outcomes.
Which Exam P study materials do you need?
Start with one main source for explanations. Add a reference or question set when it fills an identified gap. Several manuals are less useful than one source you can actively work through.
| Your current need | What to look for | Where to start |
|---|---|---|
| Learning probability from the beginning | Definitions, derivations, introductory examples and graded exercises. Inspect a sample chapter before purchasing. | The suggested readings in the SOA syllabus. A sample-answer library alone may not teach enough background. |
| Understanding a numbered SOA sample question | A setup, explanation of the chosen model, intermediate calculations and a check. | Our 718 free sample-reference solutions. Keep the official question PDF beside the independently written solution. |
| Recalling a distribution or formula | Support, parameter convention, conditions and a worked use case. | The free Exam P formula sheet and PDF, with no account required. |
| Practicing under time | Unseen mixed questions with answers withheld until the attempt is complete. | Start with the online sample exam linked from SOA's study page. For additional original sets, compare the Exam P practice exams and topic workbooks. Repeated questions can overstate readiness. |
| Keeping worked reference solutions offline | A searchable document with actual page previews and a clear statement of included material. | Preview the Probability Proof Manual — $49. It collects worked reference solutions and original variants; official question wording is not included. |
ActuaryProof sells its own study PDFs. The free resources above can be used without buying them. This is a guide to matching a format to a need, not a claim that we have tested or ranked every provider's course.
How long should you study for Exam P?
Use demonstrated knowledge and your available weekly hours to set the calendar. There is no preparation-hour total here that guarantees a pass. A recent probability course may shorten the learning stage; unfamiliar calculus or repeated setup errors may extend it.
As a planning example, eight weeks with ten available hours each week gives 80 hours. That is an availability calculation, not a sufficiency claim. Reserve time for correction and mixed practice before deciding whether your planned sitting is realistic.
| Stage | Suggested work | Evidence to move forward |
|---|---|---|
| Learn the foundations | Study a short section, close it, explain the model, then solve an unfamiliar related question. | You can choose the model and state its assumptions without copying a worked example. |
| Connect topics | Mix event, distribution, payment and joint-model questions. Revisit the topic responsible for each mistake. | You recognize the method without knowing which chapter the question came from. |
| Practice pacing | Work mixed sets without notes, then review every wrong or guessed answer. | Your results are repeatable on unfamiliar questions and calculation speed leaves time to revisit marked items. |
A four-week review in 16 sessions
If you have already learned the syllabus, this allocation keeps each topic inside the official weight range (4 of 16 sessions is 25%, 8 is 50%). It is a review schedule, not a promise that a beginner can learn the entire course in four weeks.
| Topic | Week 1 | Week 2 | Week 3 | Week 4 | Sessions |
|---|---|---|---|---|---|
| General Probability | 2 | 0 | 0 | 2 | 4 |
| Univariate Random Variables | 2 | 3 | 2 | 1 | 8 |
| Multivariate Random Variables | 0 | 1 | 2 | 1 | 4 |
Record each miss as recall, setup, calculation, or check, and repair it on a new problem rather than by rereading the worked example.
How should you use the SOA sample questions?
- Read the question in the official set and write down the requested quantity, support and assumptions.
- Attempt it before opening the matching numbered solution. Record where you first became uncertain.
- Compare the model and calculation. Classify the error as wording, model, algebra, arithmetic or time management.
- Work the original variant on the solution page. Re-solving the same numbers can test memory instead of understanding.
- Later, attempt an unfamiliar mixed question that uses the same idea without announcing it.
A correct guess still needs review. Conversely, a sound setup followed by a numerical slip calls for a different repair from choosing the wrong distribution. Write the specific correction you will make on the next attempt.
What should you check before a full timed attempt?
The November 2026 format is 30 multiple-choice questions in three hours, an average of six minutes per question. Individual questions need different amounts of time; six minutes is a planning average, not a rule to stay on every question that long.
- Can you identify each topic's basic models without a formula sheet?
- Can you finish mixed questions you have not seen recently?
- Can you explain each wrong answer you selected and solve a related fresh example?
- Can you use the permitted calculator and supplied tables without losing time to unfamiliar operations?
Use the official instructions for the permitted materials and your sitting. A score on a short diagnostic or a familiar practice set is not an estimated probability of passing.
Still choosing a first exam? Compare Exam P and FM prerequisites and subjects.