ActuarialExam PA
SOA Exam PA · Predictive Analytics
Updated August 2026

SOA Exam PA Study Plan & Practice System

Exam PA is the project-based predictive analytics exam using R. Candidates receive a dataset and business problem, build a model, and write a report. Upload your ACTEX digital edition, TIA handouts, or your own R practice notes and exclam.ai turns them into a fully guided study plan covering the R techniques, modeling decisions, and reporting framework.

Typical hours
~150
Typical weeks
~14
Topic areas
4
Pass rate
~50–60%

Exam format

Project-based computer-based test. You receive a dataset and a business problem, build a predictive model in R, and submit a written report. Significantly different format from other SOA exams.

Project-based rather than multiple-choice. You build a model and write a report.

R proficiency is essential. Practice with the SOA-provided starter code and R packages.

Reports are graded on both modeling decisions and communication clarity.

Typically taken after SRM.

Format details change periodically. Always verify the current exam structure on soa.org before your exam.

Exam PA topic areas

The 4 topic areas tested on Exam PA, with approximate weight bands from the public SOA syllabus. Click any topic for a dedicated study page.

Topic names and approximate weight bands sourced from the public SOA Exam PA syllabus on soa.org.

How exclam.ai helps with Exam PA

Upload your study manual

Drop in your ACTEX Exam PA digital edition (full-corpus PDF), scanned ASM pages, TIA handouts, or your own notes. exclam.ai parses them into structured content for a fully guided study plan.

Generate flashcards & quizzes

AI generates flashcards for formulas and concepts, plus multiple-choice quizzes targeting each of the 4 topic areas. Edit any card.

Get a weekly adaptive plan

Coverage phase, review phase, and mocks — distributed across your ~14-week timeline or whatever you choose.

Exam PA study plan

Pick the timeline that matches your exam date. exclam.ai spreads roughly 150 hours across coverage, review, and mock phases — each reading paired with its own quiz and FSRS flashcards. Switch the plan length to rebalance the schedule instantly.

Plan length
14 weeks · ~11h/week

Approximate milestones

14 weeks
CoverageMonths 1–2
  1. Problem Framing and Data Preparation~wk 2
  2. Generalized Linear Models~wk 4
  3. Decision Trees and Ensemble Methods~wk 6
  4. Model Validation and Business Communication~wk 8
ReviewMonths 2–3
Mock examsMonth 3
Examend of Month 3

Your full plan · 14 weeks to exam day

14 weeks · ~11h/week
CoverageReviewMocks· chips show the task mix for that week; buffer days are the planned rest
August 2026
  1. Wk 1Aug 24 – Aug 30Problem Framing and Data PreparationCoverageReadQuizFlashcards15–25%5 study · 2 buffer
  2. Wk 2Aug 31 – Sep 6Problem Framing and Data PreparationCoverageReadQuizFlashcards15–25%5 study · 2 buffer
September 2026
  1. Wk 3Sep 7 – Sep 13Generalized Linear ModelsCoverageReadQuizFlashcards30–40%5 study · 2 buffer
  2. Wk 4Sep 14 – Sep 20Generalized Linear ModelsCoverageReadQuizFlashcards30–40%5 study · 2 buffer
  3. Wk 5Sep 21 – Sep 27Decision Trees and Ensemble MethodsCoverageReadQuizFlashcards20–30%5 study · 2 buffer
  4. Wk 6Sep 28 – Oct 4Decision Trees and Ensemble MethodsCoverageReadQuizFlashcards20–30%5 study · 2 buffer
October 2026
  1. Wk 7Oct 5 – Oct 11Model Validation and Business CommunicationCoverageReadQuizFlashcards15–25%5 study · 2 buffer
  2. Wk 8Oct 12 – Oct 18Model Validation and Business CommunicationCoverageReadQuizFlashcards15–25%5 study · 2 buffer
  3. Wk 9Oct 19 – Oct 25ReviewReviewQuizFlashcards5 study · 2 buffer
  4. Wk 10Oct 26 – Nov 1ReviewReviewQuizFlashcards5 study · 2 buffer
November 2026
  1. Wk 11Nov 2 – Nov 8ReviewReviewQuizFlashcards5 study · 2 buffer
  2. Wk 12Nov 9 – Nov 15MocksMocksQuizFlashcards5 study · 2 buffer
  3. Wk 13Nov 16 – Nov 22MocksMocksQuizFlashcards5 study · 2 buffer
  4. Wk 14Nov 23 – Nov 29MocksMocksQuizFlashcards5 study · 2 buffer
Exam daySun, Nov 29, 2026End of week 14

Your Exam PA plan

A visual preview of how exclam.ai turns your plan into daily reading, flashcard, quiz, review, and mock streaks across the timeline you pick.

EP / day
4
daily target
Projected streak
4days
if you never miss
Study days
69
Modules
4
Per week
~11h
~150h total
To exam
14weeks

Sign in to unlock your full 14 weeks plan

Every module dated to your exam, with quizzes and FSRS flashcards generated from your own materials.

Unlock my full plan free

Week-by-week module breakdown

Your Exam PA plan walks the syllabus module-by-module before transitioning into review and mock phases. Module weeks are allocated proportional to SOA topic weights — heavier-weighted modules get more time, lighter ones either get a single week or pair with adjacent topics on intensive paces. Change the plan length above to see the schedule rebalance.

Module 1 of 4
Weeks 1–2
Exam weight 15–25%
Problem Framing and Data Preparation

Translating a business problem into a predictive modeling question, exploratory data analysis, and feature engineering. Build flashcards covering every learning objective and quiz yourself before moving to the next module.

Module 2 of 4
Weeks 3–4
Exam weight 30–40%
Generalized Linear Models

Building, evaluating, and interpreting GLMs in R for common actuarial applications. Build flashcards covering every learning objective and quiz yourself before moving to the next module.

Module 3 of 4
Weeks 5–6
Exam weight 20–30%
Decision Trees and Ensemble Methods

Fitting and tuning decision trees, random forests, and gradient boosting models in R. Build flashcards covering every learning objective and quiz yourself before moving to the next module.

Module 4 of 4
Weeks 7–8
Exam weight 15–25%
Model Validation and Business Communication

Cross-validation, test set performance, sensitivity analysis, and communicating results to non-technical stakeholders. Build flashcards covering every learning objective and quiz yourself before moving to the next module.

Phase: Review
Weeks 9–11

Review

Weak-topic drilling. exclam.ai surfaces topics where you underperformed during coverage and re-quizzes them. Daily FSRS flashcard reps across all 4 topics to prevent decay. Start doing timed question batches focused on the heaviest-weight sections.

Phase: Mocks
Weeks 12–14

Mocks

Full-length Exam PA practice exams under timed conditions. Target one mock every 3–5 days. Same-day error review: for every question you miss, re-derive the solution from scratch. Taper in the final 3 days — light flashcards only.

Popular Exam PA study manuals

exclam.ai works alongside any of these vendors. Upload paths vary, ACTEX sells full digital editions as downloadable PDFs (cleanest upload), ASM print manuals can be scanned for the full corpus, and The Infinite Actuary provides downloadable PDF handouts alongside their videos. For ASM digital (app-locked) and Coaching Actuaries (online platform), upload your own notes taken while studying. We do not distribute or reproduce any vendor content, your manual stays with the publisher.

ACTEX Study Manual
Digital edition · downloadable PDF
Coaching Actuaries
Online platform · upload your notes
SOA past project examples
Your own notes · any source

Exam PA questions

How much R do I need to know for Exam PA?

Fluent. You need to be comfortable with data manipulation (dplyr or base R), GLM fitting (glm), decision trees (rpart, randomForest, gbm), and plotting (ggplot2). Practice with the SOA sample projects before the exam.

How is Exam PA graded?

Graders score both your model choices and your written report. Clear communication of assumptions, validation, and recommendations matters as much as model accuracy. A strong model with a weak report can fail.

How long is the exam?

Historically PA has been a multi-hour sitting where you produce the project from scratch. Check current SOA guidelines on soa.org before your sitting — format has changed over time.

Start your Exam PA plan today

Upload your ACTEX digital edition, scanned ASM pages, TIA handouts, or your own notes, exclam.ai builds a fully guided study plan around your exam date.

See pricing