Exam MAS-II: Modern Actuarial Statistics II Study Guide
Advanced credibility, loss distributions, ratemaking, reserving, risk measures, and predictive modeling.
Overview
Exam MAS-II (Modern Actuarial Statistics II) covers advanced actuarial topics including credibility, loss distributions, ratemaking, reserving, risk measures, and predictive modeling for CAS candidates. It builds on MAS-I and applies statistical methods to real insurance problems. This exam replaced portions of the former Exam 4 and Exam S.
Exam Format
- Duration
- 240 minutes (4 hours)
- Questions
- 40 multiple-choice questions
- Pathway
- CAS
- Format
- Computer-based testing (CBT)
Topic Breakdown
Loss Distributions
15-20%Parametric families, coverage modifications, censoring and truncation, maximum likelihood estimation
Credibility
15-20%Limited fluctuation, Buhlmann, Buhlmann-Straub, exact Bayesian, empirical Bayes nonparametric
Ratemaking
15-20%Individual and class ratemaking, loss development, trending, expense loading, profit and contingencies
Reserving
15-20%Chain-ladder, Bornhuetter-Ferguson, Cape Cod, frequency-severity, paid vs. incurred triangles
Risk Measures
10-15%VaR, TVaR (CTE), coherent risk measures, risk-adjusted capital, reinsurance applications
Predictive Modeling
10-15%GLMs for insurance data, model selection, cross-validation, regularization (LASSO, Ridge), model interpretation
Recommended Study Approach
- 1
Master the parametric loss distributions (Pareto, Lognormal, Weibull, Gamma) and practice fitting them to data using MLE.
- 2
Study coverage modifications (deductibles, limits, inflation) and their effect on distribution parameters.
- 3
Practice credibility calculations extensively, especially Buhlmann and Buhlmann-Straub parameter estimation.
- 4
For ratemaking, work through complete rate indications including development, trending, on-leveling, and credibility weighting.
- 5
Study reserving methods by completing full triangle exercises with chain-ladder, BF, and Cape Cod methods.
- 6
Understand risk measures (VaR, TVaR, CTE) and their properties, especially subadditivity for TVaR.
Related Topics
Regression and Predictive Modeling
Linear and logistic regression, GLMs, model selection, and predictive analytics.
Loss Distributions and Modeling
Parametric loss models, severity and frequency distributions, and aggregate claims.
Credibility Theory
Buhlmann, Buhlmann-Straub, classical credibility, and empirical Bayes methods.
Ratemaking and Pricing
Pure premium method, loss ratio method, experience rating, and classification.
Reserving Methods
Chain ladder, Bornhuetter-Ferguson, and stochastic reserving techniques.
Reinsurance Mathematics
Proportional and non-proportional reinsurance, excess of loss, and stop-loss pricing.
Risk Measures and Management
VaR, TVaR, expected shortfall, coherent risk measures, and ERM frameworks.
Bayesian Statistics
Prior and posterior distributions, conjugate families, MCMC, and Bayesian estimation.
Cheat Sheets
Loss Models Cheat Sheet
Severity distributions, frequency models, aggregate loss formulas, and coverage modifications.
Credibility Theory Cheat Sheet
Buhlmann, Buhlmann-Straub, classical credibility, and empirical Bayes formulas.
Reserving Methods Cheat Sheet
Chain ladder, Bornhuetter-Ferguson, and other loss reserving techniques.