PracticeBlogFlashcardsEspañol
All Practice Exams

100+ Free CAS Exam 8 Practice Questions

Pass your CAS Exam 8 Advanced Ratemaking exam on the first try — instant access, no signup required.

✓ No registration✓ No credit card✓ No hidden fees✓ Start practicing immediately
~30% Pass Rate
100+ Questions
100% Free
1 / 100
Question 1
Score: 0/0

In classification ratemaking, what is the primary advantage of multivariate methods over one-way (univariate) analyses?

A
B
C
D
to track
2026 Statistics

Key Facts: CAS Exam 8 Exam

~30%

Historical Pass Rate

CAS published pass rates

4 hours

Exam Time

CAS Exam 8 syllabus

~$1,200

Exam Fee

CAS 2026 fee schedule

2x/year

2026 Sittings

April 14-21 and October 19-27

400-600 hrs

Recommended Study

Fellowship candidate norms

25%

GLM Pricing Weight

Largest single domain

CAS Exam 8 Advanced Ratemaking is a 4-hour computer-based, primarily essay/calculation exam offered each Spring (April) and Fall (October) by the Casualty Actuarial Society. The exam fee is approximately $1,200 and the historical pass rate is around 30%, making it one of the most demanding actuarial pricing exams. Coverage spans seven major areas: classification ratemaking (20%), GLM pricing (25%), ILFs and excess layer pricing (15%), loss distributions (10%), individual risk rating (15%), catastrophe pricing (10%), and reinsurance pricing (5%). The 2026 sittings are April 14-21 and October 19-27.

Sample CAS Exam 8 Practice Questions

Try these sample questions to test your CAS Exam 8 exam readiness. Each question includes a detailed explanation. Start the interactive quiz above for the full 100+ question experience with AI tutoring.

1In classification ratemaking, what is the primary advantage of multivariate methods over one-way (univariate) analyses?
A.Multivariate methods always produce smaller standard errors than one-way analyses
B.Multivariate methods correctly account for correlations among rating variables and avoid double-counting
C.Multivariate methods do not require credibility weighting
D.Multivariate methods eliminate the need for off-balance corrections
Explanation: One-way analyses ignore correlation between rating variables, which leads to distorted relativities when variables are not independent. Multivariate methods (such as GLMs) estimate the effect of each variable while holding others constant, removing this distortion.
2The pure premium for a class is calculated as $185, while the overall expected pure premium is $210. The class is assigned a credibility of 0.40. What is the credibility-weighted pure premium?
A.$195
B.$200
C.$205
D.$210
Explanation: Credibility-weighted pure premium = Z * class + (1 - Z) * overall = 0.40 * 185 + 0.60 * 210 = 74 + 126 = $200.
3In the loss ratio method of class ratemaking, what does the indicated relativity represent?
A.The ratio of class loss ratio to the overall loss ratio
B.The class pure premium divided by the base class pure premium
C.The ratio of class earned premium to base class earned premium
D.The class frequency divided by the base class frequency
Explanation: In the loss ratio method, the indicated change to a class relativity is proportional to the ratio of the class loss ratio to the overall (target) loss ratio. Classes with higher-than-expected loss ratios get higher relativities.
4Bailey's minimum bias procedure with the multiplicative balance principle is mathematically equivalent to which GLM?
A.Normal distribution with identity link
B.Poisson distribution with log link
C.Gamma distribution with inverse link
D.Binomial distribution with logit link
Explanation: Bailey & Simon's multiplicative minimum bias procedure produces the same parameter estimates as a GLM with a Poisson error distribution and a log link. This equivalence is the historical bridge from minimum bias methods to modern GLM ratemaking.
5Which residual is most appropriate for diagnosing GLM model fit when the response distribution is highly skewed, such as Gamma or inverse Gaussian severity?
A.Raw (response) residuals
B.Pearson residuals
C.Deviance residuals
D.Working residuals
Explanation: Deviance residuals are based on each observation's contribution to the model deviance and are approximately normally distributed even for skewed response distributions, making them preferred for diagnostic plots in GLMs with non-normal errors.
6A territory has expected pure premium of $300 and observed pure premium of $360 with credibility 0.50. The complement of credibility is the overall mean of $300. The credibility-weighted indicated relativity, with current relativity 1.10, is closest to:
A.1.05
B.1.10
C.1.21
D.1.32
Explanation: Credibility-weighted pure premium = 0.5 * 360 + 0.5 * 300 = 330. Indicated relativity to base = 330 / overall 300 = 1.10. Multiply by current 1.10 only if compounding; the question asks indicated relativity which is 330/300 = 1.10 times current 1.10 = 1.21 if expressed off-base. The new indicated rating factor is 1.10 (to overall), but applied through current it is 1.21 vs base.
7Which of the following best describes the role of an offset in a GLM for frequency modeling?
A.A parameter estimated from the data to scale predicted frequencies
B.A known per-record adjustment included in the linear predictor with coefficient fixed at 1
C.A residual term added after fitting to correct for bias
D.A weight that increases the influence of large exposures
Explanation: An offset is a known quantity (most often log of exposure for a Poisson frequency model) included in the linear predictor with its coefficient fixed at 1. This forces predicted counts to scale linearly with exposure without estimating an exposure parameter.
8For modeling pure premium directly with a single GLM, which distribution is most commonly recommended?
A.Gaussian (Normal)
B.Poisson
C.Gamma
D.Tweedie
Explanation: The Tweedie distribution with power parameter 1 < p < 2 is a compound Poisson-Gamma distribution. It places probability mass at zero (no claims) and a continuous distribution on positive losses, making it ideal for direct pure premium modeling.
9Which combination is most typical for a two-stage frequency-times-severity GLM in personal auto pricing?
A.Poisson/log for frequency and Gamma/log for severity
B.Gaussian/identity for frequency and Poisson/log for severity
C.Binomial/logit for frequency and Tweedie/log for severity
D.Gamma/inverse for frequency and Normal/identity for severity
Explanation: The standard two-stage approach uses Poisson with a log link for claim frequency (counts per exposure) and Gamma with a log link for severity (claim size). Multiplying expected frequency by expected severity yields expected pure premium.
10In a multiplicative GLM with log link, a coefficient of 0.18 on a binary indicator translates to approximately what relativity for that level versus the base?
A.0.85
B.1.18
C.1.20
D.1.82
Explanation: With a log link, the relativity equals exp(coefficient). exp(0.18) is approximately 1.197, which rounds to 1.20. This means the level has roughly 20% higher predicted pure premium than the base level holding all else equal.

About the CAS Exam 8 Exam

CAS Exam 8 Advanced Ratemaking is the upper-division CAS pricing exam on the FCAS pathway. It tests multivariate classification ratemaking, GLM-based pricing (including Tweedie pure premium and frequency-severity setups), increased limits factors and excess layer pricing, loss distribution fitting under truncation and censoring, individual risk rating (schedule, experience, retro), catastrophe pricing using EP curves and AAL, and reinsurance pricing for XOL, quota share, and profit commission structures.

Questions

100 scored questions

Time Limit

4 hours

Passing Score

Scaled

Exam Fee

~$1,200 (Casualty Actuarial Society (CAS))

CAS Exam 8 Exam Content Outline

20%

Classification & Rating Plans

Multivariate ratemaking, target loss ratio and pure premium methods, credibility blending, minimum bias procedures, residual diagnostics (Pearson, deviance, working)

25%

GLM-Based Pricing

Tweedie pure premium, Poisson frequency with log(exposure) offset, Gamma severity, two-stage frequency-severity, base rate plus relativities, interactions, smoothing splines, LRT and Wald tests

15%

Increased Limits Factors & Excess Layer Pricing

ILF construction and consistency checks, limited average severity, layer pricing using LAS differences, risk loads, ALAE treatment, fitted vs empirical tail extrapolation

10%

Loss Distributions

Pareto, lognormal, mixed and spliced severity models, truncation (left) for layered loss data, censoring (right) for policy limits, MLE under truncation/censoring, goodness-of-fit

15%

Individual Risk Rating

Schedule rating (debits, credits, IRPM caps), experience rating mod with credibility weighting (NCCI primary/excess), composite rating, retro plans (basic premium, LCF, tax multiplier, Table M and Table L)

10%

Catastrophe Pricing

Cat load as AAL divided by non-cat premium, EP curves and OEP/AEP, layered cat pricing, secondary uncertainty, demand surge, model validation against exposure-adjusted history

5%

Reinsurance Pricing

XOL pricing via burning cost and exposure rating, quota share with sliding-scale and profit commission, ROL and payback period, reinstatement premium, surplus relief, aggregate XOL

How to Pass the CAS Exam 8 Exam

What You Need to Know

  • Passing score: Scaled
  • Exam length: 100 questions
  • Time limit: 4 hours
  • Exam fee: ~$1,200

Keys to Passing

  • Complete 500+ practice questions
  • Score 80%+ consistently before scheduling
  • Focus on highest-weighted sections
  • Use our AI tutor for tough concepts

CAS Exam 8 Study Tips from Top Performers

1Master the Bailey-Simon to GLM bridge: minimum bias multiplicative procedures equal Poisson with log link, the foundation of multivariate ratemaking
2Drill ILF, LAS, and layer pricing problems until you can compute layer expected losses from limited expected values without hesitation
3Practice complete retro premium calculations with Table M and Table L lookups, including min and max premium adjustments and per-occurrence loss limit interactions
4For loss distributions, write the MLE likelihood expressions for left-truncated, right-censored, and combined truncation/censoring scenarios from memory
5Work cat layer pricing using EP curves end to end: derive AAL from event tables, layer the AAL between attachment and exhaustion, and load to a final premium

Frequently Asked Questions

What is the format of CAS Exam 8?

CAS Exam 8 Advanced Ratemaking is a 4-hour computer-based exam delivered through Pearson VUE. The exam is primarily essay and calculation rather than pure multiple choice, requiring candidates to show work for partial credit on multi-step ratemaking, retro, ILF, and reinsurance problems.

When is CAS Exam 8 offered in 2026?

Per the CAS 2026 schedule, Exam 8 is offered twice: April 14-21 (Spring sitting) and October 19-27 (Fall sitting). Candidates register through the CAS portal and schedule a Pearson VUE appointment within the relevant exam window.

What is the pass rate for CAS Exam 8?

Historical pass rates for CAS Exam 8 hover near 30%, making it one of the more demanding fellowship exams. Candidates typically have already passed Exams 5, 6, and 7 and treat Exam 8 as the capstone pricing assessment on the FCAS pathway.

How much does CAS Exam 8 cost?

The CAS exam fee for fellowship-level exams including Exam 8 is approximately $1,200 per sitting in 2026. Cost varies slightly by registration deadline; check the official CAS fee schedule before registering.

What topics carry the most weight on CAS Exam 8?

GLM-based pricing carries the largest weight at roughly 25%, followed by classification and rating plans at 20%, ILFs and excess layer pricing at 15%, individual risk rating at 15%, loss distributions at 10%, catastrophe pricing at 10%, and reinsurance pricing at 5%.

How long should I study for CAS Exam 8?

Most successful Exam 8 candidates spend 400 to 600 hours over four to six months. Time should focus on deriving GLM relativities from raw data, building ILFs and layer prices, working full retro plans with Tables M and L, and pricing XOL and quota share treaties under varying loss assumptions.