From GLMs to Comprehensive Insurance Pricing: Techniques and Challenges
By Alan Chalk, David Deacon, Montserrat Guillen, and Max Martinelli
From GLMs to Comprehensive Insurance Pricing: Techniques and Challenges expands the practical methods introduced in Generalized Linear Models for Insurance Rating (CAS Monograph No. 5) and is a complement to Penalized Regression and Lasso Credibility (CAS Monograph No. 13). This monograph focuses on the real challenges that practicing actuaries face when developing rating models.
Readers will learn how to:
- Handle high-cardinality categorical variables
- Incorporate nonlinear relationships
- Resolve issues caused by highly correlated predictors
- Blend separate analytical components—such as geographic or credit-based models—into a coherent pricing structure
Throughout, the authors emphasize sound model validation, avoiding overfitting, and the importance of actuarial judgment when using automated or machine-learning-based methods.
Three case studies add real-life context to the discussion:
- FAA–NTSB Aircraft Incident Frequency Model: Demonstrates modeling of rare events, categorical complexity, credibility, and nonlinear aging effects.
- UK STATS19 Road Accident Model: Highlights how penalized regression stabilizes models built from numerous correlated demographic variables.
- Simulated Auto-Insurance Data: Illustrates evaluation metrics, ground-truth interpretation, and the handling of restricted or correlated rating factors.
Blending statistical rigor with practical insight, this monograph offers a comprehensive, modern framework for constructing transparent, reliable, and high-performing insurance pricing models. It will help foster a deeper understanding of GLMs and their applications in today’s pricing environment.
This monograph will be applicable to Exam 8 beginning with the October 2026 administration.
About the Authors
Alan Chalk is an Actuarial Data Scientist at Sabre Insurance Limited, where he applies actuarial and machine-learning techniques to enhance pricing and fraud-detection capabilities.
David Deacon has more than 20 years of experience in the insurance industry. He is responsible for the pricing, product development, underwriting strategy, and strategic initiatives for Heritage Insurance Holdings.
Montserrat Guillen is a Full Professor of Quantitative Methods at the University of Barcelona and leads the UBriskcenter research group. She is recognized as a top global author in telematics insurance and is the most-cited woman in actuarial science worldwide.
Max Martinelli is a Lead Actuarial Data Scientist at Akur8, where he works with insurers across the North American property and casualty market to apply machine learning to pricing and rating.
ISBN 978-1-7370028-9-5
This monograph is currently available to all CAS members as a free download. The CAS Monograph Series showcases CAS members' extensive specialized expertise, helping to raise the performance standard for property and casualty actuaries through insightful research. The monographs represent just one way that the CAS provides its members with access to relevant information, research and resources that they can apply directly on the job to advance in their careers. For information on submitting a monograph, visit the Monograph Submission Guidelines page.
