Welcome to the project Regression Modeling with Actuarial and Financial Applications! The primary product of this project is a book by the same title, designed for basic actuarial education. Specifically, in North America, the content covers the material for the "Applied Statistical Methods" portion of VEE (validated by educational experience) courses, required by the Society of Actuaries and Casualty Actuarial Society.

 

Of course, there are many good introductions to regression and time series available in the literature. The approach of this project is unique in that it is designed for actuaries or, more generally, quantitative analysts interested in financial risk management problems. The two main features that distinguish this approach from others are

  • Data  - we have gathered 41 datasets that will allow readers to explore regression and time series modeling on their own

  • Examples - the index summarizes many other examples of regression and time series applications in insurance, risk management and finance

Contents. The project focuses on topics of interest to actuaries. The content consists of basic VEE topics but also includes advanced topics such as generalized linear models, two-part models, claims triangles and more, as well as general communications topics on writing and graphical design.

 

The publisher for the book is Cambridge University Press.  The first edition is available in hardback, paperback and e-format. See the Table of Contents for the design. An excerpt will give you a feel for the writing style.


 

Supplements

 

Data. To encourage readers to learn interactively, data used in the book and exercises are available at this web site. These data sets are taken primarily from insurance and actuarial applications. By studying them, the reader not only learns statistics but also becomes familiar with potential insurance applications of regression and time series methods!

 

Statistical Software Scripts. To make this content truly "applied," one learns by doing - for us, this means analyzing data using statistical software. To encourage a wide readership, the book will not be tied to a specific statistical software package. However, examples of three statistical software packages will be available on this web site in the form of a series of suggests steps, or "scripts."

  • Examples are given in terms of the statistical package SAS, used widely in the insurance industry. However, SAS has a steep learning curve and is expensive to purchase (for students and universities) - it is not the best choice for many readers.

  • The text analysis was done using the open source (free) package "R" - by featuring this package, the project will be accessible to a wide user base.

  • Examples of Excel will be given, simply because this package is widely used in universities and industry and because many actuaries use Excel almost every day. Examples of Excel are limited because this package is not really geared to provide modern statistical analyses.

Project Resources. Students learn through analytic exercises where they reinforce theory through numerical exercises and theoretical exercises and extensions. They also learn through software homeworks, performing statistical analyses on real data sets. Another method of learning is through more free-ranging "projects" where the student or a team of students is given greater latitude in using applied statistical methods on a problem of interest. This page provides some tips on how students might select a project, a structure for writing a project report and some sample student projects.

 

Instructors' Resources. A few additional resources, in the form of an instructor's manual is available at the Cambridge University Press website. In addition to a description of the the data sets that is available to all, the manual includes a sample syllabus, detailed solutions to the exercises as well as a series of presentations on the content.  These are based on lectures given at the University of Wisconsin-Madison.

 

How You Can Help. Please write to Jed Frees (jfrees@bus.wisc.edu) for information, comments or if you would like to post some data or sample student projects.

 

 

 

Date: 24 July 2010