Francis Analytics
Actuarial Data Mining Services

{LeftPanel}

 

Contact us:
215-923-1567

louise.francis@data-mines.com

 

Panels

Casualty Actuary Society Annual Meeting
Phoenix, AZ
May 15-18, 2005

COTOR's Loss Reserving from the Viewpoint of Modeling

This session will introduce a new class for practicing actuaries that is being sponsored by the Committee on the Theory of Risk. The committee plans to work with Regional Affiliates to present the class. The session will present an abbreviated version of the subject matter covered by the class.

Based on Gary G. Venter's paper "Testing The Assumptions Of Age-To-Age Factors", the class will explore the relationship of the chain-ladder method to mathematical models of the loss development process, including comparing and contrasting the assumptions in each. Following Venter's suggestions, the class will cover ways to test the reasonableness of a mathematical model of the loss development experience, and, implicitly, the reasonableness of the chain-ladder approach. Several lines of business will illustrate both reasonable and unreasonable models and methods. Participants will be able to keep the Excel workbooks they will use in the class, which include useful charts.

Moderator:
Louise A. Francis, Consulting Principal, Francis Analytics & Actuarial Data Mining, Inc.
Panelist:
Oakley E. Van Slyke, President, Capital Management Technology

 

Casualty Actuary Society
Special Interest Seminar on Predictive Modeling
Chicago,IL
October 4-5, 2004

Generalized Linear Models I

This basic introduction to generalized linear models will begin with a basic review of common statistical methods that actuaries and other analysts have used for many years. No familiarity with modeling methods will be assumed. Topics will include introduction to regression and ANOVA. Methods of performing goodness-of-fit and diagnostic tests to examine the underlying assumptions of the regression model will be presented. The session will then discuss the often-encountered problem of what to do when the model assumptions are violated, which will lead to an introduction to GLMs. Where possible, the session will illustrate the statistical principles with applications that can be performed in Excel.

Moderator:
Richard B. Moncher, Vice President & Actuary, Bristol West Insurance Group
Panelist:
Louise A. Francis, Consulting Principal, Francis Analytics and Actuarial Data Mining, Inc



Casualty Actuary Society
2004 Ratemaking Seminar
Philadelphia, PA
March 11-12 2004


Data Mining in the Property/Casualty Insurance Industry

In recent years, data mining techniques have been increasingly applied within the property/casualty insurance industry to such areas as pricing, underwriting, marketing, and claim management. This session will provide an overview of data mining in property/casualty insurance, with specific attention given to actuarial considerations relevant to data mining in insurance. Topics to be discussed include the business and statistical backgrounds of data mining, data sources, the steps involved in a typical data mining project, model validation, and potential property/casualty applications. Several popular data mining techniques, including Neural Networks, CART, MARS, and Association Rules, will be covered in some detail.

Moderator:
Cheng-Sheng P. Wu, Director, Deloitte & Touche LLP
Panelists:
Louise A. Francis, Consulting Principal, Francis Analytics & Actuarial Data Mining Inc.
Jim Guszcza, Manager, Deloitte & Touche LLP

 

Casualty Actuarial Society
Annual Meeting.
New Orleans, LA.
November 9-12, 2003

Models- Do You Trust Them?

Technological advancements and innovations have significantly contributed to the proliferation of computer models within the insurance industry. From catastrophe to DFA models, the potential applications are countless but the uncertainties behind the methods often worry the potential users of the model's output. This "black box" effect has been and will remain an intrinsic part of most modeling tools. Additional concerns come from the often large discrepancies in results between models utilizing the same underlying information.

This session aims to demystify models, review the criteria for a good model, and provides ways to interpret and handle large discrepancies between two different models. The discussion will include general and model-specific views.

Moderator/Panelist:
Stuart B. Mathewson, Lead Pricing Actuary, Employers Reinsurance Corporation
Panelists:
Louise A. Francis, Consulting Principal, Francis Analytics & Actuarial Data Mining Inc.
Jonathan B. Hayes, Senior Vice President, Guy Carpenter & Company, Inc.

 

Casualty Actuary Society
2002 Annual Meeting
Boston, MA
November 10-13, 2003

Data Mining

Data mining is a new and evolving discipline that uses advanced technologies to find patterns in data. This session will discuss what data mining is and some of the popular methods used. The panel will present some potential actuarial data mining applications like credit scoring, underwriting, loss development, and modeling financial markets. Developed procedures such as bagging and boosting for combining the results of multiple methods will be discussed. Time permitting, the panelists will also discuss the inner workings of neural networks, one of the more commonly used data mining methods.

Moderator:
Cheng-Sheng P. Wu, Director, Deloitte & Touche LLP
Panelist:
Louise A. Francis, Consulting Principal, Francis Analytics & Actuarial Data Mining
Stijn Viaene, KBC Insurance Research Chairperson, Katholieke Universeiteit, Belgium

 

Casualty Actuary Society
Loss Reserve Seminar
Arlington, VA
September 23-24, 2002


Fair Value Accounting and Actuaries in the Post-Enron World

Proposals to reform accounting rules have received a lot of attention from accounting standards organizations in recent years. Following Enron, these proposals received increased visibility. Many of the proposals would dramatically alter how insurance accounting is done. This session will discuss fair value accounting and its implications for the actuarial profession.

The session will begin with a definition of the term "fair value," and will present a summary of initiatives undertaken by various standard-setting bodies. Topics discussed will include alternatives to fair value, implementation issues, presentation issues (for example, what would an income statement look like under fair value accounting), and a critique as to the value of fair value financial statements. The discussion will include a dialogue as to the actuarial profession's readiness to implement such a system, and what may be needed to prepare the profession for a fair value world. The session will then present an introduction to procedures actuaries may consider for estimating the fair value of liabilities.

Moderator:
Scott P. Weinstein, Actuarial Senior Manager, KPMG LLP
Panelists:
Louise A. Francis, Consulting Principal, Francis Analytics and Actuarial Data Mining
Michael J. Grillaert, Partner, KPMG LLP

 

Casualty Actuary Society
Casualty Loss Reserve Seminar
Minneapolis, MN
18-19 September 2000

Fair Value of Insurance Liabilities - CAS White Paper

This session will discuss the recently issued CAS white paper on this topic (expected to be issued this summer). It will begin with a definition of the term "fair value", a summary of initiatives undertaken by various standard-setting bodies, and a brief introduction to the white paper. Each section of the paper will then be presented in summary form. Sections include fair value estimation methods, alternatives to fair value, implementation issues, presentation issues (e.g. what would an income statement look like under fair value accounting), and a critique as to the value of "fair value" financial statements. The discussion will end with a dialog as to the profession's readiness to implement such a system, and what may be needed to prepare the profession for a "fair value" world.

Moderator:
Ralph S. Blanchard, Second Vice President and Actuary, Travelers Property Casualty Corporation

Panelist:
Louise A. Francis, Consulting Principal, Francis Analytics and Actuarial Data Mining