The Latest on Mortality

Part of presentation on mortality methods and trends. This part covers grouping the data.

Transcript

1. THE LATEST ON MORTALITY SESSION 67 OCTOBER 18, 2011 Primary Competency: Technical Skills & Analytical Problem Solving GROUPING THE DATA Kevin Pledge
2. How do you decide how to group your data? OR Do you decide how to group your data? 2
3. Old way New way STUDY PROCESS 3 Collect and Group Data Process Data Publish Results Analyze Published Results Collect Data Process Analyze and Group Analyze and Publish Results
4. Whatever we had last year + Some more TRADITIONAL GROUPING DEFINED AT PROCESSING 4
5. Whatever we had last year + Some more GROUPING DATA MODERN APPROACH #1 5
6. # 2 Analyze by slice and dice based on policy and product attributes #3 Neural Network Analysis? All of the Above GROUPING DATA MODERN APPROACH 6
7. Premature aggregation 7
8. Omission of attributes Assumptions in processing Policy / Coverage EXAMPLES OF PREMATURE AGGREGATION 8
9. STUDY PROCESS RECAP 9 Collect Data Process Analyze and Group Analyze and Publish Results Groups defined as attributes Automated and Regular Multiple Groups Avoid Aggregation Regroup depending on purpose
10. • Kevin is CEO and Co-Founder of Insight Decision Solutions. • Insight provides insurance analytics on a data warehouse platform, in-house and hosted. • In 2009 Kevin was responsible for the Retirement Pension Experience Committee (RPEC) mortality study which covered over 30 million exposure years developed from data provided from various government and private pension schemes. • Kevin has written several articles and is co-author of “The Analysis of Insurance Earnings” (2008). • He has served on three SOA section councils and Microsoft’s Partner Advisory Council for 5 years. • Kevin qualified as a Fellow of the Institute of Actuaries (UK) in 1992 and a Fellow of the Society of Actuaries (US) in 2004. kpledge@insightdecision.com 905 475 3282 x 224 @kevinpledge http://ca.linkedin.com/in/kevinpledge KEVIN PLEDGE 10