Abstract
Motivation: This paper takes a multi-faceted approach to quantifying the significance of data quality issues for property/casualty actuaries, addressing both the prevalence of data quality issues across areas of practice and the significance of those issues. The conclusion gives some guidance to improve data quality.
Method:
This paper:
• describes some actual data quality disasters in non-insurance and insurance businesses;
• presents the results of a data quality survey of practicing actuaries in the United States, Canada, Great Britain and Bermuda;
• presents the results of a data quality experiment where data was altered to change its quality and the effect on analyses using the data was quantified; and
• provides advice on what can be done to improve the state of data quality, including introducing some freeware that can be used to screen data.
Results: Both the survey results and the data quality experiment suggest that data quality issues affect the accuracy and increases the uncertainty associated with actuarial estimates
Conclusions: Data quality issues significantly impact the work of property/casualty insurance actuaries; and such issues could have a material impact on the results of property/casualty insurance companies.
Availability: Excel spreadsheets containing the data used in the data quality experiment as well as the spreadsheet containing the bootstrap procedure are available here.
Keywords: Data, data quality, reserve variability, exploratory data analysis, data diagnostics
Method:
This paper:
• describes some actual data quality disasters in non-insurance and insurance businesses;
• presents the results of a data quality survey of practicing actuaries in the United States, Canada, Great Britain and Bermuda;
• presents the results of a data quality experiment where data was altered to change its quality and the effect on analyses using the data was quantified; and
• provides advice on what can be done to improve the state of data quality, including introducing some freeware that can be used to screen data.
Results: Both the survey results and the data quality experiment suggest that data quality issues affect the accuracy and increases the uncertainty associated with actuarial estimates
Conclusions: Data quality issues significantly impact the work of property/casualty insurance actuaries; and such issues could have a material impact on the results of property/casualty insurance companies.
Availability: Excel spreadsheets containing the data used in the data quality experiment as well as the spreadsheet containing the bootstrap procedure are available here.
Keywords: Data, data quality, reserve variability, exploratory data analysis, data diagnostics
Volume
Winter
Page
1-70
Year
2008
Categories
Financial and Statistical Methods
Statistical Models and Methods
Data Diagnostics
Actuarial Applications and Methodologies
Data Management and Information
Data Quality
Financial and Statistical Methods
Statistical Models and Methods
Exploratory Data Analysis
Actuarial Applications and Methodologies
Reserving
Reserve Variability
Publications
Casualty Actuarial Society E-Forum
Prizes
Management Data and Information Prize
Documents
