More Data Now Available for CAS-Sponsored Research, Including Updated Schedule P Data
The CAS Research Council is pleased to announce that the following data sets are now available to authors interested in working on CAS research projects:
Updated Schedule P Data: We have extended what was compiled previously by an additional 10 accident years (1998-2007). The update provides detailed, standardized, and extensive historical claims data on losses and premiums from major personal and commercial lines of business. Data pulled from Schedule P – Analysis of Losses and Loss Expenses in the National Association of Insurance Commissioners (NAIC) database.
S&P Data License: Provides crucial financial data, market trends, and performance benchmarks helpful for risk modeling, solvency assessment, competitive analysis, and strategic planning.
Regrid: Offers comprehensive data on U.S. residential properties sourced from tax assessor data and includes parcel boundaries and other spatial details.
VerdictSearch: Spanning January 1, 2010, through December 31, 2024, this database of jury verdicts and settlements across the U.S. includes items such as verdict/settlement dates, jurisdictions (state and county), insurers involved, and injury type.
Loss Simulation Model and Documentation: Offers an open-source R package to help actuaries perform transaction data fitting and claim-level loss reserving. Includes Excel front end as an easy-to-use interface.
ActSim (Coming Soon!): ActSim is a Python-based actuarial modeling toolkit (developed by Juntao Simon Zhang and validated by the CAS Risk Working Group) that combines distribution fitting, stochastic simulation, and synthetic claims generation into a unified workflow. Similar to CAS Loss Simulator 2.0, ActSim generates realistic claim-level data (including development over time) and provides distribution fitting to support assumption setting, making it practical for reserving analysis, scenario testing, and risk modeling.
Pricing Data Set: With nearly 4 million earned exposures, this data set offers 1 order of magnitude more exposures than freMTPL. It includes dozens of features (blinded). Confidential use required.
Example CAS Papers That Make Use of CAS Data Resources
Making Bootstrap Reserve Ranges More Realistic by David R. Clark Hang Ding, and Lianmin Zhou (Schedule P)
A Behavioral Approach to Understanding Loss Reserves by Uri Korn (Schedule P)
Enhancing Actuarial Ratemaking with Synthetic Data for Privacy Preservation by Noa Zamstein (Pricing Data Set)
How to Access
To request access to these data sets or ask questions about them, please contact CAS Research & Publications Director Elizabeth Smith. Schedule P and the Loss Simulation Model are available on the CAS Research Resources web page. ActSim will be available soon on the CAS GitHub.
