end-to-end actuarial claims reserving workflow using the Chain Ladder and Mack Chain Ladder methods in Python.
-
Updated
Jul 29, 2026 - Jupyter Notebook
end-to-end actuarial claims reserving workflow using the Chain Ladder and Mack Chain Ladder methods in Python.
A Python implementation of the Chain Ladder method for projecting IBNR claims
Production actuarial loss reserving workbench — Chain Ladder, Mack (1993), ODP Bootstrap, BF & Cape Cod methods, role-based approval workflow, full audit trail, Streamlit UI, and live demo on Streamlit Cloud.
P&C loss reserving on CAS Schedule P data: chain ladder, BF, Cape Cod, Mack, Merz-Wüthrich one-year view and ODP bootstrap, written from scratch and validated
CAS Workers Compensation Reserving Case Study. Loss Development Factors, Catastrophe/Outlier Analysis, Ultimate Loss Estimates, Frequency-Severity.
Chain-ladder and Bornhuetter-Ferguson reserve estimates with historical backtests, uncertainty analysis, and independent verification.
Historical motor pricing and reserve-uncertainty studies with exposure-aware calibration, interval evaluation and explicit source-cohort limits.
Motor insurance pricing and reserving: cross-validated GLM vs tuned XGBoost with balance correction, French vs Spanish portfolio comparison, chain-ladder and bootstrap reserves.
Can machine learning beat the chain-ladder at insurance loss reserving? A leakage-safe test on 95 US auto insurers (NAIC Schedule P) with Python, SQL and a Streamlit app.
Agent AI actuariel sur Microsoft Foundry : Chain-Ladder, erreur de Mack, Bornhuetter-Ferguson et Solvabilité II (best estimate, SCR, marge de risque, ratio de solvabilité), politique en RAG, garde-fou, évaluation 12/12.
Backtests Chain Ladder, Mack, and ODP Bootstrap reserving methods using public U.S. Schedule P insurer data.
To associate your repository with the loss-reserving topic, visit your repo's landing page and select "manage topics."