Finding
The development method has 14.54% weighted absolute percentage error across 468 masked one-step cells, compared with 44.50% for repeating the latest increment. Adding a 1.10 tail factor raises the modeled unpaid balance by $17.12 billion. The result shows substantial sensitivity to late claims development.
Why it matters
Investigate the largest tail sensitivities first. Request actuarial selections, claims-handling changes and reinsurance detail before discussing reserves or capital returns.
Who could use this
A finance and actuarial diligence committee deciding where further reserve investigation is needed.
Data and scope
4,046 paid/incurred cells from six original 2025 insurer filings. Thirty-one of forty-eight disclosed triangles meet the selected development gates.
How the analysis works
Gated paid-development chain ladder, volume and median methods, masked-diagonal benchmark comparisons and explicit tail sensitivities. Source scope and numerical cells are retained.
Limits of the evidence
The current disclosure contains revised historical cells, so this is not an as-released forecast. Older claims, expense scope and sparse late-development factors matter. The model does not justify reserve release or capital returns.