Life & Health Insurance Data for Investors & Quant Researchers

Three patterns dominate. First, longevity factors: Human Mortality Database supplies the point-in-time, country-harmonized death series, while Society of Actuaries mortality tables define the valuation basis annuity and life writers price against - the gap between observed mortality and the priced basis is the trade. Third, balance-sheet screening: NAIC statutory filings and experience reports yield loss ratios and financial-strength reads at an annual cadence.

Two caveats belong in the factor spec, and they bite harder because 7 of the 9 records are government- or society-published rather than vendor products. Actuarial experience studies are restated aggregates, so they support basis comparisons, not honest event studies.

The Human Mortality Database is the anchor: its own record describes point-in-time, country-harmonized series reaching back centuries, which means a vintage pulled today stays reproducible. Society of Actuaries mortality and annuity tables serve a related role - they document the valuation basis policies were priced on, so you can compare today's observed mortality against the priced assumption rather than a restated curve.

Only 2 of the 9 records here ship as static archives; the other seven revise on their own schedules.

Not directly. This slice holds no satellite imagery, card-panel or web-traffic feed - none of the categories investors associate with consumer-sensing alternative data qualify here.

That makes this an institutional-publishing slice end to end: statistical agencies, multilateral bodies, professional societies and one policy foundation (KFF) supply all nine records, consistent with a catalog where 781 of 1,744 datasets (44.8%) are public datasets.

History length is this slice's strongest suit. Two of nine sources ship as static archives - the Human Mortality Database reaches back centuries, and the Society of Actuaries tables persist until superseded. Four more refresh annually (MEPS, NAIC, the Census coverage program, the World Bank indicators), two monthly (Eurostat Health, KFF), and one - UK ONS Health and Social Care Statistics - weekly, the only sub-monthly cadence in the set. Against a catalog where 22.6% of datasets, zero daily sources here tells you the alpha lives in level and structure, not surprise.

Survivorship discipline still applies: annual and monthly series get revised, static archives do not. Pull vintages if your backtest requires period-correct inputs.