The Metra methodology is a two-layer public corpus, published here in full and intended to be read, audited, critiqued, and attested by credentialed actuaries. Each layer is version-pinned and separately attestable.
Current public corpus
Reviewers evaluating the full instrument should start with the Complete Actuarial Brief; the versioned documents below are the pinned, citable layers. Attestations are version-pinned per layer.
Methodology document — current version
The full mathematical specification: cohort construction and exposure units, Bühlmann credibility (k = 400 person-quarters with variance-component derivation), per-marker exposure with percentile-bootstrap confidence intervals (B = 10,000), short-horizon trajectory prediction intervals, twelve-month forecast with lag haircut (λ = 0.50) and attrition adjustment (α = 0.15), per-marker coefficient table anchored at literature lower bounds, limitations openly named, attestation protocol, bibliography.
Methodology addendum — Experience Calibration · current version
The additive Experience Calibration layer: take the group's own paid medical and Rx experience, mature it to an incurred basis (completion factor or IBNR), pool large claimants, and credibility-blend the observed per-employee rate against the carrier manual rate (Bühlmann Z = N/(N+400), the constant k = 400 derived as an independent full-credibility anchor). The blended rate is projected forward under trend and attrition, then the forward metabolic signal from v1.0 is applied as a disclosed dollar offset — returning a projected forward claims figure with two labeled uncertainty bands: an expected-claims confidence band and a realized-paid predictive interval. v1.0 is unchanged and consumed as a finished input; the two credibility weightings are kept disjoint and never stacked. v2.1 is additive to v2.0, which remains published and version-pinned.
Reading guide
A short reading guide for the methodology document, written for the actuary preparing to read it ahead of a renewal or strategy meeting. Anticipated reviewer questions, the seven most-frequently-raised technical objections and their responses with section references, a pre-meeting checklist, and a post-meeting follow-up template. The companion is a reading aid; it adds no methodology of its own.
Attestation roster
The list of credentialed actuaries who have read this methodology at the published version and attested to its defensibility. Each entry carries the actuary's name, credential, firm, the methodology version attested, and the attestation date. Attestations are version-pinned; a version bump requires re-attestation. Compensation arrangements, where any exist, are publicly disclosed alongside each entry.
Sample output
A complete generated Workforce Exposure Forecast PDF for a representative 322-employee employer cohort, produced from the demonstration corporate population on the Metra platform. The Sample carries an illustrative Demo Signer block (Marcus Chen, FSA, MAAA — Chen Actuarial Consulting), which is a demonstration entry and not a live signing engagement; the published roster on the attestations page shows this entry as a clearly marked demonstration signer with live engagement pending. The Sample also shows the cover, executive summary, per-marker exposure breakdown with bootstrap CIs, 30/60/90-day trajectory bands, the 12-month forecast, and the methodology-version footer. Useful as a reference for what the buyer-side instrument looks like at the renewal table.
Companion document
The doctrine-register companion piece read in parallel with the methodology document: the multi-marker cost coefficients with their full citations, the conservatism-and-stacking rationale, and the appendix introducing the multi-marker move from the original single-marker (weight-only) framing. Bundled with this methodology document by reviewing actuaries who want the technical specification and the doctrine framing together.
For credentialed actuaries
If you hold an ASA, FSA, MAAA, FCAS, or ACAS credential and are reading this document professionally, you are the intended primary audience. Critiques, corrections, and counter-derivations are welcomed and inform the next version. Attestation submissions and review correspondence: care@usemetra.com, subject line “Methodology Review — v1.0” or “Methodology Attestation — v1.0”.