Philidor Docs
Risk Framework

Tiers & Suitability

How 0-10 scores map to Prime, Core, and Edge tiers and to institutional suitability labels.

Tier Mapping (Score-Based)

TierScore RangeInterpretation
Prime8.0 - 10.0Strong relative resilience under current evidence
Core5.0 - 7.9Moderate resilience with meaningful constraints
Edge0.0 - 4.9Elevated or unresolved risk pressure

Tiers are relative classifications, not guarantees or recommendations.

The Prime cutoff is set at 8.0 to ensure that reaching Prime requires genuine strength across all three vectors; a vault cannot land in Prime purely on the back of one dominant vector when the others are mediocre.

Suitability Labels (Policy Layer)

Alongside the numeric tier, the methodology emits a suitability label that captures governance and evidence requirements on top of score:

SuitabilityRule
institutionalreviewed + final score ≥ 8.0 + no hard-fail flags + high-confidence evidence
qualifiedreviewed + final score ≥ 5.0
speculativereviewed or provisional but below qualified, or constrained by caps
not_assessedunreviewed

Suitability is distinct from tier:

  • Tier describes score range.
  • Suitability adds governance and evidence requirements on top of the score.

A Prime-tier vault can still fall short of institutional if evidence confidence or review status is insufficient. Conversely, institutional requires both a Prime score and clean governance/evidence state.

What Forces A Lower Outcome

Score-limiting constraints are explicit and explainable. The following may pull a score below what the weighted dimensions alone would suggest:

  • cap cascade on asset scoring (review_status, hard-fail flags, overrides, staleness)
  • unresolved address fallback to strict unreviewed behavior
  • portfolio-level concentration, correlation, and wrong-way penalties
  • asset-quality drag and hard cap for non-prime assets
  • fail-safe operational mode (degraded or fail_closed)

Practical Reading Guidance

  • A Prime tier can still be non-institutional if evidence confidence or governance criteria are insufficient.
  • A Core or Edge tier often indicates one or more binding caps; check the explainability fields (caps, flags, freshness, provenance).
  • When fail-safe is fail_closed, no score is served by design.

On this page

Raw