| BT-001 |
pass |
Track C crosscheck: FinView vs Gold Copy bond pricing
All 6 instruments within tolerance. Largest deviation on HY Corp 7Y (0.04% clean price). |
finview.L1.bond_pricer |
| BT-002 |
pass |
Track C crosscheck: duration and convexity
Duration and convexity consistent across implementations. |
finview.L1.bond_pricer |
| BT-003 |
pass |
CECL scenario differentiation: base vs stress ECL
Stress ECL 2.17x base ECL. Rating migration increases PD non-linearly, producing >2x uplift for 2x PD shock. |
finview.L3.cecl_calculator |
| BT-004 |
pass |
CECL recovery rate sensitivity
ECL responds proportionally to LGD changes (elasticity 0.95, expected ~1.0). |
finview.L3.cecl_calculator |
| BT-005 |
pass |
CECL severely adverse scenario (DFAST)
4.24x uplift under severely adverse. Consistent with crisis-level default experience. |
finview.L3.cecl_calculator |
| BT-006 |
pass |
Stress engine: parallel rate shock consistency
MV change within 3bps of analytical estimate including convexity correction. |
finview.L2.stress_engine |
| BT-007 |
pass |
Stress engine: DFAST scenario ordering
Scenario ordering confirmed: baseline (-0.8%) < adverse (-3.2%) < severely adverse (-7.1%). |
finview.L2.stress_engine |
| BT-008 |
pass |
Stress engine: credit spread shock
Spread shock MV impact within 7bps of analytical estimate. |
finview.L2.stress_engine |
| BT-009 |
review |
Stress engine: combined scenario additivity
8.3% deviation from additivity. Within tolerance but flagged for review — cross-gamma effects from rate-spread correlation may warrant monitoring. |
finview.L2.stress_engine |
| BT-010 |
review |
Firm balance-sheet tracking vs reported filings (13-firm fleet)
Asset-side balance sheet reconciles to filings (total assets, separate account 13/13). invested_assets + investment_income miss for AFL/APO/JXN — the same consolidated-NCI /... |
insmodel.L4.asset_engine |
| BT-011 |
superseded |
Aggregate policy-reserves tracking vs reported filings (13-firm fleet)
SUPERSEDED BY BT-012 (2026-06-04). This was an HONEST RESULT, NOT A PRODUCTION-ENGINE VERDICT: the e2e crosscheck assembled product reserves from MockDataProvider (notional —... |
insmodel.L4.reserve_engine_stat |
| BT-012 |
pass |
Firm-calibrated reserve reconciliation by reportable segment (L&A vertical)
RESOLVES BT-011 ROOT CAUSE. With firm-calibrated books the reserve aggregate reconciles to filings (6/8 segment-grade firms within 10%: APO/CRBG/GL/JXN/LNC/UNM; PFG ~88% and... |
insmodel.L4.reserve_engine_stat |
| BT-013 |
review |
Reserve roll-forward vs persistence baseline (predictive, QUARTERLY cadence)
QUARTERLY CADENCE (corrected from an initial annual-only read). Persistence is an even tighter baseline at quarterly frequency (~3.4% pooled q/q MAE vs ~11% annual) — the... |
insmodel.L4.reserve_engine_stat |