Q30, run 2026-11-04-2022-sonnet

Which entities had a foreign-currency issuer rating exactly at their country's ceiling on 1 October 2026?

Tests: Ceilings: which ratings the cap binds.

Expected: entities = ['E05', 'E1028', 'E1039', 'E1101', 'E1126', 'E1194', 'E1236', 'E15', 'E5012', 'E5047', 'E5091', 'E5116', 'E5124']

✅ R1 csv run 1 graded by code

13 entities have an FC rating equal to the ceiling. Valdoria's ceiling is BB (A12) and Corvenia's is A (A37), and the 13 are the VD entities at BB and the CV entities at A, listed in the JSON.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

✅ R1 csv run 2 graded by code

The entities whose FC rating equals their country ceiling (BB for VD, A for CV) are in the JSON.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

❌ R1 csv run 3 graded by code

11 entities (list in JSON). Only VD (BB) and CV (A) ceilings are in the data, and these are the FC ratings equal to them.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5124", "E15", "E1039", "E1194"]}

Why: entities: expected ['E05', 'E1028', 'E1039', 'E1101', 'E1126', 'E1194', 'E1236', 'E15', 'E5012', 'E5047', 'E5091', 'E5116', 'E5124'], got ['E05', 'E1028', 'E1101', 'E1126', 'E1236', 'E5012', 'E5047', 'E5124', 'E15', 'E1039', 'E1194']

✅ R1 csv run 4 graded by code

Thirteen entities (list in JSON).
- VD's ceiling is BB; E05, E1028, E1101, E1126, E1236, E5012, E5047, E5091 and E5124 are rated BB.
- CV's ceiling is A; E15, E1039, E1194 and E5116 are rated A.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

✅ R1 csv run 5 graded by code

13 entities have a foreign-currency rating exactly at the ceiling, listed in the JSON. The ceiling is BB for VD (A12) and A for CV (A37).

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5124", "E5091", "E15", "E1039", "E1194", "E5116"]}

✅ R4 csv+markdown run 1 graded by code

13 entities have an FC rating exactly at their ceiling (see JSON): nine in VD (ceiling BB) and four in CV (ceiling A). Domicile follows the RESEARCH record; this excludes E1245, which RESEARCH places in FR.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

✅ R4 csv+markdown run 2 graded by code

13 entities (list in the JSON). They have a foreign-currency rating equal to the ceiling: BB for VD or A for CV.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

✅ R4 csv+markdown run 3 graded by code

13 entities (listed in JSON). The VD ceiling is BB and the CV ceiling is A. The list includes E5106's neighbours only where the RESEARCH domicile applies, so E1245 (research: FR) is not counted.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

❌ R4 csv+markdown run 4 graded by code

12 entities (listed in the JSON). Ceilings are BB for VD and A for CV; these entities have an FC rating exactly at their country's ceiling, with country taken from RESEARCH where records disagree (E5106 is VD).

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5124", "E15", "E1039", "E1194", "E5116"]}

✅ R4 csv+markdown run 5 graded by code

13 entities (list in the JSON). These are entities domiciled in VD (ceiling BB) or CV (ceiling A) whose latest FC rating equals the ceiling. For domicile I used the RESEARCH record where records disagree.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

✅ R10 executed run 1 graded by code

13 entities (listed in the JSON). CV's ceiling is A and VD's is BB, and each of these has an open foreign-currency rating exactly at its country's ceiling.

JSON: {"entities": ["E15", "E1039", "E1194", "E5116", "E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124"]}

✅ R10 executed run 2 graded by code

13 entities. CV's ceiling is A (E15, E1039, E1194, E5116 are at A) and VD's is BB (E05, E1028, E1101, E1126, E1236, E5012, E5047, E5091, E5124 are at BB).

JSON: {"entities": ["E15", "E1039", "E1194", "E5116", "E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124"]}

✅ R10 executed run 3 graded by code

13 entities, listed in the JSON. Country ceilings are CV = A and VD = BB. Five entities sit at BB in VD (E05, E1028, E1101, E1126, E1236) and three more at BB in VD-domiciled group companies (E5012, E5047, E5091, E5124 make four of these; together they are the VD set). The CV entities at A are E15, E1039, E1194 and E5116.

JSON: {"entities": ["E15", "E1039", "E1194", "E5116", "E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124"]}

✅ R10 executed run 4 graded by code

13 entities with a current FC rating equal to their country's ceiling (CV: A, VD: BB), listed in the JSON.

JSON: {"entities": ["E05", "E15", "E1028", "E1039", "E1101", "E1126", "E1194", "E1236", "E5012", "E5047", "E5091", "E5116", "E5124"]}

❌ R10 executed run 5 graded by code

E15, E1039 and E1194 are at the CV ceiling of A. E05, E1028, E1101, E1126, E1236, E5012, E5047, E5091 and E5124 are at the VD ceiling of BB.

JSON: {"entities": ["E15", "E1039", "E1194", "E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124"]}

Why: entities: expected ['E05', 'E1028', 'E1039', 'E1101', 'E1126', 'E1194', 'E1236', 'E15', 'E5012', 'E5047', 'E5091', 'E5116', 'E5124'], got ['E15', 'E1039', 'E1194', 'E05', 'E1028', 'E1101', 'E1126', 'E1236', 'E5012', 'E5047', 'E5091', 'E5124']

✅ R8 csv+axioms run 1 graded by code

13 entities were exactly at their ceiling: five at CV's ceiling of A (E15, E1039, E1194, E5116) and BB entities at VD's ceiling. The full list is in the JSON.

JSON: {"entities": ["E15", "E1039", "E1194", "E5116", "E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5124", "E5091"]}

✅ R8 csv+axioms run 2 graded by code

13 entities. Nine are domiciled in VD and rated BB (the ceiling is BB), and four are domiciled in CV and rated A (the ceiling is A).

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

❌ R8 csv+axioms run 3 graded by code

12 entities (list in JSON). Each has a foreign-currency rating equal to its country's ceiling, which is BB for Valdoria and A for Corvenia. E5106 counts as VD-domiciled via its research duplicate E5169, but it is rated B-, so it is not in the list.

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5124", "E1039", "E1194", "E15", "E5116"]}

✅ R8 csv+axioms run 4 graded by code

13 entities, listed in the JSON. They are rated at the BB ceiling for VD (A12) or the A ceiling for CV (A37), for example E05 (A14) and E15 (A41).

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5091", "E5124", "E15", "E1039", "E1194", "E5116"]}

❌ R8 csv+axioms run 5 graded by code

12 entities, listed in the JSON. Eight sit at VD's BB ceiling (E05, E1028, E1101, E1126, E1236, E5012, E5047, E5124) and four at CV's A ceiling (E15, E1039, E1194, E5116).

JSON: {"entities": ["E05", "E1028", "E1101", "E1126", "E1236", "E5012", "E5047", "E5124", "E15", "E1039", "E1194", "E5116"]}