Q7, run 2026-11-04-1711-sonnet

← back to the run

Which issuers were downgraded in 2026?

Tests: Action types: outlook changes, Watch placements and withdrawals are not downgrades.

Expected: issuers = ['E08', 'E10']

❌ R1 csv run 1 graded by code

E08, E10 and E16. They were downgraded in A21/A22, A27 and A44 respectively.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R1 csv run 2 graded by code

E08, E10 and E16 were downgraded in 2026 (A21 and A22, A27, A44). Instrument downgrades such as A29 are not issuer downgrades.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R1 csv run 3 graded by code

E08, E10 and E16. They were downgraded in A21/A22, A27 and A44. E06's action (A17) is only a watch.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R1 csv run 4 graded by code

E08, E10 and E16. The issuer downgrades in 2026 are A21 and A22 (E08), A27 (E10) and A44 (E16). I05's downgrade (A29) is an instrument action.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R1 csv run 5 graded by code

E08, E10 and E16 had issuer-rating downgrades in 2026 (A21, A22, A27, A44). I05's downgrade (A29) is on an instrument, not an issuer rating.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R2 rdf run 1 graded by code

E08 (A21, A22), E10 (A27) and E16 (A44). I05's downgrade (A29) is on the instrument, so I did not count its issuer E11.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R2 rdf run 2 graded by code

E08, E10 and E16. Their issuer-level downgrades in 2026 are A21/A22, A27 and A44. I05's downgrade (A29) is an instrument rating.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R2 rdf run 3 graded by code

E08, E10 and E16. Their issuer downgrades are A21/A22, A27 and A44. I05's downgrade (A29) is an instrument rating. E16 comes from a "QA" source system, which the data does not explain.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R2 rdf run 4 graded by code

E08, E10 and E16. Their downgrades are A21 and A22, A27, and A44. I05 (A29) was also downgraded, but that is an instrument rating.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R2 rdf run 5 graded by code

E08, E10 and E16. Their 2026 downgrades are A21 and A22 for E08, A27 for E10 and A44 for E16. I05's downgrade (A29) is an instrument rating, so I excluded it.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

✅ R6 rdf+ontology run 1 graded by code

E08 and E10 were downgraded in 2026 (A21, A22, A27). A44 is on E16, a QA record, so it is ignored. A29 is an instrument downgrade, not an issuer downgrade.

JSON: {"issuers": ["E08", "E10"]}

✅ R6 rdf+ontology run 2 graded by code

E08 and E10. E08 was downgraded in A21 and A22, and E10 in A27. E16 is QA data and is ignored, and A29 concerns an instrument.

JSON: {"issuers": ["E08", "E10"]}

✅ R6 rdf+ontology run 3 graded by code

E08 and E10. They were downgraded by A21/A22 and A27. E16's downgrade (A44) is QA test data and is ignored. I05's downgrade is an instrument rating, not an issuer's.

JSON: {"issuers": ["E08", "E10"]}

✅ R6 rdf+ontology run 4 graded by code

E08 and E10. E08 was downgraded in A21 and A22, and E10 in A27. A44 (E16) is QA test data and is ignored. A29 is an instrument downgrade, not an issuer downgrade.

JSON: {"issuers": ["E08", "E10"]}

✅ R6 rdf+ontology run 5 graded by code

E08 and E10. The 2026 downgrades (DG) are A21 and A22 for E08 and A27 for E10 (E17 is the same entity). A44 is a QA record and is ignored. A29 is an instrument downgrade, not an issuer one.

JSON: {"issuers": ["E08", "E10"]}

✅ R5 csv+ontology run 1 graded by code

E08 and E10. E08 was downgraded (A21, A22) and E10 was downgraded (A27). The QA record E16 (A44) is ignored, and I05 is an instrument, not an issuer.

JSON: {"issuers": ["E08", "E10"]}

✅ R5 csv+ontology run 2 graded by code

E08 and E10. E08 was downgraded in A21 and A22, and E10 in A27; A44 is on E16, a QA record, so it is ignored.

JSON: {"issuers": ["E08", "E10"]}

✅ R5 csv+ontology run 3 graded by code

E08 (A21, A22) and E10 (A27). A29 downgrades an instrument, not an issuer, and E16's A44 is QA data, so it is ignored.

JSON: {"issuers": ["E08", "E10"]}

✅ R5 csv+ontology run 4 graded by code

E08 and E10. E08 was downgraded by A21 and A22, and E10 by A27. I05's downgrade (A29) is an instrument rating, and E16 is QA data and ignored.

JSON: {"issuers": ["E08", "E10"]}

✅ R5 csv+ontology run 5 graded by code

E08 and E10. E08 was downgraded in A21 and A22, and E10 in A27 (E17 is the same entity). E16's downgrade (A44) is QA data and is ignored; E06's action (A17) is a Watch, not a downgrade.

JSON: {"issuers": ["E08", "E10"]}

✅ R4 csv+markdown run 1 graded by code

E08 and E10. E08 was downgraded in A21 and A22, and E10 in A27 (E17 is the same entity). A29 is an instrument rating, and A44 is a QA record, so it is ignored.

JSON: {"issuers": ["E08", "E10"]}

✅ R4 csv+markdown run 2 graded by code

E08 and E10. E08 was downgraded in A21 and A22, and E10 in A27 (E17 is the same entity). A44 is on E16, a QA record, so it is ignored.

JSON: {"issuers": ["E08", "E10"]}

✅ R4 csv+markdown run 3 graded by code

E08 and E10. E08 was downgraded in A21 and A22, and E10 in A27 (E17 is the same entity as E10). E16's downgrade (A44) is QA data and is ignored. I05's downgrade (A29) is an instrument rating, not an issuer rating.

JSON: {"issuers": ["E08", "E10"]}

✅ R4 csv+markdown run 4 graded by code

E08 and E10. E08 was downgraded by A21 and A22, and E10 by A27. A44 (E16) is QA data and is ignored, and A29 is an instrument action.

JSON: {"issuers": ["E08", "E10"]}

✅ R4 csv+markdown run 5 graded by code

E08 and E10. E08 had downgrades A21 and A22, and E10 had A27. A29 downgrades an instrument (I05), and A44 is on QA record E16, which is ignored.

JSON: {"issuers": ["E08", "E10"]}

❌ R3 csv+padding run 1 graded by code

E08, E10 and E16 had issuer-rating downgrades in 2026 (A21, A22, A27, A44). I05 (A29) is an instrument downgrade, not an issuer one.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R3 csv+padding run 2 graded by code

E08, E10 and E16. A21/A22 downgraded E08, A27 downgraded E10 and A44 downgraded E16. I counted issuer ratings only, so the 2026 downgrade of the instrument I05 does not add E11.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R3 csv+padding run 3 graded by code

E08 (A21, A22), E10 (A27) and E16 (A44) were downgraded in 2026. Orrin's A17 is only a watch action, not a downgrade.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R3 csv+padding run 4 graded by code

E08, E10 and E16. Their 2026 downgrades are A21/A22, A27 and A44. A17 for E06 is only a watch, not a downgrade.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']

❌ R3 csv+padding run 5 graded by code

E08, E10 and E16 were downgraded in 2026 (A21, A22, A27, A44). E06's action was a watch, not a downgrade (A17). The I05 downgrade (A29) is an instrument action.

JSON: {"issuers": ["E08", "E10", "E16"]}

Why: issuers: expected ['E08', 'E10'], got ['E08', 'E10', 'E16']