COREP Reporting¶

The RWA Calculator generates COREP (COmmon REPorting) credit risk templates for regulatory submissions. These templates follow the EBA DPM taxonomy as defined in Regulation (EU) 2021/451.

Why COREP Matters¶

UK-regulated banks submit quarterly COREP returns to the PRA as part of ongoing supervisory reporting. The credit risk templates require firms to aggregate their exposure-level RWA calculations into standardised row/column formats by exposure class. Manual aggregation is error-prone and audit-unfriendly — generating templates directly from calculation results ensures consistency between the RWA engine output and the reported figures.

Template Overview¶

The calculator covers nine credit risk template families across SA, IRB, and geographical breakdowns. Each template is reported once per SA or IRB exposure class (or per country for geographical templates) — the exposure class acts as a filter, not a row dimension.

flowchart TD
    P["Pipeline Output<br/>(LazyFrame)"] --> SPLIT{"Split by<br/>approach"}

    SPLIT -->|SA exposures| C07["<b>C 07.00 / OF 07.00</b><br/>CR SA<br/><i>One submission per<br/>SA exposure class</i>"]

    SPLIT -->|IRB exposures| C0801["<b>C 08.01 / OF 08.01</b><br/>CR IRB Totals<br/><i>One submission per<br/>IRB exposure class</i>"]

    SPLIT -->|IRB exposures| C0802["<b>C 08.02 / OF 08.02</b><br/>CR IRB by Obligor Grade<br/><i>One submission per<br/>IRB exposure class</i>"]

    SPLIT -->|IRB exposures| C0803["<b>C 08.03 / OF 08.03</b><br/>CR IRB PD Ranges<br/><i>Fixed PD range buckets</i>"]

    SPLIT -->|IRB exposures| C0804["<b>C 08.04 / OF 08.04</b><br/>RWEA Flow Statements<br/><i>Period-over-period movements</i>"]

    SPLIT -->|IRB exposures| C0805["<b>C 08.05 / OF 08.05</b><br/>PD Back-Testing<br/><i>One submission per<br/>IRB exposure class</i>"]

    SPLIT -->|Slotting exposures| C0806["<b>C 08.06 / OF 08.06</b><br/>Specialised Lending Slotting<br/><i>By SL type</i>"]

    SPLIT -->|All IRB| C0807["<b>C 08.07 / OF 08.07</b><br/>Scope of Use<br/><i>SA vs IRB coverage</i>"]

    SPLIT -->|SA by country| C0901["<b>C 09.01 / OF 09.01</b><br/>Geo Breakdown SA<br/><i>Per country</i>"]

    SPLIT -->|IRB by country| C0902["<b>C 09.02 / OF 09.02</b><br/>Geo Breakdown IRB<br/><i>Per country</i>"]

    C0801 -.->|"Same columns,<br/>disaggregated by<br/>obligor grade"| C0802

    C0801 -.->|"Aggregated into<br/>fixed PD bands"| C0803

    style C07 fill:#fff3e0,stroke:#fb8c00
    style C0801 fill:#e3f2fd,stroke:#1e88e5
    style C0802 fill:#e3f2fd,stroke:#1e88e5
    style C0803 fill:#e3f2fd,stroke:#1e88e5
    style C0804 fill:#e3f2fd,stroke:#1e88e5
    style C0805 fill:#e3f2fd,stroke:#1e88e5
    style C0806 fill:#e8f5e9,stroke:#43a047
    style C0807 fill:#f3e5f5,stroke:#8e24aa
    style C0901 fill:#fce4ec,stroke:#e53935
    style C0902 fill:#fce4ec,stroke:#e53935

C 02.00 / OF 02.00 — Own Funds Requirements¶

C 02.00 / OF 02.00 is the master capital template — the single point at which RWEA across every risk type (credit, CCR, market, operational, CVA, securitisation, settlement) is aggregated into the firm's Total Risk Exposure Amount (TREA) and Total Own Funds Requirements. This calculator only populates the credit risk section (rows 0050–0420); all other risk-type rows are null because those risk types are outside its scope.

This is where the Basel 3.1 output floor is actually applied. The floor formula

TREA = max(U-TREA, x · S-TREA + OF-ADJ)

is evaluated row 0010 of OF 02.00, with col 0010 carrying the U-TREA components, col 0020 carrying the S-TREA components fed in from OF 02.01, and col 0030 carrying the floored result after the multiplier x and OF-ADJ are applied. See output-floor.md for the floor mechanics in detail.

Reference: CRR Art. 92 (own funds requirements); PRA PS1/26 Art. 92 para 2A / 3A / 5 (output floor); Regulation (EU) 2021/451 Annex I (CRR layout); PRA PS1/26 Annex I (OF 02.00 layout).

Column Structure¶

Row Structure¶

The row layout is significantly restructured under Basel 3.1 (PRA PS1/26 Annex II §1.3): F-IRB / A-IRB / Slotting are split into separate sub-sections, corporate and retail sub-class breakdowns are added, slotting is broken out by 5 SL types, and three new output-floor indicator rows are inserted.

CRR vs Basel 3.1 — Output Floor Delta¶

Implementation¶

OF 02.01 — Output Floor Comparison (Basel 3.1 only)¶

OF 02.01 is a Basel 3.1-only template with no CRR equivalent. It provides a dedicated output floor comparison for internal-model firms, showing modelled versus standardised total risk exposure amounts side by side by risk type. The template supplies the raw comparison data consumed by C 02.00 / OF 02.00 to apply the floor multiplier.

Reference: PRA PS1/26 Art. 92 para 2A / 3A

Column Structure¶

Row Structure¶

Implementation¶

C 07.00 / OF 07.00 — CR SA¶

Column Structure¶

The SA template has a wide column layout covering the full credit risk waterfall from original exposure through CRM to final RWEA. The column derivation flows left to right:

flowchart LR
    A["<b>0010</b><br/>Original<br/>Exposure"] --> B["<b>0030</b><br/>(-) Provisions"]
    B --> C["<b>0040</b><br/>Net Exposure"]
    C --> CRM["CRM Substitution<br/>Effects"]
    CRM --> D["<b>0050-0060</b><br/>Unfunded:<br/>Guarantees,<br/>Credit Derivs"]
    CRM --> E["<b>0070-0080</b><br/>Funded:<br/>Fin. Collateral,<br/>Other FCP"]
    CRM --> F["<b>0090/0100</b><br/>Substitution<br/>Out/Inflows"]
    F --> G["<b>0110</b><br/>Net After<br/>CRM Subst."]
    G --> FCCM["Fin. Collateral<br/>Comprehensive"]
    FCCM --> H["<b>0120</b><br/>Volatility Adj"]
    FCCM --> I["<b>0130</b><br/>(-) Cvam"]
    I --> J["<b>0150</b><br/>Fully Adjusted<br/>Exposure (E*)"]
    J --> K["<b>0160-0190</b><br/>Off-BS by CCF"]
    J --> L["<b>0200</b><br/>Exposure Value"]
    L --> M["<b>0220</b><br/>RWEA"]

    style A fill:#e8f5e9,stroke:#43a047
    style M fill:#fff3e0,stroke:#fb8c00

Full Column Reference¶

Row Structure¶

Each SA template submission (per exposure class) contains five row sections:

flowchart TD
    T["C 07.00 / OF 07.00<br/>Row Structure"] --> S1["<b>Section 1</b><br/>Total Exposures<br/><i>+ 'of which' breakdowns</i>"]
    T --> S2["<b>Section 2</b><br/>Breakdown by<br/>Exposure Types<br/><i>On-BS, Off-BS, CCR</i>"]
    T --> S3["<b>Section 3</b><br/>Breakdown by<br/>Risk Weights<br/><i>0% to 1250%</i>"]
    T --> S4["<b>Section 4</b><br/>Breakdown by<br/>CIU Approach"]
    T --> S5["<b>Section 5</b><br/>Memorandum<br/>Items"]

    style S1 fill:#fff3e0,stroke:#fb8c00
    style S3 fill:#e8f5e9,stroke:#43a047

# Shared label / group / section constants (DRY — referenced by multiple templates)
# =============================================================================

# --- SA exposure-class labels ---
_LBL_SA_CENTRAL_GOVT = "Central governments or central banks"
_LBL_IRB_CENTRAL_GOVT = "Central governments and central banks"
_LBL_SA_RGLA = "Regional governments or local authorities"
_LBL_SA_PSE = "Public sector entities"
_LBL_SA_MDB = "Multilateral development banks"
_LBL_SA_INTL_ORG = "International organisations"
_LBL_SA_MORTGAGES = "Secured by mortgages on immovable property"
_LBL_SA_DEFAULTED = "Exposures in default"
_LBL_SA_COVERED_BOND = "Covered bonds"
_LBL_SA_OTHER = "Other items"

# --- Common column / row labels ---

_LBL_COL_OBSERVED_DEFAULTS = "Observed new defaults for the period"
_LBL_COL_GEN_CRA = "General credit risk adjustments"
_LBL_COL_SPEC_CRA = "Specific credit risk adjustments"
_LBL_COL_CRA_WRITEOFFS = "Credit risk adjustments/write-offs for observed new defaults"
_LBL_ROW_TOTAL_EXPOSURES_UC = "TOTAL EXPOSURES"
_LBL_ROW_TOTAL_EXPOSURES_MC = "Total exposures"
_LBL_ROW_SFT_NETTING = "SFT netting sets"
_LBL_ROW_OF_WHICH_QCCP = "  of which: centrally cleared through a QCCP"
_LBL_ROW_DERIV_LST_NETTING = "Derivatives & Long Settlement Transactions netting sets"
_LBL_ROW_CCP_NETTING = "From Contractual Cross Product netting sets"
_LBL_ROW_CREDIT_RISK_EXCL_CCR = "Credit risk (excluding CCR)"
_LBL_ROW_OP_RISK = "Operational risk"
_LBL_ROW_CIU = "Collective investment undertakings (CIU)"
_LBL_ROW_OF_WHICH_SME_LC = "  of which: SME"

# --- Common group labels ---
_GRP_CRM_SUBST_UNFUNDED = "CRM Substitution: Unfunded"
_GRP_CRM_SUBST_FUNDED = "CRM Substitution: Funded"
_GRP_CRM_SUBST = "CRM Substitution"
_GRP_FIN_COLL_COMP = "Fin. Collateral Comprehensive"
_GRP_CCF_BREAKDOWN = "CCF Breakdown"
_GRP_EXPOSURE_VALUE = "Exposure Value"
_GRP_INTERNAL_RATING = "Internal Rating"
_GRP_CRM_LGD_UNFUNDED = "CRM in LGD: Unfunded"
_GRP_CRM_LGD_FUNDED = "CRM in LGD: Funded"
_GRP_OUTPUT_FLOOR = "Output Floor"
_GRP_COVERAGE_PCT = "Coverage %"
_GRP_RWEA_SA_BREAKDOWN = "RWEA: SA Breakdown"

# --- Section names ---

            COREPRow("0344", "  of which: SME (materially dependent)"),
            COREPRow("0350", "of which: Other real estate"),
            COREPRow("0351", "  of which: Residential (not mat. dependent)"),
            COREPRow("0352", "  of which: Residential (mat. dependent)"),
            COREPRow("0353", "  of which: Commercial (not mat. dependent)"),
            COREPRow("0354", "  of which: Commercial (mat. dependent)"),
            COREPRow("0360", "of which: Land ADC exposures"),
            # 0040 removed: replaced by detailed RE breakdown above
            COREPRow("0050", "of which: Exposures under permanent partial use of SA"),
            COREPRow("0060", "of which: Exposures under sequential IRB implementation"),
        ],
    ),
    RowSection(
        _SEC_BREAKDOWN_EXPOSURE_TYPES,
        [
            COREPRow("0070", "On balance sheet exposures subject to credit risk"),
            COREPRow("0080", "Off balance sheet exposures subject to credit risk"),

C 08.01 / OF 08.01 — CR IRB Totals¶

The IRB totals template is filtered by IRB exposure class and by own estimates of LGD/CCF (Foundation vs Advanced IRB). It covers the full IRB waterfall: original exposure, CRM substitution effects, CRM in LGD estimates (with detailed collateral breakdown), exposure value, LGD, maturity, RWEA, and memorandum items (expected loss, provisions, obligor count).

Column Structure¶

Row Structure¶

_LBL_COL_CREDIT_DERIVS = "(-) Credit derivatives"
_LBL_COL_OTHER_FUNDED = "(-) Other funded credit protection"
_LBL_COL_SUBST_OUTFLOWS = "(-) Substitution outflows"
_LBL_COL_SUBST_INFLOWS = "Substitution inflows (+)"
_LBL_COL_EXPOSURE_VALUE_LC = "Exposure value"
_LBL_COL_RWEA_PRE_SF = "RWEA pre supporting factors"
_LBL_COL_SME_SF_ADJ = "(-) SME supporting factor adjustment"
_LBL_COL_INFRA_SF_ADJ = "(-) Infrastructure supporting factor adjustment"
_LBL_COL_RWEA_POST_SF = "RWEA after supporting factors"
_LBL_COL_OF_WHICH_ECAI = "Of which: with ECAI credit assessment"
_LBL_COL_RWEA_HYPHEN = "Risk-weighted exposure amount"

            COREPRow("0020", "On balance sheet items subject to credit risk"),
            COREPRow("0030", "Off balance sheet items subject to credit risk"),
            COREPRow("0031", "  of which: off-BS by CCF 10%"),  # New in B3.1
            COREPRow("0032", "  of which: off-BS by CCF 20%"),
            COREPRow("0033", "  of which: off-BS by CCF 40%"),
            COREPRow("0034", "  of which: off-BS by CCF 50%"),
            COREPRow("0035", "  of which: off-BS by CCF 100%"),
            COREPRow("0040", _LBL_ROW_SFT_NETTING),
            COREPRow("0050", _LBL_ROW_DERIV_LST_NETTING),
            COREPRow("0060", _LBL_ROW_CCP_NETTING),
        ],
    ),
    RowSection(
        "Calculation Approaches",

C 08.02 / OF 08.02 — CR IRB by Obligor Grade¶

C 08.02 disaggregates C 08.01 by individual obligor grade or pool from the firm's internal rating system. It uses the same column structure as C 08.01 with the addition of an obligor grade identifier column.

Structure¶

# C 08.01 / OF 08.01 — IRB ROW SECTIONS
# =============================================================================

CRR_IRB_ROW_SECTIONS: list[RowSection] = [
    RowSection(
        "Total and Supporting Factors",
        [
            COREPRow("0010", _LBL_ROW_TOTAL_EXPOSURES_UC),
            COREPRow("0015", "of which: Exposures subject to SME-supporting factor"),
            COREPRow("0016", "of which: Exposures subject to infrastructure supporting factor"),
        ],
    ),

C 08.03 / OF 08.03 — CR IRB PD Ranges¶

C 08.03 aggregates IRB exposures into fixed PD range buckets for disclosure under Article 452(g). It provides a standardised view of key risk parameters (PD, LGD, CCF, RWEA, EL) across comparable PD bands. This template excludes slotting exposures (reported in C 08.06) and CCR exposures.

Column Structure¶

Row Structure¶

C 08.04 / OF 08.04 — CR IRB RWEA Flow Statements¶

C 08.04 reports quarter-over-quarter movements in IRB RWEA, decomposed into seven standardised driver categories. This template excludes CCR exposures. It is submitted once per IRB exposure class.

Column Structure¶

Row Structure¶

Implementation Notes¶

C 08.04 / OF 08.04 is implemented in v0.1.169. The pipeline provides current-period data only, so:

• Row 0090 (closing RWEA) is populated from rwa_final per exposure class
• Row 0010 (opening RWEA) and rows 0020–0080 (movement drivers) are null — they require prior-period comparison data that a single pipeline run cannot produce
• Slotting exposures are excluded (C 08.06 covers specialised lending separately)
• Template definitions: CRR_C08_04_COLUMNS, B31_C08_04_COLUMNS, C08_04_ROWS
• Generator: _generate_all_c08_04(), _generate_c08_04_for_class()
• Bundle field: COREPTemplateBundle.c08_04: dict[str, pl.DataFrame]

To populate the full flow statement, callers can supply prior-period RWEA externally and merge it with the generated template.

C 08.05 / OF 08.05 — CR IRB PD Back-Testing¶

C 08.05 reports PD model back-testing per IRB exposure class — comparing model-assigned PDs against realised one-year default rates across the same 17 fixed PD range buckets used in C 08.03. This template is the supervisory back-testing view that supports IRB model validation under CRR Art. 180. One submission is filed per IRB exposure class. Slotting and CCR exposures are excluded.

The B31 column 0010 label is explicit that the reported average PD is after the PD input floors (Art. 160(1) / 163(1)) and any exposure-level risk weight floors (Art. 161(3) / 164(5)). PD-bucket allocation, however, uses the pre-input-floor PD — the same convention as OF 08.03.

Column Structure¶

Row Structure¶

The 17 PD range buckets are identical to C 08.03 / OF 08.03. Defaulted exposures are assigned to the 100% bucket (row 0170).

Implementation Notes¶

C 08.05 / OF 08.05 is implemented in the generator. The template requires multi-year historical data that a single pipeline run cannot produce, so two columns fall back to current-period approximations when historical inputs are absent:

• Col 0010 (arithmetic average PD): populated from irb_pd_floored per PD bucket. CRR uses the floored PD for both reporting and allocation; Basel 3.1 uses irb_pd_original for allocation and irb_pd_floored for the col 0010 value.
• Col 0020 (prior-year obligor count): falls back to current-period obligor count (counterparty_reference.n_unique()) when prior_year_obligor_count is not provided.
• Col 0030 (defaulted during year): obligor-level count from is_defaulted, falling back to PD ≥ 1.0 when the default flag is absent.
• Col 0040 (observed default rate): col 0030 / col 0020.
• Col 0050 (historical annual default rate): falls back to col 0040 (current-period observed rate) when historical_annual_default_rate is not provided.
• Slotting exposures excluded (covered by C 08.06); CCR exposures excluded (covered by C 34.07 / OF 34.07).
• Template definitions: CRR_C08_05_COLUMNS, B31_C08_05_COLUMNS in src/rwa_calc/reporting/corep/templates.py.
• Generator: _generate_all_c08_05(), _generate_c08_05_for_class(), _compute_c08_05_values() in src/rwa_calc/reporting/corep/generator.py.
• Bundle field: COREPTemplateBundle.c08_05: dict[str, pl.DataFrame] (one entry per IRB exposure class).
• Excel sheet prefix: C 08.05 (CRR) / OF 08.05 (Basel 3.1), one sheet per exposure class.

To populate cols 0020 and 0050 with true multi-year data, callers can supply the optional prior_year_obligor_count and historical_annual_default_rate columns on the input LazyFrame; the generator will use them in preference to the single-period fallbacks.

C 08.06 / OF 08.06 — CR IRB Specialised Lending Slotting¶

C 08.06 reports specialised lending exposures subject to the supervisory slotting criteria under Article 153(5). Exposures are broken down by slotting category (1–5) and remaining maturity (< 2.5 years / ≥ 2.5 years). One submission covers all SL types; the SL type (project finance, object finance, etc.) acts as a filter dimension.

Column Structure¶

Row Structure¶

C 08.07 / OF 08.07 — CR IRB Scope of Use¶

C 08.07 reports the split of a firm's exposures between SA and IRB approaches, showing what proportion of each exposure class (CRR) or roll-out class (Basel 3.1) is subject to each approach. Significantly expanded in Basel 3.1 with detailed RWEA attribution by reason for SA use and materiality thresholds.

Column Structure¶

Row Structure¶

C 09.01 / OF 09.01 — CR GB 1 (Geographical Breakdown SA)¶

C 09.01 provides a geographical breakdown of SA exposures by country of obligor residence. It is submitted once at total level and once per material country (threshold defined in Art 5(5) of the Reporting (CRR) Part).

Original exposure pre-conversion factors is reported by country of the immediate obligor. Exposure value and RWEA are reported by country of the ultimate obligor (after CRM substitution effects).

Column Structure¶

Row Structure¶

C 09.02 / OF 09.02 — CR GB 2 (Geographical Breakdown IRB)¶

C 09.02 provides a geographical breakdown of IRB exposures by country of obligor residence, including PD, LGD, and expected loss parameters. Submitted once at total level and once per material country.

Column Structure¶

Row Structure¶

Overall CRR vs Basel 3.1 Template Comparison¶

Usage¶

Generate from Pipeline Results¶

from rwa_calc.reporting import COREPGenerator

generator = COREPGenerator()

# From a LazyFrame of calculation results
bundle = generator.generate_from_lazyframe(results, framework="CRR")

# From a CalculationResponse (uses cached Parquet)
bundle = generator.generate(response)

# Access templates as DataFrames
print(bundle.c07_00)      # C 07.00 SA credit risk
print(bundle.c08_01)      # C 08.01 IRB totals
print(bundle.c08_02)      # C 08.02 IRB by PD grade
print(bundle.c08_05)      # C 08.05 IRB PD back-testing (per exposure class)
print(bundle.c07_rw_breakdown)  # C 07.00 risk weight breakdown

Export to Excel¶

from pathlib import Path
from rwa_calc.reporting import COREPGenerator

generator = COREPGenerator()
bundle = generator.generate_from_lazyframe(results)

# Export multi-sheet Excel workbook
result = generator.export_to_excel(bundle, Path("corep_templates.xlsx"))
# Creates sheets: "C 07.00", "C 07.00 RW Breakdown", "C 08.01", "C 08.02"

print(result.format)      # "corep_excel"
print(result.row_count)   # Total rows across all sheets

Regulatory References¶