AnaCredit Reporting: Automated SDMX-ML File Generation

For credit institutions and foreign branches operating in Luxembourg, AnaCredit Reporting (Analytical Credit Datasets) reporting is one of the most technically demanding regulatory obligations they face. Mandated by ECB Regulation (EU) 2016/867 and transposed in Luxembourg via BCL Circular 2017/240, AnaCredit requires the monthly and quarterly transmission of highly granular credit and credit risk data in a strict SDMX-ML XML format validated against BCL-specific XSD schemas.

The complexity is real: multiple datasets, hundreds of validation rules, counterparty identifiers, LEI codes, instrument attributes, financial data, protection information — all packaged into structurally precise XML files that must pass both BCL and ECB quality checks before they are accepted.

Many compliance teams are still building these files manually, or relying on fragile in-house scripts that break with every schema update. Fund XP’s AnaCredit SDMX Reporting Generator gives your team a familiar, auditable, and fully automated path from raw credit data to submission-ready SDMX-ML files — directly from Microsoft Excel, no XML coding required.

---

What Is AnaCredit and Who Is Required to Report?

AnaCredit stands for Analytical Credit Datasets. It is a pan-European granular credit database built by the European Central Bank to collect loan-by-loan data on credit exposures granted by credit institutions to non-financial corporations and other legal entities.

Under Article 3 of ECB Regulation 2016/867, the reporting population in Luxembourg covers all credit institutions and foreign branches resident in Luxembourg, independently of their legal status. These reporting agents must transmit their data to the Banque centrale du Luxembourg (BCL), which forwards it to the ECB.

Reporting covers instruments such as loans, credit lines, overdrafts, deposits, credit card debt, revolving credit, trade receivables, and financial leases, provided the debtor exposure exceeds the applicable threshold. The data collected spans ten distinct datasets:

• Counterparty reference data (LEI, national identifier, institutional sector, legal form, enterprise size, economic activity)
• Instrument data (type, amortisation, currency, interest rate, maturity, purpose)
• Financial data (outstanding nominal amount, accrued interest, default status, arrears)
• Counterparty-instrument relationships (creditor, debtor, servicer, originator roles)
• Joint liabilities
• Accounting data (carrying amount, impairment stage, forbearance, IFRS classification)
• Protection received (collateral type, value, valuation approach)
• Instrument-protection linkages
• Counterparty risk data (probability of default)
• Counterparty default data (default status per counterparty)

---

What Makes AnaCredit Reporting So Technically Challenging?

Is the SDMX-ML XML format really that complex?

Yes — and underestimating it is one of the most common sources of reporting failures.

AnaCredit files must be submitted as SDMX-ML StructureSpecificData XML, validated against BCL-specific XSD schemas (currently version 1.0.13, covering files such as BCL_ANCRDT_T1M_v1_0_13.xsd, BCL_ANCRDT_T2Q_v1_0_13.xsd, and BCL_ANCRDT_T1_REF_v1_0_13.xsd, together with the ECB and BCL constraint overlays). The file structure requires a correctly formed header section referencing multiple BCL data structure definitions, followed by separate dataset blocks for each cube — instruments, financials, accounting, counterparty-instrument relationships, protection, and more.

Each attribute must be encoded using BCL-specific coded values. For instance, the instrument type “Loans other than overdrafts…” is transmitted as code 1004; “Non-fiduciary instrument” as 2; “Monthly” payment frequency as 8; “Variable” interest rate type as 2. A single wrong code, a missing namespace declaration, or an out-of-sequence element causes the file to fail schema validation immediately.

Three separate transmission types feed the BCL at different frequencies: T1M (monthly, covering instruments, financials, counterparty-instrument, joint liabilities, and counterparty reference), T2Q (quarterly, covering accounting data), and T2M (covering protection received, instrument-protection, counterparty risk, and counterparty default data). Each has its own schema and its own BCL survey code.

How many validation rules does AnaCredit actually enforce?

Fund XP’s tool embeds 73 structured validation checks across seven quality dimensions, directly mirroring the BCL’s own data quality framework:

• Data Compliance — submission type, sender RIAD code, reporting year/month, and survey code must be valid before any data is processed
• Uniqueness — counterparty identifiers, LEI codes, instrument key pairs (Contract ID + Instrument ID), protection identifiers, and counterparty default/risk records must all be unique within their respective datasets
• Referential Integrity — every financial record must match an instrument record; every financial record must have both a creditor and a debtor in the counterparty-instrument dataset; every accounting record must match a financial record; every instrument-protection record must reference a protection received entry; every counterparty referenced anywhere must exist in the counterparty reference dataset
• Completeness — mandatory attributes (type of instrument, currency, inception date, legal final maturity date, interest rate, default status, outstanding nominal amount, off-balance-sheet amount, impairment fields, and key counterparty attributes) must be present
• Consistency (BCL) — enterprise size and its date must both be reported together or neither; legal proceedings status and date must be consistent; other entity identifier and code must align
• Consistency (ECB) — settlement date must be on or after inception date; legal final maturity date must be on or after settlement date; inception date must not be in the future relative to the reporting period; default and past-due dates must be consistent with inception date
• Plausibility — record counts for all ten datasets are checked against a 0.66–1.5 ratio versus the prior reporting period, flagging implausible volume changes before submission

Seventeen of these checks are classified Priority 1 — Critical, meaning a failure causes file or record rejection at the BCL. A further twenty-eight are Priority 2 — High (correction required). The remainder are Priority 3 — Medium (correction recommended). The Validation Dashboard sheet recalculates automatically as data is entered, giving your team real-time visibility of the risk profile of the current submission.

What submission parameters does the BCL require?

Each transmission requires a correctly populated header: sender RIAD code, reporting agent RIAD ID, observed agent RIAD ID, reference year and month, survey code (e.g. BCL_ANCRDT_T1M), submission type (FULL_REPLACEMENT or delta), and emergency correction flag. The tool’s Header sheet exposes all of these fields in a structured form and applies validation checks TR0041, TR0120, TR-YEAR, TR-MONTH, and TR-SURVEY against them before any XML is generated.

---

How Does Fund XP’s Excel Template Solution Work?

What does the tool actually produce?

The Fund XP AnaCredit SDMX Reporting Generator is a structured Excel workbook that serves as both a data entry interface and a validation engine. It covers all ten AnaCredit datasets across the three BCL transmission types, with each dataset on its own dedicated worksheet:

• CounterParty Reference (T1R / BCL_ANCRDT_ENTTY_C)
• Instruments (T1M / BCL_ANCRDT_INSTRMNT_C)
• Financial dataset (T1M / BCL_ANCRDT_FNNCL_C)
• Counterparty Instrument Static (T1M / BCL_ANCRDT_ENTTY_INSTRMNT_C)
• Joint Liabilities (T1M / BCL_ANCRDT_JNT_LBLTS_C)
• Accounting (T2Q / BCL_ANCRDT_ACCNTNG_C)
• Protection received (T2M / BCL_ANCRDT_PRTCTN_RCVD_C)
• Instrument-protection (T2M / BCL_ANCRDT_INSTRMNT_PRTCTN_RCVD_C)
• Counterparty Risk data (T2M / BCL_ANCRDT_ENTTY_RSK_C)
• Counterparty default data (T2M / BCL_ANCRDT_ENTTY_DFLT_C)

Each column maps precisely to its corresponding SDMX attribute name (e.g. ENTTY_ID, LEI, CNTRCT_ID, INSTRMNT_ID, ANNLSD_AGRD_RT, OTSTNDNG_NMNL_AMNT) as defined in the BCL’s Mapping sheet, complete with the regex pattern that governs valid values. When the workbook is complete, the generator produces a properly structured SDMX-ML XML file ready for BCL submission.

How do coded values and dropdowns work?

One of the most error-prone aspects of manual AnaCredit reporting is the translation between human-readable labels and BCL transmission codes. The tool handles this automatically through its built-in mapping layer.

For every coded attribute, the Mapping sheet defines the full translation table. For example:

• Type of instrument: “Loans other than overdrafts…” → 1004; “Overdrafts” → 20; “Credit lines other than revolving credit” → 1002; “Credit card debt” → 51; “Trade receivables” → 71; “Finance leases” → 80
• Institutional sector: “S.11 Non-financial corporations” → S11; “S.122A Credit institutions” → S122_A
• Default status: “Not in default” → 14; “Default because unlikely to pay” → 19; “Default because more than 90/180 days past due” → 20
• Interest rate type: “Fixed” → 1; “Variable” → 2; “Mixed” → 3
• Performing status: “Performing” → 11; “Non-performing” → 1
• Type of impairment: “Stage 1 (IFRS)” → 23; “Stage 2 (IFRS)” → 24; “Stage 3 (IFRS)” → 25

Your team works entirely in plain-language values. The generator substitutes the correct XML codes at file creation time, with no manual lookup required.

Date format handling is equally important: the Mapping sheet specifies the input date format (MM/DD/YYYY) and all dates are automatically converted to the ISO 8601 format required by the BCL schemas at output.

What happens when the Validation Dashboard detects errors?

The Validation Dashboard provides a live, auto-recalculating view of all 73 checks. Each rule displays its Rule ID, category, description, priority level, impact on submission, and an error count or detail field. As data is entered or updated across the workbook’s ten data sheets, the dashboard refreshes immediately.

Priority 1 errors (Critical) must be resolved before file generation is attempted — the tool will not produce an XML output that carries known fatal issues. Priority 2 errors (High) require correction before submission. Priority 3 errors (Medium) are flagged for review but do not block generation, allowing the team to make an informed decision on whether to correct or proceed.

This sequence — enter data, review dashboard, correct errors, generate XML — replaces the painful cycle of submitting a file, receiving a BCL quality feedback report days later, deciphering hundreds of error codes, making corrections, and resubmitting.

---

What Are the Key Benefits for Compliance Teams?

Does it reduce the risk of BCL rejection and quality feedback errors?

This is where the tool delivers its most immediate value. AnaCredit quality feedbacks from the BCL can contain thousands of individual error observations, and the correction deadline — the 15th working day after the reference period — leaves little room for manual debugging.

By running the same validation logic the BCL applies, before you ever generate a file, the tool catches errors at the point of data entry rather than after submission. Referential integrity violations — such as a debtor referenced in the financial dataset who has no counterparty reference record, or a protection identifier used in an instrument-protection row that has no corresponding entry in the protection received sheet — are flagged immediately, not three weeks later.

Does it handle the specific Luxembourg reporting requirements?

Yes. The tool is purpose-built for BCL reporting, not a generic SDMX tool retrofitted for AnaCredit. It incorporates the BCL’s Luxembourg-specific requirements, including the correct survey codes for each transmission type (BCL_ANCRDT_T1M, BCL_ANCRDT_T2Q, BCL_ANCRDT_T2M, BCL_ANCRDT_T1_REF), the BCL’s RIAD code structure for reporting agents and observed agents, and the BCL’s specific treatment of national identifiers (RCS numbers for Luxembourg counterparties, CSSF matricules for investment funds and sub-funds).

The tool also handles the FULL_REPLACEMENT submission type that the BCL requires as the baseline transmission mode, validated by rule TR0041 as a Critical check — ensuring this mandatory parameter is never accidentally omitted or incorrectly set.

Can multiple observed agents be managed within the same workbook?

The Header sheet supports separate configuration of the Reporting Agent RIAD ID and the Observed Agent RIAD ID, covering the common Luxembourg scenario where a legal entity reports both its own book and the books of foreign branches under a single BCL reporting relationship. Each generated XML file carries the correctly scoped agent identifiers in its header.

---

Is the Tool Current with the Latest BCL Requirements?

Which schema version does it support?

The tool is built against BCL AnaCredit schema version 1.0.13, the current production schema. This covers all relevant XSD files distributed by the BCL: BCL_ANCRDT_T1M_v1_0_13.xsd, BCL_ANCRDT_T2Q_v1_0_13.xsd, BCL_ANCRDT_T1_REF_v1_0_13.xsd, and the BCL and ECB constraint overlays (BCL_LU_ANCRDT_C_CONSTRAINTS_v1_0_13.xsd, ECB_LU_ANCRDT_C_CONSTRAINTS_v1_0_13.xsd). The internal version of the tool is tracked on the Header sheet (V2026.01), making it straightforward to verify alignment with BCL releases.

Does it reflect recent BCL reporting instruction updates?

Yes. The tool reflects the BCL’s current reporting instructions (version 4.6, October 2025), including the updated Data Quality Indicator framework effective from November 2025. The plausibility checks (PC0010–PC0090) in the Validation Dashboard incorporate the BCL’s 0.66–1.5 period-over-period ratio thresholds for each of the ten dataset cubes, covering both monthly and quarterly transmission cycles.

---

What Does Getting Started Look Like?

How quickly can my team produce a first file?

The workbook is self-contained and requires no installation, server infrastructure, or specialist technical configuration. Your team opens the Excel file, configures the Header sheet (sender RIAD code, reporting agent and observed agent identifiers, year, month, survey, and submission type), then begins entering data into the ten dataset sheets using familiar Excel functionality.

The Mapping sheet serves as a built-in reference guide, showing every field, its column mapping, its coded value translations, and the regex pattern that defines a valid value — meaning your team has full transparency into exactly what the BCL’s schema expects, without needing to consult the XSD files directly.

For institutions importing data from a core banking system, the sheet structure and column assignments in the Mapping configuration allow straightforward CSV-to-Excel population, followed by a Validation Dashboard review before XML generation.

---

Ready to Simplify Your AnaCredit Reporting?

AnaCredit compliance does not have to mean months of XML development, repeated quality feedback cycles, and deadline pressure every 15th working day. Fund XP’s AnaCredit SDMX Reporting Generator gives your team a structured, validated, and fully auditable path to submission-ready SDMX-ML files — month after month, quarter after quarter.

Whether you are preparing for your first AnaCredit submission, recovering from recurring quality feedback issues, or simply looking to eliminate the manual effort your current process requires, Fund XP has the solution.

Contact Fund XP today at info@fund-xp.lu to request a demonstration and discuss how the tool fits your institution’s reporting structure.

---

🏛️ BCL AnaCredit Hub

Main entry point for all BCL AnaCredit documentation:  bcl.lu — AnaCredit

---

 1. Regulations, Manual, Circulars & Circular Letters

Last updated: 16 Jan 2024  Overview page

---

2. Reporting Instructions

Last updated: 19 Feb 2026  Overview page

Core operational documents:

Cross-report comparison methodology:

SDMX Schemas — current and historical:

 3. Reference Documents (BCL → ECB)

Last updated: 10 Nov 2025  Overview page

---

 4. FAQ

 BCL FAQ page

The BCL’s FAQ redirects to the ECB’s centralised Q&A portal, which is the authoritative resource for conceptual and methodological questions:  AnaCredit Q&A — ECB

---