If you're an investment firm authorised in the Netherlands, quarterly prudential reporting to De Nederlandsche Bank (DNB) isn't optional — it's a core operational requirement. Yet it's one of the areas where firms most consistently underestimate the effort involved, particularly the first time around.
This article is aimed at CFOs, COOs, and compliance officers at investment firms that are either new to DNB reporting or struggling with the process. It covers the practical mechanics: what you report, how the taxonomy works, what the DLR portal expects, and where firms typically go wrong.
The regulatory framework: IFR/IFD in brief
Since June 2021, investment firms in the EU are subject to the Investment Firm Regulation (IFR, Regulation 2019/2033) and the Investment Firm Directive (IFD, Directive 2019/2034). These replaced the previous CRR/CRD IV framework for most investment firms and introduced a proportionate prudential regime.
Under IFR/IFD, investment firms are classified into three categories:
- Class 1 — systemically important firms, essentially treated as banks. Subject to CRR/CRD.
- Class 2 — firms exceeding at least one threshold in Article 12(1) IFR. Subject to full K-factor requirements and detailed reporting.
- Class 3 — small, non-interconnected firms below all thresholds. Simplified reporting regime.
Most Dutch investment firms, including MTF and OTF operators, fall into Class 2. This means they're subject to the full set of K-factor capital requirements and the corresponding quarterly reporting obligations.
What gets reported — and when
DNB requires quarterly prudential reports from all authorised investment firms. The reporting deadline is typically 45 calendar days after the end of each quarter (e.g., 14 February for Q4, 15 May for Q1).
The core reporting templates cover:
- Own funds composition — CET1, AT1, T2 capital, deductions, and total own funds
- Capital requirements — the higher of: permanent minimum capital (€75K–€750K depending on licence type), fixed overhead requirement (25% of prior year fixed overheads), or K-factor requirement
- K-factor calculations — depending on your activities, this may include K-AUM, K-CMH, K-ASA, K-COH, K-DTF, K-NPR, K-CMG, K-TCD, K-CON
- Liquidity requirement — at least one-third of the fixed overhead requirement held in liquid assets
- Concentration risk — exposures to individual clients or groups of connected clients
- ICAAP summary — annual, but informs your quarterly capital adequacy assessment
Which K-factors apply to you?
Not all K-factors apply to every firm. An MTF operator, for instance, typically faces K-DTF (daily trading flow) and potentially K-CON (concentration risk), but not K-AUM (assets under management). Getting this mapping right from the start avoids reporting data you don't need — and missing data you do.
The XBRL taxonomy: what you're actually filing
DNB requires all prudential reports in XBRL (eXtensible Business Reporting Language) format. If you've never worked with XBRL before, the learning curve is steeper than most firms expect.
XBRL is not a spreadsheet format. It's a structured data standard where every data point is tagged with a specific taxonomy concept, linked to reporting dimensions, and validated against business rules. The taxonomy defines:
- Concepts — the individual data points (e.g., "CET1 capital after deductions")
- Dimensions — how data points are sliced (e.g., by counterparty type, instrument category)
- Filing indicators — which templates you're required to file (based on your classification and activities)
- Validation rules — cross-checks between data points and mathematical consistency rules
DNB uses a custom taxonomy aligned with EBA's IFR reporting framework but with Dutch-specific extensions. The taxonomy version changes periodically — typically annually — and each version change requires updating your mapping and potentially your reporting tool.
The DLR portal: filing mechanics
All filings go through DNB's DLR (Digitaal Loket Rapportages) portal. The process works as follows:
- Generate XBRL instance document — either using reporting software or a manual tool. The file must conform to the active taxonomy version.
- Upload to DLR — the portal accepts .xbrl files and runs immediate validation. You'll get a validation report within minutes.
- Address validation errors — the portal distinguishes between blocking errors (filing rejected) and warnings (filing accepted but flagged). All blocking errors must be resolved before the filing is accepted.
- Submit — once validation passes, you confirm submission. The filing is timestamped.
- Resubmission — if you discover errors post-submission, you can resubmit. DNB tracks the version history.
Access to the DLR portal requires eHerkenning (the Dutch government authentication system) at level EH3 or higher. Setting up eHerkenning for the first time takes 2–4 weeks, so don't leave this until your first filing deadline.
Tool choice matters
Some firms attempt to build XBRL files manually using XML editors or Excel-based tools. This works for Class 3 firms with minimal reporting, but it's impractical for Class 2 firms. Invest in proper reporting software — either a dedicated XBRL tool (e.g., Invoke, Parsity, BR-AG) or build a custom pipeline if your data volumes justify it. The upfront cost is typically recovered within two quarters of avoided manual errors and rework.
Data sourcing and the mapping problem
The biggest operational challenge in DNB reporting isn't filing the XBRL — it's sourcing the underlying data and mapping it to taxonomy concepts. This is where most of the effort goes, especially in the first year.
Typical data sources include:
- General ledger / accounting system — for own funds composition, fixed overheads, P&L data
- Trading / position management system — for K-factor inputs (trading volumes, positions, exposures)
- Risk management system — for concentration risk, counterparty exposures
- HR / payroll system — for staff-related fixed overhead components
The mapping exercise involves documenting exactly which data field from which source system maps to which XBRL taxonomy concept, with what transformation logic. For a Class 2 firm with moderate complexity, expect 150–300 individual data point mappings.
This mapping document becomes a critical operational artifact — it's the bridge between your business reality and the regulatory report. Keep it versioned, reviewed, and updated whenever source systems or taxonomy versions change.
Common filing errors
Based on our experience supporting firms through this process, the most common errors fall into five categories:
- Filing indicator mismatches. Firms either file templates they don't need to file (wasted effort) or fail to file required templates (blocking error). The filing indicators must match your firm classification and licensed activities precisely.
- Sign convention errors. The taxonomy expects specific sign conventions — some values must be positive, others negative. A capital deduction entered as a positive number instead of negative will either fail validation or produce an incorrect capital figure.
- Dimensional misalignment. XBRL data points often require specific dimensional breakdowns. Reporting a total where the taxonomy expects a dimensional split (or vice versa) causes validation failures.
- Fixed overhead calculation errors. The fixed overhead requirement uses a specific regulatory definition of "fixed overheads" that doesn't perfectly align with accounting definitions. Variable compensation, profit shares, and certain non-recurring costs must be excluded. Getting this wrong directly impacts your capital requirement.
- Stale taxonomy versions. Filing against an outdated taxonomy version is a blocking error. When DNB publishes a new taxonomy version (typically annually), you must update your reporting tool and mappings before the first filing deadline under the new version.
Building a sustainable reporting process
The goal isn't just to survive each quarter's filing deadline — it's to build a process that's repeatable, auditable, and doesn't consume disproportionate management attention. The firms that do this well share a few characteristics:
- Automation where possible. Automate data extraction from source systems, K-factor calculations, and XBRL file generation. Manual processes don't scale and introduce errors.
- Clear ownership. Assign explicit responsibility for the reporting process — from data sourcing through filing. The worst outcomes happen when reporting is a shared responsibility with no single owner.
- Pre-filing dry runs. Generate and validate your XBRL file at least one week before the deadline. This gives you time to identify and resolve issues without deadline pressure.
- Reconciliation to accounts. Your regulatory report should reconcile to your audited financial statements. Where there are legitimate differences (regulatory adjustments, classification differences), document them.
- Taxonomy change management. When DNB publishes a new taxonomy version, assess the impact on your mappings and reporting tool immediately — not a week before the first filing deadline under the new version.
Timeline for first-time filers
If you're setting up DNB reporting for the first time (e.g., after receiving your investment firm licence), a realistic timeline looks like this:
- Month 1–2: Select reporting tool, set up eHerkenning, obtain DLR portal access
- Month 2–3: Complete data point mapping exercise, configure reporting tool
- Month 3–4: Perform dry run filings, resolve validation errors, train relevant staff
- Month 4+: First live quarterly filing
Starting this process 4 months before your first filing deadline is the minimum. 6 months is more comfortable and accounts for the inevitable delays in tool procurement, eHerkenning setup, and internal data readiness.
Need help with DNB reporting?
Whether you're setting up reporting for the first time or cleaning up a process that isn't working, a 30-minute call will help clarify the path forward.
Schedule a ConsultationFree: MiFID II Compliance Checklist
50-point checklist covering all key obligations for investment firms — including prudential reporting requirements.