ESMA has been formally calling for improved data quality in Annex IV submissions since the first filing cycle in 2015. Its most recent annual statistical report on EU alternative investment funds carries the same message.
A decade of the same finding from the same regulator about the same reporting obligation is not a quality improvement programme making slow progress. It is evidence of a structural problem that the industry has repeatedly chosen to manage around rather than solve.
Understanding why the problem persists, and why AIFMD II will make it significantly more visible, requires looking past the data deficiencies themselves and at the operational architecture that produces them.
The problem is not the data. It is who owns it.
The standard framing of the Annex IV data quality challenge focuses on volume: between 400 and 800 data elements, depending on fund type, mapped to 340 template fields across 38 AIFM level and 302 AIF level data points.
The volume is real. But volume alone does not explain why the same types of errors appear in submissions year after year.
The more precise explanation is that Annex IV requires data from systems that were never designed to interoperate, assembled by people who were never given the authority or the tools to govern it end to end.
The fund administrator holds NAV, AuM and investor level data, but operates under a service agreement that defines what it maintains and what it does not. The risk analytics platform holds VaR, leverage calculations and stress test outputs, built for investment management purposes and using its own instrument master and calculation conventions.
The custodian holds position and instrument data, enriched against its own reference data vendor, which may not be the same vendor used by the administrator. The portfolio management system holds trading and exposure data, maintained to the standard required for investment decisions rather than regulatory reporting.
Annex IV requires data from all four of these sources, reconciled, enriched and mapped to a regulatory template that none of them were built to feed.
The person who can assemble it is typically a compliance or operations professional with access to all four systems and a sufficiently detailed understanding of the template requirements to know where each element needs to go. In most firms, that person is working in Excel.
Not because Excel is the right tool, but because no single system in the firm’s technology stack spans all four data sources, and no governance structure has ever formally assigned ownership of the cross system reconciliation to a specific function.
This is why the problem has persisted for a decade. The failure cost is borne by the operations team, through overtime, deadline pressure and the stress of reconciling data that was never designed to reconcile cleanly, rather than appearing anywhere on the firm’s risk register.
The process works, after a fashion, until the regulator flags it.
The Annex IV data quality problem has persisted for a decade not because firms lack expertise, but because ownership of the cross system reconciliation has never been formally assigned.
Why NCA validation and DQEF are not the same test
The operational distinction between passing NCA validation and passing ESMA’s Data Quality Engagement Framework is one that many compliance teams have not fully internalised. The difference matters.
NCA validation is a structural test. It confirms that the submitted XML is well formed, conforms to the required schema version and contains the mandatory fields in the correct format.
A file that passes NCA validation has been correctly constructed. It has not necessarily been correctly populated.
The DQEF is a substance test. It runs more than forty automated checks that cross reference reported values against external regulatory datasets, including GLEIF for LEI validity, EMIR trade repositories for derivative instrument coverage, and MiFID reference data for venue and instrument identifiers.
It also applies plausibility logic.
Does the reported leverage ratio make sense given the AuM and gross exposure figures? Does the fund’s reported liquidity profile align with the asset classes it holds? Is a fund reporting derivative trading in possession of valid, non expired LEIs in GLEIF?
A submission can satisfy NCA structural validation entirely and still generate multiple DQEF warnings.
When those warnings are passed to the NCA for follow up, which is the standard escalation path, the firm receives regulatory contact about a submission it considered closed. That sequence is more common than many compliance teams realise, because the DQEF operates downstream of the submission portal and its outputs are not always visible to the filer in real time.
Why identifier failures keep recurring
The most consistently flagged deficiencies in DQEF monitoring centre on identifier fields: missing or expired LEIs at the AIFM and AIF level, absent ISINs for portfolio instruments, and incomplete prime broker counterparty LEI data.
These are not obscure fields. They are foundational to the supervisory purpose of Annex IV. Regulators cannot aggregate and analyse cross industry data without consistent, machine readable entity and instrument identification.
The persistence of these deficiencies, a decade into the regime, reflects a specific operational failure: LEI maintenance has not been built into many firms’ continuous governance processes, so it degrades between reporting cycles and is patched, often incompletely, at filing time.
LEI expiry is a useful illustration of how the ownership gap operates in practice.
LEIs expire if not renewed annually. The GLEIF database records the expiry date, and the DQEF cross references it. An expired LEI in an Annex IV submission generates a DQEF warning regardless of whether the underlying entity information is accurate.
Renewing the LEI is a simple process. The reason it is frequently not done is that no one in the firm owns it as a standing obligation.
Compliance considers it a data management task. Operations considers it a compliance task. The fund administrator maintains what its service agreement specifies, which typically does not include LEI renewal for all entities in scope.
The result is that a trivially avoidable quality failure recurs every reporting cycle because accountability has fallen between three functions that each assume one of the others is handling it.
What ESMA’s decade of warning is actually signalling
The charitable interpretation of ESMA’s persistent data quality findings is that Annex IV is genuinely complex, the data is difficult to assemble and the industry is making incremental progress.
That interpretation is probably accurate. It is also beside the point.
ESMA’s data quality monitoring programme exists because the supervisory value of Annex IV depends entirely on the data inside it being accurate, consistent and comparable across submissions.
An Annex IV database populated with expired LEIs, missing ISINs and leverage metrics that fail plausibility checks against AuM data is not a systemic risk monitoring tool. It is a filing exercise.
The regulatory
purpose of the regime is not served by a submission that passes structural validation and fails substance testing.
Why AIFMD II raises the stakes
The direction of travel under AIFMD II makes the stakes considerably higher.
ESMA is expanding DQEF test coverage in parallel with the expansion of Annex IV’s data scope. The new reporting framework will require all instruments, all exposures, delegation data with FTE granularity and liquidity management tool disclosures.
Each of these new data categories will generate new DQEF test coverage.
Firms that arrive at the AIFMD II reporting regime with manual pipelines, fragmented data ownership and a pattern of recurring identifier deficiencies will not find that the expanded scope smooths itself out in operation.
They will find that the DQEF, now running against a significantly larger data surface, generates more warnings, more NCA follow up and more remediation cycles than they have capacity to manage.
The quality bar is not staying where it is. It is being raised at the same time as the data volume increases.
The gap between firms that have built governed, automated reporting infrastructure and firms that have not will become visible in the 2027 filing cycle in a way it has not been before.
Key takeaways
The Annex IV data quality problem has persisted for a decade because cross system data reconciliation has no formal owner in many firms. It sits in the gap between compliance, operations and the fund administrator.
NCA validation is a structural test confirming XML formation. ESMA’s DQEF is a substance test cross referencing reported values against GLEIF, EMIR and MiFID data. Satisfying the first does not mean satisfying the second.
LEI expiry is one of the most avoidable recurring quality failures in Annex IV submissions, and one of the clearest examples of what happens when governance accountability falls between functions.
ESMA is expanding DQEF test coverage in parallel with AIFMD II’s expanded data scope. Firms with manual pipelines are not facing a heavier version of the current standard. They are facing a structurally different one.
The question is not whether the data quality problem is known. It is whether anyone in your firm has been given the authority and infrastructure to solve it, or whether it continues to be managed around at each reporting cycle.
The question for SMF16 holders and COOs
For SMF16 holders and COOs, the test is specific.
Pick any single data field in your last Annex IV submission: the LEI of the largest counterparty, the ISIN of the top instrument by exposure, or the leverage figure for the most complex fund.
Now trace it.
Which system produced that value? Who validated it before it entered the template? What audit trail exists to demonstrate that validation happened?
In firms with governed reporting infrastructure, that trace is a matter of minutes.
In firms where the process lives in a spreadsheet and institutional knowledge, it is a reconstruction exercise. A regulator asking the question has already formed a view about the answer before you provide it.
AIFMD II will generate more of those questions, across a larger data surface, with less tolerance for the answer being unclear.
How Datox helps
Datox helps fund managers, fund administrators and compliance teams move from deadline driven reporting to governed regulatory reporting infrastructure.
By connecting data sources, standardising validation logic and creating a clear evidence trail from source data to final submission, Datox helps firms reduce manual effort, improve Annex IV data quality and prepare for the higher reporting standard introduced by AIFMD II.
To see how Datox can support your Annex IV reporting process, book a demo with our team.