ESMA’s 2025 Data Quality Report: Why SFTR Accuracy Matters Beyond Reconciliation

ESMA’s 2025 Report on Quality and Use of Data confirms that SFTR data is no longer just a regulatory obligation. It is now used by the ECB for secured funding and exposure analysis, by the IMF for liquidity stress testing and by NCAs to challenge firms on inconsistencies and unusual reporting patterns. Yet loan and collateral matching rates remain low and reconciliation outcomes alone do not reveal whether the underlying data is accurate.

A broader report with a broader message

The latest ESMA Report on Quality and Use of Data 2025 has a clear message: regulatory reporting data is now embedded in supervisory, policy and systemic risk analysis.

This year’s report has the broadest scope ever, reaching far beyond MiFIR, EMIR and SFTR to encompass regimes such as Prospectus Reporting, Credit Rating Agency Reporting, CCP Supervisory Reporting, Crowdfunding Reporting and DORA Reporting.

ESMA has been more explicit about data usage for policy and risk analysis beyond NCAs and ESMA itself, with the reporting detailing how the ECB uses transaction data for monetary policy, counterparty exposure and asset risk analysis, and how the IMF draws on it for international financial stability assessments.

This greater, broader scrutiny of transaction reporting data is combined with a detailed account of all of the work conducted with the European Commission around transaction reporting simplification and burden reduction and the establishment of data days.

The implication seems to be that reducing the burden on firms comes with great responsibility – if we trust firms with a cap on reporting scrutiny through a paring back of requirements, we expect a renewed push on data quality in exchange.

SFTR data is no longer just a reporting obligation

While the SFTR section contains limited detail and no EU enforcement actions for 2024, ESMA makes it clear that NCAs continue to follow up with reporting entities where there are inconsistencies or unusual patterns in the data. Three points are particularly relevant: 

“The purpose of the SFTR reporting is supervisory and macroprudential oversight to monitor risks related to the shadow-banking and securities finance markets.”

“Reporting may be performed directly by this counterparty or delegated to a third party but without transferring legal responsibility.”

“When SFTR data reveals inconsistencies or unusual patterns, NCAs follow up with the reporting entities concerned to clarify issues or require corrections.”

Why does SFTR data quality matter?

The report also makes clear that SFTR data is now being used well beyond basic supervisory monitoring. ESMA highlights the ECB’s use of SFTR data in risk analysis and policy work, particularly to assess secured funding activity, exposure concentrations, links between banks and non-banks, and vulnerabilities in collateralised funding structures.

This matters because it shows SFTR data has become an important macroprudential data source. ESMA also notes that the IMF uses SFTR data in liquidity stress testing under its Financial Sector Assessment Programme, as well as in analysis of the resilience of LDI funds.

The implication is clear: SFTR data is now part of the infrastructure used to assess market-wide risks, which raises the stakes for firms to ensure the data they report is accurate, complete and reliable.

Pairing and matching only tell part of the story

Despite the critical nature of SFTR data usage, the ESMA data quality focus remains on two-sided reporting and the TR reconciliation process. This does not fully account for the volume of single-sided reports, the proportion of delegated reports (where both sides may pair and match without necessarily being correct) and the prevalence of particular data vendors enriching a large proportion of fields (meaning a matched report sourced from a single vendor offers no independence in validation).

Loan matching was still only between 50-64% with “slight improvements after mid-2025” while collateral matching fell towards 65% by late 2025. These statistics in themselves tell you very little about data quality or the potential bias on aggregated figures used for national, regional and global macro systemic purposes.

As stated previously, two wrongs do not make a right and significant economic errors could slip by unnoticed. This could also potentially create material distortions to aggregated loan values, collateral values or the level of loan collateralisation, all of which could have serious implications for the likes of LDI resilience measures for example.

Rather than relying too heavily on reconciliation outcomes, supervisors need a clearer view of how inaccurate the underlying reports really are. That should include single-sided reports, unmatched reports and even fully matched reports where material errors may be masked by the false comfort of the TR reconciliation process. This would provide a more credible basis for NCA action against firms that misreport.

What should firms do now?

What is clear is that transaction reporting functions are set to become much more technical, intertwined and strategically important. Firms should consider whether their teams have the skills, knowledge and understanding to adapt to the changing regulatory environment. 

• If you would like to discuss the quality of your SFTR reporting, how best to prepare for change or your training plans, please get in touch.