CRR Article 430a Immovable Property Loss Data: The 2025 Hard Test

ByOfficer

On 6 July 2026 the European Banking Authority published its 2025 immovable property loss data: the annual dataset of losses and exposures for residential and commercial property across the EU and EEA. For a standardised-approach credit-risk reporting team, that dataset does real work. It is one of the inputs that decides whether the preferential risk weights you applied to property-secured exposures in your last COREP submission still hold for the year ahead.

The data sits under Article 430a of the Capital Requirements Regulation. The EBA collates it from supervisory reporting that institutions file with their national competent authorities, then presents it market by market so a bank can see the loss experience of every immovable property market it lends into. The 2025 figures arrived as a single file, the “Immovable property loss” dataset for Q4 2025, alongside an interactive view in the EBA data portal.

None of this rewrites the CRR. The EBA said so plainly: the publication does not amend the legal conditions and does not decide whether any individual exposure qualifies for a given treatment. Your institution still owns that call. What the data does is close the loop on the CRR “hard test”, the mechanism that ties certain favourable treatments for property exposures to the loss rates a market actually records. Read the wrong market’s line, or miss that a market slipped below the qualifying condition, and the risk weight in your next return is wrong.

Related reading: our guide to the CRR3 standardised approach for credit risk.

What the EBA released, and what it actually covers

The publication is a single, harmonised source for losses and exposures on lending secured by residential and commercial immovable property, covering the reference year 2025 and drawn from institutions’ own supervisory reporting. The EBA aggregates the submissions and presents them by national immovable property market. The file resolves into a set of market lines, one for each property market across the Union and the wider EEA, and a bank reads the lines for the markets it lends into.

Three things came out together. The dataset itself, delivered as the “Immovable property loss” PDF for Q4 2025. An interactive visualisation in the European Data Access Portal, where the same figures can be filtered by market. And a standing “Immovable property losses” page on the EBA site that carries the series over time. The EBA frames the exercise around transparency, simplification and consistent application of the CRR provisions, rather than any new obligation.

The practical value of the per-market presentation is easy to miss. A bank with a mortgage book concentrated in one country reads one line. A cross-border lender with residential exposures in several markets, and commercial exposures in others, has to read a separate line for each market and for each of the residential and commercial splits. The dataset does not do that mapping for you. It gives you the raw market picture and leaves the exposure-level work where the CRR puts it, inside the reporting institution.

The 2025 immovable property loss data cycle: key dates

Deadline pressure is the reason this topic sits near the top of a reporting team’s July reading. The calendar is short and worth pinning down before anything else.

  • Reference period: full year 2025, presented against a year-end 2025 (Q4 2025) reference point.
  • EBA publication date: 6 July 2026.
  • Reporting cadence: annual. Article 430a requires institutions to report the aggregate loss and exposure data to their competent authorities once a year, and competent authorities publish annually on an aggregated basis.
  • Effect: the market classification you rely on can move once a year, on the back of this cycle.

The annual rhythm is the operational trap here. Because the qualifying condition is tested against fresh loss data each year, a market that supported a favourable treatment in your 2024 reporting is not guaranteed to support it in 2026. The publication is the moment you find out. A team that treats the market classification as static, set once and left alone, will carry a stale risk weight until a validation query or a supervisor asks why.

Article 430a: what institutions actually report

Article 430a is titled “Specific reporting obligations” and it is narrow. It requires institutions to report to their competent authorities, on an annual basis, aggregate data for each national immovable property market to which they are exposed. The article lists six data points, and the structure of those six is what teams need to get right.

For residential property, institutions report losses on exposures where residential property is recognised as collateral, in each case up to the lower of the pledged amount and 55 % of the property value (unless a competent authority decides otherwise under Article 124(9)). They also report overall losses on those exposures, up to the lower of the pledged amount and 100 % of the property value, and the exposure value of all outstanding residential-collateralised exposures on the same 100 % of property value basis. For commercial property the same three points repeat, again measured against 55 % of the property value for the covered-part loss figure and 100 % for the overall-loss and exposure-value figures.

Two of those points look almost identical and are not. The “losses up to the covered part” figure and the “overall losses” figure feed different sides of the eventual loss-rate calculation, and the exposure value gives the denominator. Mapping the wrong loss figure to the wrong field is the kind of quiet error that survives a submission and only surfaces when a competent authority reconciles the aggregate against its own view. If your field mapping is uncertain, our walkthrough of common COREP reporting errors covers the discipline that keeps numerator and denominator aligned.

The reporting geography matters too. Institutions report to the competent authority of their home Member State. Where an institution runs a branch in another Member State, the data for that branch also goes to the host Member State’s competent authority. And the data is reported separately for each immovable property market within the Union, without rolling the markets together. Competent authorities then publish the aggregated figures annually, together with historical data where it is available, which is the series the EBA consolidates into the EU-wide picture.

The hard test: why a published loss line moves your risk weight

The EBA calls the mechanism the “hard test”. Under it, certain preferential treatments for exposures secured by immovable property may be applied only where the relevant loss-rate thresholds are met. The published loss data is the evidence that a market clears, or fails to clear, those thresholds. This is the reason a data release with no new rule attached can still change a capital number.

The treatment at stake is the loan-splitting approach. Under the CRR standardised approach for immovable property, an exposure can be split so that the part covered by the property gets the favourable property risk weight and the remainder is treated as an ordinary exposure to the counterparty. For income-producing real estate, meaning exposures where repayment materially depends on the cash flows generated by the property itself, the CRR applies a more cautious default. The loan-splitting approach becomes available for those income-producing exposures where the market qualifies under the hard test, which is to say where the competent authority has published loss data showing the market’s loss rates stay within the limits the CRR sets.

The EBA is specific about which CRR provisions the dataset supports, and it is worth taking those from the source rather than inferring them. The EBA identifies Article 125(2), which allows the loan-splitting approach for income-producing residential real estate exposures where the conditions are met; Article 126(2), which provides an equivalent treatment for income-producing commercial real estate exposures; and Articles 199(3) and 199(4), which allow institutions using the Internal Ratings-Based approach to recognise residential or commercial immovable property as eligible collateral, including where repayment materially depends on the property’s cash flows. The exact annual loss-rate limits attached to each provision live in the CRR articles themselves, and a reporting team confirming a treatment should read them against the current consolidated CRR text before relying on them.

One misconception is worth killing early. The hard test is a market-level gate, not an exposure-level blessing. A favourable published loss line keeps a treatment available for everyone lending into that market. It says nothing about whether your specific loan meets the CRR’s own conditions on valuation, coverage and the materiality of the property’s cash flows. The EBA was explicit that institutions remain responsible for assessing whether the relevant CRR conditions are met when applying the corresponding treatment.

What changes in your standardised-approach credit-risk reporting

The immovable property exposure class flows straight into the standardised-approach credit-risk templates in COREP. The risk weight you assign to a property-secured exposure depends on whether the exposure is income-producing and on whether the market qualifies for loan-splitting, so the hard test result feeds directly into the risk-weighted exposure amount you report. For the broader picture of where these numbers land, our COREP reporting guide sets out the standardised-approach credit-risk templates and how the exposure classes map into them.

The task the publication creates is a per-market reconciliation. For every national immovable property market where you hold property-secured exposures, confirm that the current published loss data still supports the treatment you are applying. This is the first thing I check when a fresh loss-data release lands: the individual market lines for the markets my book actually touches, because those are the only lines that can move a risk weight I have to file. The EU headline number tells me nothing I can act on. A market that quietly drops out of the favourable treatment is the exposure that turns a routine submission into a restatement.

The consequences run downstream. If a market stops qualifying, income-producing exposures in that market move to the less favourable default, risk-weighted exposure amounts rise, and the own funds requirement rises with them. Under CRR3 that increase can also interact with the output floor, so a change that starts as a single market line can touch the floored total-risk-exposure amount. Our explainer on the CRR3 output floor phase-in shows how standardised risk weights feed that calculation during the transitional period.

There is a second misconception here, and it favours the smaller lender’s comfort. This is an IRB concern and a large-bank concern alike. The loan-splitting treatment sits in the standardised-approach articles, Articles 125 and 126, which is exactly where a standardised-approach bank does its property risk weighting. A bank that never touches an internal model is still fully inside the hard test, and still depends on the published loss line for every income-producing property market it lends into.

The IRB collateral-eligibility angle: Article 199

IRB banks read the same dataset for a different reason. The EBA points to Articles 199(3) and 199(4), which set the conditions under which immovable property collateral may be recognised as eligible under the IRB approach, in particular under the Foundation IRB approach for real estate collateral securing income-producing real estate exposures. The market-level loss evidence supports that recognition in the same way it supports the standardised-approach loan-splitting treatment.

For a Foundation IRB bank, the point is that eligible-collateral recognition is not automatic for income-producing exposures. The derogations in Article 199(3) and (4) carry their own legal conditions, and the published loss data is part of the evidence base that makes the recognition defensible. A bank that recognises property collateral on IPRE exposures without the market-level support is exposed to exactly the kind of finding that IRB approaches attract. The EBA’s annual IRB benchmarking of internal approaches is a reminder of how closely collateral treatment inside internal models is scrutinised.

The two approaches are not interchangeable for reporting. A standardised-approach bank uses the data to decide a risk weight; an IRB bank uses it to decide collateral eligibility that then feeds loss-given-default. Same source, different downstream field, and a team that runs both approaches across a group should not assume one team’s read covers the other’s need.

Four markets where a different source applies

The EBA dataset is not the operative source for every market, and this is the trap most likely to catch a cross-border lender. For four Member States, the relevant reference for applying the derogations remains the loss data published by the national competent authority on its own website. The EBA file does not cover those four markets.

  • Belgium: the National Bank of Belgium (NBB), through its supervisory disclosure requirements for credit institutions.
  • France: the Autorite de controle prudentiel et de resolution (ACPR), through its supervisory disclosures.
  • Croatia: the Croatian National Bank (HNB), through its disclosure on losses from exposures collateralised by immovable property.
  • Romania: the National Bank of Romania (BNR), through its statistical data.

A bank with Belgian or French property exposures that reads only the EBA dataset for those markets is using the wrong source. The consistent, EU-wide file is convenient, and for most markets it is exactly what you want. For these four, the national publication is the reference that applies, and the two should not be substituted for one another. Build the per-market check so that BE, FR, HR and RO route to the national authority automatically, so no one has to remember the carve-out under time pressure.

What the publication does not do

The last thing to be clear about is the boundary the EBA drew around its own release. Treating the dataset as more than it is creates its own risk.

The publication does not amend the legal conditions set out in the CRR. It does not constitute a separate supervisory decision on the eligibility of any individual exposure. And it does not shift responsibility: institutions remain responsible for assessing whether the relevant CRR conditions are met when they apply the corresponding treatment. A favourable market line is a gate that stays open, not a supervisor signing off your loan.

The error to avoid is reading the annual publication as a green light and stopping there. The market qualification is necessary, not sufficient. The exposure still has to meet the CRR’s conditions on valuation, on the coverage line, and on whether repayment materially depends on the property’s cash flows, and those conditions are assessed inside your institution, exposure by exposure. The data tells you the gate is open. It does not walk your exposure through it.

Frequently Asked Questions

Does the EBA immovable property loss data change any CRR rule?

No. The EBA stated the publication does not amend the legal conditions in the CRR and does not constitute a supervisory decision on individual exposures. It is an evidence source for the hard test, and institutions remain responsible for assessing whether the CRR conditions are met.

How often is this data published, and what period does the 2025 release cover?

The cadence is annual. Article 430a requires institutions to report the aggregate data to their competent authorities once a year, and competent authorities publish annually on an aggregated basis. The release dated 6 July 2026 covers reference year 2025, presented against a year-end 2025 reference point.

My bank uses the standardised approach only. Does this affect me?

Yes. The loan-splitting treatment the data supports sits in Articles 125 and 126, which are standardised-approach articles. A standardised-approach lender depends on the published loss line for every income-producing property market it lends into, without ever touching an internal model.

We have property exposures in Belgium and France. Do we use the EBA file?

No. For Belgium, France, Croatia and Romania the relevant reference remains the national competent authority’s own publication: the NBB, the ACPR, the HNB and the BNR respectively. Use the national source for those four markets and the EBA dataset for the rest.

What is the difference between “losses up to the covered part” and “overall losses” in the dataset?

Article 430a reports both. The “up to the covered part” figure caps losses at the collateralised portion, up to 55 % of the property value for both residential and commercial. The “overall losses” figure runs up to 100 % of the property value on those exposures. The two feed different sides of the loss-rate view, and the reported exposure value gives the denominator.

Does a favourable published loss line mean our individual loan qualifies for the treatment?

No. The hard test is a market-level gate. A qualifying market keeps the treatment available for everyone lending there. Your specific exposure still has to meet the CRR’s own conditions on valuation, coverage and the materiality of the property’s cash flows.

Where does this show up in COREP?

In the standardised-approach credit-risk templates. The risk weight assigned to a property-secured exposure depends on whether it is income-producing and whether the market qualifies for loan-splitting, so a change in market qualification changes the risk-weighted exposure amount you report and, in turn, the own funds requirement.

What happens if a market’s loss rate breaches the CRR limit?

The favourable treatment ceases to be available for that market until the qualifying condition is met again in a later year. Income-producing exposures in that market move to the less favourable default, which raises risk weights and can flow through to the CRR3 output floor calculation.

Related Articles

Key Takeaways

  • On 6 July 2026 the EBA published its 2025 immovable property loss data under Article 430a of the CRR, covering residential and commercial property across the EU and EEA, presented by national market.
  • The data is the evidence base for the CRR “hard test”: certain preferential treatments for property exposures apply only where a market’s published loss rates stay within the CRR limits.
  • Article 430a requires institutions to report annually to their competent authorities six aggregate data points per national market, covering losses up to the covered part, overall losses, and exposure value, for residential and commercial property separately.
  • The EBA identifies the provisions the data supports as Article 125(2) and Article 126(2) loan-splitting for income-producing residential and commercial real estate, and Articles 199(3) and 199(4) for IRB collateral eligibility.
  • The hard test works at market level. A qualifying market keeps a treatment available, but each exposure must still meet the CRR’s own conditions, and institutions remain responsible for that assessment.
  • For Belgium, France, Croatia and Romania the operative source is the national competent authority’s own publication: the NBB, ACPR, HNB or BNR.
  • A market that stops qualifying moves income-producing exposures to a less favourable treatment, raising standardised risk weights and potentially the CRR3 output floor.
  • Run a per-market reconciliation after each annual release: read the market lines for the markets your book touches, since only those can move a risk weight you file.

Sources and References

  • European Banking Authority, press release, “The EBA publishes 2025 loss data for immovable property markets under Article 430a of the Capital Requirements Regulation”, 6 July 2026: eba.europa.eu
  • European Banking Authority, Interactive Single Rulebook, Article 430a (Specific reporting obligations): eba.europa.eu single rulebook
  • Regulation (EU) No 575/2013 (Capital Requirements Regulation), consolidated text, Articles 124, 125, 126, 199 and 430a: eur-lex.europa.eu
  • Regulation (EU) 2024/1623 (CRR3) amending Regulation (EU) No 575/2013, including the revised standardised approach for exposures secured by immovable property: eur-lex.europa.eu
  • National competent authority publications for the four carve-out markets: National Bank of Belgium (NBB), Autorite de controle prudentiel et de resolution (ACPR), Croatian National Bank (HNB) and National Bank of Romania (BNR), each on its own website as referenced by the EBA press release.

Reading the 2025 loss lines before your next return

The EBA release is a small file with a large reach. It does not change a word of the CRR, and it does not sign off a single exposure, yet the market lines inside it decide whether the property risk weights already sitting in your COREP submission survive the year. The work it creates is not glamorous: read the lines for the markets you actually lend into, route Belgium, France, Croatia and Romania to their national sources, and confirm that every income-producing property exposure still stands on a market that clears the hard test. Do that once a year, on the back of this cycle, and the annual publication becomes a checkpoint rather than a restatement waiting to happen.

Last updated: July 2026

Disclaimer: The information on RegReportingDesk.com is for educational and informational purposes only. It does not constitute legal, regulatory, tax, or compliance advice. Always consult your compliance officer, legal counsel, or the relevant supervisory authority for guidance specific to your institution.

Similar Posts