KGFS Reporting in Denmark: New Home Loans to High-Debt Borrowers

Finanstilsynet’s KGFS return, “Boligudlån for kunder med høj gældsfaktor” (home loans to customers with a high debt-to-income factor), requires in-scope Danish banks, branches of foreign banks and mortgage banks to report quarterly new lending secured on owner-occupied homes and holiday homes. The return includes a total new-lending measure and separately identifies specified lending to customers with a gældsfaktor above four and an LTV above 60 per cent. KGFS reporting has run on instructions dated 15 September 2025 from the period Finanstilsynet labels 202543, with two templates: GFNPI for banks’ mortgage-like loans and GFNRR for mortgage loans.

The return is national. It sits outside the EBA reporting framework, travels as an XML file through FIONA Online, the reporting system Finanstilsynet shares with Danmarks Nationalbank, and takes its risk-profile test from the executive order on good practice for home credit (god skik for boligkredit), its gældsfaktor definition from the guidance to the KRGS accounting return, and its property categories from the executive order on the delimitation of property categories. The KGFS instructions point at annex 7 of the KRGS guidance, which is not where that definition sits in the 10 April 2026 version, and the god-skik rule behind the risk profile has moved from § 21 to § 19 across two orders, so a team that maps loans from the KGFS instructions and the god-skik guidance alone can cite a stale annex and a superseded section number in its own documentation.

Related reading: KOFS Reporting in Denmark: Own Funds Returns and Deadlines

The legal basis of KGFS: the god-skik order and Finanstilsynet’s instructions

KGFS has no executive order of its own. The obligation to file, the scope threshold, the templates and the frequency are set out on Finanstilsynet’s KGFS page and in the instructions (anvisning) published there. Neither the page nor the instructions cites the statutory provision under which Finanstilsynet collects the return.

What the return measures comes from the rules on good practice for home credit. The current text is executive order (bekendtgørelse) no. 1443 of 27 November 2025 on good practice for home credit, issued by the Ministry of Industry, Business and Financial Affairs under, among other provisions, § 43, stk. 2, 3 and 7, and § 373, stk. 4 of the Danish Financial Business Act (lov om finansiel virksomhed), and in force from 1 January 2026. Its § 19, stk. 5 provides that a home-credit provider may not, as a starting point, enter into risky agreements on mortgage loans or mortgage-like loans with a high loan-to-value with consumers if the consumer thereby gets a high gældsfaktor. § 19, stk. 6 disapplies stk. 1 to 5 for agreements with a principal of DKK 100,000 or less, unless several loans secured on the same property together exceed DKK 100,000.

The numbers sit in Finanstilsynet’s guidance, VEJ no. 9458 of 1 June 2023. The guidance treats a gældsfaktor above four as high as a starting point, meaning total debt greater than four times annual income before tax. It names two loan types that, above 60 per cent of the property value, can as a starting point be considered risky loans with a high loan-to-value: all variable-rate loans, with or without repayment, where the rate is reset at intervals of less than five years; and interest-only variable-rate loans where the rate is reset at intervals of five years or more. The guidance also records Finanstilsynet’s expectation that a decision to grant such a loan to a high-gældsfaktor consumer is taken at a higher organisational level than usual.

That guidance was written for executive order no. 944 of 21 June 2022 and explains the rule as § 21, stk. 5. Order no. 944 was repealed by executive order no. 1514 of 6 December 2023, in force from 30 December 2023, which kept the rule in § 21. Order no. 1514 was in turn repealed by order no. 1443 of 2025, where the same wording sits in § 19. Retsinformation lists both later orders under the guidance’s subsequent changes, but the guidance text refers to § 21 throughout. A credit policy or loan-system flag labelled “§ 21, stk. 5” therefore points at the right rule under a superseded number.

One reading error is worth heading off here. § 19, stk. 5 is a starting-point rule, and the guidance says in terms that it does not prevent risky loans to consumers who thereby get a high gældsfaktor where lending is justifiable in the individual case, for example where the consumer has high positive net wealth. A non-zero amount in the high-gældsfaktor columns of KGFS is a reported fact about lending volumes. It carries no finding of a breach.

Who must report KGFS: a DKK 200 million stock test for a flow report

Finanstilsynet’s KGFS page sets the threshold separately for each institution type:

• Danish banks (pengeinstitutter) and branches of foreign credit institutions file KGFS if their total lending secured on owner-occupied homes and holiday homes, including intermediated non-group mortgage loans, exceeds DKK 200 million.
• Mortgage banks (realkreditinstitutter) file if their total lending secured on owner-occupied homes and holiday homes, excluding intermediated non-group mortgage loans, exceeds DKK 200 million.

The test runs on the stock of lending. The report then captures a flow: loans disbursed during the quarter. A bank with a large, old owner-occupied book and very little new lending is still in scope, and its quarterly figures can be close to zero.

The intermediated-loan line is where banks and mortgage banks part company. For a bank, mortgage loans it has intermediated for a mortgage bank outside its own group count toward the DKK 200 million; for a mortgage bank, the same category is left out. The page does not say on which date the threshold is measured, how often it is re-tested, or whether an institution that falls back below DKK 200 million stops filing. Those points are worth settling in writing with the Finanstilsynet contact listed on the KGFS page next to the FIONA servicedesk.

Template allocation follows the institution type. The GFNPI instructions address the bank (“Pengeinstituttet bedes indberette”), while GFNRR is addressed to the mortgage bank and “relevante pengeinstitutter”, relevant banks. The instructions do not define which banks are relevant. The most natural reading is that the phrase captures banks whose threshold includes intermediated non-group mortgage loans, because GFNRR is the template that holds those loans and GFNPI expressly leaves them out.

Three further boundaries come from the instructions. Both templates are filed at solo level, and there is no consolidated KGFS, unlike the paired KRGS and KRGC accounting returns. Only the private-customer segment is reported, meaning natural persons acting outside their trade: a self-employed person borrowing to buy a home, including a holiday home, counts as a private customer; the same person borrowing for production premises does not; and a property mainly used as a private home with a small business part stays private. And only loans secured on property in the categories “ejerboliger og fritidshuse” are in, following the executive order on the delimitation of property categories.

Cooperative housing sits outside on two counts. The god-skik guidance states that the gældsfaktor rule does not cover andelsbolig loans, because a cooperative share cannot be financed as real property, and that loans to housing cooperatives fall outside the rule as well. Nothing in the KGFS instructions brings either category back in.

What the report contains: GFNPI, GFNRR and four columns by region

The instructions state the purpose in one line: KGFS monitors credit institutions’ supply of mortgage-like loans and mortgage loans to customers with a high gældsfaktor and a loan-to-value above 60 per cent of the property value. Two templates carry it.

• GFNPI covers mortgage-like loans (realkreditlignende lån) granted by the bank, as new lending broken down by region and gældsfaktor. Intermediated non-group mortgage loans are excluded, and loans for project homes enter as new lending only once disbursed.
• GFNRR covers mortgage loans and intermediated non-group mortgage loans, as new lending broken down by region and gældsfaktor.

The KGFS instructions use “realkreditlignende lån” without defining it. The definition sits in § 2, nr. 13 of executive order no. 1443 of 2025: a loan with an agreed term of more than 10 years and a principal of at least DKK 100,000 at origination, secured on an owner-occupied home, a holiday home or qualifying agricultural property in Denmark, within the lending limits of §§ 5 and 7 of the Danish Act on mortgage loans and mortgage bonds (lov om realkreditlån og realkreditobligationer), and, for a bank loan, capable of serving as collateral for særligt dækkede obligationer, the Danish covered bond type.

Rows, columns and XML codes

Both templates share one grid. The rows are the eleven Danish regions (landsdele) plus a total, and the location of the pledged property decides the region, using Danmarks Statistik’s classification. The published KGFS schema assigns technical identifiers to the regional rows. The exact row codes, XML element names and other technical identifiers should be taken directly from the applicable 202543 schema, XSD and XML example when implementing the file. The regions are:

• København by, Københavns omegn, Nordsjælland, Bornholm
• Østsjælland, Vest- og Sydsjælland
• Fyn, Sydjylland
• Østjylland, Vestjylland
• Nordjylland, and a total row

The four columns carry the analysis:

The instructions’ worked examples treat c100 as the total new-lending column and the specialised columns as additional qualifying amounts. For example, a fixed-rate amortising top-up of DKK 0.3 million is reported in c100 only, while qualifying loans are reported in c100 and the applicable specialised column. The validation rule described further down, which only requires c100 to be at least the sum of the other three columns, is consistent with that layout. A dashboard that labels c100 “high-gældsfaktor lending” overstates the subset by the full volume of ordinary lending.

How the instructions count new lending

New lending (nyudlån) means new loans, extensions of existing loans, and extensions made on refinancing or conversion that are not caused by refinancing or conversion costs, in each case paid out by the institution in the period. A loan disbursed and redeemed within the same quarter still counts.

Refinancing is where the rules get specific, and they key off the risk profile under the god-skik order:

• A proceeds-neutral refinancing or conversion is new lending when the loan’s profile moves from non-risky to risky. For moves from risky to risky, from non-risky to non-risky and from risky to non-risky it is as a starting point not reported, except that a move from non-risky to a fixed-rate loan without amortisation, with LTV above 60 per cent and gældsfaktor above 4, is also reported as fixed-rate new lending. A profile change on refinancing without a new loan agreement is ignored.
• An extension made on refinancing goes into the total column as a starting point when the profile moves from non-risky to non-risky or from risky to non-risky. Where the move is from non-risky to a fixed-rate loan without amortisation, with LTV above 60 per cent and gældsfaktor above 4, the extension is also reported as fixed-rate new lending.
• An extension where the profile moves from risky to risky goes into both the total and the variable-rate columns, and so do extensions and proceeds-neutral refinancings that move the profile from non-risky to risky.
• Exposures transferred from another credit institution count as new lending unless the profile is treated as unchanged (risky to risky, risky to non-risky, or non-risky to non-risky) and the loan is not extended beyond set-up costs. A transfer from non-risky to a fixed-rate loan without amortisation, above 60 per cent LTV and with gældsfaktor above 4, counts.

The worked examples in the instructions turn these rules into amounts. Three from the GFNPI set are worth copying into a test pack:

• A customer with gældsfaktor 5 extends an existing variable-rate loan (fixation under five years, no amortisation) from DKK 1.0 million to DKK 1.2 million at a final LTV of 80 per cent. Only the DKK 0.2 million increase is reported, in c100 and c110.
• The same type of customer takes a fixed-rate top-up loan with amortisation of DKK 0.3 million on top of an existing loan, final LTV 80 per cent. It goes into c100 only.
• A customer moves a fixed-rate loan with amortisation of DKK 1 million from one bank to another, where it becomes a variable-rate loan with fixation under five years and with amortisation, at 80 per cent LTV and without extension. The receiving institution reports DKK 1 million in c100 and in c110.

Measuring gældsfaktor, LTV and amortisation

KGFS defines gældsfaktor as the customer’s total debt divided by the household’s gross income, with a reference to the definition in “bilag 7” of the guidance to the accounting return for credit institutions and investment firms, the KRGS guidance. The instructions then spell out the inputs. Gross income can include wages, unemployment benefits, pensions, student grants (SU), housing benefit, child benefits and child and spousal support, plus dividends, bonus, interest and rental income to the extent they are reasonably certain or stable over several years, taking the lowest stable level where rental income fluctuates. Own and employer contributions to employer-administered pension schemes are left out; released retirement savings can be spread over a number of years.

Debt is gross. Free funds cannot be netted, credit lines count at their maximum, family debt counts even when it is not being repaid, and SU loans count. Mortgage loans are taken at market price (kursværdi) and bank loans at their size. For this reporting purpose only, debt that relates solely to a self-employed person’s business may be deducted.

Loan-to-value includes prior-ranking loans plus mortgage-like loans (at nominal value) and mortgage loans (at market value), relative to the property value at the latest time of granting, with the value set under the valuation order. A loan is “with amortisation” if its instalments correspond to repayment over a maximum of 30 years; it is “without amortisation” if instalments fall short of that or if the loan carries an interest-only option at the time of granting. In GFNPI the same principles apply to overdraft facilities.

Interest-rate caps get their own test. A cap that runs for at least five years at a sufficiently low level makes a loan equivalent to fixed-rate; a shorter cap or a higher cap level leaves it variable. Sufficiently low means the six-month average of the average long yield on Danish-krone mortgage bonds, per Realkreditrådet’s bond yield statistics, plus one percentage point, which the instructions word as “opregnet til nærmeste hele procenttal” (raised to the nearest whole per cent; the wording does not state the rounding direction more precisely). A capped loan without amortisation counts as fixed-rate only if the cap runs for the whole term, and capped loans treated as fixed-rate go into c130 when they lack amortisation and exceed the LTV and gældsfaktor limits.

Units, values and exclusions

Amounts are reported in DKK thousands. Loans in all currencies are included, converted to DKK, and guarantees and similar items are excluded. Loans taken over under joint funding (fælles funding), where a loan and its mortgage deed pass to another bank or mortgage bank that issues covered bonds on it, are not reported.

The two templates value differently. GFNPI reports gross outstanding, meaning the maximum amount granted including undrawn amounts, at nominal value, while GFNRR reports outstanding amounts at market value. Two GFNPI examples settle the gross question: a DKK 1 million loan of which DKK 0.7 million is drawn reports DKK 1 million, and a DKK 1 million priority loan with a linked deposit of DKK 0.4 million also reports DKK 1 million.

KGFS reporting deadlines and reference dates on the fixed Danish calendar

KGFS is quarterly. Reference dates are the quarter-ends, and each filing reports the new lending paid out during that quarter. Finanstilsynet’s deadline list for electronic reporting (indberetningsfrister) places KGFS among the national returns submitted through FIONA Online, with four fixed deadlines: 11 February, 12 May, 11 August and 11 November. Paired with the quarter-ends in calendar order, they run as follows:

• Quarter ending 31 March: due 12 May
• Quarter ending 30 June: due 11 August
• Quarter ending 30 September: due 11 November
• Quarter ending 31 December: due 11 February of the following year

The list gives the same four dates for banks, mortgage banks and branches of foreign credit institutions. Not every national FIONA return works this way. In the same list, several quarterly national returns for banks and mortgage banks run on working-day counts instead, for example KDOLS, KRDS, KSOS and KSUS at 20 working days after quarter-end (30 working days for the fourth quarter). A reporting calendar that applies one working-day rule to all FIONA returns will compute the wrong KGFS date.

The current file set applies from the period the KGFS page labels 202543, and the page does not decode the label. The deadline list uses the same style, referring to an ESG period “202522” filed in FIONA before that return’s first e-Reg reference date of 30 June 2026, which fits a format of year, frequency digit and period number. On that reading, 202543 is the quarter ending 30 September 2025, due 11 November 2025. The next applicable reference date is always the next quarter-end, filed by the next date in the cycle above.

The deadline list does not say what happens when a deadline falls on a weekend or a public holiday, and the KGFS documents are silent on the point too.

Submission through FIONA Online: the KGFS XML file

KGFS goes through FIONA Online, the web-based reporting system run jointly by Danmarks Nationalbank and Finanstilsynet. FIONA Online offers three input routes: entering data directly, copying it from an Excel sheet, or uploading a reporting file in the format defined for the report type. Access requires a user set up for the reporting institution and a Microsoft Azure AD licence, with sign-in confirmed by the Microsoft Authenticator app or an SMS code. Technical support sits with the FIONA servicedesk at Danmarks Nationalbank (fiona@nationalbanken.dk).

The channel split matters for teams that also file EU returns. Finanstilsynet’s deadline list puts EBA returns such as COREP own funds, now the COREPOF module, and FINREP on e-Reg from the 31 March 2026 reference period, and keeps KGFS in the separate block of national returns submitted through FIONA Online. Our COREP reporting guide covers the EU side of that calendar; KGFS uses no EBA taxonomy at all.

The file is submitted in XML format. Finanstilsynet publishes an XSD and an XML example for the 202543 file set on the KGFS reporting page. The exact element names, attributes, cardinalities and validation constraints should be taken directly from the applicable XSD and XML example when implementing the file.

Several submission points are left open. Neither the KGFS page nor the instructions specify a file-naming convention, and none of the KGFS documents describes a resubmission procedure. FIONA Online lets a filer see the error lists for a report and correct errors, which is the closest the published material comes. The KGFS page also keeps the earlier schema, XML sample, XSD and instructions for periods 201744 to 202542; which file set a correction to one of those older quarters should use is not stated.

Validation rules in the KGFS error checks and schema

Finanstilsynet publishes a separate list of error checks (kontroloversigt) with the 202543 KGFS file set. It holds ten checks. Eight are total checks, one for each of the four columns in each template: the total row must equal the sum of the eleven regional rows, for SBNy, VNy, VNyUA and FNyUA in GFNPI (GFNPI_sum_1 to GFNPI_sum_3 and KGFS_GFNPI_sum_4, as the IDs are written in the control list) and for RSRNy, RVNy, RVNyUA and RFNyUA in GFNRR (KGFS_GFNRR_sum_5 to KGFS_GFNRR_sum_8). Two are column checks: KGFS_GFNPI_sum_9 requires SBNy to be greater than or equal to VNy + VNyUA + FNyUA, and KGFS_GFNRR_sum_10 requires RSRNy to be greater than or equal to RVNy + RVNyUA + RFNyUA.

The instructions’ worked examples show the logic behind the column checks: c100 carries total new lending and c110 to c130 carry the qualifying subsets, so the subsets cannot exceed the total.

None of the KGFS documents reports rejection statistics, so the useful exercise is to ask which build errors these rules would catch. A loan-level extract that applies a different date or segment filter to c100 than to c110 to c130 can push the high-gældsfaktor columns above the total and fail the column check. A regional mapping that drops loans with a missing postcode leaves the regional rows short of a total computed separately, and fails the totals check for that column.

What the control list cannot see matters as much. Nothing in it tests whether a borrower’s gældsfaktor or LTV was computed correctly, whether a refinancing was classified under the right profile rule, or whether KGFS agrees with any other return. A file can pass every check and still misstate c110.

Caveats and interactions: where KGFS departs from its neighbours

The KGFS columns and the god-skik risk list overlap without matching. Columns c110 and c120 track the guidance’s two risky categories closely: variable-rate loans with fixation under five years, and interest-only variable-rate loans with fixation of five years or more, both above 60 per cent LTV. Column c130 has no counterpart in that list. It captures fixed-rate lending without amortisation above 60 per cent LTV to high-gældsfaktor borrowers, a loan type the guidance does not name as risky. A loan sitting in c130 is therefore no evidence of a § 19, stk. 5 case on the strength of the KGFS column alone.

The gældsfaktor itself is defined in three slightly different ways:

• KGFS: the customer’s total debt over household gross income, with business-only debt deductible for this reporting purpose only.
• The god-skik guidance: total debt over annual income before tax, where debt relating only to a self-employed person’s business is left out, but debt from a parental purchase (forældrekøb) is counted together with the related rental income, even though it is treated as business debt for tax purposes.
• The KRGS guidance, in its annex on banks’ customer classification: the household’s total debt over household gross income.

A gældsfaktor stored once in a credit file and reused for all three purposes can drift from the KGFS definition on exactly these points: whose debt sits in the numerator, and how business and parental-purchase debt are treated.

The KGFS pointer to “bilag 7” of the KRGS guidance does not match the current numbering. In the KRGS guidance dated 10 April 2026, Bilag 7 is “Realkreditudlån fordelt efter ejendomskategori og efter lånetype” (mortgage lending by property category and loan type), and the gældsfaktor description sits in Bilag 5, “Pengeinstitutters klassifikation af kunder”. The KGFS instructions spell out their own operative definition, so the stale pointer matters mainly when a reviewer or auditor follows it.

Proportionality is thin. The DKK 200 million threshold is the only lever, and the KGFS documents contain no waiver, national discretion or transitional relief.

Small loans are a quieter boundary. The § 19, stk. 6 exemption for principals of DKK 100,000 or less is not repeated in the KGFS instructions. If the god-skik definition of a mortgage-like loan is applied, with its minimum principal of DKK 100,000, small bank loans fall outside GFNPI by definition. GFNRR has no equivalent floor in its text, so the instructions do not exclude a small mortgage loan that meets the column conditions, even though § 19, stk. 5 does not apply to it. The instructions do not address the point.

KGFS also sits beside other national FIONA returns. Banks and mortgage banks file the KRGS accounting return on the same four dates (group 4 institutions have a 25 February deadline for the annual report), mortgage banks file KOAS (mortgage lending with outstanding debt at maturity) on the same four dates, and KLTS (Loan-to-Value) runs annually to 11 February with a 20-working-day half-year deadline; where one loan dataset feeds several returns, consistent figures are worth checking internally. For a comparable national dataset on property lending, see our guide to ESPREP-RES real estate lending reporting in Luxembourg; for loan-level credit data at a very different granularity, see AnaCredit reporting in Germany.

Changes to KGFS: the 202543 set, a new god-skik order and an unchanged channel

The dated changes that touch KGFS are few:

• 30 December 2023 and 1 January 2026: the god-skik order was replaced twice (order no. 1514 of 2023, then order no. 1443 of 2025), and the gældsfaktor rule moved from § 21 to § 19.
• 15 September 2025: the date on the new KGFS instructions and on the KGFS page’s last-updated line; the page lists instructions, a schema and an error list applicable from period 202543. The tracked-changes version of the instructions strikes out the whole BOBPI template, which had collected banks’ gross outstanding lending secured on homes where the loan serves as primary home financing as a direct alternative to mortgage loans and mortgage-like loans. A comparison of the earlier instructions linked for periods 201744 to 202542 with the 202543 text shows no other substantive change to GFNPI or GFNRR.
• 31 March 2026 reference period: the monthly and quarterly EBA returns for banks and mortgage banks, including COREPOF and FINREP, moved from FIONA to e-Reg, while other EU returns start on e-Reg from their own reference dates, such as 30 June 2026 for ESG. KGFS, as a national return, stayed on FIONA Online.

Neither the KGFS page nor the instructions announce a further change. The last one arrived as a new period block on the KGFS page, with tracked-changes versions of the instructions and schema published alongside clean ones, and that page is the place to watch.

Frequently Asked Questions

A borrower’s gældsfaktor is exactly 4.0, or the LTV is exactly 60 per cent. Does the loan go into c110 to c130?

No. The KGFS schema’s column headers use strict inequalities, “GF > 4” and “LTV > 60”, and the instructions state the same thresholds in words (“gældsfaktor større end 4”, “LTV højere end 60 pct.”). A loan at exactly 4.0 or exactly 60 per cent stays in c100 only. The god-skik guidance draws its high-gældsfaktor line in the same place, at more than four.

A loan is approved in one quarter but paid out in the next. Which quarter reports it?

The quarter of payment. New lending is defined by what the institution “har udbetalt i perioden”, has paid out in the period, and for GFNPI the instructions add that project-home loans enter only when disbursed. The approval date plays no part in the definition.

The customer has an interest-only option but repays from day one. Is the loan “without amortisation”?

Yes, for KGFS purposes. The instructions treat a loan as without amortisation if it has the option of interest-only periods at the time of granting, whatever the customer then does. The god-skik guidance takes the same line for the gældsfaktor rule: a loan with an interest-only option counts as interest-only whether or not the option is used when the agreement is made.

The latest income data on file is two years old. Does the gældsfaktor need refreshing before reporting?

The instructions base the gældsfaktor on the most recently obtained income information (“de seneste indhentede indkomstoplysninger”) and set no maximum age for it. The KGFS value therefore follows the latest income figures the institution has obtained, whenever those were collected.

The customer owns two homes for a few months while selling the old one. Whose debt goes into the gældsfaktor?

The instructions measure debt on the future situation, after the previous home is sold. If the old home is not yet sold and its proceeds are meant to reduce the new debt, the institution applies a prudent estimate of those proceeds.

Does an in-scope bank file anything for a quarter with no new lending in the segment?

The KGFS documents do not address nil returns. Whether Finanstilsynet expects zero values or a file without template blocks is a question for the FIONA servicedesk or the KGFS contact.

Related Articles

• KOFS Reporting in Denmark: Own Funds Returns and Deadlines: how Danish COREP own-funds reporting moved from the FIONA KOFS and KOFC names to the e-Reg COREPOF module.
• ESPREP-RES Reporting in Luxembourg: CSSF Real Estate Lending Data: who files the CSSF residential and commercial real estate lending returns and what the templates hold.
• AnaCredit Reporting in Germany: Bundesbank Credit Data and Deadlines: loan-level credit reporting to the Bundesbank and the debtor threshold that triggers it.
• COREP Reporting Explained: A Practical Guide to Prudential Reporting: the EU prudential templates that Danish institutions file through e-Reg.
• FINREP Reporting Explained: the EBA financial reporting framework and how its templates are organised.

Key Takeaways

• Run the DKK 200 million test on the owner-occupied and holiday-home lending stock, adding intermediated non-group mortgage loans for a bank or foreign branch and leaving them out for a mortgage bank.
• Build GFNPI at gross granted amounts and nominal value, and GFNRR at market value, both in DKK thousands, both solo, both private customers only.
• Key the reporting calendar to 12 May, 11 August, 11 November and 11 February, independent of any working-day rule used for other FIONA returns.
• Before upload, run the validations specified in Finanstilsynet’s applicable KGFS control list, including the published consistency checks for the regional totals and reporting columns.
• Cite § 19 of order no. 1443 of 2025 for the risk-profile rule and Bilag 5 of the April 2026 KRGS guidance for the classification definition of gældsfaktor.
• Remove BOBPI from the build: it left KGFS with the 202543 set.

Sources and References

• Retsinformation: VEJ nr. 9458 af 01/06/2023, Vejledning til bekendtgørelse nr. 944 af 21. juni 2022 om god skik for boligkredit
• Retsinformation: Bekendtgørelse nr. 1443 af 27. november 2025 om god skik for boligkredit
• Retsinformation: Bekendtgørelse nr. 1514 af 6. december 2023 om god skik for boligkredit
• Finanstilsynet: KGFS, Boligudlån for kunder med høj gældsfaktor (reporting page)
• Finanstilsynet: KGFS anvisning, 15 September 2025, clean version (PDF) and tracked-changes version (PDF)
• Finanstilsynet: Anvisninger til indberetning af KGFS, earlier version for periods 201744 to 202542 (PDF)
• Finanstilsynet: KGFS schema 202543 (Excel)
• Finanstilsynet: KGFS kontroloversigt, list of error checks (Excel)
• Finanstilsynet: KGFS XSD 202543 and KGFS XML sample 202543
• Finanstilsynet: Indberetningsfrister for elektronisk indberetning, version dated 10 August 2026 (PDF)
• Finanstilsynet: Hvad er FIONA Online? and Indberetninger til FIONA Online (national returns by institution type)
• Finanstilsynet: Vejledning til regnskabsindberetning for kreditinstitutter og fondsmæglerselskaber m.fl., 10 April 2026 (PDF)

Building each KGFS quarter-end file

Each KGFS cycle starts at a quarter-end and closes on the next of 12 May, 11 August, 11 November or 11 February. The file is two grids of eleven regions and a total, in DKK thousands; its accuracy is decided upstream, in how each disbursement, extension and transfer is classified against the KGFS new-lending rules and the § 19 risk profile. Run the regional totals and the c100 column check on the loan-level extract before the XML is generated, and settle the threshold-timing and nil-return questions with Finanstilsynet in writing before the first quarter they affect.

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.