Lending rules in light of higher interest rates

Danish banks and mortgage credit institutions are subject to certain rules when granting mortgages to households, so-called borrower-based measures. The rules were implemented during a period of accommodative monetary policy and interest rates that were significantly lower than today. The purpose of the rules is to protect homeowners from risks such as rising interest rates or falling property prices, and to ensure that lenders do not loosen credit standards.¹

The rules mainly restrict the most highly indebted borrowers. The rules include limits on e.g., the loan-to-value ratio, the borrower’s debt relative to income, and the access to interest-only mortgages and variable-rate mortgages. Table 1 provides an overview of the lending rules implemented since 2013. The down payment should be at least 5 per cent and interest-only mortgages with variable rate are not permitted when the debt-to-income and loan-to-value ratios are high. Furthermore, interest-only mortgages as well as variable-rate mortgages should only be granted to borrowers who are able to pay both interest and principal on an equivalent 30-year fixed-rate repayment mortgage. The rules outlined in the so-called Growth Area Guidelines apply solely to Copenhagen and Aarhus. These rules include a so-called interest rate stress where borrowers applying for variable interest rate are assessed based on a higher rate than the rest of the country.

The rules restrict borrowers in various ways. The down payment requirement, the interest stress, and the rule regarding the ability to service a 30-year fixed-rate repayment mortgage are “hard” rules in the sense that the loan amount should not exceed the specified limits. In contrast, the debt-to-income ratio rule in the Executive Order on Good Practice for Mortgage Lending is a “soft” rule in the sense that it only prohibits certain loan types if both the loan-to-value and debt-to-income ratios exceed the specified thresholds.

The rules are implemented in consumer protection legislation (Good Practice for Mortgage Lending) and in the Executive Order on governance (the Growth Area Guidelines). Thus, the Danish lending rules do not have a legal basis in legislation specifically addressing risks to the overall financial system. In practice, lending rules with a consumer protection purpose can have the same impact on financial stability as lending rules implemented with a financial stability purpose. However, there may be instances however where this is not the case. For example, if financial stability considerations would warrant measures that are stricter than what can be justified by consumer protection, such measures could not be implemented in the current legal framework. The implementation of the lending rules in the consumer protection legislation also means that institutions have some flexibility to deviate from the measures if their credit assessments warrant it; cf. section 4.

Authorities in other countries have implemented lending rules that are similar to the Danish rules. This analysis focuses solely on the effect of higher interest rates on the Danish lending rules, but the conclusions are generally in line with experiences in other countries; cf. ECB (2023). Some of the Danish lending rules address risks related to the low interest rate environment. Low nominal interest rates – or loose monetary policy in general – can have unintended consequences for financial stability, for instance because of increased risk-taking; cf. De Nicolò et al. (2010) and Lian et al. (2019). Lending rules may dampen the transmission of monetary policy by reducing some households’ ability to increase indebtedness and thus to take full advantage of the low interest rates. However, monetary policy and lending rules can also complement each other as lending rules make the financial system more stable; cf. Laeven et al. (2022). Hence, lending rules can increase the transmission of monetary policy, as monetary policy operates through banks and the financial system more generally. Thus, lending rules contribute to making the financial system more stable. In what follows, the analysis will focus on the lending rules and how they limit borrowers when interest rates change.

The interest rate level matters for the relative bindingness of the lending rules

The higher interest rates since 2022 have impacted how and when the lending rules restrict borrowers; cf. chart 1. The chart illustrates the maximum amount a typical homebuyer can borrow until the rules become binding. The typical homebuyer is defined as a family consisting of two adults and two children, whose income and down payment corresponds to the median for the group (see box 1 for a description of assumptions). To illustrate how interest rate developments affect the bindingness of the rules, we keep incomes and down payments constant. Hence, only interest rates vary over time.

At low nominal interest rate levels, as in 2020, the first binding lending rule was the rule regarding the debt-to-income in the Executive Order on Good Practice for Mortgage Lending. Thus, the typical homebuyer could borrow kr. 3.6 million without being restricted by any rule. At a larger loan amount, the debt-to-income ratio would exceed 4, at which point the debt-to-income rules apply.²

At higher interest rates, as in 2023, the first binding rule is instead the interest rate stress from the Growth Area Guidelines. The interest rate stress applies to homebuyers in Copenhagen and Aarhus opting for a variable-rate mortgage. In the rest of the country – or for homebuyers not opting for a variable-rate mortgage – it is the rule requiring the homebuyer to be able to service a 30-year fixed-rate repayment mortgage that is the first binding rule with interest rates as in 2023. In general, the rule requiring the homebuyer to be able to service a 30-year fixed-rate repayment mortgage constitutes an upper limit for the loan amount that credit institutions can offer.³

The change in the relative bindingness of the rules primarily stems from the fact that the typical homebuyer can service a lower debt when interest rates increase – and thus, also a lower debt according to the credit assessment in the interest rate stress. Conversely, the loan amount under the debt-to-income rules is not directly affected by interest rate developments; here, only pre-tax income matters.⁴

The interest rate level thus affects which rule will be the first constraint. The debt-to-income rules are often the first constraint when interest rates and borrowing costs are relatively low. When interest rates are higher, it is typically the rules related to the borrower’s ability to service the debt that will be the first constraint.

Before the lending rules: Interest-only loans could imply highly indebted homebuyers

Some of the lending rules restrict access to interest-only mortgages. That is the case for the first rule introduced in 2013, which requires borrowers to be able to service a fixed-rate repayment mortgage to get an interest-only mortgage. Prior to 2013, the individual credit institution would themselves determine the minimum credit standards when offering interest-only mortgages.

The rule that borrowers must be credit-assessed based on a fixed-rate repayment mortgage significantly reduced the maximum loan amount for interest-only mortgages with variable rate; see chart 2. With interest rates and incomes as in 2020, the typical homebuyer could theoretically sustain a modest consumption expenditure while simultaneously servicing a first-year payment equivalent to a total mortgage debt of kr. 10.6 million or 12 times the pre-tax income of the typical homebuyer. The combination of interest-only and variable-rate could thus enable some very large mortgages. The chart illustrates a possible mortgage given data on incomes and down payment. It does not show whether credit institutions would grant such a mortgage or whether homebuyers would apply for it. However, in the years before the financial crisis, when there were no rules regarding interest-only loans, credit institutions granted many large loans relative to borrowers’ incomes; cf. Danmarks Nationalbank (2017).

Overall, the lending rules restricted borrowers’ ability to increase leverage. The impact was greater when interest rates were relatively low. Accordingly, the lending rules have contributed to preventing the very low nominal interest rates during the period 2010-2021 to fully translating into larger mortgages. As a result, some homebuyers have had to borrow less than they would have been able to if there were no lending rules in place. The lending rules have thus moderated the extent to which the low interest rate environment affected both debt accumulation and housing prices.

Many homebuyers choose to borrow less than the maximum loan amount according to the lending rules. Where the previous section illustrated the maximum loan amount for each of the lending rules, this section instead illustrates how the lending rules restrict homebuyers depending on how much of their income they use to service the housing debt. Hence, the starting point is not the maximum loan amount according to the lending rules, but rather how much the typical homebuyer can borrow given a fixed amount spent on servicing the housing-related debt relative to the after-tax income.

The lending rules do not constitute a restriction for the typical homebuyer with a debt service-to-income ratio corresponding to the median; cf. chart 3. The debt service-to-income (DSTI) is defined as the sum of interest payments and instalments divided by the after-tax income. The horizontal lines represent the loan amount where the first lending rule is binding, as illustrated in the previous section. The loan amount for the typical homebuyer with a median DSTI is below the maximum loan amount according to the lending rules in 2020 as well as in 2023. Hence, a typical homebuyer with a DSTI corresponding to the median can in principle borrow more according to the lending rules but chooses not to.

For the typical homebuyer with a DSTI of 30 per cent, the interest rate level inversely affects whether the lending rules constitute a constraint. When interest rates were low as in 2020, a DSTI of 30 per cent would imply that the limits in one or more lending rules would be exceeded. The reason is that homebuyers can service a large debt relative to their incomes when interest rates are low. Therefore, the rules regarding borrowers’ DTI ratio will typically be binding during periods of low interest rates. Conversely, when interest rates are higher, the lending rules do not restrict the typical homebuyer with a high DSTI. This is because a mortgage interest rate of around 5 per cent and a DSTI of 30 per cent entail a loan amount that is less than four times the income of the typical homebuyer. At the same time, it falls within the limits set by the other lending rules.

When borrowing is cheap, more homebuyers are restricted by the lending rules

In the following, the starting point is how homebuyers with a high DSTI of 30 per cent are restricted by the lending rules. How lending rules constrain the typical homebuyer over time is illustrated in chart 4. The maximum loan amount according to the lending rules is when the typical homebuyer has only a minimum consumption expenditure, see box 1 and the turquoise curve in chart 4.

The red curve illustrates the loan amount for the typical borrower with a DSTI of 30 per cent. Around 2021, this loan amount starts decreasing. The declining trend reflects increasing long-term interest rates. The grey curve illustrates the loan amount until the first rule becomes binding. Up to and including the first quarter of 2022, the loan amount corresponding to a DSTI of 30 per cent exceeds the loan amount until the first rule becomes binding. After this point, the lending rules do not restrict the typical homebuyer with a high DSTI.⁵ The lending rules have thus contributed to reducing risks related to the low-interest rate environment: When interest rates are low, more borrowers are restricted by the lending rules.⁶

The lending rules have been relatively robust in terms of restricting the maximum loan amount on interest-only mortgages with variable interest rate

The purpose of some of the lending rules is to restrict the access to interest-only mortgages. Chart 5 illustrates how the typical homebuyer who applies for a full interest-only mortgage with variable rate is restricted by the rules over time. Chart 5 differs from charts 3 and 4, which illustrated how the lending rules restrict the typical homebuyer choosing a 30-year fixed-rate repayment mortgage.

With low interest rates as in 2020-21, the typical homebuyer with a DSTI of 30 per cent would be able to service the first-year payments corresponding to an interest-only mortgage of around kr. 8 million; cf. chart 5. In practice, the typical homebuyer would borrow a significantly smaller amount, as the first lending rules become binding at a loan amount of kr. 3.6 million; cf. the grey curve.

Over time, the lending rules have been relatively robust in terms of limiting the access to full interest-only mortgages with variable interest rates. Thus, the typical homebuyer with a high DSTI of 30 per cent has been restricted by the rules throughout the period, while even a typical homebuyer with a DSTI corresponding to the median was restricted when interest rates were low, as in 2020-21. After the interest rates have increased, the lending rules do not generally constitute the same restriction in relation to interest-only mortgages. This is mainly because the difference in payment between an interest-only mortgage and a repayment mortgage decreases when interest rates rise.

Data for all homebuyers: Higher interest rates have implied that fewer homebuyers with high DSTIs are restricted by the debt-to-income rules

Whereas the previous sections were based on an analysis of a typical homebuyer, the following section is based on data for all homebuyers over the period 2020-23. The starting point is still whether the lending rules will constitute a restriction if the homebuyer chooses a debt service-to-income ratio of 30 per cent. The picture is basically the same as in the analysis of the typical homebuyer: The lending rules restrict fewer homebuyers with a high DSTI today than before; cf. chart 6. In 2020-21, most homebuyers would thus be restricted by one or more lending rules if the chosen loan amount would correspond to a DSTI of 30 per cent. In 2023, on the other hand, only around a quarter of homebuyers with a high DSTI would be restricted by the lending rules.

It is especially the debt-to-income rules that do not constitute the same restriction after interest rates have increased. The down payment requirement as well as the interest rate stress, on the other hand, both constitute a relatively constant restriction for homebuyers with a high DSTI. The relatively constant share of homebuyers that are restricted by e.g., the down payment requirement can be explained by the fact that incomes and down payments do not vary much from quarter to quarter. Accordingly, the share of homebuyers restricted by the down payment requirement at a DSTI of 30 does not vary much over period either.⁷ Two factors can explain why the higher interest rates do not imply that more homebuyers become restricted by the interest rate stress or the rule that borrowers should be able to service a 30-year fixed-rate repayment mortgage. On the one hand, higher interest rates imply that the two rules become binding at a lower loan amount. On the other hand, higher interest rates also imply a lower loan amount for the homebuyer opting for a mortgage corresponding to a debt service-to-income ratio of 30 per cent. Hence, increasing interest rates do not necessarily imply that more homebuyers become restricted by either the interest rate stress or the rule requiring borrowers to be able to service a 30-year fixed-rate repayment mortgage, when viewed relative to how rising interest rates affect homebuyers with a fixed budget to pay interest and instalments of 30 per cent of their income.

Not much of a difference between how young and older homebuyers are restricted by the lending rules compared to other homebuyers

The lending rules restrict homebuyers differently because homebuyers have different incomes and wealth. However, all homebuyers share the common experience that the lending rules become more restrictive when interest rates are low, as illustrated in charts 7 and 8. This applies both to first-time buyers under the age of 35 and to homebuyers older than 60.

Relatively more homebuyers over the age of 60 are restricted by the lending rules if the loan amount corresponds to a DSTI of 30 per cent.⁸ This is mainly because homebuyers older than 60 as a group have a lower income. Accordingly, they will get a DSTI of 30 per cent at a lower loan amount than other homebuyers. However, homebuyers above the age of 60 often borrow less than other groups of homebuyers. Therefore, there is not much of a difference in the extent to which the lending rules bind different groups of homebuyers who have a DSTI corresponding to the median for each group; cf. chart 8. First-time buyers under 35 years with a DSTI corresponding to the median for first-time buyers are more often restricted by the lending rules than the other groups. This is because they tend to have higher DSTIs compared to other homebuyers.

This section highlights the development in some of the variables that are affected by interest rates and lending rules. Where the previous section illustrated that the lending rules generally restrict more homebuyers when interest rates are low, this section describes the development of aggregated variables such as the debt-to-income ratio, the debt service-to-income ratio, etc.

The higher interest rates since 2022 have gone hand in hand with a declining share of homebuyers with a debt-to-income ratio greater than four; cf. chart 9. As a share of total new lending for house purchase, new lending to homebuyers where the total debt exceeded four times the household income before tax, fell from around 30 per cent in 2020-21 to less than 15 per cent in 2023. The decreasing debt-to-income ratios may be a result of debt being more expensive to service. As a result, homebuyers on average tend to borrow less. Also, with a fixed loan amount, the debt-to-income ratio will naturally decease following a decline in housing prices, particularly as wage inflation has remained relatively high during the same period.

The higher interest rates have also affected homebuyers' debt service-to-income-ratios. In 2023, homebuyers used a larger share of their disposable income to pay interest and instalments on their housing debt compared to previous years. This applies both to the median homebuyer as well as the 90th percentile. The higher DSTIs can in part, be attributed to lending rules being less of a restriction for homebuyers, when interest rates rise, as discussed previously.

If homebuyers spend a fixed share of their incomes on paying interest and instalments on the housing debt, higher interest rates will imply that they must buy a cheaper home. Instead, the higher observed DSTIs mean that homebuyers have been willing and able to spend a larger share of their incomes on servicing housing debt. The increasing debt service-to-income ratios of homebuyers since 2022 may help to explain why the fall in housing prices in Denmark has been relatively modest since the monetary policy was tightened, cf. e.g. Danmarks Nationalbank (2024).

The higher debt service-to-income ratios as interest rates have increased may, on the one hand, reflect changed preferences among homebuyers regarding the size of debt servicing expenditures as a share of their monthly budget. On the other hand, the higher DSTIs since 2022 may also illustrate that some homebuyers borrowed less during a low interest rate environment than they would ideally have wanted to, because the lending rules would restrict more homebuyers when interest rates were low.⁹ Restrictions on the debt-to-income ratios implies, for example, lower DSTIs all else being equal.

Finally, higher DSTIs may also reflect the relatively high price and wage increases since 2022. This is because higher inflation implies that the real value of the debt is eroded faster. High inflation after the debt is originated turns a relatively high DSTI into a smaller DSTI in real terms after a few years, thus making it easier for the homeowner to service the same nominal debt. Homebuyers may therefore be more willing to accept high initial interest burdens in periods of relatively high inflation because the real interest burden is expected to be lower after a few years.

Increasing share of interest-only mortgages with variable rates

The trend of falling debt-to-income ratios and increasing debt service-to-income ratios have coincided with an increasing number of homebuyers who are granted interest-only mortgages with variable rate; cf. chart 10. The proportion of homebuyers with this type of mortgage has thus increased from approximately 10 per cent before the interest rate increases to 15-25 per cent after. The development reflects, on the one hand, that the issuance of mortgages with variable rates has increased, while on the other hand, the total issuance of interest-only mortgages has fallen. The observation that relatively more homebuyers are granted interest-only mortgages with variable rates may reflect that fewer homebuyers have a debt-to-income ratios greater than four. Lower debt-to-income ratios imply that the lending rules are less of a restriction when granting interest-only mortgages with variable rate.

While the issuance of interest-only mortgages with variable rates has increased, the total new lending of interest-only mortgages has fallen. This can be due to several factors. Firstly, the relative difference between the monthly payment on an interest-only mortgage and a repayment mortgage decreases when interest rates go up. This may encourage borrowers to choose repayment mortgages to a larger extent. Secondly, instalments are a kind of savings, and when the return on savings increases due to higher interest rates, repayment mortgages become more attractive.

The legal framework provides credit institutions with some flexibility to grant loans even if the limits of the lending rules may be breached. Some of the possibilities to deviate are explicitly mentioned in the set of rules. For example, the credit assessment of older homeowners applying for an interest-only loan to withdraw home equity does not necessarily need to be based on a repayment mortgage. Another example is that the down payment requirement should be ”assessed in relation to the specific customer and can be revised both upwards or downwards”; cf. the Danish Financial Supervisory Authority (2023a). A third example is that young customers can be exempted from the DTI rule of Growth Area Guidelines if they have good job prospects and choose a 30-year fixed-rate repayment mortgage. The flexibility of the lending rules differs from the rules in the Act on Mortgage Loans and Mortgage Bonds, where the upper LTV-limit for mortgages financed by a mortgage credit institution issuing covered bonds is 80 per cent, regardless of whether an assessment of a specific customer may justify a higher LTV. ¹⁰ However, the two sets of rules have different purposes. The purpose of the Act on Mortgage Loans and Mortgage Bonds is to protect bond investors, while the purpose of the lending rules in this analysis is consumer protection etc.

Over the period 2020-23, credit institutions have granted a relatively constant share of loans beyond the limits of the down payment requirement and the DTI rules; cf. chart 11.¹¹ Thus, around 9-12 per cent of the total new lending for home purchases was issued to homebuyers with either a low down payment or a high DTI and a loan type not permitted according to the rules in the Order on Good Practice for Mortgage Lending. There has been a downward trend in 2023, reflecting fewer home purchases resulting in a DTI ratio greater than four.¹²

Issuing loans beyond the limits of the lending rules does not necessarily imply that the lender does not comply with the rules, cf. the flexibility of the rules. The relatively constant share of deviations over time may indicate that credit institutions have had the same approach to the granting of loans both before and after the interest rate increases. Hence, the higher interest rates do not seem to have changed this part of the credit policy.

First-time buyers and homebuyers with low incomes are more often granted loans beyond the limits ​​of the lending rules

Some groups of homebuyers are on average more likely to be restricted by the lending rules than other groups. A statistical model can illustrate which groups that are more likely to be granted a loan beyond the limits; cf. table 2. The table shows characteristics of homebuyers that are more frequently granted loans beyond the limits of the lending rules. A positive sign means that the variable is associated with a higher frequency when the other variables are taken into account. As in chart 11, only the down payment requirement and the rule are considered. 

Table 2 shows that lenders more often grant loans beyond the limits ​​of the lending rules to first-time buyers, who make up 65 per cent of the homebuyers beyond the limits of the rules ​​and 41 per cent below. Younger homebuyers as well as homebuyers with relatively low incomes and low gross wealth are also more frequently approved for loans exceeding the limits. These groups may experience higher income growth after buying a home which may justify why credit institutions approve loans even if some limits in the lending rules are exceeded. The negative sign of income and gross wealth may indicate that homebuyers with higher incomes and assets have less need to borrow up to the limit. The flexibility in the lending rules is thus used to accommodate homebuyers with relatively low incomes, which is in line with the rules’ legal basis in the consumer protection legislation.

Homebuyers in households consisting of two adults, generally have a higher probability of getting a loan beyond the limits of the lending rules, all other things being equal. Likewise, a higher probability is associated with having more children living at home. However, there is not much difference between the average family size across homebuyers who are granted a loan beyond the limits and those who are not; cf. the second and third column.

There are also geographical differences in how credit institutions grant loans beyond the limits of the lending rules. All else being equal, the frequency is greater for home purchases in Greater Copenhagen and Aarhus as well as in municipalities outside the larger cities. Conversely, the probability is correspondingly lower in the larger provincial cities, which are the reference group in the regression. However, there are no large geographical differences in loans that have been granted above or below the limits of the loan regulations.

Danmarks Nationalbank (2017), Risks are building up in the financial sector, Danmarks Nationalbank Analysis (Financial Stability – 2nd half 2017), no. 23.

Danmarks Nationalbank (2024), Inflation is on track but some inflationary pressure persists, Danmarks Nationalbank Analysis (Outlook for the Danish economy), no. 2, March. 

De Nicolò, G., G Dell’Ariccia, L. Laeven and F. Valencia (2010), Monetary Policy and Bank Risk Taking, IMF staff position note, SPN/10/09.

ECB (2023), Evolution of mortgage lending standards at the turn of the housing market cycle, Macroprudential Bulletin – Focus, no. 22

Ministry of Industry, Business and Financial Affairs (2023), Bekendtgørelse nr. 1514 af 6. december 2023 om god skik for boligkredit (only in Danish).

Danish Financial Supervisory Authority (2023), Vejledning nr. 9458 af 1. juni 2023 til bekendtgørelse om god skik for boligkredit (only in Danish).

Danish Financial Supervisory Authority (2023), Vejledning nr. 9299 af 20. april 2023 om forsigtighed i kreditvurderingen ved belåning af boliger i vækstområder mv. (only in Danish).

Laeven, L. A. Maddaloni and C. Mendicino (2022), Monetary policy, macroprudential policy, and financial stability, ECB discussion papers, no. 2647.

Lian, C., Y. Ma, C Wang (2019), Low Interest Rates and Risk-Taking: Evidence from Individual Investment Decisions, The Review of Financial Studies, 32, 6, pp. 2107-2148.

The Systemic Risk Council (2017), Recommendation on limiting risky loan types at high levels of indebtedness.

The Committee on the causes of the financial crisis (2013), The financial crisis in Denmark - causes, consequences and lessons, Ministry of Industry, Business and Financial Affairs.

The analysis consists of a Danish and English version. In case of doubt as to the correctness of the translation, the Danish version will prevail.

Editing completed on 23 May 2024