MONETARY POLICY REPORT | SUMMER 2023 (box 2)
(author: Eva Hromádková)
A substantial part of monetary policy transmission in the household segment takes place via the effect of interest rates on the volume of genuinely new loans for house purchase. However, as the period since the CNB’s policy rates (and subsequently also client rates) went up increases in length, another channel is gaining in importance – the impact of the increased rates on current loan instalments, which are refinanced[1] over time. This box uses individual data on mortgages to quantify this impact in the segment of mortgage loans to households, which makes up the largest part of households’ debt burden.
The main data source used in this analysis is the Survey of new loans secured by residential property (the “Survey”), which is conducted every six months by the CNB’s Financial Stability Department. It includes information on all new bank loan agreements leading to a change in the volume of loans in the economy, i.e. new loans including loan increases and loans refinanced with different institutions.[2]
However, the Survey does not contain data on loans for which the rate was refixed with the original bank without any increase in loan size.[3] In commercial banks’ statistics, these refixed loans fall into the “other new mortgage agreements” category and have long accounted for around 30% of all new loans. Since the start of 2022, however, their share has increased to 60% owing to a substantial decrease in genuinely new loans. The total volume of new mortgage agreements used to calculate the volume of loan instalments and to create possible scenarios is shown in Chart 1. It contains both data from the Survey and other new mortgage agreements.[4]
Chart 1 – The largest volume of genuinely new mortgage loans was agreed in 2021; other new mortgage agreements now have a significant share
total volume of new mortgage agreements in CZK billions
Note: In 2020, the moratorium – involving agreements to temporarily suspend repayments – was reflected in the volume of other new mortgage agreements.
According to the data from the Survey, the most popular fixed-rate period is five years. It was agreed for 53% of total mortgage loans in the whole period under review (see Chart 2). However, the volume of loans with longer fixed-rate periods rose significantly in 2019 – the share of ten-year fixes reached 28% and that of seven-year fixes 31%. This led to an increase in the average fixed-rate period to 7.6 years. This can be explained by a decrease in banks’ cost of funds at the time (see the 5Y IRS rate in Chart 3) and an easing of banks’ credit standards applied to loans for house purchase compared to 2017 and 2018.[5]
Chart 2 – Clients most frequently agreed a five-year fixed-rate period
share of loans by fixed-rate period; in %
Note: The chart shows the five fixed-rate periods with the largest mortgage volumes. The rest of the fixed-rate periods account for 3% of all mortgage agreements on average.
Chart 3 – Interest rates on new mortgages rose sharply for all fixed-rate periods in 2022
interest rate on new mortgages and 5Y IRS rate in %; by fixed-rate period
Conversely, after monetary policy rates were raised in early 2021 and client interest rates subsequently rose sharply for all fixed-rate periods (see Chart 3), the average fixed-rate period fell to 5.1 years, with clients not wanting to commit to high rates for a long time.
Chart 4 shows the estimated time of future refinancing broken down by year of loan origination.[6] Loans totalling around CZK 420 billion, provided mainly in 2018 and 2019 at an average current interest rate of 2.7%, are expected to be refinanced over the next two years. The expected change in instalments was quantified using two versions of the possible path of interest rates. The first assumes that the average rate upon refinancing stays the same during 2023 and 2024 as it was in 2023 Q1. The second version ties the predicted client interest rate on refixed mortgages to the five-year IRS rate,[7] which represents the cost of funds for the most frequent fixed-rate period.
Chart 4 – The volume of loans for refinancing will peak in 2026 and 2027
volume of loans for refinancing in CZK billions; by year of origination
The overall macroeconomic impacts are limited. Assuming refinancing at current rates (version 1), the estimated total cumulative impact on household expenditure this year is around CZK 2.1 billion, with the volume of additional repayments gradually rising at the year-end. This amount represents the lower limit on the estimate due to the lack of information on part of the mortgages from the Survey. From the perspective of the costs of households with mortgages, this represents an almost 35% increase in instalments (CZK 3,200 on average), which may be a financial burden for them. The implications of version 2 are more moderate; the cumulative impact on expenditure this year is around CZK 1.3 billion, representing a rise in instalments of around 20% for households with mortgages (CZK 1,800 on average). In line with the CNB’s current forecast, this version assumes a relatively sharp decrease in reference rates this year and the next.
Another output of this analysis is also important from the long-term perspective. Chart 4 shows that large volumes of loans provided between 2020 and 2022 (CZK 907 billion in total) are to be refinanced in 2025–2028. These loans were granted at interest rates of around 2%. The slope of the yield curve in this period will thus have a significantly greater potential effect on households’ expenditure than in 2023–2024.
[1] For the purposes of this box, refinancing means both a change in the interest rate at the original bank (refixation) and the refinancing of a loan at a different bank than the original one.
[2] The survey covers around 936,000 loans provided between July 2015 and February 2023. Mortgages that were granted before 2015 and have not been refinanced to date (i.e. loans with rates fixed for more than eight years) were not included in the sample. In 2014, these loans accounted for only 2%–3% of all new loans, so their volume is currently negligible. The type of data also determined the analytical approach taken: as no information is available on the parameters of the current stock of loans, we have to track the flow of new mortgage agreements over time and estimate the moment at which they will be due for refinancing based on their fixed-rate period.
[3] For the purposes of this analysis, we use information on the overall volume of, and average interest rates on, refixed loans from commercial banks’ statistics. The fixation and maturity structures are approximated using individual data from the Survey for the category of loans refinanced with other banks.
[4] The volume of mortgages used in the calculations was adjusted for a one-off administrative factor – the loan moratorium (involving agreements to temporarily suspend repayments in 2020). The calculations based on individual loans (such as the average repayment increase) are based exclusively on data from the Survey.
[5] See the results of Bank Lending Survey I/2020.
[6] The calculation assumes no early repayment or refinancing of the existing loan and full refinancing at the end of the fixed-rate period. Early repayment or a decrease in the loan volume at the end of the fixed-rate period through partial repayment would lead to a smaller volume of loans, so the estimated rise in loan instalments represents the upper limit on the impact on repayments of existing loans. On the other hand, the analysis abstracts from loans that will be provided in the years ahead and will thus increase the refinancing volume.
[7] The second version implicitly assumes the same interest rate path as that in the baseline scenario of the CNB’s summer forecast, namely a gradual decline in market interest rates over the entire outlook.