The global surge in inflation following the COVID-19 pandemic prompted many central banks to raise interest rates sharply. A key channel through which monetary policy transmits to the economy is mortgages (e.g. Cloyne et al. 2021). In many countries, mortgages typically have fixed rates for five, ten, or even more years. Fixed rates slow the pass-through from higher market rates to existing loan rates, but they do not eliminate it – in many cases, households eventually need to refinance their mortgage when the fixation period ends. In such settings, the timing and magnitude of the transmission of rate changes to household finances depend crucially on whether borrowers are aware of and prepare for future increases in their mortgage rate. If households adjust in advance, they may dampen transmission; if they do not, strains on household finances and risks to financial stability can materialize.

In a recent paper (Fuster et al. 2026), we study how borrowers prepare for and respond to changes in their mortgage rates. We do so in the German context, combining three sources of evidence: (i) loan-level administrative data from a large bank, (ii) an online survey of its mortgagors, and (iii) a randomised controlled trial (RCT) in which the bank sent an informational letter to a subset of mortgage customers. A growing literature in economics employs similar approaches (e.g. Gorodnichenko et al. 2021, Schnorpfeil et al. 2023).

Institutional setting: Predictable resets, but room for preparation

A key feature of German mortgages is the separation between the overall maturity of the loan and the length of the interest-rate fixation period. When the fixation ends, borrowers typically refinance onto a new contract.  An important feature of the German setting is that borrowers can lock in interest rates up to four years before fixation ends with a ‘forward mortgage’. Upon refinancing, they can also pay down part of the balance or adjust the loan term.

This institutional setup makes the prospective ‘payment reset’ both predictable and, to a meaningful extent, manageable. Because the reset is anticipated, borrowers have incentives to adjust their balance, contract choices, and refinancing plans in advance, potentially dampening the realised change in monthly payments when the fixation expires. However, such adjustments require that borrowers are aware of the rate environment and their contract structure.

Bank data: Realised payments increase much less than under passive refinancing

Our dataset from the partner bank covers mortgagors observed through May 2025. The data include contract terms and payment information, and allow us to link an expiring to a new loan if refinancing occurs internally at the partner bank.

To quantify the pass through of rate increases, we construct a ‘passive counterfactual’: what would happen to monthly payments if borrowers simply rolled over their remaining balance at expiration, without changing contract terms. We then compare this passive benchmark to the realised payment change observed when borrowers refinance.

Figure 1 summarises the result. Borrower actions mitigate the impact of rate changes on payments. For loans expiring after the second half of 2022 – that is, during the high-rate environment – monthly payments would have increased substantially under passive refinancing (about €87 per month, or €1,044 per year). Instead, the realised increase is much smaller, at around €23 per month (or €276 per year).

Figure 1 Changes in monthly payments following internal refinancing

Notes: Average change in monthly payments upon internal refinancing: realised (red) vs passive counterfactual (blue). The passive counterfactual assumes borrowers roll over remaining debt at expiration without extra payments or contract adjustments.

This gap between the passive counterfactual and realised payments implies that the cash-flow consequences of rate hikes are not purely mechanical. Instead, borrowers often adjust along multiple margins: they reduce their loan balance, slow down debt repayment, and lock in interest rates in the early phase of the 2022-23 policy tightening.

Note also that during the low-rate period until early 2021, borrowers on average took actions that increased their new monthly payment relative to the counterfactual. This mostly reflects a shortening of the loan term upon refinancing, such that the mortgage amortises faster. Such behaviour is consistent with monthly payment targeting (Argyle et al. 2020).

Survey evidence: Borrowers are generally informed and actively prepare

Because the observational patterns provide only a partial view of borrower behaviour and their drivers, we fielded a survey among a subset of borrowers in late 2024. Three survey patterns are particularly relevant for interpreting the limited realised payment response:

1. High awareness of the rate environment. Most respondents are well-informed about the recent increase in mortgage rates and their current level.
2. Widespread preparation for fixation expiration. Most borrowers state that they prepare for the end of the fixation period. Preparatory margins include acquiring information and increasing savings, which are only partially visible in the bank data.
3. Causal evidence from vignettes. An experimental component reveals that borrower preparation is sensitive to the size of the rate increase – a pattern that aligns with the notion of forward-looking, state-contingent preparation.

The survey also indicates that the minority of borrowers who do not prepare are more likely to be unaware of rate increases – suggesting that informational frictions exist, even if average awareness is high – and have smaller balances, consistent with a smaller incentive to acquire information.

Letter RCT: Limited effects on beliefs, subgroup acts, selection into reading

To causally study the role of attention and awareness, we partnered with the bank to run a letter RCT in November 2024. The bank sent a letter (on official letterhead) to over 35,000 mortgagors who needed to refinance at some point. The letter explained the rise in rates, illustrated how higher rates can affect monthly payments after refinancing, and described coping options (including partial repayments and increasing savings).

The survey conducted about one month later reveals that letter receipt did not affect beliefs about mortgage rates or expected payment changes – consistent with high ex-ante informedness about rates. However, the letter did significantly increase familiarity with specific mortgage features and options, notably partial prepayments and forward loans.

We then examine bank outcomes in the six months following letter receipt. Average treatment effects are small, but there is meaningful heterogeneity by proximity to the refinancing date: among borrowers close to loan expiration – those with immediate scope to act – the letter strongly increased the likelihood of refinancing.

Finally, we document selection into reading the letter (as reported in the survey): borrowers with larger loans, higher income and self-reported financial literacy, and more accurate beliefs about rates were substantially more likely to report reading the letter. This selection limits the effectiveness of communication aimed at reaching the least informed.

Conclusion

Our evidence implies that borrowers’ anticipatory actions substantially dampen the realised payment pass-through from market rate increases, relative to a passive-refinancing benchmark. In our setting, borrowers are generally aware of higher rates and often prepare for fixation expiration, which helps explain why realised payment increases are modest even after large rate hikes. This result contrasts to some extent with findings of limited awareness and understanding of US borrowers (e.g. Bucks and Pence 2008).

At the same time, informational frictions are not absent: targeted communication can raise awareness of specific coping options and, when refinancing is imminent, can measurably shift behaviour. Yet the limited share of borrowers who engage with such communication (and the selection into engagement) suggests that reaching less aware households remains challenging.

References

Argyle, B S, T D Nadauld, and C J Palmer (2020), “Monthly Payment Targeting and the Demand for Maturity”, Review of Financial Studies 33: 5416–5462.

Bucks, B and K Pence (2008), “Do borrowers know their mortgage terms?”, Journal of Urban Economics 64: 218–233.

Cloyne, J, C Ferreira, and P Surico (2020), “Monetary policy when households have debt: new evidence on the transmission mechanism”, Review of Economic Studies 87: 102-129.

Fuster, A, V Gianinazzi, A Hackethal, P Schnorpfeil, and M Weber (2026), “The response of debtors to rate changes”, CEPR Discussion Paper 21093.

Gorodnichenko, Y, M Weber, and O Coibion (2021), “How inflation expectations affect households’ spending decisions”, VoxEU.org.

Schnorpfeil P, M Weber, and A Hackethal (2023), “Households’ response to the wealth effects of inflation”, VoxEU.org.