Past rate hikes will not hold back eurozone consumers

Households’ debt servicing costs have increased since the start of the European Central Bank’s rate hiking cycle, but the pass-through of higher policy rates to households in different eurozone member states varies widely owing to differences in borrowing habits. Whereas households with variable rate debt have already seen their interest costs increase sharply, only some holders of debt with longer fixed-rate periods have experienced an increase in costs. Over the past months several ECB officials, including Chief Economist Philip Lane, have suggested that some further tightening of financial conditions should be expected as more fixed-rate loans reset to a new, higher rate.

In this note we focus on the transmission of the ECB’s past hiking cycle to the cost of household debt. For the eurozone five largest member states, we look at the change in interest costs over the past two years and attempt to quantify the shock that is still expected. We find that higher rates may have dampened consumer demand somewhat, but we find no convincing evidence that the remaining impact of rate hikes will substantially affect consumer spending in the coming quarters. This, we argue, reduces the ECB’s urgency to cut rates and allows for a gradual easing cycle.

The ECB’s rapid hiking cycle has driven interest rates to their highest point in more than a decade. Yet, the impact on households’ debt servicing costs to date differs substantially across countries.

Households across the eurozone have seen an increase in interest payments in the past two years (see figure 1).[1] In aggregate, this increase was the sharpest in Spain where quarterly interest payments almost doubled between the end of 2021 and the end of 2023. Italy comes in second, with an increase of as much as 60%. Meanwhile, the value of interest payments has remained almost flat in France. The differences between countries are mostly driven by the amount of household debt, the share of variable rate loans, and the extent to which changes in the policy rate are passed on to retail interest rates.

Italian and Spanish households have lower debt than the euro area average (see figure 2), but variable rate loans are much more prevalent than in their northern counterparts. The interest rates on new mortgage loans have also risen relatively fast in Italy, and more than in its peers (see figure 3). By contrast, French households have more debt, but interest rates tend to be fixed for the entirety of the loan. This obviously slows the transmission of higher interest rates to existing debt. On top of that, a clear transmission lag can be observed in French borrowing costs due to the usury rate, which caps the rates commercial banks can charge their clients. The French central bank sets this usury rate based on the interest rates that prevailed in the prior quarter, thereby limiting the pace of any upside adjustment to lending rates.[2]

Meanwhile, Dutch households by far have the largest debt burden among the largest five member states, but interest rates on mortgages tend to be fixed for 10 years or longer – after which they generally reset to another fixed term, if the loan has not matured yet. The latter also holds for Germany. So even though German lending rates have risen faster than in its peers (except for Italy), the pass-through to debt holders has been mitigated by the prevalence of long fixed-rate periods.

[1] Note: Figure 1a is based on interest paid by households before financial intermediation services indirectly measured (FISIM) allocation, according to the national/sectoral accounts. It is not adjusted for mortgage (interest) tax deductions, which can substantially reduce the impact of higher interest rates on households’ finances. In the Netherlands, taxpayers with a mortgage on their primary owner-occupied residence are allowed to deduct up to 43.6% (2024) of their mortgage interest payments if the mortgage is being repaid within 30 years on a straight-line or annuity basis. In Italy, one can deduct 19% of paid interest on a first-home mortgage, up to a maximum of EUR 4,000 per year. So the maximum tax credit amounts to EUR 760. In Spain, the measure was abolished in 2013 (with some regional exceptions), but it is still applicable on mortgages on a main residence bought before January 1, 2013. In those cases, taxpayers can deduct 15% of at most EUR 9,040 in mortgage interest and repayments, annually.

[2] Between February and December 2023, the usury rate was temporarily set every month due to the rapid rate hikes.

Eventually, rate hikes should also affect households in countries where fixed-rate loans are more prevalent, but that adjustment process is very gradual and may take years. Borrowers who took out a loan between 2016 and 2021 with an interest rate reset this year will probably feel the biggest impact.

Those who took out a loan in the past two years and those who borrow against variable rates have already felt the largest impact of higher rates. Owing to the inverted curve, this group may even be able to lock in lower rates than they currently pay – provided that this option is actually available to them. In the coming 12 to 18 months, the biggest interest rate shock will still be felt by those households with older loans who see their fixed-rate period expire.

That is particularly true for households that managed to lock in very low interest rates, that is, those who took out a loan between 2015 and 2021, or refixed their interest rate in that period. Meanwhile, those who originally borrowed for 10 years or more may not see as much of an interest rate shock this year, due to the rates that prevailed over a decade ago. In fact, some of these mortgages may still reset to lower rates.

Fixed rates limit the near-term impact

While individual households may feel the impact of the hiking cycle, we estimate that the aggregate impact is fairly small. The fact that only a very specific set of loans is affected by higher rates in the near term limits the transmission of higher rates to consumers.

Households across the eurozone have, on average, already felt the largest impact of higher rates on the interest costs of existing loans. We estimate that this interest rate shock peaked last year. Households that had to re-fix their interest rate, or saw their loans reset to variable rates, between July 2022 and December 2023 faced an increase in borrowing costs of up to 250 basis points.

That’s a substantial interest rate increase for the affected households, but bear in mind that this only applies to those loans that actually reset. Accounting for the fact that only a part of the entire loan portfolio resets at any given point in time, we estimate that the impact on the total amount of outstanding loans has been much smaller (see figure 4).

And considering that variable rate loans have largely been reset to higher rates by now, the increase in borrowing costs over the coming period will have to be driven by the more gradual resetting of fixed-rate loans. Meanwhile, depending on the speed of the ECB’s cutting cycle, some variable rate loans may even reset to lower rates from 2025 onwards. This implies that the impact will be substantially smaller.

Big differences across countries

This average eurozone rate increase is not very representative for individual countries. As noted above, Spanish borrowers have in aggregate seen a far bigger jump in interest costs. However, on an individual basis, the increase in retail lending rates may have hit German households relatively harder. We estimate that the difference in interest rates before and after the interest rate reset date is actually the highest in Germany and France (see figure 5). The relatively low share of loans with an interest rate reset in any given year mitigates this impact on the total stock of outstanding loans (see figure 6).

By comparison, the small interest rate shock in the Netherlands stands out. We believe that this is largely due to the low amount of consumer credit in the country; mortgages make up a very large majority of household borrowing. The share of consumer credit is significantly higher in both France and Germany. Consumer credit will, on average, have shorter periods of rate fixation. Moreover, Dutch mortgage rates have risen less sharply (see figure 3), which limits the impact on those loans that see their fixed rate expire.

Not immune

The relatively small interest rate shock does not mean that Dutch, German, and French households are immune to higher interest rates, but our estimates do suggest that the impact is more delayed and spread out by long fixed-rate periods. We argue that it could take until 2027 or 2028 before the impact becomes more noticeable, as that is when 10-year fixed-rate loans that were originated during the low-rates environment of 2017 to 2021 start to reset. This matches DNB estimates for the Dutch market: The central bank has computed that the interest rate on the bulk of Dutch loans will not reset until 2026 to 2031.

Of course, prevailing interest rates at that time will determine just how big of an impact will still be felt by then. However, despite the imminent easing cycle, it is fairly safe to assume that interest rates will not be as low as they were in the late 2010s and early 2020s.[3] Meanwhile, for Spanish households in particular, the prospect of rate cuts may bring some relief next year. The extent of that relief hinges on the amount of rate cuts though.[4]

[3] Furthermore, our calculations focus on the impact of rate resets of existing loans and do not take into account the impact of the rate differential between new and maturing loans. Data is missing to accurately calculate this impact, but the available data suggests that the rate on new debt will remain higher than the interest rate on maturing loans in the coming years.

[4] A counterfactual analysis based on the forward curve as at early February underscores the sensitivity of Spanish variable-rate loans to the outlook for monetary policy: Earlier this year, it looked as though interest payments of Spanish households could start decreasing again in 2025, as variable-rate loans were about to reset to a lower rate than their current levels. However, with far fewer rate cuts priced into the interest curve currently, this windfall disappears.

To put the projected change in interest payments into perspective, we compare the past and expected interest rate shocks with household consumption in 2023. This indicates how much household consumption could suffer or benefit from the expected change in interest payments in the coming years, all else equal. (Households may, of course, also benefit from increased interest income. (See Appendix “Mitigating factors.”) Our estimates above equate to an increase in interest payments of at most 0.5% of consumption in Spain this year, following 1.1% in 2023. This may have caused some consumer weakness over the past year, relative to a counterfactual without rate hikes.

But in most countries the estimated pass-through is lower and in many cases also more protracted. So past rate hikes will probably continue to lift aggregate debt servicing costs, even though the ECB looks set to start a rate cutting cycle in June. On balance, the expected impact on consumption is fairly small and spread out over multiple years.

Admittedly, the impact on household finances could be higher if one takes into account that new loans will be issued at higher rates, but then we also have to take into account the various mitigating factors that will soften the burden of existing debt. Higher wage growth, for example, may be a particularly effective offsetting factor in those countries where the impact of higher rates is spread out over time.

All in all, higher rates may have dampened consumer demand somewhat. However, we find no convincing evidence that debt servicing costs will prevent or substantially slow a recovery in consumer spending in the coming quarters. This, we argue, reduces the ECB’s urgency to cut rates and allows for a relatively gradual easing cycle.

In reality we cannot treat the pass-through of a change in (policy) rates to interest payments to consumption as an isolated development. Several other factors may mitigate the impact.

Interest income

Firstly, rate hikes have not only increased household interest payments, but also their interest income. As with interest payments, developments in interest income have differed widely across countries. Such differences have been driven by the size and type of financial assets, as well as the speed with which higher policy rates have translated into higher deposit rates, for example.

Increased interest income has certainly dampened the negative impact of higher costs in Italy and Spain, but it actually grew faster than payments in the Netherlands, and especially in Germany and France. Hence, especially in the latter two countries, higher interest rates seem to have actually benefitted households more than they have caused pain – at least in aggregate. Accordingly, the development in remuneration of financial assets going forward will be important as well.

Wages, inflation, and saving preferences

Other variables in the mix are wage gains and inflation. Indeed, (real) wage growth could dampen the increase in the interest burden, measured as interest payments over income. Income growth is expected to slow on the back of lower employment growth, while inflation is projected to come in much lower this year – although the drop will be much more striking in Italy than Spain, for example.

Furthermore, households could opt to lower their saving rate rather than cut back consumption in response to higher interest costs. While saving rates have come down from their pandemic peak, they are still (or again) higher than pre-pandemic in all countries but Italy. Of course, households’ saving preferences may also be influenced by interest rates on deposits, for example. Theory suggests that, up to a certain point, higher remuneration of savings should induce households to cut back on spending and start saving.