Real estate markets in advanced economies in the face of monetary policy tightening
The real estate market is one of the main channels through which monetary policy is passed on to the real economy. Tighter financial conditions are feeding through to mortgage interest rates and are cooling down the demand for housing. Given the notable tightening of monetary policy over the past year in many of the advanced economies, in this article we document the change in trend in international real estate markets and analyse the extent of the adjustment that may lie ahead.
In many developed economies, house prices had been rising substantially for years, a trend that accelerated during the pandemic (due to a change in residential preferences, expansionary fiscal and monetary policies, limited housing supply, etc.). Global house prices went from 2.8% growth in 2019 to 4.3% in 2020 and accelerated to 11.8% in 2021 in nominal terms (1.4%, 3.5% and 8.6% respectively, in real terms).26 In the chart below, we can see that this upward trend slowed down in 2022, especially in real terms (–0.3% year-on-year in Q4 2022). The recovery in house prices was much greater than the trend for household income, leading to notable overvaluation in some markets. According to the Dallas Fed’s Exuberance Indicator, 12 of the 25 countries analysed showed signs of overvaluation in Q3 2021.27
- 26. Global house price indicator constructed from 25 advanced countries. See the International Housing Observatory.
- 27. The price correction observed since last summer in many economies has already cooled down some of these markets, so that in Q4 2022 only two countries still showed signs of overvaluation (New Zealand and Israel).
The most worrying cases are those in which the recovery in price was accompanied by an increase in household debt (which was already at high levels in some cases). This group includes New Zealand, Canada, Australia and the US. In Europe, the cases of note are the Netherlands, Luxembourg, Sweden, Norway and Denmark, countries in which, moreover, variable-rate mortgages predominate, making newly mortgaged households more vulnerable to rising interest rates. A second group is made up of Portugal, Germany and the UK, countries that have also recorded strong growth in house prices since the outbreak of the pandemic as well as, in the case of Germany and the UK, an increase in household debt.
Housing markets in several advanced economies began to adjust by mid-2022, against a backdrop of rising interest rates and eroding real household disposable income. The markets recording the biggest declines in house prices since their peak (with data up to March-April 2023, depending on the country) are Sweden (–10.5%), Canada (–8.9%), Denmark (–7.2%), Australia (–7.0%), New Zealand (–6.8%), the Netherlands (–6.1%), Germany (–5.3%) and the US (–4.5%).
Despite these declines, house prices in all these countries are well above their pre-pandemic level, as can be seen in the chart below. In addition, the most recent data for some countries indicate a certain pause in this price adjustment (especially in Australia, Canada and the US) while, in the case of Norway, a marked upward trend has resumed. House sales, meanwhile, have fallen at double-digit year-over-year rates in several of these markets so far in 2023: Australia (–42%), Sweden (–30%), the US (–23%), the UK (–19%), New Zealand (–15%) and the Netherlands (–11%).28
- 28. Average year-on-year change for the months of 2023 with available data.
Although the most recent data show a pause in the adjustment of house prices in some countries, we believe the downward trend will continue in the coming quarters as the tightening of financial conditions feeds through to the real economy.29 The speed and intensity of this adjustment in the various real estate markets will depend largely on the imbalances accumulated during the expansionary phase of the cycle. Consequently, corrections should be more extensive in those markets where: (i) house prices have grown more strongly, outstripping household income (overvalued markets); (ii) household debt is high and has been rising in recent years, and (iii) the financing of home purchases has become tougher due to the interest rate hikes by central banks in recent quarters.
To assess the potential house price adjustment that could occur in these markets, we have analysed in-depth the pressure exerted by more limited affordability and rising interest rates. Our calculations point to potential corrections (for the average of the countries analysed) of close to 15%. Let’s look at this in more detail.
- 29. Generally speaking, house prices tend to be relatively resistant to downward adjustment in changes of cycle. However, the sharp declines seen in the number of sales suggest that the adjustment in house prices will continue in the coming quarters.
1 To determine the force that both factors might exert on these markets, we have carried out two complementary exercises that provide insights into the potential house price adjustment that could occur. First, we calculated the fall in house prices required to bring the housing affordability ratio (house price in relation to household disposable income) back to pre-pandemic levels over the next two years. Second, we produced an econometric model that indicates the sensitivity of house prices to an interest rate shock for each of the countries.
According to the first methodology,30 and for the 25 countries analysed as a whole, the average adjustment in house prices over the next two years from their peak could be around 13%.331 However, there are notable differences between countries. The US and New Zealand are the markets where we estimate that house price adjustment would need to be the greatest (close to 20% peak-to-bottom) to regain the affordability ratio of late 2019. In both countries, prices began to fall in the middle of last year but a significant portion of the estimated total adjustment required is still pending. On the other hand, in Italy we estimate that house prices still have room to grow over the next two years as the expected increase in household income alone would already bring the affordability ratio to converge at the Q4 2019 level by the end of 2024. Denmark has already seen considerable house price adjustment (–7.2%), bigger than the estimated figure.
Sweden, Australia and Canada, which were among the markets overheating the most, have carried out about half the adjustment required. In contrast, Luxembourg and the Netherlands, with a similar adjustment potential (around 15%), are only a quarter of the way there.
- 1. We have assumed a period of eight quarters for this adjustment to take place since this is the time it usually takes for the impact of tighter monetary policy to peak after central banks raise interest rates.
- 30. This calculation assumes that household income in 2023-2024 (the denominator of the affordability ratio) grows at the same rate as real GDP and as per IMF forecasts (WEO October 2022). We have also carried out alternative calculations using GDP per capita and nominal GDP forecasts. The results are qualitatively similar. In other words, the adjustment required in order to regain the Q4 2019 affordability ratio is achieved, in part, thanks to an improvement in the denominator (income).
- 31. We have assumed a period of eight quarters for this adjustment to take place since this is the time it usually takes for the impact of tighter monetary policy to peak after central banks raise interest rates.
Although the potential adjustments estimated are large, it is also important to note that house prices would still be above their December 2019 level in all countries by the end of 2024. In other words, this is only a partial correction after the strong upturns recorded during the pandemic (see the chart below).
The second methodology32 indicates that the monetary tightening observed in recent quarters entails, on average, a potential price adjustment of close to 5%.33 Our analysis suggests that it takes about eight quarters for interest rate hikes to be fully passed on to house prices, pointing to a gradual cooling down of real estate markets that could continue into the second half of 2023 and 2024. Furthermore, this exercise enables us to separate the final impact of monetary tightening between two forces: on the one hand, the sensitivity of each economy to a given interest rate hike; and, on the other hand, the rise in interest rates actually observed in recent quarters. Similar to the first exercise, in some economies we estimate a potential impact of close to 15% while in others we predict hardly any impact at all.
- 32. We have reproduced and extended, to the rest of the countries, the Euro area model provided by Battistini et al. (2022), «The impact of rising mortgage rates on the euro area housing market», ECB, Economic Bulletin 6/2022. Specifically, we have estimated an autoregressive vector between house prices, residential investment and the mortgage interest rate, adding control variables (real GDP, CPI, short-term interbank interest rates and real estate credit), for the period 2003-2022. The impulse-response functions have been calculated using the local projection method.
- 33. We analysed a group of Anglo-Saxon countries (US, UK, Canada and Australia), a group of European countries (Euro area aggregate, Germany, France, Italy, Spain, Netherlands and Ireland) and a group of Scandinavian countries (Norway, Denmark and Finland).
For example, in the US, the estimated sensitivity of real estate prices to interest rates is not among the highest in the sample but monetary tightening has been extensive, resulting in a relatively high potential adjustment (almost 20%). The picture for the Euro area as a whole is similar, although slightly more moderate due to lower interest rate tensions (potential slightly below 15%). However, the estimated impact among the four large European economies (Germany, France, Italy and Spain) is clearly lower (between –2% and –10%). In the sample as a whole, the markets most affected by this exercise are the Anglo-Saxon markets, with Canada, Australia and the aforementioned case of the US, as well as the Netherlands in Europe.
Looking beyond the figures, there are two important caveats to be made here. First, the estimated adjustment is not abrupt but rather occurs over a two-year time period. Secondly, there are considerable differences between countries, with greater adjustment potential in the Anglo-Saxon economies (in particular, the US, Australia and Canada) and some heterogeneity in Europe. In this respect, it should be pointed out that Spain’s real estate market is in a good position, both in relative and absolute terms. It should also be noted that the calculations carried out gauge the potential drop in house prices in those countries where imbalances are most pronounced, but these are illustrative exercises which do not take all relevant factors into account and are therefore more a measure of risk than a forecast per se.
In fact, there are several factors that could support real estate markets. One notable factor is the good performance of the labour market, which is widespread among the economies analysed, as this boosts household income, helping to mitigate affordability problems and limiting the number of forced sales. In addition, households have healthier balance sheets and banks have ample capital buffers to absorb potential losses without triggering feedback mechanisms that push down prices. However, tougher housing affordability ratios and the tightening of monetary policy highlight the great challenges posed by the current economic environment.