Temporary dip in house prices and sales expected from 2021

While macro-economic figures are historically low as a result of the coronavirus crisis, the Dutch housing market is still booming. An important reason why the corona crisis has had no effect appears to be due to the Dutch government’s support measures, which so far have been fairly effective in preventing sharp rises in unemployment among the group of potential homebuyers. Even so, we expect that this group’s position on the employment market will eventually be affected, and so the housing market will not remain unscathed. Strong activity in in the first seven months of this year mean that house prices and sales will be much higher in 2020. However, from 2021 we expect a temporary dip in prices and sales.

The housing market figures published by Statistics Netherlands and the Land Registry show the situation on the housing market some three months previously, due to the delay between signing the provisional purchase agreement and the actual notarial transfer of ownership. The fact that house prices were still rising strongly in the second quarter was therefore not surprising, since those figures largely related to the situation before the coronavirus crisis took hold. This makes the trends in the Dutch housing market in July more remarkable, since these figures related to houses that were sold around the end of March and early April – the period in which the corona crisis had us firmly in its grip. The order to stay at home as much as possible did not deter house hunters from going in search of their new home: more than 22,500 homes changed ownership in July, 7.3 percent more than in the previous year. An average of 7.4 percent more was paid for these homes than in July last year.

The loss of jobs and incomes, or the fear of these losses, is the main factor by which the coronavirus crisis will affect the housing market. Up to now, unemployment has risen fast, but more slowly than we had expected. The effectiveness of the Dutch government’s support measures would appear to be the reason for this. However, our forecasts for unemployment remain pessimistic: now that the government will gradually be phasing out the large-scale stimulus packages, we believe that the number of reorganizations and bankruptcies will rise in the coming quarters. This will mean not only more employees on flexible contracts losing their jobs, but also those with a permanent contract. We therefore expect unemployment numbers to peak somewhat later, but will remain high for longer (Figure 2).

The adjusted unemployment estimates has also led us to adjust our view of trends in house prices and sales. As unemployment will peak later, we expect the impact of the crisis on the housing market will also be felt a little later. Where we first assumed falling house prices in the last quarter of 2020, we now expect the first price falls to occur in the first quarter of 2021. As a result, 2020 will again be a strong year on the housing market, in which we believe that houses will be on average some seven percent more expensive. These strong house price rises in 2020 will also be reflected in the annual figures for 2021 by way of a spillover effect[1]. Despite four quarters of falling prices, the annual figure is expected to show a fall of ‘only’ 0.8 percent (Figure 3) overall. And although we expect only two quarters of falling house prices in 2022, year-on-year growth will in fact be much slower due to the sharp fall in 2021.

[1] The rise in house prices is usually reported as a year-on-year change. This produces what economists refer to as a spillover effect. To illustrate: in 2019 house prices rose very fast, as a result of which homebuyers paid considerably more in December 2109 than those who bought their house in January 2019. Let us assume that house prices do not rise further after December 2019. In that case, house prices in January 2020 are just as expensive as houses in December 2020. Although prices have not risen month-on-month, compared to a year earlier, i.e. January 2019, houses are substantially more expensive. Although the rise in house prices is in fact levelling off, the year-on-year rise is still considerable.

So far the job losses have been most severe among young people under 25 years old, and affect mostly school pupils and students with a part-time job (Figure 4). Since this group only accounts for a small proportion of house sales, the effect of rising unemployment on the number of house sales was still limited in July. Until the end of July, more than 129,000 existing homes changed ownership: 6.7 percent more than in the first 7 months of 2019. We expect fewer houses to be sold in the second half of 2020 than in the second half of last year. Nevertheless, the strong first seven months of 2020 lead us to expect that no fewer than 220,000 homes will be sold this year (0.6 percent more than in 2019). In 2021 we expect that job and income losses, or the fear of such, among a larger group of first-time buyers and homeowners moving house, as well as caution among investors and those moving house, will depress the number of sales to around 190,000 (Figure 5).

In the transaction figures for the first seven months of 2020, we can see the growing proportion of homebuyers in the age group younger than 35 years, largely at the expense of the age group older than 55 years (Figure 6). Although the age group younger than 35 years may also include those moving house, it is likely that a large proportion of these are first-time buyers. Before and at the start of the housing crisis of 2008 to 2014, we also saw a rise in the proportion of young homebuyers. At the start of a downturn, first-time buyers are possibly less likely to experience competition from hesitant homeowners who postpone a move – whether or not until the moment that their own house has been sold – or investors pulling out of the market, making it easier for them to find a house to buy. This would seem to be the case now too. Reduced competition from investors may also explain the growing supply in the major student cities (Figure 7). The absence of tourists, expats and students from the Netherlands and abroad as a result of the corona crisis has made investment in flats to let a less attractive proposition. And because investors are most active in the student cities according to research by the Land Registry, the effect of lower buy-to-let activity on the supply is precisely most marked in these cities.

Alongside the rising supply in the student cities, the initial effects of the coronavirus crisis can also be seen in the confidence indicators. In June the Eigen Huis Market Indicator was at 93, the lowest level for six years (Figure 8). The indicator has crept up slightly since then, to 97 in August. In view of trends in consumer confidence, however, it cannot be ruled out that confidence in the housing market will decline again: consumers were even more pessimistic in August about the expected unemployment, and felt it to be a less favorable moment to make large purchases. At present the majority of the Dutch are pessimistic about the situation on the housing market, but even so, confidence in the housing market is still well above the long-term average of 88.9.

All in all, during the second quarter the proportion of those who felt it was a good time to buy a house remained reasonably constant (Figure 9). But underlying this, the coronavirus crisis has indeed caused a shift in people’s views of the housing market. We see this above all in the reasons they give when deciding whether it is actually a good or bad time to buy. Where in recent years the lack of affordability and limited supply of houses for sale were the main arguments against buying a house, the unfavorable economic situation and pressure on household incomes are now the main reasons not to buy. On the other hand, for those who feel it is a good time to buy, low mortgage interest rates and the ease with which they can sell their existing house are the two main arguments. But at the same time there is a sharp increase in the proportion of people who believe it is a good moment to buy because they expect owner-occupied homes to become more affordable but mortgage rates to rise.

To conclude, although no coronavirus effect is discernible yet in the price and transaction figures, the first cracks are starting to appear in the supply and confidence figures. Once unemployment begins to rise faster, we also expect to see an effect of the coronavirus crisis on trends in Dutch house prices and sales. Provided that the coronavirus is kept under control and there will not be a second strict national lockdown in the Netherlands, prices this year are expected to be on average seven percent higher, followed by a fall of 0.8 percent in 2021. As regards the number of house sales, we expect there to be 220,000 this year and 190,000 next year.