STATISTICAL REFLECTIONS 9 November 2018 Contents Summary...1 Changes in property transactions...1 Annual price index...1 Quarterly pure price index...2 Distribution of existing home transactions...2 Regional characteristics of the market of existing...3 International data...3 Methodological notes...4 Summary In 2018, the housing market continued to expand. The 7.1 rise in the estimated number of home sales was entirely due to an increase in existing home sales, as no significant was seen in new home sales compared to a year ago. In the second quarter of 2018, prices grew year-on-year by 10.6 for existing and by 10.4 for new. In the first and second quarters of 2018, Hungarian housing prices, in inflation adjusted terms, were close to the pre-crisis level of 2008. Compared to ten years ago, inflation adjusted prices were only 0.4 lower for new and 1.6 lower for existing. At the same time, the housing market still shows diverging trends. Inflation adjusted house prices were significantly higher than ten years ago in Budapest and the county seats of more developed regions and lower in other parts of the country. Changes in property transactions Preliminary data show about 59 thousand home-purchase contracts concluded in the first and second quarters of 2018, a 7.1 growth compared to data of similar processing levels aggregated a year earlier. Only 2.9 of the dwellings sold during this period were newly built. In the first two quarters of 2018, a of 3,432 were built for sale, but data received so far show only 1,739 new home sales. Number of home sales and built for sale Year, quarter Home sales secondhand Of which: new Table 1 (Thousand units) built for sale 2007 191.2.... 17.9 2008 154.1 140.0 14.1 17.4 2009 91.1 82.9 8.3 16.9 2010 90.3 85.5 4.8 10.7 2011 87.7 83.9 3.9 4.8 2012 86.0 83.3 2.6 3.5 2013 88.7 86.4 2.3 3.2 2014 113.8 110.5 3.3 3.4 2015 134.1 130.7 3.4 3.1 2016 146.3 141.4 4.9 5.2 2017 153.8 147.7 6.1 7.3 Quarters 1 2 2018 59.0 57.3 1.7 3.4 Annual price index In the first two quarters of 2018, home prices grew by 8.2 in the second-hand market compared to 2017. It means that housing prices would have been that much higher if the same had been sold as a year earlier. However, the composition of sold shifted towards lower value resulting in a slower increase of 3.1 in the average price of second-hand actually sold compared to 2017. 1 Compared to the base year of 2015, the price index of second-hand reached 130. The quality composition was still lower than in the base period, indicating that the share of settlements with lower home prices grew in existing home sales. This is explained by the fact that housing market demand shifts toward new under construction in settlements with more favorable housing market position and especially in Budapest. In the first two quarters of 2018, sales prices grew year-on-year by 7.6 in the new home market, showing a 26 rise compared to 2015. 1 If we multiply the composition effect and the pure, we will get the index of.
2 Statistical reflections Also, it can be seen here that the price increase already exceeds the value for the whole of the previous year, so the rate of price increase has also accelerated since 2017. As built for sale are heavily concentrated in large cities, the quality index of new continues to exceed the base value so the average price of new has grown faster at 34 than the index of pure price since 2015. Table 2 Trends and factors of annual price s Year, quarter composition effect pure composition effect Second-hand pure () Previous year=.0 2008.7 102.2 102.9 88.6 101.8 90.1 2009 101.6 98.2 99.7 94.3 94.5 89.1 2010 102.9 93.6 96.3 109.8 97.9 107.5 2011 99.7 96.7 96.4 98.7 96.4 95.2 2012.7.0.7.4 96.2 96.6 2013 98.9 101.0 99.8 101.2 97.1 98.3 2014.3 104.4 104.7 102.6 104.2 106.9 2015 99.7 108.0 107.7.9 111.4 112.4 2016 97.4 110.5 107.6 92.9 113.3 105.3 2017 109.7 106.0 116.3 102.6 105.9 108.7 Quarters 1 2 2018 99.3 107.6 106.8 95.3 108.2 103.1 2015=.0 2007 94.4 97.4 92.0 102.1 101.6 103.7 2008 95.1 99.5 94.6 90.4 103.4 93.5 2009 96.5 97.7 94.3 85.2 97.8 83.3 2010 99.4 91.5 90.9 93.6 95.7 89.6 2011 99.1 88.4 87.6 92.4 92.3 85.3 2012 99.7 88.4 88.2 92.8 88.8 82.4 2013 98.6 89.3 88.0 93.9 86.2 81.0 2014 98.9 93.2 92.2 96.3 89.8 86.6 2015.0.0.0.0.0.0 2016 97.4 110.5 107.6 92.9 113.3 105.3 2017 106.8 117.1 125.1 95.4.1 114.5 Quarters 1 2 2018 106.1 126.0 133.6 90.9 130.0 118.1 In parallel, prices for new increased year-on-year by 10.4 in the second quarter of 2018. This represents a 1.8 increase over the first period of the year. To observe price s in the new build market, for the time being, we have data only about 600 new-home transactions for the second quarter, so we will analyse only the market of existing in more detail. Figure 1 Price trend in the housing market pure price (2015=) 140 130 110 90 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Second-hand Distribution of existing home transactions The price distribution of second-hand has seen a striking in recent years. Ten years ago, more than two-thirds of were sold for between HUF 4 and 14 million, in 2018 this rate was only 45. Previously, sales above HUF 20 million accounted for one tenth of all existing home sales, now this figure stands at 23. After 2008, the unimodal distribution of home transactions peaking at HUF 8 million gradually disappeared and it was replaced by a more even distribution with a longer slope. This transformation has been ongoing over the past ten years and was not affected by the boom after 2015. Deepening regional differences are behind this phenomenon: as it can be seen on the chart the share of both low and high value sales has grown while the share of modus range transactions declined. Figure 2 Price distribution of existing home transactions 12 10 8 Inflation had a significant impact on price developments: the 30 base index of second-hand is 24 in real terms, while the real value of new has risen by 20 since 2015. In the first two quarters of 2018, national housing prices, in inflation adjusted terms, were close to the pre-crisis level of 2008. Real-term home prices were 0.4 lower in the new build market and 1.6 lower in the second-hand market compared to ten years ago. Quarterly pure price index In quarter 2 2018, the level of second-hand housing prices grew year-onyear by 10.6 and quarter-on-quarter by 2.3. 6 4 2 0 0 5 10 15 20 25 30 35 40 45 50 55 million HUF 2008 Quarters 1 2 2018
Statistical reflections 3 Regional characteristics of the market of existing In the first two quarters of 2018, housing markets in larger cities continued to see a sustained buyer interest in new under construction. Consequently, the share of smaller settlements increased in existing home sales. In the first two quarters of the year, Budapest and the county seats accounted for 46 of all home sales and smaller settlements for 54. During that period, in Budapest one fifth, in the county seats one third of all sold were in prefabricated housing estates. In these settlements the share of detached houses is low (6.2 and 19 respectively), while in smaller towns the majority of sold are detached houses (56). Home sales in villages, which were almost without exception house sales, accounted for about one fourth of all sales in Hungary. Figure 3 Distribution of existing home sales by settlement and building type in the first two quarters of 2018 Villages 22.6 Towns, multi dwelling 13.5 Towns, detached houses 16.9 Budapest, detached houses 1.5 Budapest, non-prefabricated multi-dwelling 18.3 Budapest, prefabricated 5.2 detached houses 4.1 non-prefabricated multi-dwelling 9.8 prefabricated 8.1 In the first two quarters of 2018, second-hand sold in Budapest were on average HUF 26.6 million, HUF 2.2 million more than in 2017. Meanwhile, the average price increased by HUF 1.3 million to HUF 14.3 million in county seats. Average home prices by square metre were HUF 477 thousand in Budapest and HUF 233 thousand in county seats, up by 14 and 13 respectively compared to the previous year. In the period under review, average home prices slightly fell to HUF 11.1 million in towns. Average prices fell, along with a rise in specific prices (from HUF 151 thousand to HUF 153 thousand), due to the smaller size of marketed. In contrast, villages have seen a fall in both and square metre prices since 2017. Average square metre prices stood at HUF 77 thousand in villages and HUF 58 thousand in villages outside agglomerations, amounting to one third and one fourth respectively of the national average. The divergence between the price level of smaller settlements and the national average prevailing since the start of observations continued in recent years. The price level remained steadily close to the national average in county seats and grew significantly above average in Budapest, where home prices, as a percentage of the national average, grew from 160 between 2008 and 2015 to more than 200 in the first half of 2018. Figure 4 Deviation of existing home prices from the national average by type of settlement 60 40 20 0 20 40 60 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Q 1 2 2018 Budapest County seats Towns Villages Diverging local processes can be detected behind the inflation adjusted national home price index, which is close to its pre-crisis level. Between 2008 and the first half of 2018, inflation adjusted existing home prices grew in Budapest and the county seats of more developed regions, but fell in the smaller settlements of these regions. Western Transdanubia is the only exception, where smaller settlements saw a modest real price rise of 3. All settlements, including the county seats, were below the inflation adjusted price level of 2008 in Southern Transdanubia and Northern Hungary and the same is true for the Pest region's housing market. Figure 5 Regional deviation of real home prices from the level of 2008 in the first two quarters of 2018 Budapest Pest region Central Transdanubia, county seats Western Transdanubia, county seats Southern Transdanubia, county seats Northern Hungary, county seats Northern Great Plain, county seats Southern Great Plain, county seats 30 20 10 0 10 20 30 40 International data The housing market index of Eurostat shows the aggregated price trends of second-hand and new flats. In the second quarter of 2018, the overall housing market index of the EU Member States accounted for 113.0 of
4 Statistical reflections the 2015 base value. Within the Eurozone, the housing price index was below the EU average with a value of 111.6. The lower price rise of the euro area is also reflected in deflated housing price indices. According to Eurostat's aggregate, the real housing price index calculated for the EU as a whole has exceeded the corresponding Eurozone value since 2015. 2 Figure 6 Combined housing price index in the European Union and Hungary (2015=) 140 130 110 90 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 EU average Euro area Hungary In the second quarter of 2018, the Hungarian value of the house price index aggregated according to the Eurostat methodology was 131.2. At that time, Iceland had the highest base index exceeding 140 followed by Hungary, Ireland (with over 130) and Latvia. The strong and continuous rise in home prices is in line with Scandinavian trends in Iceland and followed sharp post-2008 downturns in the other above mentioned states. Of the neighbouring countries, a price rise of 0.9 was recorded in Austria, 3.1 in Romania, 1.2 in Slovakia and 4.2 in Slovenia in the first two quarters of 2018. This latter was the highest quarterly price increase among EU Member States during this period. Methodological notes Starting from the publication of data from the first quarter of 2017, we use the 2015 base in accordance with Eurostat's data dissemination practices. We also implemented several methodological s simultaneously with the base. The most important of these was the program that was developed for linking administrative data for statistical purposes (ESS VIP ADMIN). Through the development, information on housing market transactions was supplemented with information available in the statistical registers of HCSO and relevant for the housing market processes. This gives us more accurate information on the type of dwellings sold on the market and their immediate environment. The calculation models of the housing market price index starting from the 2015 base are based on these new and more accurate information. The revised 2015 and 2016 price index does not override previously published price trends, but its concrete values have d. The cumulative values of the published housing market indices are also included in the housing price indices of Eurostat. 3 Due to the harmonised methodology, these data are fully comparable across the European countries as well as with the aggregated indices of the EU member states. The source of price observations is the stamp duty database of the National Tax and Customs Administration of Hungary (NAV), from where the anonymized stamp duty data are taken over on a monthly basis directly after their receipt. All home sales concluded by private individuals are subject to this data transfer including home sale prices and the most important characteristics. At present, there are data series of uniform structure comparable in every respect from 2007, which make it possible to analyse s in home prices in a more detailed and exact way. From 2016 onwards, data received include the nationality and birth year of the given home buyer. The gradually completed data base still allows only preliminary information on the processes of 2017. Our compilation's data for the period prior to 2017 are final. As a result of missing data, 1 per cent of all cases were excluded from calculations. In those cases, where there were no data on the floor area of the given dwelling, but all other data were available, the floor area was estimated using the home price and its other characteristics, then we used this estimated value to further calculate. Following this, a log linear regression model was used to analyse the data. Major data used in this model: floor area of the given dwelling, character of the building, specific geographical, administrative and income indicators of the given settlement (or district in Budapest) and the characteristics of the immediate neighborhood zone and the residential building. New dwellings were separated by NAV based on benefits used to buy a new dwelling. Based on the findings of the first model estimation a further 5 per cent of dwellings were filtered out as outliers from further index calculations. After the exclusion of outliers, based on repeated model estimations, s were broken down by the composition effect and pure s. As a result of the log linear method the released price indices resulted from the geometrical average of the given prices in all cases. However, the average prices of this publication are always arithmetical averages, which were calculated after the completion of the outlier filtering. The Eurostat s aggregated housing price index is the weighted average of the price indices of second-hand and new presented in our publication. The weights are the aggregate values of home sales realized in the previous year. The most recent Hungarian data published by Eurostat are always preliminary results based on the data recorded by the end of the second month following the reference period, while to this present publication we have used data received for the complete quarter following the reference period. 2 http://ec.europa.eu/eurostat/statistics-explained/index.php/housing_price_statistics_-_house_price_index. 3 https://ec.europa.eu/eurostat/web/housing-price-statistics/methodology/.
Statistical reflections 5 Quarterly aggregate housing market price index in some European countries (2015=.0) Denomination 2017 2018 1 2 3 4 1 2 Austria 111.5 114.2 114.8 116.5 117.5 118.5 Belgium 105.0 104.7 108.2 107.8 107.7 108.7 Bulgaria 112.7 115.4 117.6 119.5.7 124.0 Croatia.9 104.6 105.1 108.5 109.4 109.3 Cyprus 99.6 102.7 102.5 105.2 103.3 104.0 Czech Republic 116.2 119.1 121.2 122.3 125.0 128.0 Denmark 106.6 110.5 111.0 109.1 114.0 115.8 Estonia 108.1 108.4 112.1 113.5 115.2 116.4 Finland 101.6 102.7 102.2 101.8 101.7 103.4 France 102.4 103.4 105.7 105.8 105.4 106.3 Germany 107.4 109.5 111.0 112.8 113.6 115.1 Hungary 115.2 118.7 122.3 124.3 128.3 131.2 Iceland 122.1 130.1 135.7 136.9 138.8 140.5 Ireland 113.4 115.9 122.2 125.3 127.3 130.5 Italy 98.9 99.1 98.6 98.7 98.6 99.4 Latvia 113.1 119.5 119.7 119.9 126.0 129.9 Lithuania 111.1 114.5 116.7 116.9 119.8 123.0 Luxembourg 109.8 112.2 112.3 113.6 116.8 117.9 Malta 105.2 108.7 112.8 115.7 111.5 115.1 Netherlands 109.7 111.7 115.7 117.9 119.9 121.8 Norway 114.9 115.5 112.4 112.2 113.6 116.1 Poland 103.3 105.4 106.5 107.9 109.5 112.0 Portugal 111.9 115.5 119.6 121.1 125.6 128.5 Romania 109.0 114.3 112.4 113.8 116.1 119.7 Slovakia 107.1 113.1 115.6 116.2 119.6 121.0 Slovenia 106.9 111.4 111.9 116.1 121.2 126.3 Spain 108.2 110.4 112.4 113.4 115.0 117.9 Sweden 114.0 116.2 117.8 114.4 113.5 114.1 United Kingdom 108.8 111.3 113.8 114.2 113.5 114.9 EU average 106.4 108.2 109.9 110.7 111.4 113.0 Euro area 105.3 106.8 108.5 109.4 110.1 111.6 quarter Table 3 () Further information, data (links): Tables Contact details: kommunikacio@ksh.hu Contact us! Telephone: +36 1 345 6789 www.ksh.hu HUNGARIAN CENTRAL STATISTICAL OFFICE, 2018 All rights concerning the layout, graphics and design work of this publication are reserved for HCSO. Any kind of reproduction of them has to be approved by HCSO. Any secondary publication is allowed only by the indication of source.