On several occasions, the two indices have shown different developments in the housing market. Why?
The article was originally published on the website of Statistics Norway.
Real estate Norway's house price index has given 5 percentage points higher price inflation than Statistics Norway's house price index after the fall in oil prices in 2014. Large regional price growth differences over the last three to four years, especially between the West of Norway and the Oslo area, explain the gap between the indices.
The housing market plays an important role in the Norwegian economy. The rise in house prices has, among other things, a major impact on the development of household wealth, and thus the development in households' purchases of goods and services. Such purchases account for more than half of mainland GDP. House price trends are also important for housing investment developments. In addition, the trend in house prices affects general inflation and financial stability, and could thus have an impact on Norges Bank's interest rate setting.
Real Estate Norway's monthly house price index and Statistics Norway's quarterly house price index measure in each way the price development of second-hand housing nationwide. They are both important statistics that are used and mentioned regularly by several forecasting communities in Norway. In this article, we take a closer look at why these two indices are lagging and give a rather different picture of house price growth after the fall in oil prices in 2014.
Similarities and differences between the indices
The underlying price data material is the same for both indices, and this covers about 70 per cent of the turnover in the housing market. However, the indices are different in how the price trends for different types of housing by geographical region are weighted together. While Real Estate Norway's house price index is based on revenue weights, that is, the value of all housing transactions, Statistics Norway's house price index is based on inventory weights, that is, the value of all housing. An important assumption behind the weighting of Statistics Norway's index is that the price trend for housing that is not sold is the same as the price trend for housing that is sold.
If the purpose of the index is to measure the price trend for homes that are traded and purchased by households, it will be relevant to use turnover weights, such as in the Real Estate Norway´s Index. If, on the other hand, the purpose of the index is to measure the price trend for households' total housing stock, it will be relevant to use inventory weights, such as in Statistics Norway's index.
Norges Bank uses Real Estate Norway's house price index as a starting point for housing price forecasts in the Monetary Policy Report. Because we want a measure of the development in housing wealth, our forecasts in the cyclical trends are based on Statistics Norway's house price index. Since housing can be used as collateral for household loans, the development of housing wealth will affect households' ability to finance loans for consumption. Therefore, the development in housing wealth in the macroeconomic model KVARTS is based on Statistics Norway's house price index.
Sprawling development after the fall in oil prices
The two indices are in the first quarter of 2003–2. In the quarter of 2014, a very similar development pattern was shown and thus for a long time gave a nearly identical picture of the rise in house prices in Norway. Chart 1 shows that the indexes after the second quarter of 2014, that is after the fall in oil prices hit the Norwegian economy, have seen a striking development and given a rather different picture until the top in the second quarter of 2017. Real Estate Norway's index gives about 5 percentage points stronger house price inflation than Statistics Norway's index during this period. At the same time, Real Estate Norway's index shows a fall in house prices of about 0.2 per cent in the period Q4 2016–4. Q2 2017, while Statistics Norway's index shows growth of around 0.7 per cent in the same period. This makes a difference of almost 1 percentage point.
To explain this gap, we have calculated an alternative index of turnover weights based on price data from StatBankbanken. Chart 1 shows that the alternative index of turnover weights is almost coincident with the Real Estate Norway Index throughout the first quarter of 2011–4. quarter of 2017. The remaining difference between the indices can be explained by some differences in both method choice and regional classification. In addition, the weights are calculated at different times and one is a monthly index and the other a quarterly index.
The difference between Real Estate Norway's house price index and Statistics Norway's index is therefore mainly related to different weight sets. We can thus compare the alternative index with Statistics Norway's index of inventory weights to find which regions contribute most to the gap between the indices over the last three to four years. In the text box we give a more detailed account of the calculations that form the basis of this comparison.
Large regional price growth differences
Let us first explain why different weight sets can give rise to striking developments between the property price indices of Real Estate Norway and Statistics Norway. In general, the turnover of housing is greater in big cities than what the stock of housing would indicate. This is because apartments, which sell more than detached houses, are the most common type of housing in big cities. Therefore, when we use turnover weights instead of inventory weights, big cities will have a higher weight in the house price index for the country as a whole than more rural regions with less turnover.
If housing prices in a region rise, this could increase the turnover of housing. An increase in regional housing prices can therefore result in higher house price growth for the country as a whole when we use turnover weights rather than inventory weights. The opposite will usually apply in case of a fall in house prices. These effects can be large if there are large differences in house price trends between regions.
Figure 2 shows that regional price growth differences have been significant after the fall in oil prices in 2014. Oslo and Bærum experienced by far the largest price growth with a full 40 per cent in the second quarter of 2014–2. quarter 2017, followed by Akershus outside Bærum with close to 35 per cent in the same period. Stavanger and Agder and Rogaland outside Stavanger, areas that were particularly hard hit by the economic downturn that followed the fall in oil prices, on the other hand, experienced a fall in prices of just over 5 per cent and almost zero growth in house prices in the second quarter of 2014–2. quarter 2017.
In comparison, house price growth for the country as a whole was about 20 percent in the same period. Oslo and Bærum also experienced the largest fall in prices during the second half of last year, at around 6.5 per cent.
Oslo and Bærum mean the most
Figure 3 shows that the difference between the alternative index and Statistics Norway's index of holding weights in average quarterly growth in the period Q2 2014–2. Q2 2017 of just over 0.2 percentage points is mainly due to the strong price increase in Oslo and Bærum, which has weighted more than 30 percent of total housing sales. The strong fall in prices in Stavanger in the wake of the fall in oil prices, unlike Oslo and Bærum, has made a negative contribution to the difference in average quarterly growth. However, this contribution is small due to relatively modest sales weight. Due to higher inventory weights than turnover weights, Hedmark and Oppland, Western Norway outside Bergen and Northern Norway also made a negative contribution of a total of 0.1 percentage points in the period Q2 2014–2. quarter 2017.
Similarly, but with the opposite sign, the difference in average quarterly growth for the third and fourth quarters of 2017 of barely –0.2 percentage points can largely be attributed to price developments in the same regions. As a result of the fall in prices in Oslo and Bærum, this region has made a negative contribution of almost 0.5 percentage point to the difference in average quarterly growth over the last six months of last year. At the same time, the fall in Hedmark and Oppland, Western Norway outside Bergen and Northern Norway in addition to Agder and Rogaland outside Stavanger, again due to higher inventory weights than turnover weights, made a positive contribution of 0.4 percentage points in the same period.
The difference in average quarterly growth between the two indices doubles to 0.4 percentage points if the period from the second quarter of 2014 ends in the first quarter and not in the second quarter of 2017. Likewise, the difference in average quarterly growth is close to –0.5 percentage points when the last half year in last year is extended by the second quarter. This doubling and, well, the difference in average quarterly growth is mainly due to the fact that Oslo and Bærum, as the only region, experienced a fall in prices in the second quarter of 2017.
What should we keep in mind?
Almost the whole gap between the property price indices of Real Estate Norway and Statistics Norway since the second quarter of 2014 can be explained by using turnover weights and inventory weights respectively. With large regional differences in price developments, as we have seen in recent years, different sets of weights can give rise to striking developments in the two indices. This is important to keep in mind when considering general house price developments in Norway or making house price forecasts based on one or the other index.