Statistical processing

Source data

Administrative registers from the data distributor (Datafordeleren), which, among other things, collect information from the electronic deed in the electronic land registration system.

Frequency of data collection

Monthly.

Data collection

Data for these statistics are collected electronically from administrative registers on the Data distributor (Datafordeleren), where a large number of public basic data is available.

Data validation

Data from the Data Distributor is initially checked for errors by Geodatastyrelsen (responsible for the owner register) and the Tax Authorities (responsible for the appraisals). When Statistics Denmark receives data, further checks are conducted, for example, the validity of dates and geographical information.

Data compilation

In the statistics for sales of real estate property, the individual real estate property transactions are divided according to category of real estate property, region, type of transfer and period. Then aggregated figures are calculated for number of sales, average prices and the ratio between purchase price and appraisal value (spar-value). Finally, the price index is calculated.

The number of sales in the price calculation is calculated as all sales except extreme observations and outliers. The sales that are sorted out are, for example, partial sales, sales of several properties, services in addition to cash purchase price (inheritance/gift), sales where the seller is a public authority, sales where sales information is missing or sales with extreme price. An extreme price is defined as the spar-value (ratio of price and appraisal value) is less than 0,4 or higher than 3. Price index, average prices and spar-value is calculated on the transactions that are included in the price calculation.

The average price per property is calculated as a simple arithmetic mean of the properties included in the price calculation. Purchase sum per hectage/square meter is only available for resp. agriculture (hectares) and land (square meters). One hectare corresponds to 100 x 100 square meters, equal to 10.000 square meters.

The purchase sum in percent of taxable value, the so-called the spar-value, is used to calculate the price index. The average spar-value is calculated as the average price per property divided by the average appraisal value. In this way, the appraisal value is used as the weight basis. This also means that a property with a high appraisal value has a higher weight than a property with a lower value (value-weighted spar-value).

The price index is calculated on the basis of the development in the spar-value (using the so-called value-weighted SPAR method), which seeks to take into account differences in the quality of the sold properties. The calculation is described in more detail in this note Beregning af prisindeks for ejendomssalg. This note is only available in Danish.

The estimated number of sales is use to calculate the number of registered sales for the preliminary figures. The number of sales is estimated in the preliminary figures because the statistics is calculated before all sales have been registered. The variable includes all sales, inclusive extreme observations and outliers. There is no estimation (enumeration) of the final figures.

Statistics Denmark have made analyzes that shows that a real property transaction is typically registered close to the takeover data, and than more expensive properties on average have a longer takeover period than cheaper properties. This means that more expensive properties are included later in the database than cheaper properties, as the transactions comes from the electronic registration system of sold properties. The analyzes also shows that the purchase sum in percent of taxable value (spar-value) is higher on the more expensive properties than on the cheap properties. In addition, the expensive properties weight most in the calculation of the spar-value. In order to take into account the potential bias in the preliminary price indices, the observed skewness in the registration pattern is corrected. The method is described in more detail in the note Revisioner af ejendomssalgsstatistikken, only available in Danish.

A first-time buyer is defined as a person who has not been a dwelling owner for at least the previous three years. The purpose of the three-year limit is to ensure that the group of first-time buyers has not been on the housing market recently. You could also define a first-time buyer quite literally as someone who has never owned a home. However, it is difficult to calculate how many middle-aged and older dwelling buyers have never owned a dwelling. In the calculation of first-time buyers, the person is counted several times in the statistics if the same person has acquired several owner-occupied dwelling the same year. Cooperative housing does not count here as owner-occupied housing. If you have owned a cooperative dwelling and bought an owner-occupied dwelling during the period, you are counted as a first-time buyer. Only trades in ordinary free trade are included. If there is at least one non-first-time buyer in the transaction, the transaction is categorized as "non-first-time buyers" and none of the buyers count as first-time buyers, even if one of the buyers has not been a dwelling owner in the three previous years. Only people who enter into housing transactions, where the entire group of buyers are first-time buyers, count as first-time buyers.

Adjustment

Seasonal adjustments are made to both price index and the number of sales in the StatBank table EJ12. Seasonal adjustment has the challenge that the latest figures are the most difficult to seasonally correct, but at the same time it is these figures that are of greatest interest. A general assumption for seasonally correcting is that the seasonal pattern is stable over time. Regarding housing prices this concerns stability in multiplicative seasonal factors. In this connection , one must be aware of that the seasonal correction of the latest observations is partly based on projection of the observed series, and sudden changes in the course of the series increase the uncertainty and the risk of misrepresentation. Especially when the latest observations deviate from the expected course. Major revisions of published figures may occur when the forecast values of the seasonal adjustment are replaced by observed values. Statistics Denmark has assessed that it is too uncertain to seasonally adjust the monthly house prices, therefore only the quarterly house prices are seasonally adjusted. Read more about this subject in the analysis of Season in the housing market, only available in Danish.