2 Necessary data and sources

For the production of the residential property price index (IMPI), the FSO is only interested in a specific part of Switzerland’s property stock, namely changes in the prices of private residential properties (single-family houses and condominiums). The survey only takes into account residential properties sold at market prices in the respective time period.

The fact that properties are extremely heterogeneous makes comparing real estate prices difficult. No two properties are identical. The comparison of prices is only meaningful if we consider the different property characteristics. In addition to comparing the prices paid for residential property in the survey period and the previous period, the characteristics of the properties sold in the current and previous period must also be compared (section 3.2).

In order to track price changes over time, the FSO requires a whole set of information: primarily data on which properties were sold at which time. It is also necessary to ensure that we are looking at the right property type. The purchase price must also be a market price. Finally, the FSO requires certain information on the characteristics of the single-family houses and condominiums sold (figure G1) Precise information on the characteristics of the transaction, address, use, structure and location variables (format, categories) can be found on the website: https://github.com/bfs-preis/impi/wiki .

2.1 Property type

The IMPI measures price changes in single-family houses and condominiums. Single-family houses include all houses with one apartment and single-family houses with a granny flat. Condominiums generally refer to individually-owned apartments in an apartment block. Separate hobby rooms acquired as part of the dwelling and parking spaces in an underground garage belonging to a multi-family house are counted respectively as extra rooms and parking spaces (figure G2). It is rare for a two-family or three-family house to be included in the sample. This can occur if the owner uses one of the apartments themself. Strictly speaking, these are multi-family houses. They cannot be ruled out in these exceptional cases, however, as they are financed by the reporting mortgage-lending institutions according to the same criteria as owner-occupied property. Unlike with investment properties, in such cases the mortgage-lending institutions do not record the number of apartments in their databases, but instead treat the properties as if they comprise one single apartment. ­According to the data suppliers, the number of such cases is very small. With its statistical processing of extreme values, the FSO can exclude any distortion of the data.

2.2 Transaction date

The purchase of a property can take several months: from the listing of the property to searching for an institute to finance the transaction, estimating the market value and negotiating the price, to entry of the new owner in the land register (figure G3). The FSO defines the date on which the risks and benefits of a property are transferred to the new owner as the transaction date. This transfer of ownership coincides with entry in the land register. The date of the purchase agreement and the transaction date do not need to be identical.

The data suppliers for the IMPI are mortgage-lending institutions. The closest date in banks’ management systems to the transfer of ownership in the land register is the date of credit clearance. On this date the mortgage bond, which backs the loan from the bank, is also recorded in the land register. By searching in its database for new loans by payment date, the mortgage-lending institution ensures that the purchase or sale actually took place in the period in question. This is the same distinction that the reporting mortgage-lending institution also makes for the Swiss National Bank’s survey on new mortgages ­(HYPO-B). Only data on new mortgages for the purchase of residential property are supplied to the FSO. Transactions involving remortgaging are excluded. It not only made a mortgage offer, along with competitor institutions, but it provided the financing, usually as the sole institution. In this way, duplicated values are excluded from the supplied data.

2.3 Price

The FSO records the purchase price, as entered in the purchase agreement. To measure price changes in residential property, the FSO only records purchase prices on the open market. Market prices are the result of what are known as ‘arm’s length transactions’. This refers to a transaction where the purchase price is negotiated between a buyer and seller, who are both acting in their own self-interest. Sales at preferential prices or between friends or family members (e.g. pre-inheritance and partial donations) are excluded from the residential property price index. In the course of a real estate transaction, the same property has different prices and estimated values: the seller and estate agent set the asking price. The mortgage-lending institution, from which the potential buyer has requested a mortgage offer, estimates the market value and sets a lending limit for the property. During the negotiations between the buyer and seller, the price can also change several times. The final purchase price is the one recorded in the notarised purchase agreement. For regulatory reasons, the mortgage bank must document the final purchase price. Additional valuations are carried out by the cantonal building insurer and the tax administration. They generally vary significantly from the current market price (figure G3).

2.4 Use

The FSO also includes in its survey information on whether single-family houses and condominiums are «owner-occupied» or «rented». However, not all data suppliers can provide the FSO with information on «owner-occupied/rented». The number of transactions is insufficient to calculate price changes in the two property segments separately. There is an interest on the part of the real estate sector and internationally (e.g. at Eurostat) to be able to observe whether the residential property market is split in two, into one segment for owner-occupied residential ­property and one market segment for single-family houses and condominiums that are acquired by private investors purely for investment purposes. The data on whether a property is used as a first or second home by the owner is available across the whole sample. The type of use of a residential property is relevant in all municipalities in which the second home percentage threshold of 20% is reached or exceeded. It is expected that such locations will have a two-tier property market.

2.5 Property characteristics and location

Only in very rare cases is the same property sold in consecutive quarters. Price changes therefore have to be derived from the sale of different single-family houses and condominiums. When comparing two transactions, we need to be able to distinguish whether a change in property price from one period to the next is due to a change in the quality of the sold properties, or if a price change has actually taken place. There is a price change if more or less was paid for the same property characteristics on the open market. In order to ensure that like is compared with like, the FSO also records the objective characteristics of the houses and apartments. This is information on the property type and size, building quality and age, condition and location. Even if their architecture is identical, two properties can at least be distinguished by their location. No two views are the same. Public transport connections, travel time to the nearest town centre, noise pollution and tax burden are also characteristics that are linked to a property’s location and for which buyers are willing to pay. By quantifying their location characteristics, the FSO ensures maximum comparability of residential properties.

2.6 Data sources

The FSO is legally obliged to minimise the burden on data suppliers. It obtains the data required for the IMPI from various sources and makes the greatest possible use of administrative data. The IMPI data suppliers are Switzerland’s largest mortgage-lending institutions, the Federal Register of Buildings and Dwellings, and a range of federal offices with publicly-accessible geographical information.

2.6.1 Mortgage-lending institutions

The FSO collects the data for the IMPI from Switzerland’s 26 largest mortgage-lending institutions (lending banks). Through the 26 banks, the FSO covers around 90% of the market, measured in terms of total mortgage volume . The ranking of the 26 largest mortgage-lending institutions (banks) was compiled on the basis of outstanding mortgage volume. The comparison is based on the figures published in their 2017 annual reports for the Swiss market. They are represented in all regions of Switzerland and can provide a large amount of high-quality data that are necessary for production of the property price index, as they report on which properties changed hands in a given time period. They are also able to draw a distinction between the market segments whose economic momentum the IMPI is designed to illustrate. For business purposes, they distinguish between owner-occupied property and investment property.  Owner-occupied property and investment property are evaluated by the banks using different methods and are subject to different financing rules. Mortgage market regulations require Swiss lending banks to enter the definitive purchase price of properties they finance in their systems. They further filter the information by residential property and arm’s length transaction. In order to check the market value of properties, they work with electronic valuation programs. The property details they feed into these tools are the same that interest the FSO.

2.6.2 Public data on location quality

The FSO supplements the data from the banks on properties sold with various location information. In order to establish its microlocation (precise address and neighbourhood) and macrolocation (canton, municipality), the FSO draws on publicly-accessible administrative data collections (textbox).

2.6.3 Land registers

Alongside the survey of mortgage institutes, the FSO conducts a second survey of cantonal land registers. Using the land registers, the FSO is the first to count the total number of property purchases per year. The FSO will calculate a regional weighting that is representative of the Swiss real estate market based on the number of transactions. The total volume of all real estate sold cannot be surveyed in the land registers for the time being. Up to now, purchase prices have only been included in the electronic land registers in a small number of cantons. Land register law also precludes a nationwide survey of prices from purchase agreements. Through the survey of banks, the FSO only records purchases that came about with the help of credit financing. The FSO assumes that this is by far the most common type of transaction. Its actual share of the total market should be revealed by the land register survey, which is scheduled for launch in 2022. It will build on the annual long-term preservation of land register data. The survey automatically filters out the data required for the statistics from the land register data sets transferred by the cantons to the Swiss Federal Archives. The additional workload for the land registers is therefore minimal.