The bullish cycle of the real estate sector consolidated in 2017, as shown by more than 460,000 homes being sold in Spain, up by 14.6% on the previous year. Will house purchases continue to rise at such a considerable rate? According to Google search data, the outlook is promising.

Google Trends is a free tool from Google that provides information on the online searches carried out for different themes and geographical areas. Specifically, Google Trends provides monthly information on trends in the popularity of search terms chosen by users (such as «buy flat»). It can also provide data based on a certain timeframe, country or search category (for example, «real estate market»).1

A statistical analysis of the data trends provided by Google shows that the popularity of internet searches using the term «comprar piso» (buy flat) in Spain is closely related to the actual house purchases made in the following quarters (see the first chart).2 This means that searches carried out today in Google can provide many clues as to the house purchases we can expect within one year.

The popularity of the search term «comprar piso» was lower but nevertheless still high in 2017. This suggests that house purchases will continue to grow in Spain over the coming quarters, albeit at a somewhat slower rate than last year. Growth could be around 8% in 2018 which would result in half a million transactions (see the second chart). A somewhat more detailed analysis based on the popularity of search terms such as «comprar piso en Barcelona» (buy flat in Barcelona) and «comprar piso en Madrid» (buy flat in Madrid) suggests that the growth in sales of these two zones will outperform Spain’s average and could reach 10%.

The use of Google search data opens up a new paradigm for the forecasting of real estate trends, a notable achievement given the sector’s cyclical nature. Being able to more accurately predict changes in house purchase trends before these actually happen could improve the capacity of housing supply to react to changes in demand. This is particularly important for sectors such as real estate where supply tends to react slowly. This results in a large amount of real estate for sale accumulating during periods of economic slowdown, aggravating the downward pressure on property prices.

1. This tool does not provide the total number of searches but the popularity of each search term based on a random sample of data that
is regularly altered. This could limit the scope of the analysis since the index’s relative popularity may change as a result of other events not related to the real estate sector. Historical data may also differ slightly depending on the date they are downloaded from the website as random sampling is used.

2. A time series regression analysis confirms that the relationship between the two series is statistically significant.