Predicting Housing Prices with Google Searches in Finland

Widgrén Joona

Abstract

This report examines whether Google search queries can be used to predict the present and the near future house prices in Finland. Compared to a simple benchmark model, Google searches improve the prediction of the present house price index by 7.5 % measured by mean absolute error. In addition, search queries improve the forecast of near future house prices. Predicting the present and near future house prices is relevant information to many agents, such as realtors and political decision makers.

Publication info

Series
ETLA Raportit - Reports 63
Date
14.12.2016
Keywords
Google Trends, Internet, Nowcasting, Forecasting, Housing market, Time series
ISSN
2323-2447, 2323-2455 (Pdf)
JEL
C1, C22, C43, C53, C82, E27
Pages
36
Language
Finnish