Working Paper: CEPR ID: DP15875
Authors: Stig Møller; Thomas Pedersen; Erik Christian Montes Schütte; Allan Timmermann
Abstract: We develop a new housing seach index (HSI) extracted from online search activity on a limited set of keywords related to the house buying process. We show that HSI has strong predictive power for subsequent changes in house prices, both in-sample and out-of-sample, and after controlling for the effect of commonly used predictors. Compared to the stock market, online search has much stronger predictive power over house prices and its effect also lasts longer. Variation in housing search is a particularly strong predictor of subsequent price changes in markets with inelastic housing supply and high speculation.
Keywords: Internet search; Housing markets; Housing demand; Forecasting; Inelastic housing supply
JEL Codes: C10; E17; G10; R3
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Housing Search Index (HSI) (R31) | Housing Demand (R21) |
| Housing Search Index (HSI) (R31) | Future House Prices (R31) |
| Housing Search Index (HSI) (R31) | Changes in House Prices (R31) |
| Changes in HSI (C43) | Changes in House Prices (R31) |