Forecasting House Prices And Rents: Combining Dynamic Factor Models and Machine Learning

Farley Ishaak, Peng Liu, Egbert Hardeman and Hilde Remoy

ERES from European Real Estate Society (ERES)

Abstract: Recent literature has shown how dynamic factor models (DFM) can be used successfully to predict real estate price returns. In this paper, we take it a step further. In a two-step approach we estimate (1) a dynamic factor model over multiple markets to extract a few common trends, and (2) estimate a per-market Autoregressive Distributed Lag (ARDL) model including the dynamic factors, in a LASSO framework. In total we estimate 7 different variants (for example by also utilizing macroeconomic explanatory variables) of this model for rents and prices for a selection of Polish cities. Compared to a vanilla ARDL model, our LASSO-DFM augmented ARDL, reduces the prediction error by more than 60% on average. What is more, the prediction errors are relatively "stable." With this we mean that the size of the error is comparable over time and over markets, without any large outliers. This holds true even for forecasts over very long horizons.

Keywords: Autoregressive Distributed Lag; LASSO; Poland (search for similar items in EconPapers)
JEL-codes: R3 (search for similar items in EconPapers)
Date: 2024-01-01
New Economics Papers: this item is included in nep-big, nep-ets and nep-ure
References: Add references at CitEc
Citations:

Downloads: (external link)
https://eres.architexturez.net/doc/oai-eres-id-eres2024-207 (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:arz:wpaper:eres2024-207

Access Statistics for this paper

More papers in ERES from European Real Estate Society (ERES) Contact information at EDIRC.
Bibliographic data for series maintained by Architexturez Imprints ().