Artificial intelligence and rental markets: Revealing value potential in residential real estate through algorithm-driven hedonic models
Artificial intelligence (AI) and especially machine learning (ML) methods are gradually revolutionizing research and practice, not only in the real estate sector. In this study, we apply these so-called intelligent methods to investigate whether they are more suitable for evaluating residential rental markets than traditional hedonic methods.
Our results can be summarized as follows: ML methods perform better than traditional ones both in understanding and forecasting market rents. Furthermore, we find that investors think "linearly" and cannot realize value-creation potential in their portfolios, due to their traditional and thus imprecise market analyses to date. To the best of our knowledge, we are the first to provide a comprehensive framework for rental markets, based on these new methods, and offer valuable insights for investors, researchers and other market participants.