Location

260-005, Owen G. Glenn Building

Start Date

12-15-2014

Description

This study adopts data mining methods to analyze the short- and long-term dynamic between news message content and property prices in Spain and the United States. We construct news sentiment indices based on various text mining methods which exhibit remarkable similarities to the respective property prices. Comparing dictionary-based and dynamic approaches, our results indicate that static methods produce the best estimators for investor sentiment in real estate markets. Using a Vector Error Correction Model framework, we analyze similarities between real estate markets both in the short- and long-run. The main finding of this study is a significant relationship between news messages and property prices in the long-run. Our results are stable, including a number of fundamental variables, and are underlined by forecast error variance decompositions and impulse response estimates, which additionally highlight the appropriateness of news sentiment as a crucial determinant for decision making.

Share

COinS
 
Dec 15th, 12:00 AM

Long- and Short-Term Impact of News Messages on House Prices: A Comparative Study of Spain and the United States

260-005, Owen G. Glenn Building

This study adopts data mining methods to analyze the short- and long-term dynamic between news message content and property prices in Spain and the United States. We construct news sentiment indices based on various text mining methods which exhibit remarkable similarities to the respective property prices. Comparing dictionary-based and dynamic approaches, our results indicate that static methods produce the best estimators for investor sentiment in real estate markets. Using a Vector Error Correction Model framework, we analyze similarities between real estate markets both in the short- and long-run. The main finding of this study is a significant relationship between news messages and property prices in the long-run. Our results are stable, including a number of fundamental variables, and are underlined by forecast error variance decompositions and impulse response estimates, which additionally highlight the appropriateness of news sentiment as a crucial determinant for decision making.