Scholarlink

<p>This study reexamines the impact of out-of-town (OOT) buyers on housing price dynamics by examining whether news sentiment can act as a fundamental driver of these market movements. Using housing news headlines from South Korea, we construct a News Sentiment Index (NSI) via a fine-tuned KR-FinBERT model and incorporate it into a Bayesian VAR framework alongside OOT transaction data. Empirical results show that when NSI is included, OOT transactions no longer Granger-cause housing prices. Indeed, incorporating NSI leads to a substantial reduction in OOT’s price contribution, whereas NSI shocks exhibit a remarkable, immediate impact on price formation. Importantly, this informational influence is heterogeneous between OOT buyers and locals; NSI replaces a substantial share of the own explanatory power of OOT transactions and accounts for nearly half of the influence of OOT activity on local transactions, while having little effect on the dynamics of local transactions. The main analyses suggest that NSI may help alleviate informational frictions faced by geographically distant participants, whereas the outlet-based stale–novelty decomposition indicates that news also conveys genuinely new information, thereby generating more persistent effects on valuations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Is the Out-of-Town Buyer Effect Overemphasized? Evidence from News Sentiment as an Information Proxy

  • 期刊论文

  • 发表日期：

    2026年6月3日

查看全文
  • Sangwon Seo,
  • Youngjae Hwang

摘要

This study reexamines the impact of out-of-town (OOT) buyers on housing price dynamics by examining whether news sentiment can act as a fundamental driver of these market movements. Using housing news headlines from South Korea, we construct a News Sentiment Index (NSI) via a fine-tuned KR-FinBERT model and incorporate it into a Bayesian VAR framework alongside OOT transaction data. Empirical results show that when NSI is included, OOT transactions no longer Granger-cause housing prices. Indeed, incorporating NSI leads to a substantial reduction in OOT’s price contribution, whereas NSI shocks exhibit a remarkable, immediate impact on price formation. Importantly, this informational influence is heterogeneous between OOT buyers and locals; NSI replaces a substantial share of the own explanatory power of OOT transactions and accounts for nearly half of the influence of OOT activity on local transactions, while having little effect on the dynamics of local transactions. The main analyses suggest that NSI may help alleviate informational frictions faced by geographically distant participants, whereas the outlet-based stale–novelty decomposition indicates that news also conveys genuinely new information, thereby generating more persistent effects on valuations.