In this paper, we use a Global Vector Autoregression (GVAR) model to assess the spatio-temporal mechanism of house price spillovers, also known as “ripple effect”, among 93 Italian provincial housing markets, over the period $2004-2016$. In order to better capture the local housing market dynamics, we use data not only on house prices but also on transaction volumes. In particular, we focus on estimating, to what extent, exogenous shocks, interpreted as negative housing demand shocks, arising from 10 Italian regional capitals, impact on their house prices and sales and how these shocks spill over to neighbours housing markets. The negative housing market demand shock hitting the GVAR model is identified by using theory-driven sign restrictions. The spatio-temporal analysis carried through impulse response functions shows that there is evidence of a “ripple effect” mainly occurring through transaction volumes.

在本文中，我们使用全球向量自回归（GVAR）模型评估了93个意大利省级住房市场在此期间的房价溢出的时空机制，也称为“涟漪效应”。 $\mathrm{2004年}-\mathrm{2016年}$。为了更好地掌握当地房地产市场的动态，我们不仅使用房价数据，还使用交易量数据。特别是，我们着重于估计在多大程度上由10个意大利区域性首都引起的，被解释为负面的住房需求冲击的外部冲击，对它们的房价和销售的影响以及这些冲击如何扩散到邻国的住房市场。通过使用理论驱动的符号限制可以确定对GVAR模型造成的负面住房市场需求冲击。通过冲激响应函数进行的时空分析表明，有证据表明“涟漪效应”主要通过交易量发生。