Speculative and Wake-Up Call Contagion across International Housing Markets

Abstract

This paper considers the three international house price cycles in the period 1975-2013. Motivated by (i) the observed increased international synchronization of house prices during full up- and downturns of the aggregate international house price cycle of OECD countries, and (ii) the fact that the US house price seems to be the first-mover. The goal is to investigate whether there is evidence of international speculative and wake-up call contagion from the US housing market to the housing markets of a group of OECD countries. We follow Masson (1998)’s distinction of contagion as a residual effect after controlling for (inter)national fundamentals, and combine it with Rigobon (2000)’s notion of a significant increase in the extent of spill-overs during certain times. Our interpretations of speculative and wake-up call contagion are thus essentially narrow versions of shift contagion (Rigobon, 2002). The analysis is done through panel data in an Error-Correction Mechanism Model. Our approach is unique in several ways. First, we are the first to distinguish between speculative and wake-up-call contagion in the housing market. Second, we also assess contagion during periods of international house price increases and therefore are – to our knowledge – among the first to presume contagion to be a potential explanation for this synchronization. Preliminary results suggest that, (i) a US house price boom is contagious; (ii) the wake-up-call that induces the US real house price growth to correct from the peak to fundamentally justified rates, is contagious; (iii) speculation that induces the US real house price growth to undershoot its fundamentally justified rate, is contagious – and likely to follow up on a pre house price correction. The essence for authorities to consider real house price developments in the US and to create (contagion-preventive) policies for their domestic housing markets, becomes clear. Finally, we find that our model that consists of mainly macroeconomic variables is unable to explain the behaviour of German real house prices – which are also uninfluenced by developments in the US.