This paper investigates whether bank integration measured by cross-border bank flows can capture the co-movements across housing markets in developed countries by using a spatial dynamic panel model. The transmission can occur through a global banking channel in which global banks intermediate wholesale funding to local banks. Changes in financial conditions are passed across borders through the banks' balance-sheet exposure to credit, currency, maturity, and funding risks resulting in house price spillovers. While controlling for country-level and global factors, we find significant co-movement across housing markets of countries with proportionally high bank integration. Bank integration can better capture house price comovements than other measures of economic integration. Once we account for bank exposure, other spatial linkages traditionally used to account for return co-movements across region -such as trade, foreign direct investment, portfolio investment, geographic proximity, etc. -become insignificant. Moreover, we find that the co-movement across housing markets decreases for countries with less developed mortgage markets characterized by fixed mortgage rate contracts, low limits of loan-to-value ratios and no mortgage equity withdrawal. JEL Classifications: C23, G15, F36, R3 Abstract This paper investigates whether bank integration measured by cross-border bank flows can capture the co-movements across housing markets in developed countries by using a spatial dynamic panel model. The transmission can occur through a global banking channel in which global banks intermediate wholesale funding to local banks. Changes in financial conditions are passed across borders through the banks' balance-sheet exposure to credit, currency, maturity, and funding risks resulting in house price spillovers. While controlling for country-level and global factors, we find significant co-movement across housing markets of countries with proportionally high bank integration. Bank integration can better capture house price comovements than other measures of economic integration. Once we account for bank exposure, other spatial linkages traditionally used to account for return co-movements across region -such as trade, foreign direct investment, portfolio investment, geographic proximity, etc. -become insignificant. Moreover, we find that the co-movement across housing markets decreases for countries with less developed mortgage markets characterized by fixed mortgage rate contracts, low limits of loan-to-value ratios and no mortgage equity withdrawal. JEL Classifications: C23, G15, F36, R3