This paper studies the role of local banking systems in the propagation of ECB Quantitative Easing (QE) programs. I firstly document that local deposit markets are fragmented across country lines, but the assets held by banks backing the deposits are in more integrated markets. I then consider a multi-country New Keynesian model with heterogeneous banking sectors but common monetary policy. All banks can access collateral from the same union-wide asset market, using them to back liquid deposit liabilities that are issued locally. QE has real effects if it increases the quantity or quality of collateral available to the banking sector. I find that QE has a powerful effect across the currency union, raising output and inflation by 62bps and 60bps, respectively. The pass-through is very similar across countries, despite fragmented deposit markets, as all banks face the same reduction in the cost of collateral from the union-wide asset market. The overall impact increases significantly if the beginning of QE coincides with adjusting the policy rate rule to be a weaker counteracting force by making it less responsive to inflation.