We estimate the role of monetary policy, net capital inflow and credit supply shocks for house prices, residential investment and durable consumption. These fundamental shocks account for three leading hypotheses about the causes of the recent housing bubble in the US: loose monetary policy, a `global saving glut' and loose credit standards for mortgage borrowers. Shocks in credit standards are stemming from variations in the assets of security broker-dealers, as these channel the securitized credit supply to the real economy. We show that the `global saving glut' hypothesis does not assert housing activity and consumption. Indeed, monetary policy and credit supply shocks have significant effects on housing variables and durable consumption, explaining a similar share of the variations in housing variables. The credit supply shock contributes three times more to the variation in durable consumption due to the effect on subprime mortgage borrowers, which is not captured by the monetary policy shock. A significant negative response of current account deficit to monetary policy shocks and a positive response to credit supply shocks implies that the current account deficit might be driven by domestic factors rather than by foreign savings. The results from the counterfactual experiment indicate that monetary policy shocks are transmitted to durable consumption to a large extent indirectly through housing collateral and employment effects. In contrast, credit supply shocks have a direct effect on durable consumption in the short run.