Can a DSGE model replicate the nancial crisis eects without assuming unprecedented and implausibly large shocks? Starting from the assumption that the subprime crisis triggered the nancial crisis, we introduce balance-sheet eects for housing market borrowers and for commercial banks in an otherwise standard DSGE model. Our crisis experiment is initiated by a shock to subprime lending risk, which is calibrated to match the observed increase in subprime delinquency rates. Due to contagion of prime borrowers and to the ensuing adverse effect on banks balance sheets, this apparently small shock is sufficient to trigger a decline in housing investment comparable to what was observed during the nancial crisis. The adverse effect of sub-primers risk on commercial banks' agency problem is a crucial driver of our results.