Coastal cities around the world have experienced large costs from major flooding events in recent years. Climate change is predicted to bring an increased likelihood of flooding due to sea level rise and a higher frequency of severe storms. In order to plan future development and adaptation, cities must know the magnitude of losses associated with these events, and how they can be reduced. Often losses are calculated by adding up insurance claims or surveying flood victims. However, this largely neglects the loss due to the interruption of economic activity caused by a flood. There have been some attempts to account for the output losses using input-output techniques, but these do not account for the mitigation achieved through flexible prices, changes in output composition, and factor substitution. Here, we use a computable general equilibrium (CGE) model to study how a local economy responds to a flood, focusing on the subsequent recovery/reconstruction. Initial damage is modelled as a shock to the capital stock and recovery requires rebuilding that stock. We apply the model to Metro Vancouver by considering a flood scenario causing total capital damage of $14.6 billion spread across five municipalities. Transportation and Warehousing are most severely impacted, followed by Manufacturing and Wholesale Trade. We find that the GDP loss relative to a scenario with no flood is 1.9% ($2.07B) in the first year after the flood, 1.7% ($1.97B) in the second year, 1.5% ($1.70B) in the fifth year and 1.1% ($1.42B) in the twentieth year. We also find that the losses tend to scale approximately linearly with the damage rate.