We explore Lithuanian credit register data and two bank closures to provide a novel estimate of firms’ bank-switching costs and a novel identification of the hold-up problem. We show that when a distressed bank’s closure ...