Identifying empty creditors with a shock and micro-data

Firms with credit-default swaps (CDS) traded on their debt may face "empty creditors" as hedged creditors have less incentive to participate in firm restructuring. We test for the existence of empty creditors by employing an exogenous change to the bankruptcy code in Germany that effectively removes their potential impact on CDS firms. Using a unique dataset on bank-firm CDS net notional and credit exposures we find that the probability of default for CDS firms drops when the effect of empty creditors is removed. This effect increases in the average CDS hedge position of a firm's creditors and in the concentration of the firm's debt. Firms with longer credit relationships, with higher average collateral ratios of their debt, and financially safer firms are less affected by empty creditors. Banks that are not capital constrained and that are liquidity constrained embed the empty creditor effect into their probability of default estimates of affected firms to a larger extent. So do banks that monitor their creditors less and that earn a smaller portion of their income from interest activities.