The Demise of Double Liability as an Optimal Contract for Large-Bank Stockholders

This paper tests the optimal-contracting hypothesis, drawing upon data from a natural experiment that ended during the Great Depression. The subjects of our experiment are bank stockholders. The experimental manipulation concerns the imposition of state or federal restrictions on the contracts they write with bank creditors. We contrast stockholders that were subject to the now-conventional privilege of limited liability with stockholders that faced an additional liability in liquidation tied to the par value of the bank's capital. Our tests show that optimal contracting theory can provide an explanation both for the long survival of extended-liability rules in banking and for why they were abandoned in the 1930s.