Abstract

Christopher L. Colvin (Queen’s University Management School, Queen’s University Belfast), Abe de Jong (Rotterdam School of Management, Erasmus University) and Philip T. Fliers (Rotterdam School of Management, Erasmus University)

Why do some banks fail in financial crises while others survive? This paper answers this question by analysing the consequences of the Dutch financial crisis of the 1920s for 143 banks of which 37 banks failed. Banks’ choices in balance sheet composition, corporate governance practices and shareholder liability regimes were found to have a significant impact on their chances of experiencing distress. Banks bore a higher probability of failing if, on the eve of the crisis, they: were highly performing; were highly leveraged; had fewer interlocking directorates with non-banks; and concentrated their managerial interlocks with highly profitable banks. Banks which chose to adopt shareholder liability regimes with unpaid capital were more likely to experience distress, but could mitigate this risk by keeping higher portions of their equity unpaid. Receiver operating characteristic analysis shows that interlock characteristics in particular have a high predictive power.