Experimental study of bank runs in the context of loss aversion

The goal of this study is to verify whether loss aversion, used as a motivator, has a different impact on tested subjects than a classical financial incentive usually used in the bank runs experiments. This novelty, based on the loss aversion literature in the context of bank runs, has never been examined before, according to our knowledge. In particular, it might significantly affect decision-making of subjects. In order to verify this hypothesis, we conduct a laboratory experiment which is aimed to shed more light on this issue. In fact, a detailed analysis of bank runs is crucial for further understanding of the banking crisis, with an opportunity to address important policy implications. The experiment and the modified pay off function inspired partially by Arifovic and Jiang (2015) are built upon fundamentals of the Diamond and Dybvig model. The main variable is the “coordination parameter” defined as the minimum fraction of depositors required to wait so that waiting entails a higher payoff than withdrawing. This experiment will take the form of a simultaneous game, and the coordination parameter will have three values fixed throughout the experiment (0.2; 0.7 and 0.9). In other words, we will conduct six sessions – two of them for each value of the coordination parameter with total number of subjects equal to 60. During the experiment, live subjects will be informed by fictitious messages, indicating the number of people from the last round who have withdrawn in the first or in the second period. Subjects will be motivated by facing the loss rather than by financial gains. At the beginning of the experiment, subjects will get the maximum payoff, which they can earn during the session. Consequently, while playing the game, subjects will be exposed to potential loss caused by their suboptimal choices. This should ensure that subjects might converge to more efficient outcomes than otherwise. This argument appears relevant in the light of a recently conducted field experiment by Fryer, Levitt, List, Sadoff, (2018), in which case the group of teachers facing the loss was performing significantly better than the group facing the incentives in the form of classical gains.