A classroom experiment with bank equity, deposit insurance, and bailouts

ABSTRACT

In this classroom experiment, students see how low bank equity requirements can interact with deposit insurance to encourage excessive risk-taking. The experiment fills a niche Admati and Hellwig (2013) have noted: citizens in a democracy must understand why bank owners argue for low equity requirements and why society as a whole is better off with higher requirements. Instructors can run the experiment in principles of economics courses to introduce the topics of banking and financial crisis, or in advanced courses to promote discussion of financial reform. It takes about 45 minutes to run and debrief, and requires no computerization.

KEYWORDS:

• Bank capital
• classroom experiment
• deposit insurance
• financial crisis
• moral hazard

JEL CODES:

• A2
• E02
• G01
• G21
• G28
• G32
• L51

Acknowledgements

The author gratefully acknowledges Whitman College for its sabbatical support, and thanks Clay Callahan, Jeanette Weber, Joshua Foster, Brian Griffith, Gambhir Kunwar, and two anonymous referees for their helpful comments.

Notes

1. The author would be happy to e-mail a digital copy of the instructions and analysis questions to instructors who request them.

2. The loan has a 2/3 chance of paying out $110 and a 1/3 chance of paying out $50, so the expected payout is (2/3)($110) + (1/3)($50) = $90.

3. In this experiment, depositors are the only government-guaranteed bank creditors. Their artificially cheap funding is the sole source of the moral hazard problem. In a more general setting, the moral hazard problem could also arise from the government explicitly or implicitly providing a guarantee to other bank creditors, for instance, by deeming the bank too-big-to-fail.

4. This experiment uses 3 percent and 15 percent equity requirements because 3 percent was effectively the requirement in place before the 2008 global financial crisis, and 15 percent is in the 15 percent to 25 percent range advocated by Admati and Hellwig (2013) and Myerson (2014).

5. Expense is the main reason for having only two groups operate their banks. An instructor willing to spend more could alter the instructions to say that the instructor will randomly select, say, four groups to operate their banks. An instructor wanting to spend less could alter the instructions to have only one group operate their bank. Each extra group that operates a bank requires about three minutes of class time, for the random selection of the group, the drama of that group's die roll, a moment of attention for their joy or dismay in their success or failure, and finally the distribution of their dividends or collection of their lost equity.

6. Each project has a 1/3 chance of failing, so there is a 1/9 chance of both projects failing.

7. In my experience of running this experiment twice, all the students in Country A wanted to start a bank, and none in Country B wanted to start one. In each session, one project succeeded, and one failed.