An exchange rate risk experiment with multiple currencies

Abstract

In this article, the authors describe a classroom experiment on exchange rates appropriate for undergraduate courses in macroeconomics, international economics, and money and banking. Student teams compete by managing virtual portfolios of six foreign currencies over a period of several weeks. Trading requires a few minutes in class. Students gain an understanding of currency movements, financial risk, and portfolio management. The experiment allows for a test of the efficient markets hypothesis. A single class spreadsheet is used to record the history of trades and portfolio balances, assist in formulation of team strategy, and provide the raw data for students to analyze their performance in an end-of-semester reflection paper. Sample outcomes from classes in money and banking and international economics are discussed.

Keywords:

• Classroom experiment
• cooperative learning
• exchange rates
• portfolio choice

JEL CODES:

• A22
• F31
• G11

Notes

1 See Jones, Sackley, and Watson (2017).

2 If risk diversification is not a major focus of the course, then a different experiment with only two currencies will suffice, narrowing the emphasis to predictions about exchange rate dynamics based on national politics and comparative economic trends.

3 The actual class time required can be reduced still further if, after class as students exit the room, one designated member from each team goes to the instructor and dictates the desired portfolio holdings. Alternatively, groups may turn an index card with their desired dollar holdings of each currency in to the instructor. The drawback of these approaches is that groups don’t get to observe the decisions of the other teams in real time.

4 These labels are chosen because “$out” represents the dollar value that came out of the previous trade period, and “$in” represents the dollar value that teams are putting into each currency.

5 Rousu et al. (2015) provide evidence that real monetary incentives are generally important for classroom experiments. In this case, real money reinforces the role of being a financial trader.

6 Rather than debriefing the entire experiment at the end of the semester, some instructors may wish to discuss particular parts of the experiment as they come up in lecture. For our experiment, the entire debrief was conducted on the last trading day.

7 The 1-year sample horizon is not based on any hard and fast rule from economic theory; however, we also constructed a minimum variance portfolio based on 5 years’ worth of data and it did not perform as well as the portfolio based on 1 year of data. The reason for this is that exchange rate regimes may change and political events may shock the exchange rate. Neither of these changes should provide any predictive power going forward. In our sample, for example, the 5-year horizon included the Brexit vote and a softening of the Chinese peg.

8 Appendix C is available upon request.