SA475: Research Seminar
Recent Examples of Seminars
- Understanding Asset Prices and Returns
- Experimental Economics
Understanding Asset Prices and Returns
Instructors: Professors Gary Fowler (Math) and Katherine Smith (Econ)
Primary question: Are U.S. capital markets efficient (competitive)?
In order to answer this question, we first need to understand what market efficiency means. This is done by developing a simple theoretical model of how assets are priced. By solving a single (representative) agent's utility maximization problem, the first order conditions reveal a general asset pricing equation that can be used to price any asset. Given certain assumptions, market efficiency and the random walk hypothesis fall out of this basic equation.
While the theoretical model is illustrative, it cannot be used to test for market efficiency empirically in that the determinants of asset prices are unobservable. Students will therefore come up with their own econometric models (based on the theoretical model) which are empirically tractable. They will then gather the relevant data and use their model to test whether capital markets are indeed efficient.
Instructors: Professors Charles Mylander (Math) and Pamela Schmitt (Econ)
Game theory is broadly interpreted as the study of multiperson decision problems. The disciplines most involved in game theory are economics, mathematics, and military science. Students working on a project using game theory will use concepts and skills acquired in other mathematics and economics courses to formulate a matrix game model and estimate the payoff matrix for a scenario of their choice. For 2-person game models optimal min-max solutions must be found. Examples of areas using matrix games in analysis include ways to efficiently govern common pool resources; ways to efficiently provide public goods; oligopoly pricing decisions; coordination games; tactics in sports, and anti-submarine tactics.
Experimental economics involves the design of experiments using human subjects to support or refute the predictive powers of economic models and theories. Experiments have also been used to compare the behavior of economic decisions makers (subjects in an experiment) to a game theoretic solution (i.e. the Nash Equilibrium). For example, game theory indicates that the Nash Equilibrium solution to the voluntary contribution public goods game with is for a player to “free-ride”, thereby not contributing to the public good. Experimental economics allows economists to determine if this equilibrium is consistent with human behavior. Finally, economic experiments have also been designed to test the assumptions of consumer and producer behavior in economic models. Students working in this area will choose an economic model or game model and design an experiment using midshipmen playing the roles of the economic agents. The results of the experiment will be analyzed, usually by statistical methods, and compared with the results (or hypotheses) predicted by the theoretic model. In many cases the economic experiment will be a variation of an experiment reported in the literature.
The semester is devoted to working on individual research papers. The research papers will either focus on linear programming methods or experimental methods. The research papers focused on experimental methods must include: (1) a review of the relevant literature, (2) a presentation of the experimental procedures used and the research hypotheses, and (3) an analysis of data and empirical evaluation of hypotheses using appropriate statistical techniques. During the research phase, weekly progress reports are required. Interim grades (6 and 12 week) are based on seminar participation and progress reports. The final course grade is driven primarily by the grade on the paper and your class presentation.