Date of Award

2015

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Applied Economics

First Advisor

Kent Smetters

Abstract

In the first chapter, Christian Goulding and I present a model of asset prices with recursive preferences and the simple consumption growth dynamics of Mehra and Prescott (1985) but relax the assumption that preference parameters are constant over time. We show that rare, temporary, and plausible fluctuations in the elasticity of inter-temporal substitution (EIS) and risk aversion (RA) can quantitatively explain numerous regularities in U.S. asset prices including: the equity premium and risk-free rate puzzles, excess return and consumption growth predictability, a counter-cyclical risk premium and an upward-sloping real yield curve. A novel implication is that time-varying EIS is more important than time-varying RA for explaining many of these regularities, suggesting a new source of risk in investors' ability to plan their consumption over long horizons. In addition, our model can accommodate a behavioral interpretation of psychological factors (e.g. fear) that drive fluctuations in asset prices beyond traditional risk factors.

The second chapter is an empirical study of the value of star college athletes. Collegiate athletes in the U.S. are not allowed to be paid directly for their athletic ability. Under the current regulations imposed by the NCAA, any compensation beyond scholarships and grant-in-aid to cover some basic living expenses is forbidden. This artificial constraint on athletes' wages, when university athletic programs are generating significant revenues, has sparked much recent debate over the compensation of college athletes. To help inform this debate, I quantify the value of a NCAA Division 1 FBS (Football Bowl Subdivision) star football and NCAA Division 1 star basketball players by estimating their marginal revenue product using a novel dataset of individual player and team performance statistics and publicly available athletic program revenue data. I find that a star college football player is worth up to $1.2-$2.1 million while star college basketball players are worth up to $655,000-$1.1 million a year. Interestingly, I also find evidence suggesting that a college recruiter's ability to identify revenue generating star players is limited and that the marginal revenue product of star college players declines as the team's media coverage increases.

Share

COinS