Date of Degree
2009
Document Type
dissertation
Degree Name
PhD (Doctor of Philosophy)
Department
Computer Science
First Advisor
Kasturi Varadarajan
Abstract
It is demonstrated that for certain markets where traders have constant elasticity of substitution utility (CES) functions, the existence of a price equilibrium can be determined in polynomial time. It is also shown that for a certain range of elasticity of substitution where the CES market does not satisfy gross subsitutability that price equilibira can be computed in polynomial time. It is also shown that for markets satisfying gross substitutability, equilibria can be computed in polynomial time even if the excess demand is a correspondence. On the experimental side, equilibrium computation algorithms from computer science without running time guarantees are shown to be competitive with software packages used in applied microeconomics. Simulations also lend support to the Nash equilibrium solution concept by showing that agents employing heuristics in a restricted form of Texas Holdem converge to an approximate equilibrium. Monte Carlo simulations also indicate the long run preponderance of skill over chance in Holdem tournaments.
Pages
xiii, 146
Bibliography
139-146
Copyright
Copyright 2009 Benton John McCune
Recommended Citation
McCune, Benton John. "Algorithmic game theory and the computation of market equilibria." dissertation, University of Iowa, 2009.
http://ir.uiowa.edu/etd/405.