Computation in Macroeconomic Asset Pricing

Abstract

This dissertation investigates computational methods for macroeconomic asset pricing models. It demonstrates that advances in economic modeling often require advances in computation and highlights a particular case where more demanding computational methods are required to solve an economic model. It also discusses advances in computational technology that allow researchers to utilize solution methods that would have been previously infeasible. In particular, it demonstrates the wide applicability and potential gains of GPU computing, a parallel computing framework, and applies those tools to a computationally challenging model which investigates trading volume in a general equilibrium, complete-markets economy where agents have heterogeneous beliefs.