Abstract
This project proposes various new methods for the theoretical and empirical analysis of arbitrage-free asset markets in the presence of frictions and/or pricing errors. The first part of the project proposes new model-free technologies to (i) test asset pricing models in the presence of pricing errors, (ii) select asset pricing factors incorporating the no-arbitrage constraints implied by the Arbitrage Pricing Theory (APT) and (iii) detect global factors in international asset markets. The second part of the project focuses on dynamic economies with transaction costs. It proposes new technologies for solving optimal portfolio problems and for characterizing equilibrium asset prices in the presence of multiple traded assets and multivariate state dynamics. Empirically, we adopt linear pricing rules represented by stochastic discount factors that are compatible with general transaction cost structures. In this way, we introduce more realistic testing procedures for asset pricing models and factor detection approaches that are (i) economically motivated by economic assumptions regarding, e.g., market frictions or asymptotic no-arbitrage conditions in the APT and (ii) more suitable for the analysis of large asset markets. On the theoretical side, we aim to develop technologies to solve partial equilibrium and general equilibrium asset pricing problems with transaction costs for a broader class of economic settings than previously achieved in the literature. By their nature, such technologies can give rise to powerful tools to improve the understanding of the properties of, e.g., asset prices, volume, and trading behavior in financial markets with frictions.