Dynamics, Sparsity, and Optimization in High-Dimensional Portfolio Decisions

• Activity Number: 395
• Type: Invited
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Mike West* and Kaoru Irie
• Companies: Duke University and Duke University
• Keywords: Bayesian decision analysis ; Bayesian forecasting ; Multi-step forecasting ; Portfolio decisions

Abstract

I discuss challenges and recent developments in Bayesian decision analysis in portfolio studies, touching on a number of general questions and decision theoretic perspectives. This draws on recent work in Bayesian forecasting and decision theory emerging from personal and corporate interests in improved financial portfolio construction that involve multi-step forecasts and decisions in financial time series. I discuss new classes of economically and psychologically relevant portfolio utility functions, and solutions to the resulting Bayesian expected utility optimization problem. The latter involves a novel idea of mapping the technical structure of (some) optimization problems to those of parallel, synthetic Bayesian inference problems, thus providing access to standard Bayesian computational methods. Applications in sequential portfolio studies with multivariate financial (currency, commodity and stock index) time series illustrate some of the practical utility and benefits of the new Bayesian decision analysis framework and methodology.