Bayesian Econometrics in Finance

Eric Jacquier
MIT Sloan and HEC Montreal

Nicholas Polson
Booth Graduate School of Business, University of Chicago


This chapter surveys Bayesian Econometric methods in finance. Bayesian methods provide a natural framework for addressing central issues in finance. In particular, they allow investors to assess return predictability, estimation and model risk, formulated predictive densities for variances, covariances and betas. This can be done through decision theoretic problems, such as option pricing or optimal portfolio allocation. Bayesian predictive distributions are straightforward to calculate and summarize the investorŐs future views for return distribution and expected utility computation. Nonlinear functionals, such as market efficiency measures and Sharpe ratios, are easily dealt with from a Bayesian perspective. A central theme in this chapter is the use of simulation-based estimation and prediction via Markov Chain Monte Carlo (MCMC) and particle filtering (PF) algorithms. We provide detailed applications of these methods to the central issues in finance.