Efficient Bayesian Inference in High Dimensional Time Series: Examples from Macroeconomics and Finance

Abstract:

Statistical inference for dynamic models in many dimensions often comes along with a huge amount of parameters that need to be estimated.

Thus, to handle the curse of dimensionality, suitable regularization methods are of prime importance, and efficient computational tools are required to make practical estimation feasible. In this talk, I exemplify how these two principles can be implemented for Bayesian models of importance in macroeconomics and finance. To begin with, a focus is placed on sparse factor stochastic volatility models to estimate and predict large dynamic covariance matrices. In what follows, I discuss vector autoregressive models with time-varying contemporaneous correlations that are capable of handling vast dimensional information sets. To conclude, I propose an algorithm to carry out inference in large heteroskedastic dynamic regression settings with mixture innovation components for each coefficient in the system. All models are illustrated with empirical applications.