Dynamic Mixture Vector Autoregressions with Score-Driven Weights
51 Pages Posted: 16 Feb 2022 Last revised: 27 Nov 2024
There are 2 versions of this paper
Dynamic Mixture Vector Autoregressions with Score-Driven Weights
Dynamic Mixture Vector Autoregressions with Score-Driven Weights
Date Written: November 27, 2024
Abstract
We propose a novel dynamic mixture vector autoregressive (VAR) model in which the time-varying mixture weights are driven by the predictive likelihood score. Intuitively, the state weight of the k-th component VAR model is increased in the subsequent period if the current observation is more likely to have been drawn from this particular state. The model is not limited to a specific distributional assumption and allows for straightforward likelihood-based estimation and inference. We conduct a Monte Carlo study and find that the score-driven mixture VAR model is able to adequately filter and predict the mixture dynamics from a variety of different data generating processes, which other observation-driven dynamic mixture VAR models cannot handle appropriately. Finally, the empirical performance of the approach is illustrated by two applications: (i) the conditional joint distribution of stock and bond returns, and (ii) the regime-dependent connection of economic and financial conditions.
Keywords: Dynamic Mixture Models; Generalized Autoregressive Score Models; Macro-Financial Linkages; Nonlinear Vector Autoregressions; Stock and Bond Return Dynamics
JEL Classification: C32, C34, G17
Suggested Citation: Suggested Citation