Forecasting under model uncertainty:Non-homogeneous hidden Markov models with Polya-Gamma data augmentation

TL;DR

This paper introduces a Bayesian approach using Pólya-Gamma data augmentation for forecasting univariate time series with non-homogeneous hidden Markov models, effectively handling predictor uncertainty and improving predictive accuracy.

Contribution

It develops a novel latent variable scheme with Pólya-Gamma augmentation for inference in NHHMMs, enabling better predictor selection and forecasting under model uncertainty.

Findings

Improved forecast accuracy over benchmarks on volatility data

Effective predictor selection via reversible jump MCMC

Enhanced model mixing and convergence properties

Abstract

We consider two-state Non-Homogeneous Hidden Markov Models (NHHMMs) for forecasting univariate time series. Given a set of predictors, the time series are modeled via predictive regressions with state dependent coefficients and time-varying transition probabilities that depend on the predictors via a logistic function. In a hidden Markov setting, inference for logistic regression coefficients becomes complicated and in some cases impossible due to convergence issues. In this paper, we aim to address this problem using a new latent variable scheme that utilizes the P\'{o}lya-Gamma class of distributions. We allow for model uncertainty regarding the predictors that affect the series both linearly -- in the mean -- and non-linearly -- in the transition matrix. Predictor selection and inference on the model parameters are based on a MCMC scheme with reversible jump steps. Single-step and multiple-steps-ahead predictions are obtained by the most probable model, median probability model or a Bayesian Model Averaging approach. Using simulation experiments, we illustrate the performance of our algorithm in various setups, in terms of mixing properties, model selection and predictive ability. An empirical study on realized volatility data shows that our methodology gives improved forecasts compared to benchmark models.