A Comparison of Forecasting Procedures for Macroeconomic Series: The Contribution of Structural Break Models

Bauwens, Luc; Korobilis, Dimitris; Koop, Gary; Rombouts, Jeroen V.K.

doi:10.2139/ssrn.1748703

Cited by 15 publications

(21 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Following studies such as Bauwens, et al (2011), we compute p(π o t |π (t−1) , M i ) from the simulated predictive density, using a kernel smoother to estimate the empirical density (from draws of forecasts). Finally, in computing the log predictive likelihood, we sum the log values over different samples, detailed below.…”

Section: Model Assessmentmentioning

confidence: 99%

A Bayesian Evaluation of Alternative Models of Trend Inflation

Clark

Doh²

2011

Working Paper (Federal Reserve Bank of Cleveland)

View full text Add to dashboard Cite

Section: Model Assessmentmentioning

confidence: 99%

A Bayesian Evaluation of Alternative Models of Trend Inflation

Clark

Doh²

2011

Working Paper (Federal Reserve Bank of Cleveland)

View full text Add to dashboard Cite

“…Uncertainty about the number of regimes is easily incorporated in a Bayesian context by setting a maximum number of breaks, say K max , and allowing the data to determine the 'true' number of estimated breaks K, where 1 ≤ K ≤ K max . In Bauwens et al (2011) we give exact implementation details on forecasting with a univariate version of this model, which I follow closely in this multivariate extension. Estimation details are provided in the Appendix.…”

Section: Structural Breaks Varmentioning

confidence: 99%

“…Subsequently one could also attempt to implement variable selection in the coefficients of each regime separately. Nevertheless, given the uncertainty about drawing the latent break dates in these models (see Bauwens et al, 2011), having to sample latent variable selection indicators in each regime can result in a very inefficient posterior sampler. Jochmann et al (2010) allow selection of coefficients in different regimes in the context of the stochastic search variable selection algorithm (see next subsection); however, they do so using crucial simplifying assumptions to increase stability of their posterior sampler and reduce uncertainty of the break dates.…”

Section: Structural Breaks Varmentioning

confidence: 99%

VAR Forecasting Using Bayesian Variable Selection

Korobilis

2009

SSRN Journal

Self Cite

View full text Add to dashboard Cite

This paper develops methods for automatic selection of variables in Bayesian vector autoregressions (VARs) using the Gibbs sampler. In particular, I provide computationally efficient algorithms for stochastic variable selection in generic linear and nonlinear models, as well as models of large dimensions. The performance of the proposed variable selection method is assessed in forecasting three major macroeconomic time series of the UK economy. Data-based restrictions of VAR coefficients can help improve upon their unrestricted counterparts in forecasting, and in many cases they compare favorably to shrinkage estimators.

show abstract

“…We denote this model as BMA-RW-SV (see Table 1 for details. 5 4 The specification put forward here is more flexible than break models applied in Koop and Potter (2009) and Bauwens, Koop, Korobilis, and Rombout (2015). Specifically, we have the flexibility of allowing for different parameters to change at different points in time.…”

Section: Frameworkmentioning

confidence: 99%

Modeling and forecasting aggregate stock market volatility in unstable environments using mixture innovation regressions

Nonejad

2017

Journal of Forecasting

View full text Add to dashboard Cite

We perform Bayesian model averaging across different regressions selected from a set of predictors that includes lags of realized volatility, financial and macroeconomic variables. In our model average, we entertain different channels of instability by either incorporating breaks in the regression coefficients of each individual model within our model average, breaks in the conditional error variance, or both. Changes in these parameters are driven by mixture distributions for state innovations (MIA) of linear Gaussian state‐space models. This framework allows us to compare models that assume small and frequent as well as models that assume large but rare changes in the conditional mean and variance parameters. Results using S&P 500 monthly and quarterly realized volatility data from 1960 to 2014 suggest that Bayesian model averaging in combination with breaks in the regression coefficients and the error variance through MIA dynamics generates statistically significantly more accurate forecasts than the benchmark autoregressive model. However, compared to a MIA autoregression with breaks in the regression coefficients and the error variance, we fail to provide any drastic improvements.

show abstract

A Comparison of Forecasting Procedures for Macroeconomic Series: The Contribution of Structural Break Models

Cited by 15 publications

References 21 publications

A Bayesian Evaluation of Alternative Models of Trend Inflation

A Bayesian Evaluation of Alternative Models of Trend Inflation

VAR Forecasting Using Bayesian Variable Selection

Modeling and forecasting aggregate stock market volatility in unstable environments using mixture innovation regressions

Contact Info

Product

Resources

About