Forecasting in Dynamic Factor Models Subject to Structural Instability*

Stock, James H.; Watson, Mark W.

doi:10.1093/acprof:oso/9780199237197.003.0007

Cited by 147 publications

(151 citation statements)

References 15 publications

Supporting

Mentioning

138

Contrasting

Order By: Relevance

“…We work with a small TVP-VAR with three variables, a medium TVP-VAR with seven and a large TVP-VAR with 25. Following, e.g., Stock and Watson (2008) and recommendations in Carriero, Clark and Marcellino (2011) we transform all variables to stationarity.…”

Section: Datamentioning

confidence: 99%

“…Some series in the database were observed only on a monthly basis and quarterly values were computed by averaging the monthly values over the quarter. All variables are transformed to be approximately stationary following Stock and Watson (2008). In particular, if z i;t is the original untransformed series, the transformation codes are (column Tcode below):…”

Section: A Data Appendixmentioning

confidence: 99%

See 1 more Smart Citation

Large time-varying parameter VARs

Koop

Korobilis

2013

Journal of Econometrics

366

385

View full text Add to dashboard Cite

In this paper, we develop methods for estimation and forecasting in large time-varying parameter vector autoregressive models (TVP-VARs). To overcome computational constraints, we draw on ideas from the dynamic model averaging literature which achieve reductions in the computational burden through the use forgetting factors. We then extend the TVP-VAR so that its dimension can change over time. For instance, we can have a large TVP-VAR as the forecasting model at some points in time, but a smaller TVP-VAR at others. A final extension lies in the development of a new method for estimating, in a timevarying manner, the parameter(s) of the shrinkage priors commonly-used with large VARs.These extensions are operationalized through the use of forgetting factor methods and are, thus, computationally simple. An empirical application involving forecasting inflation, real output and interest rates demonstrates the feasibility and usefulness of our approach.

show abstract

Section: Datamentioning

confidence: 99%

Section: A Data Appendixmentioning

confidence: 99%

Large time-varying parameter VARs

Koop

Korobilis

2013

Journal of Econometrics

366

385

View full text Add to dashboard Cite

show abstract

“…To provide an empirical application of our tests, we use Stock and Watson's (2009) ) the estimated number of factors ranges from 2 to 6, therefore we implement our test forr = 2-6. The Sup-Wald test is applied since no priori break date is assumed to be known.…”

Section: An Empirical Applicationmentioning

confidence: 99%

“…The estimated break date is seems to be around 1979 or 1980 (second oil price shock), rather than 1984, which is the only candidate considered by Stock and Watson (2009) as a potential break date in their empirical application with the same data set. One possible explanation for this break date could be the Iranian revolution at the beginning of 1979 and its subsequent impact on monetary policy in the US (see Fernández-Villaverde et al (2010)).…”

Section: An Empirical Applicationmentioning

confidence: 99%

Detecting big structural breaks in large factor models

Liang

Dolado

Gonzalo

2014

Journal of Econometrics

116

107

View full text Add to dashboard Cite

a b s t r a c tTime invariance of factor loadings is a standard assumption in the analysis of large factor models. Yet, this assumption may be restrictive unless parameter shifts are mild (i.e., local to zero). In this paper we develop a new testing procedure to detect big breaks in these loadings at either known or unknown dates. It relies upon testing for parameter breaks in a regression of one of the factors estimated by Principal Components analysis on the remaining estimated factors, where the number of factors is chosen according to Bai and Ng's (2002) information criteria. The test fares well in terms of power relative to other recently proposed tests on this issue, and can be easily implemented to avoid forecasting failures in standard factor-augmented (FAR, FAVAR) models where the number of factors is a priori imposed on the basis of theoretical considerations.

show abstract

“…Stock and Watson (2008) and Banerjee et al (2008) point out the possibility of changing factor loadings that could affect factor estimation and, thus, the forecasting ability of these models, up until now there are no papers that empirically contrast this possibility.…”

Section: Introductionmentioning

confidence: 99%