A dynamic factor model framework for forecast combination

Yeung Lewis Chan, James H. Stock, Mark W. Watson

Research output: Contribution to journalArticlepeer-review

63 Scopus citations


A panel of ex-ante forecasts of a single time series is modeled as a dynamic factor model, where the conditional expectation is the single unobserved factor. When applied to out-of-sample forecasting, this leads to combination forecasts that are based on methods other than OLS. These methods perform well in a Monte Carlo experiment. These methods are evaluated empirically in a panel of simulated real-time computer-generated univariate forecasts of U.S. macroeconomic time series.

Original languageEnglish (US)
Pages (from-to)91-121
Number of pages31
JournalSpanish Economic Review
Issue number2
StatePublished - Jul 1 1999

All Science Journal Classification (ASJC) codes

  • General Economics, Econometrics and Finance


  • JEL classification: C32, C22
  • Key words: Combination forecasts, principal component regression, James-Stein estimation


Dive into the research topics of 'A dynamic factor model framework for forecast combination'. Together they form a unique fingerprint.

Cite this