A comparison of Direct and Iterated AR Methods for Forecasting Macroeconomic Series h-Steps Ahead

53 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Forecasting Macroeconomic Variables Using Diffusion Indexes in Short Samples with Structural Change (2008, unknown): “2 Two other examples of the usefulness of factor models with fewer than 50 time-series observations are Matheson (2005) for the case of New Zealand and Breitung and Eickmeier (2005) who provide an example of the use of dynamic factor models in macroeconomic analysis for the case of the Euro area usi”
Working Paper Series (2019, unknown): “Marcellino et al. (2006) used 171 series for the sample 1959:1-2002:12 to assess this and alternative forecasting methodologies.”
Combining forecast densities from VARs with uncertain instabilities (2010, methodology): “Each individual model is used to produce h-step ahead forecasts via the direct approach; see the discussion by Marcellino et al. (2003). Hence, the macro variables used to produce an h-step ahead forecast density for τ are dated τ − 1 − h.”
Large time-varying parameter VARs (2013, methodology): “Both ways provide us with βT+h and we simulate draws of yτ+h conditional on βT+h to approximate the predictive density.7 The alternative would be to use direct forecasting, but recent papers such as Marcellino et al. (2006) tend to find that iterated forecasts are better.”