Skip to main content
. 2021 Jan 16:utab001. doi: 10.1093/ectj/utab001

Table 3.

GDP forecast performance of the TVi-CSV-MF-VAR relative to two other models.

Horizon in months 0 2 4 6 8
A. Relative to Antolin-Diaz et al. (2017)
Rel. RMSFEs 0.94 0.93 0.97 0.99 1.03
Rel. logPLs 0.11 0.08 –0.03 –0.03 –0.02
B. Relative to Cimadomo and D’Agostino (2016)
Rel. RMSFEs 0.94 1.00 1.01 1.06 1.04
Rel. logPLs 0.17 0.09 0.02 0.05 0.05

Note: Comparison of our parsimonious mixed-frequency model with time-varying intercepts and CSV (TVi-CSV-MF-VAR) and two different models: (A) The model by Antolin-Diaz et al. (2017); (B) The model by Cimadomo and D’Agostino (2016) with Inline graphic (GDP, CPI, FFR) and Inline graphic.