October 3, 2008
Evaluating CPB's published GDP growth forecasts; a comparison with individual and pooled VAR based forecasts
We compare the accuracy of our published GDP growth forecasts from our large macro model, SAFFIER, to those produced by VAR based models using both classical and Bayesian estimation techniques.
We employ a data driven methodology for selecting variables to include in our VAR models and we find that a randomly selected classical VAR model performs worse in most cases than the Bayesian equivalent, which performs worse than our published forecasts in most cases. However, when we pool forecasts across many VARs we can produce more accurate forecasts than we published. A review of the literature suggests that forecast accuracy is likely irrelevant for the non-forecasting activities the model is used for at CPB because they are fundamentally different activities.
Downloads
Authors
![Adam Elbourne Photo of Adam Elbourne](https://www.cpb.nl/sites/default/files/user/pictures/picture-89682-1549549857.jpg)
![Henk Kranendonk Photo of Henk Kranendonk](https://www.cpb.nl/sites/default/files/user/pictures/picture-89433-1544015889.jpg)
![Rob Luginbuhl Photo of Rob Luginbuhl](https://www.cpb.nl/sites/default/files/user/pictures/picture-89631-1549458057.jpg)
![Bert Smid Photo of Bert Smid](https://www.cpb.nl/sites/default/files/user/pictures/picture-89535-1544098252.jpg)
Martin Vromans