Working Paper: NBER ID: w31419
Authors: Menzie D. Chinn; Baptiste Meunier; Sebastian Stumpner
Abstract: We nowcast world trade using machine learning, distinguishing between tree-based methods (random forest, gradient boosting) and their regression-based counterparts (macroeconomic random forest, gradient linear boosting). While much less used in the literature, the latter are found to outperform not only the tree-based techniques, but also more “traditional” linear and non-linear techniques (OLS, Markov-switching, quantile regression). They do so significantly and consistently across different horizons and real-time datasets. To further improve performances when forecasting with machine learning, we propose a flexible three-step approach composed of (step 1) pre-selection, (step 2) factor extraction and (step 3) machine learning regression. We find that both pre-selection and factor extraction significantly improve the accuracy of machine-learning-based predictions. This three-step approach also outperforms workhorse benchmarks, such as a PCA-OLS model, an elastic net, or a dynamic factor model. Finally, on top of high accuracy, the approach is flexible and can be extended seamlessly beyond world trade.
Keywords: Nowcasting; World Trade; Machine Learning; Forecasting
JEL Codes: C53; C57; E37
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| regression-based machine learning techniques (C29) | prediction accuracy (C52) |
| three-step approach (Y20) | prediction accuracy (C52) |
| preselection (D79) | prediction accuracy (C52) |
| factor extraction (C38) | prediction accuracy (C52) |
| machine learning regression techniques (C45) | forecasting world trade (F17) |
| traditional methods (C90) | prediction accuracy (C52) |