Deus Ex Machina? A Framework for Macro Forecasting with Machine Learning

Bolhuis, Marijn; Rayner, Brett

doi:10.2139/ssrn.3579665

Cited by 6 publications

(5 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using a dataset of 19 variables, he shows how Elastic Net and Random Forest can be used to nowcast Lebanese GDP growth before official data is released. Bolhuis and Rayner (2020) apply alternative machine learning algorithms to nowcast the Turkish GDP growth. In addition, to further lower forecast errors, the authors combine individual machine learning models into ensembles.…”

Section: Literature Reviewmentioning

confidence: 99%

“…A decision tree is a non-parametric approach that in each step iteratively splits a sample into two groups chosen by the algorithm to yield the largest reduction in the forecast error of the variable of interest (Bolhuis and Rayner (2020)), the so-called recursive partitioning. Regression tree is nonparametric regression method that allows for prediction of continuous variables (Chart [x]).…”

Section: Random Forestmentioning

confidence: 99%

“…If C = 0, the algorithm disregards individual deviations and constructs the simplest hyperplane for which every observation is still within the acceptable margin of 𝜀𝜀. For sufficiently large C, the algorithm will construct the most complex hyperplane that predicts the outcome for the training data with zero error(Bolhuis and Rayner 2020).Chart [x]:A simple SVM to nowcast GDP Note: The orange hyperplane is the prediction, while blue dots are actual observations. The model contains 40 regressors, while to visualize, we only use term premium and economic sentiment.…”

mentioning

confidence: 99%

“…Bolhuis and Rayner (2020) define support vector machine (SVM) as an algorithm that constructs hyperplanes to partition predictor combinations and make a point forecast for each of the sections.©International Monetary Fund. Not for Redistribution…”

mentioning

confidence: 99%

See 3 more Smart Citations

Nowcasting GDP - A Scalable Approach Using DFM, Machine Learning and Novel Data, Applied to European Economies

Dauphin¹,

Sanjani²,

Suphaphiphat³

et al. 2022

IMF Working Papers

View full text Add to dashboard Cite

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Section: Random Forestmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Nowcasting GDP - A Scalable Approach Using DFM, Machine Learning and Novel Data, Applied to European Economies

Dauphin¹,

Sanjani²,

Suphaphiphat³

et al. 2022

IMF Working Papers

View full text Add to dashboard Cite

show abstract

“…First, this paper relies on a range of data-selection machine-learning approaches, related to Lasso and elastic net. Advances in data computing have resulted in an increased use of machine learning and other big data techniques in the data selection process (Bolhuis & Rayner, 2020;Chakraborty & Joseph, 2017;Jung et al, 2018;Proietti & Giovannelli, 2020;Woloszko, 2020).…”

Section: Literaturementioning

confidence: 99%

Nowcasting services trade for the G7 economies

Jaax,

Mourougane,

Gonzales

2023

World Economy

View full text Add to dashboard Cite

The growing importance of services trade in the global economy contrasts with the scarcity of timely data on this part of international trade. This paper develops models to nowcast aggregate services imports and exports, as measured by monthly services trade data, for the G7 countries. The methodology relies on machine‐learning techniques and dynamic factor models and combines traditional high‐frequency data with the use of Google Trends search data. The estimated services trade nowcasting models display a higher out‐of‐sample predictive power than a simple benchmark model. However, there does not seem to be one approach that outperforms other model specifications. Rather, a weighted average of the best models, combining machine‐learning with dynamic factor models, seems to be a promising avenue. The best models improve one step ahead predictive performance relative to a simple benchmark by 30%–35% on average across the G7 countries and across trade flows. Models capture approximately 67% of the fall in services exports following the COVID‐19 shock and 60% of the fall in imports on average across G7 economies.

show abstract