Nowcasting GDP: what are the gains from machine learning algorithms?

June 4, 2024
Issue 2024-06


We compare several machine learning methods for nowcasting GDP. A large mixed-frequency data set is used to investigate different algorithms such as regression based methods (LASSO, ridge, elastic net), regression trees (bagging, random forest, gradient boosting), and SVR. As benchmarks, we use univariate models, a simple forward selection algorithm, and a principal components regression. The analysis accounts for publication lags and treats monthly indicators as quarterly variables combined via blocking. Our data set consists of more than 1,100 time series. For the period after the Great Recession, which is particularly challenging in terms of nowcasting, we find that all considered machine learning techniques beat the univariate benchmark up to 28 % in terms of out-of-sample RMSE. Ridge, elastic net, and SVR are the most promising algorithms in our analysis, significantly outperforming principal components regression.

JEL classification:
C53, C55, C32, C37
Nowcasting, Forecasting, Machine learning, Rridge, LASSO, Elastic net, Random forest, Bagging, Boosting, SVM, SVR, Large data sets

Additional files

Related content


  • Milen Arro-Cannarsa

  • Dr. Rolf Scheufele

Your settings

Required: These cookies (e.g. for storing your IP address) cannot be rejected as they are necessary to ensure the operation of the website. These data are not evaluated further.
Analytics: If you consent to this category, data such as IP address, location, device information, browser version and site visitor behaviour will be collected. These data are evaluated for the SNB's internal purposes and are kept for two years.
Third-party: If you consent to this category, third-party services (used, for example, to add social multimedia content to the SNB's website) will be activated which collect personal data, process these data, disclose them abroad - worldwide - and place cookies. The relevant data protection regulations are linked in the 'Privacy statement for the website of the Swiss National Bank'.

Choose your preferred settings:

This website uses cookies, analytics tools and other technologies to provide requested features, content and services, to personalise the content shown, to provide links to social media, and to analyse the use of the website in anonymised form for the purposes of improving usability. Personal data are also disclosed abroad - worldwide - to video service providers and the analytics tools of these providers are used. More information is available under 'Manage settings'.