New UNCTAD website monitors world trade and GDP in real time

The site enhances the usefulness and accessibility of UNCTAD nowcasts to the international community.

© Shutterstock/Avigator Fortuner

UNCTAD launched a new interactive website on July 1 that provides real-time estimates of global trade and GDP, showing the impacts of multiple global economic shocks.

The website provides data-driven estimates, known as nowcasts, that highlight lingering economic pressures from the war in Ukraine and continuing high inflationary concerns. They offer timely information to help guide policy responses.

UNCTAD has long produced nowcasts of quarterly world merchandise export growth. These have provided a timely picture of how global trade has evolved since the onset of the COVID-19 pandemic.

But their dissemination via the end-of-term bulletins was limited.

Latest data available

To generate the real-time estimates, the latest available data and data revisions are fed into the models on a weekly basis to update and revise the nowcasts and provide insight into current economic and business conditions well in advance of the release of forecasts. final figures several months late.

Estimates are derived from models for the three UNCTAD trade series covering world merchandise exports expressed in values ​​and volumes and services exports. And for the first time, the estimates include annual growth in global GDP.

In addition to estimates, the website displays the evolution of nowcasts over time, showing how data releases have shaped them as economic conditions have evolved.

The models underlying nowcasting are known as short-term memory (LSTM) artificial neural networks, which surpass the previous approach of UNCTAD’s dynamic factor model.

The methodology and information on the accompanying open-source multi-programming language library is available in an UNCTAD research paper.

A second UNCTAD research paper details the performance of the methodology during the pandemic, while a third research paper compares its performance to other common nowcasting and machine learning methodologies.

Sherry J. Basler