Tracking Global Trade from Space: IMF’s Satellite-Based Nowcasting Breakthrough

A pioneering working paper from the International Monetary Fund (IMF), authored in collaboration with researchers from the University of Wisconsin and Delft University of Technology, introduces a revolutionary model for real-time monitoring global trade. Titled “Nowcasting Global Trade from Space” (WP/25/93), the study uses satellite-based data from the Automatic Identification System (AIS) to track vessel movements and generate near-instantaneous estimates of maritime trade flows. This approach offers a much-needed solution to the delays and gaps in traditional trade data, which often take weeks or months to become available. As maritime transport accounts for nearly 80 percent of global merchandise trade by volume, the model’s timely insights provide a powerful lens for economic analysis, forecasting, and policy planning.

PortWatch: The Engine Behind the Model

At the heart of this innovation lies the IMF’s PortWatch platform, a publicly accessible tool launched in beta in late 2023. PortWatch collects high-frequency AIS signals, transmitted by tankers, bulk carriers, and containerships, across 1,666 ports and 24 strategic maritime chokepoints worldwide. These signals provide details about ship positions, drafts, and cargo flows. Since its launch, the platform has been significantly upgraded. It now includes a broader port database, refined data estimation methods, and adjustments for two major technical issues: the netting effect in containerized trade and the use of ballast water. The netting effect occurs when containerships simultaneously load and unload cargo at ports, leading to underestimated volumes in AIS data. This was corrected using container throughput data from major ports and statistical bootstrapping techniques. The ballast water adjustment ensures that the weight of non-cargo water carried by ships is excluded from trade calculations, improving accuracy.

From Ship Signals to Trade Values: A Three-Step Process

The nowcasting model operates through a clear and structured three-step methodology. The first step involves aggregating physical cargo volumes by vessel type across all monitored ports. The second step translates this volume into trade value by applying average unit values, dollars per metric ton, sourced from the CEPII BACI database. These values are updated using global commodity and manufacturing price indices, as well as supplementary CPI data from the U.S. The third step deflates the trade values using Laspeyres-type indices to derive trade volumes in constant prices, a method widely used by statistical agencies in Europe and the United States. The result is a set of global and regional trade indices, benchmarked to a 2019 base year and available just seven working days after the reference month.

Tracking Crises and Shifts: From COVID to Geopolitics

Validation against official trade data reveals the model’s impressive performance. Compared with the Netherlands Bureau for Economic Policy Analysis’s (CPB) World Trade Monitor, which covers 81 countries, the IMF’s nowcast correlates at 0.95 for trade value and 0.80 for trade volume. These results are particularly striking during periods of global disruption. The model successfully tracked the sharp drop in trade in early 2020 during the COVID-19 outbreak, the subsequent recovery, and the slowdown in 2023. Even during the volatile early months of the Ukraine conflict, the model captured changes in global trade flows. Regional estimates also show a strong fit, especially for advanced economies and major emerging markets. Beyond trade, the nowcast aligns closely with global industrial production data, with a correlation of 0.78. Seasonal adjustments using the X12-ARIMA method further allow for meaningful month-over-month and quarter-over-quarter comparisons, with a correlation of 0.97 for trade value in the adjusted three-month series.

A Window into Fragmentation and Regionalization

One of the most compelling applications of the model is its ability to track geopolitical and structural shifts in global trade. The study finds strong evidence of growing fragmentation between U.S.- and China-aligned blocs. Since 2019, the share of trade between these two blocs has declined, while trade with nonaligned countries has expanded, supporting the idea of “connector countries” facilitating trade amid rising geopolitical tensions. The transformation of Russia’s oil trade after Western sanctions is a vivid case in point: AIS data show a surge in oil shipments to China, India, and Brazil, replacing exports to Europe. Meanwhile, Europe has increased its reliance on imports from the U.S. and Norway. On the question of regionalization, however, the picture is more nuanced. While East Asia and Europe account for the bulk of intra-regional maritime trade, the global trend since 2019 shows no strong or consistent move toward near-shoring. Temporary dips during the pandemic and rebounds after the Ukraine war suggest reactive, not structural, shifts.