A Riskier World: How Global Shocks Alter Economic Outlooks in Fragile Nations

In a bold step toward improving macroeconomic forecasting under uncertainty, the World Bank’s Economic Policy Global Department has released a new Policy Research Working Paper titled “Forecast Sensitivity to Global Risks: A BVAR Analysis” (May 2025). Authored by Heather Ruberl, Remzi Baris Tercioglu, and Adam Elderfield, and shaped by insights from the International Bank for Reconstruction and Development and research precedents such as those developed in the World Bank’s Uruguay country team, the paper introduces a rigorous method for quantifying how external macroeconomic shocks affect developing countries. Drawing on Bayesian Vector Autoregression (BVAR) and integrating it with the Bank’s macro-structural forecasting model MFMod, the researchers simulate the impacts of global risk factors such as oil prices, interest rates in advanced economies, remittances, and global demand on 115 developing economies. Their objective: to move beyond rigid point forecasts and offer policymakers a fuller understanding of potential economic trajectories under realistic global volatility.

Modeling Uncertainty with BVAR and MFMod

At the heart of the analysis is the BVAR model, a statistical framework well-suited to modeling the complex and often intertwined relationships between macroeconomic variables. The authors simulate 1,000 alternative global shock scenarios per country, feeding these scenarios into the World Bank’s MFMod model to generate distributions of possible economic outcomes for each country. Instead of predicting a single path for GDP or inflation, the model shows a spectrum of possible futures. The underlying variables in the BVAR include U.S. and Eurozone long-term interest rates, GDP of major economies (including the U.S., China, India, and the Eurozone), global oil prices, trade-weighted export demand, and commodity prices, alongside country-specific remittance flows. Each simulation reflects the historical joint distribution of global shocks, meaning the model draws on real-world past volatility, including the global financial crisis and the COVID-19 pandemic, to generate plausible near-term futures.

This setup enables the researchers to answer a pressing question in development forecasting: how much of a country’s forecast is sensitive to forces beyond its control? The answer, it turns out, is significant. In most years, global shocks shift GDP levels in developing countries by up to ±2% from baseline forecasts. In about 30% of years, this impact widens to ±4%. These fluctuations, while not necessarily catastrophic, can have meaningful effects on fiscal planning, monetary strategy, and external balances, especially for economies that lack deep financial buffers.

Country Cases: How Global Risks Play Out Differently

To illustrate the model’s practical value, the report zooms in on four countries: Poland, Angola, the Philippines, and The Gambia. Each represents a different level and type of exposure to global economic forces. In Poland, a highly open and trade-driven economy, the simulations show GDP growth uncertainty ranging from –3.6% to +10.3% annually. The most significant risks stem from export market demand, followed by import price fluctuations. These effects are amplified by Poland’s deep integration into global supply chains and capital markets.

Angola tells a very different story. As an oil-dependent exporter, its economic fate hinges heavily on oil prices. The simulations reveal that Angola faces more asymmetric risk, higher probability of strong upside gains when oil prices surge, but also steep fiscal and current account losses when they fall. Non-oil exports, being marginal, offer little buffer. Interestingly, the model also shows that the effects of individual shocks do not always sum neatly when combined, reflecting the complex interdependencies among variables like oil and other commodity prices.

In the case of the Philippines, the simulations show a more stable economic forecast with smaller uncertainty bands. This resilience is due to the country’s diversified economy and relatively consistent growth history, excluding COVID-related disruptions. While export market fluctuations and commodity prices still pose risks, they are less pronounced. In contrast, The Gambia’s economy is significantly shaped by remittance flows, which account for a large share of household income and external financing. Here, global shocks affecting the incomes of Gambians abroad, particularly in the U.S. and Europe, have a disproportionately large impact on domestic outcomes.

Why Traditional Forecasts Fall Short

What sets this paper apart is its challenge to the status quo in economic forecasting. Traditional models often rely on deterministic assumptions, fixed oil prices, stable interest rates, or smooth growth trajectories in advanced economies. While helpful for basic scenario planning, such models fail to account for the real-world volatility that policy planners must navigate. This BVAR-based approach introduces uncertainty not as a footnote, but as a central element of the forecast. It reflects the fact that most developing countries are “price takers” in global markets, with little control over key variables that shape their macroeconomic outcomes.

Limits, but Valuable Lessons for Policymakers

The authors are careful to note the limitations of their method. Since BVAR models are based on historical data, they cannot predict black swan events or structural breaks not seen in the past. Furthermore, the analysis includes only a defined set of global shocks, leaving out potentially important domestic risks or socio-political disruptions. The responses also depend on the internal structure of MFMod, which may impose certain linearities or behavioral assumptions that do not fully reflect the dynamism of real-world economies.

Despite these constraints, the value of this modeling approach is clear. It provides policymakers with a much-needed probabilistic understanding of the future, equipping them to prepare not just for the most likely scenario but for a range of possible outcomes. Whether it’s building up fiscal buffers, adjusting exchange rate policies, or targeting monetary interventions more precisely, this framework strengthens the toolkit for managing economic resilience. In an era of polycrisis and global interdependence, anticipating the plausible is more important than ever, and this paper provides a rigorous and practical way to do just that.