Sweden’s central bank, the Riksbank, is embracing a transformative shift in economic forecasting, as recent findings from its Monetary Policy Department reveal that artificial intelligence (AI)-based models are outperforming traditional time series models in predicting Swedish GDP and inflation.
This development marks a significant leap forward in the application of advanced technology to monetary policy and economic analysis, offering new tools and insights that could reshape how central banks approach forecasting.
According to a survey conducted by Riksbank economists, AI models—particularly random forests and neural networks—demonstrate superior accuracy compared to conventional methods like autoregressive models and dynamic factor models.
The study, detailed in the report “AI-based forecasting in Sweden”, highlights the ability of AI to capture non-linear relationships, process vast and diverse datasets, and automate forecasting tasks.
These capabilities provide a clear edge over traditional models, which often struggle to adapt to complex economic dynamics or incorporate real-time data effectively.
The implications of this shift are profound.
Traditional time series models, while reliable in stable conditions, can fall short when faced with sudden economic shifts or unprecedented events.
AI, by contrast, excels at identifying patterns in large datasets and adjusting to new information, making it particularly valuable in an era of rapid global change.
For Sweden, a small, open economy sensitive to international trends, such adaptability is crucial for maintaining economic stability and informing monetary policy decisions.
The Riksbank’s findings suggest that AI-based models not only enhance the accuracy of GDP and inflation forecasts but also pave the way for broader integration into economic analysis.
As these models continue to evolve, they are expected to complement—and in many cases surpass—traditional approaches.
This could lead to more proactive and precise policy responses, helping the Riksbank navigate challenges like inflationary pressures, supply chain disruptions, or shifts in global demand.
Beyond the Riksbank’s research, Sweden’s economic landscape has been under scrutiny from international bodies.
In February, the International Monetary Fund (IMF) completed its annual Article IV consultation, a comprehensive review of Sweden’s economy.
The final report, now available on the IMF’s website, offers an external perspective on the country’s fiscal health, growth prospects, and policy framework.
While the IMF report does not specifically address AI forecasting, its insights into Sweden’s economic resilience and challenges provide a backdrop for understanding why advanced forecasting tools are increasingly vital.
Meanwhile, Sweden’s commitment to financial stability was evident in a separate development.
In September 2024, the Nordic-Baltic Stability Group (NBSG) conducted a large-scale financial crisis simulation involving nearly 450 participants.
Coordinated by Denmark’s Finansiel Stabilitet, the exercise tested the region’s ability to respond to systemic shocks.
The recently published report on the simulation underscores the importance of preparedness in an interconnected financial system—a context in which accurate forecasting, such as that enabled by AI, could prove invaluable.
As AI models become more refined and accessible, their role in economic forecasting is likely to expand, not just in Sweden but globally.
For the Riksbank, this technological advancement offers a powerful tool to enhance decision-making, ensuring that monetary policy remains agile and progressive.