Anatomy of an AI News Shock in Global Markets
A new working paper from Dutch central bank researchers isolates AI news shocks by combining daily equity factors with narrative restrictions in a Bayesian vector autoregression. The study traces how revisions in AI's expected economic value flow through United States and euro area economies, using sign, magnitude, and narrative filters to separate AI-driven moves from broader financial noise.
Distinguishing AI news from general market moves
The authors construct a daily equity factor that captures differences in firms' AI exposures. By applying sign and magnitude constraints alongside narrative descriptions of AI developments, they identify revisions to AI's expected economic value. This method allows them to distinguish AI-specific news from general risk-on risk-off swings that often dominate daily market headlines.
US investment-led adjustment takes hold first
Monthly local projections show that in the United States, positive AI news shocks trigger an early and persistent rise in capital-goods orders. Broader real activity responds more gradually and with greater uncertainty. The investment response arrives before hiring or output expansions, suggesting that firms commit to AI capital before they see corresponding revenue gains.
Price pressures move from producer to consumer
Producer prices increase before consumer prices after an AI news shock. This lead-lag pattern indicates that investment and resource-demand pressures emerge in the supply chain ahead of final retail-level price changes. The delay reflects the time needed for AI-integrated inputs to propagate through production cycles and reach end users.
Spread from AI-network sectors to the wider economy
Initially, the effects concentrate in sectors closely connected to the AI production network—semiconductor fabrication, data-center construction, specialized software firms. Over subsequent months, the expansion ripples outward to manufacturing, logistics, and other broader industries. This diffusion pattern mirrors how general-purpose technologies typically take hold in an economy.
Euro area: equity gains, currency depreciation, import costs first
In the euro area, AI news shocks produce a different initial pattern. Equity prices rise, and the euro depreciates. Import costs and energy prices feed into inflation before real activity strengthens. Inflation expectations and nominal interest rates also increase. The sequence reflects the region's openness to global supply chains and its reliance on imported technology inputs.
Monetary policy dilemma arrives before productive gains
The findings highlight a timing problem for policymakers. Financial spillovers and imported price pressures from AI news arrive before the broader productive benefits of AI adoption become visible in GDP or employment data. This window creates a constraint: tightening policy to counter early inflation could dampen the very investment that later generates productivity growth.
Practical takeaways for teams tracking AI's economic impact
For analysts and investment teams, the paper suggests watching capital-goods orders and producer-price indices as leading indicators of AI's macro effect. In Europe, currency moves and import-cost data offer similar early signals. The international spillover channel means that AI news from the United States does not stay contained—euro area variables adjust ahead of domestic real activity. Tracking these sequences helps separate signal from the daily noise of AI hype cycles.