Trump’s War Jolts Global Central Banks From Fed to ECB to BOJ

An ECB blog post reports that artificial intelligence has so far had no negative impact on euro‑zone employment; the heaviest users of AI have actually added staff. The finding is descriptive with no quantified magnitudes and is unlikely to move markets immediately, though it could inform future ECB policy discussions if the trend persists.

Analysis

Heavy AI adoption is creating a complementary labor market rather than replacing it: firms spending on models are concurrently expanding MLOps, data-engineering and product teams, which lifts demand for cloud compute, systems integrators and specialist training services. Expect hyperscaler cloud revenue tied to AI workloads to grow materially faster than baseline (+20–40% YoY in active AI projects over 6–24 months in our scenario), with outsized capex and service uptake concentrated at a handful of providers and chip vendors.

Second-order supply-chain winners are semiconductor equipment and memory suppliers that enable scale-out model training — more racks of GPUs means more wafers, more EUV runs and higher ASML/TSMC/ Micron utilization; conversely, legacy on-prem software vendors that don’t pivot to MLOps risk margin compression. At the firm level, large consultancies and systems integrators capturing integration, deployment and retraining spend will expand revenue per client faster than classic staffing firms, which face increasing competition from reskilling platforms.

Key catalysts that can reverse the trend include a meaningful step-function improvement in foundation models that sharply reduces compute per unit of value (months–years) or swift regulatory responses that tax AI automation. Macro angle: persistent skill-driven wage inflation in node clusters could keep regional central banks more cautious — a multi-quarter tail that compresses valuation multiples for long-duration software names.

Consensus underweights the persistence of high-skill labor demand and the resulting skew to capital goods (fabs/EUV) and services (integration, training). Monitor cloud GPU utilization, job postings for MLOps, and hyperscaler capex guides as high-signal indicators to time positions.

Key Decisions for Investors

• Long NVDA via a 12–18 month call spread (buy Jan-2027 $700 calls, sell Jan-2027 $1,200 calls) — allocate 1–2% NAV. Rationale: capture continued GPU demand for model training while financing premium; target 2.5x nominal premium, stop if NVDA falls 25% from entry.
• Buy ACN (Accenture) stock, 6–12 month horizon — overweight systems-integration exposure to AI deployments. Target +20–30% upside as per-client AI services expand; set tactical stop-loss at -10% to protect against macro multiple compression.
• Pair trade: long COUR (Coursera) and short MAN (ManpowerGroup), equal-dollar, 12–24 month horizon — reskilling platforms should outgrow traditional staffing if firms internalize AI hires. Target relative return of +40% (COUR +50% and MAN -10%); exit if pair moves against by 15%.
• Buy ASML (ASML) on pullbacks, 12–24 months — exposure to sustained wafer fab capex driven by AI GPU demand. Position size 1–2% NAV; upside target 15–25%, stop-loss 12% to limit cycle risk.