Since 1815 The Dollar Has Changed Its Anchor Twice, This Chart Suggests That A Third Shift Is Underway

The U.S. is positioning artificial intelligence as a potential long-term anchor of dollar strength, comparable to the historic roles of gold and oil. The strategy centers on maintaining control of AI-critical data centers, semiconductors, cloud platforms and payment systems, while extending U.S. influence across global technology supply chains. Its success could reinforce U.S. technological leadership and the dollar’s international role, though the article provides no near-term financial metrics or policy actions.

Analysis

The investable expression is not a near-term USD trade: reserve-currency demand is determined primarily by Treasury-market depth, real-rate differentials, fiscal credibility and cross-border funding needs. AI-related capital expenditure can reinforce U.S. growth and equity inflows, but it will not offset a sharp rise in Treasury term premium or a Fed easing cycle. Treat broad dollar strength as a second-order, 6-18 month possibility rather than a direct implication of AI infrastructure leadership.

The nearer earnings sensitivity sits in the physical bottlenecks: NVDA/AVGO compute, VRT/ETN power and thermal management, CEG/VST generation, and EQIX/DLR data-center capacity. The key second-order constraint is power availability rather than chips; interconnection delays can defer data-center revenue while simultaneously increasing the value of contracted generation and grid equipment. That favors a barbell of semiconductor exposure and electrification infrastructure over software names whose AI monetization remains less measurable.

Consensus may be underpricing the capital-intensity hangover. Hyperscaler spending supports suppliers over the next 1-3 quarters, but if enterprise AI revenue fails to catch up during 2027 budget planning, depreciation, energy and lease costs could pressure margins at MSFT, AMZN, GOOGL and META despite continued top-line cloud growth. This is particularly relevant if AI capacity is built ahead of utility transmission upgrades, producing stranded or underutilized capacity in specific regions.

Falsification points: a meaningful reduction in hyperscaler capex guidance, sustained easing in GPU lead times, or utility load forecasts being revised down would weaken the infrastructure thesis within days. Conversely, incremental long-term power contracts, higher data-center preleasing, and upward revisions to grid-capex plans would extend the cycle through 2027.

Key Decisions for Investors

• Favor a 3-6 month pair trade long VRT and ETN versus short an equal-dollar basket of MSFT/GOOGL: the pair isolates power-density and cooling bottlenecks from broad AI enthusiasm. Target 15-20% pair upside; exit if either hyperscaler cuts aggregate 2027 capex guidance by more than 10%.
• Accumulate CEG or VST on pullbacks for a 6-18 month horizon, but only where incremental contracted data-center load is independently disclosed. The upside is repricing of long-duration power scarcity; the principal risk is regulatory intervention, falling power prices, or delayed load connection.
• Maintain core NVDA exposure rather than chase a broad AI-software basket into the next earnings cycle; hedge with 3-6 month SMH puts if semiconductor lead-time data normalize. This preserves exposure to sustained accelerator demand while limiting downside from a capex-reset narrative.
• Do not initiate a structural long DXY solely on AI themes. Use a USD long only tactically around widening U.S. real-rate differentials or improving capital-flow data; a declining rate differential or rising Treasury term premium would invalidate the setup.