2026 Industrials Outperformance Forecast Driven by AI Tech CAPEX and Infrastructure

This analysis is based on a December 18, 2025 Seeking Alpha article [7] predicting industrial ETFs XLI and VIS will outperform the S&P 500 in 2026, driven by AI-related CAPEX and data center infrastructure needs. As of the event date, the U.S. industrials sector showed a -0.22% daily change, underperforming consumer defensive but outperforming technology and utilities sectors [0]. The core driver identified is the surge in data center power demand—projected to triple U.S. power requirements by 2030 (aligning with upper-end industry forecasts) [4][5]. This demand will accelerate spending on grid infrastructure, industrial equipment, automation systems, HVAC solutions, and related services.

UBS forecasts global AI CAPEX will reach $571 billion in 2026 [1], with Bloomberg highlighting that the AI boom is extending the industrials demand cycle [3]. Data center energy consumption projections vary: the Washington Post reports 60-150% growth by 2030 [4], while Forbes projects data centers will use 9-12% of U.S. electricity by 2030 (up from ~4% in 2024) [5]. This trend benefits industrial subsectors including electrical equipment (grid solutions), HVAC, automation, and construction. Key players exposed to this demand include Eaton Corporation (ETN, electrical systems), Johnson Controls (JCI, HVAC), and Rockwell Automation (ROK, automation). XLI’s top holdings (GE Aerospace, Caterpillar, RTX, Uber) have varying exposure: RTX’s Collins Aerospace provides data center cooling, and Caterpillar’s construction equipment supports facility builds [6].

1. Cross-sector convergence
   : The AI tech boom is directly driving industrial demand, marking a shift from traditional infrastructure to tech-enabled industrial spending. This convergence creates new client segments (tech hyperscalers) and expands opportunities for firms with energy transition capabilities.
2. Demand cycle extension
   : Despite short-term concerns about AI CAPEX peaks [2], Bloomberg’s analysis indicates the AI boom will extend the industrials demand cycle [3], providing medium-term support.
3. Competitive landscape shifts
   : Specialized firms with data center-focused solutions may lower entry barriers, while traditional industrial players will need to expand digital infrastructure offerings. M&A activity could accelerate as firms acquire AI-enabled industrial capabilities.
4. Policy and regulatory implications
   : Grid modernization (identified as a priority by PJM Interconnection and ERCOT [5][8]) and stricter energy efficiency regulations for data centers will shape industrial demand trajectories.

• Growth in AI-related CAPEX (reaching $571 billion in 2026 [1]) driving demand for industrial goods and services.
• Grid modernization investments to support data center power needs.
• Expansion of energy-efficient HVAC and automation solutions for data centers.
• Partnerships between industrial firms and tech hyperscalers.

• Supply chain instability for components like transformers and semiconductors could hinder meeting increased demand.
• Regulatory uncertainties around energy transition and data center energy efficiency.
• Slower-than-expected AI adoption could dampen CAPEX growth.
• Short-term peaks in AI CAPEX [2] may lead to demand volatility.

The analysis highlights a potential outperformance opportunity for U.S. industrial sector ETFs (XLI, VIS) in 2026, driven by AI-related CAPEX and data center infrastructure demands. Data center power consumption is projected to surge significantly by 2030, requiring accelerated grid modernization and industrial equipment spending. Key industrial subsectors and players exposed to this trend include electrical equipment (Eaton), HVAC (Johnson Controls), and automation (Rockwell Automation). The AI boom is expected to extend the industrials demand cycle, though risks related to supply chains, regulations, and AI adoption rates remain. Stakeholders should monitor AI CAPEX trends, infrastructure policy, and energy transition progress for context.