McKinsey chairs: Building a more resilient industrial base may require $2 trillion in investment

US imports of manufactured goods total ~$3T annually, with ~25% tied to at least two trade vulnerabilities and ~5% exposed to all three. Replacing more sensitive future-shaping goods (e.g., semiconductors, data center servers) would require production capacity to quintuple or more and roughly $2T of capacity-creating investment, versus ~$660B of potential output for less sensitive categories (about half the trade deficit). The article argues ramp-up will need coordinated upgrades across energy, workforce, permitting, and supply-chain visibility, with early-acting firms better positioned as the economy shifts from cost-led to resilience-led manufacturing.

Analysis

The market is likely to misprice this as a generic industrial-onshoring story, when the real trade is in bottlenecks: power delivery, factory automation, process controls, and supply-chain software. Those businesses can re-rate earlier than steel, chemicals, or broad manufacturers because they monetize the first wave of redesign work and capex planning, not just finished output. The second-order effect is a sustained lift in orders for grid gear, transformers, cooling, and robotics while import-heavy assemblers face margin pressure from higher domestic input costs.

Time horizon matters. Over the next 1-3 months this is mostly narrative support, so the best expression is relative value rather than outright beta; the earnings impact only shows up if capex commitments turn into purchase orders and permits. Over 6-18 months, the structural winners are the semiconductor equipment, electrical infrastructure, and industrial automation ecosystems; the losers are low-value-added import intermediaries and offshore contract manufacturers whose pricing advantage erodes as customers diversify sourcing.

Contrarian view: the consensus may be overestimating how much production can be reshored quickly and underestimating how much of the spend will leak into non-manufacturing enablers. If rates stay restrictive or permitting/workforce constraints persist, the headline reshoring thesis will stay aspirational and the valuation premium could fade. What would falsify the trade is a lack of follow-through in capex guidance, power-grid order growth, and semiconductor tool demand over the next two quarters.

Key Decisions for Investors

• Long PAVE / short XRT for 3-6 months as a relative-value expression of domestic infrastructure-capex beneficiaries versus import-dependent consumption; target a 5-7% spread, cut if capex data and industrial order flow fail to inflect by the next two monthly prints.
• Add to BOTZ or ROBO on pullbacks for a 6-18 month horizon; this is a cleaner way to own the AI/robotics layer that makes reshoring economically viable. Risk is that automation spend gets delayed if credit conditions tighten or manufacturers defer projects.
• Prefer an enablers basket (ETN, EMR, HUBB, PWR) over broad industrials if entering new exposure; these names should capture the earliest margin-accretive orders from grid, controls, and construction bottlenecks. Falsify if order growth does not accelerate despite public reshoring announcements.