Hassett forecasts 4% growth, says AI boom and tax incentives driving US investment surge

Kevin Hassett said he is highly confident the U.S. economy will print around 4% growth for the rest of the year, citing an AI productivity boom, rising capital spending and factory construction. He also said tax incentives such as full expensing and bonus depreciation are fueling a rush to invest in U.S. manufacturing before key provisions expire. The comments are broadly supportive of growth-sensitive sectors and reinforce a pro-investment, pro-capex policy backdrop.

Analysis

The market is likely underpricing how much of this “growth boom” is really a capex cycle, not a classic demand-led expansion. That matters because capex-led upturns tend to concentrate gains in a narrow set of beneficiaries first: semiconductor equipment, industrial automation, power/cooling, and logistics nodes near major builds. The second-order effect is a widening performance gap between firms with domestic manufacturing exposure and those relying on consumer demand or global trade volumes.

The most interesting implication for TSM is less about the headline demand for wafers and more about its strategic pricing power in an environment where customers are racing to secure capacity. That supports utilization and contract discipline, but the nearer-term risk is that the market has already capitalized part of this AI-premium, so upside likely comes from estimate revisions rather than multiple expansion. For NVS, the signal is broader U.S. localization: companies with balance sheets and U.S. manufacturing footprints can turn policy incentives into margin support, while peers with less domestic capacity face a relative cost disadvantage.

The key contrarian risk is timing. A capex boom can look like GDP acceleration before it shows up in cash flows, and if financing conditions tighten or policy expectations shift, some projects get deferred without necessarily destroying long-run demand. The cleanest tell over the next 1–3 months is whether industrial order data, AI infrastructure capex guides, and semiconductor lead times keep improving; if not, the narrative can roll over quickly even if macro prints stay strong.

Consensus is likely too linear on the “4% growth” framing and too complacent about crowding in the AI trade. If the AI buildout broadens, the winners should migrate from the obvious mega-cap tech names into picks-and-shovels industrials and power infrastructure; if it narrows, the trade becomes more vulnerable to valuation compression. That creates an attractive relative-value setup rather than a simple outright risk-on call.

Key Decisions for Investors

• Long TSM into the next 1-2 earnings cycles; use pullbacks of 5-7% to add. Thesis: capacity scarcity plus AI-driven demand should protect revenue revisions, but trim if management commentary suggests customers are pulling forward less volume than expected.
• Long NVS vs. a basket of global pharma with weaker U.S. manufacturing exposure over the next 3-6 months. The trade benefits if U.S. localization incentives translate into durable margin support and lower effective tax friction.
• Pair trade: long semicap / industrial automation basket, short broad consumer discretionary index for 2-4 months. Capex-led growth should hit suppliers before household demand, creating a divergence in earnings momentum.
• Buy 3-6 month call spreads on TSM rather than outright calls to reduce valuation risk. The catalyst is estimate revision, not multiple re-rating; cap gains if the market continues to price in AI capacity scarcity.
• Watch for reversal trigger: if U.S. industrial production and AI capex guideposts stop accelerating within 4-8 weeks, reduce exposure to the AI/investment complex by 25-50%.