US core capital goods orders blow past expectations in March

U.S. core capital goods orders jumped 3.3% in March, well above the 0.5% forecast, while shipments rose 1.2%, signaling stronger business equipment spending in Q1. The report supports GDP tracking near 2.3% annualized growth, with AI-related capex cited as a key driver, though Iran-related oil and commodity price risks could curb future investment. The macro readthrough is important ahead of the Fed's two-day meeting and Thursday's GDP release.

Analysis

The signal here is not just “business spending is okay”; it is that the capex mix is becoming more concentrated in AI-adjacent compute rather than broad-based industrial demand. That is a positive for a narrow set of beneficiaries with pricing power and capacity constraints, but it also means the durability of the upswing is more fragile than the headline print suggests. If AI spend is doing most of the work, any pause in hyperscaler budgets would hit the marginal buyer much faster than in a normal upcycle.

For Goldman, the implication is less about a near-term earnings catalyst and more about a valuation floor: stronger growth and firmer rate expectations reduce the odds of a multiple unwind in financials, especially if the market starts pricing fewer cuts. The bigger second-order effect is on duration-sensitive growth names: if capex remains hot while rates stay higher for longer, the market may continue to favor “picks and shovels” AI infrastructure over long-duration software. That’s constructive for compute, networking, and power-chain vendors, but not for anything relying on easy financing or delayed monetization.

The geopolitical overlay matters because higher commodity prices can become a tax on the same capex cycle that is supporting GDP. A sustained oil move higher would pressure corporate margin guidance within one or two quarters and could finally force a reassessment of AI project ROIs outside the largest platforms. In other words, the current setup is good for growth-sensitive equities only if energy stays contained; if not, the market may rotate from “AI growth at any price” to “cash flow now,” which is a very different factor regime.

Contrarianly, the market may be underestimating how much of this is a timing effect from delayed data rather than a true acceleration in demand. That means the upside surprise could fade into a softer summer print, especially if tariffs, oil, or Fed hawkishness compress CFO confidence. The cleanest trade is to express the AI capex winner versus the broad macro lever, not to chase the macro print itself.

Key Decisions for Investors

• Long GS vs short IWM for 4-8 weeks: stronger growth and a still-higher-for-longer rates backdrop should support financials relative to small caps, with better earnings sensitivity on the long leg and weaker financing sensitivity on the short leg.
• Long SMCI or a SMCI/QQQ pair for 1-3 months: if AI capex remains the dominant driver, hardware vendors should outperform broad tech; use a tight stop because this is the most crowded expression and vulnerable to any hyperscaler capex pause.
• Buy APP on pullbacks for a 2-4 month horizon: if markets keep rewarding cash-generative AI-adjacent winners over pure software duration, APP remains a cleaner beneficiary than lower-quality growth names; risk/reward is attractive only on weakness, not strength.
• Short a basket of high-duration, unprofitable software names against long AI infrastructure for the next quarter: the data supports continued factor rotation toward tangible spend beneficiaries if rates stay elevated and growth holds.
• Set up a conditional hedge: buy XLE calls or energy exposure as a macro hedge against the AI-capex trade; if oil shock persists, it likely compresses corporate investment sentiment and is the cleanest way to offset a growth/capex long book.