What's happening
US core capital goods orders jumped 3.3% in March 2026, delivering the strongest monthly gain since June 2020 and far surpassing economist expectations of a 0.5% increase. The surge was led by computers and electronic products, which shot up 3.7% as businesses accelerated investments in AI infrastructure and data center equipment. Core capital goods shipments rose 1.2% in March after climbing 1.3% in February, indicating sustained business equipment demand.
Overall durable goods orders rebounded 0.8% in March after dropping 1.2% in February. The strength in capital goods orders contrasted with mixed housing data, where single-family housing starts jumped 9.7% to a seasonally adjusted annual rate of 1.032 million units, but permits for future construction fell 3.8% to 895,000 units. The goods trade deficit widened to $87.9 billion as imports increased $9.6 billion to $299.3 billion while exports rose $5.2 billion to $211.5 billion.
Why it matters for markets
The 3.3% surge in core capital goods orders provides concrete macroeconomic validation of the AI investment boom that has driven technology valuations higher. Stephen Stanley, chief U.S. economist at Santander U.S. Capital Markets, noted that "the stunning degree of strength during a month when firms would have had valid reason to be cautious attests to the substantial energy in business investment that was bottled up last year due to policy-related uncertainty." The 3.7% jump in computers and electronic products orders specifically signals robust demand for the semiconductors, servers, and networking equipment powering AI data centers.
This data arrives as AI hardware leaders trade at elevated valuations, with NVIDIA commanding a 42.7 price-to-earnings ratio on its $5.09 trillion market capitalization and Broadcom trading at 79.2 times earnings with a $1.92 trillion market cap. The strong orders data supports the thesis that enterprise AI spending can justify these premium valuations, particularly as hyperscalers like Amazon Web Services and Microsoft Azure expand their infrastructure to meet AI compute demand.
The sustained momentum in capital goods shipments, rising 1.2% in March after 1.3% in February, suggests the AI capex cycle has legs beyond the initial rush. This sequential strength indicates businesses are moving from planning to actual deployment of AI infrastructure, creating a more durable revenue stream for equipment suppliers than one-time purchasing spikes.
Sectors and assets to watch
Semiconductor companies stand to benefit most directly from the AI infrastructure buildout driving capital goods orders. NVIDIA, with $215.94 billion in revenue and dominant positions in AI training chips like the H100 and upcoming Blackwell architecture, remains the primary beneficiary of data center GPU demand. Broadcom, generating $68.28 billion in revenue from networking chips and custom ASICs for hyperscalers, provides the connectivity infrastructure essential for AI workloads.
Hyperscale cloud providers are the primary purchasers driving the capital goods surge. Amazon Web Services, part of Amazon's $716.92 billion revenue base, and Microsoft Azure, contributing to Microsoft's $305.45 billion in revenue, are expanding their data center footprints to support AI services. These companies translate the capital equipment investments into recurring cloud revenue, making them both beneficiaries and drivers of the AI infrastructure cycle.
What to watch next
Monitor April core capital goods data to confirm whether the 3.3% March surge represents sustained AI investment momentum or a one-time catch-up from delayed projects. Key leading indicators include semiconductor billings data, hyperscaler capital expenditure guidance in upcoming earnings reports, and data center construction permits. Ben Ayers from Nationwide expects some near-term economic uncertainty, noting potential pullbacks in construction activity, which could affect the broader capital goods outlook if businesses become more cautious about major infrastructure investments.