Citi's research points to Chinese optical module makers as the clearest beneficiaries of a 79 percent surge in US cloud vendor capital spending.
Citi's research points to Chinese optical module makers as the clearest beneficiaries of a 79 percent surge in US cloud vendor capital spending.

Citi's research points to Chinese optical module makers as the clearest beneficiaries of a 79 percent surge in US cloud vendor capital spending.
US cloud vendors' second-quarter capital expenditure jumped 79 percent year-over-year, with full-year guidance raised and demand visibility extended to 2028, creating a structural tailwind for Chinese optical module suppliers, Citi said.
The four hyperscalers — Alphabet, Amazon, Microsoft, and Meta — combined for $182 billion in second-quarter capex, with cloud revenue growth accelerating to 48 percent year-over-year from 39 percent in the prior quarter, according to Goldman Sachs' mid-earnings-season report. Analysts now project hyperscaler capex to exceed $1 trillion by 2027, an upward revision of more than $125 billion from prior forecasts.
Citi identifies Chinese optical communication companies as structural beneficiaries of this spending cycle. The bank suggests the buying window may open around 2026 first-half earnings disclosures, when supply chain companies report their order books and margin trajectories. Near-term events to watch include the China International Optoelectronic Expo (CIOE) in September, Alibaba Cloud's Yunqi Conference, and the European Conference on Optical Communication (ECOC).
The stakes are significant for the AI infrastructure supply chain. AI infrastructure-related stocks accounted for roughly one-third of the S&P 500's second-quarter EPS growth, with Alphabet, Amazon, Micron Technology, and Nvidia as the top contributors. Goldman Sachs maintains its S&P 500 target of 8,000 for end-2026, with a 2027 EPS forecast of $385.
The raised guidance from US cloud vendors is notable for its duration. Demand visibility now extends to 2028, which means optical module suppliers can plan capacity expansion with greater confidence. This is particularly relevant for Chinese manufacturers, which have been expanding production capacity to meet AI data center demand.
The capex-to-cash-flow gap is widening. In the second quarter, the four hyperscalers (including Oracle) disclosed $182 billion in capex against roughly $5 billion in free cash flow, and raised about $101 billion through debt and equity. Goldman Sachs credit strategists expect these companies to issue about $400 billion in global investment-grade bonds next year to support continued AI infrastructure expansion.
Management teams outlined their financing strategies during earnings calls. Alphabet said it has significantly expanded its debt portfolio and recently conducted equity financing to maintain a strong balance sheet. Meta stated it is evolving its capital structure, increasing leverage and diversifying long-term funding sources including collaboration with BlackRock. Amazon said it will flexibly evaluate financing instruments as business grows. Oracle disclosed plans to raise about $40 billion in fiscal 2027 through debt and equity, including a previously announced $20 billion immediate equity raise.
While the demand outlook is strong, cost pressures remain a key risk. The median S&P 500 stock's net margin has been essentially flat over recent quarters, and consensus forecasts for third-quarter margins have been revised downward in most sectors. For optical module companies, the key question is whether pricing power can offset rising input costs as capacity scales.
Citi's timing suggestion — buying around 2026 first-half earnings disclosures — reflects the historical pattern where supply chain companies often see share price volatility around earnings season even when fundamentals are strong. The upcoming CIOE and ECOC exhibitions could provide early visibility into order momentum and technology roadmaps.
For investors, the trade is straightforward: Chinese optical module makers offer direct exposure to the AI infrastructure buildout at a time when hyperscaler capex is accelerating. The sector's fortunes are tied to the $1 trillion-plus capex trajectory through 2027, with Nvidia's August 26 earnings report serving as the next major directional event for the AI trade. Nvidia, the S&P 500's largest company yet to report results, could offer further clarity on demand for AI infrastructure components.
This article is for informational purposes only and does not constitute investment advice.