US companies have reduced new investment in China and are increasingly taking earnings out of existing Chinese operations, according to a Federal Reserve staff analysis published on 25 September. The research finds a broad pullback, but does not assign it to one cause or forecast that US businesses will leave China altogether.
The note examines foreign direct investment, or FDI, which is investment that gives a company a lasting interest in an overseas business. It includes building a new facility, acquiring a foreign company and reinvesting profits in an existing subsidiary.
Its central finding is that the slowdown appears across several measures: greenfield projects, acquisitions and capital spending by Chinese subsidiaries of US multinationals.
Why the usual bilateral figures miss part of the picture
Official statistics show direct investment from the United States to China. But they can miss money routed through Hong Kong and other investment hubs before it reaches an operation in mainland China.
Federal Reserve economist Cody Kallen combined official US data with project, deal and subsidiary information to look through those intermediate holding companies. The analysis tracks developments through 2025 for several measures, although some subsidiary data end earlier.
The broader measure shows that US companies’ exposure to China rose until 2020 before beginning to fall in real terms and as a share of US outward investment. That is suggestive evidence of a change in business links, not a count of factories that have closed.
Our explanation of the difference between direct-investment positions and annual flows sets out why an accumulated ownership position should not be confused with a single year’s new spending.
New projects and acquisitions have weakened
The note finds that US-announced greenfield projects in China were around 300 a year between 2003 and 2013. They fell in the middle of the 2010s, dropped sharply during the Covid pandemic and did not recover through 2025. The decline was broad across high technology, advanced manufacturing, other manufacturing and services.
US acquisitions of Chinese companies also fell notably in 2022 and remained low, the research says. The analysis includes acquisitions routed through non-US intermediaries whose ultimate parent is American, reducing the risk that a corporate structure hides the buyer’s nationality.
Deal values tell a slower-moving story. They did not fall until 2024, partly because large acquisitions can take time to complete and can dominate the total value in a given year.
Existing operations are retaining less profit
The analysis also examines the choices made by US companies that already have Chinese subsidiaries. Since 2015, dividend payouts from those operations have overtaken reinvested earnings, reversing the earlier pattern in which companies retained more profit locally for future expansion.
Capital-investment rates have declined, while the shares of subsidiaries reporting asset sales or negative growth in property, plant and equipment have risen. The note describes those trends as signs of a pullback in US multinationals’ Chinese operations.
A dividend payout is not automatically a divestment. A company may send cash home for many reasons. The significance lies in the direction of several indicators together, rather than any one financial measure.
Several forces may be involved
The note links the timing of the change to trade tensions after 2018, the Covid pandemic and geopolitical risks after Russia’s invasion of Ukraine. It also identifies investment screening and other barriers as possible factors.
It does not attempt to separate the effect of each one. That matters because a company may be responding to tariffs, supply-chain resilience, weaker demand, regulation or a combination of those issues. The evidence documents a pattern of investment fragmentation; it does not prove one explanation for every firm’s decision.
For businesses, the shift can mean that China remains an important market while new production, research and supplier investment is spread more widely across countries. For policymakers, it raises questions about how trade policy, security rules and investment decisions are changing long-standing corporate ties.
The analysis is a Federal Reserve staff note by Cody Kallen. It reflects the author’s analysis and does not represent a Federal Reserve policy statement.
