Decoupling Direct Investment: American Firms’ Retreat from China

September 25, 2026

Decoupling Direct Investment: American Firms' Retreat from China

Cody Kallen

For decades, China was a major destination for foreign direct investment (FDI) by U.S. companies. In the last decade, however, trade tensions, geopolitical risks, and investment screening have interrupted these FDI flows and led firms to shift production toward third countries (Reyes-Heroles et al., 2020). Recent research has documented preliminary signs of FDI fragmentation-the reorientation of investment away from geopolitically or geographically distant countries-worldwide, and with emphasis on the relationship between the U.S. and China (Kallen, 2025; Aiyar et al., 2023; Gopinath et al., 2025). By comparison, this note focuses specifically on U.S. investment in China and comprehensively examines different forms of investment.

Official bilateral FDI statistics show a modest reduction of U.S.-to-China FDI flows. However, these official statistics understate the true exposure of U.S. firms and investors to China due to investment routing through Hong Kong and other investment hubs. To correct for this mismeasurement and provide a comprehensive picture of US investments into China, I complement the official data with data on individual FDI projects, deals, and subsidiary investments. This approach allows us to look through investment hubs, decompose FDI by type, and track developments through 2025.

I find broad-based declines in U.S.-to-China greenfield FDI, U.S. acquisitions of Chinese firms, and the capital investments by Chinese subsidiaries of U.S. multinationals. Moreover, U.S. multinationals are reducing their existing footprints in China, with rising dividend payout rates (retaining less earnings) and increasingly reporting sales of physical capital or negative capital stock growth. These developments have begun to reverse decades of deepening integration between the U.S. and China, especially through the Chinese operations of American companies.

Data

This analysis draws on five sources of data on foreign direct investment. The Bureau of Economic Analysis (BEA) provides official measures of bilateral financial flows for FDI, FDI positions, and income from existing FDI. Although these official bilateral measures are timely, measurement of these flows can be biased by the role of investment hubs, as direct investments made in China that are routed through a subsidiary in Hong Kong or the Cayman Islands would not appear as U.S.-to-China FDI. Accordingly, I also draw on BEA data on the Activities of U.S. Multinational Enterprises (AUSMNE), which provides details on the production and investment activities of U.S. multinationals and their majority-owned foreign affiliates, coming from a survey that allows us to look past these intermediate subsidiaries.

In addition to these aggregate data sources, I draw on three types of FDI-relevant microdata and disaggregated data. For greenfield FDI-where a foreign firm constructs new facilities and operations-I use data on the number of announced investment projects from fDi Markets. I analyze cross-border M&A using microdata from LSEG Data and Analytics. Finally, to analyze the physical investments of Chinese subsidiaries of U.S. multinationals, I use firm-level microdata from Capital IQ, identifying firms as subsidiaries of U.S. multinationals using the S&P Business Entity Cross-Reference Service. Notably, for each of these data sources, we can identify American firms investing in China regardless of the use of any intermediate affiliates in investment hub countries.

Financial Flows from the U.S. to China

Figure 1 presents the bilateral financial flows from the U.S. to China for foreign direct investment (red line), which have shifted lower since the Covid pandemic. Because many of these investments are implemented using subsidiaries in Hong Kong, I also include a measure combining China and Hong Kong (black line). The left panel shows that these flows, while volatile, have declined in recent years, although flows to Hong Kong have recovered more than flows directly to China.

The right panel shows the existing U.S.-to-China (and to China and Hong Kong) FDI positions. The blue line presents the outstanding net equity of majority-owned Chinese subsidiaries of U.S. multinationals; this captures the equity of all these majority-owned subsidiaries, regardless of whether they have an intermediate parent in another country (e.g., Hong Kong), net of equity in other affiliates of the multinational group. This measure reveals that the official U.S.-to-China bilateral FDI position substantially understates the true exposure of U.S. multinationals to China, and the combined position for China and Hong Kong is a better proxy. These positions rose until 2020 before beginning to decline-in real levels and as shares of total U.S. outward FDI positions-suggestive but not conclusive evidence of the decoupling of FDI between the U.S. and China.

By comparison, Chinese data on FDI utilization by country suggest that the U.S. declined from over 10 percent of all FDI into China in 2002 to 2 percent in 2011 and remained low thereafter. However, this measure can miss cross-border M&A and reinvested earnings, which have become the primary forms of U.S. FDI into China.

Subcomponents of U.S.-to-China FDI

Foreign direct investment can be decomposed into new FDI-greenfield FDI and cross-border mergers and acquisitions-and the reinvestment activities of multinationals' existing foreign affiliates.

For the subcomponents of FDI, I first investigate the number of greenfield FDI projects in China announced by U.S. investors, shown in Figure 2. Between 2003 and 2013, these were consistently high, around 300 projects per year. These greenfield FDI projects declined in the mid-to-late-2010s, which may reflect growth issues in 2015 and 2016, before dropping off sharply in 2020 amid Covid and never recovering. Figure 2 also splits these projects into industry groups, consisting of high tech, advanced manufacturing (besides high tech), other manufacturing, and all other industries. The drop-off during and after the Covid pandemic was broad-based across all these sectors.

Although the level of greenfield FDI declined worldwide in 2020, the dropoff in FDI between the U.S. and China was more severe and persistent than elsewhere. The second panel of Figure 2 shows that U.S.-to-China greenfield FDI projects also declined as a share of outward U.S. greenfield FDI and as a share of all greenfield FDI into China. In sum, the U.S. has become a less important source of investment into China, and China has become a less important destination for U.S. investment.

Turning to cross-border M&A, Figure 3 presents the number of deals (left panel) and the aggregate value of deals (right panel). For each deal, LSEG Data and Analytics identifies the countries of the target and the acquirer, as well as the countries of their ultimate parent companies. I define direct acquisitions as cases where a U.S. company acquires a Chinese company. I define hidden acquisitions as cases where a Chinese company is acquired by a buyer not in the U.S. that has an American ultimate parent company; these deals would not appear in bilateral financial measures of U.S.-to-China FDI. I also identify divestments as cases where the target of the deal is Chinese, but with an American ultimate parent divesting from it (typically to a Chinese buyer).

After the Great Recession, U.S. acquisitions of Chinese companies began to decline in number. Between 2016 and 2021, this trend reversed, primarily reflecting a U.S.-led M&A boom. In 2022, the numbers of U.S.-to-China acquisitions-both direct and hidden-dropped notably and have remained low. This drop cannot be explained the global M&A cycle, as U.S. acquisitions of Chinese target firms have declined relative to all foreign acquisitions of Chinese targets and relative to the foreign acquisitions of U.S. investors. Although the number of divestments did not rise as might be expected with the decline of new investment, this may reflect potential difficulties in obtaining approval for divestments.

M&A values show greater volatility and did not drop until 2024, lagging behind deal volume. This reflects the time between deal announcement and completion, particularly for large deals that account for most of the aggregate values.

Beyond new investments, I also examine how U.S. firms are managing the returns from their existing Chinese operations. For U.S.-to-China FDI, BEA data through 2023 (discontinued for subsequent years) reported the income from existing investments and reinvested income, with the difference representing dividends and other distributions to U.S. parents. Figure 4 reports the three-year moving average reinvestment rate (reinvested earnings as a share of total income) and dividend payout rate. Between 2010 and 2015, the reinvestment rate rose and the dividend payout rate fell, as U.S. investors retained more of their income in China to invest. Since 2015, however, these have reversed, with dividend payouts overtaking reinvestments in recent years. By comparison, excluding China and investment hubs, reinvestment rates dropped in 2018 but have since recovered to their historical norm near 50 percent.

U.S. Multinationals' Capital Investments

Finally, I investigate the capital investments of Chinese subsidiaries of U.S. multinationals. The left panel of Figure 5 presents the moving average mean and median of these subsidiaries' capital expenditures, scaled by their existing property, plant and equipment. For comparison, I report the same ratio constructed from aggregate data on capital expenditures and PPE for China in the BEA's Activities of U.S. Multinational Enterprises data (excluding 2009 due to a change in the set of included industries).

These measures of capital expenditures broadly tell a similar story. Capital investment rates were strongest in the 2000s, but have decreased in recent years.

The right panel of Figure 5 explores indicators of potential divestment by Chinese subsidiaries of U.S. multinationals, measured as the shares of firms reporting sales of PPE or negative PPE growth. The share of Chinese subsidiaries reporting sales of PPE has gradually risen. The share reporting negative PPE growth has been more volatile, rising in 2015 and 2016 amid Chinese growth concerns and again since 2021.

Taken together, declining capital expenditure rates and rising indicators of divestment suggest a pullback in U.S. multinationals' operations in China.

Conclusions

This note provides the most comprehensive documentation to date of FDI fragmentation between the U.S. and China. Across multiple data sources and forms of FDI-official bilateral statistics, greenfield projects, cross-border M&A, and subsidiary-level financial data-I document broad evidence of declining U.S.-to-China FDI, especially since U.S. tariffs on China in 2018, Covid in 2020, and the Russian invasion of Ukraine in 2022. These patterns reveal both a slower pace of new investment and the active strategic reduction of American multinationals' Chinese operations, reversing decades of deepening economic integration.

While these patterns provide clear evidence of FDI fragmentation between the U.S. and China in recent years, these results do not distinguish between their potential drivers: trade tensions, geopolitics, or investment screening and other barriers. The timing and breadth of declines suggest multiple reinforcing drivers, following the 2018 tariffs, Covid, and the Russian invasion of Ukraine. Whether this fragmentation continues amid recent drastic changes to the international trade landscape and geopolitical risks remains an open question for policymakers, researchers, and businesses navigating global supply chain disruptions and geoeconomic uncertainty.

References

Aiyar, Shekhar, Anna Ilyina, and others (2023). "Geoeconomic Fragmentation and the Future of Multilateralism," International Monetary Fund Staff Discussion Note SDN/2023/001.

Gopinath, Gita, Pierre-Olivier Gourinchas, Andrea F. Presbitero, and Petia Topalova (2025). "Changing Global Linkages: A New Cold War?" Journal of International Economics, vol. 153, https://doi.org/10.1016/j.jinteco.2024.104042.

Kallen, Cody (2025). "Breaking Up: Fragmentation in Foreign Direct Investment," International Finance Discussion Papers 1413. Washington: Board of Governors of the Federal Reserve System.

Reyes-Heroles, Ricardo, Sharon Traiberman, and Eva Van Leemput (2020). "Emerging Markets and the New Geography of Trade: The Effects of Rising Trade Barriers," IMF Economic Review, vol. 68 no. 3, pp. 456-508, September.