The concept of foreign direct investment (FDI) has become foundational to theories of globalization, interdependence, and political risk. Yet these theories rest on an often untested assumption: that “foreign” capital is genuinely foreign. This paper challenges that premise by distinguishing round-tripped investment — domestic capital returning through offshore intermediaries - from truly foreign-originated investment. We argue that misclassifying these categories biases both empirical inference and the informing of theory about how firms respond to geopolitical risk and reward. We illustrate the value of better conceptualization and measurement of FDI through the case of China. Using data from the Foreign Invested Enterprises in China (FIEC) dataset, we deploy a random-forest model to leverage known firm attributes to identify likely round-tripped investors. We estimate that between 10–30% of nominally foreign-invested firms are round-tripped mainland firms. Reclassifying investors by ultimate corporate owners and replicating earlier work on divestment yields two results. First, the 2018–2019 U.S.-China Trade War increased exits for both round-trip and foreign firms. Second, round-trip firms are significantly less likely than genuinely foreign MNCs to divest when tariff-exposed. Aggregate FDI statistics thus overstate the scale of MNC exit and underestimate the targeted effect of tariffs on the exit of genuinely foreign MNCs. The results demonstrate that widely cited indicators of “foreign” disengagement from China are in fact illusory. This underscores the need for more faithful conceptualization of FDI coupled with the measurement of firm-level ownership data for understanding of the causes and effects of FDI.