A sizable, sensible downward revision to official measures of foreign direct investment in the US

The official estimate of the market value of foreign direct investment (FDI) in the United States at end-2025 has been revised down by close to 30%—from $20.3 trillion to $14.5 trillion—following a sensible change in the methodology employed by the U.S. Bureau of Economic Analysis (BEA) to derive such estimates. The figure below shows the old and the revised FDI in the U.S. series, scaled by US GDP.¹

What happened? In short, the BEA measures the market value of FDI using stock prices. In the past, the BEA relied on broad stock market indices, which recently have been inflated by the soaring stock prices of hyperscalers like Nvidia and Apple, while most foreign direct investment in the U.S. is in industries such as chemicals and financial services whose stock prices have not risen as much. The revisions take all that into account.

When valuing financial claims and liabilities vis-à-vis the rest of the world, dealing with shares or bonds held by domestic or foreign investors is generally straightforward—there is a market price for those financial instruments, and the end-year market value (the one captured in the annual international investment statistics) is easily established. But U.S. affiliates of foreign multinationals, or foreign affiliates of U.S. multinationals, are typically not listed companies—they are often fully owned or majority-owned by their parent companies. The book value of the equity that the parent holds can be obtained from company accounts, but how should the market value be determined?

Until recently, the BEA revalued equity positions in affiliates at historical cost with the stock prices of the country where the affiliate was located. Therefore, equity in U.S. affiliates held by foreign multinational corporations were revalued using U.S. stock prices. This can be problematic when stock price changes differ greatly across sectors and countries, as I pointed out several times, including in a 2023 paper and a recent blog. For U.S. FDI abroad, affiliates are very often holding companies located in financial centers, whose value has little to do with the stock price of said financial center. FDI in the U.S. is concentrated in such sectors as financial services, chemicals, wholesale and retail trade, and autos, while the huge increase in stock prices has been driven by the tech sector. To caricature: The methodology used the stock price of Apple, Alphabet, Meta, and Nvidia to revalue Deutsche Bank, BASF, or Toyota affiliates in the U.S., while it used foreign stock prices to revalue the affiliates that Apple, Alphabet, Meta, and Nvidia hold around the world. With U.S. stock prices rising enormously more than foreign stock prices during the past decade, the result of this approach was a massive increase in the estimated market value of FDI in the U.S. compared to U.S. FDI abroad: the value of the former, until a month ago, exceeded the value of the latter by $6.4 trillion dollars. In contrast, other estimates of FDI published by the BEA that do not adjust equity values for changes in stock prices (historical cost or current cost, which adjusts book value for inflation) showed that the 2025 value of U.S. FDI abroad still exceeded the value of FDI in the U.S. by around $1.27 trillion and $900 billion, respectively.

One consequence of this revision is a large improvement in the official measure of the U.S. net international investment position (NIIP) at the end of 2025, from minus $27.5 trillion (-90% of GDP) to minus $21.9 trillion (-71.6% of GDP). The NIIP is the difference between financial claims on nonresident entities held by U.S. residents and the financial claims of nonresidents on U.S. residents.

A related consequence of the previous estimation method was the build-up of a massive global discrepancy in FDI statistics, with global FDI assets falling considerably short of global FDI liabilities, because partner countries did not use the same methodology to revalue their FDI in the U.S. The revisions to U.S. data for 2019-25 have virtually eliminated this discrepancy (see figure below). U.S. FDI abroad and FDI in the U.S. are a sizable fraction of global FDI claims and liabilities, and hence the revision had a big impact on global figures.

So how did the BEA revise the figures? In an April note, the BEA highlighted the growing misalignment between the industry composition of the direct investment equity positions and that of the S&P 500 index used for revaluation to market value, to which I alluded above. The revaluation is now based on “industry-specific index components from S&P Dow Jones, weighted to align with the industry composition of the historical-cost equity positions.” This is a sensible approach, and the BEA should be commended for addressing this issue. The U.S. remains a large net debtor even after the revision, but the headline figure now has firmer grounding.