The True Owners of Foreign Capital

Explaining the sources and destinations of capital flows is a key focus of research in international finance. But capital flows between countries can flow through financial centers before they arrive at their ultimate destination, and these intermediary flows distort the record of the actual ownership of investments. Two recent papers seek to provide a more accurate picture of the true sources of foreign finance.

Jannick Damgaard of Danmarks Nationalbank, Thomas Elkjaer of the International Monetary Fund and Niels Johannesen of the University of Copenhagen differentiate between “phantom” and “real” foreign direct investment in their 2019 IMF working paper, “What Is Real and What Is Not in the Global FDI Network?”  Phantom FDI flows to shell companies that do not engage in any business activities, and are used to minimize corporate taxation before the funds are channeled to their final destination. Among the host countries that receive a significant amount of phantom investment are the Netherlands, Luxembourg, Hong Kong, Switzerland, Singapore and Ireland. The phantom FDI overstates the actual amount of investment that takes place and obfuscates the ultimate ownership of foreign capital.

Damgaard, Elkjar and Johannsen use several sources of data in order to uncover the actual owners of FDI. These include the IMF’s Coordinated Direct Investment Survey, which reports foreign investments in 110 countries by the country of the immediate owner; the OECD’s Foreign Direct Investment Statistics, which differentiates between FDI in Special Purpose Entities (SPEs), a form of shell company, and non-SPE investment, and also includes information on the ultimate owners of investment; and Orbis, a global database of corporate data, including ownership information. Since the OECD data are incomplete, they estimate the share of real FDI in total FDI by using the negative relationship of real FDI/total FDI and total FDI/GDP.

Their results show that in 2017 global FDI of almost $40 trillion included real FDI of $25 trillion and phantom FDI of about $15 trillion. Moreover, the share of phantom FDI in total FDI has risen from above 30% in 2009 to just below 40% in 2017. Luxembourg reported the largest amount of phantom FDI of $3.8 trillion, followed by the Netherlands with around $3.3 trillion. The largest stock of real FDI, on the other hand, was located in the U.S., which also owned the largest amount of outward FDI. China has been a significant recipient of inward FDI (but see below), as were the United Kingdom, Germany and France. The authors also found evidence of “round tripping,” i.e., supposedly inward foreign investment that is actually held by domestic investors. In the case of China and Russia about 25% of real FDI is owned by investors in those countries.

Another investigation of the data on international capital was undertaken by Antonio Coppola of Harvard, Matteo Maggiori of Stanford’s Graduate School of Business, Brent Neiman of the University of Chicago’s Booth School of Business and Jesse Schreger of the Columbia Business School, and they report their results in “Redrawing the Map of Global Capital Flows: The Role of Cross-Border Financing and Tax Havens.” Global firms have increasingly issued securities through affiliates in tax haven, and these authors seek to uncover the ultimate issuers of these securities. Their results allow them to distinguish between data reported on a “residency” basis based on the country where the securities are issued versus a “nationality” basis, which shows the country of the ultimate parent.

The authors begin with data from several databases that allows them to uncover global ownership chains of securities through tax haven nations such as Luxembourg and the Cayman Islands.  They use this mapping to determine the ultimate issuers of securities held by mutual funds and exchange traded fund shares that are reported by Morningstar. Finally, they use their reallocation matrices to transform residency-based holdings of securities as reported in the U.S. Treasury’s International Capital data and the IMF’s Coordinated Portfolio Investment Survey to nationality-basis holdings.

Their results lead to a number of important findings. Investments from advanced economies to emerging market countries, for example, have been much larger than had been reported. For example, U.S. holdings of corporate bonds in the BRIC economies (Brazil, Russia, India and China) total $99 billion, much larger than the $17 billion that appears in the conventional data. U.S. holdings of Chinese corporate bonds alone rises from $3 billion to $37 billion, and of Brazilian bonds the total increases from $8 billon to $44 billion. These figures are even higher when the U.S. subsidiaries of corporations in emerging markets which issue securities in the U.S. are accounted for. Similarly, holdings of common equities in the emerging markets by investors in the U.S. and Europe are much larger when the holdings are reallocated from the tax havens to the ultimate owners. This is particularly evident in the case of China.

The reallocation also shows that the amount of corporate bonds issued by firms in the emerging markets has been more significant than realized. While the issuance of sovereign bonds is accurately reported, the issuance of corporate bonds has often occurred via offshore subsidiaries. These bonds are often denominated in foreign currencies, so their reallocation to their ultimate issuers results in an increase in foreign currency exposure for their home countries.

As in the previous study, Coppla, Maggiori, Neiman and Schreger also find that some “foreign” investment represents domestic investment routed through a tax haven, such as the Cayman Islands. These flows are particularly significant in the case of the U.S. In addition, some FDI flows to China should be classified as portfolio, since they reflect foreign participation in offshore affiliates that is channeled to China. FDI positions are not revalued as often as portfolio holdings, and as a result the authors claim that China’s net foreign asset position is overstated.

The results of these ground-breaking papers have important implications. First, the international ownership of capital is more concentrated than realized. The “Lucas paradox” of international capital flowing from developing to advanced economies was based on misleading data. The U.S. and several other advanced economies have large stakes in the emerging markets. Second, some of emerging markets are more vulnerable to currency depreciations than the official data suggest because their corporations have issued debt through subsidiaries in ta haven countries. Third, multinational corporations have been successful in shielding their income from taxation by using tax havens. The OECD has been working to bring this profit shifting under control, but effective reform may require a fundamental change in how multinationals are taxed by national governments.