Standard international macroeconomics predicts that capital should flow from capital-rich countries, where the marginal return on capital is low, to capital-poor countries, where the marginal return is high. By implication, such capital flows would improve the international allocation of capital and raise global output. Furthermore, there would be no difference between gross and net capital flows, as capital movements are unidirectional.
The patterns of international capital flows observed in the past 20 years, however, stand in stark contrast to these predictions. First, Lane and Milesi Ferretti (2006) show that developing countries have accumulated large stocks of net foreign assets over the past decade, while industrialized countries, and the U.S. in particular, have increased their net foreign liabilities. Prasad, Rajan, and Subramanian (2006, 2007) document that the average per-capita income of countries running current account surpluses has trended downwards in the past three decades, while that of deficit countries has trended upwards. Since 1998, the relative per-capita income of the surplus countries has actually been below that of the deficit countries, i.e., net capital flows have been “uphill” from poor to rich countries. Second, many emerging markets have experienced much faster growth of their gross external financial assets and liabilities than of their net positions (Lane and Milesi Ferretti, 2001, 2006, 2007). Ju and Wei (2007) observe that many developing economies, including China, Malaysia, and South Africa, are net importers of foreign direct investment (hereafter, FDI) and net exporters of financial capital at the same time, while developed countries such as France, the United Kingdom, and the United States are net exporters of FDI and net importers of financial capital. Third, despite of a negative net international investment position since 1986, the U.S. have been consistently receiving a positive net international investment income until 2005 (Gourinchas and Rey, 2007; Hausmann and Sturzenegger, 2007; Higgins, Klitgaard, and Tille, 2007). Given the huge net foreign liabilities of USD 2.5 trillion, the U.S. paid out only USD 3.4 billion in net income payments to foreign investors during the first half of 2006, implying payments on net international liabilities at the annual interest rate of 0.27%. In other words, the U.S. is a significant net debtor actually for free!
Two strands of literature have recently offered explanations of these empirical puzzles. One is based on the risk-sharing investors can achieve by diversifying investment globally, see Caballero, Farhi, and Gourinchas (2008); Devereux and Sutherland (2007); Mendoza, Quadrini, and R?os-Rull (2007). In these models, international capital flows are determined by the cross correlation patterns of the shocks hitting individual economies. However, there is no substantive distinction between FDI and portfolio investment in these models. Empirically, the coexistence of negative international investment position and positive international investment income in the U.S. can be explained by the rate of-return differential between the U.S. international assets and liabilities. For example, the rate of return on U.S. FDI assets abroad was 8% during the first half of 2006, while a rate of return on foreign FDI assets in U.S. was just 5.1%; rates of return on other U.S. international assets and liabilities were about the same, much lower than the rates of return on FDI mentioned above. Such patterns of interest rates also prevailed in earlier years (Higgins, Klitgaard, and Tille, 2007; Kitchen, 2007). However, FDI and its related equity premium has received much less attention in the recent theoretical literature.
The other strand of literature focuses on financial market frictions. Gordon and Boven berg (1996) show that asymmetric information between firms and investors in different countries can reduce international capital mobility and lead to persistent interest rate differentials between countries. Gertler and Rogoff (1990) show that different degrees of capital market imperfections can dampen and even reverse capital flows from rich to poor countries. Boyd and Smith (1997) and Matsuyama (2004) show in the overlapping generations framework that, in the presence of credit market imperfection, financial market globalization may lead to an equilibrium with multiple steady states where fundamentally identical countries end up with different levels of income, a result Matsuyama (2004) calls “symmetry breaking”. It challenges the traditional notion that international capital mobility leads to the convergence of living standard across countries. More recently, several authors have studied the implications of agency costs for the welfare effects of liberalizing capital account (Aoki, Benigno, and Kiyotaki, 2007; von Hagen and Zhang, 2006, 2008). Notably, Ju and Wei (2006, 2007) show the bypass effect in a static model, i.e., when both FDI and financial capital flows are allowed, all financial capital leaves the country with more severe credit market imperfections, while FDI flows into this country. In this framework, the distinction between financial capital and FDI is crucial.
Our paper extends this second strand of literature. We develop a two-country over lapping generations model, which is a dynamic extension of the framework used in Ju and Wei (2007) and von Hagen and Zhang (2007). It allows us to endogenize the marginal return on investment and provides a richer but still tractable framework to analyze the direction and composition of international capital flows as well as the implications for production and welfare. Our model builds on the notion that individuals in an economy differ in their ability of using physical capital for production. From the efficiency perspective, it would be desirable to transfer all capital to the most productive individuals to maximize aggregate output. Due to financial frictions, however, the most productive individuals are subject to borrowing constraints. The constrained aggregate credit demand has a general equilibrium effect, keeping the rate of return on loan (hereafter, the loan rate) lower and the rate of return on equity (hereafter, the equity rate) higher than the marginal rate of return on investment. While, if there were no financial frictions, the two interest rates would be equal to the marginal rate of return on investment. This way, financial frictions distort the two interest rates and generate an equity premium in this deterministic model.
Matsuyama (2004) assumes that changes in aggregate investment takes place on the extensive margin, i.e., each productive project has a fixed investment size, while the number of individuals in a country who can invest in these projects is endogenously determined. Our model setting differs from Matsuyama (2004) only in one aspect. We assume that changes in aggregate investment takes place on the intensive margin, i.e., the number of individuals in a country who can invest is fixed, while the size of the project investment is endogenously determined. As our first result, we show that under international capital mobility, there exists a unique and stable steady-state equilibrium even in the presence of credit market imperfections. Thus, Matsuyama’s symmetry-breaking property depends critically on the assumption of the fixed investment size of entrepreneurial project and thus, changes in aggregate investment take place only on the extensive margin.
In our model, financial development by relaxing the borrowing constraints of entrepreneurs raises aggregate domestic credit demand and thus, the loan rate rises while the equity rate declines under international financial autarky. In other words, cross-country difference in financial development results in cross-country differences in the loan rate as well as in the equity rate under international financial autarky. Under perfect capital mobility, these differences drive international capital flows. In the steady state, the financially less (more) developed country endogenously becomes the poor (rich) country; financial capital flows from the poor to the rich country, while FDI flows in the opposite direction; net capital flows are from the poor to the rich country; despite of its negative net international investment position, the rich country receives a positive net international investment income, because the rich country earns a higher rate of return on its FDI than it pays out on its foreign debts. This way, our model explains the three empirical facts mentioned above.
Ju and Wei (2007) assume cross-country differences in terms of capital-labor ratio, financial development, corporate governance, and property right protection for generating two-way capital flows. In contrast, the two countries in our model differ fundamentally only in the level of financial development. Thus, cross-country difference in financial development alone is sufficient to generate two-way capital flows. This way, we minimize the set of assumptions needed for explaining this empirical fact. In addition, the static framework in Ju and Wei (2007) is used for analyzing the immediate impacts of capital account liberalization but not the transitional and long-run effects, while the overlapping-generations framework enables us to discuss the short-run and the long-run effects.
Financial capital flows as well as FDI affect the owners of credit capital and equity capital in the different ways, implying that liberalizing capital flows has distributional effects within a country and that the welfare effects may be ambiguous. Based on our model, we can make welfare predictions of capital account liberalization, depending on the levels of financial development in the two countries. We also show that, in the OLG frame work, liberalizing capital flows affects the intergenerational distribution of income due to transitional effects. Thus, our framework may explain why capital account liberalization often encounters both support and opposition in a given country.
In contrast to the predictions of the standard models of international macroeconomics, liberalizing capital flows reduces world output in our model, same as in Matsuyama (2004). It results from the concavity of aggregate production and our assumption that lenders do not have other savings alternatives except lending to entrepreneurs who are most productive in the use of real capital. von Hagen and Zhang (2009) relax this assumption and show numerically that international liberalizing capital flows can be globally efficiency enhancing. In this paper, we stick to this assumption for tractability purpose.
The rest of the paper is structured as follows. Section 2 sets up the model and discusses the long-run patterns of the loan rate and the equity rate with respect to financial development under international financial autarky. Section 3 analyzes the long-run and short-run production and welfare implications under two scenarios of capital mobility. Section 4 concludes with the main findings. Appendix collects the proofs of propositions and discusses the robustness of our results under two alternative model specifications.
Download
PDF Ebook Financial Development and the Patterns of International Capital Flows
