Foreign Debt, Foreign Direct Investment and Volatility

Abstract

We investigate how foreign debt and foreign direct investment (FDI) affect the growth and welfare of a stochastically growing small open economy. First, we find that foreign debt influences the growth of domestic wealth by lowering the cost of capital, while FDI affects the country's welfare by providing an additional source of permanent income. Second, a decline in domestic investment may improve domestic welfare as FDI replaces the gap. Even when the welfare deteriorates, its magnitude is mitigated, leaving more room for discretionary fiscal policy. Third, a fiscal policy aimed to stabilize domestic output fluctuations needs to be conducted not to crowd out the welfare benefit of FDI too much. Fourth, an economy with both types of foreign capital experiences wider welfare swings by external volatility shocks than the one with foreign debt alone, while the welfare effects from domestic volatility shocks are mitigated. The welfare effects of fiscal shocks are much smaller with both types of foreign capital. Lastly, the first-best labor income tax covers the government absorption by the labor's share of total output, and the capital income tax covers the rest. Investment is penalized or subsidized depending on the social marginal cost-gain differential.

Keywords:

• Foreign direct investment
• foreign debt
• inelastic debt supply
• volatility

JEL Classifications:

• F21
• F34
• F36

Notes

¹Our definition of FDI is different from the one by the IMF's Balance of Payments Manual, 5th edition (BPM5), which describes FDI as foreign acquisition of at least 10% ordinary shares or voting power of an enterprise abroad. We assume a full foreign ownership because, following the BPM5, it is difficult to tell FDI in the model from foreign portfolio investment (FPI), which generally has a shorter duration than FDI. Furthermore, the share of FPI is negligibly small in the composition of foreign capital inflows to developing countries. Kraay et al. (2005) reports net foreign equity liabilities of 1.9%, compared to FDI's 27.4% during 1990s. Lane and Milesi-Ferretti (2001) reports 10% and 45%, respectively, over the same period. Also it requires an introduction of an additional price variable, the relative price of the foreigners' claims on the flow of domestic output. It tends to make the model analytically intractable when we analyze fiscal policy implications. For a model in this direction, see Kraay et al. (2005).

²See Schmitt-Grohé and Uribe (2003), and Turnovsky and Chattopadhyay (2003).

³If we assume utility-generating government expenditure, consumption tax may influence the social optimum portfolio choice. See Turnovsky (1999).

⁴τ and τ* (and τ′ and τ*′) are not necessarily equal because, in many countries, FDI firms receive various types of tax benefits as part of the host countries’ efforts to attract foreign capital.

⁵Due to the formulation of the wealth constraint, (PZ)<0 when the agent is saving.

⁶dM* includes the foreigner's all possible income sources other than FDI return and interest income. Although its existence is not crucial for the implications of this model, it is required to satisfy the transversality condition for the equilibrium, which implies a positive C*/W* ratio.

⁷With foreign debt alone, n _i =ω _i (i=k, z) and no distinction is necessary.

⁸Formal derivation of Ĉ/W is provided in Appendix B. Once Ĉ/W is obtained, plugging it into equation (12a) yields ψˆ.

⁹Under mild conditions on the risk-free world interest rate and the interest premium function, it is straightforward to show that the mean interest rate on foreign debt, r _z , is lower than its counterpart in autarky, . For a range of reasonable parameter values, calibration of the model shows lower r _z than .

¹⁰The denominator of equations (17a) and (17b) changes to , where and a nonlinear combination of the foreigner's equity share and volatility terms will be added, which may not allow closed-form solutions as we have in equations (17a) ∼ (17f).

¹¹Ramey and Ramey (1995) show empirical evidence on the tradeoff between volatility and growth. For a positive correlation, see Kormendi and Meguire (1985). Imbs (2007) provides evidence on a positive correlation between sectoral volatility and growth and a negative correlation between macroeconomic volatility and growth.

¹²From dΩ = γΩ d(C/W)/d(C/W)/d(C/W) − Ω[d(c/W) − (η − τ _a )fd(K ^A /W)]/[C/W − (η − τ _a ) f(K ^A /W)], the first term is the direct channel and the second one is the indirect channel.

¹³It is natural to assume substantially high since FDI firms tend to have less information about the host economy. However, it does not necessarily require to be greater than .

¹⁴We choose ˘ω _k by setting the partial derivative of with respect to ω _k at zero. In this case, however, the domestic agent is a creditor, not a debtor. See Turnovsky and Chattopadhyay (2003) for a detailed discussion.

¹⁵From the definition of wealth allocation equation (14), ∂ ωˆ _i /∂ x=(ωˆ _k (∂[ncirc] _z /∂ x)−ωˆ _z (∂[ncirc] _k /∂ x))/([ncirc] _k −[ncirc] _z ), for . It is obvious that the adjustment in the composition of aggregate capital does not necessarily lead to a change in the allocation of domestic wealth.

¹⁶The labor income and the consumption-wealth ratio are reported as their percentage in domestic wealth. The change in welfare is in percentage change from the benchmark equilibrium welfare level.

¹⁷ World Investment Report (UNCTAD, 2007).

¹⁸Sample countries were categorized in low, medium-low, medium, medium-high, and high income groups. Global Development Finance Online (World Bank, 2006a); World Development Indicators Online (World Bank, 2006b); and International Financial Statistics Online (IMF, 2007).

¹⁹In order to satisfy the transversality condition and match the domestic consumption–wealth ratio, we set the fixed return m* in the foreign country at 0.15, which is close to the share of labor income in the domestic economy.

²⁰Empirical evidence on the relationship between FDI and volatility (particularly the real exchange rate volatility) is mixed. For a positive relationship, see Cushman (1988) and Goldberg and Kolstad (1995). For a negative one, see Bénassy-Quéré et al. (2001).

²¹Without FDI, less investment means lower productivity of labor, hence lower labor income and consumption, which uniformly deteriorates welfare.

²²One reason is that, in our current setup, government spending is pure consumption. It only crowds out private expenditure without affecting the private agents' consumption and investment decisions via consumption-externality or provision of productive infrastructure. For discussions in these directions in stochastic growth models, see Corsetti (1997) for a closed economy, and Turnovsky (1999) for an open economy with perfect capital mobility.

²³The expressions are for the deterministic component, and for the stochastic component, where .