The shift in the composition of capital inflows into developing countries toward relatively more bank credit and less foreign private investment is likely to have increased their vulnerability to various economic disturbances. Total income payments on foreign investment tend to move more closely with a country’s ability to service them than do interest payments on external debt, which continue even if the original borrowing financed unprofitable investments or consumption. In this sense, the greater the share of foreign investments in a country’s portfolio of external liabilities, the greater is the share of risk associated with economic disturbances that is borne by foreign investors. In addition, since foreign investment can be sensitive to changes in a host country’s relative competitiveness, as well as to its interest rate and credit policies, a higher proportion of such investment in total capital flows can increase their responsiveness to a country’s adjustment policies.
Since the greater risk-bearing associated with foreign investment generally needs to be compensated by higher expected returns, total service payments may be higher, the greater the share of foreign instruments in the portfolio. (Although this does not imply that a host country would necessarily need to raise the expected rate of return to foreign investors in order to attract a greater volume of foreign investment, since a removal of restrictions on such inflows would probably be enough to generate increased investment at existing returns.) The desired composition of the portfolio will depend on the desired trade-off between risk and return. The combination of risk and return that a country is willing to accept will be determined not only by individual preferences within the country, but also by the costs associated with maintaining service payments on foreign liabilities when economic conditions deteriorate. These costs generally result from the need to restore a sustainable current account position, either by reducing aggregate expenditures or by switching resources from nontraded to traded goods sectors. The relatively low levels of per capita consumption and limited supply responses in many developing countries mean that the costs of making large adjustments over a short time can be substantial. However, although a country’s long-term ability to service its total external liabilities depends on the size of total service payments (relative to its total output and its ability to earn or save foreign exchange) the way in which it adjusts to economic disturbances in the short term will also depend on the composition of those service payments. In particular, service payments on direct investment consist of both dividend remittances and reinvested earnings and the costs of adjustment may differ, depending on which is most affected by economic disturbances.
The impact of the composition of a country’s external liabilities on the costs of adjustment can be illustrated by considering the different effects of economic disturbances on two economies in which investment is financed by external debt and by external equity investment, respectively. An external economic disturbance that affects foreign exchange earnings (such as a decline in the terms of trade or a fall in volume of exports) would not alter interest payments due on external debt. Future expenditures would have to be reduced and resources switched from the nontraded to the traded sectors to generate the foreign exchange to meet the interest payments. Profits on equity investment would be likely to decline, however, either because they were affected directly by the external economic disturbance (if the investment were in the export sector), or indirectly by policies adopted to restore external equilibrium. It is, of course, possible that in some circumstances adjustment policies could increase profits on foreign investment—for example, as a result a large devaluation if the foreign investments were concentrated in the import-substituting sector and if output were not affected by shortages of imported inputs. These effects are discussed in greater detail later in this section. But generally the required reduction in future expenditures to generate resources for service payments would be less when investment is financed by equity than by debt. However, whether the immediate foreign exchange outflow was also lower than for debt-financed investment—which would reduce the need for a transfer of resources between traded and nontraded sectors—would also depend on whether the decline in profits resulted in lower dividend remittances overseas or in lower reinvested earnings. Some limited evidence on this aspect will also be discussed later in this section.
Comprehensive empirical comparisons between the service payments on foreign investment and those on external debt, and a country’s ability to make those payments, are hampered by the lack of reliable key information in many developing countries, in particular of time series on reinvested earnings. However, there is some evidence that total returns on direct investment are more correlated with a country’s ability to service its external liabilities than are interest payments on external debt. For a group of non-oil developing countries with sufficiently long time series on reinvested earnings, the estimated annual rate of return on direct investment was positively associated with the annual rate of growth of GDP. An above- (or below-) average rate of growth of GDP was associated with an above- (or below-) average return on direct investment in all but one year between 1974 and 1982. There was a similar, but much weaker, positive association between rates of return on direct investment and rates of growth of exports. By contrast, there was little association between the rate of growth of GDP and exports in these countries, and the average interest rate paid on their outstanding external debt. The difference in movements in rates of return and interest rates was particularly marked during the recent recession. Similar results were obtained for rates of return on direct investment from the United States in the manufacturing sectors of developing countries; these returns tended to move more closely with growth rates in non-oil exports and non-oil GDP in the host countries than did interest rates on world financial markets (see Chart 3). These results are also discussed in more detail in Appendix III.
Developing Countries: Rates of Return on U.S. Direct Investment and Related Variables, 1973–82
(In percent)
1London Interbank Offered Rate on three-month deposits.2In U.S. dollars.3From U.S. Department of Commerce, Survey of Current Business, various issues.There is, therefore, some evidence that total returns paid on direct investment are, in general, more positively correlated with changes in a country’s ability to service those payments than are interest payments on its external debt. This should ease the process of adjustment to economic disturbances in countries with a large proportion of direct investment in total external liabilities. This is illustrated by an examination of the relative importance of direct investment in the total liabilities of countries that have encountered debt-servicing difficulties in recent years. For 28 developing countries that rescheduled part of their external debt during 1983, the stock of direct investment accounted for an average of only 14 percent of their total external liabilities (i.e., direct investment plus external debt) at the end of 1983, compared with an average of 24 percent for those 49 developing countries with available data that did not reschedule debt.31
However, the way in which variations in profits affect adjustment also depends on their distribution between remitted dividends and reinvested earnings, since this influences the immediate foreign exchange outflow. As discussed in Section II, a large share of the earnings from direct investment is generally reinvested, and constitutes a substantial proportion of total direct investment in developing countries. For 12 non-oil developing countries with relatively long time series on reinvested earnings (Bolivia, Brazil, Cameroon, Colombia, Costa Rica, El Salvador, Honduras, Israel, Jamaica, Mexico, Morocco, and Sierra Leone), they constituted an average of some 39 percent of total direct investment during 1973–82. The share of earnings that are reinvested, however, fluctuates substantially with changes in economic conditions. For instance, reinvested earnings of U.S. affiliates in manufacturing in developing countries were much less stable than their gross dividend payments (Table 4). Like those of companies in industrial countries, the dividend payments of the affiliates were to a large extent unaffected by short-term fluctuations in profitability. This was particularly true in 1982 and 1983 when earnings of manufacturing affiliates fell sharply while dividend payments remained unchanged. Other elements of the sources and uses of funds must have adjusted to the lower level of reinvested earnings: either new capital expenditures were reduced or affiliates increased their borrowing from sources other than the parent company.
U.S. Direct Investment in Developing Countries: Trends in and Distribution of Earnings, 1973–83
(In billions of U.S. dollars)
Before host-country withholding taxes on dividends.
Reinvested earnings as a proportion of total earnings.
U.S. Direct Investment in Developing Countries: Trends in and Distribution of Earnings, 1973–83
(In billions of U.S. dollars)
| All industries | Manufacturing | |||||||
|---|---|---|---|---|---|---|---|---|
| Earnings | Distributed1 | Reinvested | Reinvestment ratiot2 | Earnings | Distributed1 | Reinvested | Reinvestment ratio2 | |
| Incorporated affiliates only | ||||||||
| 1973 | 3.0 | 1.4 | 1.6 | .52 | 0.9 | 0.3 | 0.6 | .66 |
| 1974 | 3.6 | 2.1 | 1.5 | .43 | 1.0 | 0.3 | 0.7 | .70 |
| 1975 | 4.0 | 0.9 | 3.1 | .78 | 1.3 | 0.4 | 0.9 | .71 |
| 1976 | 3.6 | 2.3 | 1.2 | .34 | 1.2 | 0.5 | 0.7 | .54 |
| 1977 | 3.9 | 1.7 | 2.3 | .58 | 1.3 | 0.5 | 0.8 | .63 |
| 1978 | 4.8 | 1.9 | 2.9 | .60 | 1.9 | 0.6 | 1.3 | .68 |
| 1979 | 6.1 | 2.5 | 3.6 | .59 | 2.2 | 0.9 | 1.3 | .58 |
| 1980 | 7.2 | 2.8 | 4.4 | .61 | 2.6 | 0.7 | 1.8 | .72 |
| 1981 | 8.1 | 3.1 | 5.0 | .62 | 2.3 | 1.0 | 1.3 | .54 |
| 1982 | 6.3 | 3.4 | 2.9 | .46 | 0.9 | 1.0 | -0.1 | -.17 |
| Incorporated and unincorporated affiliates | ||||||||
| 1982 | 11.4 | 7.8 | 3.6 | .31 | 0.8 | 1.1 | -0.3 | -.37 |
| 1983 | 9.8 | 7.1 | 2.7 | .28 | 0.5 | 1.0 | -0.5 | -1.00 |
Before host-country withholding taxes on dividends.
Reinvested earnings as a proportion of total earnings.
U.S. Direct Investment in Developing Countries: Trends in and Distribution of Earnings, 1973–83
(In billions of U.S. dollars)
| All industries | Manufacturing | |||||||
|---|---|---|---|---|---|---|---|---|
| Earnings | Distributed1 | Reinvested | Reinvestment ratiot2 | Earnings | Distributed1 | Reinvested | Reinvestment ratio2 | |
| Incorporated affiliates only | ||||||||
| 1973 | 3.0 | 1.4 | 1.6 | .52 | 0.9 | 0.3 | 0.6 | .66 |
| 1974 | 3.6 | 2.1 | 1.5 | .43 | 1.0 | 0.3 | 0.7 | .70 |
| 1975 | 4.0 | 0.9 | 3.1 | .78 | 1.3 | 0.4 | 0.9 | .71 |
| 1976 | 3.6 | 2.3 | 1.2 | .34 | 1.2 | 0.5 | 0.7 | .54 |
| 1977 | 3.9 | 1.7 | 2.3 | .58 | 1.3 | 0.5 | 0.8 | .63 |
| 1978 | 4.8 | 1.9 | 2.9 | .60 | 1.9 | 0.6 | 1.3 | .68 |
| 1979 | 6.1 | 2.5 | 3.6 | .59 | 2.2 | 0.9 | 1.3 | .58 |
| 1980 | 7.2 | 2.8 | 4.4 | .61 | 2.6 | 0.7 | 1.8 | .72 |
| 1981 | 8.1 | 3.1 | 5.0 | .62 | 2.3 | 1.0 | 1.3 | .54 |
| 1982 | 6.3 | 3.4 | 2.9 | .46 | 0.9 | 1.0 | -0.1 | -.17 |
| Incorporated and unincorporated affiliates | ||||||||
| 1982 | 11.4 | 7.8 | 3.6 | .31 | 0.8 | 1.1 | -0.3 | -.37 |
| 1983 | 9.8 | 7.1 | 2.7 | .28 | 0.5 | 1.0 | -0.5 | -1.00 |
Before host-country withholding taxes on dividends.
Reinvested earnings as a proportion of total earnings.
Consequently, it appears that—at least during the recent recession—part of the automatic adjustment that occurred when returns on foreign direct investment fell with deteriorating economic conditions resulted not from a decline in foreign exchange outflows for dividend payments, but from a reduction in the level of domestic aggregate demand. This may have been brought about directly (as capital expenditures of affiliates declined) or indirectly (as affiliates” increased demand for credit to maintain capital expenditures and dividend payments crowded out other borrowers). Such adjustment may have involved short-term costs similar to those that would have been involved in maintaining service payments on external debt, although, in the longer term, the reduced level of reinvested earnings implies a lower level of foreign liabilities. However, the decline in the share of reinvested earnings was much less marked for affiliates outside the manufacturing sector, and during earlier recessions. There appear to have been a number of special factors that tended to reduce reinvestment of earnings during the last recession; in particular, the sharply increased debt burden and the associated risk of restrictive government policies—such as controls on foreign exchange transfers—may have discouraged reinvestment of earnings. Consequently, it is difficult to judge the extent to which the experience provided by the recent recession is indicative of a more general tendency for reinvestment ratios to decline during periods of economic difficulty. Much of the decline in reinvestment of earnings occurred in some of the larger Latin American countries that had heavy burdens of external debt.
Although direct investment flows grew much less rapidly than bank lending to developing countries over the last decade, they have generally been a more stable component of resource inflows, particularly during 1982–83. Direct investment inflows have tended to fall during periods of adverse economic conditions, because of declining opportunities for profitable investment and tighter cash-flow positions of the parent company and its affiliates. Nevertheless, they held up rather better during the recent period of recession and widespread debt-servicing difficulties than did private, and especially commercial bank, lending. Direct investment also has the added advantage that the maturity structure is more in line with the underlying investments than is that of commercial loans. This helps a country avoid the debt-servicing problems that can arise when longer-term investments are financed by short-term bank loans and a deterioration in a country’s external financial position makes banks reluctant to roll the loans over.
A larger share of direct investment in capital inflows is likely to make the latter more sensitive to the adjustment policies undertaken by a developing country. For instance, although the major impact of exchange rate policy will be on the current account balance, movements in the exchange rate and domestic prices and costs affect the profitability of direct investment. A depreciation of the real exchange rate will tend to increase the profits and output of an enterprise, provided that its output is more traded than the inputs used to produce it.32 Most enterprises established through foreign direct investment probably fall into this category, and will therefore be encouraged by an exchange rate policy that maintains international competitiveness. A real exchange rate depreciation may decrease the profitability of direct investments in enterprises whose output is less traded than inputs. Investment in public utilities or in the production of final goods for domestic markets protected by quantitative import controls on the basis of large-scale imported inputs is probably in this category. However, a policy of maintaining an overvalued exchange rate is unlikely to encourage a substantial inflow of such direct investment, since the probability of periodic adjustments in exchange rates to more appropriate levels increases the uncertainty associated with such investment. Available evidence for direct investment flows between industrial countries indicates that, on balance, direct investment inflows into a country increase when its relative competitive position is improved.33
Moreover, direct investment flows are only one component of the overall financing of the total capital expenditures of a foreign-controlled firm, and they can be strongly affected by the host country’s interest rate and credit policies. A policy of increasing interest rates toward market-clearing levels is likely to reduce domestic financing of a firm’s expenditures and encourage foreign direct investment, particularly during the initial period as the firm’s stock of liabilities adjusts to the new interest rate differentials. Conversely, where direct investment in a country is substantial, it can significantly increase the costs of inappropriate policies. Affiliates of foreign-controlled companies have substantial opportunities to engage in short-term intracompany lending in response to shifts in interest rate differentials and exchange rate expectations. This can make capital movements sensitive to monetary and exchange rate policy even in countries with rudimentary capital markets and severe restrictions on most capital movements.
