A long with the rise in portfolio capital inflows, developing countries have experienced a surge in foreign direct investment inflows during the 1990s. Like portfolio capital, these inflows have gone to a relatively small number of countries in Latin America and Asia. Their sharp rise is in part explained by factors that are likely to exert an influence over a limited period of time, such as the privatization of government assets or the rebalancing of asset holdings in response to economic reforms in developing countries. Although foreign direct investment is often perceived as a relatively stable source of financing, a review of experience in the 1980s of a sample of highly indebted developing countries suggests a more cautious view. When balance of payments difficulties are encountered, the net impact of all transactions associated with foreign direct investment (including both current and capital account transactions) may serve to exacerbate external imbalances.
Recent Trends in Direct Investment
Since the mid-1970s, when only modest flows were recorded, net inflows of foreign direct investment to developing countries have risen rapidly.39 Nevertheless, there have been substantial fluctuations around this upward trend (Table A21 and Chart 15). For example, a surge in inflows has occurred since 1990. Over the period 1991–93, cumulative net inflows amounted to $134 billion. In real terms, net foreign direct investment flows to developing countries in the early 1990s were almost two and a half times their average level in the 1980s.
Net Foreign Direct Investment to Developing Countries
(In billions of U.S. dollars)
Source: World Economic Outlook data base.The surge in foreign direct investment is notable not only for its size, but also for its coincidence with the growth of other private capital flows. The rate of increase in net foreign direct investment in developing countries since 1990 has been comparable with that of portfolio flows. This simultaneous upturn in both direct and portfolio inflows contrasts sharply with previous experiences of private capital surges to developing countries. For example, bond financing was dominant in the 1920s and 1930s, but defaults during the interwar period led in the first twenty years of the postwar period to foreign direct investment replacing bonds as the primary form of capital inflow. A new surge in portfolio flows, specifically in the form of bank lending, began in the 1960s and peaked in the early 1980s, but it was accompanied by a steep decline in foreign direct investment flows.40
Although many countries have experienced increases in direct investment inflows, as was the case with previous surges, the latest upturn has been concentrated in only a few countries. Three countries—Brazil, Indonesia, and Mexico—accounted for over 60 percent of all foreign direct investment flows to developing countries between 1971 and 1981.41 At the end of the 1980s, 63 percent of the total stock of foreign direct investment in developing countries was held in five countries—Brazil, China, Egypt, Malaysia, and Mexico. In contrast, the five developing countries with the most external debt—Argentina, Brazil, India, Indonesia, and Mexico—accounted for only 33 percent of the total stock of debt42 Between 1990 and 1993, the bulk of foreign direct investment flows went to two regions: Latin America and Asia.
In Latin America, the total inflow doubled after 1990. It reached $14 billion in 1992 before declining modestly to around $13 billion in 1993. Over this period, average annual inflows were two to three times higher than in the 1980s. Two thirds of the total foreign direct investment inflow to Latin America during the period went to Mexico and Argentina, with Chile, Venezuela, Brazil, and Colombia accounting for a large share of the remainder. Developing countries in Asia received an estimated $66 billion in foreign direct investment inflows in 1991–93, roughly half of the total flow to developing countries. China was by far the largest recipient, accounting for more than 40 percent of the inflows to Asia. Malaysia and Singapore were also major recipients of foreign direct investment, but at less than half of the level of flows to China. In the case of Malaysia, this represented a substantial pickup in inflows, while for Singapore it represented a relatively steady inflow of investment. Thailand and Indonesia also saw sizable foreign direct investment inflows over the period. Inflows to the economies in transition in Central and Eastern Europe also rose sharply over the period 1991-93, amounting to approximately 10 percent of total foreign direct investment inflows to developing countries. Finally, annual flows to Africa were largely unchanged in 1991–93, with almost all of the funds going to Nigeria and South Africa.
The nature of and motivation for foreign direct investment suggest that longer-term considerations play a role in explaining these flows. As a result, direct investment might be expected to exhibit greater stability than other types of private capital flows.43 Moreover, there are factors at work in the global economy, such as growing trade, greater market homogeneity, and improved communications technology, that should provide long-term impetus for increased flows of foreign direct investment.44 While all of this may be true in general, the recent experience of developing countries may be explained by a number of factors that can be expected to have an impact over a more limited period of time.
Far-reaching economic reforms—including the removal of legal restrictions on capital movements and on nonresident holdings of domestic assets—have eliminated many of the barriers that formerly acted to deter foreign direct investment. At the same time, the alleviation of debt burdens has sharply lessened risks for all would-be investors.45 Thus, the response of foreign direct investment to these changes may to an extent represent a stock adjustment—implying a onetime rebalancing of the pattern of corporate asset holdings in response to policy changes in host countries.
Foreign direct investment related to privatization has also been an important factor in the recent upturn, particularly in flows to Latin America. The potential for further foreign direct investment inflows from new privatizations, however, will inevitably depend on the size of the remaining stock of public sector assets that can be earmarked for sale. In general, the pace of this type of inflow can be expected to slow as countries near the end of their privatization programs. There may be further inflows related to the restructuring of newly privatized enterprises, but these are not likely to continue on the scale of the initial inflows associated with the sale of the enterprises.
Beyond such specific factors, general economic expansion in the largest recipient countries also contributed to the recent surge in foreign direct investment. Domestic growth not only provides foreign firms with enhanced investment opportunities, it also provides a pool of funds for reinvestment. In part because enterprises typically view such internally generated resources as cheaper than funds raised in the capital market, reinvested earnings now account for a large proportion of total foreign direct investment capital flows. Reinvested earnings, nonetheless, appear to move procyclically; a slowdown in economic growth has a dampening effect.
Behavior of Foreign Direct Investment Transactions in a Crisis
Foreign direct investment capital flows have been viewed as potentially providing a more stable form of financing for economic development, as well as a substitute for reduced flows of commercial bank financing.46 An important question remains, however, about the behavior of such flows when a country encounters balance of payments difficulties 47 It is widely argued that the longer-term motivations for foreign direct investment, together with the substantial costs usually entailed in liquidating fixed assets, would result in these flows being less responsive to adverse short-run macroeconomic developments.
The stability of foreign direct investment capital flows during crisis periods would, indeed, appear to be largely confirmed by a review of balance of payments data reported to the IMF.48 Chart 16 shows the composition of total net private capital flows to six heavily indebted countries that experienced external payments difficulties during the 1980s.49 The data are expressed in real terms, deflating U.S. dollar values by an index of the unit price of exports for industrial countries. While small in magnitude compared with other private short- and long-term flows, net foreign direct investment capital flows were relatively stable throughout the debt crisis. Moreover, after rising substantially in the late 1970s, those flows returned to previous levels in the 1980s. Throughout the period, inflows of foreign direct investment were recorded, while sizable outflows of other forms of investment occurred in the mid-1980s.
Focusing solely on recorded capital flows, however, may not fully capture the influence of foreign direct investment on a country’s external position. Whether foreign direct investment contributes to or alleviates a balance of payments “shock” depends on the behavior of the net flow resulting from both current and capital account transactions between a domestic affiliate and its foreign parent. Conceptually, all earnings from affiliated companies are assumed to accrue to their foreign parents and are included in current account transactions; the portion of earnings that is reinvested by the foreign parents in their affiliates is recorded as an inflow in the host country’s capital account. Thus, the net impact of foreign direct investment transactions is measured by associated capital inflows (comprising both reinvested earnings and “new” investment flows) less total earnings of foreign-owned companies. There is an a priori reason to expect that, in addition to the stability of inflows through the capital account, foreign direct investment may also exert a stabilizing influence during a crisis through its effects on the current account. Current account outflows in the form of repatriated earnings may be expected to decline during a crisis, since the crisis is likely to reduce the total earnings of foreign-owned companies.
To capture fully the influence of foreign direct investment transactions on a country’s external position, it is important that the data accurately record total earnings and total capital flows (i.e., reinvested earnings are fully accounted for). Given uncertainties about the coverage of balance of payments data in many countries, U.S. data on U.S. direct investment abroad have been used to examine the overall balance of payments effects of foreign direct investment transactions for the six heavily indebted countries discussed above.50 While the U.S. data capture only the activity of U.S. firms, they have the advantage of including all components of direct investment and, thus, may better capture short-run changes in the behavior of foreign investors. The data are also likely to be broadly representative, owing to the fact that U.S. companies account for roughly half of the foreign investment in the group of six heavily indebted developing countries.
For U.S. affiliates, total earnings fell when external payments difficulties of the host countries became more severe. These affiliates, however, responded to the crisis by reducing reinvested earnings by more than the decline in total earnings. Repatriated earnings, thus, remained relatively stable, and actually increased throughout the period. In 1983, repatriated earnings exceeded income, generating negative reinvested earnings (i.e., a reduction in the assets of the U.S. affiliates in these countries).
The sharp drop in reinvested earnings during the crisis was accompanied by a decline in flows of other foreign direct investment capital to the six countries. With the more complete coverage of reinvested earnings in the U.S. data, it appears that there was substantially more volatility in foreign direct investment inflows during the debt crisis than is suggested by the balance of payments data for the six countries. Moreover, the foreign direct investment capital inflows shown in the U.S. data appear to have behaved in a manner generally consistent with domestic investment as a whole in the host countries. For the group of countries surveyed, domestic investment fell from 24 percent of GDP in 1981 to 18 percent of GDP in 1985, reflecting poor output performance, low profitability, and an unstable financial environment. Net resource transfers back to the parents increased sharply beginning in 1983 and continued until 1991. Over the period 1983-89, these transfers averaged around 7 percent per year of the total stock of U.S. direct investment in the six countries.
This analysis contrasts with the perception that direct investment flows may help stabilize the balance of payments during a crisis. The persistence of large net resource transfers associated with foreign direct investment out of the six countries sampled here may be related to the balance of payments difficulties these countries encountered. In common with all other external creditors, foreign investors faced the risk that the unallocated loss implied by these debt-service difficulties could fall on them, and they appear to have been reluctant to commit to new investment until expected losses to creditors declined, as signaled by increases in the secondary market value of debt.51 The recent surge in direct investment inflows into these countries has largely corresponded with a return to creditworthiness and renewed access to foreign capital markets in general.
Net foreign direct investment inflows are defined as foreign direct investment capital inflows less capital outflows for direct investments abroad by domestic residents.
See Lizondo (1991).
See, for example, Goldsbrough (1986) and Cardoso and Dorn-busch (1988) for discussion of these issues.
See Claessens, Dooley, and Warner (1993); the authors find that broadly speaking foreign investment is no less volatile on a year-on-year basis for a given country than other capital flows.
These data conceptually include reinvested earnings of domestic affiliates of foreign firms. In practice, however, the recording of reinvested earnings is incomplete in many developing countries. Foreign direct investment statistics are usually compiled by the central bank using actual cross-border flows and generally fail to adequately capture the reinvested portion of domestic affiliates’ earnings.
Argentina, Brazil, Chile, Mexico, the Philippines, and Venezuela.
U.S. Department of Commerce, Survey of Current Business, various issues. All data have been converted into 1985 U.S. dollars using an index of the unit value of industrial country exports. Earnings data and repatriated earnings data are expressed net of with-holding taxes and include interest received from loans. Beginning in 1981, earnings and reinvested earnings data exclude increases in the dollar value of stocks either through exchange rate changes or through capital gains or losses.
