1 Capital Inflow, Saving and Investment: The Facts

Three important facts about the inflow of foreign capital into a group of medium-income developing countries – the countries that received substantial inflows of foreign private capital – in the period 1990-2005 are brought out by the data in Table 1.2 First, the inflow of total foreign private capital (TFPC) grew steadily to reach a peak in 1997 – the year the east Asian economic crisis set in. The inflow then declined till 2002 but grew steadily again in the subsequent years. Except in a few years, the inflow was substantial. Second, foreign direct investment (FDI) has been the major form of inflow of private capital into developing countries. Also, FDI inflow held up quite well even in the aftermath of the east Asian crisis. The observed fluctuations in the inflow of foreign

Table 1: Net Inflow of Foreign Saving into Medium-Income Developing Countries

(1990-2005)

Net Inflow as Percentage of GDP
FDI	PFEQ	PBL	TFPC	FAID	TER
1990	0.8	0.1	-0.9	0.0	0.7	0.7
1991	0.9	0.2	-0.9	0.2	0.6	0.8
1992	1.0	0.3	-0.5	0.8	0.2	1.0
1993	1.6	1.1	0.0	2.7	0.2	2.9
1994	1.9	0.8	-0.2	2.5	-0.1	2.4
1995	2.0	0.4	-0.5	1.9	0.1	2.0
1996	2.2	0.5	0.1	2.8	-0.4	2.4
1997	2.6	0.6	-0.4	2.8	-0.1	2.7
1998	2.7	0.0	-0.9	1.8	0.1	1.9
1999	3.1	0.2	-2.0	1.3	0.0	1.3
2000	2.6	0.3	-2.4	0.5	-0.2	0.3
2001	2.7	0.1	-2.4	0.4	0.1	0.5
2002	2.3	0.1	-2.2	0.2	-0.3	-0.1
2003	1.9	0.5	-2.1	0.3	-0.3	0.0
2004	1.9	0.6	-1.6	0.9	-0.2	0.7
2005	2.1	0.9	-1.4	1.6	0.0	1.6

FDI - foreign direct investment; PFEQ - portfolio equity capital; PBL - private bank loans; TFPC - total foreign private capital; FAID - foreign aid; TER - total external resources Source: Author’s estimates based on data from (i) UNCTAD’s World Investment Report database, and (ii) World Bank’s Global Development Finance database. See Appendix 1 for the details.

Economic & Political Weekly

EPW

december 10, 2011 vol xlvi no 50

private capital are explained by fluctuations in the inflows of private bank loans (PBL) and, to a lesser extent, of portfolio equity capital (PFEQ). Third, the inflow of foreign aid (FAID) into this set of countries was quite insignificant over this period; foreign private capital displaced foreign public capital.

The period since 1990, then, has been a period when the mediumincome developing countries received substantial inflow of foreign private capital, the bulk of it in the form of FDI. The standard theory would lead us to expect the investment in this group of countries to have been significantly higher than the domestic saving during the period. The fact that much of the inflow was in the form of FDI only strengthens the expectation. The actual experience, unfortunately, has been contrary to this expectation.

The data relating to saving and investment in the same group of medium-income developing countries are set out in Table 2. These data bring out three remarkable facts. First, gross investment (GI) always fell short of gross aggregate saving (GAS); not all the resources available for investment were in fact invested. Second, GI was only slightly above the gross domestic saving (GDS) in the pre-east Asian crisis period and below it in the postcrisis period. Given that investment financed by foreign saving was significant, it follows that a part of the domestic saving remained unused throughout the period.

Table 2: Saving and Investment in Medium-Income Developing Countries (1990-2005)

GI GDS GAS GI GDS GAS

1990	24.9	24.8	25.5	1998	24.8	24.7	26.6
1991	24.2	23.9	24.7	1999	24.6	25.1	26.4
1992	24.6	23.7	24.7	2000	24.9	25.1	25.4
1993	26.1	24.1	27.0	2001	24.8	25.3	25.8
1994	26.8	25.4	27.8	2002	25.8	26.0	25.9
1995	27.4	25.9	27.9	2003	27.5	28.8	28.8
1996	26.5	25.2	27.6	2004	29.5	30.5	31.2
1997	26.4	25.5	28.2	2005	29.9	31.5	33.1

GI: gross investment (gross capital formation); GDS: gross domestic saving; GAS = GDS + TER (in Table 1): gross aggregate saving Source: Author's estimates based on data from (i) Table 1, and (ii) World Bank's World Development Indicators database. See Appendix 1 for the details.

This evidence certainly does not suggest that inflow of foreign capital pushed the investment rate above the domestic saving rate in developing countries. Quite the contrary; it actually suggests that inflow of foreign capital pushed domestic investment below what could have been supported by domestic saving alone. In short, what the evidence really tells us is that foreign capital actually displaced or crowded out domestic capital.

Table 3: Foreign Currency Reserves of Medium-Income Developing Countries (as % of Their GDP, 1990-2005)

1990	4.9	1996	10.0	2001	12.6
1991	6.5	1997	10.3	2002	15.2
1992	6.9	1998	11.0	2003	17.7
1993	7.9	1999	12.0	2004	20.1
1994	8.5	2000	11.2	2005	21.3
1995	9.2

Source: Author’s estimates based on data from World Bank’s World Development Indicators database. See Appendix 1 for details.

In parallel, there was rapid growth in foreign currency reserves of this group of countries over the same period. As the data in Table 3 show, the combined foreign currency reserve of these countries, as a percentage of their combined GDP, increased steadily from less than 5% in 1990 to more than 21% in 2005.

2 Exogenous Capital Inflow, Crowding-Out and Reserve Accumulation

In what follows, we develop a theoretical argument to explain why foreign capital tends to displace domestic capital in developing countries. The argument, in outline, is as follows. At least since the mid-1980s, most developing countries have been pursuing export-oriented economic growth. This makes it necessary for them to maintain a fixed or depreciating real exchange rate; any appreciation hurts export growth and hence undermines economic growth. Understandably, therefore, preventing nominal exchange rate appreciation and restraining inflation have emerged as the most important goals of macroeconomic policy in these countries. In such a context, substantial inflow of foreign capital presents a developing country with rather awkward problems of macroeconomic management. For, substantial monetary expansion is required to prevent appreciation of the nominal exchange rate, but this risks fuelling inflation. A way out of this dilemma lies in sterilisation, which is effective insofar as money supply is controlled directly rather than through manipulation of the interest rate. Sterilisation, however, has its costs; it diverts the flow of domestic saving away from domestic investors to the central bank.

Thus when macroeconomic policy is focused on maintaining a fixed or depreciating real exchange rate, substantial inflow of foreign private capital tends to generate a growing surplus of domestic saving, which is accumulated by the government as foreign currency reserve. Inflow of foreign saving, because it induces outflow of domestic saving, has little impact on the overall investment rate. What happens is an increase in the share of foreign investment in total investment in the recipient country.

A formal presentation of the argument requires outlining a macroeconomic model for an open developing economy. The simple model that I propose is characterised by the following equilibrium conditions:

Y = C(DY) + Id(r) + If + G – [M(Y,q) – X(q)], If = Sf.e/Pd …(1)

+ -+ - + MS+ ĮSf. e = Pd. L(Y, r) + Sf. e, 0 d Įd 1 …(2)

+ where Y is aggregate output; DY (= Y-T) is private disposable income; T is exogenously given tax revenue of the government; C is private consumption; Id is investment (private and public) financed by domestic saving (private and public); r is an index of money market tightness (further explained below); If is investment

100 financed by foreign saving; G is exogenously given government consumption; M represents imports of goods and services; X represents exports of goods and services; q is real exchange rate; MS is the part of money supply that would be there if net inflow of foreign saving is zero; Į is an index of sterilisation; Sf is net inflow of foreign saving (in units of foreign currency); e is nominal exchange rate (units of domestic currency per unit of foreign currency); Pd is domestic price level; and L is the domestic demand for money. Real exchange rate q is defined as (e. Pf / Pd), where Pf is the exogenously given foreign price level. Y, C, Id, If, G, M and X are all in real terms. The signs below variables indicate the signs of the relevant partial derivatives.

Equation (1) defines the equilibrium in the product market. Total investment in the economy is the sum of its two components: investment undertaken by domestic entrepreneurs who depend on domestic saving and investment undertaken by foreign entrepreneurs who depend on foreign saving. The former is affected by money market tightness while the latter is exogenously given. The postulation that investment financed by domestic saving is a decreasing function of money market tightness rather than simply of interest rate reflects recognition of the fact that monetary authorities in developing countries control money supply more through direct means such as manipulation of the reserve ratio and open market operations than through manipulation of the interest rate.

It is assumed that capital inflow is exogenous so that a recipient country can neither know a priori nor influence through policy the quantity of foreign private capital it might receive over a certain period. It is further assumed that the entire amount of foreign finance received by a country in any given period is actually invested (i e, external resources do not finance domestic consumption). Private consumption is assumed to depend only on the level of private disposable income. The import and export demand functions are fairly standard and require no explanation.3

Equation (2) defines the asset market equilibrium. The righthand side of the equation shows that any inflow of foreign capital generates additional demand for money, the quantity of which depends on the nominal exchange rate. The extent to which money supply actually expands in response to the additional demand depends on the extent to which monetary authorities resort to sterilisation. This is shown on the left-hand side of the equation, where Į (0 d Į d 1) represents the degree of sterilisation; the value of Į is 1 when there is no sterilisation and 0 when there is full sterilisation. Obviously, sterilisation can only affect the credit supply available to domestic investors; the demand for money from foreign investors must always be met.4

Equations (1) and (2) together define an IS-LM framework that incorporates the specificities of developing country situations.5 The conventional question in open economy macroeconomics concerns the effects of monetary and fiscal policies on output and current account under fixed and floating exchange rate regimes. The question is appropriate in the conventional framework in which changes in monetary and fiscal policies are autonomous and induce capital inflow or outflow. But the question is not appropriate in the context of developing economies, capital inflow to which is exogenous in character. Here changes in the capital

december 10, 2011 vol xlvi no 50

EPW

SPECIAL ARTICLE

account are autonomous and define the problem that monetary, fiscal and exchange rate policies must address. The appropriate question, therefore, is: what sort of monetary, fiscal and exchange rate policies can developing countries sensibly pursue in the face of exogenous changes in capital inflow?

The best way of answering the question is to consider the effects of exogenous changes in capital inflow, in the nominal exchange rate and in the price level on output and money market tightness. By using the IS-LM framework outlined above, the three following propositions can be shown to hold when there is no sterilisation:

It is clear that increased capital inflow brings economic benefits in the form of increased investment and output when the nominal exchange rate and the domestic price level are either fixed or falling. Strikingly, however, money market tightness increases even when there is no sterilisation so that investment financed by domestic saving must fall. Thus increased inflow of foreign saving, through its effect on output, increases domestic saving but, at the same time, renders a part of it surplus. However, this outcome can, in principle, be avoided by increasing the money supply by a quantity sufficiently larger than the domestic currency value of capital inflow at the given exchange rate to induce a decline in money market tightness. But this is worthwhile only if there are no inflationary consequences. In short, if the possibility of inflation can be ruled out so that money supply can be expanded as necessary, any amount of foreign capital can be absorbed as additional investment with beneficial effect on output.

It is also clear that if either the nominal exchange rate appreciates or the domestic price level rises, increased capital inflow brings no benefit and inflicts cost in the form of reduced output. The reason, of course, is that these changes increase the trade

Economic & Political Weekly

EPW

deficit. Moreover, they reduce exports and hence undermine industrialisation. Thus, there are good reasons why developing countries should want to ensure that increased inflow of foreign capital does not cause either appreciation of the nominal exchange rate or inflation.

There is a potential linkage, however, between exchange rate movements and price movements, which has so far been left out of account. A depreciation (appreciation) of the nominal exchange rate tends to increase (reduce) the domestic price level by increasing (reducing) the prices of imported goods, which, in the case of developing countries, consist largely of capital and intermediate goods. This means that depreciation or appreciation of the nominal exchange rate need not necessarily translate into depreciation or appreciation of the real exchange rate. This, in turn, means that increased capital inflow has uncertain effects on output and money market tightness when it is associated with either depreciation or appreciation of the nominal exchange rate. It is arguable, however, that prices in developing countries tend to be upwardly flexible but downwardly rigid. Thus, price rises in consequence of currency depreciation are more likely than are price declines in consequence of currency appreciation. So the positive effects of currency depreciation are less certain than are the negative effects of currency appreciation.6

On the whole, therefore, there is a fairly strong case for maintaining a fixed nominal exchange rate in the face of exogenous rises in capital inflow.7 But a fixed nominal exchange rate does not by itself ensure a fixed real exchange rate. For, maintaining a fixed nominal exchange rate in the face of an exogenous increase in capital inflow necessarily means an expansion of money supply and hence in aggregate demand. Such expansion of aggregate demand almost invariably generates inflationary pressure because there are important supply bottlenecks in non-traded sectors of developing economies.8 Hence efforts to maintain a fixed nominal exchange rate result in appreciation of the real exchange rate.

This problem can be dealt with through a policy of sterilisation, which can be used to maintain a fixed domestic price level. But this has a price. Using our framework, the following proposition can be shown to hold:

Given fixed nominal exchange rate and domestic price level, the positive effect of an increased inflow of foreign capital on output grows weaker and the positive effect on money market tightness grows stronger as the degree of sterilisation rises, i e, as the value of Į declines from 1 toward 0 (Case 1, Appendix 2).

Thus the higher the degree of sterilisation that is required to prevent a rise in the domestic price level, the greater is the squeeze on availability of domestic credit and hence the lower is the investment undertaken by the domestic investors. Essentially, sterilisation restrains the growth of aggregate demand for a given increase in capital inflow through a crowding-out of investment financed by domestic saving.

The main point that emerges from the analysis is that an exogenous rise in capital inflow usually results in a crowding-out of investment financed by domestic saving because developing countries need to maintain both a fixed nominal exchange rate and a low rate of inflation. Accumulation of foreign currency reserve is a consequence of this crowding-out. This is intuitively

SPECIAL ARTICLE

obvious, but can also be made explicit. The balance of payments equilibrium for the economy is given by: 'FCR.e/Pd = Sf. e/Pd – [M(Y,q) – X(q)] …(3)

+ -+ where 'FCR is change in foreign currency reserve (expressed in foreign currency unit). The condition for product market equilibrium can also be written as:

Id(r) + If = Sp(DY) + (T – G) + [M(Y,q) – X(q)] …(4)

-+ + -+ where Sp is domestic private saving. Recognising that total domestic saving (S) is the sum of private saving Sp and government’s revenue surplus (T – G) and that If = Sf. e/Pd, and rearranging the terms, we get:

'FCR.e/Pd = Sf. e/Pd – [M(Y,q) – X(q)] = S(Y) – Id(r) …(5)

+ -+ + -

Thus a positive change in foreign currency reserve results from an increase in the surplus of domestic saving over domestic investment, and the growth of this surplus is a direct consequence of the crowding-out of domestically financed investment by foreign investment.9 As we have seen, an exogenous increase in capital inflow increases both output and money market tightness even in the absence of sterilisation so that domestic saving rises while investment financed by domestic saving declines. Thus, even in the absence of sterilisation, we should expect increased capital inflow to be associated with increased foreign currency reserve. Sterilisation strengthens the tendency; for a given increase in capital inflow and a fixed nominal exchange rate, the higher the degree of sterilisation required for holding the domestic price level constant, the higher is the increase in foreign currency reserve.

3 Conclusions

It is an old argument that inflow of foreign private capital benefits developing countries by raising their investment rates above their saving rates. The actual experience of the 1990s, when private capital inflow to developing countries increased sharply, effectively refutes this argument. Much of the increased inflow went to relatively advanced and high-saving developing countries, and served to increase their foreign currency reserves rather than their investment rates.

Empirical evidence also suggests that the traditional argument is based on a wrong premise, namely, that it is the existence of

Notes

1 An indirect confirmation of this comes from the fact that empirical studies have failed to find any positive effect of foreign capital on economic growth in recipient developing countries. See, for example, Kose, Prasad, Rogoff and Wei (2009); Prasad, Rajan and Subramanian (2007); and Rodrik and Subramanian (2009).

2 A detailed list of the medium-income developing countries included in the group is provided in Appendix 1.

3 Following the standard practice, it is assumed that the Marshall-Lerner condition is satisfied.

4 Sterilisation involves two operations by the central bank; it first creates money to purchase the foreign currency that foreign investors bring in and then sells government securities to withdraw money from circulation. It is obviously unlikely that the foreign investors who have just

102 excess demand for investment finance in developing countries that induces inflow of foreign private capital. From the recipient countries’ point of view, inflow of foreign private capital is in fact exogenous in character. This means that even limited openness makes the capital account the main driver of macroeconomic balance. It also means that developing countries’ efforts to “attract” foreign private capital through provision of special incentives are seriously misplaced.

The reason why exogenous rises in capital inflow increase the foreign currency reserves of the recipient countries rather than their investment rates is relatively simple. Developing countries need stable real exchange rates in order to pursue export-oriented growth. This obliges them to guard against appreciation of their currencies on the one hand and rise in inflation on the other. An exogenous increase in capital inflow, therefore, poses a rather awkward problem of macroeconomic management: pre-emption of nominal exchange rate appreciation requires substantial expansion of money supply, which threatens to accelerate inflation. The standard policy response to this problem is sterilisation. But sterilisation leads to crowding-out of domestic capital by foreign capital so that increased capital inflow adds little to aggregate investment. Increase in foreign currency reserve is the counterpart of this crowding-out and can be viewed as officially mediated capital outflow.

Inflow of foreign private capital, therefore, cannot augment investment in recipient developing countries. It can at best help in modernisation and restructuring of the already existing capital stock, and crowding-out of domestic capital can be seen as a part of this process. It is possible that in certain situations, where a developing country already possesses a sizeable but technologically dated capital stock, the benefit of restructuring exceeds the cost of crowding-out. What is certain is that poor countries, with little capital stock and low saving rates, cannot expect to achieve accelerated economic growth with the help of foreign private capital. By allowing unrestricted inflow of foreign private capital, they can only undermine their own domestic investors, lend their own savings to developed countries and increase their dependence on foreign investors for sustaining any kind of economic growth. Maintaining controls over capital flows is a sound policy stance for most

developing countries.

december 10, 2011

Calvo and Mishkin (2003) discuss the reasons why the “fear of floating” tends to be pervasive.

8 This is an aspect that receives much emphasis in the so-called Australian model of the open economy. See, for example, Salter (1959), Swan (1963) and Corden (1977).

9 It should be noticed that a saving surplus also arises when a country runs a trade surplus. So the foreign currency reserve of a developing country can increase either because of increased inflow of foreign capital or because of increased trade surplus or because of both.

References

Calvo, Guillermo and Carmen Reinhart (2002): “Fear of Floating”, Quarterly Journal of Economics, Vol 117, No 2.

Calvo, Guillermo and Frederic S Mishkin (2003): “The Mirage of Exchange Rate Regimes for Emerging

vol xlvi no 50

EPW

SPECIAL ARTICLE

Market Economies”, Journal of Economic Perspectives, Vol 17, No 4.

Corden, W M (1977): Inflation, Exchange Rates and the World Economy (Oxford: Clarendon Press).

Hausmann, Ricardo, Ugo Panizza and Earnesto Stein (2001): “Why Do Countries Float the Way They Float?”, Journal of Development Economics, Vol 66, No 2.

Kose, M A, E Prasad, K Rogoff and S-J Wei (2009): “Financial Globalisation: A Reappraisal”, IMF Staff Papers, Vol 56, No 1.

Prasad, E, R Rajan and A Subramanian (2007): “Foreign Capital and Economic Growth”, Brookings Papers on Economic Activity, 1: 153-209.

Rodrik, D and A Subramanian (2009): “Why Did Financial Globalisation Disappoint?”, IMF Staff Papers, Vol 56, No 1.

Salter, W E G (1959): “Internal and External Balance: The Role of Price and Expenditure Effects”, Economic Record, Vol 35, pp 226-38.

Swan, Trevor W (1963): “Longer-run Problems of the Balance of Payments” in Heinz W Arndt and W Max Corden (ed.), The Australian Economy: A Volume of Readings (Melbourne: Cheshire).

Appendix 1

El Salvador, Guatemala, Guyana, Honduras, Jamaica, Mexico, Paraguay, Peru, Uruguay

Asia-Pacific: China, Fiji, India, Indonesia, Korea (Republic), Malaysia, Pakistan, Papua New Guinea, Philippines, Sri Lanka, Thailand

Middle East-Africa: Cameroon, Congo (Republic), Côte d’Ivoire, Egypt, Ghana, Kenya, Mauritius, Morocco, Tunisia, Turkey, Zimbabwe

Data, Definitions and Sources

The data on inflow and outflow of FDI are from UNCTAD’s Handbook of Statistics. FDI is defined as investment that is made to acquire a lasting management interest (10% or more of voting stock) in an enterprise in a country other than that of the investor (defined according to residence), the investor’s purpose being an effective voice in the management of the enterprise.

The data on net official inflow are from World Bank’s Global Development Finance database and correspond to what are called “official net transfers”. The latter are defined as official flows on long-term debt to official creditors (excluding IMF) plus official grants (excluding technical cooperation) minus “principal repayments” minus “interest payments”.

r Pd[(1 – CY + MY). L + Id . LY ] Pd[(1 – CY + MY). L + Id . LY]

From the first equation, it can be seen that when dT = dSf = de = dPd = dG = dq = ,if Į = 1. Declines as Į declines. Ambiguous when Į=

dY Id for all values of Į if L”Id.

rrr

= GU ±0 (1 – CY + MY) (Į±. e . dSf LY . dSf drEQ = – The IS curve is downward sloping. Similarly from the second equation, Pd[(1 – CY + MY). L + Id . LY ] Pd[(1 – CY + MY). L + Id . LY]

GU / Case 2: de , dPd = dG = dT = dSf = dMS = The LM curve is upward sloping. (– Mdq + Xdq + Sf . de) . L (Į – 1). de . Sf .Id

Rearranging the terms and writing in a matrix form, we get: qqrr

dYEQ = + 1 – CY + MY Idr dY (1 – CY + MY). L + Id . LY (1 – CY + MY). L + Id . LY