An Information-Based Model of Foreign Direct Investment: The Gains from Trade Revisited

Peter Isard, Andrew Rose, and Assaf Razin
Published Date:
January 2000
  • ShareShare
Show Summary Details


Eitan Berglas School of Economics, Tel Aviv University, Tel Aviv 69978, Israel


Eitan Berglas School of Economics, Tel Aviv University, Tel Aviv 69978, Israel


School of Economics and Finance, University of Hong Kong, Pokfulam Road, Hong Kong


The financial aspects of foreign direct investment (FDI) are the focus of this paper. The gains from trade argument (applied to intertemporal trade) is re-examined in this case of informational-asymmetry-driven FDI. FDI is observed to be a predominant form of capital flows to emerging economies, especially when they are liquidity-constrained internationally during a global financial crisis. We analyze the problem of channeling domestic savings into productive investment in the presence of asymmetric information between the managing owners of firms and portfolio stakeholders. We explore the role played by FDI in reviving equity-financed capital investment for economies plagued by such information problems. In the presence of information asymmetry, the paper identifies how FDI gives rise to foreign overinvestment as well as domestic undersaving. We show that the gains from trade could be sizable when the domestic credit market is either under-developed or failing as a result of a financial crisis. But with a well-functioning domestic credit market, the gains turn into losses. Surprisingly, capital may flow into the country even when the autarkic marginal productivity of capital in the domestic economy falls short of the world rate of interest. In such a situation, capital should have efficiently flown out rather than in, and FDI becomes a social loss-generating phenomenon.

1. Introduction

The financial turmoil in East Asia was both a consequence of, and a trigger for, severe international illiquidity.1 Despite their being liquidity-constrained internationally when foreign bank lending and foreign portfolio equity flows dry up, the Asian crisis economies continue to be recipients of large foreign direct investment (FDI) flows, which remarkably have not declined at all (see Figure 1). Similarly, the liquidity problems associated with the debt crisis in the early 1980s and with the currency collapse in Mexico in the mid-1990s demonstrate also the resilience of FDI to financial crises. These striking episodes underscores an essential role of FDI: it serves as a major link between the domestic capital market and the world capital market when other types of international financial investment become deficient. This paper develops a model aimed at highlighting this important role of FDI, and addresses the issue of whether this type of international capital market link through FDI is in general also beneficial from the social welfare perspective.

Figure 1.FDI flows, foreign portfolio flows, and other flows to developing countries.

Source: International Monetary Fund, World Econmic Outlook Database

1 Aggregate Flows to Argentina, Bolivia, Brazil, Chile, Columbia, Cote d’lvoire, Ecuador, Mexico, Morocco, Nigeria, Peru, the Philippines, Uruguay, and Venezuela

2 Aggregate Flows to Indonesia, Korea, Malaysia, the Philippines, and Thailand

In a formal sense, foreign acquisition of shares in domestic firms is classified as FDI when the acquired shares exceed a certain fraction of ownership (usually, 10-20%). From an economic point of view, however, FDI is not just a purchase of a sizable share in a company but, more importantly, is an actual exercise of control and management. The exercise of control enables the foreign direct investors better access to hard-to-get information about the acquired firm’s current and potential performance than what is generally available to a minority shareholder.

Frequently, there is a significant asymmetry in information between the managing stockholders (owner-managers) and other portfolio stakeholders (such as the debt and small equity holders). In the absence of FDI, this informational asymmetry causes a market failure which can be quite severe in the case of equity-financed capital investment. In such case, the equity market shrinks to a “lemons” market a laAkerlof (1970), implying a severe shortage of financing for capital investment. We show here that FDI has an essential role to play in restoring the proper functioning of the domestic equity market for capital investment. As indicated, though, such a market will not be fully efficient because it leads to foreign over-investment and domestic under-saving.2

The inefficiencies associated with FDI finance may sometimes dominate the traditional gains from trade emanating from directing capital from the world market (where the rate of interest is relatively low) to the domestic capital market (where the return to capital is relatively high). Furthermore, the inefficient domestic capital market may attract FDI even when the domestic social rate of return to capital is below the world rate of interest. In such a case, financial liberalization may even misdirect the world capital flows. Accordingly, we demonstrate how the existence of informational-asymmetry-driven FDI can actually turn the gains from international (intertemporal) trade into strict losses.

Our focus on the welfare consequences of FDI from a finance-based perspective is different from most of the existing literature, which focuses on the various explanations for the widespread flows of FDI in the world economy. The international trade literature has examined closely the interactions between FDI and international trade, highlighting the role of multinationals and of FDI in explaining intra-industry trade (see Helpman (1984) and an empirical analysis by Wickham and Thompson (1989)). Brainard (1993a, 1993b) provides a useful empirical assessment of the tradeoff between the proximity to the market and the advantages of concentration in production. FDI is the main instrument that affects this tradeoff. Goldberg and Klein (1997) investigate the relation among trade, FDI, and the real exchange rate. As an extension to this literature, we identify in this paper a distinct mechanism associated with the financial aspects of FDI that turns out to be important in determining its welfare consequences as well.

The organization of the paper is as follows. Section 2 develops an FDI-equity model without a domestic credit market and examines the welfare gains (losses) from FDI. Section 3 introduces a well-functioning domestic credit market and reexamines the gains (losses) from FDI. Concluding remarks are provided in Section 4.

2. Foreign Direct Investment and Equity Finance

In this section and the next, we assume a two-period model of a small, capital-importing country, referred to as the home country. It is assumed that capital imports are channelled solely through foreign direct investment (FDI), due to home-bias in portfolio flows as in Gordon and Bovenberg (1996) and Razin, Sadka, and Yuen (1998a, 1999a). The economy is small enough so that, in the absence of any government intervention, it faces a perfectly elastic supply of external funds at a given risk-free world rate of interest, r*.

Suppose there is a very large number (N) of ex ante identical domestic firms. Each firm employs capital input (K) in the first period in order to produce a single composite good in the second period. We assume that capital depreciates at the rate δ. Output in the second period is equal to F (K)(l + ε), where F (·) is a production function exhibiting diminishing marginal productivity of capital and ε is a random productivity factor. The latter has zero mean and is independent across all firms, (ε is bounded from below by—1, so that output is always nonnegative.) We assume that ε is purely idiosyncratic, so that there is no aggregate uncertainty. Through optimal portfolio decisions, consumer-investors will thus behave in a risk-neutral way.

Investment decisions are made by the firms before the state of the world (i.e., ε) is known.3 Since all firms face the same probability distribution of ε, they all choose the same level of investment. They then seek funds to finance the investment. At this stage, the ownermanagers of the firms are better informed than the outside fund-suppliers. There are many ways to specify the degree of this asymmetry in information. In order to facilitate the analysis, however, we simply assume that the owner-managers, being “close to the action,” observe ε before they make their financing decisions; but the fund-providers, being “far away from the action,” do not.

In this section where investment is equity-financed, the original owner-managers observe ε while the new potential shareholders of the firm do not. The market will be trapped in the “lemons” situation described by Akerlof (1970). At the price offered by the new (uninformed) potential equity buyers, which reflects the average productivity of all firms (i.e., the average level of ε) in the market, the owner-manager of a firm experiencing a higher-than-average value of ε will not be willing to sell its shares and will pull out of the market completely. The equity market will fail to serve its investment financing functions efficiently. We therefore turn to consider another source of equity finance—viz. international capital flows in the form of FDI.

2.1. The FDI-Equity Market Equilibrium

In a formal sense, foreign acquisition of shares in domestic firms is classified as FDI when the shares acquired exceed a certain fraction of ownership (say, 10-20%). From an economic point of view, we look at FDI not just as ownership of a sizable share in a company but, more importantly, as an actual exercise of control management and acquisition of inside information (the value of ε in our model).

Suppose that foreign direct investors purchase domestic companies from scratch, at the “greenfield” stage, i.e., before any capital investment is made.4 For the sake of simplicity and in order to focus on FDI, we ignore, with no loss of generality, all other sectors of the economy in which information is symmetric, and assume that foreign direct investors acquire all the greenfield investment sites. Any single home investor who lacks access to foreign capital markets cannot challenge the foreign direct investors for these sites.5 Note that if a home investor uses only her own fund, she can purchase only a tiny fraction of a greenfield investment site. In such a case, she cannot gain control of the site and the informational advantage entailed by such control.

Upon acquisition and before ε is known, these foreign investors make their capital investment decisions. The realized value of £ is then revealed to them, but not to the potential new equity holders who are solicited to finance the capital investment. Being unable to observe ε, domestic investors will offer the same price for all firms, reflecting the average productivity for the group of low productivity firms they purchase. On the other hand, the foreign direct investors who do observe ε will not be willing to sell at this price the firms which experience high values of ε. Therefore, there will be a cutoff level of ε, say εo, such that all firms which experience a lower value of ε than the cutoff level will be purchased by domestic investors. All other firms will be retained by the foreign direct investors.

Define e as the mean value of ε realized by the low productivity firms:

i.e., e is the conditional expectation of ε, given that ε ≤ εo. For later use, we also denote by e+o) the conditional expectation of ε, given that ε ≤ εo:

Note that the weighted average of eo) and e+o) must yield the average value of ε, that is:

where Φ(·) is the cumulative probability distribution of ε, i.e., Φ(εo) = prob (ε ≤ εo). Equation (3) also implies that eo) < 0 while e+o) > 0, i.e., the expected value of ε for the “bad” (“good”) firm is negative (positive).

The cutoff level of ε is then defined by:

where r¯ is the domestic consumer rate of interest (return). The value of a typical domestic firm in the second period is equal to its output, plus the undepreciated capital, i.e., F (K)(1 + ε) + (1—δ)K. Since domestic equity investors will buy only those firms with ε ≤ εo, the expected second-period value of a firm they buy is F (K) [l + eo)] + (1—δ)K, which they then discount by the factor 1+r¯ to determine the price they are willing to pay in the first period. At equilibrium, this price is equal to the price that a foreign direct investor is willing to accept for the firm which experiences a productivity value of εo. The cutoff price is equal to the expected value of the marginal εo-firm, F (K)(1 + εo) + (1—δ)K, discounted by a factor 1 + r*. Firms that experience a value of ε higher than εo are retained by the foreign direct investors. This explains the equilibrium condition (4).

As eo) < εo, an interior equilibrium (i.e.,—1 < εo < 1) requires that the foreigners’ rate of return (r*) be higher than the residents’ rate of return (r¯). In some sense, this means that domestic investors are “over-charged” by foreign direct investors for their purchases of domestic firms. These foreign investors will not accept a price below {F (K)(1 + εo) + (1—δ)K}/(1 + r*) for the low productivity firms. Note the crucial role of FDI in allowing for an international rate-of-return differential(viz.,r*>r¯). This differential is essential for the existence of an equity market. In an autarkic situation without FDI, there is no rate-of-return differential between the original owner-managers and potential equity buyers. Without this differential, the market will collapse to one of “lemons” and it will be difficult to equity-finance capital investment.

Consider the capital investment decision of the firm that is made before ε becomes known, while it is still owned by foreign direct investors. The firm seeks to maximize its market value, net of the original investment. With a probability Φ(εo), it will be sold to domestic investors, who pay {F(K)[1+e(εo)]+(1Δ)K}/(1+r¯). With a probability [1—Φ(εo)], it will be retained by the foreign investors, for whom it is worth on average {F (K) [1 + e+o)] + (1 - δ)K] /(1 + r*). Hence, the firm’s expected market value, net of the original capital investment, is:

where K—(1—δ)K0 is gross investment and K0 is the initial stock of capital. Maximizing this expression with respect to K yields the following first-order condition:

Equation (6) implies that:

A formal proof of these inequalities is provided in Appendix A.

Notice that the “textbook presumption” is that in the absence of capital flows (i.e., in a financial autarky) the domestic net-of-depreciation marginal productivity of capital (i.e., F’(Ko)—δ) exceeds the world rate of interest r*. We have shown that the flows of FDI bring down the net-of-depreciation marginal productivity of capital (i.e. F’(K)—δ) below the world rate of interest. Thus, FDI transforms the initial shortage of domestic capital into an overly abundant stock of domestic capital; that is, we have foreign over-investment. Furthermore, the more surprising effect of FDI occurs when the “textbook presumption” of an initial shortage of capital does not hold (i.e., when F’(Ko)—δ < r*). In this case, capital inflows are not fundamentally needed. Nevertheless, the odd nature of incentives which emanate from the domestic equity market, plagued by asymmetric information, elicits capital inflows. These misdirected flows of capital widen the gap between the relatively high world rate of interest and the relatively low net-of-depreciation marginal productivity of capital.

The (maximized) value of V in (5) is the price paid by the foreign direct investors at the greenfield stage of investment. Since the value of s is not known at this point, the same price is paid for all firms. Note, however, that some of the firms are then resold to domestic savers, so that net foreign direct investment (FDI) is equal to:


(Recall that {V[1Φ(εo)]V̂}/Φ(εo) is the price at which the low-productivity firms are resold to domestic savers. It is assumed that the foreign direct investors import capital to finance investment in the high-productivity firms which they retain.

Employing (9), equation (8’) reduces to:

The remaining equilibrium conditions are standard. In the first period, the economy faces a resource constraint stating that FDI must suffice to cover the difference between domestic investment (viz. N[K—(1—δ)Ko]) and national savings (viz., the difference between the first period output and saving, NF (Ko)—c1:

Since foreigners will be able to extract from the home country an amount of 1 + r* units of output in the second period for each unit that they invest in the first period, the home country faces the following second-period budget constraint:6

That is, the second period gross national output (namely, NF (K)—FDI (l+r*)), plus the undepreciated capital, (namely, (1—δ)NK), must suffice to support private consumption (C2). Employing (10), one can rewrite (11) in present value terms as:

Naturally, c1 and c2 are determined by the utility-maximizing consumer-saver. But since they do not have access to the world capital market and can only borrow/lend from the domestic market, these consumer-savers will choose c1 and c2 by equating their intertemporal marginal rate of substitution to the domestic default-risk-free (rather than the world) rate of return:

where u is the consumer’s utility function.

In this model the six equations—(4), (6), (8), (10), (11) and (13)—determine the six endogenous variables—εo, r, K, FDI, c1, and c2.

2.2. Gains from Trade

We have demonstrated the crucial role of FDI in sustaining a non-“lemon”-type domestic equity market, albeit not a fully efficient one. There is also a second role, the traditional “gains from trade” role of directing foreign savings into domestic investment. To flash out in a simplified manner the kind of gains or losses brought about by FDI, we compare the laissez-faire allocation in the presence of FDI with the closed economy laissez-faire allocation.

The laissez-faire allocation in the presence of FDI is characterized by:

(see equation (7)). The first inequality states that the return accruing to domestic savers (i.e., r¯) is below the return to the economy (i.e., F’(K)—δ). This indicates that domestic households under-save. The second inequality states that the return accruing to the economy from the inflow of foreign capital (i.e., F’(K)—8) is below its cost to the economy (i.e., r*), indicating excessive capital inflows. Therefore, the laissez-faire allocation with FDI is characterized by domestic under-saving and foreign over-investment.7

In a closed economy, the original owner of a greenfield investment site cannot finance capital outlays on her own. She has to appeal instead to the domestic equity market. In this case the economy will be trapped in a “lemons”-type equilibrium. Any owner-manager of a firm realizing a higher-than-average productivity factor (ε) will pull out of the market which prices all firms according to their average productivity. The market thus shrinks to “lemons,” the shares of the firms with the lowest productivity. Strictly speaking, no new investment can be financed through the equity market and all firms will simply produce with their initial stock of capital which is Ko.8 In this autarkic economy there is no capital market and there is neither saving nor investment.

In this case with no domestic credit, FDI has conflicting effects on welfare. Its first crucial (and unique to this model) role discussed above is to facilitate the channelling of domestic saving into domestic investment. This, by itself, is welfare-enhancing. But, as we have already indicated, FDI is driven also by distorted incentives and its traditional role of directing foreign savings into domestic investment generates an excessive stock of domestic capital. (Either when capital inflows were not all needed or when they were needed to start with, too much of them took place). This foreign over-investment (coupled with domestic under-saving) tends to reduce welfare.

We use numerical examples to illustrate the total effect of FDI on welfare. In these examples, we employ a logarithmic utility function (u (c1, c2) = In (c1) + γ In (c2)), with a subjective discount factor γ, a Cobb-Douglas production function (F (K) = AKα), and a uniform distribution of ε defined over the interval [—a, a].9 The welfare gain (loss) is measured by the uniform percentage change (in c1 and c2) which is needed in order to lift the autarkic utility level to the FDI utility level. These welfare gains (losses) are calculated for various levels of the world rate of interest r*, ranging from 0 to 0.45. (These rates correspond to real annual rates of interest, ranging from 0% to 1.25% on a 30-year period interval.) As shown in Figure 2, the welfare gain numbers are reasonably large. At levels of r* lower than the autarkic net-of-depreciation marginal productivity of capital (F’(Ko)—δ = 0.06), these numbers reflect a combination of the traditional gains (i.e., capital cost saving) and the two conflicting welfare effects of FDI mentioned in the preceding paragraph. The latter is still strongly positive, equal to almost a 3.7% increase in lifetime consumption, even when the former is absent, i.e., when r* = 0.06. These huge gains are due to the large inflow of capital, amounting to about 60% of GDP, that finances about 50% of the firms with FDI-equity and indirectly the consumption loans made by the domestic households (at a saving rate of—43%). Naturally, the gains come down as the gap between the autarky interest rate and the world interest rate widens. Observe, though, that even when the gap is negative so that there is no economic need for capital inflows to start with, there may still exist welfare gains.10

Figure 2.Welfare gain/loss from FDI without domestic credit

3. FDI with Domestic Equity and Credit Markets

In the preceding section, we have demonstrated that the gains from trade brought about by FDI can be quite sizable. In fact, FDI may fulfill several roles. It may create an active (albeit distorted) domestic stock market that facilitates the channelling of domestic savings to finance new domestic investment. It may also facilitate the channelling of foreign savings to the domestic stock market to help finance part of the new investment. This is why the gains from trade through FDI can be rather substantial.

However, when a domestic credit market is doing most of the job of channelling domestic savings into domestic investment, the role of FDI diminishes. In fact, it is often observed that FDI is highly leveraged domestically. After gaining control of the domestic firm, a foreign direct investor usually resorts to the domestic credit market to finance new investment and possibly trade shares of the firm in the domestic equity market later on after profits from its original investment are realized. We thus extend in this section the model of the preceding section to include a domestic credit market. We then demonstrate, somewhat surprisingly, that not only the gains from trade through FDI diminish, but they can well be significantly negative.

3.1. The FDI-Equity-Credit Equilibrium

The sequencing of firm decisions is as follows. Before ε is revealed to anyone (i.e., under symmetric information), foreign investors bid up domestic firms from their original domestic owners, investment decisions are made, and full financing through domestic credit is secured. Then, s is revealed to the owner-managers (who are all foreigners), but not also to domestic equity investors. At this stage, shares are offered in the domestic equity market and the ownership in some of the firms is transferred to the domestic investors. The foreign direct investors are able in the initial stage (i.e., before e is revealed to anyone) to outbid the domestic savers because the latter lack access to large amounts of funds necessary in order to seize control of the firms while the former, by assumption, are not liquidity constrained.

Since credit is extended ex ante, before ε is revealed, firms cannot sign default-free loan contracts with the lenders. We therefore consider loan contracts which allow for the possibility of default. We adopt the “costly state verification” framework a laTownsend (1979) in assuming that lenders make firm-specific loans, charging an interest rate of rj to firm j (0 ≤ j ≤ N).11 The interest and principal payment commitment will be honored when the firms encounter relatively good shocks, and defaulted when they encounter relatively bad shocks. The loan contract is characterized by a loan rate (rj), with possible default, and a threshold value (ε¯j) of the productivity parameter as follows:

When the realized value of εj is larger than ε¯j, the firm is solvent and will thus pay the lenders the promised amount, consisting of the principal Kj—(1—δ)Koj plus the interest rj[Kj - (1 - δ)Koj] as given by the right-hand-side of (14). If, however, εj<ε¯j, the firm will default. In the case of default, the lenders can incur a cost in order to verify the true value of εj and to seize the residual value of the firm. This cost, interpretable as the cost of bankruptcy, is assumed to be proportional to the firm’s realized gross return, μ[F (Kj)(1+εj)+(1-δ)Kj], where μ ≤ 1 is the factor of proportionality. Net of this cost, the lenders will receive (1 - μ)[F (Kj)(1 + εj) + (1 - δ)Kj].

Since there is no aggregate risk, the expected rate of return required by domestic consumersavers, denoted by r¯, can be secured by sufficient diversification,. Therefore, the “default” rate of interest, rj, must offer a premium over and above the default-free rate, r¯, according to:

The first term on the left-hand-side of (15’) is the contracted principal and interest payment, weighted by the no-default probability. The second term measures the net residual value of the firm, weighted by the default probability. The right-hand-side is the no-default return required by the domestic lender. Observe that (14) and (15’) together imply that:

Since e(ε¯j)<ε¯j and μ ≥ 0, it follows that rj>r¯, the difference being a risk-premium (which depends, among other things, on Kj, ε¯j, and μ).

The firm in this setup is competitive (price-taker) only with respect to r¯, the market default-free rate of return. This r¯ cannot be influenced by the firm’s actions. However, rj, Kj, and ε¯j are firm-specific and must satisfy equations (14) and (15’). In making its investment, Kj—(1—δ)Koj, and its financing (loan contract) decisions, the firm takes these constraints into account. Since these decisions are made before ε is known, i.e., when all firms are (ex-ante) identical, they all make the same decision. We henceforth drop the superscript j.

The remainder of this section proceeds along similar lines as the preceding section. In the equity market which opens after ε is revealed to the (foreign) owner-managers, there is a cutoff level of ε, denoted by εo (generally different than the corresponding εo of the preceding section), such that all firms experiencing a value of ε above εo will be retained by the foreign direct investors and all other firms (with ε below εo) will be sold to domestic savers. This cutoff level of ε is given by:


is the conditional expectation of the ε’s between ε¯ and εo.

Notice that firms that experience a value of ε below ε¯ default and have zero value. These firms are not retained by the foreign direct investors; hence εoε¯. All other firms generate in the second period a net cash flow of [F (K)(1 + ε) + (1 - δ)K] -(1+ r)[K - (1 - δ)Ko]. The left-hand-side of (16’) represents the marginal (from the bottom of the distribution) firm retained by foreign investors. The right-hand-side of (16’) is the expected value of the firms that are purchased by domestic savers. With a conditional probability of [Φ(εo)Φ(ε¯)]/Φ(εo), they generate a net expected cash flow of F(K)[1+ê(ε¯,εo)]+(1δ)K}(1+r)[K(1δ)Ko], with a probability of Φ(ε¯)/Φ(εo) they generate a zero net cash flow. This explains equation (16’).

We can substitute equation (14) into (15’) and (16’) in order to eliminate r and then rearrange terms to obtain:


Consider now the capital investment decision of the firm that is made before ε becomes known, while it is still owned by foreign direct investors. With a probability of Φ(εo)Φ(ε¯) it will be sold to domestic savers who pay a positive price equalling

which reduces to F(K)[ê(ε¯,εo)ε¯]/(1+r¯), where use is made of (14). With a probability of 1—Φ(εo), it will be retained by the foreign investors for whom it is worth

where use is made of (14). Hence, the firm seeks to maximize

subject to constraint (15), by choice of K and ε¯,. given εo.12 This maximization yields the following first-order conditions:


where λ is a Lagrange multiplier. Our numerical simulations reported below suggest that in this case too there will be domestic undersaving and foreign overinvestment: r¯<F(K)δ<r*.

The (maximized) value of V in (18) is the price paid by the foreign direct investors at the greenfield stage of investment. Since the value of ε is not known at this point, the same price is paid for all firms. As in the preceding section, the low-ε firms are then (after ε is revealed to the foreign direct investors) resold to domestic savers, all at the same price, because ε is not observed by these savers. Net capital inflows through FDI are given by: (see equation (18)). Unlike the preceding section (with no domestic credit), in this section all capital outlays are financed domestically and FDI consists only of the price paid for the ownership and control of the high-ε firms.

The remainder of the equilibrium conditions is standard. The first-period resource constraint is given by:

The second-period resource constraint is

Note that the last term on the left-hand-side of (23) reflects the existence of real default costs. Finally, utility maximization implies that

In this model the eight equations ((15), (16), (19)-(24)) determine the eight endogenous variables (K,ε¯,εo,c1,c2,r¯,FDI,λ)..

3.2. Gains from Trade

Unlike the preceding case of no domestic credit, an autarkic economy in this case can utilize domestic savings to debt-finance domestic investment. The crucial role of FDI as a vehicle for sustaining a domestic equity market through which domestic savings are channelled into domestic investment is thus substantially diminished. Consequently, the negative effect of FDI associated with the distorted incentives emanating from the domestic equity market dominates, and altogether there may exist a net welfare loss from trade. Figure 3 illustrates the welfare gains and losses occurring at various levels of the world rate of interest, r*, for the same set of parameter values as in Figure 2.13 Except for levels of r* ranging from 4.1 to 4.7 (equivalent to an annual real rate of 5.58% to 5.97%) where some minimal welfare gains of 0.04% to 0.55% are recorded, welfare losses are prominant (about—2% at lower levels of r* and increasing to more than—20% when r* exceeds 4.7) Compared with the case of no domestic credit, the role of FDI in financing domestic investment is much less important. At r* ≤ 4.7, the FDI-GDP ratio is about 6-8% whereas the fraction of firms financed by FDI-equity is less than 40% (and domestic savings are positive, i.e., no FDIfunded consumption loans). Note that the autarkic risk-free interest rate of 2.9 falls short of all the values of r* considered here. So here again, we have the possibility that although the FDI flows are not fundamentally needed, they do nevertheless flow in.

Figure 3.Welfare gain/loss from FDI with domestic credit

4. Concluding Remarks

International capital markets are notoriously imperfect. Indeed, under asymmetric information, the equity market may be plagued by the Akerlof-type lemons problem. In the absence of a well-developed domestic credit market, in which case domestic savings cannot be efficiently channelled into domestic investment, FDI can play a double role. First, it provides a vehicle for reviving the domestic equity market as a channel to direct domestic savings into domestic investment. Second, it supplies foreign savings on top of domestic savings to finance domestic investment in a capital-hungry economy.

The second role of FDI generates the traditional gains from trade to the domestic country, However, its first role, though crucial to the financing of investment projects in the domestic country, is not costless. As the equity market is characterized by asymmetric information, it does not always transmit the “correct” signals about the social rates of return to domestic capital. As a result, there are some welfare losses that may offset some or all of the gains stemming from the mere channelling of domestic savings into domestic investment.

When a well-developed domestic credit market exists, through which domestic savings can be channelled into domestic investment even in the absence of an equity market, the first role played by FDI will no longer generate any gain. On the contrary, the “incorrect” signalling effect entails a strict welfare loss. When FDI can be leveraged domestically (through the domestic credit market), the traditional gains from trade associated with the second role of FDI will be severely curtailed. As a result, the total net effect of FDI on the welfare of the domestic economy could well be negative. As usual in normative economics, social losses are not inconsistent with individual maximization. In particular, although there are possibly losses for the society as a whole due to the information-based distortions, each individual domestic saver is utility-maximizing and the individual foreign direct investors all gain from their investment in the domestic economy.

The recent debate over international capital market liberalization generally focuses on the possible negative effects of short-term capital flows. Some economists even advocate a levy on short-term capital inflows in order to reduce the magnitude of these inflows on the ground that such large inflows could turn around abruptly into uncontrollable capital outflows during financial crises. Although unrelated to financial crises, we have seen here that at least one type of long-term capital inflows—namely, FDI—may also entail some welfare losses.14

Evidently, the gains from trade discussion in this paper has focused entirely on the financial aspects of inward FDI. In the absence of a well-developed credit market, our theory shows that FDI helps revive the domestic equity market and, in so doing, creates incentives for expansion in domestic investment. Interestingly, there is some evidence that shows that FDI in developing countries does tend to “crowd in” domestic investment (see Borensztein, De Gregorio, and Lee (1998)). However, the reader should not lose sight of the potential gains from FDI even when its contribution to the total size of capital inflows is not all that big, as when FDI is heavily domestically leveraged. The potential gains stem from the additional role of FDI as a vehicle for technology transfer from the industrial countries to the emerging market economies. In addition, some recent evidence suggests that FDI may promote competition. The large size of and the advanced technology possessed by multinationals often enable them to compete in industries in which barriers to entry, such as large capital requirements, limit the access of potential firms. Well-known also is the hypothesis that, as a vehicle for technology transfer, FDI contributes to productivity growth. (See World Bank (1999).) In a companion paper (Razin, Sadka, and Yuen (1999b)), we extend the model of this paper to provide a theoretical framework under which gains from FDI can be based not only on the inside information argument, but also on its role in the promotion of competition and growth of productivity in the domestic economy. In many cases, the losses from the former may well be outweighed by the gains from the latter.

Furthermore, if the government applies corrective taxes to counteract the information based distortions associated with FDI, then the otherwise closed economy will be distortionfree and introducing capital flows through FDI will restore the usual gains from trade. These corrective taxes, which take the form of an ex ante corporate subsidy (to stimulate savings in domestic equity), coupled with a tax on FDI (to discourage excessive foreign investment), can be implemented without endowing the government with information about the productivity levels of the firms.15 This underscores the importance of intervention through the imposition of taxes/subsidies in the capital market also in the presence of FDI.

Appendix A: Proof of Inequality (7)

Substituting for 1/(1+r¯) from (4) into (6) and rearranging terms we get:

where x=[F(K)(1+εo)+(1δ)K]/{F(K)[1+e(εo)]+(1δ)K}>1 since εo>e(εo) It follows from (A1) that:

because the term in the curly brackets is equal to one (see equation (3)). This proves the inequality at the right end of (7).

Substitute for 1 + r* from (4) into (6) and rearrange terms to get:

Since x > 1, it follows from (A2) that

which completes the proof of (7).

Appendix B: An Alternative Autarkic Model of Domestic Equity Market with No Domestic Credit

One can restore some investment in our closed economy by modifying the decision-making process of the firm. We can envisage a two-stage decision rule imposed by firm owners on the managers. In the first period, firms determine their investment rules in the planning stage while the actual investment and its funding are delayed to the implementation stage. These investment rules are approved by the owners of the firms before ε is known. The management then implements these rules by seeking funds from the domestic equity market to finance the investment, after ε is known. They are also empowered by the owners not to invest at all when their ε is higher than some threshold level. For further discussion of the rationale behind this sequence of firm decisions, the reader is referred to Razin, Sadka, and Yuen (1998b).

When the value of ε that is revealed to the manager is high enough, she would prefer to employ just the existing capital (i.e., Ko) rather than to raise equity in a market that will not pay a premium for a high value of ε. Thus, there exists a cutoff level of ε, denoted by εo (generally different from the εo in the FDI case), such that all firms which experience a value of ε above εo will not make any new investment, while all other firms (i.e., the low-ε firms) will equity-finance their new investments at a price reflecting the average value of the “lemons.” This cutoff level of ε is defined by:

where K is the stock of capital of the low-ε firms that do make new investments, and (1—δ)Ko is the stock of capital for the high-ε firms that do not make any new investment. Thus, in contrast with the autarkic equilibrium of section 2.2, we have here a positive level of new investment.


Research on this paper was conducted while the authors were visiting the IMF, Razin and Yuen at the Research Department and Sadka at the Fiscal Affairs Department. We wish to thank Evgeny Agronin for very competent research assistance, and Joshua Aizenman and other conference participants for useful comments. Financial support from the RGC through a HKU-CRCG grant and an earmarked grant is gratefully acknowledged.

General Discussion

Tamim Bayoumi noted that most of the existing literature on FDI had focused on attempting to explain its causes. By contrast, Razin, Sadka, and Yuen (RSY) had focused on the welfare consequences of FDI, which was praiseworthy. In Bayoumi’s view, the limited attention that the profession had paid to the welfare consequences of FDI reflected a tendency of macroeconomists to ignore the distinction between GDP and GNR In this context, it would be very interesting to analyze the welfare consequences of the heavy subsidization of inward FDI by Ireland during the 1980s. It is generally assumed that the subsidies resulted in significant welfare gains for Ireland. But Bayoumi noted that part of Ireland’s economic growth over the past decade could be attributed to the process of European integration (as was also the case, for example, in Portugal), and that a careful study that isolated the effects of the subsidization policy might well conclude that Ireland’s GNP had been negatively affected, ceteris paribus.

James Boughton asked how the analysis might be affected by the addition of mutual funds to the set of investors that RSY had considered. Razin answered that without positions of control over firms, mutual funds could not exploit the inside information that the paper had highlighted as a source of distortion. Subsequent discussion involving Razin, Sadka, and Aizenman clarified that the adverse welfare effects of FDI in the RSY model were not entirely attributable to asymmetric information, and that these adverse effects could be mitigated through mutual funds or other means to address what Aizenman had referred to as “financial repression.”

Razin emphasized that the authors did not want their analysis to be interpreted as a negative verdict on FDI. In seeking to flesh out the financial aspects, the RSY model had abstracted from some widely-recognized gains from FDI.


    AizenmanJoshua. (1998). “Capital Mobility in a Second-Best World—Moral Hazard with Costly Financial Intermediation.”Dartmouth College, mimeo.

    AkerlofGeorge. (1970). “The Market for ‘Lemons’: Qualitative Uncertainty and the Market Mechanism.”Quarterly Journal of Economics89488500.

    BarryFrank and JohnBradley. (1997). “FDI and Trade: The Irish Host-Country Experience.”Economic Journal10717981811.

    BrainardS. L. (1993a). “An Empirical Assessment of Factor Proportions Explanation of Multinationals Sales.”NBER Working Paper No. 4580.

    BrainardS. L. (1993b). “An Empirical Assessment of the Proximity-Concentration Tradeoffs Between Multinational Sales and Trade.”NBER Working Paper No. 4269.

    BorenszteinEduardoJose DeGregorio and Jong-whaLee. (1998). “How Does Foreign Direct Investment Affect Economic Growth?”Journal of International Economics45115135.

    ChangRoberto and AndresVelasco. (1998). “The Asian Liquidity Crisis.”NBER Working Paper No. 6796.

    FrenkelJacob A.AssafRazin and EfraimSadka. (1991). International Taxation in an Integrated World Economy. CambridgeMA: MIT Press.

    GoldbergLinda S. and Michael W.Klein. (1997). “Foreign Direct Investment, Trade and Real Exchange Rate Linkages in Southeast Asia and Latin America.”NBER Working Paper 4344.

    GordonRoger H. and A. LansBovenberg. (1996). “Why is Capital So Immobile Internationally?: Possible Explanations and Implications for Capital Income Taxation.”American Economic Review8610571075.

    HelpmanElhanan and AssafRazin. (1983). “Increasing Returns, Monopolistic Competition and Factor Movements: A Welfare Analysis.”Journal of International Economics14263276.

    HelpmanElhanan. (1984). “A Simple Theory of International Trade with Multinational Corporations.” Journal of Political Economy.

    RazinAssafEfraimSadka and Chi-WaYuen. (1999a). “Implications of Home Bias: A Pecking Order of Capital Inflows and Corrective Taxation.” InAssafRazin and EfraimSadka (eds.) The Economics of Globalization: Policy Perspectives from Public Economics. Cambridge University Press.

    RazinAssaf (1999b). “Host Country Benefits from FDI: Two Non-traditional Views on Gains from Trade.” Paper prepared for an NBER EASE conference in HawaiiJune12-141999.

    RazinAssaf (1999c). “Excessive FDI Flows Under Asymmetric Information.” Paper prepared for a Federal Reserve Bank of San Francisco conference.

    RazinAssaf (1998a). “A Pecking Order of Capital Inflows and International Tax Principles.”Journal of International Economics444546.

    RazinAssaf (1998b). “Capital Flows with Debt- and Equity-Financed Investment: Equilibrium Structure and Efficiency Implications.”IMF Working Paper WP/98/159.

    SosnerNathan. (1998). “The Sequence of Investment Decisions as a Solution of an Agency Problem.” The Eitan Berglas School of Economics Tel-Aviv University mimeo.

    StiglitzJoseph E. and AndrewWeiss. (1981). “Credit Rationing in Markets with Imperfect Information.”American Economic Review71393410.

    TownsendRobert M. (1979). “Optimal Contracts and Competitive Markets with Costly State Verification.”Journal of Economic Theory21265293.

    United Nations. (1998). World Investment Report 1998: Trends and Determinant. New York and Geneva.

    WickhamE. and H.Thompson. (1989). “An Empirical Analysis of Intra-Industry Trade and Multinational Firms.” InP. K. M.Tharakan and J.Cole (eds.) Intra-Industry Trade: Theory Evidence and Extensions. London: Macmillan.

    World Bank. (1999). Global Development Finance.

See, for example, Chang and Velasco (1998), who trace the emergence of international illiquidity to the shortening of the foreign debt structure, the mismatch in currency denomination between assets and liabilities, and a vulnerability triggered by financial liberalization.

See Frenkel, Razin and Sadka (1991) for a discussion of these two wedges of inefficiency associated with capital flows and their corrective tax implications.

For a principal-agent foundations for such an economic structure in which investment is precommitted before the realization of the productivity parameter, see Sosner (1998).

For instance, Barry and Bradley (1997) report that “… most of the FDI into Irish manufacturing entails the construction of entirely new state-of-the-art factories on green-field sites...” (p. 1809). This is not the case in Central and Eastern Europe where most of the FDI flows to “brown-field” sites.

The existence of wealthy individuals or families in the home country may possibly limit the scope of our analysis to the extent that they can compete with the foreign direct investors on control over these greenfield investment sites. Our analysis will carry over, however, if they form joint ventures with the foreign direct investors. On the other hand, the foreign direct investors need not be excessively resourceful. Even a small technological advantage they may enjoy over and above the domestic investors will enable them to bid up all these investment sites from the domestic investors and to gain control of these industries.

Note that aggregate output is NF (K), since E{ε) = 0.

One can show that it is possible to partially restore a first-best efficiency (i.e., r¯=F(K)δ=r* by employing a Pigouvian corrective policy which consists of a corporate income subsidy and a tax on capital income of non-residents.

One can restore some investment into this closed economy by modifying the decision-making process of the firm; see Appendix B.

We set the parameter values as follows: γ = 0.295, α = 0.333, δ = 0.723, N = 1, A = 1, Ko = 1, and a = 0.99. Since we think of each period as constituting half of the lifetime of a generation (i.e., about 25 years), the values of γ and δ are chosen to reflect this assumption.

For another example of misdirected capital flows in the context of international trade with differentiated products, see Helpman and Razin (1983).

The εo-condition, as given by equation (16), is determined by equilibrium in the equity market. As such, it will not be taken into account by the price-taking firms when choosing their investment levels.

Here, in order to obtain sensible solutions, we vary r* from 3.85 to 5.35 (equivalent to an annual real interest rate of 5.4% to 6.36%). In addition, we assume that the bankruptcy cost parameter has a value μ = 0.05. Note that there could be two equilibria for levels of r* between 4.6 and 5.0. For a more detailed explanation of this probability of multiple equilibria, see Razin, Sadka, Yuen (1999c).

See also Aizenman (1998) for an alternative modelling of financial intermediation based on moral hazard that generates losses from capital inflows.

Similar analysis of an optimal tax design has been carried out in Razin, Sadka, and Yuen (1998a).



This interesting paper has two main contributions. First, it models FDI in the presence of asymmetric information, where being “close to the action” leads to information advantages. Second, it identifies conditions under which FDI may be welfare reducing, questioning the presumption that long term investment is welfare improving.

The first part of my comments will provide a diagrammatic interpretation of the paper, using a demand/supply, consumer/producer surplus welfare analysis. The second part will put the present paper in the context of the second best literature, and will discuss possible policy interpretations.

Supply/Demand Interpretation of the Model

The insight of the paper may be traced with the help of the familiar diagrammatic analysis of saving and investment decisions. To simplify, let the depreciation rate be zero. A useful benchmark is the full information case, with unfettered access to the capital market, summarized in Figure 1. Output is produced by a large number of small firms. The production of firm i is

Figure 1.Complete information

where εi is i.i.d. shock, with zero mean. Curve MPK plots the expected marginal productivity of capital as a function of the stock of capital; the latter is measured horizontally from point O. Let the initial stock of capital in the emerging market economy be Ko. We plot domestic saving (curve S) using Ko as the modified origin. With risk neutrality, the equilibrium investment is characterized by equating the expected marginal productivity of capital with the interest rate. Hence, the autarky interest rate is r¯A, and the autarky investment is Io. Suppose that this interest rate exceeds the global rate (r*), as is frequently the case in emerging markets. The open economy equilibrium is obtained in panel II of Figure 1. The inflow of capital increases the stock of capital to K*. The drop in the interest rate reduces domestic saving, and the gap between the higher investment and the lower saving is met by the inflow of foreign capital, depicted by I*. The welfare gain from capital inflows is provided by the dotted triangle.

Figure 1 provides us with the benchmark for the more complex case investigated in the present paper. Suppose that domestic agents lack access to the foreign saving market—they can not save at the global interest rate, apparently due to financial repression. In financial autarky, all domestic agents face the same interest rate. The authors show that the information advantage of the owner-manager leads to a lemon equilibrium—the collapse of the market for new capital, akin to Akerlof’s (1970) seminal paper. More precisely, an agent who purchases a firm has an information advantage, being “close to the action,” observing the realized i.i.d. shock. His attempt to sell the firm would signal the inferiority of his firm. As all agents face the same interest rate, no buyer would be interested in buying such a firm, leading to a zero investment equilibrium.

Suppose now that we allow multinationals to invest in the economy. Foreign agents have access to the “deep pocket” of global capital, and they can purchase the stock of capital prior to the realization of productivity shocks. In contrast, domestic agents are too small to bid for the exclusive control of production sites at the green field level. As before, the advantage of gaining such control is being “close to the action.” Inflows of capital introduce a new dimension of heterogeneity, as foreign agents face an interest rate that may differ from the domestic one. The key condition characterizing the resultant equilibrium is equation (4). The inflow of capital leads to a domestic interest rate that is below the foreign one. Foreign agents will retain the control of the superior production sites—firms the realized productivity of which is above a cutoff level of εo. They will sell the inferior sites—firms the productivity of which is below that cutoff level. Domestic agents are willing to buy these firms, because they discount future income using a lower interest rate. The asymmetric information implies that the expected productivity of the firms sold corresponds to the expected productivity of firms below the cutoff rate, εo, denoted by e–(εo). As the expected productivity of all firms is zero, it follows that e–(εo) < 0. The gains from trade stem from the differences in firms’ valuation—domestic agents facing a lower interest rate are valuing more a given future income stream than the corresponding valuation of the same income stream by the foreign owners. The cutoff productivity rate is determined by equating the value of the average firm bought by domestic agents [the LHS of (4)] to the value of the marginal firm sold by the multinationals, i.e., the value of the firm the productivity of which is £0 [the RHS of (4)].

Figure 2.Asymmetric information.

The resultant equilibrium can be traced with the help of Figure 2. In the absence of FDI, the lemon equilibrium would prevail, and the stock of capital would be Ko. With multinationals, the expected marginal productivity of the firms bought by domestic agents is MPK(1 + eo)), where MPK(1 + eo)) < MPK. The modified expected marginal productivity curve is depicted in Figure 2, panel II by the dashed curve to the left of the original curve. The resultant equilibrium is characterized by relatively large FDI. The inflow of foreign capital depresses the expected marginal productivity of domestic capital to r˜, below the international interest rate r* but above the domestic one, r.¯ An interesting feature of this equilibrium is that the marginal productivity of capital is below the level observed with full financial integration and complete information equilibrium (see Figure 1, panel II), manifesting excessive investment even relative to the complete information benchmark.

Figure 2, panel II also identifies the welfare effects of FDI from the vantage point of domestic agents, using the autarky, zero investment as the benchmark. FDI has 2 fundamental welfare effects. First, it leads to the ‘export’ of rents, an effect that may be referred to as “cherry picking” by multinationals. Having access to the “deep pocket” enables multinationals to buy firms at the green field stage, to retain the control of the superior firms, and to sell the inferior firms to domestic agents. This is equivalent to an adverse productivity shock from the point of view of domestic agents, reducing welfare by the dotted trapezoid in Figure 2, panel II (denoted by—). Second, FDI enables the formation of capital, increasing the domestic ownership of capital to Kd. This increases the domestic surplus by the dotted triangle (denoted by +). Hence, FDI is a mixed blessing in such an economy, and the simulation in the paper shows that the ultimate welfare effect is ambiguous. Figure 2, panel II suggests that the balance between ‘rent export’ and capital formation would determine the ultimate welfare effects of FDI. Indeed, in the second part of the paper the authors show that if the domestic financial market is functioning in autarky, supporting domestic investment, then the “capital formation” gain associated with the inflow of capital shrinks, and FDI unambiguously reduces welfare.


This paper is a contribution to the second best literature, showing that FDI can reduce welfare due to existing distortions. The interest in these issues has been revitalized following the financial crises of the nineties, opening the debate regarding the desirability of encouraging free capital flows between the OECD countries and the emerging market economies. This remains a disputable issue, as the ultimate welfare contribution of capital mobility remains elusive [see Rodrik (1998)].

A selective tour of the second best literature in international economics should start with the ‘founding fathers’—Bhagwati and Johnson. Their seminal papers on tariffs and immiserizing growth focused attention on the impact of distortions on growth. They showed that growth magnifies the welfare costs of existing distortions, leading to ambiguous net welfare effects. Brecher and Diaz Alejandro (1977) were among the first to focus on capital mobility in a second best world. The events in Chile in the eighties triggered a lively discussion about the desirable order of reforms. Contributions by McKinnon (1982) and Edwards and van-Wijnbergen (1986) showed that the sequencing of reforms in a second best universe is not obvious, and deserves further attention. More recently, Kohn and Marion (1992) showed in the context of a knowledge-based endogenous growth framework, that opening capital markets does not necessarily improve welfare for the nation or for the world as a whole. In the context of costly risk monitoring by banks, Aizenman (1998) showed that excessive risk undertaken in the presence of moral hazard leads to a distortion, implying that unrestricted capital mobility tends to reduce (increase) welfare if banks are relatively inefficient (efficient).

The novel aspect of the present paper is the focus on FDI, the inflow of which frequently has been viewed more favorably than short term capital inflows. The authors are identifying conditions under which even FDI may be welfare worsening. As is the case with all the above literature, the peril of dealing with second best is that there is no unique policy prescription that follows from the second best equilibrium. Specifically, the policy implications are conditional on the initial distortions, and on the degrees of freedom available to the policy maker.

In the context of the present paper, it is not obvious whether this paper calls for restrictions on FDI. The adverse welfare effects of FDI in the present paper are due to the combination of asymmetric information and implicit financial repression, where domestic agents are not allowed to save at the world interest rate. This separation of markets is needed to support an equilibrium where the domestic interest rate is below the international one. Hence, one may argue that the present paper calls either for limiting FDI in the presence of financial repression, or for removing financial repression in the first place, prior to FDI.1 Obviously, which of these options is more feasible will be determined by both efficiency and political economy considerations, and may vary from country to country.


As is shown by the experience of various countries (including Korea), financial repression may be sustained for long periods, despite the presence of various ‘leaks’ that may mitigate its effects.


    Aizenman J. (1998). “Capital Mobility in a Second Best World—Moral Hazard With Costly Financial Intermediation.”NBER Working paper no. 6703.

    AkerlofG. (1970). “The Market for Lemons: Quality Uncertainty and the Market Mechanism.”Quarterly Journal of Economics89488500.

    BhagwatiJ. N. (1956). “Immiserizing Growth: A Geometrical Note.” Review of Economic StudiesJune.

    BrecherA. R. and C. F.Diaz Alejandro. (1977). “Tariff, Foreign Capital and Immiserizing Growth.”Journal of International Economics7317322.

    EdwardsS. and van-WijnbergenS. (1986) “The Welfare Effects of Trade and Capital Market Liberalization.” International Economic Review 27 (1)February141148.

    JohnsonG. H.“The Possibility of Income Losses from Increased Efficiency or Factor Accumulation in the Presence of Tariffs.” The Economic JournalMarch151154.

    KohnM.MarionN. (1992). “The Implications of Knowledge-Based Growth for the Optimality of Open Capital Markets.”Canadian Journal of Economics25 (4) November pp. 865883.

    McKinnonR. I. and (1982). “The Order of Economic Liberalization: Lessons from Chile and Argentina.” Carnegie Rochester Conference Series on Public Policy17(0) Autumn159186.

    RodrikD. (1998). “Who Needs Capital Account Convertibility?”Manuscript Harvard University.

    Other Resources Citing This Publication