This section reviews three strands of theoretical literature on the effects of financial liberalization: the classic case for liberalized financial markets, models of asymmetric information, and models of domestic distortions.
The Classic Case for Capital Mobility
The classic case for international capital mobility is well known but worth restating. Flows from capital-abundant to capital-scarce countries raise welfare in the sending and receiving countries alike on the assumption that the marginal product of capital is higher in the latter than the former. Free capital movements thus permit a more efficient global allocation of savings and direct resources toward their most productive uses.
Capital mobility creates opportunities for portfolio diversification, risk sharing, and intertemporal trade. Households, firms, and, for that matter, entire countries can borrow when incomes are low and repay when they are high, smoothing consumption. The ability to borrow abroad can thus dampen business cycles by relieving households and firms of the need to sharply compress consumption and investment spending and to thereby depress domestic demand, when domestic production and income turn down (see Greenwald, Stiglitz, and Weiss, 1984). By holding claims on foreign countries, households and firms can diversify risks associated with disturbances that impinge on the home country alone. Companies can protect themselves against cost and productivity shocks by investing in branch plants in several countries across which such shocks are imperfectly correlated. Capital mobility can thereby enable investors to achieve higher risk-adjusted rates of return. Income levels in recipient countries should also rise as a result of the capital inflows. Higher rates of return can encourage saving and investment that deliver faster rates of economic growth.24
In addition, by analogy to the case for free trade in goods, free capital flows allow the global economy to reap the efficiency gains created by specialization in the production of financial services. Some countries will find it more efficient to import than to produce financial services, in return for which they will export other goods or services. The fact that the production of many financial services, wholesale financial services in particular, is characterized by economies of scale and scope implies that their production will be concentrated in certain countries on efficiency grounds. Capital account liberalization can also promote the dynamic efficiency of the financial sector. Increased international competition in the provision of financial services can force domestic producers to become more efficient, stimulate innovation, and improve productivity.
A final argument for capital account liberalization, emphasized by Quirk and Evans (1995) and Cooper (1998), is that, with the growing difficulties of enforcement, policies designed to limit international capital flows will have to be increasingly invasive and distortionary in a world of highly developed capital markets.
Asymmetric Information
The classic case for unfettered capital markets is predicated on the assumption that they deliver an efficient allocation of resources. A large literature on “efficient markets” draws out the implications of this assumption. In contrast, those who emphasize the pervasiveness of information asymmetries dispute that this assumption is appropriate. Under the alternative assumption that information is asymmetric, inefficiencies can arise.25 These are the consequence of three problems: adverse selection, moral hazard, and herding behavior.
Adverse selection can occur under asymmetric information because lenders have incomplete knowledge of borrower quality and borrowers who are bad credit risks have a strong incentive to seek out loans. When incomplete information prevents lenders from being able to evaluate credit quality, they will only be willing to pay a price for a security (or lend at an interest rate) that reflects the average quality of firms issuing securities (or applying for loans), where that price is likely to be less than the fair market value for high-quality firms but above fair market value for low-quality firms. Because the owners and managers of high-quality firms realize that their securities are undervalued (equivalently, their borrowing costs are excessive), they will not wish to seek financing on the market. The only firms that will wish to sell securities will be low-quality ones because they know that the price of their securities is greater than their value. Because high-quality firms will issue few securities, many projects with a positive net present value will not be undertaken, while other projects whose net present value is lower than the opportunity cost of funds will in fact be financed. Under these circumstances, a liberalized capital market will not deliver an efficient allocation of resources.26
Moral hazard can occur under asymmetric information because borrowers can alter their behavior after the transaction has taken place in ways that the lender regards as undesirable.27 Borrowers will wish to invest in relatively risky projects in which they do well if the project succeeds but the lender bears most of the loss if the project fails; lenders, in contrast, will wish to limit the riskiness of the project. Hence, borrowers will attempt to alter their projects in ways that increase their risk after the financial transaction has taken place, and information asymmetries will facilitate their efforts to do so. Under these circumstances, many of the investment projects actually undertaken will be excessively risky. Lenders, as a result, will be reluctant to make loans, and levels of intermediation and investment will be suboptimal.
Finally, in financial markets characterized by incomplete information, lenders may be prone to engage in herding behavior, whereby they try to follow the lead of those whom they believe to be better informed. Such behavior gives rise to sudden market movements and volatility. Herding can be rational in the presence of information cascades, when agents optimally infer information from the actions of other agents and therefore act alike; in the presence of payoff externalities, when the payoff to an agent adopting an action increases with the number of other agents adopting the same action; and in the presence of principal-agent problems, under which investors managing money for others may have an incentive to “hide in the herd” in order not to be easily evaluated.28 Herding can arise when incompletely informed investors infer that a security is of lower (or higher) quality than previously thought from the decisions of other, presumably better informed, investors to sell (or buy) it. Such behavior can work to amplify price movements and precipitate sudden crises. Herding can also arise when investors are incompletely informed about the quality of those who manage their funds. Low-quality money managers may then find it rational to emulate the investment decisions of other managers in order not to be found out. Again, this could amplify asset price volatility. Finally, herding can be rational in the presence of payoff externalities, when the payoff to an agent adopting an action increases in the number of other agents adopting the same action. Obstfeld (1996) presents a model in which individual currency traders are too small to exhaust the central bank’s reserves and force devaluation of the currency, but simultaneous sales of that currency by several traders can have that effect. Krugman (1996) shows how this payoff externality can result in herding behavior.29
Beyond these theoretical analyses of herding behavior, a good deal of empirical research, in both domestic and international markets, suggests that asset prices fluctuate with more volatility than can reasonably be explained by economic fundamentals (see, for example, Shiller 1990 and 1993). Although these empirical results have been disputed, there is enough substance to be concerned about the issue even if herding behavior and other asymmetric information problems are not the full explanation. Moreover, economic fundamentals clearly do fluctuate and sometimes by large amounts. Experience as well as theory suggest that when these fluctuations coincide with weak financial systems, damaging economic and financial crises can easily be the result.
When adverse selection or moral hazard deters desirable transactions that would otherwise have occurred had information asymmetries been less severe, the result is economically suboptimal but relatively unlikely to be associated with important systemic risks. In other circumstances, however, adverse selection or, especially, moral hazard can potentially be associated with problems of systemic dimension. The most important problem in this area is almost surely the moral hazard created by extensive explicit and implicit government guarantees for financial institutions (and sometimes other types of enterprises), without adequate safeguards against imprudent risk taking by such institutions or adequate incentives for market discipline to effectively police excessive risk taking.
Herding behavior and the associated distortions to market dynamics are probably to some extent an intrinsic feature of most active financial markets, domestic and international. The key questions are how to recognize when these problems may become serious and how to contain or counteract them in these situations. Theory (and experience) provides no complete answers. In principle, better information about the buildup of an excessive concentration of debt and other sources of systemic risk should contain herding enthusiasms, perhaps partly by provoking earlier and less damaging corrections. Macroeconomic policies also have a role to play in resisting the buildup of systemically important imbalances in financial markets, as well as in countervailing the damaging effects of financial crises as the herd stampedes to the downside.
Domestic Distortions
Even when information is complete, international financial liberalization can be welfare reducing in the presence of domestic distortions. This result is a straightforward application of the theory of the second best.30
A classic illustration is Brecher and Diaz-Alejandro (1977), who consider the effects of capital flows in the presence of trade distortions. They set out a model of a small, open, relatively labor abundant country that protects its capital-intensive industries; an example is a developing country better endowed with labor than capital that protects or subsidizes its motor vehicle and aircraft industries. Because protection of these relatively capital intensive industries boosts the rate of return to capital invested in the country, capital will flow in, leading capital-intensive sectors to expand and labor-intensive sectors to contract. At world prices, the value of domestic production declines, because the misallocation of resources between the capital and labor-intensive sectors is magnified.31 Capital having flowed from higher-value uses abroad to lower-value uses in the capital-importing country, world welfare declines as well. Cooper concludes from this analysis that the “free movement of capital is likely to become allocatively efficient only after trade barriers have come down substantially, particularly barriers on capital-intensive activities in labor-rich countries” (Cooper 1998, p. 13).
McKinnon and Pill (1997) apply a variant of this model to the case where policymakers subsidize or guarantee foreign loans to the domestic banking system.32 Government guarantees against the possibility of domestic bank failures encourage excessive capital inflows. If the domestic economy is relatively well endowed with labor, but the financial sector is relatively capital intensive, capital inflows will have the same income- and welfare-reducing effects described above.33
Summary and Implications
This review suggests that the theoretical presumption in favor of the liberalization of domestic and international financial markets is weakened by the presence of asymmetric information and domestic distortions. Whether liberalization is welfare enhancing depends on the nature of those distortions, on the extent of information asymmetries, and on the severity of the adverse selection, moral hazard, and market inefficiencies that result. However, these problems are not a given, and public policy has the capacity to correct or ameliorate them.
For the domestic distortions discussed in the previous subsection that are fundamentally the result of misguided government policies, the solution is straightforward: correct the policies. In particular, if capital inflows threaten to reduce welfare because they will flow into heavily protected domestic industries where the true (social) return to capital is less than the cost of foreign borrowing, then eliminate or reduce protection to correct this distortion. After all, even without inflows of foreign capital, it makes no sense to invest large amounts of domestic capital in industries where the true rate of return is low or even negative. In some situations, government guarantees perform a desirable function. For example, modest amounts of sovereign borrowing to help finance investments in human capital, such as primary schooling and basic health care, can be justified as these are high rate of return activities for which, unfortunately, private market mechanisms work poorly. Financial guarantees for inefficient state enterprises that enable them to expand and waste more resources are a different matter. But such problems are not intrinsically problems of financial liberalization. Their cure is obvious and desirable regardless of the nature of the financial system; liberalization only makes the cure more urgent.
Some problems of asymmetric information can also be addressed reasonably directly through public policy, notably, policies that encourage adherence to world-class standards for accounting, auditing, and information disclosure; that facilitate enforcement of sound rules of corporate governance; and that protect investors and lenders from fraud and unfair practices (including through a credible judicial system and efficient bankruptcy procedures). Even when conditions are optimal, however, some problems of asymmetric information remain intrinsic to financial markets and institutions. When asymmetric information deters desirable transactions that would occur in an ideal world, financial liberalization is not able to produce all of the potential benefits relative to this ideal, but there is no threat of harm. When it encourages undesirable transactions that may be made easier in a liberalized system, the concern must be to limit the potential damage to an acceptably low level.
Theory alone provides little guidance on how to contain such dangers. It tends to send one chasing off after the phantoms of thousands of possible information asymmetries. Experience is a better teacher. While the continued development of financial markets, institutions, and instruments will probably turn up some new problems or new guises for old problems, experience points to the key issues that must be addressed to contain risks of serious misbehavior by liberalized financial systems. Strong, market-based incentives for prudent risk management by businesses and especially by financial institutions are critical. Because of their special character and place in the economic system, financial institutions must be subject to proper prudential supervision and regulation by an appropriate government agency. The government, usually the central bank, needs to be prepared to act as a lender of last resort to the financial system in the event of a systemic crisis. The government must contain the moral hazard associated with this activity by limiting such intervention to cases of systemic threat; by ensuring that financial institutions that make losses absorb them; and by imposing on shareholders, managers, and subordinated creditors the grief they deserve when an institution approaches insolvency. Macroeconomic policy needs to resist the buildup of serious financial imbalances and to counteract disorderly markets and limit economic contraction in a downturn. The public debt must be kept within reasonable bounds, and its maturity and currency structure must be prudently managed. Beyond the sovereign, excessive leverage, especially in the financial system but also in the business or household sectors, may be a cause for concern. Foreign currency debt may be a particular problem.
Notably, except for this last point, everything on the above agenda (including the points in the preceding two paragraphs) is as relevant for domestic financial liberalization as it is for capital account liberalization because the key potential problems with financial liberalization are germane to both its domestic and international dimensions. Yet, the presence of these problems has not led policymakers to suppress domestic financial liberalization and proclaim the virtues of highly repressive financial systems in their economies. Indeed, as discussed earlier, the clear trend has been toward domestic financial liberalization. The existence of information-related and other problems in domestic markets is clearly not perceived as sufficient justification to reverse this trend.
Although it does not lend itself to statistical analysis, this experience across a wide range of industrial and developing countries is probably the most important and powerful empirical evidence about the bottom-line questions of whether, on balance, financial liberalization generally and capital account liberalization specifically are beneficial. Countries that have successfully liberalized their financial systems domestically and internationally seek to preserve and develop these systems. Countries that have temporarily lost access to international capital markets actively seek to regain it. Countries that have never enjoyed such access seek to acquire it. Even countries that have had some bitter experiences with financial liberalization are not in general moving back to financial repression. Rather, they seek to learn what went wrong so that they can correct it and thereby enjoy more of the benefits of financial liberalization with fewer of the risks. Against this powerful evidence from experience, it remains to be seen what a more statistical analysis can supply.
See Goldsmith (1969), McKinnon (1973), and Shaw (1973). In theory, financial development can raise an economy’s growth rate in two ways: by increasing the rate of capital accumulation and by spurring technological innovation. The financial system can stimulate growth by facilitating the trading, hedging, diversifying, and pooling of risk; by enhancing the allocation of resources; by better enabling investors to monitor managers and exercise corporate control; by mobilizing savings; and by facilitating the exchange of goods and services (Levine, 1997). Obstfeld (1994a) provides a model of these relationships.
A synthetic treatment of these issues in the developing country context is Mishkin (1996).
A seminal theoretical treatment of this issue is Stiglitz and Weiss (1981). Whether additional liberalization will be beneficial, starting from an initial position where domestic or international financial transactions are restricted, depends on the precise nature and magnitude of the distortions involved. This is a question to which it is difficult to give general answers, although some of the most important issues are taken up below.
Moral hazard in financial markets can also result from sources other than asymmetric information; as discussed further below, an important potential source of moral hazard arises from the possibility that investor behavior will be altered by the extension of government guarantees that relieve investors of some of the consequences of risk taking.
Devenow and Welch (1996) summarize the literature on models of rational herding.
Calvo and Mendoza (1997) provide a model—closely related to the arguments above—of how capital account liberalization and the globalization of securities markets can increase the extent of herding. Briefly, their argument is that globalization, by increasing the menu of financial assets available to investors and promoting portfolio diversification, at the same time reduces the returns to investing in acquiring information on individual assets and thereby aggravates the problem of incomplete information that encourages herding. The same phenomena arise in Bacchetta and van Wincoop’s (1998) model of international capital flows in the presence of incomplete information and learning. But where Calvo and Mendoza argue that capital market liberalization, to the extent that it occurs simultaneously in many countries, undermines individual incentives to gather information and thereby permanently increases herding behavior, Bacchetta and van Wincoop argue that incomplete information is a transitional problem associated with recent liberalization (namely, that international investors will have the least information about recently liberalized markets), so that the problem should be overcome as learning takes place over time.
More generally, the applicability of the theory of the second best in the context of capital controls and capital account liberalization is discussed in Dooley (1996a).
Because labor and capital are attracted to the relatively low value added (at world prices) import-competing activity.
See also Mathieson and Rojas-Suarez (1993), Dooley (1996b), and Krugman (1998).
A final possible undesired effect of capital account liberalization is its distributional consequences. Theory suggests that increased international capital mobility raises the effective price elasticity of the supply of capital to any single national economy, making it more difficult to tax capital and shifting the incidence of taxation toward labor services (Rodrik, 1997). This effect is often referred to as the tendency for capital mobility to increase international tax competition, especially in capital taxation. National authorities may regard these distributional consequences as undesirable. The optimal response to this problem is not capital account restrictions, however, but lump-sum redistribution.