سبد سرمایه گذاری مبتنی بر تئوری استقراض خارجی بیش از حد و کنترل سرمایه در یک اقتصاد کوچک باز

We develop a simple model of portfolio choice in a mean variance framework to address the issue of international borrowing and financial crisis. Instead of adverse selection or moral hazard of lending and borrowing activities we emphasise the role of exchange rate movement. Syndicated borrowing by way of internalising the aggregate effect tends to restrict excessive borrowing from external source. However, this may undermine the welfare consequences by further aggravating the extent of risk undertaken in the process. There is a built-in externality in the model that leads to over exposure to foreign currency debt and readily calls for intervention by the government. Government intervention by way of a tax on foreign borrowing may help restrain the amount of external debt and implement the first best.

Since the Asian crisis hit the world, economists, financial experts and policy makers have been busy reinvestigating the reasons and consequences of intervention in the financial markets. Most of such discussions take the form of exploring what went wrong in the fastest growing segment of the global economy. Generally the germ of the crisis has been traced to the classic problems of moral hazard and adverse selection (Krugman, 1998) associated with financial intermediation, etc. and/or to the careless attitude of macroeconomic policies as reflected in overvaluation of exchange rates, associated current account problems, composition of external debt, etc. (see, for example, Burnside et al., 2001 and Rakshit, 2002). Although the aftermath of the crisis has drawn attention of the researchers all around the world, the potential of such crises has been explored extensively in the earlier works of financial economists such as Diamond and Dybvig (1983).1 In fact regulatory mechanisms which prevent run on banks or Tobin's tax which penalises speculation in foreign exchange markets are pretty well known in the areas of banking, finance and international finance. Typically arguments in favour of unhindered free flow of goods and capital often undermine the role of intervention. In case of a closed economy regulations of financial markets, banks, capital markets, etc. may distort the environment within which trade and exchanges take place. Cause of intervention here has to be justified in terms of some pre-existing distortions which markets cannot fully internalise. In case of a small open economy standard text books do not prescribe restricting trade or capital flows. In trade theoretic terms such intervention does not yield any terms of trade advantage for a typical small economy. They only tax and distort consumption, making it more expensive. However, intervention does occupy a prime position in the theory of international trade and a host of interesting results on theory of tariffs, quotas, trade related subsidies are available for nations which can favourably alter their trading positions through restrictive regulatory policies. But a small country cannot do much in this regard. A lesson seems to have emerged from the recent discussions on the issue of financial crisis, bank failures and balance of payments problems. It seems that (mis)information has played havoc in Asia. The debt structure of commercial banks was not properly monitored by the regulatory authorities. Mutual funds or those financing real estate boom kept the lenders in the dark. As a result the consequences of the crisis were largely liquidity induced—in the face of capital flight firms could not roll over short-term debt.2 While rushing to convert Asian assets into dollars, investors hardly looked at the sound macroeconomic indicators of the economies in trouble—a reflection of what is known as “herd-behaviour” (Banerjee, 1992). In case of earlier financial crises (for example in Mexico in 1987), economists have blamed over valuation of exchange rate in a pegged exchange rate regime (Krugman, 1979 and Rakshit, 2002). In his study, Rakshit (2002) has, however, argued that informational problems were found to be more important in generating the financial crisis in Asia. Once started in Thailand it spread through contagion effect. In some other studies economists have also expressed similar views (see, for example, Allen and Gale, 1998, Allen and Gale, 2000 and Allen and Gale, 2002). Informational problems being at the epicentre of such crises, the necessity of intervention evolves around smoothening the process of information flows. If clients know about the “capital adequacy” index of a bank or a financial intermediary or its true ‘net worth’ well ahead in time, they should take precautionary action and jumps are not likely to take place. As Krugman (1998) puts it, the limited liability clause coupled with a belief that in time of a crisis the government will step in and save the ailing intermediary leads to too much investment in risky projects. To the extent the regulatory authorities are able to enforce ‘transparent’ behaviour, the system has lesser of a chance to be hit by a crisis. Without informational problems such as ‘noise’ of any kind, markets should take care of the anomalies and except idiosyncratic jumps there should not be much of a problem. The issue of intervention loses all its charm in that context. It is ironic that a long-standing hypothesis regarding intervention in the face of a market failure has been accepted by the economists and the basis of such action existed independent of informational complexities. A substantial segment of the literature on trade theory has been devoted towards justifying government action in a competitive economy which trades with the rest of the world. For example, a bumper coffee production in Africa or Latin America has always results into a crash in coffee prices, because of an excess supply in coffee export. Exporters may export ‘too much’ simply because each is too small to internalize the damaging effect of an export boom on the terms of trade. The unique feature of international trade is that it naturally identifies the domain of intervention. Policies that maximize national welfare are often those which restrict global trade and do not allow the universally first-best to be achieved. In this paper we analyse a case where intervention becomes necessary because atomistic agents do not internalise the impact of the aggregate outcome. We deal with the problem of external borrowing by local financial intermediaries. Our idea is related to the literature on exchange rate collapse, bank failures and macroeconomic recession. A recent paper by Burnside et al. (2001) argues that interest guarantees provided by the government adversely affect the incentives of the banks or intermediaries to hedge exchange risks. Hence, the financial system becomes more fragile and cannot cope with a potential crisis. Guarantees make the banks shy away from the forward markets. This possibly explains the lack of forward trading in the emerging markets. The lack of forward trading may also be a consequence of being able to cover the resulting interest rate risk implied from forward contracts, because incomplete markets create inefficiencies and so a high price. Although one must note that in some of the markets in the developing countries explicit regulations prevent extensive forward trading from taking place. For example in India forward transactions are restricted to a large extent and even if they are not, imperfect markets may not allow full hedging of the risks. However, the interesting point is that there may not be much of a demand for forward transactions if interest guarantees are provided by the government or because of other inefficiencies due to incomplete markets. We do not highlight the role of imperfect information or contagion effect in this paper. Instead we approach the problem of intervention from a different angle with risk-averse banks and a perfect forward market with exchange rate movements responding to capital inflows and outflows. Our basic argument is borrowed from trade theory and runs as follows. Banks allocate their borrowing between external and internal sources according to the principle of risk-aversion in the face of random movements in exchange rate. Typically individual bank does not internalise the exchange rate implication of total capital inflow or outflow, i.e. its impact on the deterministic component of the exchange rate. This tends to distort the portfolio choice. In particular while choosing the optimum proportion of external borrowing in the total portfolio, banks tend to undervalue the current exchange rate and overvalue the future exchange rate. The consequence is that foreign currency borrowing remains unhedged since the repayment will be made with overvalued local currency. It has often been argued that a major problem in pre-crisis Asia was overvalued local currency. This in turn leads to ‘excessive’ external borrowing relative to the “first-best” level. The argument is similar in nature with the one which suggests that competitive exporting units will over export because they will not be considering the aggregate movement in the terms-of-trade. This is well illustrated in the standard textbooks of Caves et al. (1993) and Helpman and Krugman (1989). Drawing on the experience of South Korea, Noland (2005) has argued that in its early part of development it has experienced rapid sustained growth in the presence of control of international capital and delinking of its domestic and international financial markets. As a matter of fact Noland (2005) has questioned South Korea's capital account liberalization programme so far as it affected the timing, magnitude and particulars of 1997 crisis. Kapur (2005) has also expressed similar view citing the examples of Hong Kong and Singapore. It has been argued that the policy of ‘non-internalization’ of Singapore dollar and its exchange market policy have helped Singapore to emerge relatively unscathed from 1997 crisis. Bordo and Meissner (2005) has compared the role of foreign currency debt in propagating the major financial crises during the period 1880–1913 and more recent episodes during 1972–1997. They have concluded that debt in foreign currency may not necessarily lead to a crisis, however, high it might be. This is true both in the late 19th or early 20th century as well as in more recent time. What matters most in propagating a debt crisis, currency crisis or a banking crisis is a mis-managed foreign currency debt. This empirical evidence calls for an intervention by the government as private agents (banks or large corporations) cannot do it on their own as in any other case of externalities. Byström et al. (2005) has shown that finance and securities firms suffered more than the manufacturing firms in Thailand during the 1997 crisis. It has also been found that the smaller firms were more severely affected than the larger firms. Typically a small open economy faces an exogenously given world interest rate. The effective interest burden of the debt contains an exchange rate component which can fluctuate and therefore the effective interest rate is likely to differ from the nominal rate in the rest of the world. Individual borrowers would hardly internalise the exchange rate movements resulting from aggregate borrowing. When funds flow in, exchange rate tends to appreciate which makes every dollar borrowed fetch greater amount of local currency. When funds flow out, exchange rate depreciates effectively increasing the interest burden of the debt. Both of these should reduce the amount of external borrowing relative to the one where the exchange rate is assumed not to respond to individual borrowing. We develop a simple model of portfolio choice with risk aversion to highlight this basic point. To drive our point home we assume no uncertainty regarding the lending or the borrowing rates and the entire random activity is associated with a part of the exchange rate movement. Under such an adjustment mechanism syndication of borrowing activity tends to restrict excessive borrowing from an external source. However, such a resolution may undermine the welfare consequences by further aggravating the extent of risk undertaken in the process. Regulatory intervention in this case may help to restrain the amount of external debt and implement the first best at the same time. The paper is laid out as follows. The Section 2 describes the model and the results and the last section concludes.