Cambridge Journal of Economics 2008, 32, 665681 doi:10.

1093/cje/ben028 Advance Access publication 27 June, 2008

Is the accumulation of international reserves good for development?

Moritz Cruz and Bernard Walters*
International reserves accumulation has been the preferred policy recently adopted by developing economies to achieve nancial stability. The aim of this policy is to increase liquidity and thus reduce the risk of suffering a speculative attack. The main concern expressed in the literature has been related to its cost. Most of the studies conclude that the opportunity cost of international reserves accumulation is around 1% of GDP. However, these studies have not analysed whether this strategy is, or could be, more broadly supportive of development, an issue that must be of central interest for developing economies. The aim of this paper is to show that the stockpiling of international reserves is not optimal for developmental purposes and that there exist alternative policies that can be applied to achieve nancial stability, policy autonomy and a better performance in terms of development. Key words: International reserves, Financial stability, Speculative attacks, Capital management JEL classications: O11, O16, E12, E63, F41

1. Introduction
The factor identied by most commentators as triggering recent nancial crises across the developing world has been illiquidity in the wake of an interruption and/or reversal of capital inows (see Allen, 2004, Calvo, 2001, Feldstein, 1999, 2002). The sequence has been of speculative attacks inexorably exhausting international reserves, forcing countries to increase their domestic interest rate and eventually to oat their foreign exchange rates. The economic policy response to forestall this cycle of speculative attackcapital ight nancial crisis, adopted by both crisis-affected and other non-affected emerging economies since Mexicos 199495 nancial crisis, but particularly since the 199798 East Asian crisis, has been simply to increase liquidity through the accumulation of international reserves. In 2001, the Report of the High-Level Panel on Financing for Development to the United Nations stressed that since the Asian crisis international reserves in emerging
Manuscript received 13 October 2006; nal version received 6 May 2008. Address for correspondence: Moritz Cruz, Universidad Nacional Autonoma de Mexico (UNAM), Instituto de Investigaciones Economicas, Circuito Mario de la Cueva s/n, CU 04510, Mexico; email: aleph3_98@ yahoo.com * Universidad Nacional Autonoma de Mexico, Mexico and University of Manchester, UK, respectively. The authors are very grateful to two referees for their comments, which have improved the paper substantially. The usual disclaimers apply.
The Author 2008. Published by Oxford University Press on behalf of the Cambridge Political Economy Society. All rights reserved.

666

M. Cruz and B. Walters

economies had increased by around 60%. Furthermore, developing countries international reserves have risen from 68 percent of GDP during the 1970s and 1980s to almost 30 percent of GDP by 2004 (Rodrik, 2006, p. 255). Earlier concerns about the issue of reserve accumulation focused on their optimum size. More recent concerns have addressed the cost of excess reserves in terms of the difference between the market return and the much lower return from holding reserves in (typically) US treasuries. On this basis, some studies suggest that the excess of international reserves has a cost of around 1% of gross domestic product (GDP) (Bird and Rajan, 2003; Rodrik, 2000, 2006). This is, as Rodrik (2006, p. 262) states, a large number by any standard. It is a multiple of the budgetary cost of even the most aggressive anti-poverty programs implemented in developing countries. Other studies, moreover, have identied potential costs in terms of detrimental effects on the nancial sector of either moral hazard problems, created by very liquid banking systems channelling excess credit to overheated asset markets and/or domestic macroeconomic risks such as ination, high intervention costs and monetary imbalances (Garca and Soto, 2004; Mohanty and Turner, 2006; Schiller, 2007). The dominant explanation for the build-up of reserves, and therefore for accepting their associated cost, is that it represents precautionary behaviour designed to provide insurance against the far higher output and other costs associated with earlier crises (see Cerra and Sexena, 2008). Countries following this strategy are evidently hoping to emulate those economies that escaped speculative attacks and/or capital ight (e.g. Chile, Columbia, China and India) and maintained policy autonomy, thereby providing a way of reducing the risks of future crises and of minimising the need to turn to the IMF [International Monetary Fund] if crises occurred (Bird and Mandilaras, 2005, p. 85). Another alternative but not exclusive, explanation for reserves accumulation identies it with an active policy of export-led industrialisation characteristic of several countries in East Asia (especially China). This is the mercantilist interpretation of reserve accumulation, in which increased reserves are a by-product of maintaining a competitive exchange rate designed to expand tradable production. In summary, as well as insurance against the opportunity costs associated with crises, there are benets claimed for the policy of international reserves accumulation in terms of the policy autonomy necessary to pursue independent industrialisation strategies. However, this justication is not entirely compelling. A large literature demonstrates that it was not international reserves accumulation per se that deterred speculative attacks and provided policy autonomy for the countries that avoided the earlier crises; these aims were attained through the implementation of capital controls (see Agosin and FfrenchDavis, 1996; Athukorala, 2003; De Gregorio et al., 2000; Doraisami, 2004; Edwards, 1999, 2003; Joshi, 2003; Ocampo, 2002; Palma, 2002). Furthermore, the export (and growth) record of the more successful of these economies (particularly their export-led growth success) has not been primarily the result of a short term competitiveness strategy based on the maintenance of a competitive real exchange rate (RER) through the accumulation of international reserves. On the contrary, it has been the result of a long term strategy that has included supply side policies such as the building of technological capabilities. Thus, as Aizenman (2006, p. 2) observes, the mercantilist case for hoarding international reserves, as an ingredient of an export led growth strategy, is lacking empirical evidence. In addition, for most countries the adoption of a reserve accumulation strategy was taken in the context of the decision to adopt or reinforce the neo-liberal strategy of rapid

Effect of accumulation of international reserves on development

667

nancial liberalisation, unrelated to the development of either deep nancial markets or mature and effective regulatory structures. The evidence suggests that this, in fact, leads to a reduction in policy autonomy, a high degree of nancial instability and very little (if any) contribution towards development (see Chang and Grabel, 2004; Grabel, 1995; Kose et al., 2006). Governments are evidently in a position where they see little alternative to the world of increased nancial liberalisation but have no faith in the insurance provided by the IMF. In this context, the policy of reserve accumulation is seen as providing the only option to navigate between the Scylla of a nancial crisis and the Charybdis of an IMF package. The justication for the policy of reserve accumulation is, therefore, bounded by the acceptance of the liberalised nancial architecture with a narrow interpretation of the opportunity costs and benets available from its pursuit. A full evaluation should address at least two further questions. First, whether this strategy, through the policy autonomy it provides by preventing nancial crises and the consequent involvement of the IMF, contributes to development in the sense of aiding the process of industrialisation. Strikingly, the studies that have addressed the question of reserves accumulation have not analysed this question. Second, whether there are alternative policy options able to provide nancial stability at a lower opportunity cost than stockpiling international reserves. The aim of this paper is to argue precisely that the policy of amassing reserves is misguided and developing countries should (re)consider alternative policies that are feasible and effective, which might provide nancial stability while maintaining the policy autonomy necessary to pursue growth and industrialisation goals. The paper is set out as follows. Section 2 explains why developing countries accumulate international reserves. Section 3 considers whether this strategy is worthwhile and argues that at least three important issues need to be considered before embarking on this course of action: the tradeoffs involved in the stockpiling of international reserves; the factors that predispose an economy to nancial crisis; and nally, whether there are policies better suited to the promotion of nancial stability and policy autonomy that also allow sufcient space to promote industrialisation. Section 4 argues, based on historical evidence, that such alternatives exist and have been successfully used in the past to achieve and maintain nancial stability, which, in turn, has allowed the pursuit of active policies supporting industrial development, with fewer costs associated with their implementation. Finally, Section 5 presents the concluding remarks.

2. What is driving developing countries international reserve accumulation? Precautionary, policy autonomy and mercantilist motives
The issue of the accumulation of international reserves has regained relevance and interest1 (see Aizenman, et al., 2004; Aizenman and Marion, 2002; Bird and Rajan, 2003; Mendoza, 2004; Wijnholds and Kapteyn, 2001) since the mid 1990s because of the boom in nancial crises in the developing world, notably in Mexico, Thailand, Korea, Malaysia, Philippines, Indonesia, Brazil, Turkey, Russia and Argentina. The majority of these crises were associated with the adoption of the neo-liberal nancial liberalisation strategy of free mobility of capital (for further discussion see Allen, 2004; Arestis and
1 In an earlier period, during the 1960s, as a consequence of the plans to provide the international nancial system with greater liquidity, the debate focused on dening the optimal level of international reserves necessary to maintain the value of a currency within the xed exchange rate system. A decade later, during the 1970s, when most countries adopted freely oating foreign exchange rate regimes, international reserves were seen as a buffer to absorb a transitory current account shock (see Edwards, 1983; Garca and Soto, 2004).

668

M. Cruz and B. Walters

Glickman, 2002; Cruz et al., 2006; Grabel, 1996; Palma, 2003; Singh, 2003), which increased the vulnerability of the capital account of the balance-of-payments, a source ignored during the earlier era of capital controls. The observed increased vulnerability in the modern era of nancial liberalisation has meant, rst and foremost, that stockpiling international reserves has been seen as the central policy option that a country can pursue to avoid a nancial crisis and its high economic costs (Bird and Rajan, 2003). This choice is reected in the evolution of world international reserves. Data from the World Development Indicators indicate that world wide accumulation of international reserves registered a marked break in 199495, in the aftermath of Mexicos nancial crisis. From that year, international reserves sharply increased, with an increase of 19.4% in 1995. Interestingly, if Organisation for Economic Cooperation and Development (OECD) countries1 are eliminated from the sample, given that their liberalisation occurred (long) after the development of mature economic and nancial structures, the pattern described becomes more marked, with an increase in foreign exchange holdings in 1995 of 22.1%. In fact, around two thirds of international reserves are currently held by developing countries (Aizenman, 2007). Thus, as Rodrik (2006) stresses, developing economies are deeply involved in international reserves accumulation. The evidence therefore suggests that the accumulation of international reserves was seen initially as a source of protection or insurance, and more recently as a permanent buffer stock against the total or overall vulnerability of the balance-of-payments, arising from both the capital and the current accounts. Of course, in the current era of free mobility of capital, much more attention has been placed on the vulnerability arising from the capital account. A developing countrys reserves is related to changes not in real quantities (such as imports or output) but in nancial magnitudes (Rodrik, 2006, p. 257, emphasis in original). The results reported by Mendoza (2004), for a sample of 65 developing economies, are consistent with the hypothesis that, since the boom in nancial crises, the strategy of reserve accumulation has been linked to the pursuit of nancial stability. In addition, the econometric work of Aizenman and Lee (2005) supports the view that developing countries mainly demand and accumulate foreign exchange for precautionary reasons. Notwithstanding the nancial stability that international reserves might provide, it is also clear that developing countries consider the increase in international reserves holdings as having other, non-precautionary purposes. These can be categorised as the policy autonomy and mercantilist motives for holding reserves. The policy autonomy argument is straightforward: avoiding nancial crises avoids the interference of and dependence on international agencies. The eruption of nancial (but also debt and political) crises has led to the involvement of international multilateral institutions, mainly the IMF (but also the World Bank), and their conditional assistance packages, with an inevitable loss of policy autonomy. Countries judge that by preventing a crisis through the accumulation of liquidity, they will also minimise IMF conditional assistance and gain policy autonomy, even in the event of a crisis actually occurring (see Bird and Mandilaras, 2005). Therefore, developing countries see in a strategy of reserve accumulation a source of policy independence or sovereignty. An alternative but not exclusive motive is provided by the mercantilist approach. This centres on the maintenance of a competitive RER as a form of active industrial strategy.
1

Except, for obvious reasons, Mexico, Korea and Turkey.

Effect of accumulation of international reserves on development

669

The aim is to achieve and then maintain a RER that is sufcient to promote a high export growth rate (see Aizenman 2006; Aizenman and Lee, 2005); the accumulation of international reserves is an inevitable counterpart to such a strategy. The mechanism is explained in a recent study of Frenkel and Ros (2006). They argue that if the rate of accumulation in the tradable goods sector is a positive function of protability, and protability in that sector is a positive function of the RER, then a competitive RER will lead to faster growth of the traded goods sector. And they further note that a competitive RER operates as an industrial policy designed to distort relative prices in favour of tradable goods activities (like any industrial policy, such as export promotion industrialisation). In this sense, a more depreciated currency is equivalent to a uniform tariff on imports, with the advantage that a depreciated currency does not distort relative prices against exports because it simultaneously implies a subsidy (an income transfer) of the same size (p. 636). Thus, by maintaining a competitive RER with the concomitant reserve accumulation, the protability of the tradable sector can be promoted and, in turn, rms will invest and expand production. It is important to mention, nally, that additional benets may attach to the accumulation of international reserves. For example, large stocks of reserves may reduce the volatility of the foreign exchange rate (see Aizenman, 2006; Aizenman and RieraCritchton, 2006; Cady and Gonzalez-Garcia, 2006). In addition, the cost of foreign borrowing may also be reduced. In summary, the precautionary, mercantilist and policy sovereignty motives are the driving forces behind developing countries hoarding of international reserves. The benets claimed for this strategy, which can be classied with respect to development purposes, policy autonomy and nancial stability, seem, at least at rst glance, very attractive. Furthermore, there appears to be a presumption that the costs associated with the policy are small compared with the cost of being less liquid. Accordingly, this policy is seen as the best means to achieve nancial stability, provide policy autonomy and aid industrialisation. However, this impression is deceptive. First, there are large direct and potential opportunity costs inherent to the adoption of this policy. Second, some of the benets enumerated may not be achieved, or even be achievable, in practice. Third, there are policy alternatives available to pursue these targets. When the proposed benets are set against a fuller analysis of all the opportunity costs in terms of development purposes, policy autonomy and nancial stability the policy of reserve accumulation, particularly in the long term, loses much of its attractiveness. The next section presents a more complete analysis of the wider costs associated with this strategy.

3. The drawbacks of pursuing a policy of international reserves accumulation 3.1 Direct and potential costs
It is clear that following the strategy of reserves accumulation comes with considerable direct costs. Holding reserves incurs an opportunity cost, which is the difference between what the reserves could have earned and what they actually earn (Ramachandran, 2004, p. 365). This opportunity cost has been estimated for the Report of the High-Level Panel on Financing for Development to the United Nations in 2001 to be of the order of 8%, which represents the differential between the yield on the reserves and the marginal cost of borrowing.

670

M. Cruz and B. Walters

In addition to establishing the differential, estimating the cost also requires identifying the size of the excess holdings relative to some optimum level. So far, international reserves accumulation seems unrelated to any clear notion of what might constitute an optimal level of reserves. The so-called GuidottiGreenspan rule,1 which proposes the maintenance of reserves equivalent to 12 months of a countrys total foreign obligation, which includes but is not limited to imports (Mendoza, 2004, p. 76) does not adequately explain the recent evolution of international reserves. This is unsurprising as this sort of criterion is based on a rule-of-thumb,2 rather than any formal theory.3 Nevertheless, as we have already stressed, there are clear precautionary reasons to hold reserves but quantifying optimum reserves is. . . not straightforward since it is difcult to estimate the adjustment costs and output losses that reserves may enable a country to avoid (Bird and Mandilaras, 2005, p. 86).4 In the absence of a robust denition of the optimum level of reserves a number of ad hoc ratios have been used. Rodrik (2000) and Bird and Rajan (2003), for example, have estimated that the excess of reserves holdings above the level dened by the conventional ratio R(reserves)/M(imports), to be around 1% of GDP, a result conrmed by Rodrik (2006). Cruz and Walters (2007), applying the notion of a maximum sustainable external threshold,5 estimate this cost at around 0.19% of GDP for the Mexican economy during the 19962004 period. Mendoza (2004), on the other hand, estimates that the quarterly cost of reserves accumulation, for a sample of 65 developing economies during the 19982001 period, was approximately US$450 million.6 In summary, it is clear that, based even on a narrow criteria dening excess and cost, as Rodrik (2006, p. 261) observes, developing countries are paying a very high price to play by the rules of nancial globalization. In addition to these considerable direct costs, it is important to note that reserve accumulation can actually be counterproductive, potentially generating further long term problems. For example, large reserves stocks may create moral hazard problems that could weaken the nancial system of a country. This, in turn, could make crises to be deeper. . .
1 Proposed initially by Pablo Guidotti (then deputy nance minister of Argentina) and then rened by former US Federal Reserve Chairman Alan Greenspan in 1999. 2 The same can be said about the ratio of international reserves to imports (R/M), which suggests that an adequate level of reserves can be established as that level of reserves which are able to cover at least 3 or 4 months of imports. More importantly, however, the criterion suggested by the ratio R/M became inappropriate when most economies moved to a freely oating exchange rate regime and/or when they could borrow foreign currency in the international markets. In a freely oating system, the need to keep reserves is reduced considerably (even for fully open economies) because, in theory at least, the external imbalances can be corrected through adjustments of the exchange rate or, in the last resort, through the ability to borrow from the international markets. 3 Despite the lack of a theory for reserve adequacy, there are studies that have applied econometric techniques to try to determine it. These include, among others, Ben-Bassat and Gottlieb (1992) for a selected group of 13 economies, Ramachandran (2004) for the case of India and Garca and Soto (2004) for Chile and other Asian economies. 4 In the absence of consensus about the optimal level of reserves, the data is best characterised as following what has come to be called Mrs Machlups wardrobe theory (Machlup, 1966). According to Bird and Rajan (2003, p. 877), this suggests that the acquisitive characteristics of monetary authorities in terms of adding to their reserves resembled those of Mr. Machlups wife in terms of clothes. According to this idea no level of reserves was ever enough. 5 The central idea of this threshold comes from the fact that during the boom of crises in the 1990s, both the current account decit and the short term external debt, each expressed as a fraction of GDP, reached levels beyond which the historical record indicates nancial markets start to get nervous and, on that basis, decide to withdraw their capital out of a country. In this sense, the authors suggest that a level of reserves that could maintain calm amongst nancial investors would be of an order of around 56% of GDP. Any level of international reserves in terms of GDP above that threshold can be considered an excess and this, in turn, allows an estimate of its cost. 6 A cost that, as Mendoza himself stresses (p. 73), needs to be taken in the proper context. For example, for a large country like Brazil or Malaysia it may be innitesimal, but large for a small country like Uganda.

Effect of accumulation of international reserves on development

671

(Garca and Soto, 2004, pp. 1718). This is because currency intervention injects liquidity into domestic money markets producing very liquid market systems that can spill over into overheated assets markets and perhaps distort the banking system (see Schiller, 2007). In addition, according to Mohanty and Turner (2006, p. 40), large and prolonged reserve accumulation aimed at resisting or delaying currency appreciation can create a range of domestic macroeconomic risks through its effects on the balance sheets of the central bank and the private sector. These risks include near-term ination, high intervention costs and monetary imbalances. In their study, the authors argue that these potential risks have been controlled so far only because many countries accumulating reserves over the past few years have faced conditions of substantial excess capacity and low ination, which has meant that policy could be eased in the face of upward pressure on the currency. In these circumstances, reserve accumulation did not create the dilemma for policymakers of choosing between their ination and their exchange rate objectives.

3.2 Does international reserves accumulation promote development?

The logic of accumulating large amounts of any resource, like reserves, when a country is most in need of them for alternative productive projects, is questionable.1 This is particularly true when the economy has suffered a deep economic slowdown as a result of a crisis and thus the need for rapid and sustained growth is more urgent than ever. There is evidence that crisis-affected economies suffer not only large and persistent output losses, but they never recover from such negative shocks in the sense of output losses being reversed (Cerra and Sexena, 2008). Thus, the broadest interpretation of the opportunity cost of holding excess reserves should allow the cost of reserves to be considered in terms of foregone developmental projects. Although the approach is contentious,2 in a recent study, Cruz (2006) estimated that, had the excess of international reserves been absorbed by the Mexican economy during the 19962004 period (either through investments in infrastructure or in sectors with a high rate of return), the upper bound of Mexicos growth would have been 4.3% per year instead of the 3.7% observed. In addition to the various costs enumerated above, the empirical signicance of using international reserves to promote growth (the mercantilist view) is contested. The econometric analysis of Polterovich and Popov (2002) suggests that countries with growing foreign exchange reserve ratios to GDP, all other things being equal, exhibit higher investment/GDP ratios, higher trade/GDP ratios, higher capital productivity and higher rates of growth. However, the results of Aizenman and Lee (2005) indicate that although the variables associated with the mercantilist motive are statistically signicant their economic importance in accounting for reserves hoarding is close to zero and is dwarfed by other variables. In a similar vein, Aizenman (2006) concludes that the management of international reserves should not be seen as a panacea, particularly for an export-led growth strategy. Furthermore, as previously highlighted, large and prolonged
1 In this respect, for example, the Bank of England highlights (p. 10) that even a positive return may not be optimal; the key question is whether higher returns, after allowance for risk, could be made elsewhere (e.g. through investment in the countrys domestic infrastructure). See Handbooks in Central Banking No. 19 of the Bank of England, available at: http://www.bankofengland.co.uk/education/ccbs/handbooks/ccbshb19. htm#top 2 According to Rodrik (2006, p. 8) the process of accumulating international reserves. . . makes clear that the relevant counterfactual in most instances is not one dollar of additional public investment, but one less dollar of short-term foreign debt. Also, so far, there is no evidence that international reserves have been used to nance infrastructure projects, especially when the aims of reserves accumulation are liquidity and protection (see Singh, 2006).

672

M. Cruz and B. Walters

reserve accumulation aimed at delaying currency appreciation can create a range of domestic macroeconomic risks. It is also important to emphasise that, if it is to promote growth and development effectively, a policy of maintaining a competitive RER through reserves accumulation needs to be coupled with appropriate long-run supply side policies, while a competitive RER can only be a short term policy. Japan, Korea, Singapore and China [according to Polterovich and Popov (2002), examples of successful development strategies using growing foreign exchange reserves], are manufacturing export leaders. However, this is an outcome that has involved much more than a non-appreciated currency strategy. They have, in fact, long since developed an export structure that allows them to enjoy the benets of trade. Their superb export performance is not something recent that has occurred since the boom of nancial crises. It is rather the story of East Asian economies during the last 50 years. The corollary is that adopting a large reserves holdings strategy will not, on its own, deliver the benets of export-led growth from a developmental point of view.1 Thus, it is not appropriate to embark on reserves accumulation and its effects on the foreign exchange rate and export promotion, especially in the long term, without other complementary policies aimed at creating the supply side structures that are needed for (high value added) export growth.

3.3 Do international reserves guarantee nancial stability and policy autonomy?

Even though the balance of the discussion so far might suggest that prevailing patterns of reserve holdings are far from crazy in view of the signicant costs of being less liquid (Rodrik, 2006, p. 10), at least two further questions should be considered before deciding to embark on a strategy of stockpiling reserve holdings. In the rst place, the factors that open the door to nancial crises, or, in other words, to a rapid drain of liquidity, should be considered. The empirical evidence indicates that there are two main routes by which an economy can suffer a nancial crisis. Directly, by opening the capital account of the balance-of-payments and simultaneously deregulating domestically, i.e., by implementing the neo-liberal strategy of nancial liberalisation. And indirectly, via contagion, which is, however, to a considerable extent an inevitable consequence of the increasingly tight commercial and nancial links between countries.2 The implication, therefore, is that only if policymakers are planning to implement (or reinforce) the neo-liberal nancial integration strategy3 will it be worth considering the strategy of international reserves accumulation.
1 Mexico, for example, has increased reserves by around 300% since the 199495 peso crisis. However, manufactured exports have risen, on average, at a 12% rate during 19962005, which is much lower than the 20% rate achieved during the 198194 period. 2 Recall that the danger of the contagion risk, dened as the threat that a country will fall victim to nancial and macroeconomic instability originat[ing] elsewhere (Grabel, 2003, p. 320), will depend upon the vulnerability of the recipient emerging economy (i.e. its current levels of ight, currency and fragility risk) relative to the specic form and size of the shock (trade, foreign income or nancial, i.e. cost of nance, investor herd behaviour) as well as any responses by policymakers and investors (see Chui et al., 2004 and ECLAC, 1998). Additionally, the effect of the contagion risk will vary in duration and intensity according to the linkages that the country under consideration maintains with the country (or region) where the crisis originates. 3 Recall, however, that the neo-liberal strategy of nancial liberalisation is growth-distorting because it promotes the creation of new opportunities for risky investment practices and a corresponding misallocation of credit toward speculative activities, with destabilising macroeconomic effects (Grabel, 1995, p. 129). Furthermore, abandoning nancial repression may lead to an explosion of government debt, economic instability and lower economic growth (Fry, 1997, p. 768).

Effect of accumulation of international reserves on development

673

As we have stressed, however, a large body of literature suggests that, among other things, nancial liberalisation has preceded nancial crises with the associated huge costs and that it has reduced signicantly the space for and the autonomy to formulate policies in the pursuit of national development objectives. In this sense, neo-liberal nancial liberalisation is not supportive of industrialisation, while it sharply increases nancial instability.1 Furthermore, it is important to stress that even an emerging-market economy with an idealized external position would remain highly exposed to the punishing vicissitudes of liberalised nancial ows (DArista, 2000, p. 3). In other words, there is no guarantee for a poor economy that even holding large amounts of international reserves will completely eliminate the risks that emanate from hot money because it has proved impossible to identify either exactly what moves international liquidity holders to attack a currency or what determines their timing. International reserves holdings can be seen, therefore, as simply an ad hoc protection against speculative attacks. The second question that needs to be asked when considering a strategy of reserve accumulation is whether there are feasible policy alternatives that can allow countries the policy space to apply their own approaches to development, understood as the structural transformation of their economies in the direction of a modern, industrialised state. Fortunately, the historical evidence of both developed and newly industrialised economies shows that there are indeed several alternatives that have proven to be effective in these common aims, among which are restrictions on currency convertibility and the famous (Chilean) management of capital controls (see Grabel, 2003).2 These policies are addressed in the next section. Table 1 summarises the costs and benets of the strategy of international reserves accumulation with respect to development purposes, policy autonomy and nancial stability. As can be seen, the benets of this strategy are very attractive but, as we have argued, they can only be enjoyed on a short term basis: as soon as the neo-liberal strategy of nancial liberalisation is applied, nancial stability is subject to the vicissitudes of hot money and policy autonomy for development purposes disappears. Also, the competitive RER policy can only be sustained as long as monetary and macroeconomic imbalances can be managed and, in the same vein, long term export promotion depends on supply side policies. In this sense the costs, especially the long term costs, outweigh the benets of the reserves accumulation strategy. Finally, the need to hold extremely large international reserves can be reduced substantially if alternative policies that can provide policy autonomy, nancial stability and the space to support industrialisation are implemented.

4. Financial stability, policy autonomy and development: the alternative strategies

Recent nancial crises are strongly associated with the degree of freedom with which capital is able to leave the country given an economic or political change. This suggests that any strategy that aims at reducing nancial vulnerability without trying to reduce agents
1 Only the most optimistic studies emphasise the fact that, despite currency crises, nancial liberalisation can be linked to boombust cycles (see Tornell et al., 2004) or argue that developing countries can benet from nancial liberalisation, but with many nuances (see Kose et al., 2006). 2 It is important to mention that Grabel (2003, 2004) also proposes the trip wires and speed bumps approach as a policy alternative that can also offer nancial stability and aid the attainment of development goals. The idea consists of providing policymakers (and investors) with measurable variables that can indicate if the country is approaching dangerous levels of risk of suffering capital ight or a speculative attack, and tools to reduce such risks.

674

M. Cruz and B. Walters

Table 1. Costs and benets of the international reserve accumulation strategy Benets Costs Forgone resources for developmental purposes. Moral hazard problems. Macroeconomic risks (ination, high intervention costs, monetary imbalances). No room for industrial and growth policies. The economy remains highly exposed to hot money, particularly to capital outows.

Development purposes

Foreign exchange stability. Export growth (through a competitive RER). Increased creditworthiness. Avoid reliance on the IMF in the case a crisis occurrs. Prevents speculative attacks.

Policy autonomy Financial stability

RER, real exchange rate; IMF, International Monetary Fund.

nancial freedom to transfer funds across borders is likely to fail. As Keynes (quoted by Davidson, 2002, p. 214) warned loose funds may sweep round the world disorganising all steady business. Nothing is more certain than that the movement of capital funds must be regulated. Moreover, it is easy to forget that the Golden Age was, among other things, an era of effective national economic regulation (Crotty and Epstein, 1996, p. 118). It is clear that only strategies aimed at controlling the ow of capital and thus the behaviour of agents provide options that are likely to offer greater nancial stability. In addition, reducing the freedom of capital can increase policy autonomy because freedom to [move money across the borders] is one of the main sources of capitalist political power (Crotty and Epstein, 1996, p. 120). Policy autonomy cannot be trimmed to avoid IMF assistance when collapses occur; it needs to be broader and should imply, among other things, the capacity of the authorities to address monetary and scal policy towards industrialisation and growth goals and also be able to smooth business cycles. In other words, . . . countries must be allowed to impose capital controls in the interest of demand management (Bhaduri, 2002, p. 45). Furthermore, there is evidence that such measures have also contributed in the process of development through, amongst other channels, subsidising imports, stabilising the exchange rate and improving the allocation of investments. For these reasons, here we focus on presenting and analysing the feasible alternative proposals, summarised by Grabel (2003), which can be applied by domestic ` policymakers to achieve nancial stability. They include the management of capital ows (a la Chile) and restrictions on currency convertibility. Before discussing these options, it is important to make two important observations. First, conclusions from the literature, both for and against capital controls, suggest that they involve tradeoffs. Thus, the issue is not whether the policies have costs but whether policymakers can ensure that these costs do not offset the benets. In addition, although these policies may have been proven to offer nancial stability and policy autonomy in the past, this need not imply that they will now operate as they were originally applied; they need to be adjusted to the present context of globalisation. This raises the question of which policy or combination of policies are most appropriate. Evidently this depends on the particular circumstances of the economy (such as the level of development) and the prevailing external conditions.

Effect of accumulation of international reserves on development

675

Second, a growing body of literature has suggested an alternative approach to achieve nancial stability, and thus promote development, through the involvement of the international community.1 These proposals essentially view nancial stability as a global public good (see Kaul et al., 1999, for a denition of global public goods). The idea is that poor people may lack income and barely participate in nancial markets. Yet nancial crises, through multiple transmission channels, can hurt them badly. Thus, international nancial stability is important for all people (Grifth-Jones, 2003). However, notwithstanding the remarkable value of this insight, our emphasis in this paper is on policies that are feasible in the sense that they can be applied by domestic authorities. If developing countries really want to protect themselves and accelerate industrialisation they need to apply, sooner rather than later, policies that are effective but which are under their control. In this sense, they cannot depend on . . . international organizations nor expect that a new nancial international architecture that will make the world less dangerous (Feldstein, 1999, p. 1). This does not mean, of course, that they should stop the search for mechanisms to attain global nancial stability.

4.1 The management of capital ows (the Chilean model)

The so-called management of capital ows is one of the strategies that have been shown to be effective in reducing speculative attacks and capital ight and provide policy autonomy whilst contributing to development goals.2 This is because, as Thirlwall (2003, p. 95) points out, capital controls, in whatever form. . . allow countries to manage their exchange rate effectively and provide a greater degree of monetary independence (cf. Joshi, 2003, for the case of India). In other words, capital controls provide a solution to the policy dilemma of reconciling internal and external equilibrium. Although there may be some national differences in design, the mechanisms by which capital controls operate are to lengthen the maturity structure and stabilise capital inows, mitigate the effect of large volumes of inows on the exchange rate and exports, and protect the economy from the instability associated with speculative excess and the sudden withdrawal of external nance (Grabel, 2003, p. 326). Moreover, capital controls insulate a countrys nancial system from the contagion effects of nancial crises and enhance the autonomy of pro-growth and social policy (Epstein et al., 2003). They also promote development by attracting favoured forms of foreign investment and by maintaining foreign exchange stability, which is crucial to subsidise capital imports. In other words, the
1 See, for example, Eichengreen, 1999, and Davidson, 2002, for proposed changes in the international nancial architecture; Mendoza, 2004, and Bird and Rajan, 2003, for strategies that involve measures to redene and readjust international nancial instruments (like the SDRs) to the needs of emerging economies; Bird and Mandaliras, 2005, for mechanisms to achieve global balance-of-payments equilibrium; see also Grifth-Jones, 2003, for a review about why global nancial stability has not been achieved and what needs to be done. 2 The management of capital ows was applied successfully in Chile and Colombia during the 1990s (see Agosin and Ffrench-Davis, 1996; De Gregorio et al., 2000; Edwards, 1999, 2003; Palma, 2002), in India and China and, for a short period during the East Asian crisis, in Malaysia (see Athukorala, 2003; Doraisami, 2004; Joshi, 2003). The strategy allowed these economies to pass through the era of nancial crises with low levels of nancial and macro instability (see David, 2007; Ocampo, 2002). Also, in the early post-war years capital controls for macroeconomic reasons were generally imposed as part of policy packages dealing with balance-of-payments difculties so as to avoid, or at least reduce, the need for devaluations. Rich and poor countries alike used controls on capital inows for long term development reasons. When freer capital movements were allowed from the 1960s onwards, rich countries, such as Germany, Holland and Switzerland, when destabilised by large capital inows, imposed controls such as limits on non-residents bank deposits and their purchase of local securities (Khor, 2001).

676

M. Cruz and B. Walters

risk of currency collapse is reduced through the adoption of a crawling band exchange rate regime coupled with capital inows control1 and the probability of a sudden exit of investors is reduced by allocating investment towards longer term activities.2 In summary, the management of capital ows has proved to be a highly effective strategy to mitigate speculative attacks and capital ight and to enhance policy autonomy for progrowth goals (see David, 2007; Epstein et al., 2003). Furthermore, it is evident that those economies that have managed capital inows and outows have recorded more stable growth and a marked move towards further industrialisation. The main cost stressed by critics is that their adoption could inhibit access to nancial resources as well as investment, thereby reducing economic growth. However, there is little evidence supporting the view that portfolio investment encourages productive investment and growth (through nancing productive projects). On the contrary, there is empirical evidence suggesting that foreign direct investment does not necessarily promote growth; it is not, as Chang and Grabel, 2004, point out, the Mother Theresa of capital ows.3 In addition, the volume of capital inows does not necessarily decrease in the face of capital controls. In fact, evidence suggests that the only change is a positive one, with a larger share of capital going to longer term activities (for evidence in the case of Chile, Colombia, Taiwan, Singapore, Malaysia in 1998, India and China see Chang and Grabel, 2004, and Epstein et al., 2003). This suggests that resources can indeed be efciently allocated. Other emerging economies recently seem to have realised the benets of this strategy. For example, in June 2005, Argentina decided to apply measures to retain 30% of portfolio capital inows, with the specic aim of stabilising the exchange rate. More generally, capital management should represent a developmental option for developing economies as it did in the past for todays industrialised ones; it cannot be accepted, as Chang (2002) stresses, that in the name of market efciency the same state action could be considered an intervention in one society and not in another. Finally, many criticisms of capital management arise from their potential microeconomic costs. The argument is that capital controls have the potential to harm medium sized enterprises because the strategy forces domestic lenders to raise domestic costs during the economic boom (see Forbes, 2003, 2004). However, severe limitations on data availability in these studies suggest that the results be treated cautiously. In addition, neither study presents their analysis in the necessary comparative context, nor do they show clear evidence that the microeconomic costs are likely to offset the macroeconomic benets of capital management.

4.2 Restrictions on currency convertibility

The management of foreign exchange currency transactions provides another feasible policy measure to achieve nancial stability, increase policy autonomy and support development goals. Free currency convertibility allows investors to move their money
1 According to Thirlwall (2003, p. 79), the historical experience of the last 30 years or so points to an important exchange rate regime policy conclusion: intermediate positions between rigidly xed rates (or hard pegs) and oating (what might be called soft pegs) are not sustainable without capital controls. 2 Edwards (1999) provides evidence that these goals were achieved in the case of Chile and Ocampo (2002) illustrates their successful application in Colombia and other emerging economies. Importantly, Feldstein (1999) highlights this policy as a means of increasing liquidity, thus further improving policy autonomy. 3 See Singh (2003) for a comprehensive exposition of the disadvantages of foreign direct investment and why it should be regulated.

Effect of accumulation of international reserves on development

677

freely from one nancial centre to another regardless of the purpose of the conversion or the identity of the holder. It is this capability that allows investors to put the domestic currency under pressure cheaply and easily, further decreasing the value of assets and potentially causing a nancial crisis. The corollary is that currencies that cannot be exchanged freely for other currencies or assets denominated in them cannot be put under such pressure (Chang and Grabel, 2004). Thus, the greater the restrictions on exchanging the currency, for both national and foreign residents, the lower is the chance of the currency being subject to a speculative attack. Furthermore, capital ight is also reduced because restricted convertibility discourages investors from acquiring the kind of domestic assets that are more prone to ight because they cannot be easily and quickly converted to their own currency (particularly portfolio assets). These effects translate into a reduction in foreign exchange pressures, thus increasing policy autonomy and the ability of the authorities to pursue development goals.1 It is worth highlighting that, as with the management of capital ows, there is considerable evidence to show that restrictions on currency convertibility are capable of providing nancial stability and improving policy autonomy. The experiences of Taiwan, China and India demonstrate that such measures support the process of industrialisation by allowing the subsidy of capital imports, implementing counter cyclical scal and monetary policies and directing resources towards the development of a productive structure. Equally striking is the fact that many of todays industrialised economies only removed restrictions on full exchange convertibility for capital movements very recently: the USA, Canada, Germany and Switzerland in 1973, while Britain and Japan only did so in 1979 and 1980, respectively, and France and Italy made the transition as late as 1990 (Nayyar, 2003, p. 66). Importantly, these economies not only made this transition when they considered that their currencies (and their economies) were strong enough to resist foreign exchange pressures, but also only after this strategy had fullled its function within the development process. The debate about the likely effectiveness of this policy measure have emphasised the Achilles heel of restrictions on currency convertibility: leakages between the commercial and the nancial rate, after a dual exchange rate system has been set up. According to Eichengreen and Wyplosz (1996), experience suggests that dual exchange rates work well when the gap between the commercial and nancial rates is smallmeaning that they work least well during crises. Furthermore, the IMF has traditionally been hostile to these sorts of exchange rate practices, seeing them as an interference with the free market in goods and
1 Among the mechanisms that a government can apply are restrictions on capital account transactions, particularly restrictions on both domestic residents and foreigners and on on the form of foreign direct or portfolio investment in national assets. This measure can decrease abrupt capital ight. Alternatively, charging a higher domestic price for foreign currencies than the ofcial rate for investing abroad in capital assets such as shares and properties might inhibit ight risk (see Thirlwall, 2003).The authorities can also authorise currency convertibility for trade transactions, repayment of loans or prots repatriation previously authorised. This mechanism is similar to the adoption of a dual exchange rate system, where a specic exchange rate is set up for capital account transactions and another, cheaper exchange rate, is set up for the current account. Multiple exchange rate systems, of which the dual is a special case, imply different exchange rates for different transactions on either the current or nancial account (see Mikesell, 2001; Thirlwall, 2003). Importantly, according to the IMF Articles of Agreement (specically Article 8) this kind of selective exchange rate convertibility is allowed (Chang and Grabel, 2004). For a more complete discussion concerning the benets of the dual exchange rate system see Mikesell (2001). Moreover, the government could manage convertibility by requiring that investors apply for a licence that entitles them to exchange currency for a particular reason. This mechanism would allow the government to inuence the pace of currency exchanges.

678

M. Cruz and B. Walters

capital. This view is consistent with studies suggesting that low levels of restriction on currency convertibility (and capital controls) are associated with higher levels of economic development, efciency and integration of the nancial sector (see Johnston, 1999).1 Edwards (1999, p. 69), points out that these types of outow controls have been largely ineffective, particularly because the market nds ways of evading them; he also stresses that evidence has shown that in almost 70% of the cases where controls on outows have been used as a preventive measure a signicant increase in capital ight has been detected after the controls had been put in place. Finally Eichengreen (2000, p. 1110) has also stressed that not too much should be expected of outows controls in times of crisis, given the strong incentives that then exist for avoidance. However, Chang and Grabel (2004, p. 173) stress that critics of this type of policy generally focus on their high costs, [B]ut they overlook the fact that the resources devoted to these wasteful activities are generally dwarfed by the resources wasted in the currency speculation that frequently occurs in liberal nancial environments. Moreover, the economic and social costs of nancial instability and crisis tend to be much greater than the economic costs of convertibility restrictions. In summary, the balance of the argument and evidence supports the view that, despite evident costs, alternative policies to maintain nancial stability, policy autonomy and the pursuit of national developmental objectives exist and are feasible.

5. Concluding remarks
This paper has considered whether the widespread policy among developing economies of amassing large amounts of foreign exchange reserves is, in any sense, optimal from the perspective of supporting the process of industrialisation, achieving nancial stability and extending policy autonomy. It argues that this policy cannot be justied in these terms, particularly in the long term. This is because, even when the policy of reserve accumulation can be, and has been, defended as being legitimate relative to the enormous costs associated with nancial crises and their aftermath, this defence fails to take account of a number of issues. First, it accepts, albeit reluctantly, the necessity and, indeed, inevitability of neo-liberal nancial liberalisation of both domestic and capital markets. The paper points out that this is contestable and that the evidence shows that the neo-liberal nancial liberalisation strategy is deeply implicated in precipitating such crises (and their associated output costs) and reducing policy autonomy. Second, it is possible to defend such a strategy as a reection of an active export growth policy of maintaining a competitive exchange rate. However, it was argued that the empirical evidence does not unambiguously support this conclusion and even in situations where this strategy appears to have worked, this has been accompanied by a range of complementary, supply-side policies to promote the long term expansion of exports. It was concluded, therefore, that no general endorsement of the policy was provided by these examples. Finally, it might be argued that the costs of such a reserve accumulation strategy would be acceptable if there were no feasible alternatives available to governments. However, both current and historical evidence suggests that a range of policies based on controlling the free ow of capital are, as a matter of fact, both feasible and effective, allowing the
1

However, these studies do not show causality.

Effect of accumulation of international reserves on development

679

achievement of nancial stability, policy autonomy, and the promotion of industrialisation and growth. These policies include: the management of capital ows according to the Chilean model and a variety of restrictions on currency convertibility. It was accepted that these policies were not without cost and would need to be consistent with the new context of increasing globalisation. However, it was argued that they provided an effective alternative to holding large reserves while providing nancial stability, increasing policy autonomy and supporting the industrialisation process. In summary, this paper has argued that the policy of amassing large foreign exchanges reserves is not consistent with developmental goals. Countries seeking nancial stability and an autonomous nancial strategy should resist full capital account liberalisation and impose restrictions on the ability of capital holders to liquidate their positions without cost or delay. In other words, developing countries need the freedom of action that developed countries allowed themselves when they were in their earlier stages of development.

Bibliography
Agosin, D. and Ffrench-Davis, R. 1996. Managing capital inows in Latin America, pp. 16191 in Haq, M., Kaul, I., and Grunberg, I. (eds), The Tobin Tax. Coping with Financial Volatility, Oxford, Oxford University Press Aizenman, J. 2006. International Reserves Management and the Current Account, NBER Working Paper Series, no. 12734, 152 Aizenman, J. 2007. Large Hoarding of International Reserves and the Emerging Global Economic Architecture, NBER Working Paper Series, no. 13277, 121 Aizenman, J. and Lee, J. 2005. International Reserves: Precautionary Versus Mercantilist Views, Theory and Evidence, NBER Working Paper Series, no. 11366, 128 Aizenman, J. and Marion, N. 2002. The High Demand for International Reserves in the Far East: Whats Going On?, NBER Working Paper Series, no. 9266, 144 Aizenman, J. and Riera-Critchton, D. 2006. Real Exchange Rate and International Reserves in the Era of Growing Trade and Financial Integration, NBER Working Paper Series, no. 12363, 136 Aizenman, J., Lee, Y. and Rhee, Y. 2004. International Reserves Management and Capital Mobility in a Volatile World: Policy Consideration and a Case Study of Korea, NBER Working Paper Series, no. 10534, 127 Allen, R. (ed.) 2004. The Political Economy of Financial Crises, Cheltenham, Edward Elgar Arestis, P. and Glickman, M. 2002. Financial crisis in Southeast Asia: dispelling illusion in the Minskyan way, Cambridge Journal of Economics, vol. 26, 23760 Athukorala, P. 2003. Crisis and recovery in Malaysia: the role of capital controls, Cheltenham, Edward Elgar Bhaduri, A. 2002. Nationalism and economic policy in the era of globalization, pp. 1949 in Nayyar, D. (ed.), Governing Globalization, Oxford, Oxford University Press Ben-Bassat, A. and Gottlieb, D. 1992. Optimal international reserves and sovereign risk, Journal of International Economics, vol. 33, 34562 Bird, G. and Mandilaras, A. 2005. Reserve accumulation in Asia, World Economics, vol. 6, no. 1, 8599 Bird, G. and Rajan, R. 2003. Too much of a good thing? The adequacy of international reserves in the aftermath of crises, World Economy, vol. 26, no. 6, 87391 Cady, J. and Gonzalez-Garcia, J. 2006. The IMF Reserves Template and Nominal Exchange Rate Volatility, IMF Working Paper, no. 627, 124 Calvo, G. 2001. Capital ow volatility: issues and policies, Journal of African Studies, vol. 10, supplement 1, 1635 Cerra, V. and Sexena, S. C. 2008. Growth dynamics: the myth of economic recovery, American Economic Review, vol. 98, no. 1, 43945 Cruz, M. 2006. Pueden las reservas internacionales contribuir al crecimiento mexicano? EconomaUNAM, vol. 3, no. 8, 11524

680

M. Cruz and B. Walters

Cruz, M. and Walters, B. 2007. Acumulacion de reservas internacionales: costos y alternativas para la economa mexicana durante el periodo 19952004, pp. 7794 in Calva, J. (ed.), Macroeconoma del Crecimiento Sostenido. Agenda Para el Desarrollo, vol. 4, Mexico, Porrua Cruz, M., Amann, E. and Walters, B. 2006. Expectations, the business cycle and the Mexican peso crisis, Cambridge Journal of Economics, vol. 30, 70122 Crotty, J. and Epstein, G. 1996. In defence of capital controls, pp. 11849 in Panitch, L. (ed.), Are There Alternatives?, London, Social Register Chang, H.-J. 2002. Breaking the mould: an institutionalist political economy alternative to the neo-liberal theory of the market and the state, Cambridge Journal of Economics, vol. 26, 53959 Chang, H. and Grabel, I. 2004. Reclaiming Development. An Alternative Policy Manual, London, Zed Books Chui, M., Hall, S. and Taylor, A. 2004. Crises Spillovers in Emerging Market Economies: Interlinkages, Vulnerabilities and Investor Behaviour, The Bank of England Working Paper, no 212, 144 DArista, J. 2000. International foreign exchange reserves: an update, Capital Flows Monitor, June, 14 David, A. 2007. Controls on Capital Inows and External Shocks, World Bank Policy Research Working Paper, no 4176, 126 Davidson, P. 2002. Financial Markets, Money and the Real World, Edward Elgar, London De Gregorio, J., Edwards, S. and Valdes, R. 2000. Controls on capital inows: do they work? Journal of Development Economics, vol. 63, 5983 Doraisami, A. 2004. From crisis to recovery: the motivations for and the effects of Malaysian capital controls, Journal of International Development, vol. 16, 24154 ECLAC 1998. Impacto de la crisis Asiatica en America Latina, available at www.eclac.cl/ publicaciones (19 September 2006 last accessed) Edwards, S. 1983. The demand for international reserves and exchange rate adjustments: the case of LDCs, 19461972, Economica, vol. 50, 26980 Edwards, S. 1999. How affective are capital controls? Journal of Economic Perspectives, vol. 13, no. 14, 6584 Edwards, S. 2003. Financial instability in Latin America, Journal of International Money and Finance, vol. 22, 1095106 Eichengreen, B. 1999. Towards a New Financial Architecture: a Practical post-Asian Agenda, Washington, Institute for International Economics Eichengreen, B. 2000. Taming capital ows, World Development, vol. 28, no. 6, 110516 Eichengreen, B. and Wyplosz, C. 1996. Taxing international nancial transactions, pp. 1539 in Haq, M., Kaul, I., and Grunberg, I. (eds), The Tobin Tax. Coping with Financial Volatility, Oxford, Oxford University Press Epstein, G., Grabel, I. and Jomo, K. S. 2003. Capital management techniques in developing countries: an assessment of experiences from the 90s and lessons for the future, Paper prepared for the XVIth Technical Group Meeting of the G-24 in Port Spain, Trinidad and Tobago Feldstein, M. 1999. Self-protection for Emerging Market Economies, NBER Working Paper Series, 123 Feldstein, M. 2002. Economic and Financial Crises in Emerging Market Economies: Overview of Prevention and Management, NBER Working Papers Series, 136 Forbes, K. 2003. On Cost of the Chilean Capital Controls: Increased Financial Constraints for Smaller Trade Firms, NBER Working Paper Series, no 9777, 151 Forbes, K. 2004. Mud in the Wheels of Market Discipline, NBER Working Paper Series, no 10284, 124 Frenkel, R. and Ros, J. 2006. Unemployment and the real exchange rate in Latin America, World Development, vol. 34, no. 4, 63146 Fry, M. 1997. In favour of nancial liberalization, Economic Journal, vol. 107, no. 442 Garca, P. and Soto, C. 2004. Large Hoardings of International Reserves: Are They Worthy?, Central Bank of Chile Working Papers, no 299, 139 Grabel, I. 1995. Speculative-led economic development: a post-keynesian interpretation of nancial liberalization programmes in Third World, International Review of Applied Economics, vol. 9, 12747

Effect of accumulation of international reserves on development

681

Grabel, I. 1996. Marketing the third world: the contradictions of portfolio investment in the global economy, World Development, vol. 24, no. 11, 176176 Grabel, I. 2003. Averting crisis? Assessing measures to manage nancial integration in emerging economies, Cambridge Journal of Economics, vol. 27, 31736 Grabel, I. 2004. Trip wires and speed bumps managing nancial risks and reducing the potential for nancial crises in developing economies, Intergovernmental Group of Twenty-Four (G-24) Research Papers, 130 Grifth-Jones, S. 2003. International nancial stability and market efciency as a global public good, pp. 43554 in Inge, K., Conceicao, P., Le Goulven, K., and Mendoza, R. (eds), x Providing Global Public Goods, New York, Oxford University Press Johnston, R. 1999. Exchange Rate Arrangements and Currency Convertibility. Development and Issues, Washington, International Monetary Fund Joshi, V. 2003. India and the impossible trinity, World Economy, vol. 26, no. 4, 55583 Kaul, I., Grunberg, I. and Stern, M. 1999. Dening global public goods, pp. 219 in Inge, K., Grunberg, I., and Stern, M. (eds), Global Public Goods, New York, Oxford University Press Khor, M. 2001. Rethinking Globalisation, London, Zed Books Kose, M., Prasad, E., Rogoff, K. and Wei, S.-J. 2006. Financial Globalization: A Reappraisal, IMF Working Paper, 194 Machlup, F. 1966. The need for monetary reserves, Banca Nazionale del Lavoro Quarterly Review, Sept, 5875 Mendoza, R. 2004. International reserve-holding in the developing world: self insurance in a crisis-prone era, Emerging Markets Review, vol. 5, 6182 Mikesell, R. 2001. Dual exchange markets for countries facing nancial crises, World Development, vol. 29, no. 6, 103541 Mohanty, M. and Turner, P. 2006. Foreign exchange reserve accumulation in emerging markets: what are the domestic implications? BIS Quarterly Review, September, 3952 Nayyar, D. 2003. Globalization and development, pp. 6182 in Chang, H. -J. (ed.), Rethinking Development Economics, London, Anthem Press Ocampo, J. 2002. Capital-account and Counter-cyclical Prudential Regulations in Developing Countries, WIDER Discussion Papers, no 82, 125 Palma, G. 2002. The three routes to nancial crises. The need for capital controls, pp. 297338 in Eatwell, J. and Taylor, L. (eds), International Capital Markets. Systems in Transition, New York, Oxford University Press Palma, G. 2003. The Three routes to nancial crises: Chile, Mexico, and Argentina [1], Brazil [2]; and Korea, Malaysia and Thailand [3], pp. 34776 in Chang, H. -J. (ed.), Rethinking Development Economics, London, Anthem Press Polterovich, V. and Popov, V. 2002. Accumulation of Foreign Exchange Reserves and Long Term Growth mimeo, Moscow, New Economic School Ramachandran, M. 2004. The optimal level of international reserves: evidence for India, Economics Letters, 36570 Rodrik, D. 2000. Exchange Rate Regimes and Institutional Arrangements in the Shadows of Capital Flows, mimeo, pp. 119, available at http://ksghome.harvard.edu/,drodrik/contact.html (10 September 2006 last accessed) Rodrik, D. 2006. The social cost of foreign exchange reserves, International Economic Journal, vol. 30, no. 3, 25366 Singh, A. 2003. Capital account liberalization, free long-term capital ows, nancial crises and economic development, Eastern Economic Journal, vol. 29, no. 2, 191216 Singh, C. 2006. Should India use foreign exchange reserves to nance infrastructure?, Economic and Political Weekly, vol. 43, 51725 Schiller, T. 2007. How Asia could have vibrant capital markets, Financial Times, April 25. Thirlwall, A. 2003. Trade, the Balance of Payments and Exchange Rate Policy in Developing Countries, Cheltenham, Edward Elgar Tornell, A., Westermann, F. and Martinez, L. 2004. The Positive Link Between Financial Liberalization, Growth and Crises, NBER Working Paper Series, no 10293, 153 Wijnholds, J. and Kapteyn, A. 2001. Reserve Adequacy in Emerging Market Economies, IMF Working Paper, no 143, 148