The Fund's Role Regarding Cross-Border Capital Flows

Global capital flows have multiplied many times over in recent years, mainly between advanced economies but increasingly also to emerging markets, reflecting the general reduction in regulatory and informational barriers. Thus, with international asset positions now dwarfing output, global portfolio allocations and reallocations have profound effects on the world economy, as demonstrated by recent boombust episodes of both global reach (e.g., the transmission of the 2001 IT shock and the 2008 mortgage market shock from the United States) and regional significance (in Asia, Latin America, and Central and Eastern Europe). Such cycles and reversals in cross-border capital flows should not be surprising, given that these flows - more so than domestic ones - imply crossing informational barriers, currency and macroeconomic risks, and regulatory regimes.


Global capital flows have multiplied many times over in recent years, mainly between advanced economies but increasingly also to emerging markets, reflecting the general reduction in regulatory and informational barriers. Thus, with international asset positions now dwarfing output, global portfolio allocations and reallocations have profound effects on the world economy, as demonstrated by recent boombust episodes of both global reach (e.g., the transmission of the 2001 IT shock and the 2008 mortgage market shock from the United States) and regional significance (in Asia, Latin America, and Central and Eastern Europe). Such cycles and reversals in cross-border capital flows should not be surprising, given that these flows - more so than domestic ones - imply crossing informational barriers, currency and macroeconomic risks, and regulatory regimes.

I. Overview1

1. Context. Global capital flows have multiplied many times over in recent years, mainly between advanced economies but increasingly also to emerging markets, reflecting the general reduction in regulatory and informational barriers (Figure 1). Thus, with international asset positions now dwarfing output, global portfolio allocations and reallocations have profound effects on the world economy, as demonstrated by recent boom-bust episodes of both global reach (e.g., the transmission of the 2001 IT shock and the 2008 mortgage market shock from the United States) and regional significance (in Asia, Latin America, and Central and Eastern Europe). Such cycles and reversals in cross-border capital flows should not be surprising, given that these flows—more so than domestic ones—imply crossing informational barriers, currency and macroeconomic risks, and regulatory regimes.

Figure 1.
Figure 1.

Cross-border Capital Flows, 1985-2009 1/

In trillions of US dollars, unless otherwise specified

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

Source: WEO.1/ Gross capital inflows defined as inward FDI, portfolio and other investment inflows from the balance of payments. Net flows defined as net FDI plus portfolio and other investment balances.

2. Role of the Fund. Insofar as volatile capital flows reflect deeper forces (e.g., different interest rates, currency regimes, growth prospects, and regulatory practices), the policy debate arguably should be about underlying drivers, not whether capital flows should or should not be controlled. This is certainly true. However, it is also true that the sheer size of international flows often eclipses domestic flows and that the global architecture that enables or hinders capital flows is a patchwork of bilateral, regional, and other arrangements with differing and discriminatory provisions. In contrast to trade in goods and services, there are no widely accepted “rules of the game” for international capital flows—this despite being the principal conduit for the transmission of global shocks. Unfortunately, and notwithstanding its mandate to oversee the stability of the international monetary system, the Fund has been hamstrung in its efforts to forge such rules, reflecting perceived ambiguity in its Articles, divergent attitudes among members, and the legacy of a failed attempt to confront the issue in the late 1990s. As a result, although Fund advice has been offered in individual cases and some conceptual markers have been laid (e.g., strengthening financial regulation ahead of opening up to capital flows), the Executive Board has not had a broad ranging discussion of capital account liberalization and controls since 1997.

3. Post-crisis challenges. In the aftermath of the global crisis, and especially now with resurgent capital flows requiring a considered policy response, it is not tenable for the Fund to remain on the sidelines of a debate so central to global economic stability. Various parts of the membership are taking divergent approaches to surging capital inflows—among other things, currency intervention, taxes, reserve requirements, and prudential measures. The appropriateness of these responses, both for the member itself and for the wider global good (e.g., they may only divert flows to other countries or exacerbate other imbalances), have not been fully evaluated. Yet doing so is nearly impossible in the absence of an institutional line and consistent framework. It is not surprising that the IMFC and the G-20 have identified volatile capital flows as a key element in the reform of the international monetary system.2

4. Goal and approach. The aim here is to delve into these institutional issues, exploring the role the Fund could play in developing rules of the game for global capital flows and in fostering multilateral, nondiscriminatory, approaches that look to the interest of both the originators and recipients of capital. At the risk of over-simplifying, the main line of argumentation is as follows:

  • Section II. The academic evidence on the impact of capital flows is mixed, with the benefits of higher investment and risk diversification also coming with a destabilizing tendency for capital floods and droughts, especially in emerging markets. Such problems are hardly confined to cross-border financial transactions, but they are more difficult to handle at the country level given the important role of global drivers and externalities in individual country policies. Moreover, tracking the risks has been made harder by significant information gaps. As such, there is a strong case for a multilateral approach that takes into account global and domestic stability.

  • Section III. Cross-border capital flows take place without global “rules of the game.” At the Fund, and in contrast to the obligation to liberalize payments and transfers for current international transactions, the Articles of Agreement explicitly grant members the right to “exercise such controls as are necessary to regulate international capital movements.” Since the failure to amend this provision of the Articles (Article VI, Section 3) in 1997, the Fund’s work in this area has focused on analytical and conceptual issues, with policy advice not always offered consistently, as pointed out by the Fund’s own Independent Evaluation Office. With more countries looking for guidance on how to handle surging capital inflows, and apt to go in divergent directions, Fund staff recently articulated some of the hitherto implicit agreed wisdom on capital controls in a staff position note, but the Executive Board has yet to pronounce on these issues.

  • Section IV. A more pro-active and systematic role for the Fund with respect to global capital flows seems desirable. In developing an institutional view, the first step might be for the Executive Board to articulate its views on responding to capital surges and reversals—in what circumstances, through what kind of instruments, and over what time horizons—in the context of the upcoming discussion of Cross-Cutting Themes from Recent Country Experience with Capital Flows. Thereafter, the Board could take up the question of endorsing principles on these issues. These principles, derived from cross-country experience and further analytical work, would be applied with due regard to country-specific circumstances in the context of bilateral surveillance. The difficulty of agreeing on such principles should not be underestimated, and the Fund should consult widely as part of that process. But neither should the cost of remaining silent, including potentially having case-by-case pronouncements lead to disparate and conflicting approaches. The Fund could also take a number of steps to support multilateral cooperation on macroeconomic, macro-prudential and structural policies affecting capital flows, taking into account the roles of both the recipients and suppliers of capital and building on efforts already under way in various fora.

  • Section V. In the longer term, to provide a more complete framework to address the complex issues related to international capital flows, consideration could be given to amending the Articles of Agreement. This could take the form of an obligation to ultimately liberalize capital movements, subject to safeguards and routine exclusion of prudential measures. Alternatively, and more neutrally, one might consider an amendment calling on members to collaborate with the Fund and others to ensure that capital movements are consistent with international monetary stability.

5. Caveats. This paper leaves specific guidance and principles on capital account policies to upcoming papers. Neither does this paper dwell on necessary efforts to address global imbalances, which are germane to the debate on global capital flows but are the subject of specific initiatives in which the Fund is involved, such as spillover reports and the G20 Mutual Assessment Process. No presumption is made here that capital account liberalization is a goal in itself in all cases, but rather that a broader range of tools and advice, contemplating both the elimination and imposition of controls, may be more appropriate for domestic and systemic stability. It is recognized that international capital flows are only a type of financial flow—the type that crosses borders—and that the overall approach must fit in a broader vision of macro-prudential regulation and supervision. As such, this is an area in which the Fund will need to collaborate with others if it is to be effective, including the Financial Stability Board (FSB), regional and national stability boards, and other regulators.

II. Challenges Calling for Collective Solutions

6. Returns and risks. Financial globalization enhances countries’ access to financing for productive domestic activity and, in principle, improves global resource allocation, thereby fostering higher levels of income and development. Increased exposure to international capital flows can also induce competition and develop the financial sector, deepen domestic capital markets, and overall generate efficiency gains in the growth process. In practice, abundant evidence of these benefits exists. Henry (2007), for example, argues that capital account liberalization, by allowing a temporary increase in investment and growth, generates permanent welfare gains. Nonetheless, country experiences with financial globalization have not universally been evaluated as positive (see Box 1, and Dell’Arriccia et al, 2008). This reflects five perceived challenges with cross-border capital flows—volatility, interconnectedness, size, their global drivers, and information gaps—none of which can be handled exclusively at the recipient country level. The case for collective action to address these challenges and thus preserve and extend the benefits of capital flows is strong.

A. Volatility

7. Stop and go. Both advanced and emerging market countries have often been subject to strong cycles in capital inflows that impose challenges on macroeconomic and prudential policy (Figures 1 and 2). While inflows played a role in the era’s prosperity in many countries, persistent surges during 2003–07 were also a factor (among others) in large real appreciations, unmanageable credit expansions and even misallocated investments and asset price bubbles. The process unwound quickly and disruptively with the sudden stop of capital inflows in 2008—Reinhart and Rogoff (2009) document several such “capital flow bonanzas” for the U.S., U.K., and several other advanced and emerging economies. While “sudden stops” in international capital flows have played a role in some advanced economy crises (e.g., in the Nordic countries in the early 1990s and Iceland in 2008), they have been more frequent and damaging for emerging markets, at least until the 2008 crisis. However, strengthened economic fundamentals and more robust external positions—with more domestic-currency denominated debt—have increased resilience in many emerging markets, contributing to their better weathering the crisis and avoiding the worst consequences of the capital flow reversal. Some low-income countries also became recipients of private capital flows during the mid-2000s, suffering a marked contraction during the crisis.

Figure 2.
Figure 2.

Percent Change in Gross Capital Inflows, 1999-2009

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

Source: Internatio nal Financial Statistics, IMF; staff estimates.

8. Procyclical bank funding. During cyclical upswings, bank lending typically increases more quickly than stable funding bases such as customer deposits (see Shin and Shin, 2010 and Adrian and Shin, 2010, who also discuss similar patterns in leverage, and the implications for financial risks). The additional financing required is provided through wholesale markets, which, for emerging economies with shallower domestic capital markets, mostly takes the form of cross-border borrowing from financial centers. Even in advanced economies, very large cross-border positions can build over time—for example, European banks reliance on short-term dollar funding from U.S. institutions in the years before the crisis. While this borrowing can represent an efficient recycling of liquidity, it has also often been in foreign currency or, when hedged, embedding a significant maturity mismatch.3 In general, wholesale financing has also been more volatile than “core deposits” of domestic, nonfinancial depositors. Significant outflows have resulted when apprehensions about counterparty risk or generalized “flights to safety” have occurred, or alternatively when domestic downturns have led to bank deleveraging. This has been an important contributor to capital flow volatility.

Evolving Views on Capital Account Liberalization

With the consensus among professional economists that more liberal trade flows have been welfare-enhancing, many have advocated similar views on capital flows. In a study focused on emerging markets, Frankel (2010) summarizes five main arguments that capital account liberalization can enhance long-term growth prospects through allowing: (i) access to finance for investment at lower cost; (ii) consumption-smoothing in the face of adverse shocks; (iii) diversification of assets and liabilities across countries; (iv) “technology transfer” from foreign banks; and (v) greater discipline on macroeconomic policy. Kose, Rogoff, and Wei (2006) also offer a nuanced, but supportive argumentation for capital account liberalization, while Henry (2007) presents a more clearly favorable theory and interpretation of the evidence (this study also provides a comprehensive literature review, citing authors on both sides of the debate).

Nevertheless, empirical findings have not been clear-cut. Some work has found evidence of benefits (see, for example, Edwards, 2001 and Arteta, Eichengreen, and Wyplosz, 2001). While making a positive overall assessment, these two studies show that a subset of economies—those with better institutions—tended to capture the gains, while in others, the results were less clear. Other studies have not found benefits: Rodrik (1998) suggests that capital account openness does not lead to stronger growth, while Prasad, Rajan, and Subramaniam (2007) go further and argue that there may even be evidence that countries that eschew foreign capital may enjoy a growth premium. Gochoco-Bautista (2010) argues that the qualified nature of the conclusions in those studies that show a positive relationship is evidence that capital account liberalization yields benefits only under a narrow range of conditions. However, Henry (2007) argues that studies finding no effects are methodologically flawed, and the academic debate continues. Developing countries have in fact exported substantial amounts of capital (particularly over the past decade), and there is a large volume of literature emphasizing the costs of volatility in capital flows (e.g., Calvo, 2005 and references therein). Some strong advocates of free trade have conceded that similar arguments are not applicable to financial liberalization (e.g., Bhagwati, 1998).

In parallel to this debate, some have found that financial development appears to make maintenance of a relatively closed capital account progressively more difficult, with rising ease of evasion and cost administering controls (see Ishii and Habermeier, 2002, and Bakker and Chapple, 2002), and possibly declining benefits (Wei and Zhang, 2007). Meanwhile, the potential pitfalls of liberalization alongside domestic financial reform have been recognized for at least a generation (Diaz-Alejandro, 1985). In this light, Ishii and Habermeier (2002) advocate a carefully sequenced liberalization framework that considers:

  • Pace. There is no consensus on the appropriate speed of liberalization. Overly hasty loosening appears to have contributed to financial instability in a number of cases. While gradualism allows agents to adjust, interest groups may use delays to try and frustrate reform.

  • State of financial regulation. Weak regulation, especially of banks, is associated with financial crises following liberalization and suggests a more measured pace of liberalization, with regulatory reform running in parallel. Specific risk exposures of financial institutions should also be considered.

  • Macroeconomic fundamentals. Weaknesses in fiscal or external conditions, or questions on debt sustainability, could result in premature capital account liberalization resulting in a rapid build-up in imbalances or an increase in vulnerability to large reversals in capital flows.

B. Interconnectedness

9. Shock transmission. The recent global crisis has illustrated that international capital flows between financial firms can engender vulnerabilities as well as efficiencies. In the run-up to the crisis, large, short-term cross-border liabilities were accumulated between financial institutions, ensuring the rapid transmission of distress when the quality of assets held by many of these institutions came into question (Figure 3 maps the most central nodes in the web of financial connections). While flows between advanced economies dominated, global banks played a role in transmitting liquidity shocks to emerging markets through both their internal capital markets and their cross-border lending to domestically owned banks (Cetorelli and Goldberg, 2010). These banks did also play a critical role in stabilization in some cases, through specific commitments on debt rollovers and recapitalization (e.g., under the European Bank Coordination—“Vienna”—Initiative). Among the advanced economies, with much larger flows, reliance on funding in foreign currencies complicated the policy response, necessitating the creation of the swap lines between central banks (Maguire and von Peter, 2009). Mutual funds and other investment vehicles in the “shadow banking system” also played a critical role in cross-border financial flows (see Figure 3 and Understanding Financial Interconnectedness (IMF Policy Papers, October 4, 2010) for more details).

Figure 3:
Figure 3:

Interconnected Financial Systems

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

C. Magnitudes

10. Trends. Capital flows between advanced economies reached unprecedented levels in the years before the crisis, with the cumulation of flows leading to very high cross-border asset and liability positions in relation to GDP for many countries (Figure 4). Net capital inflows to emerging markets also accelerated through this period peaking in 2007, at US$550 billion (Figure 5). In relation to GDP, similarly sized net flows occurred in the mid-1990s, but gross inflows reached double their previous peak at over 11 percent of emerging market GDP.4 A strong resumption in flows appears to be under way.

Figure 4:
Figure 4:

External Assets plus Liabilities

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

Source: Lane and Milesi-Ferretti (2006), data update to 2009.
Figure 5.
Figure 5.

Capital flows to emerging markets

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

Source: WEO

11. Absolute size. At a systemic level, the size and nature of cross-border financial liabilities between major financial centers implied that unprecedented international coordination was required to mobilize resources sufficient to stem the impact of the liquidity shock. Given that usage of the U.S. Federal Reserve swap lines rose to US$600 billion during the crisis, it is pertinent to ask how future crises might be tackled should capital flows, and the attendant balance sheet exposures, resume their former growth trends.

12. Absorption constraints. Lack of depth of local banking and financial markets, and weaker institutions have been found to increase the risks associated with large capital flows (Dell’Ariccia et al, 2008, Kose et al, 2006). Research has found that shallower markets have allocated foreign capital less efficiently, which could contribute to boom-bust cycles in credit, investment, and the broader economy. They also have a lower capacity to absorb inflows without large changes in relative prices. Findings also suggest that financial integration is associated with greater volatility in domestic consumption for countries below certain thresholds in financial depth and institutional development in areas such as financial supervision or corporate governance. Low income recipients of flows may face particular weaknesses. And while ensuring strong “fundamentals” (or at least stronger than those of peer countries) can strengthen capacity to handle capital flows and avoid sudden stops, they also risk attracting yet more inflows. Figure 6 shows net capital flows in relation to GDP and measures of banking sector depth, illustrating that flows have often been very large and that large variations in net inflows are typical (even flows falling within the middle 50 percent of the range of observations vary by 10 percentage points or more of GDP in some cases). In the current upswing in flows to emerging markets, growing evidence of a switch from bank to portfolio bond flows could represent an additional policy challenge in terms of coordinating creditors or the effects of leverage for some countries, should the tide reverse.

Figure 6.
Figure 6.

Net capital inflows, selected countries

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

Source: IFS, WEO, staff calculations.1/ Except Croatia (1992-2009), Czech Republic (1995-), Estonia (1993-), Latvia (1992-), Lithuania (1992-), Russian Federation (1992-), Slovak Republic (1993-), Slovenia (1992-), Ukraine (1992-). 2/ Except Albania (2002-2009), Algeria (2001-), Hong Kong (2001-), Croatia (2001-), Cyprus (1997-2007), Czech Republic (2002-), Dominican Republic (2001-), Latvia (2003-), Malta (1990-2007), Morocco (2001-), Philippines (2006-2008), Romania (2001-), Slovak Republic (1990-2008), Slovenia (1990-2006), Thailand (2003-), Venezuela (1990-2008).

D. Global Drivers

13. Structural factors. Strong, structural, economic reasons underpin much cross-border investment: ageing populations in advanced economies, sustained differential in growth potential between emerging markets and advanced economies, steadily improving access to information and declining home bias in investment all suggest capital flows will keep increasing, albeit not necessarily in a steady fashion.5 For example, the declining home bias of Japanese investors has resulted in total capital investments abroad rising from US$3 trillion in 2000 to US$5 trillion in 2006, with some of these flows invested in higher-yielding assets, including emerging markets.6 Aligning asset allocations to commonly used benchmarks could lead to substantial new flows to emerging markets for the foreseeable future; a recent survey of U.S.-based institutional investors found they had between 3 and 5 percent of their portfolios allocated to emerging markets, compared to the 13 percent weight in the MSCI all-country index, a standard benchmark for global portfolios (Institutional Investor, 2010). A one percent reallocation of institutional investors’ assets—which sum about US$46 trillion (BIS 2007)—to emerging markets would lead to new portfolio inflows of US$460 billion. This compares to total net inflows of US$550 billion at the peak in 2007. Moreover, all other things equal, these flows will go disproportionately to countries with relatively more open capital account policies.

14. Global liquidity. The global nature of capital flow cycles (Figures 12, 5) suggests that cross-border flows may be driven by common factors such as global liquidity conditions or overall increases in risk appetite. In particular, the tendency of whole columns in Figure 2 to turn red (surge) or blue (reversal) is striking, speaking to the role of global factors and interconnectedness. While there is no widely accepted definition of “global liquidity,” it is usually assumed to be influenced by policy choices in key financial centers, and its multiplication and transmission largely determined by the incentives, regulations, and business decisions of global financial institutions headquartered in those financial centers. For example, low interest rates in financial centers are cited as motivating “carry trade” speculative financial flows, with little regard for destination country fundamentals. Empirical studies of capital flows to emerging markets have typically proxied global “push” factors using U.S. interest rates or high-yield spreads and growth in advanced economies, and found them to be significant determinants, although less so than domestic “pull” factors.7 Abundant liquidity in core financial markets has been found to have a positive impact on equity returns and a negative impact on real interest rates in emerging markets (IMF 2010a).

15. Risk externalities. Financial institutions will typically not consider the systemic vulnerabilities their actions engender, and single-country regulation can simply push financial activity elsewhere, without reducing systemic risk. Within economic cycles, banks may generate an externality by taking on too much risk as asset prices rise and the economy expands, without taking into account the consequences to the broader economy when the cycle turns down (see e.g., Shin, 2010). Across cycles, large, complex financial institutions can hold claims on some countries that are large in relation to the economy (or financial sector) of the debtor, yet small for the institution as a whole (Figure 7). This implies small shifts in assets within the institution can generate large capital flows for the debtor countries, and that neither the financial institution nor their home regulators will attach much weight to risks taken in these economies.

Figure 7.
Figure 7.

Cross-border claims of BIS reporting banks, selected countries

December 2009

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

Source: BIS Consolidated banking statistics, ultimate risk basis

16. Global recycling. Emerging markets have been significant net exporters of capital overall in recent years, despite being recipients of strong FDI, portfolio, and bank inflows. This is due to stronger current account balances than in earlier economic cycles, and high reserve accumulation reflecting precautionary “self-insurance” from capital flow volatility, or exchange rate policies. Thus, private inflows (on current or financial accounts) are being intermediated and reinvested overseas through reserve accumulation, with the attendant costs in foregone consumption and investment for the intervening country, and, given the concentration of reserves holdings in few eligible assets, some negative implications for the international monetary system (see How Did Emerging Markets Cope in the Crisis? (IMF Policy Papers, June 15, 2010) and Reserve Accumulation and International Monetary Stability (IMF Policy Papers, April 13, 2010)).

17. Markets microstructure. Some of the original causes of volatility may lie with the microeconomic structure of financial markets:

  • “Contagion” may arise if liabilities of different markets are held by common investors: problems in one market can lead to sales of holdings in other markets to generate liquidity to meet redemptions or meet margin calls.8

  • “Herd behavior” may result if information on a debtors’ true solvency is costly to gather, and resource-constrained investors “rush for the exit” following the lead of specialists perceived as having better information (regardless of the latter’s motive for selling).9 While studies have focused on emerging market sovereign debt, the crisis illustrated how wholesale funding markets in systemic financial centers were subject to similar bank run dynamics.

  • Rewarding fund managers based on short-term performance, or the adoption of similar risk-return assessment technologies across firms, may create incentives to invest and divest similarly despite fundamentally different investment objectives.10

  • Regulatory arbitrage in core financial markets may also have played a role in increasing fragility for all. Obstfeld and Rogoff (2009) argue that European financial institutions’ demand for highly rated structured products backed by U.S. housing loans was motivated by a desire to reduce their capital requirements, allowing them to progressively increase their leverage as the economic cycle matured. Funding of these positions with short-term dollar debt turned out to be a critical vulnerability (see also Understanding Financial Interconnectedness (IMF Policy Papers, October 4, 2010).

E. Information Gaps

18. Identifying risks. Robust and timely data on capital flows are essential to effective surveillance. Over the past decade there has been a significant improvement in the availability of data collected on capital flows through the financial account of the balance of payments statistics, particularly on the related stock positions through international investment position (IIP) data and exercises such as the Coordinated Portfolio Investment Survey (CPIS) (see Annex 1). But the crisis revealed the need for further work to identify potential risks from cross-border flows in several dimensions.

19. Frequency. Balance of payments data are typically produced quarterly (although some countries do produce balance of payments data on a monthly basis), with some lag. However, the crisis underlined the importance of more timely and higher-frequency real and financial indicators, at least for systemically important countries and financial institutions (Claessens et al, 2010).

20. “Gross” or “net”. Financial account entries in the balance of payments statistics are reported on a net basis, summing flows in both directions (e.g., flows under portfolio liabilities are purchases by nonresidents of securities issued by residents, minus amortizations and other payments of principal—and can therefore be negative). Differences between this and the underlying flows in both directions can be very large (see Figure 8 for an example with respect to debt flows). This does not allow for an assessment of the growth in the magnitude of assets or liabilities that could generate fragilities.

Figure 8.
Figure 8.

External borrowing in the BOP

Emerging markets, in percent of GDP

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

Source: WEO

21. Level. Current data focus on cross-border flows in individual countries, while less has been done to identify the systemic conditions—global liquidity and structural factors—that drive such flows and have implications for international financial stability. Aggregating across countries, data on gross, underlying capital flows (i.e., one-way inflows without netting payments in the other direction for that line) could provide insights into trends in global liquidity, and potentially help in identifying threats to global financial stability. Further, better data can also inform how any policy actions might affect capital flows, and identify where risks may lie in the event of a liquidity shock.

III. The Legal and Institutional Framework for Cross-Border Capital Flows

22. No global framework. The common roots and potential solutions for risks associated with capital flows arise in the context of an absence of any universal frameworks for addressing them. In contrast to trade and related payments, for example, there is no universal framework that governs or otherwise oversees international capital movements. Rather, the frameworks in this area are mainly regional and bilateral, and most of them do not approach capital account issues from the perspective of macroeconomic stability, or consider the effects their provisions may have on global stability (see Annex 2).

A. The Fund’s Legal Framework: an Historical Overview

23. Capital/current account asymmetry. The Fund’s mandate with respect to international capital movements is much more limited than its comprehensive mandate vis-à-vis payments and transfers for current international transactions. Under Article VI, Section 3, members are expressly given the right to “exercise such controls as are necessary to regulate international capital movements” (see Annex 3 for a more detailed discussion of this provision). In contrast, Article VIII, Section 2(a) imposes a general obligation on members to refrain from imposing restrictions on the making of payments and transfers for current international transactions unless they are authorized by the Fund.11

24. Original context. The asymmetry described above reflects the overriding—almost exclusive—emphasis that was placed on international trade when the Fund was established. It also reflects certain assumptions regarding the division of labor for international economic activity among the relevant international organizations: bodies such as the General Agreement on Tariffs and Trade (GATT—now the World Trade Organization, WTO) were to be responsible for the liberalization of trade in goods and now services, while the Fund would ensure that members liberalized the payments and transfers associated with such trade. Conversely, the absence of a framework that would guide the liberalization of capital movements reflected the rather negative view of capital flows that then prevailed, premised on the belief that speculative capital movements had contributed to the instability of the pre-war system and that it was necessary to control such movements. It also reflected the fact that, at the time, almost all members maintained comprehensive capital controls that the drafters of the Articles assumed would remain in place for the foreseeable future (in contrast to the controls then maintained on current transactions, which were expected to be eliminated over a shorter time frame under the multilateral system administered by the Fund).

25. Second Amendment. By the time of the adoption of the Second Amendment of the Articles in 1978, the world had changed. Many advanced economies were liberalizing their capital accounts and it was recognized that international capital movements had begun to play an important role in the functioning of the international monetary system (IMS). Indeed, the key provision of the Articles that was introduced at that time—Article IV—explicitly recognizes that “the essential purpose of the international monetary system is to provide a framework that facilitates the exchange of goods, service, and capital among countries.” Although the text of Article VI, Section 3 was not changed, the Fund adopted policies recognizing in the context of the Second Amendment that the right of members to regulate capital movements under Article VI was now qualified by their newly established obligations under Article IV relating to the stability of the system of exchange rates. More specifically, the 1977 Decision on Surveillance Over Exchange Rate Policies recognized that members’ use of capital controls could give rights to a breach of their obligations under Article IV, Section 1(iii) to avoid manipulating their exchange rates in order to prevent effective balance of payments adjustment or to gain an unfair competitive advantage. Box 2 discusses the state of play with regard to capital account policies and their relevance to members’ domestic and external stability.

26. 1990s reform effort. The central proposal of the late 1990s reform effort was to amend the Articles of Agreement to eliminate the asymmetric treatment of current and capital flows. Consistent with the approach taken with respect to current payments, the amendment would have imposed upon members a general obligation to liberalize capital movements. The reform embodied a strong presumption that there were both individual country and systemic benefits from capital account liberalization in almost all states. Even then, however, it was expressly understood that the liberalization obligation under the Articles would be subject to important qualifications. First, transitional rules would have ensured that members would not be required to liberalize until they had in place both macroeconomic and financial regulatory policies and institutions that would ensure that liberalization did not compromise financial stability. Second, recognizing that in the context of a financial crisis, the magnitude of capital outflows could outstrip both the adjustment capacity of the member and the financial resources of the Fund, approval policies would be in place to ensure that members would be able to impose controls on a temporary basis in these circumstances and for reasons of macroeconomic management more generally. Additional approval policies were also contemplated for restrictions imposed for prudential purposes and for market and institutional evolution reasons.

27. Outcome. In the event, this reform effort did not come to fruition, largely due to the reluctance of key members to cede sovereignty in this important area, coupled with the belief held by many that the Asian crisis—with which the world was still wrestling at the time—had its roots in premature capital account liberalization.12 The failure, however, proved consequential, as liberalization de facto continued, but without a global organizing framework and with the Fund more reluctant to articulate forceful positions on capital account issues—indeed seemingly operating under a perception that it should only tread carefully with these issues, if at all.

28. The 2007 Decision. The 2007 Decision on Bilateral Surveillance over Members’ Policies introduced the concept of “external stability” (i.e., a balance of payments position that does not, and is not likely to, give rise to disruptive exchange rate movements) as an organizing principle for bilateral surveillance. The 2007 Decision also specifies that the Fund’s bilateral surveillance will focus on policies that significantly influence present or prospective external stability, and includes financial sector policies among those that are always to be covered in this context; it further requires the Fund’s assessment of balance of payments developments to include an evaluation of the size and sustainability of capital flows. Moreover—and similar to the 1977 Decision—the 2007 Decision identifies certain developments, including some related to capital controls and flows, that could signal the need for a closer review or a discussion with members regarding their observance of the exchange rate principles set forth in the 2007 Decision. The 2007 Decision, however, does not contain any specific principles or recommendations with respect to capital account policies as such.

B. Fund Practice

29. Ambiguity. The Independent Evaluation Office’s (IEO) 2005 Evaluation Report on “The IMF’s Approach to Capital Account Liberalization” concluded that it was generally understood that the Fund had a surveillance role over capital account policies but that ambiguity remained, and that this had led to some lack of consistency in advice across the Fund’s membership. The IEO report also noted that through the 1990s staff advice generally followed a pro-liberalization line, albeit not on the basis of Board-endorsed policy guidance, and thereafter became gradually less prescriptive.

30. Current surveillance practice. In recent years, the practice has been to address capital account issues primarily in the context of bilateral surveillance, where the Fund has provided ad hoc policy advice regarding the merits of liberalization in the context of selected Article IV consultation discussions. An internal poll of desk officers covering around 50 emerging markets conducted in January 2010 showed that staff discussed policy responses to ongoing increases in capital inflows with country authorities in 14 cases, and found that advice had been varied and tailored to individual countries’ circumstances, although in nearly half the cases there was no discussion in a staff report. Macroeconomic aspects of capital inflows have been discussed in the World Economic Outlook (e.g. Macroeconomic Implications of Capital Inflows, October 2007), and global factors affecting capital flows are also regularly discussed in the Global Financial Stability Report (GFSR—an example is a chapter on Global Liquidity Expansion: Effects on Receiving Economies and Policy Response Options in April 2010), although until recently without emphasis on policy response. While the October 2010 GFSR highlighted proposals to address systemic drivers of capital flows (Chapter 2), the onus of “policy response” remained on individual countries affected by capital flow volatility rather than on source factors.

31. What advice? That said, there is still no common analytical framework or guiding principles to underpin the Fund’s advice on capital account policies given in individual cases. There has been little discussion in surveillance of the impact of individual country macroeconomic or financial policies on capital outflows, barring crisis or potential crisis cases where outflows posed a threat to domestic stability. Regarding capital account policies, earlier orthodoxies in favor of liberalization and opposing capital controls—which even then were not encoded but expressed through speeches and research—have become more nuanced (Ostry et al, 2010), and neither the benefits of liberalization nor the costs and effectiveness of capital controls are well-established in the empirical research (see Box 1). Even the Fund’s efforts to track the degree of openness of members’ capital accounts through the AREAER have been of limited usefulness because of the absence of information on enforcement or severity of measures. Review by staff, management, and the Board are relied upon to ensure consistency in country advice. However, this approach has limits.

32. Crisis controls. Controls on capital outflows have been included in economic programs supported by the Fund in capital account crises where large outflows have threatened to overwhelm emergency financing (including under Fund arrangements) and deplete international reserves. Examples include Argentina in 2002 and Iceland in 2008. These concerns have overridden arguments that controls on outflows are particularly distortionary (e.g., by dissuading inflows in a nondiscriminatory fashion, whereas price or instrument-based controls on inflows still allow for high-value investments).

C. Other International Arrangements and their Implications for the Fund

33. A patchwork of agreements. Many Fund members have assumed legal obligations to liberalize capital movements under a broad range of international agreements with varying objectives and scope, including in particular the OECD Code of Liberalization of Capital Movements, the Treaty on the Functioning of the European Union, the General Agreement on Trade and Services, and between 2,500 and 3,000 bilateral or regional investment treaties or free trade agreements with investment chapters (see Annex 2). Indeed, the IEO report notes that the liberalization process for Fund members has often been driven by agendas such as OECD and EU ascension and commitments under bilateral and regional agreements. Most of these bilateral and regional agreements do not take into account macroeconomic and financial stability, and the patchwork they form for the regulation of international capital movements is generally not optimal for IMS stability or for a multilateral approach. For example:

Capital account policies in the Fund’s membership and their Relevance to Domestic and External Stability

Overall trend

Liberalization, maintenance, or re-imposition of capital controls has responded to domestic policy priorities, and has had often far-reaching consequences for domestic stability. These policies have also had implications for partners, peer countries and, collectively, the IMS as a whole. Over the past decades, there has been a trend toward capital account liberalization (Figure 9, see also IMF, [forthcoming]). This trend has declined in recent years, with emerging markets tightening controls on inflows at the margin, although this has not yet made an appreciable difference overall. More generally, for many emerging markets the process of integration with international capital markets has been “stop and go”, with periods of liberalization and growing capital inflows punctuated by reversals, and a less linear approach to liberalization, with some reversals in policy and a more frequent use of capital controls or prudential policies with a similar ultimate effect. The liberalization process has also varied across types of financial instruments, with a larger share of countries maintaining controls on debt creating capital inflows (see table—however, the data does not discriminate between light or severe measures, the rigor of their enforcement or possible declining effectiveness over time). Using a smaller but richer sample through to 2005, Schindler (2009) also finds a steady decline in restrictions, with a modest increase in the last years regarding some (mainly debt) inflows.

Figure 9:
Figure 9:

Trends in Capital Account Openness

Citation: Policy Papers 2010, 106; 10.5089/9781498336512.007.A001

Source: Chinn and Ito (2008). Where data for 1985 was not available, earliest available was used.

Controls by instrument, 2009

Percent of members with no controls

article image
Source: AREAER

Potential effects of capital account policies on domestic and external stability

  • Prolonged maintenance of controls: could effectively protect the financial sector from competition, and act as a tool of financial repression. Long-run efficiency costs could be high. This idea lies behind the intellectual and policy trend toward liberalization.

  • Poorly sequenced or “too rapid” liberalization: financial stability can be compromised by rapid liberalization in contexts of weak supervision or regulation, or unbalanced macroeconomic conditions (see Box 1). Some examples of crises where liberalization and its consequences played a role could include Chile in the early 1980s and the Nordic crisis of the early 1990s. Liberalization in accordance with E.U. accession requirements contributed to several emerging economies in Europe experiencing strong inflows and subsequent reversals in recent years.

  • Crisis controls: Once such a crisis arrives, controls on capital outflows may be necessary in circumstances where the potential level of outflows exceeds both the member’s adjustment capacity and the financing available from the official sector.

  • Externalities from controls: to the extent capital flows to emerging markets are determined by “push factors”, their overall magnitude may be relatively invariant to conditions in individual markets. Imposition of capital controls may simply displace flows to other economies with similar attributes. Policymakers have raised this concern in recent weeks, though it may be difficult to show empirically.

  • Bilateral and regional investment agreements in most cases do not allow for the introduction of restrictions on capital outflows in the event of a balance of payments crisis. The problem is exacerbated by the fact that the investments covered by these agreements are often very broad, including short-term capital flows.

  • The free transfer provisions of many of the existing investment agreements effectively limit the ability to impose controls on inflows, notwithstanding the fact that such controls may be desirable if other policy options are constrained, as discussed above and in Box 3. Moreover, market access commitments under these agreements may press the pace of liberalization (e.g., for financial services) beyond that which is optimal, taking into account the need to sequence liberalization appropriately (see Box 1).

  • As current international agreements do not establish a legal framework of universal applicability, the liberalization obligations under these agreements risk giving rise to the discriminatory application of restrictions among Fund members.13 Such differences could lead to potential distortions in the direction of capital flows because of unequal protection, or because of the related discriminatory application of restrictions on capital outflows during a crisis. These provisions may also undermine the ability of the country in question to mobilize external financing, if the international community were to be less willing to provide assistance where a party to a bilateral or regional agreement, in order to avoid acting inconsistently with it, imposes capital restrictions only on investors of nonparties to the agreement.

34. Challenges for the Fund. The existing configuration of agreements presents particular challenges for the Fund, including through its role in providing financing to address capital account crises (e.g., these may be triggered by premature liberalization). Given its mandate to oversee the stability of the IMS, it is critical that any framework applicable to capital movements give adequate weight to the macroeconomic and balance of payments implications of such movements.

IV. Articulating an Enhanced Fund Role

35. Missing pieces. Given the critical relevance of cross-border capital flows to global and domestic macrofinancial stability, illustrated particularly vividly by the swings observed in the last few years, a better articulated role for the Fund with respect to capital flows seems desirable. As discussed further below, the Fund could clarify its role on both capital account policies (i.e., policies directed toward international capital flows) and other macroeconomic and financial policies affecting capital flows, within its current mandate:

  • Bilateral engagement: The Fund could issue principles for the guidance of members on the appropriate design of policies to deal with international capital movements.14

  • Multilateral advice: The Fund could play a greater role in assessing global liquidity, and how macroeconomic and financial sector policies in financial centers are affecting capital flows. A more symmetrical approach to dealing with capital flows—incorporating economies that originate and intermediate capital flows as well as recipient countries—requires coordination and creation of forums for discussion. Another area is in assessing cross-country spillovers from capital account policies.

  • Market reforms and data: The Fund could take a more active role in encouraging consensus on structural reforms that could reduce volatility; and in collecting, improving, and expanding the information available on capital flows, in collaboration with relevant partner institutions, in particular the FSB.

  • Scope: The term “capital controls” is used broadly and sometimes loosely to cover a very broad range of official measures affecting international capital movements. While typically applied to limits on the rights of residents or nonresidents to enter into underlying capital transactions (e.g., limits on the external debt of residents), they can also apply to the payments and transfers associated with these transactions. A key question in articulating an enhanced Fund role would be to provide clarity and a methodological framework for determining the nature and scope of the “capital controls” that are relevant for these purposes. The “restrictions” typology developed by the Fund in the 1990s reform discussion provides a useful starting point (Box 4).

36. Mandate. The proposed approach would rely on the Fund’s existing bilateral and multilateral surveillance mandate:

  • Developing an institutional view for bilateral surveillance. As already noted, a member’s right to impose capital controls under Article VI is qualified by its obligations under Article IV, including its obligation to collaborate with the Fund and other members to assure orderly exchange arrangements and promote a stable system of exchange rates (the latter being referred to in the 2007 Surveillance Decision as “systemic stability”). The Fund could rely upon this collaboration obligation as the basis for issuing principles on the appropriate design of capital account policies that would identify actions that members should take or refrain from taking in specified circumstances in order to assure orderly exchange arrangements and/or promote a stable system of exchange rates.15 Such principles could extend beyond capital account policies to cover other policies, particularly macro-prudential ones, that affect capital flows and have potential implications for the member’s own domestic and external stability (and thereby systemic stability). The principles adopted for these purposes would thus provide guidance to members on the scope of their obligations under Article IV, Section 1.

  • Multilateral surveillance. The Fund’s approach would be guided by the objective of such surveillance under Article IV, Section 3, namely to oversee the effective operation of the IMS, with the focus thus being on the multilateral implications of members’ policies in two dimensions: firstly, the impact of macroeconomic and financial policies on capital flows, and secondly the impact of their policy response to capital flows—be it in the form of capital controls or macro-prudential measures. Although, as discussed in a previous paper,16 members do not have substantive obligations (i.e., obligations regarding the conduct of economic policies) that arise from the Fund’s multilateral surveillance mandate, they may be called upon to engage in policy discussions with the Fund under multilateral surveillance, and the Fund could make recommendations that, while not providing guidance to members on the scope of their obligations under the Articles (as do the bilateral surveillance principles) could nonetheless encourage them to take certain actions. Members could also be required to provide the Fund with the information that it needs for multilateral surveillance purposes.

Staff Approach to Advice on Capital Account Policies

Recent staff advice on capital account policies has reflected member country circumstances and intentions, and has been ad hoc. Developing a more consistent approach should not relegate country circumstances as the primary factor, but act as a framework in which to develop and discuss policy advice on a continuing basis. In thinking about such guidelines, relevant themes from the economic literature include that (i) basic theory suggests liberalization should be welfare-enhancing, while the costs of capital flow volatility can be high; (ii) empirical evidence is not conclusive on whether there are gains from liberalization per se; and (iii) sequencing liberalization and structural reforms to help mitigate risks is important, and should be taken into account in giving advice on short-term responses to capital flows.

Ostry et al (2010) suggest following a structured decision-making process, recognizing country circumstances at each step, with guidelines for policy advice including:

  • Allow the real exchange rate to adjust. If the real exchange rate appears misaligned, allowing the exchange rate to adjust may be optimal. Clearly, misalignment is a matter of judgment with many uncertainties underpinning exchange rate evaluation models (see, e.g., Lee, Milesi-Ferretti, Ostry, Prati and Ricci, 2008), and this assumes exchange rate policy options are unconstrained (e.g., by concerns over fragility in the financial system, dollarization or prior commitments to stability such as those for euro entry). The pace of adjustment should be considered.

  • Reserve accumulation. If the exchange rate is judged to be close to its equilibrium, and there is a danger that further inflows could lead to real appreciation, reserve accumulation can be considered. The costs of accumulating and holding reserves should be considered—including sterilization costs if inflation is a threat, and the implications for overall international monetary stability.

  • Run a countercyclical macroeconomic policy. Capital inflows and the potential for sudden stops can give additional impetus to cyclical forces raising the benefits from tightening policy in good times, to increase policy options in bad. Macroeconomic policy may have to be more proactive where reserve accumulation or exchange rate flexibility is not appropriate. The appropriate policy mix will have to be considered given domestic conditions and the scope for higher interest rates, or better public debt dynamics, to attract further inflows.

  • Alter the pattern of inflows. Policies favoring less volatile types of capital—FDI, especially of a nonfinancial nature, has been the least risky, followed by equity investment, local currency debt, and foreign currency debt flows—may be helpful. Examples include more stringent limits or capital requirements against open foreign exchange positions by domestic banks. Broader reporting requirements on unhedged borrowing in foreign currencies could inform timely policy changes (while corporate losses resulting from derivatives exposures in Brazil, Korea, and Mexico in 2008 suggest attention to disclosure in this area is also warranted).

  • Macroprudential policies. Countercyclical prudential policy tools such as dynamic provisioning may have a role in reducing the risks from rising aggregate leverage or asset price misalignments, although practical experience is limited. Recently, some countries have adopted prudential measures designed to address specific risks that have also served to deter, or make less volatile, capital inflows (e.g., Korea has imposed restrictions on hedging contracts—funded through offshore dollar borrowing—in relation to capital, or Indonesia’s imposition of a one-month holding period on residents’ and nonresidents’ portfolio investments in government debt, and raising of reserve requirements).

  • Controls. In the event that other policy options are exhausted, controls on capital inflows may be appropriate. Realism is required in assessing what controls can feasibly achieve (they may slow inflows and divert the funds to less risky instruments) and what they probably will not (stemming currency appreciation, or averting the need for more structural reforms). One concern is that controls will be eroded over time through diversion of flows to uncontrolled instruments, implying the effectiveness of controls rises in their comprehensiveness. Putting in place the necessary infrastructure can be difficult and costly, particularly when trying to minimize distortions. Some members (e.g., E.U. members) may face other constraints on imposing controls (see Annex 2).

Considerations Related to a Typology for Capital Controls

During the 1990s reform discussions, the Fund made progress in developing an analytical framework for use in determining when a capital control would gave rise to a “restriction” for purposes of the Fund’s proposed jurisdiction. This framework drew heavily on principles developed under international agreements such as the OECD’s Code of Liberalization of Capital Movements, as well as concepts used by Fund in its Article VIII jurisprudence. In the current context of considering an enhanced Fund role over capital flows, an updated typology of this kind could be a useful device to aid the categorization and analysis of the myriad kinds of measures often grouped under the heading of “capital controls”. The following is a summary of key aspects of the 1990s methodology.

Many types of capital controls are not restrictive and, in particular, do not materially affect the allocation of resources by market participants (e.g., a requirement that capital transfers be channeled through the banking system, or approval requirements for certain kinds of investments where approval is routinely and rapidly granted once appropriate documentation is provided). Accordingly, in analyzing the economic impact of capital controls, it is important to focus on those measures that may be considered restrictive; it is also important to consider the intensity of controls used.

For underlying capital transactions, the key criterion identifying a restriction under the 1990s methodology was whether the measure discriminates between domestic and international transactions, and the difference is not justified by differences in circumstances related to international transactions.1 There would thus be a restriction if a member were treating certain capital transactions between residents and nonresidents less favorably than the same transactions between two residents, and this differential treatment was not justified by increased risks associated with the international transactions. A restriction could also arise if a particular transaction between two residents were prohibited only where a foreign capital asset is involved; or if residents were permitted to buy certain types of capital assets domestically, but prohibited from buying that same type of asset if located abroad. Conversely, even if a nonresident is prohibited from engaging in a transaction, there is no restriction so long as the same transaction is also prohibited between residents.

The 1990s methodology would have excluded most prudential measures as such measures often do not treat international transactions less favorably and, where they do, such treatment is usually justified by relevant circumstances and therefore not a restriction.

Under the 1990s methodology, three broad types of discrimination were relevant: (1) explicit discrimination (the law or regulation treats international transactions less favorably than domestic and the less favorable treatment is not justified by relevant circumstances); (2) discriminatory authority (a governmental agency or other administrator has open-ended discretion to discriminate between international and domestic transactions); and (3) implicit discrimination (the law or regulation does not distinguish between domestic and international transactions, but nevertheless has a discriminatory effect, e.g., because it is more difficult for international transactions to meet certain conditions).

For payments and transfers related to international capital transactions, the 1990s methodology envisaged that restrictions would be identified by applying the principles that have guided the Fund’s determination of whether a measure gives rise to exchange restrictions on the making of current payments and transfers, with a modification to also cover the receipt of such payments and transfers. Accordingly, there would be a restriction if there were a prohibition on making payments, or on nonresidents’ repatriation abroad of such payments. Similarly, while approval requirements are not necessarily restrictive, a restriction would result if these measures were to give rise to undue delays in the ability to make the payment or transfer.

1A transaction is international if it is (i) between a resident and a nonresident, (ii) between two residents involving a foreign capital asset, or (iii) between two nonresidents involving a domestic capital asset.

A. Bilateral Surveillance

37. Institutional view. A premise for the envisaged Fund role under bilateral surveillance is that the Fund would have an institutional view on (i) what conditions should be in place before a member liberalizes its capital account; (ii) when the imposition of capital controls (whether on inflows or outflows) or adaptation of existing restrictions may be an appropriate response to balance of payments or macroeconomic pressures; and (iii) what are the most effective ways of designing such measures. A clear view on how domestic macroeconomic and financial policies can affect capital flows and when this might have implications for external and systemic stability could similarly prove useful. Country-specific factors would remain of primary importance, while global financial conditions and structural factors behind flows as well as the potential impact on other countries would also be considered. Steps to articulate this view are discussed below.

38. Board views. The summings up from the current paper as well as from the upcoming Board discussion on Cross-Cutting Themes from Recent Country Experiences with Capital Flows will provide an opportunity to articulate and disseminate a Fund view on capital flows.

39. Surveillance Principles. Beyond these first steps, and more pointedly, the Fund (i.e., the Executive Board) could adopt “rules of the road” on international capital flows. These could take the form of principles under its bilateral surveillance authority. The principles would be used, on a systematic basis, to assess members’ capital account policies in the context of Article IV consultations. These principles would provide guidance to members on the scope of their obligations under Article IV, Section 1 as noted earlier, but a member’s failure to adhere to the principles would not in itself give rise to a breach of obligations under that provision or create any presumption of such a breach. Rather, the principles envisaged would be in the nature of recommendations. Further—and as a consequence of members’ right to regulate international capital movements under Article VI, Section 3—the Fund could not adopt a subsequent decision requiring the imposition or removal of a control unless it could be demonstrated that this was the only available means by which the member could adhere to its obligations under Article IV. Such an approach is necessary in order to ensure that Article VI, Section 3 continues to be given substantive effect, notwithstanding its qualification by the obligations under Article IV.

B. Multilateral Level

40. Sharing experiences. Collating information on country experiences with managing capital flows, whether through controls or macro-prudential measures, could be valuable to the membership. Awareness of cross-country experience could help countries contemplating adopting measures, particularly in alerting authorities to differences in conditions that may result in similar measures having dissimilar results.

41. Externalities. At the multilateral level, the Fund could provide analysis and advice on the impact of domestic policies on outward capital flows, particularly to the extent they materially influence global liquidity. Spillover reports are a natural venue for considering the impact of economic and financial policies of systemically important members on others through capital flows. Assessing the impact of capital account policies on flows to other countries warrants further research. In addition, the Fund could monitor the aggregate effects of divergent national macroprudential policies.

42. Global liquidity. There are substantial difficulties in assessing whether global liquidity is adequate and appropriately distributed. A first step for the Fund could be to develop a research agenda on measuring global liquidity and assessing its adequacy (including exploring existing data such as changes in the IIP, or through proposing new data if needed), and assessing the role of interest rates and monetary policy in core financial markets. Developments could be tracked systematically and potential risks communicated through existing multilateral surveillance products. Combining a greater understanding of global liquidity conditions and drivers with the identification of transmission channels for liquidity shocks stemming from work on financial interconnectedness would provide a powerful new lens for financial surveillance, both for individual countries and multilaterally.

43. Multilateral dialogue and coordination. Different forms of multilateral coordination of policies affecting capital flows could be envisaged. Beyond Board papers, enhanced collaboration with the FSB could be useful particularly given overlaps between macroprudential and capital account policies, as could multilateral consultations or “working groups” organized by the Fund and including country representatives, in particular:

  • Between recipients and large financial centers that are the source (or intermediaries) of flows, to consider questions such as using macroprudential regulation in both source and recipient countries to support financial stability in recipient countries or other modalities of home-host collaboration (see Box 5 for examples). A clear view on how domestic macroeconomic and financial policies can affect capital flows and when this might have implications for systemic stability could similarly prove useful.

  • Among recipient countries dealing with volatility, or episodes of large inflows. The effectiveness of policy measures may be enhanced by reducing the potential for diversion of capital flows to others. More fundamental changes in investor behavior may be induced by altering incentives across a broader range of destinations and instruments than those of one country. The Fund could catalyze regional—or more broadly defined—groupings for cooperation.

  • Among source countries, to consider microprudential policies that could address some underlying causes of volatility or spillovers from macro or prudential policies.

44. Structural reforms. The Fund could more systematically bring the perspective of global financial stability—as opposed to institutional soundness or national systemic stability—to discussions of financial sector regulatory reform (Schinasi and Truman, 2010, give some specific examples of where such an approach would add value). It could leverage cross-country experience to propose measures—both country-based and multilateral—that would promote financial deepening in emerging markets, thereby helping them deal with the effects of size and volatility of capital flows. The Fund could also seek to foster a global consensus on structural reforms that could reduce volatility—similar to the consensus that eventually emerged for the inclusion of collective action clauses in sovereign bonds. Examples include:

  • Creating legal space for the reimposition of controls under other agreements: More systematic inclusion in bilateral and regional investment and free trade agreements of “safeguards” clauses, giving the right to temporarily impose restrictions on capital flows (outflows or inflows), where necessary for balance of payments or financial stability purposes (see discussion in Section III and Annex 2) would strengthen the overall framework for managing capital flows. In designing these clauses, careful consideration would be needed to balance the desire to provide adequate policy space to manage volatility and other vulnerabilities, while not undermining the liberalization obligations voluntarily assumed under these international agreements. Subject to Directors’ views, staff would propose to undertake further work on this issue, including consultations with the United Nations Conference on Trade and Development (UNCTAD) and other multilateral and regional bodies involved in the design and promotion of the international frameworks in this area.

  • Institutional reforms in capital markets: Reducing the uncertainty associated with debt repayments in cases of crisis or distress could lower the incentives for “panic selling” when fears of default rise. For example, GDP-linked sovereign bonds explicitly linking repayment burdens with capacity to pay, could serve this purpose. Strengthening regulatory structures in source capital markets could help counteract incentives that contribute to herd behavior and regulatory arbitrage, and market practices conducive to contagion.

    Multilateral Collaboration on Macroprudential Issues

    Collaboration between source (or intermediaries) and recipients of capital flows could prove useful. Some examples of areas where Fund-sponsored working groups could collaborate include:

    • For crisis management purposes. Voluntary “private sector involvement” already occurs on a case-by-case basis, with the European Bank Coordination Initiative (the “Vienna initiative”) being a recent, successful, case in point. In this case, parent banks with a large presence in several eastern and central European countries with Fund arrangements entered agreements on maintaining exposures to these countries in the crisis, with generally good compliance. Previous experiences include efforts to coordinate bank rollovers of short-term debt to Korea in 1997/98. Finding an external coordinating agent in a crisis can be critical to avoid banks withdrawing in the anticipation that others will withdraw first, depleting available international reserves, and can thus play an important role in stabilizing outflows. Incentives to participate in such initiatives vary depending on exposures and the particular circumstances of the crisis; however, there may be room to explore more automatic procedures for coordinating relevant parties. A good ex ante understanding of interconnectedness—in which the Fund’s surveillance could play an important role—would help ensure timely intervention, even when, in the absence of a Fund arrangement say, other bodies are more natural coordinators.

    • Macroprudential policies. Tackling the accumulation of systemic risk through procyclical expansion of bank balance sheets could be better achieved with multilateral measures. Shin (2010) proposes a tax on “noncore” liabilities of banks, to better align bank incentives with the broader social objective of low systemic risk (see paragraph 8).17 Where wholesale funding markets are thin—as in most emerging markets—this proposal would generally imply taxing bank borrowing from nonresidents. Global application would reduce scope for circumvention (although such a tax could be implemented effectively at national level). Another example could be imposing a graduated scale of maximum loan-to-value ratios for real estate lending across jurisdictions, depending on measured real estate price increases in preceding periods.

    Enhancing the Fund’s Role—Initiatives for Consideration

    The Fund could consider the following steps:

    • Articulate an institutional view on capital account policies and other policies affecting capital flows.

    • Adopt surveillance principles for the guidance of members on the appropriate design of capital account policies; principles could also cover other policies affecting capital flows that have external and systemic stability implications.

    • Foster multilateral policy coordination on all policies influencing capital flows including through:

      • Multilateral dialogue or working groups (i) between recipients and economies originating or intermediating the flows, particularly in the area of macro-prudential regulation; and (ii) among recipients to identify side-effects on others and maximize effectiveness.

      • Surveillance of global liquidity through (i) research on measuring and assessing global liquidity, including by developing new data, if necessary; (ii) tracking developments systematically and communicating potential risks through multilateral surveillance products; and (iii) combining insights and tools from work on financial interconnectedness to track shock transmission and risk concentrations.

      • Assessing externalities from systemic countries’ policies on others, e.g., through spillover reports, and developing research on the impact of capital account policies on flows to other countries.

      • Collating country experiences in managing capital flows as a resource for policymakers and to inform bilateral policy discussions.

    • Improve data. Address gaps through participation in existing and consideration of new data initiatives.

    • Propose structural reform to build consensus on (i) more systematic inclusion of safeguards allowing the use of capital controls for balance of payments and financial stability purposes in bilateral and regional investment and trade agreements; (ii) institutional and regulatory reforms in capital markets to reduce volatility (e.g., GDP-linked debt); (iii) measures to encourage less volatile and risky flows (e.g., adoption of more uniform accounting standards and contract enforcement, and reduction of incentives favoring debt over equity flows).

  • Affecting the composition of flows: Progress on international standards in accounting or contract enforcement could enable better oversight of investments by nonresidents, encouraging a shift in composition of capital flows toward less volatile FDI or long-term debt. Progress on phasing out restrictions or incentives favoring more volatile sources of funding (e.g., tax rules favoring debt over equity) could work similarly.

45. Data. Some of the data gaps discussed above are being addressed, at least partially, through existing initiatives—for example through the joint IMF-Financial Stability Board (FSB) work requested by the G-20 on identifying data gaps that masked key vulnerabilities in the run-up to the financial crisis. Nonetheless, and while the costs and benefits of new data requests must be considered in the light of competing objectives, incomplete coverage remains an issue (see Annex 1).

V. A Potential Longer-Term Mandate Reform

46. Amendment of the Articles? A final question is whether, as a longer-term reform, consideration should be given to amending the Articles of Agreement to address more directly the systemic stability responsibilities of members and the Fund in respect of international capital flows. The central surveillance approaches proposed in this paper would be integrated into the Fund’s current bilateral and multilateral surveillance mandate, thereby avoiding the need for an amendment. But while this approach would represent a significant improvement over the current Fund role as regards capital flows, and would also enable the Fund to begin expeditiously to implement its enhanced role, it also has its limitations.

47. Increased effectiveness. An amendment would provide a more complete framework within which to address the complex issues related to international capital flows. As noted earlier, while members’ right under Article VI is qualified by their systemic stability obligations under Article IV, it would not be possible in light of Article VI, Section 3 to require a member to impose or remove a control unless the Fund could demonstrate that this action was the only feasible means by which the member could adhere to its obligations under Article IV, which will rarely be the case. Accordingly, if the desire is to establish a comprehensive legal framework to govern members’ reliance on capital controls, then it would be necessary to modify Article VI, Section 3.

48. Multilateral framework. An additional advantage of an amendment is that it could enhance the status of the Fund’s assessments on capital flow issues vis-à-vis other international institutions and frameworks, thereby facilitating a hierarchy among the current disparate arrangements in this area and ultimately strengthening the multilateral approach to capital flows. The experience under the GATS provides a useful illustration in this respect. Under the GATS, the Fund’s approval of controls on current payments and transfers under Article VIII, Section 2(a) has the automatic effect of ensuring that these controls are consistent with the member’s obligations under the GATS. In the case of capital controls, however, while the Fund provides some input, its absence of jurisdiction in the area of capital gives it a more limited role, with the WTO making its own determination regarding the acceptability of capital measures. The degree to which other international agreements will in the future defer to the Fund can obviously only be determined by those agreements, but it is very likely that the Fund’s views would be given greater weight if the Articles were to be amended to give the Fund jurisdiction over capital controls.18

49. A liberalization amendment option. There are various ways in which such an amendment could be designed. A liberalization-centric approach would follow the route envisaged during the 1990s: the Fund’s Articles would be amended to establish specific obligations for members to liberalize capital movements, subject to strong safeguards that would allow for the phased elimination of restrictions, the routine exclusion or approval of measures imposed for prudential reasons, and the temporary approval of restrictions imposed for balance of payments and macroeconomic stability purposes. The difficulty with this approach is that, notwithstanding the breadth of the safeguards that would be expected to be put in place, it signals a presumption regarding the benefits of liberalization that is still widely debated. Such an approach would also not give the Fund the authority to require a member to maintain or impose controls in circumstances when such controls are deemed necessary for systemic stability.

50. A more neutral option. Given these considerations, an alternative approach would be to replace the existing text of Article VI, Section 3, with a provision (similar to the chapeau of Article IV) that requires members to collaborate with the Fund and other members in seeking to ensure that international capital movements support both sustainable economic growth and the stability of the IMS. This approach would reflect a more neutral stance regarding capital flows, as it would give the Fund the authority to require a member to either eliminate or impose restrictions, depending on what actions were considered appropriate in the circumstances of the member, taking into account the objectives of both securing sustainable growth and the stability of the IMS.

VI. Issues for Discussion

  • Do Executive Directors support an enhanced Fund role on cross-border capital flows and capital account policies, including an articulation of an institutional Fund view?

  • Do Directors support, as a central feature of this enhanced role, developing formal bilateral surveillance principles to provide guidance to members on their capital account policies?

  • Among the various multilateral missions and analytical areas discussed in the paper that the Fund could take up in that context (see Box 6), which ones do Directors consider as priorities?

  • As a longer-term initiative, should an amendment of the Articles of Agreement be considered to provide a more complete framework for addressing the complex issues in this area? If so, which of the two options presented would Directors prefer at this time?


  • Adrian, T., and H.S. Shin, 2010, “Liquidity and Leverage,” Journal of Financial Intermediation, Vol. 19, No. 3, pp 418437 (July).

  • Arteta, C., B. Eichengreen, and C. Wyplosz, 2001, “When Does Capital Account Liberalization Help More Than It Hurts?,” NBER Working Paper No. 8414 (August).

    • Search Google Scholar
    • Export Citation
  • Bakker, A. and B. Chapple, 2002, “Advanced Country Experiences with Capital Account Liberalization,” IMF Occasional Paper 214, Washington: International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Bhagwati, J., 1998, “The Capital Myth: The Difference Between Trade in Widgets and Dollars,” Foreign Affairs, Vol. 7, pp. 712 (May/June).

    • Search Google Scholar
    • Export Citation
  • Bank for International Settlements, 2007, “Institutional Investors, Global Savings and Asset Allocation,” Committee on the Global Financial System Paper, No. 27 (February).

    • Search Google Scholar
    • Export Citation
  • Bank for International Settlements, 2010, “Funding Patterns and Liquidity Management of Internationally Active Banks,” Committee on the Global Financial System Paper, No. 39 (May).

    • Search Google Scholar
    • Export Citation
  • Calvo, G. A., 1998, “Capital Flows and Capital-Market Crises: The Simple Economics of Sudden Stops,” Journal of Applied Economics, Vol., 1 No. 1, pp. 3554 (November).

    • Search Google Scholar
    • Export Citation
  • Calvo, G. A., 2005, Emerging Capital Markets in Turmoil, Cambridge, Massachusetts: MIT Press.

  • Calvo, G. A. and E. G. Mendoza, 2000, “Rational Contagion and the Globalization of Securities Markets,” Journal of International Economics 51, pp 79113.

    • Search Google Scholar
    • Export Citation
  • Cetorelli, N. and L. Goldberg, 2010, “Global Banks and International Shock Transmission: Evidence from the Crisis,” NBER Working Paper, No. 15974 (May).

    • Search Google Scholar
    • Export Citation
  • Cecchetti, S., I. Fender and P. McGuire, 2010, “Toward a Global Risk Map,” BIS Working Paper, No. 309 (May).

  • Chinn, M.D. and H. Ito, 2008, “A New Measure of Financial Openness,” Journal of Comparative Policy Analysis: Research and Practice, Vol. 10, No. 3, pp. 309322 (September).

    • Search Google Scholar
    • Export Citation
  • Claessens, S., G. Dell’Ariccia, D. Igan, and L. Laeven, 2010, “Lessons and Policy Implications from the Global Financial Crisis,” IMF Working Paper 10/44 (February).

    • Search Google Scholar
    • Export Citation
  • Dell’Ariccia, G., J. di Giovanni, A. Faria, A. Kose, P. Mauro, J. D. Ostry, M. Schindler and M. Terrones, 2008, “Reaping the Benefits of Financial Globalization,” IMF Occasional Paper 264, Washington: International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Diaz-Alejandro, C., 1985, “Good-bye Financial Repression, Hello Financial Crash” Journal of Development Economics, Vol. 19, pp 124 (September).

    • Search Google Scholar
    • Export Citation
  • Edwards, S., 2001, “Capital Mobility and Economic Performance: Are Emerging Economies Different?,” NBER Working Paper, No. 8076 (January).

    • Search Google Scholar
    • Export Citation
  • Eichengreen, B., 2009, “The Financial Crisis and Global Policy Reforms,” paper given at the Asia Economic Policy Conference, Federal Reserve Bank of San Francisco (October).

    • Search Google Scholar
    • Export Citation
  • Felices, G., B. E. Orskaug, 2008, “Estimating the Determinants of Capital Flows to Emerging Market Economies: A Maximum Likelihood Disequilibrium Approach,” Bank of England Working Paper, No.354 (November).

    • Search Google Scholar
    • Export Citation
  • Frankel, J. A., 2010, “Monetary Policy in Emerging Markets: A Survey,” NBER Working Paper, No. 16125 (June).

  • Gochoco-Bautista, M. S., 2010, “To What Extent Should Capital Flows Be Regulated?” in J. D. Sachs, M. Kawai, J. Lee, and W. T. Woo (eds.), The Future Global Reserve System—An Asian Perspective, ADB, Manila: Asian Development Bank. Paper available from:

    • Search Google Scholar
    • Export Citation
  • Henry, P. B., 2007, “Capital Account Liberalization: Theory, Evidence and Speculation,” Journal of Economic Literature, Vol. XLV, pp. 887-935 (December).

    • Search Google Scholar
    • Export Citation
  • IMF Staff and FSB Secretariat, 2009, “The Financial Crisis and Information Gaps, Report to the G-20 Finance Ministers and Central Bank Governors” (October).

    • Search Google Scholar
    • Export Citation
  • IMF, 2010a, Global Financial Stability Report, Washington D.C. (April).

  • IMF, 2010b, Global Financial Stability Report, Washington D.C. (October).

  • IMF, 2010c, A Fair and Substantial Contribution by the Financial Sector, Final Report for the G-20 (June).

  • IMF, forthcoming, “Triennial Report on Exchange Arrangements.”

  • Independent Evaluation Office, 2005, “Report on the Evaluations of the IMF’s Approach to Capital Account Liberalization” (May).

  • Institutional Investor, 2010, “Emerging Markets Go From Niche to Mainstream,” Institutional Investor Magazine (June) issue. Article available at

    • Search Google Scholar
    • Export Citation
  • Ishii, S. and K. Habermeier, 2002, “Capital Account Liberalization and Financial Sector Stability,” IMF Occasional Paper 211, Washington: International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Korinek, A., 2010a, “Regulating Capital Flows to Emerging Markets: An Externality View” (May). Available at:

    • Search Google Scholar
    • Export Citation
  • Korinek, A., 2010b, “Capital Flows, Crises and Externalities: A Primer” (June). Available at:

  • Kose, A., E. Prasad, K. Rogoff, and S. Wei, 2006, “Financial Globalization: A Reappraisal,” IMF Working Paper 06/189 (August).

  • Lane, P.R. and G.M. Milesi-Ferretti, 2006, “The External Wealth of Nations Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970-2004,” IMF Working Paper 06/69 (March).

    • Search Google Scholar
    • Export Citation
  • Lee, J., G. M. Milesi-Ferretti, J. D. Ostry, A. Prati, and L. Ricci, 2008, Exchange Rate Assessments: CGER Methodologies, IMF Occasional Paper, No. 261, Washington: International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Mcguire, P. and G. von Peter, 2009, “The U.S. dollar Shortage in Global Banking,” BIS Quarterly Review (March).

  • Mody, A., M. P. Taylor and J. Y. Kim, 2001, “Modeling Fundamentals for Forecasting Capital Flows to Emerging Markets,” International Journal of Finance and Economics, No. 6, pp 201216.

    • Search Google Scholar
    • Export Citation
  • Morris, S. and H. S. Shin, 2004, “Liquidity Black Holes,” Review of Finance No. 8, pp 118.

  • Obstfeld, M., 2009, “Lenders of Last Resort in a Globalized World”; speech given at the conference on “Financial System and Monetary Policy Implementation,” Institute for Monetary and Economic Studies, Bank of Japan (May).

    • Search Google Scholar
    • Export Citation
  • Obstfeld, M. and K. Rogoff, 2009, “Global Imbalances and the Financial Crisis: Products of Common Causes,” paper prepared for the Federal Reserve Bank of San Francisco Asia Economic Policy Conference, Santa Barbara, CA (October 18–20).

    • Search Google Scholar
    • Export Citation
  • Ostry, J. D., A. Ghosh, K. Habermeier, M. Chamon, M. S. Qureshi and D. B. S. Reinhardt, 2010, “Capital Inflows: The Role of Controls,” IMF Staff Position Note 10/04 (February).

    • Search Google Scholar
    • Export Citation
  • Persaud, A., 2003, “Liquidity Black Holes: What Are They and How are they Generated?,” published in the Singapore Foreign Exchange Market Committee Biennial Report, 2001–2002, available at:

    • Search Google Scholar
    • Export Citation
  • Prasad, E., R. Rajan and A. Subramaniam, 2001, “Foreign Capital and Economic Growth,” Brookings Papers on Economic Activity, Vol. 1:2007, pp. 153209.

    • Search Google Scholar
    • Export Citation
  • Reinhart, C.M. and K. Rogoff, 2009, This Time is Different, Princeton, New Jersey: Princeton University Press.

  • Rodrik, D., 1998, “Who Needs Capital-Account Convertibility?” in Peter Kenen (ed), Should the IMF Pursue Capital Account Convertibility?, Essays in International Finance, No. 207 (May).

    • Search Google Scholar
    • Export Citation
  • Schinasi, G.J. and E.M. Truman, 2010, “Reform of the Global Financial Architecture,” Peterson Institute for International Economics Working Paper WP 10-14 (September).

    • Search Google Scholar
    • Export Citation
  • Schindler, M., 2009, “Measuring Financial Integration—A New Dataset,” IMF Staff Papers, Vol. 56 No. 1, pp. 222-238.

  • Shin, H. S., 2010, “Noncore Liabilities Tax as a Tool for Prudential Regulation” (February). Available at:

    • Search Google Scholar
    • Export Citation
  • Shin, H. S. and K. Shin, 2010, “Procyclicality and Monetary Aggregates” (August). Available at:

    • Search Google Scholar
    • Export Citation
  • Sørensen, B.E., Y.T. Wu, O. Yosha, Y. Zhu, 2007, “Home Bias and International Risk Sharing: Twin Puzzles Separated at Birth”, Journal of International Money and Finance, 26 (2007), pp. 587605.

    • Search Google Scholar
    • Export Citation
  • Truman, Edwin, 2010, “Dealing with Volatile Capital Flows: The Case for Collective Action” (May). Available at:

    • Search Google Scholar
    • Export Citation

Annex 1. Data Initiatives Under Way and Remaining Gaps

Initiatives include joint IMF-FSB work on data gaps masking vulnerabilities in the run-up to the financial crisis. In April 2009 the Group of 20 economies (G-20) requested staff of the International Monetary Fund (IMF) and the Financial Stability Board (FSB) Secretariat to explore information gaps, and provide appropriate proposals for strengthening data collection. In response, the staff of the IMF and the FSB secretariat presented a report on “The Financial Crisis and Information Gaps” at the meeting of the G-20 finance ministers and central bank governors in November 2009 outlining twenty recommendations for closing information gaps. Gaps identified include on (i) the build-up of risks in the financial sector; (ii) cross-border financial network exposures; and (iii) vulnerabilities facing domestic sectors and markets. The report also describes information on linkages between individual financial institutions as of critical importance to global financial stability. An action plan for closing the gaps, endorsed by G-20 Finance Ministers and Central Bank Governors, was provided in May 2010. The following actions have been taken:

  • The Principal Global Indicators (PGI) website was developed by the Inter-Agency Group on Economic and Financial Statistics to provide more timely, higher frequency real and financial indicators from G-20 economies. At present, the website contains external debt, IIP, and other data for the G-20 economies.

  • In March 2010, the Executive Board endorsed proposals to modify the Fund’s Special Data Dissemination Standard (SDDS). The SDDS will include IIP data with quarterly periodicity on a mandatory basis, from 2014. Also, selected financial soundness indicators (regulatory Tier 1 capital to risk-weighted assets and to total assets, nonperforming loans net of provisions to capital, nonperforming loans to total gross loans, return on assets, liquid assets to short-term liabilities and net open position in foreign exchange to capital) and external debt on a remaining maturity basis have been included in the SDDS on an “encouraged” (i.e., not mandatory) basis.

  • The Balance of Payments and International Investment Position Manual, sixth edition (BPM6), which member countries are implementing over the coming years, includes as enhancements the identification of cross border financial activities of nonbank financial institutions, the currency composition of assets and liabilities, and, on an encouraged basis, the remaining maturity of debt. Also, BPM6 recognizes that gross flows are analytically useful and recommends that data on drawings and repayments on loans, or acquisitions or disposals of other instruments, be made available on a supplementary basis (see paragraph 8.9 of the BPM6).

  • Internationally coordinated exercises such as the Coordinated Portfolio Investment Survey (CPIS), the forthcoming Coordinated Direct Investment Survey (CDIS), and the International Banking Statistics (IBS) collected by the Bank for International Settlements (BIS) provide data to support analysis of cross-border flows (e.g., year-end holdings of portfolio investment securities). Currently 74 Fund members participate in the CPIS (though more have expressed interest), while the IBS covers 43 countries under the “locational reporting system” and 30 countries under the “consolidated banking statistics” report. The latter also reflects only information provided by regulated entities to participating regulators. The G-20 data gaps initiative includes recommendations to enhance both the CPIS and the IBS.

  • The FSB has set up a working group to develop a common template to capture and better understand the interconnections of systemically important global financial institutions with national financial sectors and markets. This work is also part of the G-20 data gaps initiative.

  • The IMF, BIS, and others have undertaken work related to better understand drivers of global liquidity positions, as well as linkages between financial participants (BIS, 2010 and Cecchetti et al, 2010).

Gaps remain to the extent that the voluntary nature of many existing efforts results in some significant flows going unreported, or reported with long delay: for instance, not all G-20 members participate in the CPIS and coverage of total reserves in COFER has been declining for years. Also, nonbank financial institutions such as mutual funds play a critical role in cross-border financial flows, and significant gaps exist. Consolidated data for some critical financial institutions operating in systemic financial centers could provide key insights into important capital movements and possible vulnerabilities. In working to address some of these gaps, and in addition to work with the FSB, IMF staff is collaborating with national authorities and international bodies though the Inter-Agency Group on Economic and Financial Statistics (which includes representatives from the BIS, European Central Bank, Eurostat, IMF (Chair), OECD, United Nations, and the World Bank). Costs and benefits of new data requests must be considered in the light of competing objectives and constrained resources.

Annex 2: Overview of Current International Agreements1

This Annex provides an overview of existing international agreements that promote the liberalization of capital movements. The Annex broadly examines: (i) the scope of the capital transactions that are covered under these agreements, (ii) the extent to which the agreements cover both inward and outward cross-border capital movements, and (iii) the extent to which liberalization obligations under these agreements allow flexibility in respect of international capital movements (e.g., exceptions for balance of payments difficulties). The agreements covered under the Annex range from bilateral and regional agreements with more limited mandates (inward, but not outward investments) to the Treaty on the Functioning of the European Union, which establishes the most far-reaching obligations of the agreements summarized in this Annex.

Bilateral and regional foreign investment and free trade agreements

There are a large number of bilateral and regional investment agreements and free trade agreements with a focus on the promotion and legal protection of foreign investment. These include over 2500 bilateral investment treaties (BIT),2 additional regional investment agreements, and a number of bilateral and regional free trade agreements (FTA) with foreign investment chapters (e.g., NAFTA, CAFTA).3 These agreements typically serve to liberalize inward, but not outward, investments and as such normally specify obligations for signatories with respect to the investors of other signatories, but not regarding the signatory’s treatment of its own investors (e.g., a country could restrict the ability of its own residents to invest overseas, even while committing itself to allowing foreign investors to make investments in its domestic markets). The type of investments covered is often (but not always) broad, including direct investment, portfolio investment and debt instruments, including sovereign debt in some cases.

These agreements often establish comprehensive post-investment obligations, including for the fair and equitable treatment of foreign investors and the free repatriation of amounts invested. The main substantive obligations are those of “national treatment” (foreign investors must be treated no less favorably than a host country’s own investors) and “most-favored nation treatment” (host country must treat the foreign investors of a signatory no less favorably than the foreign investors of other countries).

While regional agreements generally have more expansive provisions in this respect, only a small proportion of BITs and bilateral FTAs contain exceptions (sometimes referred to as “derogations”) that excuse a signatory from the performance of their obligations for balance of payments purposes. To the extent that such balance of payments exceptions are provided for, their scope varies. Some regional agreements such as the ASEAN Comprehensive Investment Agreement and NAFTA contain specific balance of payments derogation provisions with links to the country’s obligations under the Fund’s Articles.4 On the bilateral level, a small proportion of BITs and FTAs either allow for a temporary safeguard on capital inflows and outflows to prevent or mitigate a financial crisis or completely “carve out” host country legislation on capital controls (therefore deferring to such legislation); other BITs simply provide that transfers may be restricted for balance of payments purposes.

Agreements regarding the liberalization of services

The GATS, the Financial Services Chapter of the NAFTA and other recent FTAs governing trade in services help to liberalize both inward and outward cross-border movement of capital.5 First, by setting forth obligations regarding the ability of foreign service suppliers to provide services through the establishment of a commercial presence in the territory of another signatory, they liberalize the admission of direct investment (inward capital transfers relating to the supply of the service committed). Second, because the obligations also apply to the cross-border supply of services (i.e., where the foreign provider of a service operates from his own country rather than through the establishment of a local commercial presence in the territory of another signatory), the provision of financial services may, in itself, involve both inward and outward capital movements (e.g., bank located in the territory of one signatory would be able to lend to a resident of another signatory, and would also be able to receive deposits from the same resident).

The liberalizing impact of these agreements is limited in certain respects. First, where the obligations only apply to capital movements associated with the provision of services, they are, as such, somewhat limited in scope. Second, to the extent that the obligations only apply to foreign services suppliers, a signatory is not required to remove restrictions on the ability of its own service suppliers to provide services abroad. Third, by virtue of the legal framework of the GATS, the most important obligations are not of a general nature (WTO members only incur obligations to remove restrictions if they have made specific commitments in the service sector in question, and these commitments are subject to periodic rounds of negotiation and may be of a qualified or unqualified nature).6 Fourth, prudential carve-outs from financial services obligations are common in many of these agreements and can limit their liberalizing impact.7

Both the GATS and the Financial Services Chapter of the NAFTA have detailed balance of payments provisions that provide a role for the Fund.8 With respect to the GATS, the following provisions are of relevance:

  • Article XI of the GATS generally prohibits GATS signatories from applying restrictions on international transfers and payments for current transactions relating to their specific commitments. Article XI provides, however, that nothing in the GATS will “affect the rights and obligations” of Fund members under the Articles, including the use of exchange actions which are “in conformity” with the Articles of Agreement. Deference to members’ rights and obligations under the Articles is qualified by an important proviso: Article XI provides that GATS signatories may not impose restrictions on any capital transactions inconsistently with their specific commitments regarding such transactions, except under Article XII of the GATS or “at the request of the Fund”.

  • Article XII of the GATS sets forth a number of conditions under which a signatory can impose restrictions in the “event of serious balance-of-payments and external financial difficulties or threat thereof’. One of the conditions set forth in this Article is that the restrictions “be consistent with” the Fund’s Articles.

The relationship between Article Xl and XII of the GATS and the current Articles has been described extensively in previous staff papers.9 The essential elements can be summarized as follows:

  • Regarding restrictions on current payments and transfers, Articles XI ensures that the exercise by a member of its rights under the Articles to impose or maintain such restrictions (either because they have been approved for temporary balance of payments reasons or because their maintenance is authorized under the Fund’s transitional provisions) will not give rise to a breach of a signatory’s obligations under the GATS. Conversely, the GATS is precluded from permitting a signatory to impose a restriction on a current payment relating to a commitment under the GATS if the restriction is not consistent with the Articles because, for example, it has not been approved by the Fund.

  • Regarding restrictions on capital movements, as a general rule, these restrictions will only be permitted if they meet all of the conditions set forth in Article XII of the GATS. The exception to this general rule would be restrictions imposed “at the request” of the Fund. The only provision of the existing Articles that explicitly gives the Fund the authority to request the imposition of restrictions is Article VI. Section I, which authorizes the Fund to request their imposition as a means of protecting the Fund’s resources from being used to finance large or sustained capital outflows. These restrictions do not need to meet the additional conditions set forth in Article XII.

OECD Code of Liberalization of Capital Movements (the “Capital Code”)

The Capital Code is the only legally binding instrument in existence that focuses comprehensively and exclusively on international capital movements. It is comprehensive in a number of respects. First, it serves to liberalize the making of both inward and outward investments, as it establishes obligations not only with respect to the ability of nonresidents to engage in capital transactions in a local market (including, for example, the making of an investment or issuance of securities), but also with respect to the ability of residents to engage in such capital transactions abroad. Second, it covers almost all types of capital transactions. Third, the obligations apply not only to transactions but also to the payments and transfers associated with these transactions. The Capital Code’s focus on international transactions is also exclusive, however, in that it does not establish obligations with respect to transactions between residents, including post-establishment obligations. Thus, while OECD members would be obligated under the Code to allow a nonresident to obtain a controlling interest in a local enterprise, the Code (unlike bilateral investment treaties, for example) does not establish obligations with respect to how the host country is to regulate the activities of these enterprises.

The Capital Code has limited authority for members to lodge reservations with respect to specific transactions. With respect to a number of financial transactions considered short-term in nature, such reservations can be introduced at any time. Regarding all other transfers, however, reservations may normally only be made at the time a country adheres to the Capital Code.

The Code also provides a very broad level of temporary derogation for economic purposes. Specifically, it allows members to derogate temporarily from their obligations and impose capital controls on outflows for balance of payments reasons, and also allows them to impose controls on inflows for reasons arising from “serious economic and financial disturbances”. The Code, however, does not have a provision similar to GATS Article XI that specifically defers to Fund members’ rights and obligations under the Fund’s Articles.10

The Treaty on the Functioning of the European Union

The Treaty on the Functioning of the European Union (“Treaty”) generally prohibits all restrictions on the movement of capital, both between EU member states and between those member states and third countries, but it also allows for derogations from this obligation. An important distinction is made in this context between those in the euro area and those outside of it. Member states outside the euro area may be authorized by the EU to take protective measures when faced with balance of payments difficulties (which could include restrictions on capital movements) or may adopt such measures unilaterally when confronted with a sudden crisis in the balance of payments, provided that the European Council may later decide that such a measure is to be amended, suspended or abolished. In contrast, members who have adopted the euro have no express authority for such unilateral protective measures. However, when in exceptional circumstances, movements of capital to or from third countries cause, or threaten to cause, serious difficulties for the operation of economic and monetary union, the EU (as opposed to individual states) may under certain limitations take safeguard measures with regard to those third countries.

There are other qualifications to members’ capital liberalization obligations under the Treaty. Specifically, member states may take measures to prevent infringements of tax laws and prudential regulations, or measures that are justified on grounds of public policy or public security, provided these measures do not constitute a means of arbitrary discrimination or a disguised restriction on the free movement of capital. More generally, the EU (but not individual member states) may regulate capital flows to and from third countries with respect to direct investment (including acquisition of real estate), establishment, financial services or the admission of securities to capital markets.

The Treaty’s general prohibition on capital controls and the relevant qualification thereto should be seen in the context a wider system of economic and legal integration. Free movement of capital is supported by the EU-wide harmonization of regulatory frameworks and by financing arrangements that support capital inflows in qualifying member states. Furthermore, the Treaty’s enforcement mechanism is also noteworthy. Specifically, the Treaty’s freedom of capital movement requirements is part of a comprehensive body of law that is ultimately enforced by an independent judiciary (the European Court of Justice).

Annex 3. Article VI, Sections 1 and 3: Key Legal Aspects1, 2

1. The general scope of Article VI, Section 3. The Fund’s jurisdiction to ensure that members liberalize the payments and transfers associated with current international transactions does not extend to capital movements. To the contrary, members have broad authority to limit capital flows, as Article VI, Section 3 gives them the right “to exercise such controls as are necessary to regulate international capital movements,” so long as these controls do not restrict payments for current transactions or unduly delay transfers of funds in settlements of commitments. The right under Article VI, Section 3 to exercise controls on capital movements includes controls on both inward and outward capital movements. More generally, the Fund has deferred to members’ discretion for purposes of determining whether controls imposed are “necessary” to regulate international capital movements. Article VI, Section 3 thus stands in significant contrast to Article VIII, Section 2(a), under which members may not, without the consent of the Fund, impose restrictions on the making of payments and transfers for current international transactions.

2. What is “capital”? While the Articles do not define what constitutes “capital” for purposes of Article VI, Section 3, the Fund’s jurisdiction under Article VIII already extends, to a limited extent, to the capital account. The Articles define the “payments for current transactions” that are subject to the Fund’s jurisdiction under Article VIII, as “payments which are not for the purpose of transferring capital” (Article XXX(d)), and on this basis the Fund has treated as “capital” any payment or transfer that does not fall within the definition set out in that provision. Article XXX(d) further includes in its definition of “payments for current transactions” certain items that are capital in nature from an economic perspective, including (i) payments of moderate amount for amortization of loans or for depreciation of direct investments, (ii) moderate remittances for family living expenses, and (iii) normal short-term banking and credit facilities. Accordingly—and notwithstanding Article VI, Section 3—these capital transfers and transactions are subject to the Fund’s jurisdiction under Article VIII, and restrictions on these transfers or transactions require Fund approval.

3. Implications of Article VI, Section 3 for Article VIII. Article VI, Section 3 has influenced the interpretation of certain obligations of members under Article VIII. Specifically:

  • While Article VIII, Section 3 prohibits members from engaging in multiple currency practices (MCPs) or discriminatory currency arrangements without Fund approval, the Fund (on the basis of Article VI, Section 3) has declined to assert its jurisdiction over MCPs that relate solely to capital transactions, and has also determined that members may engage in discriminatory currency arrangements related solely to capital transactions whenever such arrangements “may be reasonably needed.”

  • A member’s obligation under Article VIII, Section 4 to convert certain official balances held by other members is expressly subject under the Articles to that member’s right to impose capital controls (Article VIII, Section 4(b)(i)).

  • Article VIII, Section 2(b) provides that exchange contracts involving the currency of any other member and which are contrary to the exchange control regulations of that member maintained or imposed consistently with the Articles of Agreement shall be unenforceable in the territories of that member. There was previously broad agreement among scholars and in the case law that a member’s obligation under this provision included exchange controls related to capital transactions. Some courts, however, have now held that, because Article VIII, Section 2(a) is concerned only with restrictions on current transactions, the words “exchange control regulations” in Article VIII, Section 2(b) must also be read to be limited to exchange controls related to this class of transactions.

4. Implications of Article VI, Section 3 for Fund conditionality. Two broad issues arise in considering the relationship between Article VI, Section 3 and Fund conditionality: (i) whether the Fund can require members to impose capital controls as a condition for the use of its resources, and (ii) whether the Fund can require members to remove capital controls as a condition for access to the Fund’s resources. The following may be noted in this context:

  • Article V, Section 3(a) provides that the Fund shall adopt policies on the use of its resources that will assist members to solve their balance of payments problems in a manner consistent with the provisions of the Articles and that will establish adequate safeguards for the temporary use of the Fund’s resources. This provision has been understood to permit the Fund to require members, as a condition of access to Fund resources, to impose restrictions on capital inflows (e.g., Fund-supported programs have long included performance criteria on external borrowing). As discussed in the main paper, the Fund has also supported members’ economic programs that included limits on outflows, in cases where the member’s potential needs in the absence of such controls would have outstripped both its adjustment capacity and the financing available from the Fund and other sources.

  • The right of members to control capital movements under Article VI, Section 3 has been interpreted as generally precluding the Fund from requiring the removal of capital controls as a condition for access to the Fund’s resources. As a limited exception to this principle, however, the Fund has established program conditionality on the nonaccumulation, reduction or elimination of external payments arrears, including arrears evidencing capital restrictions. This exception is based on the requirement in Article VI, Section 3(a) that the Fund establish policies that will assist members to solve their balance of payments problems and will establish adequate safeguards for the use of the Fund’s resources: arrears are particularly pernicious and their elimination and nonaccumulation are vital for effective balance of payments adjustment and safeguarding of the Fund’s resources, the very purposes of Fund conditionality under the Articles.

5. Limitation on Use of the Fund’s Resources to Finance Large or Sustained Outflows of Capital. Separate from the right to impose controls under Article VI, Section 3, Article VI, Section 1 provides that “[a] member may not use the Fund’s general resources to meet a large or sustained outflow of capital.” It should be noted that the drafters of the Articles, the staff and the Executive Board have consistently refrained from identifying or imposing any quantitative limits for the concept of “large or sustained.” No precise figure can be assigned to the concept of large, nor any precise time frame to the concept of sustained. Rather, it was repeatedly emphasized that the application of this limitation required the exercise of considerable judgment, taking into consideration the circumstances of each particular case and a number of relevant factors, including the following:

  • The limitation set forth in Article VI, Section 1 reflects, in part, a concern that priority be given to financing current account deficits. Such financing would assist in the elimination or avoidance of current account restrictions, which fall within the Fund’s mandate. In contrast, restrictions on capital transfers do not require Fund approval and the Fund may even request the imposition of such restrictions to prevent the use of its resources when the outflow is large or sustained. This priority also explains why in 1961 the staff stated that, in order to determine whether a particular capital outflow was “large,” an assessment had to be made as to whether the financing of such outflow would affect the Fund’s, ability to finance current account transactions of the member making the request or of other members. Therefore, the Fund’s overall liquidity and the size of the member’s quota were seen as relevant aspects that should be taken into account.

  • Although the only explicit limitation to the financing of capital outflows is the reference to “large or sustained” capital movements, the Fund’s purposes are also relevant in the context of the financing of a capital account deficit, for two reasons. First, any use of Fund resources must be consistent with the purposes of the Fund. Second, while the consistency with the Fund’s purposes and the concept of “large or sustained” are separate conditions for the use of Fund resources, in practice, the exercise of the Fund’s judgment in the context of “large or sustained” has focused on the consistency of the use of Fund resources with the Fund’s purposes as the main element in the Fund’s decision to provide financial assistance. A large or sustained outflow will often be a consequence of inappropriate monetary or fiscal policies, for instance, with respect to the member’s exchange rate or interest rates. In view of the fact that Fund resources are to be used to help members correct their balance of payments difficulties, a determination as to whether the capital outflows are “large or sustained” would involve an analysis of the causes of the outflows, and in particular an assessment as to whether the member’s policies are such that the use of Fund resources would contribute to—rather than delay—the resolution of these difficulties.


This paper was prepared by a team led by I. Mateos y Lago (SPR) and R. Weeks-Brown (LEG) and comprising A. Piris, N. Raman, S. Ramachandran, and S. Segal (all SPR) and K. Kwak (LEG), under the guidance of Sean Hagan (LEG) and Ranjit Teja (SPR). R. Duttagupta (RES) and R. Heath (STA) made substantive contributions.


October 2010 IMFC communiqué available at October 2010 G-20 communiqué available at


Shin and Shin (2010) illustrate with data for South Korea, while Maguire and von Peter (2009) discuss the case for European banks.


Data are from the WEO database, with “net” inflows defined as the sum of foreign direct investment, portfolio, and other investment balances. Gross inflows are defined as the sum of inward FDI, portfolio liabilities, and other investment liabilities.


Sørensen et al (2007) documents declining home bias for a sample of 24 OECD countries.


2007 Staff Report for the Article IV Consultation with Japan (IMF Country Report 07/280).


Calvo (2005); Chapter 5: “Capital Market Contagion and Recession: An Explanation of the Russian Virus.” The failure of the Long-Term Capital Management hedge fund is one example: losses on an advanced economy debt convergence strategy led to large sales in a range of other assets (shares in Royal Dutch Shell for example).


Persaud (2003) argues that small numbers of actors using similar risk-assessment technologies lie behind sudden falls in liquidity in major foreign exchange markets. Morris and Shin (2004) develop a model where short-horizon investors display herd behavior generating large price volatility.


The Articles permit members to impose restrictions on current payments and transfers if: (i) they have been temporarily approved by the Executive Board for balance-of-payments reasons; (ii) their maintenance is under the transitional provisions of Article XIV; or (iii) they have been imposed for reasons of international or national security pursuant to Executive Board Decision No. 144-(52/53).


Other previous attempts to establish universal agreements to govern international capital flows have similarly not been successful. Negotiations on the Multilateral Agreement on Investment were discontinued in 1998, and the attempt to negotiate an International Investment Framework in the WTO Doha Round was dropped.


The OECD Code does not contain firm obligations in this area, but it does specify that OECD members should “endeavor” to extend the measures of liberalization contained in that Code to all IMF members (i.e., regardless of lack of OECD membership).


The Fund could also continue, at the request of members, to provide technical assistance on a range of capital account policies, including the appropriate sequencing and timing of liberalization, and the analysis of restrictive measures.


The actions that the Fund could call on members to take or refrain from taking are not limited to the specific obligations listed in Article IV, Section 1 (i)-(iv), as it is recognized that the general obligation of collaboration under Article IV is broader in scope than the sum total of the four specific obligations set forth in that provision. The capital account policies covered by these principles would relate primarily to members’ external policies (rather than their domestic policies), but would go beyond “exchange rate policies” as contemplated under the 2007 Surveillance Decision. For a more detailed overview of the scope of Article IV, see Article IV of the Fund’s Articles of Agreement—An Overview of the Legal Framework (


The Fund’s Mandate—The Legal Framework (IMF Policy Papers, February 22, 2010) and Modernizing the Surveillance Mandate and Modalities (


IMF (2010c) proposes a Financial Stability Contribution, with somewhat similar structure, linked to improved resolution mechanisms. Korinek (2010a, 2010b) proposes taxation on a broader range of capital inflows.


Under Article XI of the GATS, restrictions on the making of payments and transfers for current international transactions that have been approved by the Fund cannot (by definition) give rise to a country’s breach of its GATS obligations. Separately, GATS members may not impose restrictions on capital transactions except under Article XII, which allows for restrictions imposed for balance of payments purposes or “at the request of the Fund”, and also allows for Fund input regarding the member’s balance of payments situation. With respect to Fund “requests”, Article VI, Section 1 of the Fund’s Articles authorizes it to request controls as a means of protecting Fund resources from being used to finance large or sustained capital outflows, but this provision has never been invoked by the Fund.


This Annex is not meant to provide an exhaustive summary of the scopes, features and implications of the myriad and sometimes complex agreements in these areas. A recent staff paper (see Reference Note on Trade in Financial Services, contains a summary of key features of the main legal and institutional arrangements governing the liberalization of trade in financial services, including the General Agreement on Trade in Services (GATS), preferential trade agreements (PTAs) and bilateral investment treaties. A more comprehensive examination and analysis of the key agreements in this area could be considered as part of the work program related to the Fund’s role in international capital flows.


Recent Developments in International Investment Agreements, UNCTAD (2008-June 2009), UNCTAD IIA MONITOR No. 3 (2009). At the end of 2008, the total stock of BITs in the world rose to 2676.


The number of these agreements has continued to increase. See UNCTAD (2008-June 2009), UNCTAD IIA MONITOR No. 3 (2009)


Specifically, the NAFTA allows for temporary derogations with respect to a restriction imposed for balance of payment reasons if the restriction is temporary, nondiscriminatory and consistent with the Fund’s Articles. In addition to these general rules, the NAFTA provides for a more specific rule with respect to current transfers; i.e., restrictions imposed on payments for current international transactions must be consistent with Article VIII, Section 3 of the Fund’s Articles. The ASEAN Comprehensive Investment Agreement contains a similar provision. Under Article 13 of the ASEAN Agreement, it is provided that, nothing in the ASEAN Agreement will “affect the rights and obligations” of Fund members under the IMF Articles, including the use of exchange actions which are in conformity with the Articles, and that a Member shall not impose restrictions on any capital transactions inconsistently with its specific commitments under the ASEAN Agreement regarding such transactions, except: (a) at the request of the IMF; (b) for balance of payments reasons (under Article 16 of the ASEAN Agreement); or (c) where movements of capital cause, or threaten to cause, serious economic or financial disturbance in the member.


In particular, some recent bilateral agreements also have robust provisions related to trade in services (e.g., EU association agreements, U.S. FTAs concluded with ASEAN members).


Some agreements, however (e.g., NAFTA and other FTAs), adopt the converse “negative list” approach by providing general obligations that are subject to reservation.


Both the GATS and the NAFTA establish a “prudential carve out” and clarifies the extent to which activities of central banks or monetary authorities are excluded from the scope of these treaties.


On its face, the GATS does not appear to allow for restrictions on capital inflows.


See, e.g., The Relationship of the World Trade Organization with the Fund—Legal Aspects (


Rather, the Code provides that “[n]othing in this Code shall be regarded as altering the obligations undertaken by a Member as a Signatory of the Articles of Agreement of the International Monetary Fund” (Article 4 of the Capital Code).


This Annex summarizes legal issues in the complex area of international capital movements solely as background for the discussion in the main paper. Previous Board papers contain a more detailed discussion of the relevant considerations in this area.


This Annex does not discuss issues related to the relationship between Article VI, Section 3 and Article IV, as these are discussed in the main paper.

The Fund's Role Regarding Cross-Border Capital Flows
Author: International Monetary Fund