Journal Issue
Share
Article

Net Capital Flows, Financial Integration, and International Reserve Holdings: The Recent Experience of Emerging Markets and Advanced Economies

Author(s):
International Monetary Fund. Research Dept.
Published Date:
July 2009
Share
  • ShareShare
Show Summary Details

Global holdings of international reserves have increased rapidly in recent years. This increase has been especially dramatic in emerging markets, both in absolute terms as well as in comparison to the reserves held by advanced countries. At the end of 2005, the average reserves-to-GDP ratio reached 19 percent in emerging markets, compared with a ratio of 10 percent in the advanced countries. Emerging markets have accumulated reserves well above the levels suggested by traditional rules of thumb based on current account transactions and short-term external liabilities.1 Also, the recent record pace of reserve accumulation in emerging markets is at odds with the prediction of a standard reserve holding model (IMF, 2003).

With increasing financial liberalization and openness to cross-border transactions, managing a country’s liquid assets to facilitate current and future international transactions—what we call “sovereign liquidity”—has become a key element in macroeconomic management. Clearly, the desired level of reserves and the availability of liquidity depend on a sovereign’s access to international capital markets. This paper examines the relationship between reserve holdings and net capital flows (capital inflows minus capital outflows) in the changing global financial environment.

In recent decades, currency and/or financial crises accompanied by reversals in capital flows have become more frequent and severe. With increased financial integration, countries are more vulnerable to contagion from within and outside their regions (for example, Kaminsky and Reinhart, 2000). In response, central banks in developing countries have accumulated reserves to cushion extreme events, the bunching of external debt maturities, or other shocks that could affect the foreign exchange market and the domestic economy.

Holding large reserves is costly, but the perceived cost may be small relative to the economic and social cost of a crisis. Many of the currency and financial crises of the past 10 years have been associated with the contractionary effects of currency depreciation, with substantial output losses, especially through balance sheet channels (for example, Choi and Cook, 2004; and Frankel, 2005). After the Asian financial crisis, emerging markets have reduced short-term external debt, and stockpiled reserves to reduce their vulnerability to a crisis (for example, Aizenman and Marion, 2004; Rodrik, 2006; and Jeanne, 2007). Because a reserve holding country can always opt not to use its reserves for debt service, reserves have an insurance value specific to the country (Van Wijnbergen, 1990). Recently, Durdu, Mendoza, and Terrones (forthcoming) suggested that financial globalization with endogenous sudden stops can explain the large buildup of reserves.

This paper examines the “external financing” of a reserve buildup using net capital flows. Given frictions and information problems in international capital markets, countries facing uncertain growth prospects and volatile capital flows have a stockpiling motive in reserve management, even at the cost of increased external debt. Sovereign liquidity holdings are analogous to corporate liquidity holdings in that they help cope with uncertain income streams and cash flows. In the face of external finance premia, the behavior of corporations suggests that the value of having liquid assets is disproportionately high for financially weak firms (Kim, Mauer, and Sherman, 1998; and Almeida, Campello, and Weisbach, 2004). This external financing view of a reserve buildup is in contrast with New Mercantilist views (for example, Dooley, Folkerts-Landau, and Garber, 2004) that a reserve buildup is financed by internal cash flows from current account surpluses as a by-product of a strategy to protect competitiveness in countries dependent on exports for output growth.

With increasing financial integration, and greater exposure and dependence on international capital, sovereign liquidity management has become crucial for macroeconomic stability. The fact that more capital does not flow from rich to poor countries—the paradox of too little flow (Lucas, 1990)—might be substantially attributable to credit-market imperfections (Reinhart and Rogoff, 2004b). For the high-growth emerging markets, the paucity of capital has been ameliorated by financial globalization. Nonetheless, emerging markets may take advantage of the wave of capital inflows to stockpile reserves, as in other times external financing may be expensive due to credit-market imperfections. With increased capital and financial account transactions, the risk of capital flow reversals that can result in huge output losses has increased the option value of holding reserves. As sovereign liquidity shortages may lead to expensive borrowings or a forced reduction in consumption and investment, countries have an incentive to fend off the risk of binding liquidity constraints in the future by hoarding reserves today. Clearly, countries that can borrow at reasonable risk premia will need a smaller stock of reserves, and the differential in reserve holdings across countries should depend on the degree of access to international financial markets.

To examine how the availability of external finance has affected the link between net capital flows and the accumulation of reserves, we estimate panel data models over the period 1980–2005 for 36 emerging markets and 24 advanced economies. Financial globalization is captured by using the stock value of foreign assets and liabilities taken from Lane and Milesi-Ferretti (2006) or by period dummies. The analysis reveals that the effect of net capital flows on reserves for emerging markets has changed dramatically over time, and their reserve holding pattern has been quite different from that of advanced countries. In the 1980s, reserve holdings were negatively associated with net capital flows for emerging markets, while such a negative link was less pronounced for advanced countries. In recent years, net capital flows have had a strong positive effect on reserves for emerging markets but a negative effect for the advanced countries. This finding supports our external financing view. In the face of increasing financial globalization, with heightened concerns about the risks of “sudden stops” (Calvo, 1998) and loss of access to international capital markets, emerging markets have built up sovereign liquidity when funds are available at low costs.

I. Determinants of International Reserves

Buffer Stocks and the Precautionary Motive

The pure buffer stock/precautionary model focuses on the opportunity cost, the adjustment cost, and volatility. First, the standard measure of the opportunity cost is the differential between the country’s own-interest rate and the interest rate on comparable U.S. treasuries. Most empirical studies, however, do not find a significant negative effect for the opportunity cost (Flood and Marion, 2002; IMF, 2003; and Aizenman and Marion, 2004).2 Second, if the import share in output is a proxy for the marginal propensity to import, and thus its inverse is an indication of the required output adjustment to produce a particular level of reserves, then the relationship would be negative as countries can use their reserves to reduce such adjustment costs. However, the relationship between reserves and the import share is ambiguous (Heller and Khan, 1978). If the import share represents the economy’s openness and vulnerability to external shocks (Edwards, 1984; and Aizenman and Marion, 2004), the relationship would be positive as the more open is a country to external shocks, the greater is the need for reserves. Also, the use of the ratio of reserves to imports as an indicator of reserve adequacy for facilitating current account transactions implies that the relationship is positive. Third, reserve holdings increase with the volatility of international transactions (Frenkel and Jovanovic, 1981; Flood and Marion, 2002; and Aizenman and Marion, 2004). This volatility is usually measured by the standard deviation of the trend-adjusted changes in reserves over some period.3 An alternative measure is the volatility of export receipts used by Edwards (1985) and Aizenman and Marion (2004).

Flood and Marion (2002) find that buffer-stock reserve models work about as well in the modern floating exchange rate period as they did during the Bretton Woods period. The IMF (2003), using a standard buffer stock model based on Aizenman and Marion (2004), suggests that the rapid accumulation of reserves in emerging markets between 1997 and 2001 was broadly in line with fundamentals, but the surge in reserves in 2002 and 2003 was above the level predicted by the model. Recent studies have examined the rapid reserve accumulation in emerging markets over the last 5 years, focusing on the risk of sudden reversals in capital flows. It has been suggested that emerging markets may hold large reserves with a precautionary motive to provide self-insurance against a capital account crisis (Aizenman and Marion, 2003; Aizenman and Lee, 2007; and Jeanne, 2007).

Net Capital Flows and Financial Globalization: External Financing View

We propose that financial market imperfections and the risk of financial distress at a country level give rise to a precautionary motive of sovereign liquidity buildup when liquidity flows are available at low costs. In a loose analogy, the finance literature suggests that corporate liquidity demand responds to cash flows under financial market frictions: Almeida, Campello, and Weisbach (2004) show that financially constrained firms have a propensity to use cash flows to build up cash reserves or liquid assets because such assets enable a firm to reduce its dependence on costly external funding for future production activities—a precautionary or stockpiling motive.

The paper emphasizes the external financing of reserves in contrast with the New Mercantilism that typically concerns internal financing only. The relationship between reserves and net capital flows can be viewed in light of the identity in the balance of payments: “current account balance plus net capital flows (the latter being the capital and financial account balance) equals the change in reserves (a positive value means an increase).” This identity implies three types of relationships between net capital flows and the stock of reserves: (a) a negative (zero) correlation when current account balance is partially (fully) offset by net capital flows—current account financing by net capital flows; (b) a positive correlation when net capital flows determine whether reserves increase or decrease regardless of current account balance (that is, current account balance is more than fully offset by net capital flows or current account balance and net capital flows move in the same direction); and (c) no systematic correlation when there is little capital mobility. This identity suggests two sources of financing reserves. Types (a) and (c) pertain to the “internal” financing of reserve accumulation using cash flows from current account surpluses. This internal financing of reserves is in line with the New Mercantilist view that emphasizes export-driven reserve accumulation. Type (b) pertains to the “external” financing of a reserve buildup using net capital flows, and this is likely to become more important with increasing financial globalization.

The external financing view suggests that sovereign liquidity demand will be related to country cash flows (that is, net capital flows), but the relationship between sovereign liquidity and net capital flows may shift over time. The cost of restocking reserves through selling assets or expanding liabilities depends on a country’s access to international capital markets and global financial conditions:4 in bad times countries with low creditworthiness, facing restricted access to international capital markets, find it more expensive to borrow than in good times. Financial globalization reduces barriers to cross-border capital flows and increases access to international financial markets, and also raises the potential risk of country cash flows. Developing countries have a heightened concern about sudden reversals in capital flows and the risk of financial crises (Calvo, 1998; Aizenman and Marion, 2003; Edwards, 2004; and Caballero and Panageas, 2005). Emerging markets, concerned about sudden stops in capital flows, may have increased their reliance on net capital flows to build up reserves in the face of financial globalization. In contrast, advanced countries have been able to raise external funds at relatively low costs through international financial markets. With increased financial globalization, advanced countries have become more active external liquidity providers and lowered their reserve levels relative to GDP. Therefore, sovereign liquidity demands may entail time-varying responses to net capital flows.

The demand for reserves also depends on a country’s credit quality as it affects the cost of external finance and the degree of access to international markets. Countries with low credit ratings face higher cash flow risk and have to pay a higher rate for external funds compared with countries with greater creditworthiness.5 Thus, lower rated countries will have a stronger incentive to build up reserve buffers to withstand adverse external shocks compared with higher rated countries, consistent with the precautionary liquidity demand story that liquidity holdings are positively related to cash flow risk (Kim, Mauer, and Sherman, 1998; and Almeida, Campello, and Weisbach, 2004). Country credit quality can be measured by sovereign ratings, which summarize and supplement information on macroeconomic indicators, default history, as well as social and political factors (see, for example, Cantor and Packer, 1996). Thus, sovereign ratings should have a negative impact on reserve holdings. Note that higher reserves could reduce the risk of sovereign default (Ben-Bassat and Gottlieb, 1992; and Mellios and Paget-Blanc, 2006), implying a positive correlation between reserves and sovereign rating. In our empirical analysis we take account of the endogeneity of sovereign ratings.

Exchange Rate Flexibility and Other Considerations

Exchange rate flexibility: Conventional wisdom holds that greater exchange rate flexibility should reduce reserves because central banks then do not need a large reserve stockpile to maintain a peg or to enhance the peg’s credibility (Flood and Marion, 2002; and Aizenman and Marion, 2004). However, with volatile capital flows, the need for reserves to temper asset price fluctuations may increase with exchange rate volatility. Also, central banks, to dampen the appreciation of their currencies (New Mercantilist views), may accumulate reserves (Frankel and Dornbusch, 1995; and Dooley, Folkerts-Landau, and Garber, 2004).

Other considerations: First, the literature on the interest rate–exchange rate nexus suggests that countries facing the risk of capital flight use domestic interest rates to counter exchange rate movements and prevent capital outflows and/or a lowering of reserves. Second, the net effect of the fiscal stance on reserves will depend on a combination of political-economy factors and fiscal vulnerabilities. Aizenman and Marion (2003, 2004) argue that a greater chance of opportunistic behavior by future policymakers or political corruption reduces the demand for reserves, and may lead to higher external borrowing for increasing the current consumption of special interest groups, suggesting a positive link between reserves and fiscal surpluses. Conversely, as fiscal vulnerabilities increase the risk of sovereign default and a crisis, countries in fiscally fragile situations may hold larger reserve cushions, implying a negative link between reserves and fiscal surpluses (Reinhart and Rogoff, 2004b). Lastly, to obtain higher returns countries may prefer to invest national resources in foreign assets other than reserves (Devereux and Sutherland, 2007).

II. Empirical Reserve Demand Models and Estimated Results

Data

The empirical analysis uses annual data for 36 emerging markets and 24 advanced countries for the 1980–2005 period (Appendix I). Most of the data have been collected from the IMF’s International Financial Statistics and World Economic Outlook databases, and from the CEIC. The descriptive statistics are reported in Appendix II—for further details, see Choi, Sharma, and Strömqvist (2007). As in IMF (2003) and Aizenman and Marion (2004), international reserves are defined as gross reserves net of gold. Reserves relative to GDP over the entire period are higher for emerging markets (12.5 percent) than for advanced countries (9.6 percent). Net capital flows, measured by the capital and financial account of the balance of payments, are higher for emerging markets than for advanced countries by about 2.5 percent of GDP. Sovereign ratings from Standard and Poor’s are higher on average for the advanced countries (22 compared with 14 for emerging markets).

Figure 1 (panel A) depicts the averages for reserves and the reserves-to-GDP ratio by country group for the 1980–2005 period. In terms of the reserves-to-GDP ratio, emerging markets caught up with the advanced countries in the early 1990s. Thereafter, while for advanced countries this ratio fluctuated around 10–11 percent, for emerging markets it rose substantially after the Asian crisis until it reached 20 percent by 2003, and then slowed down to 19 percent by 2005. Panel B shows the different patterns of net capital flows for the country groups. For emerging markets, as capital accounts were liberalized, net capital flows increased steadily over the 1991–96 period but fell dramatically during the 1997–98 Asian crisis period. For advanced countries, net capital flows showed a downward trend since 1991, a sharp increase after the Asian crisis, which reflects the funding of the large U.S. current account deficits. When net capital flows are normalized by GDP, the United States does not change the pattern of net capital flows for advanced countries, even though due to its size it has an important effect on the average level of net capital flows for advanced countries. As shown in panel C, emerging markets exhibit a downswing in the current account for 1991–96 (which mirrors an upswing in net capital flows), and a rising trend in the current account during recent years. For advanced countries, the current account relative to GDP fluctuates around a rising trend but was on average negative until the early 1990s. The negative current account average for advanced countries since 1998 is attributable to the U.S. current account deficit: excluding the United States from the sample, the cross-sectional average has been positive.

Figure 1.International Reserves, Net Capital Flows, and Current Accounts

Note: The figure depicts the cross-sectional averages of variables for emerging markets and advanced countries (see Appendix I for the country group list and Appendix II for the definition of variables). Dotted lines with symbols in panels B and C indicate the cross-sectional average of variables excluding the United States.

Following Lane and Milesi-Ferretti (2006), we consider two de facto measures of global financial integration based on actual capital flows: one is the ratio of total foreign assets and liabilities to GDP, and the other is the sum of the stock of portfolio equity assets and liabilities and the stock of foreign direct investment (FDI) assets and liabilities to GDP. Over 1970–2004, this measure gradually increased during the 1970s and 1980s but accelerated, especially in advanced countries, during the mid-1990s. Figure 2 depicts the cross-section average of the Lane and Milesi-Ferretti measures for the country groups that we use in this paper. For advanced countries, both measures show an upward trend with a sharp increase starting in 1998. For emerging markets, they show a gradual increase over the 1980s, some fluctuations with no apparent trend over the 1990s, and then a modest rise in recent years.

Figure 2.Measures of Financial Integration

Note: Panel A (Panel B) depicts the cross-sectional average of the percentage ratio of total foreign assets and liabilities (the sum of the stock of portfolio equity assets and liabilities and the stock of foreign direct investment (FDI) assets and liabilities) to GDP for 1980–2004 for emerging market and advanced country groups. Luxembourg was excluded from the sample because the data were available only from 2000.

Figure 3 shows the group means of the reserve/GDP ratios under different exchange rate regimes. On the basis of the de facto regime classification used by Reinhart and Rogoff (2004a), regimes are divided into three categories: a fixed regime (index = 1), an intermediate regime (index = 2), and a floating regime (index = 3).6 For emerging markets, consistent with the conventional wisdom the more flexible regimes are associated with lower reserves. For advanced countries, the floating regime is associated with the lowest reserves-to-GDP ratio, but the intermediate regime has the highest ratio, perhaps reflecting that more reserves are required for ameliorating exchange rate volatility under a closely managed float. Notably, the reserves-to-GDP ratio shows an upward trend under all regimes for emerging markets but only under the intermediate regime for advanced countries. Such shifts suggest that factors other than exchange rate regimes have an important effect on reserves.

Figure 3.Reserves-to-GDP Ratio by Exchange Rate Regime Type

Note: The figure depicts the group means of the reserve-to-GDP ratio by regime type over subperiods for emerging markets, advanced countries, and the pooled sample. Regimes are indexed as follows: fixed regime = 1; intermediate regime = 2; and floating regime = 3 (see footnote 6). The subperiods are whole (1980–03); Y80_90 (1980–90); Y91_96 (1991–96); Y97_00 (1997–2000); and Y01_03 (2001–03).

Reserve Regressions

Our regression model incorporates the aforementioned determinants of reserve holdings including sovereign ratings and net capital flows.7 For country i at time t, the ratio of reserves to output in regression model 1 is given by

where IR/GDP is the ratio of international reserves to GDP (gross domestic product in U.S. dollars), and SIZE is the natural log of population, a supplementary scale factor in the model as we scale the dependent variable, reserves, by GDP. σC and σT are volatility measures of the growth rate of the nominal exchange rate. We take the growth rate of the exchange rate to eliminate the level effects in the nominal exchange rate due to currency denominations and control for the nonstationarity of the exchange rate. Cross-country volatility, σC, is measured by the standard deviation of the growth rate of the nominal exchange rate for country i over the sample period. Time-varying volatility, σT, is measured by the conditional standard deviation predicted from a GARCH (1,1) model for emerging markets and an ARCH(2) model for advanced countries—the exchange rate growth series is typically more persistent (yet stationary) for emerging markets than for advanced countries. In these models, we control for fixed effects by allowing for different means across countries and, for the sake of parameter parsimony, impose the same slope coefficient within the same country group. IM/GDP is the import share in GDP, and SR is the sovereign rating assigned by Standard and Poor’s. The parameter α denotes fixed-country effects, and ε is the error term.

CF/GDP is the ratio of net capital flows to GDP. If net capital flows are used mainly for financing current account deficits (stockpiling reserves), βCF will be negative (positive). To account for the effect of financial globalization on reserves, we allow for the effect of net capital flows to vary over subperiods. For emerging markets, net capital flows fell drastically during the Asian financial crisis of 1997–98, reversed in 2001, and increased thereafter (Figure 1). Financial globalization was gradual during the 1980s but accelerated in the mid-1990s for advanced countries (Figure 2). We thus interacted the capital flow variable with four time dummies: D(80–90), D(91–96), D(97–00), and D(01–05) took on the value 1 for 1980–90, 1991–97, 1998–2000, and 2001–05 respectively, and were zero otherwise.

The net capital flow and sovereign rating variables may be correlated with the error term due to reverse causality, which may arise, for example, when countries borrow from abroad to meet a reserves target and when ample reserve holdings have a favorable impact on the assessment of sovereign ratings. To deal with the endogeneity of regressors, we use instrumental variables (IV), with the lagged values of sovereign ratings and net capital flows as the instruments.8 A first-stage regression showed that the instruments were highly correlated with the endogenous variable. The Hansen test for overidentification suggested that the regression model was correctly specified and the instruments were valid.9 Statistical inferences about coefficient estimates are based on heteroscedasticity and autocorrelation consistent standard errors.

Table 1 summarizes regression model 1 for emerging markets and advanced countries, along with the pooled sample. For the pooled sample, when country heterogeneity is taken into account by fixed country effects, the adjusted R-squares, R¯2, show that the model explains about 88 percent of the variation in reserves. However, if the variation explained by the fixed country effect is excluded, R¯2 drops to 57 percent, suggesting that a large part of the variation is picked by country-specific heterogeneity. For the group-specific regressions, the explanatory power of the model is much higher for advanced countries than for emerging markets, as indicated by their R¯2(0.94 vs. 0.79). Excluding the fixed country effects, R¯2 in the ordinary least squares (OLS) regressions drops to 0.41 for emerging markets and 0.72 for advanced countries, suggesting much larger heterogeneity in reserve holding behavior for the former group.

Table 1.Model 1: The Time-Varying Effects of Net Capital Flows on Reserves
Independent VariablesPooledEmergingAdvanced
OLSIVOLSIVOLSIV
SIZE0.246***0.254***0.319***0.358***0.132***0.068
(5.74)(5.02)(5.64)(5.08)(2.78)(1.37)
σC–0.920***–0.928***–0.003***–0.002***–0.615***–0.547***
(–8.36)(–8.26)(–3.92)(–3.60)(–6.27)(–5.69)
σT0.002*0.0010.001**0.001**0.1090.123 (0.64)
(1.62)(1.30)(2.29)(2.09)(0.64)
IM/GDP0.232***0.249***0.238***0.161**0.131*0.144**
(4.75)(4.98)(3.49)(2.01)(1.75)(2.18)
SR–0.004*–0.005**–0.004**–0.007**–0.013***–0.015***
(–1.93)(–2.18)(–2.15)(–2.17)(–3.51)(–3.60)
D(80–90)*(CF/GDP)0.1210.134–0.488**–1.086**–0.006–0.320**
(0.21)(0.83)(–2.33)(–2.32)(–0.06)(–2.02)
D(91–96)*(CF/GDP)0.151*0.344**0.1030.145–0.169–0.561**
(1.69)(2.13)(0.81)(0.68)(–1.16)(–2.05)
D(97-00)*(CF/GDP)–0.075–0.1270.1180.265–0.166*–0.442****
(–1.09)(–1.02)(1.27)(1.06)(–1.72)(–2.87)
D(01–05)*(CF/GDP)0.0720.1270.438***0.835***–0.332***–0.502***
(0.99)(1.02)(5.25)(2.67)(–2.75)(–4.11)
R¯210.8890.8870.7950.7810.9360.931
[0.597][0.596][0.413][0.388][0.724][0.723]
No. of observations897897407407490490
Hansen’s test: χ2(4) 21.0951.4813.217
[0.895][0.830][0.522]
Note: This table shows the results of regression model 1 for emerging markets, advanced countries, and the pooled sample. The reserves-to-GDP ratio (IR/GDP) is regressed on the log of population (SIZE), cross-sectional volatility (σC), time-varying exchange rate volatility (σT), import share (IM/GDP), sovereign rating (SR), and net capital flows (CF/GDP) interacted with period dummies (D). The conditional standard deviation predicted from a GARCH (1,1) model for emerging markets and that from an ARCH(2) model for advanced countries were merged to measure σT for the pooled sample. Both OLS and IV regressions include fixed-country effects. Instruments for IV estimations are the one-period-lagged values of sovereign ratings and the two lags of each of the net capital flow variables. The z-ratios in parentheses are based on standard errors robust to heteroscedasticity and autocorrelation (Bartlett kernel; bandwidth = 2). ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively.

R¯2 is the adjusted R-squared, and the figure in brackets excludes variance explained by the fixed-country effects.

Hansen’s test is an overidentification test for all instruments. Under the null hypothesis that the model is correctly specified and the instruments are valid, the test statistic is distributed as a chi-square with the degree of freedom equal to the number of overidentifying restrictions (with p-values in square brackets).

Note: This table shows the results of regression model 1 for emerging markets, advanced countries, and the pooled sample. The reserves-to-GDP ratio (IR/GDP) is regressed on the log of population (SIZE), cross-sectional volatility (σC), time-varying exchange rate volatility (σT), import share (IM/GDP), sovereign rating (SR), and net capital flows (CF/GDP) interacted with period dummies (D). The conditional standard deviation predicted from a GARCH (1,1) model for emerging markets and that from an ARCH(2) model for advanced countries were merged to measure σT for the pooled sample. Both OLS and IV regressions include fixed-country effects. Instruments for IV estimations are the one-period-lagged values of sovereign ratings and the two lags of each of the net capital flow variables. The z-ratios in parentheses are based on standard errors robust to heteroscedasticity and autocorrelation (Bartlett kernel; bandwidth = 2). ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively.

R¯2 is the adjusted R-squared, and the figure in brackets excludes variance explained by the fixed-country effects.

Hansen’s test is an overidentification test for all instruments. Under the null hypothesis that the model is correctly specified and the instruments are valid, the test statistic is distributed as a chi-square with the degree of freedom equal to the number of overidentifying restrictions (with p-values in square brackets).

A key finding is that net capital flows have a time-varying effect on reserves that differs across emerging markets and advanced countries. For emerging markets, reserves decreased with net capital flows in the 1980s, suggesting that, before financial integration took off, net capital flows were used to finance current account deficits. In contrast, during the subsequent periods, net capital flows either had little effect or a positive effect on reserves. During the 1991–96 period, a substantial net flow of capital into emerging markets led to only small increases in reserves, reflecting the use of net capital flows primarily to finance domestic expenditures. A positive but insignificant coefficient for 1997–2000 may reflect the mix of capital flight that caused reserve losses during the Asian crisis and the return of capital that helped the restocking of reserves in its aftermath. The relatively high and significant coefficient for 2001–05 implies that net capital flows led to a substantial accumulation of reserves. In contrast, for advanced countries, net capital flows were associated with lower reserve levels relative to GDP, especially in recent years. This suggests that financial globalization helped emerging markets, which had greater concern about the risk of sudden stops after the Asian financial crises, to build up their reserve levels through external financing, whereas it helped advanced countries to become more active external liquidity providers, reducing reserves.

For variables other than net capital flows, the analysis shows, first, that the population coefficient is positive and larger for emerging markets than for advanced countries.10 One could argue that the need for reserves relative to output initially increases as an economy grows to a certain threshold, and then flattens out or even declines. Second, the cross-country exchange rate volatility coefficient is negative, consistent with Flood and Marion’s (2002) finding that countries with greater flexibility in the exchange rate hold lower reserves. The coefficient estimate is much smaller for emerging markets than for advanced countries, partly because the measured volatility on average is much larger for the former than for the latter (see Appendix II). The coefficient on time-varying exchange rate volatility has a positive sign, suggesting that reserve holdings increase to moderate exchange rate volatility: it is significant at the 5 percent level for emerging markets but insignificant for advanced countries. Third, the import share coefficient is positive for all country groups. In addition, the sovereign rating has a negative effect on reserves, implying that the higher a country’s credit-worthiness, the less the need for reserve holdings, with a more pronounced effect for advanced countries than for emerging markets.11

In the pooled regressions, net capital flows do not have any discernable effects owing to their opposing effects in the two country groups, except that they had somewhat significant positive effects in the 1990s before the Asian crisis. As the main point of this paper is to show how the effects of capital flows have differed across emerging markets and advanced countries, we henceforth report results only for the two different country groups. Also, we do not report the regressions including the standard opportunity cost measure of reserves (the differential between the domestic interest rate and the rate of return on U.S. treasuries), because, as in previous studies, we did not find that it had any discernible effect on reserves.12

The Lane and Milesi-Ferretti measure of financial globalization is employed in place of the time dummies interacted with the capital flow variable in regression model 2:

where GLOBi,t denotes the logarithm of the country-specific globalization measure for country i at time t. We use two alternative country-specific globalization measures: the two-year average of the ratio of total foreign assets and liabilities to GDP; and the two-year average of the ratio of the sum of the stock of equity assets and liabilities and the stock of FDI assets and liabilities to GDP (see Appendix II for descriptive statistics). As shown in Table 2, the two alternative measures give the same qualitative results. The effect of net capital flows per se—measured by βCF—is significantly negative for emerging markets but positive for advanced countries.13 Importantly, the coefficient net capital flows interacted with the financial globalization measure (βGL) is significantly positive for emerging markets but negative for advanced countries. This finding reinforces the earlier results: the effect of net capital flows on reserve accumulation drifts up over time (from negative to positive) for emerging markets, and shifts down over time for advanced countries. The effects of other variables are similar to those in regression model 1.

Table 2.Model 2: Effects of Net Capital Flows and Financial Globalization
Independent VariablesEmergingAdvanced
OLSIVOLSIV
A. Globalization measure based on total foreign assets and liabilities
SIZE0.289***0.288***0.089**0.034
(4.47)(3.99)(2.01)(0.72)
σC-0.003***-0.003***-0.433***-0.312**
(-5.16)(-5.12)(-4.20)(-2.36)
σT0.001**0.001**0.1760.178
(2.49)(2.53)(1.09)(1.03)
IM/GDP0.298***0.294***0.110*0.109**
(4.62)(4.27)(1.72)(2.06)
SR-0.004**-0.004-0.014***-0.019***
(-2.25)(-1.56)(-4.82)(-5.69)
CF/GDP—1.467***-1.579**2.658***3.460***
(-3.42)(-2.55)(5.84)(4.51)
GLOB*(CF/GDP)0.343***0.362***-0.495***-0.657***
(3.79)(2.81)(-5.78)(-4.93)
R¯210.810 [0.461]0.810 [0.437]0.949 [0.814]0.946 [0.813]
No. of observations356356461461
Hansen’s test: χ2(1) 20.296 [0.587]0.077 [0.782]
B. Globalization measure based on the stock of equity and foreign direct investment
SIZE0.305***0.303***0.087*0.030
(4.95)(4.66)(1.83)(0.59)
σC-0.003***-0.003***-0.460***-0.358***
(-5.53)(-5.62)(-4.57)(-2.99)
σT0.001**0.001***0.1830.183
(2.52)(2.82)(1.11)(1.05)
IM/GDP0.325***0.293***0.116*0.119*
(4.78)(4.00)(1.63)(1.94)
SR-0.003**-0.003-0.014***-0.018***
(-1.89)(-1.49)(-4.44)(-5.12)
CF/GDP-0.639***-1.338***1.243***1.448***
(-3.23)(-3.40)(5.12)(2.80)
GLOB*(CF/GDP)0.252***0.436***-0.318***-0.398***
(4.48)(4.31)(-5.18)(-3.60)
R¯210.818 [0.459]0.810 [0.427]0.946 [0.813]0.943 [0.810]
No. of observations356356461461
Hansen’s test: χ2(1) 22.023 [0.155]0.399 [0.527]
Note: In this table, regressors include net capital flows (CF/GDP) interacted with a constant and the logarithm of the globalization measure (GLOB) as specified in regression model 2. Both OLS and IV regressions include fixed-country effects. Instruments for IV estimations are the one-period lagged values of sovereign ratings, the two lags of net capital flows, and the globalization measure (GLOB) multiplied by the one-period lagged CF/GDP (as we treat GLOB as exogenous). Luxembourg was excluded from the sample since the globalization measure was available only from 2000. The z-ratios in parentheses are based on standard errors robust to heteroscedasticity and autocorrelation (Bartlett kernel; bandwidth = 2). ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively.

See footnote one in Table 1.

See footnote two in Table 1.

Note: In this table, regressors include net capital flows (CF/GDP) interacted with a constant and the logarithm of the globalization measure (GLOB) as specified in regression model 2. Both OLS and IV regressions include fixed-country effects. Instruments for IV estimations are the one-period lagged values of sovereign ratings, the two lags of net capital flows, and the globalization measure (GLOB) multiplied by the one-period lagged CF/GDP (as we treat GLOB as exogenous). Luxembourg was excluded from the sample since the globalization measure was available only from 2000. The z-ratios in parentheses are based on standard errors robust to heteroscedasticity and autocorrelation (Bartlett kernel; bandwidth = 2). ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively.

See footnote one in Table 1.

See footnote two in Table 1.

Next, by first differencing Equation (2) and including lagged dependent variables, we also estimate a dynamic panel regression model to account for possible dynamic adjustments of the reserves-GDP ratio to its time-varying determinants. The cross-sectional exchange rate volatility reflects a “fixed” characteristic over the sample period. The population variable also shows little variability around a trend over time and is not suitable for a dynamic setting. Hence, variables for fixed country effects, population, import share, and cross-sectional exchange rate volatility are dropped from the dynamic panel model. The dynamic panel counterpart of Equation (3) is represented by regression model 3:

where the sovereign rating and net capital flows are assumed to be endogenous. We estimate this model, using Arellano and Bond’s (1991) generalized method of moments (GMM) estimator.

Table 3 shows how financial globalization may have changed the effect of net capital flows on reserve accumulation. The results are generally in line with those obtained in the OLS and IV regressions in Table 2, suggesting that the effect of net capital flows on reserves has increased with financial integration for emerging markets, whereas the converse is true for advanced economies. The coefficient on the one-period lagged dependent variable has a statistically significant coefficient of 0.775, indicating substantial persistence in reserve ratios. The effect of exchange rate volatility (over time) on reserves is positive and significant at the 5 percent level for both country groups. Sovereign ratings had a negative effect on reserves for both country groups, indicating that smaller reserve cushions were needed by economies with higher ratings. The Wald test statistics indicate the joint significance of the regressors. As the Arellano-Bond estimator assumes first-order autocorrelation and no second-order autocorrelation of the residuals, the serial correlation tests do not indicate misspecification.

Table 3.Model 3: Dynamic Panel Regressions
Independent VariablesEmergingAdvanced
Δ(IR/GDP)–10.775***0.762***
(7.78)(14.98)
Δ(IR/GDP)-2–0.0740.101*
(–1.09)(1.66)
ΔσT0.0005**0.146**
(2.54)(2.23)
ΔSR–0.004***–0.006**
(–2.63)(–1.97)
Δ(CF/GDP)–0.1020.740***
(–0.59)(4.64)
Δ[GLOB*(CF/GDP)]0.101***—0.149***
(2.75)(–3.99)
Constant0.004***–0.000
(3.74)(–0.35)
Wald testξ2(6) = 162.2ξ2(6) = 449.4
No. of observations333439
Arellano-Bond tests: order 1—2.98 (0.003)—3.26 (0.001)
order 2—0.47 (0.641)0.69 (0.490)
Note: This table shows the results of regression model 3 for emerging markets and advanced countries over the period 1980–2004 using the Arellano and Bond (1991) generalized method of moments. The last year in the sample is dictated by the availability of the globalization measure that is based on the sum of the portfolio equity assets and liabilities and the stock of foreign direct investment assets and liabilities (the other measure based on total assets and liabilities gave similar qualitative results). The sovereign rating and net capital flows are treated as endogenous. Luxembourg was excluded from the sample since the globalization measure was available only from 2000. The z-ratios in parentheses are based on standard errors robust to heteroscedasticity. Significance at the 1, 5, and 19 percent level is shown by ***, **, and *, respectively. The statistics for the Arellano-Bond tests are based on the null hypothesis of no autocorrelation of order 1 and 2 (with p-values in parentheses).
Note: This table shows the results of regression model 3 for emerging markets and advanced countries over the period 1980–2004 using the Arellano and Bond (1991) generalized method of moments. The last year in the sample is dictated by the availability of the globalization measure that is based on the sum of the portfolio equity assets and liabilities and the stock of foreign direct investment assets and liabilities (the other measure based on total assets and liabilities gave similar qualitative results). The sovereign rating and net capital flows are treated as endogenous. Luxembourg was excluded from the sample since the globalization measure was available only from 2000. The z-ratios in parentheses are based on standard errors robust to heteroscedasticity. Significance at the 1, 5, and 19 percent level is shown by ***, **, and *, respectively. The statistics for the Arellano-Bond tests are based on the null hypothesis of no autocorrelation of order 1 and 2 (with p-values in parentheses).

Some Additional Hypotheses

In this subsection, we examine some additional hypotheses related to reserve accumulation. First, we show that the external financing view is relevant even if there is some truth to the new mercantilism (that countries accumulate reserves as a by-product of resisting currency appreciation to guard competitiveness). In doing this, we examine whether countries with current account surpluses and appreciating currencies have accumulated more reserves than other countries. Table 4 (columns 2–3) indicates that the effects of net capital flows are largely unaffected by exchange rate appreciations and current account balances. Current account surpluses, currency appreciations, and net capital inflows after the Asian crisis are associated with a replenishing of reserve stocks in emerging markets, whereas they have been associated with a lowering of reserves in advanced economies.

Table 4.Instrumental Variables Regressions with Additional Factors
Independent VariablesCurrency Appreciation and Current Account StatusInterest Rates-Exchange Rate Nexus and Fiscal StanceForeign Asset Position
EmergingAdvancedEmergingAdvancedEmergingAdvanced
SIZE0.329***0.0670.491***0.324**0.453***0.180***
(4.83)(1.32)(4.21)(2.37)(5.25)(3.18)
σC–0.003***–0.560***–0.003***–0.651***–0.002***–0.245*
(–4.18)(–5.54)(–3.50)(–5.14)(–3.68)(–1.58)
σT0.001**0.1280.001**–0.2140.001**0.398**
(2.08)(0.62)(2.01)(–0.75)(1.97)(2.37)
IM/GDP0.168**0.137**0.0960.177**–0.0010.052
(2.11)(2.05)(1.00)(2.20)(–0.01)(0.76)
SR–0.006**–0.015***–0.008**–0.018***–0.017**–0.008**
(–2.11)(–3.63)(–1.96)(–3.10)(–3.51)(–2.20)
D(80–90)*(CF/GDP)–1.001**–0.381**–1.211**–0.194
(–2.21)(–2.23)(–2.53)(–1.41)
D(91–96)*(CF/GDP)0.240–0.610**0.285–0.3190.255–0.347
(1.09)(–2.17)(0.89)(–0.79)(1.16)(–1.50)
D(97–00)*(CF/GDP)0.341–0.410***0.355–0.499**0.792**–0.395**
(1.39)(–2.70)(1.12)(–2.05)(2.19)(–2.38)
D(01–05)*(CF/GDP)0.935***–0.560***0.934**–0.400**1.507***–0.686***
(2.63)(–4.23)(2.26)(–2.16)(3.03)(–4.49)
I(appreciation)*I(surplus)0.053***–0.014**
(4.02)(–2.24)
I(appreciation) *(1–I(surplus))–0.0100.001
(–1.00)(0.19)
Nexus0.001**0.013
(2.12)(0.70)
Fiscal balance–0.050–0.205**
(–0.016)(–2.29)
Foreign assets0.023***–0.007*
(4.05)(–1.65)
R¯210.797 [0.451]0.932 [0.723]0.773 [0.436]0.954 [0.374]0.810 [0.435]0.939 [0.854]
No. of observations407490252315320443
Hansen’s test: χ2(4)21.811 [0.771]3.172 [0.529]0.623 [0.891]0.340 [0.952]4.147 [0.386]3.861 [0.425]
Note: This table reports the IV estimation results of regression model 1 with additional factors for emerging markets and advanced countries. Regressions in columns 2-3 include an indicator function for exchange rate appreciation interacted with another indicator function for current account balance: I(appreciation)—1 if exchange rate growth > 0, and zero otherwise; and I(surplus) = 1 if the current account balance > 0, and zero otherwise. Regressions in columns 4–5 include a nexus variable defined as (1 + the money market rate minus LIBOR)/(1 + exchange rate growth) and a fiscal balance variable measured by the overall fiscal balance to GDP ratio, and are estimated for 1990–2006 since the money market rate is often unavailable in the 1980s for emerging markets. Regressions in columns 6–7 include the ratio of total foreign assets net of reserves to GDP for the period 1980–2004. All regressions use instrumental variables (IV) and include fixed-country effects. The z-ratios in parentheses are based on standard errors robust to heteroscedasticity and autocorrelation (Bartlett kernel; bandwidth = 2). ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively.

See footnote one in Table 1.

See footnote two in Table 1.

Note: This table reports the IV estimation results of regression model 1 with additional factors for emerging markets and advanced countries. Regressions in columns 2-3 include an indicator function for exchange rate appreciation interacted with another indicator function for current account balance: I(appreciation)—1 if exchange rate growth > 0, and zero otherwise; and I(surplus) = 1 if the current account balance > 0, and zero otherwise. Regressions in columns 4–5 include a nexus variable defined as (1 + the money market rate minus LIBOR)/(1 + exchange rate growth) and a fiscal balance variable measured by the overall fiscal balance to GDP ratio, and are estimated for 1990–2006 since the money market rate is often unavailable in the 1980s for emerging markets. Regressions in columns 6–7 include the ratio of total foreign assets net of reserves to GDP for the period 1980–2004. All regressions use instrumental variables (IV) and include fixed-country effects. The z-ratios in parentheses are based on standard errors robust to heteroscedasticity and autocorrelation (Bartlett kernel; bandwidth = 2). ***, **, and * indicate significance at the 1, 5, and 10 percent levels, respectively.

See footnote one in Table 1.

See footnote two in Table 1.

Second, we consider how monetary policy may have affected reserve accumulation. To this end, we include a “nexus” variable defined as the spread between domestic and foreign interest rates relative to exchange rate growth. In doing this, we control for the (overall) fiscal balance because the attractiveness of higher interest rates may be affected by the presence of fiscal liabilities (Flood and Jeanne, 2005). Table 4 (columns 4–5) shows that the nexus coefficient is positive and significant for emerging markets but insignificant for advanced economies, implying that an increase in the spread relative to exchange rate growth helped emerging markets increase reserve holdings, likely through attracting capital inflows. On the other hand, the fiscal balance coefficient is not statistically significant for emerging markets, perhaps implying that the positive effect of political-economy factors counterbalances the negative effect associated with the precautionary motive that stems from fiscal fragility. For advanced economies, the coefficient is negative and statistically significant, suggesting that the precautionary motive outweighs political-economy factors.

Third, to see how investment in foreign assets other than reserves affects reserve holdings, we used the ratio of foreign assets (net of reserves) to GDP from the Lane and Milesi-Ferretti data set. Table 4 (columns 6–7) shows that the coefficient on this ratio is positive for emerging markets but negative (significant at the 10 percent level) for advanced economies. This suggests that advanced economies, which are less liquidity-constrained than emerging markets, prefer to hold a smaller proportion of their foreign assets in reserves.

Fourth, we considered the influence of exchange rate regimes and world interest rates (for the detailed results, see Choi, Sharma, and Stromqvist, 2007). Consistent with Figure 3, emerging markets tend to hold smaller reserves under the floating regime than other regimes, whereas advanced economies tend to hold larger reserves under the intermediate regime than other regimes. The world interest rate (measured by LIBOR), a proxy for external financing costs, tends to be positively correlated with country-risk premia (for example, EMBI spreads) for emerging markets with limited access to international financial markets. For emerging markets, the world interest rate coefficient is significantly negative, suggesting that lower external financing costs in recent years may have contributed to higher reserve accumulation.

III. Concluding Remarks

This paper suggests that, despite greater financial integration and moves toward more flexible exchange rate arrangements, emerging markets have used capital inflows to build up large reserve stocks. In contrast, advanced economies, given their better access to international capital markets, have not shown this pattern. Our findings are consistent with the external financing view: emerging markets, as they integrate into the international financial system and face the risks of sudden stops in capital flows, have built up reserves after the Asian financial crisis, whereas advanced economies have balanced reserves accumulation with investments in higher yielding foreign assets.

An important issue in the context of global financial stability is how to make an assessment of comfortable reserve levels and put the savings of emerging markets to better use without compromising their financial stability. Arrangements among central banks for liquidity risk sharing (for example, through reserve pooling and currency swaps) and better access to international financial markets will help improve sovereign liquidity management in the face of potentially volatile capital flows. As emerging markets mature, they will probably have smaller reserve cushions and prefer to hold a larger proportion of their foreign assets in higher return investments. The increasing creation of sovereign wealth funds is already leading to a move in this direction, and the adoption of investment and operational norms for such funds could accelerate the process. It would also be interesting in future research to examine how the sterilization operations conducted by emerging market central banks, to manage domestic monetary conditions and prevent exchange rate appreciations, have facilitated the accumulation of reserves.

Appendix I. Country Group List

The emerging market country group (36) comprises Argentina, Brazil, Bulgaria, Chile, China, Colombia, Croatia, Czech Republic, Egypt, Estonia, Hungary, India, Indonesia, Israel, Jordan, Kazakhstan, Latvia, Lithuania, Malaysia, Mexico, Pakistan, Peru, Philippines, Poland, Romania, Russia, Slovak Republic, Slovenia, South Africa, South Korea, Taiwan Province of China, Thailand, Turkey, Ukraine, Uruguay, and Venezuela. The advanced country group (24) comprises Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Hong Kong SAR, Ireland, Italy, Japan, Luxembourg, Netherlands, New Zealand, Norway, Portugal, Singapore, Spain, Sweden, Switzerland, United Kingdom, and United States. Our country groups correspond to the groups “emerging market countries” and “industrial countries” in the IMF Research Department’s Global Data Source. Our emerging market group is similar to the “emerging market countries” except that it includes Croatia, Egypt, Jordan, Kazakhstan, and Uruguay and excludes Hong Kong SAR and Singapore. Our advanced country group includes Hong Kong SAR and Singapore because their per capita incomes have been well above the sample mean of the advanced economies for at least the last 10 years. Also, their financial integration measures have been among the highest over the period considered.

Appendix II. Descriptive Statistics

See Table A1 for descriptive statistics of key variables.

Table A1.Descriptive Statistics
VariableUnitsNMeanSDMinMax
A. Emerging markets
ReservesBillions of U.S. dollars79818.049.00.045821.5
Reserves/GDPRatio7830.1250.1040.0020.726
ln (Population)9343.1841.5820.2857.182
Exchange rate volatility9217.23216.8660.05283.322
Import/GDPRatio8010.3560.2160.0461.276
Net capital flows/GDPRatio9260.0250.056–0.2300.483
Current account/GDPRatio782–0.0140.051–0.1570.212
S&P ratingNumber (1–23)45413.73.62.023.0
GLOB (total)6874.5190.5921.7106.132
GLOB (equity)6792.6641.132–1.4975.221
Fiscal balance/GDPRatio751–0.0260.038–0.2670.099
M2/GDPRatio7890.5260.3420.0902.121
B. Advanced economics
ReservesBillions of U.S. dollars61030.266.00.0834.3
Reserves/GDPRatio6070.0960.1380.0020.747
ln (Population)6242.5821.415–1.0225.698
Exchange rate volatility6220.0960.0390.0000.146
Import/GDPRatio6240.4530.3850.0712.159
Net capital flows/GDPRatio608–0.0010.052–0.2810.156
Current account/GDPRatio6060.0050.055–0.1920.290
S&P ratingNumber (1–23)53422.01.814.023.0
GLOB (total)5525.2300.8213.3577.527
GLOB (equity)5523.8231.1361.1456.525
Fiscal balance/GDPRatio620–0.0410.441–0.2080.169
M2/GDPRatio5310.7270.3910.0002.658
Note: This table shows descriptive statistics (units; number of observations (N); mean; standard deviations (SD), minimum (Min); and maximum (Max)) of key variables for 1980–2004 for two country groups in Appendix I. Population is in million persons. Reserves are defined as gross reserves net of gold in U.S. dollars. Exchange rate (cross-section) volatility is the standard deviation of nominal exchange rate growth, and import is measured by imports of goods and services. S&P rating is Standard and Poor’s sovereign ratings (annual average). GLOB (total) is the logarithm of the two-year average of the percentage ratio of total foreign assets and liabilities to GDP, and GLOB (equity) is the logarithm of the two-year average of the percentage ratio of the sum of the portfolio equity assets and liabilities and the stock of foreign direct investment assets and liabilities to GDP. Fiscal balance is overall fiscal balance.
Note: This table shows descriptive statistics (units; number of observations (N); mean; standard deviations (SD), minimum (Min); and maximum (Max)) of key variables for 1980–2004 for two country groups in Appendix I. Population is in million persons. Reserves are defined as gross reserves net of gold in U.S. dollars. Exchange rate (cross-section) volatility is the standard deviation of nominal exchange rate growth, and import is measured by imports of goods and services. S&P rating is Standard and Poor’s sovereign ratings (annual average). GLOB (total) is the logarithm of the two-year average of the percentage ratio of total foreign assets and liabilities to GDP, and GLOB (equity) is the logarithm of the two-year average of the percentage ratio of the sum of the portfolio equity assets and liabilities and the stock of foreign direct investment assets and liabilities to GDP. Fiscal balance is overall fiscal balance.
References

    AizenmanJoshua and JaewooLee2007International Reserves: Precautionary Versus Mercantilist Views, Theory and Evidence,Open Economies Review Vol. 18 pp. 19114.

    AizenmanJoshua and NancyMarion2003The High Demand for International Reserves in the Far East: What Is Going On?Japanese and International Economies Vol. 17 pp. 370400.

    AizenmanJoshua and NancyMarion2004International Reserve Holdings with Sovereign Risk and Costly Tax Collection,Economic Journal Vol. 114 pp. 56991.

    AlmeidaHeitorMurilloCampello and MichaelS. Weisbach2004The Cash Flow Sensitivity of Cash,Journal of Finance Vol. 59 pp. 1777804.

    ArellanoManuel and StephenBond1991Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,Review of Economic Studies Vol. 58 pp. 27797.

    Bar-IlanAvnerNancyMarion and DavidPerry2007Drift Control of International Reserves,Journal of Economic Dynamics and Control Vol. 31 pp. 311037.

    Ben-BassatAvraham and DanielGottlieb1992Optimal International Reserves and Sovereign Risk,Journal of International Economics Vol. 33 pp. 34562.

    CaballeroRicardo J. and StavrosPanageas2005A Quantitative Model of Sudden Stops and External Liquidity Management,NBER Working Paper No. 11293 (CambridgeMassachusetts, National Bureau of Economic Research).

    CalvoGuillermo A.1998Capital Flows and Capital-Market Crises: The Simple Economics of Sudden Stops,Journal of Applied Economics Vol. 1 pp. 3554.

    CantorRichard and FrankPacker1996Determinants and Impact of Sovereign Credit Ratings,FRBNY Economic Policy Review Vol. 2 (Federal Reserve Bank of New YorkOctober) pp. 3753.

    ChoiWoon Gyu and DavidCook2004Liability Dollarization and the Bank Balance Sheet Channel,Journal of International Economics Vol. 64 pp. 24775.

    ChoiWoon Gyu and DavidCook Sunil Sharma and Maria Stromqvist2007Capital Flows, Financial Integration, and International Reserve Holdings: The Recent Experience of Emerging Markets and Advanced Economies,IMF Working Paper 07/151 (WashingtonInternational Monetary Fund).

    DevereuxMichael and AlanSutherland2007A Portfolio Model of Capital Flows to Emerging Markets,paper prepared for the Conference on New Perspectives on Financial Globalization held at the International Monetary FundWashingtonApril.

    DooleyMichael P.DavidFolkerts-Landau and PeterGarber2004The Revived Bretton Woods System: The Effects of Periphery Intervention and Reserve Management on Interest Rates and Exchange Rates in Center Countries,NBER Working Paper 10332 (CambridgeMassachusetts, National Bureau of Economic Research).

    DurduBoraEnriqueMendoza and MarcoTerronesforthcomingPrecautionary Demand for Foreign Assets in Sudden Stop Economies: An Assessment of the New Mercantilism,Journal of Development Economics.

    EdwardsSebastian1984The Demand for International Reserves and Monetary Equilibrium: Some Evidence from LDCs,Review of Economics and Statistics Vol. 66 pp. 495500.

    EdwardsSebastian1985On the Interest-Rate Elasticity of the Demand for International Reserves: Some Evidence from Developing Countries,Journal of International Money and Finance Vol. 4 pp. 28795.

    EdwardsSebastian2004Thirty Years of Current Account Imbalances, Current Account Reversals and Sudden Stops,IMF Staff Papers Vol. 51(Special Issue) pp. 149.

    FloodRobert and NancyMarion2002Holding International Reserves in an Era of High Capital Mobility,IMF Working Paper 02/62 (WashingtonInternational Monetary Fund).

    FloodRobert and OlivierJeanne2005An Interest Rate Defense of a Fixed Exchange Rate?Journal of International Economics Vol. 66 pp. 47184.

    FrankelJeffrey A.2005Mundell-Fleming Lecture: Contractionary Currency Crashes in Developing Countries,IMF Staff Papers Vol. 52 No. 2 pp. 149192.

    FrankelJeffrey A. and RudigerDornbusch1995The Flexible Exchange Rate System: Experience and Alternatives,in On Exchange Ratesed. byJeffrey A.Frankel (CambridgeMassachusetts, MIT Press).

    FrenkelJacob A. and BoyanJovanovic1981Optimal International Reserves: A Stochastic Framework,The Economic Journal Vol. 91 pp. 50714.

    HellerRobert H. and MohsinS. Khan1978The Demand for International Reserves under Fixed and Floating Exchange Rates,IMF Staff Papers Vol. 25 No. 4 pp. 62349.

    International Monetary Fund (IMF)2003Three Current Policy Issues in Developing Countries,World Economic Outlook (WashingtonInternational Monetary FundSeptember).

    JeanneOlivier2007International Reserves in Emerging Market Countries: Too Much of a Good Thing?Brookings Papers in Economic Activity Vol. 1 pp. 155.

    KaminskyGraciela and CarmenM. Reinhart2000On Crises, Contagion, and Confusion,Journal of International Economics Vol. 51 pp. 14568.

    KimChang-SooDavidMauer and AnnSherman1998The Determinants of Corporate Liquidity: Theory and Evidence,Journal of Financial and Quantitative Analysis Vol. 33 pp. 33559.

    LanePhilip R. and GianMaria Milesi-Ferretti2006The External Wealth of Nations Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970–2004,IMF Working Paper 06/69 (WashingtonInternational Monetary Fund).

    LipschitzLeslieMiguelMessmacher and AlexandrosMourmouras2006Reserve Adequacy: Much Higher than You Thought?” (unpublished; WashingtonInternational Monetary Fund).

    LucasRobert1990Why Doesn’t Capital Flow from Rich to Poor Countries?American Economic Review Vol. 80 pp. 926.

    MelliosConstantin and EricPaget-Blanc2006Which Factors Determine Sovereign Credit Ratings?European Journal of Finance Vol. 12 pp. 36177.

    ReinhartCarmen M. and KennethS. Rogoff2004aThe Modern History of Exchange Rate Arrangements: A Reinterpretation,Quarterly Journal of Economics Vol. 119 pp. 148.

    ReinhartCarmen M. and Kenneth S.Rogoff2004bSerial Default and the ‘Paradox’ of Rich-to-Poor Capital Flows,American Economic Review Vol. 94 pp. 538.

    RodrikDani2006The Social Cost of Foreign Exchange Reserves,NBER Working Paper No. 11952 (CambridgeMassachusetts: National Bureau of Economic Research).

    StockJames H. and MotohiroYogo2005Testing for Weak Instruments in Linear IV Regression,in Identification and Inference for Econometric Models Essays in Honor of Thomas Rothenberg (Cambridge and New YorkCambridge University Press) pp. 80108.

    Van WijnbergenSweder1990Cash/Debt Buy-Backs and the Insurance Value of Reserves,Journal of International Economics Vol. 29 pp. 12331.

Woon Gyu Choi is a senior economist at the IMF Institute; Sunil Sharma is the director of the IMF–Singapore Regional Training Institute; and Maria Strömqvist, a doctoral student at the Stockholm School of Economics, was a summer intern at the IMF Institute in 2005. The authors thank David Cook, Enrica Detragiache, Michael Devereux, Robert Flood, Brenda Gonzalez-Hermosillo, Leslie Lipschitz, Enrique Mendoza, Jaihyun Nahm, Jorge Roldos, Kwanho Shin, Evan Tanner, and an anonymous referee for helpful comments. The authors are also grateful to participants at the conference on “Korea and the World Economy” held in Seoul in 2006, and to the IMF Institute’s weekly seminar for discussions. Si-Yeon Lee provided valuable research assistance in an earlier stage of the project, and Anastasia Guscina assisted in the collection of data.

Before the 1990s, emerging market reserves fluctuated between 3 and 4 months of imports. At the end of 2005, they stood at an average of 5.8 months of imports. Even in terms of the Guidotti – Greenspan rule that countries hold reserves to cover short-term external debt (Rodrik, 2006), emerging market reserve cover has been very high: for example, the ratio was greater than 3.0 for Korea and 6.5 for China in 2005. However, recent research suggests that a broader metric should be used for assessing reserve adequacy (see Lipschitz, Messmacher, and Mourmouras, 2006).

Aizenman and Marion (2004) note that the opportunity cost variable may not be properly measured because the composition of reserves is not adequately reflected, and until the early 1990s, most emerging markets did not have market-determined domestic interest rates.

Such a reserve volatility measure, however, can be contaminated because it combines jumps in reserves owing to reserve restocking and sudden declines with speculative attacks (Flood and Marion, 2002).

The traditional buffer stock model assumes that the liquidation cost of assets for restocking reserves is known and Bar-Ilan, Marion, and Perry (2007) suggest that reserve adjustment costs are important in explaining the time path of reserve accumulation.

As creditors in sovereign debt markets have limited enforcement rights (for example, few possibilities for seizing country assets), political and credit risks play an important role in capital flows (Reinhart and Rogoff, 2004b).

Our index corresponds to the Reinhart-Rogoff classification as follows: index value 1 has categories from “no separate legal tender” to “de facto peg,” index value 2 has categories from “pre-announced crawling peg” to “managed floating,” and index value 3 has categories from “freely floating” to “freely falling.” We exclude sample observations for the category: “dual market in which parallel market data is missing.”

Aizenman and Marion (2004) find the volatility of export receipts to be statistically insignificant. We exclude this variable from our regressions because the sign of its coefficient varies depending on which other variables are included in the model and is highly sensitive to normalizations.

To account for endogeneity (high reserve holdings may reduce exchange rate variability), we also used the lagged values of exchange rate regime dummies as instruments. The results were qualitatively the same.

We ensure that the instruments are uncorrelated with the model error as follows: (i) an instrument is regressed on the endogenous regressor and country dummies to derive the component of the instrument that is uncorrelated with the regressor; (ii) the dependent variable is regressed on a set of exogenous variables including country dummies to derive the component of the dependent variable that contains the model error; and (iii) regress the filtered dependent variable from (ii) on the filtered instrument from (i), and test if the filtered instrument is statistically insignificant using a robust covariance matrix. Also, the Stock and Yogo (2005) test supports the validity of our instruments: the test rejected the null hypothesis of weak instruments at the 1 percent level for any three endogenous variables.

To account for the financial dimension of international transactions, we also use the M2-GDP ratio in place of the population variable. Like the population variable, the M2-GDP ratio has a statistically significant positive effect only for emerging markets. This result suggests that the level of monetization is not a determinant of reserve holdings for advanced economics, which already have relatively well developed financial markets.

Since the data on sovereign ratings are often not available for emerging markets in the 1980s, we also estimated regressions for the 1990–2005 period. The results were very similar to those reported in this paper.

To account for the cost of acquiring international currencies for building up reserves, we also used the EMBI and EMBI plus spreads for Latin American and non-Latin American countries for 1992–2005 as a proxy for the opportunity cost in emerging markets—the IV estimate was negative but not statistically significant.

The average total effect of net capital flows during a period can be measured by βCF + βGL*GLOBAVE, where AVE denotes the period average value. In particular, for 2001-05, the total effect based on the IV estimator in panel B is 0.127 (=—1.338 + 0.436 x 3.36) and—0.506 (=1.448—0.398x4.91) for emerging markets and advanced economics, respectively. However, its effect at very low levels of financial integration as in the 1980s can be positive for advanced economics, compared to a negative effect during the 1980s in Table 1. Thus, for advanced economics, the effect of net capital flows on reserves is mixed in the early stages of financial integration.

Other Resources Citing This Publication