This Background Paper on the United States examines the effect of fiscal deficit reduction in the context of the IMF’s multicountry simulation model, on the current account and the real exchange rate. The simulations suggest that, other things being equal, fiscal consolidation will tend to cause the real exchange rate to depreciate in the short term. The paper also estimates a long-term relationship between the real effective exchange rate for the U.S. dollar and a number of variables.

Abstract

This Background Paper on the United States examines the effect of fiscal deficit reduction in the context of the IMF’s multicountry simulation model, on the current account and the real exchange rate. The simulations suggest that, other things being equal, fiscal consolidation will tend to cause the real exchange rate to depreciate in the short term. The paper also estimates a long-term relationship between the real effective exchange rate for the U.S. dollar and a number of variables.

V. Capital Flows and Exchange Rate Volatility: Evidence from the United States 1/

1. Introduction

Under the Bretton Woods fixed exchange rate regime, a broad array of capital controls in almost all industrial countries kept cross-border capital flows at modest levels. For the United States, the capital account balance averaged a mere -0.4 percent of GDP from 1960 to 1971. However, since the breakdown of the Bretton Woods system and the dismantling of capital controls by most industrial countries, cross-border movements of capital have risen dramatically. Driven by a marked shift in private capital flows during the mid-1980s, the capital account balance of the United States swung by about 5 percentage points of GDP (Chart V-1); large, although less dramatic, swings in private capital flows have continued to characterize the 1990s (Chart V-1 and Table V-1).

CHART V-1
CHART V-1

UNITED STATES EXTERNAL ACCOUNT INDICATORS

(Four-quarter moving sum as a percent of GDP)

Citation: IMF Staff Country Reports 1995, 094; 10.5089/9781451839470.002.A005

Table V-1.

United States: Capital Account

(In billions of dollars)

article image
Source: Bureau of Economic Analysis, U.S. Department of Commerce (supplied by Haver Analytics).

First quarter at an annual rate.

These substantial fluctuations in cross-border capital flows often have been blamed for the increased exchange rate volatility of the 1980s and 1990s. 2/ Short-term speculative capital, or “hot money,” has been singled out particularly as the most volatile of capital flows and strongly linked with exchange rate instability. 3/ Concerns that large and volatile capital flows increase exchange rate volatility have led some to advocate a transaction tax on purchases and sales of foreign exchange as a way of reducing speculative foreign exchange transactions and, therefore, “excess” exchange rate volatility. 4/ These arguments stress a positive relationship between the volume of cross-border capital flows and the volatility of the exchange rate.

Another strand in the literature has focused on the effects of greater exchange rate uncertainty on the demand for domestic and foreign assets. A conclusion that emerges from this theoretical literature is that, for risk-averse agents, greater exchange rate volatility creates a bias towards domestic assets. 5/ An increase in exchange rate volatility should therefore reduce the size of net cross-border capital flows and increase the correlation between domestic saving and investment. With this in mind, several developing countries that have recently experienced a surge in capital inflows have widened their exchange rate flotation bands. The aim of this policy shift was to make short-term rates of return more uncertain for foreign investors and, thus, discourage short-term capital inflows. 1/

A different, but related, literature has focused on the effects of increased exchange rate volatility on trade and the export sector. The focus of the empirical work in this area has been to test the hypothesis that the increased exchange rate volatility of recent years has had an adverse effect on trade; the findings of these studies for the United States and some of its major trading partners have been mixed. 2/

There has been relatively little empirical work on the interaction between exchange rate volatility and capital flows, despite the growing importance of this issue and despite the fact that the capital account is the mirror image of the current account. 3/ A few studies have examined how exchange-rate uncertainty affects foreign direct investment (FDI). 4/ Some of these studies have argued that increased exchange rate uncertainty increases FDI, as manufacturers use FDI to substitute for reduced exports. There is some evidence for the United States suggesting that the increase in exchange rate uncertainty during the mid-1980s had a positive effect on gross FDI flows, although the implications for net flows is less conclusive. 5/

This chapter takes a step toward filling that gap by focusing on capital account developments. The aim is to review the basic stylized facts that characterize the U.S. capital account and to examine the links between capital flows and their potential determinants--including exchange rate and interest rate uncertainty. The next section reviews how the composition of net capital flows has evolved over time, the volatility profile of the various types of capital account transactions, and the linkages between official and private flows. Section 3 examines the relationship between exchange rate volatility and cross-border capital movements; the following section revisits an earlier literature that links capital flows to rate-of-return differentials. The patterns of official intervention in foreign exchange markets that have an effect on the capital account are reviewed in Section 5, while the last section summarizes the main findings.

2. Capital account developments and characteristics

Private capital transactions can be grouped into four main categories: FDI, portfolio investment, and net changes in U.S. banks’ and nonbank liabilities to foreigners. 1/2/ As Chart V-2 highlights, until the early 1980s, the United States consistently recorded net outflows of FDI. Net portfolio investment flows during the 1960s and 1970s were quite modest and accounted for less than half a percent of GDP. 3/ Bank flows, which presumably is the component of the capital account with the shortest maturity, accounted for most of the variation in the capital account balance until the surge in FDI and portfolio investment in the early 1980s.

CHART V-2
CHART V-2

UNITED STATES COMPONENTS OF PRIVATE CAPITAL FLOWS

(Four-quarter moving sum as a percent of GDP)

Citation: IMF Staff Country Reports 1995, 094; 10.5089/9781451839470.002.A005

Source: Bureau of Economic Analysis, U.S. Department of Commerce (supplied by Haver Analytics).

In 1994, two developments played a prominent role in accounting for the rising surplus in the capital account. First, U.S. banks borrowed heavily from their foreign branches by issuing large time deposits in the form of “senior bank notes” (see Table V-1). Secondly, acquisition of foreign securities by U.S. investors slowed markedly. This latter development, which persisted into the first quarter of 1995, reflected in part the effects of rising U.S. interest rates and in part the winding down of “emerging-markets fever” from its 1993 peak (Table V-2).

Table V-2.

United States: Net Purchases of Long-Term Foreign Securities by U.S. Investors in Selected Emerging Markets

(In billions of dollars)

article image
Source: Treasury Bulletin, various issues.

First quarter at an annual rate.

Before turning to the question of how capital flows are linked to other macroeconomic variables, it is useful to examine how the various components of the capital account interact and how their volatility has evolved over time. There are several observations that stand out from Chart V-3. First, the volatility of all the components of private capital flows, even relative to nominal GDP, has increased over time, with the upturn becoming more pronounced since the mid-1980s. 4/ Second, bank flows are considerably more volatile than other private flows. Third, changes in net foreign official reserves are far more volatile than U.S. official reserves, suggesting a much more active role in foreign exchange market intervention by foreign central banks than by the U.S. monetary authorities. Lastly, there is little difference between the volatility of FDI and that of portfolio investment. This last observation runs counter to the commonly-held view that FDI is a “more stable” form of foreign Investment while portfolio investment is thought to have more “hot money” characteristics. 1/

CHART V-3
CHART V-3

UNITED STATES VOLATILITY OF NET CAPITAL FLOWS

(Mean absolute quarterly deviation as a percent of GDP)

Citation: IMF Staff Country Reports 1995, 094; 10.5089/9781451839470.002.A005

Sources: Bureau of Economic Analysis, U.S. Department of Commerce (supplied by Haver Analytics); and staff estimates.

Simple pairwise correlations between the various components of the capital account provide information on the interaction between different types of flows. For instance, Table V-3 suggests that changes in foreign official holdings of reserves are significantly and negatively correlated with private capital flows, with the total as well as with two of the three largest components; this may reflect a leaning-against-the-wind policy in foreign exchange markets by foreign central banks. By contrast, U.S. official reserves are not significantly correlated with any component of the private capital account. The positive and significant correlation between portfolio investment and FDI also suggests that although conceptually distinct, these two types of flows may be difficult to distinguish in practice. The high correlation between the private capital account balance and bank flows reflects the relative importance of bank flows in driving the capital account throughout most of the sample period.

Table V-3.

United States: The Interaction of Private and Official Capital Flows--Contemporaneous Pairwise Correlations 1972:1-1995:1

article image
Notes: Standard errors are in parentheses. The private capital account balance, column (1), equals the sum of its components, columns (2) through (5).

Significant at the 5 percent confidence level or higher.

The next section examines whether there is any systematic relationship between exchange rate volatility and capital flows.

3. Exchange rate volatility, excess returns, and capital flows

If large and volatile capital flows increase exchange rate volatility, then one would expect that a measure of exchange rate volatility would be positively correlated with the absolute level of flows. That is, exchange rate volatility would ensue from both large inflows and large outflows. If, alternatively, the greater exchange rate volatility creates a bias towards domestic assets, then exchange rate volatility should be negatively correlated with the absolute level of capital flows. As regards official capital flows, one rationale for central bank intervention in foreign exchange markets is to insure orderly market conditions. Because volatility rises in periods of turmoil, then a positive correlation may be expected between absolute changes in reserves and exchange rate volatility.

To obtain a measure of exchange rate volatility, daily exchange rate data is employed; volatility is calculated as the mean absolute daily percent change in the value of the dollar against a particular currency (i.e., the yen or deutsche mark) or a basket of currencies over the month (or the quarter). 2/ The exchange rate index used is that reported by the Federal Reserve. 3/ Chart V-4 plots the monthly volatility measures as well as a 12-month moving average. In all three panels the latter half of the 1980s and the 1990s reveal a pattern of higher peaks and higher troughs than during the 1970s. 1/

CHART V-4
CHART V-4

UNITED STATES EXCHANGE RATE VOLATILITY

(Moving average of mean absolute daily percent change)

Citation: IMF Staff Country Reports 1995, 094; 10.5089/9781451839470.002.A005

Table V-4 reports the contemporaneous correlations for the various components of the capital account with exchange rate volatility. The total capital account balance (including the statistical discrepancy) is positively and significantly correlated with exchange rate volatility. Although other factors are not controlled for and no “causal” interpretation can be drawn from a simple correlation, this positive relationship would appear to be in line with the hypothesis that larger capital flows result in more volatile exchange rates. With the exception of FDI, all the components of the private capital account are positively correlated with the volatility measure. Among the individual components, however, the strongest correlation is with portfolio flows. Neither U.S. nor foreign official reserves appear related to exchange rate volatility. As shown in Section 5, this absence of a systematic relationship may result from the use of quarterly data, which may be too aggregated to detect systematic patterns in official flows.

Table V-4.

United States: Exchange Rate Volatility and Capital Flows--Contemporaneous Pairwise Correlations

1972:1-1995:1

article image
Notes: Standard errors are in parentheses. The private capital account balance, column (2), equals the sum of its components, columns (3) through (6). The capital account balance, column (1), equals the sum of columns (2), (7), and (8) plus the statistical discrepancy.

Significant at the 5 percent confidence level or higher.

To examine whether there was any kind of temporal causal relationship between exchange rate volatility and the volatility of capital flows, a simple bivariate vector autoregression (VAR) was estimated in these variables. 2/ In one set of VARs the exogenous variable was limited to a constant (these results are reported in Table V-5), in a second set of VARs a time trend was included as well. The results of both sets of estimates were qualitatively the same. None of the categories of capital flows helped explain future exchange rate volatility (i.e. no causal relationship was detected). However, past exchange rate volatility did help explain portfolio flows and nonbank flows, but not in the way predicted by theory. The results presented in Table V-5 suggest that an increase in exchange rate volatility raises the volatility of portfolio investment and nonbank flows. 3/

Table V-5.

United States: Exchange Rate Volatility and Capital Flows--Causality Tests

1972:1-1995:1

article image

Significant at the 5 percent confidence level or higher.

As noted earlier, the prediction of models with risk-averse agents and exchange rate uncertainty is the opposite: capital flows (both inflows and outflows) would be smaller when exchange rate volatility increases. Of course, these results could simply reflect the influence of an omitted variable. For instance, if during periods in which exchange rate volatility is high rates-of-return differentials tend to be wider, then the higher volatility of portfolio flows is due to larger desired adjustments in the agents’ portfolios and not to increased exchange rate volatility per se.

The next section examines the relationship between the various categories of capital flows and some of its determinants.

4. Excess returns, interest rate volatility, and capital flows

In a traditional portfolio-balance model where domestic and foreign assets are imperfect substitutes, capital flows are explained by excess returns,

ii*Δse

where i stands for the domestic interest rate, i* is a comparable foreign interest rate, and the last term denotes the expected change in the spot exchange rate. If rational expectations are assumed, then the expectations term may be replaced by the ex post changes in the exchange rate. Alternatively, expectations could be measured directly by using survey data. 1/ Risk considerations also will affect the direction of capital flows. If domestic interest rates are more volatile than foreign interest rates, other things equal, more capital will flow abroad.

The interest rate on a 3-month Treasury bill is used as a measure of i; an index of 3-month foreign treasury bills that matches the country composition and uses the same weights as the exchange rate index is the measure of i*; and the actual change in the multilateral spot exchange rate index between the current period (t) and the period in which the t-bill matures (t+3) is our measure of Δse.

The contemporaneous correlations between excess returns and the components of the capital account are reported in the first row of Table V-6. Both the total balance and the balance of the private capital account show a positive and significant correlation with excess returns. The two components of private flows that show the strongest links with the excess return measure are bank flows and portfolio investment. The absence of a significant correlation for FDI is due, in all probability, to our use of a short-term interest rate differential; as argued elsewhere, a more relevant rate of return may be a bond yield, which may more closely reflect the rate of return on capital. 2/ Official flows show no correlation with excess returns. This last observation is not surprising, since the motivation for central bank intervention has less to do with taking advantage of existing arbitrage opportunities than with stabilizing exchange markets in times of turmoil or achieving some exchange rate objective.

Table V-6.

United States: Capital Flows, Excess Returns, and Interest Rate Volatility--Contemporaneous Pairwise Correlations

1972:1-1995:1

article image
Notes: Standard errors are in parentheses. The private capital account balance, column (2) equals the sum of its components, columns (3) through (6). The capital account balance, column (1), equals the sum of columns (2), (7), and (8) plus the statistical discrepancy.

Significant at the 5 percent confidence level or higher.

To assess the potential impact of domestic and foreign interest rate uncertainty on cross-border capital movements, first a measure of interest rate volatility was constructed along the lines of the measure used for the exchange rate. Daily data were used to construct the monthly and quarterly indices. Chart V-5 plots the volatility indices for the domestic and foreign rate, as well as a comparable index for a U.S. 10-year Treasury note (not used in the econometric analysis). Since what should matter for the decision whether to invest domestically or abroad is relative volatilities, the bottom left-hand panel of Chart V-5 shows the ratio of domestic to foreign volatility.

CHART V-5
CHART V-5

UNITED STATES INDICATORS OF INTEREST RATE VOLATILITY

(Moving average of mean absolute daily deviation)

Citation: IMF Staff Country Reports 1995, 094; 10.5089/9781451839470.002.A005

Sources: Board of Governors of the Federal Reserve System; and staff estimates.1/ Index of foreign 3-month treasury bilis. The country composition and weights are the same as those used in the exchange rate Index.

Several features are worth noting. First, as is well known, domestic interest rate volatility rose sharply during the period when the Federal Reserve targeted nonborrowed reserve (October 1979-1982). Second, domestic short-term interest rate volatility declined in the 1990s to its lowest level during the post-1972 period, while the volatility of long rates remains consistently higher in the 1990s than throughout most of the 1970s. Third, domestic short-term interest rates tend to be more volatile than foreign ones.

The third row of Table V-6 reports the correlations between our proxy for relative interest rate volatilities and the measures of capital flows. Bank flows, which are predominantly short-term flows, show a significant and negative correlation. Perhaps, not surprisingly, for banks not only the level but also the variability of interest rates matter.

5. Foreign official Intervention

To examine further what drives foreign official capital flows and, in particular, what factors lead to foreign official foreign exchange intervention, we incorporate monthly data (which are available since mid-1989) on foreign official assets held at the Federal Reserve in the analysis. Foreign official assets held at the U.S. Federal Reserve provide a useful proxy for foreign reserve holdings of U.S. dollars since central banks routinely invest dollar reserves in U.S. Treasury securities, which are held with the Federal Reserve. 1/ The data are used to revisit some of the questions as to how foreign reserve holdings of U.S. dollars interact with exchange rate volatility and exchange rate trends.

While, as noted earlier, no systematic relationship (contemporaneous or causal) was detected between exchange rate volatility and the absolute value of changes in reserves using quarterly data, such a relationship is present at the monthly frequency. Consistent with the view that intervention increases when exchange markets become volatile, the correlation between the two variables is positive and significant (Table V-7, top panel). The relationship is contemporaneous, as no temporal causal patterns are evident.

Table V-7.

United States: Foreign Official Intervention, Exchange Rate Volatility, and Changes in the Dollar

1989:6-1995:6

article image
Notes: An increase in the dollar index denotes an appreciation. The optimum number of lags according to the Akaike and Schwarz criteria is two. Standard errors are in parentheses.

Besides intervening to calm disorderly markets, intervention by many foreign central banks often takes the form of systematically leaning against the wind, suggesting a negative relationship between changes in the dollar and changes in reserves. This systematic negative relationship shows up as both a significant contemporaneous correlation (Table V-7, top panel) and as a causal relationship from exchange rate changes to reserve changes (Table V-7, bottom panel). This pattern of intervention was particularly evident in early 1995; as the dollar came under pressure against the yen and the deutsche mark, Federal Reserve holdings of government securities for foreign official institutions rose sharply (Chart V-6). 1/

CHART V-6
CHART V-6

UNITED STATES FOREIGN OFFICIAL ASSETS AND THE DOLLAR

Citation: IMF Staff Country Reports 1995, 094; 10.5089/9781451839470.002.A005

Sources: Board of Governors of the Federal Reserve System; and staff estimates.1/ A decline in the index denotes a depreciation. Index of weighted-average exchange value of the U.S. dollar against the currencies of ten industrial countries.

6. Concluding remarks

With regard to the patterns and determinants of private capital flows in the U.S. balance of payments, several empirical regularities emerge from the previous analysis.

  • All the major categories of private capital flows, as well as exchange rates, have tended to become more volatile over time.

  • There is little evidence to suggest that foreign direct investment is any less volatile than other components of the capital account. Indeed, the positive and significant correlation between portfolio investment and FDI suggests that, although conceptually distinct, these two types of flows may be difficult to distinguish in practice.

  • The volatility of some types of capital flows, particularly portfolio investment and nonbank flows, is positively associated with exchange rate volatility.

  • Bank flows and portfolio investment appear to be significantly linked with short-term interest rate differentials, more so than other components of the capital account.

  • There appears to be a relationship between domestic and foreign short-term interest rate volatility and bank flows; other things equal, a rise in the volatility of domestic interest rates relative to the volatility of foreign rates tends to increase capital outflows.

As to the characteristics and determinants of official capital flows, the main observations that are suggested by the foregoing analysis are that:

  • The behavior of foreign official reserves is very different from the behavior of U.S. official reserves.

  • Foreign official reserve changes appear to be driven in part by a leaning against the wind strategy in foreign exchange markets. Specifically, foreign official reserve changes are negatively and significantly correlated with private capital flows and with changes in the exchange value of the dollar.

  • The evidence also suggests that foreign central banks’ intervention increases with exchange rate volatility.

References

  • Anderson, Margot, and David Alan Grier, Robust, Non-Parametric Measures of Exchange Rate Variability,Applied Economics, Vol. 24 (1992), pp. 9518.

    • Search Google Scholar
    • Export Citation
  • Asea, Patrick K., and Colin Rose, “Sharks, Speculative Attacks and the Hump-Shaped Distribution,” (University of California, Los Angeles, mimeographed, 1995).

    • Search Google Scholar
    • Export Citation
  • Bacchetta, Philippe, and Eric van Wincoop, Net Capital Flows Under Exchange Rate and Price Volatility,Studienzentrum Gerzensee Working Paper No. 94.03, (1994).

    • Search Google Scholar
    • Export Citation
  • Board of Governors of the Federal Reserve System, Federal Reserve Bulletin, Vol. 64 (August 1978), p. 700.

  • Caves, Richard E., Exchange-Rate Movements and Foreign Direct Investment in the United States,Harvard Institute of Economic Research Discussion Paper No. 1383, (May 1988).

    • Search Google Scholar
    • Export Citation
  • Claessens, Stijn, Michael P. Dooley, and Andrew Warner, Portfolio Capital Flows: Hot or Cool?,” in Portfolio Investment in Developing Countries, Stijn Claessens and Sudarshan Gooptu eds., World Bank Discussion Papers No. 228, (Washington: The World Bank, December 1993).

    • Search Google Scholar
    • Export Citation
  • Cushman, David, O., Real Exchange Rate Risk, Expectations, and the Level of Direct Investment,Review of Economics and Statistics, Vol. 67 (May 1985), pp. 297308.

    • Search Google Scholar
    • Export Citation
  • Cushman, David, O., Exchange-Rate Uncertainty and Foreign Direct Investment in the United States,Weltwirtschaftliches Archive, Vol. CXXIV (1987), pp. 32235.

    • Search Google Scholar
    • Export Citation
  • Eichengreen, Barry, Andrew K. Rose, and Charles Wyplosz, Exchange Market Mayhem: The Antecedents and Aftermath of Speculative Attacks, (prepared for the Economic Policy Panel, April 21-22, 1995).

    • Search Google Scholar
    • Export Citation
  • Eichengreen, Barry, Andrew K. Rose, and Charles Wyplosz, James Tobin, and Charles Wyplosz, Two Cases for Sand in the Wheels of International Finance,The Economic Journal, Vol. 105 (January 1995), pp. 16272.

    • Search Google Scholar
    • Export Citation
  • Dominguez, Kathryn, and Jeffrey Frankel, “Foreign Exchange Intervention: An Empirical Assessment,” in On Exchange Rates, (Cambridge: MIT Press, 1993), pp. 32746.

    • Search Google Scholar
    • Export Citation
  • Frankel, Jeffrey A., International Capital Mobility and Exchange Rate Volatility,” in International Payments Imbalances in the 1980s, Norman S. Fieleke ed., Conference Series No. 32, (Boston: Federal Reserve Bank of Boston, 1988a), pp. 16288.

    • Search Google Scholar
    • Export Citation
  • Frankel, Jeffrey A., “International Capital Flows and Domestic Economic Policies,” in The United States and the World Economy, Martin Feldstein ed., (Chicago: University of Chicago Press, 1988b).

    • Search Google Scholar
    • Export Citation
  • Obstfeld, Maurice, and Kenneth Rogoff, The Exchange Rate in Monetary Policy,Journal of Economic Perspectives, (forthcoming).

  • Qian, Ying, and Panos Varangis, Does Exchange Rate Volatility Hinder Export Growth? Additional Evidence,Empirical Economics, Vol. 19 (1994), pp. 37196.

    • Search Google Scholar
    • Export Citation
  • Reinhart, Carmen M., and Steven Dunaway, Monetary, Exchange Rate, and Fiscal Policy Responses to Previous Surges of Capital Flows,” in International Capital Markets: Developments. Prospects, and Policy Issues, (Washington: International Monetary Fund, 1995).

    • Search Google Scholar
    • Export Citation
  • Reinhart, Carmen M., and Steven Dunaway, and R. Todd Smith, Controls on Capital Flows: The Experience with Quantitative Measures and Capital Flow Taxation,” in International Capital Markets: Developments. Prospects, and Policy Issues, (Washington: International Monetary Fund, 1995).

    • Search Google Scholar
    • Export Citation
  • Turner, Philip, Capital Flows in the 1980s: A Survey of Major Trends,BIS Economic Papers No. 30, (Basle: Bank for International Settlements, April 1991).

    • Search Google Scholar
    • Export Citation
  • U.S. Department of Commerce, Bureau of Economic Analysis, The Balance of Payments of the United States: Concepts. Data Sources, and Estimating Procedures, (Washington: U.S. Government Printing Office, May 1990).

    • Search Google Scholar
    • Export Citation
  • Wihlborg, Clas, Currency Risks in International Financial Markets,Princeton Studies in International Finance, No. 44 (1978).

  • Wrightson Associates, Money Market Observer, (March 31, 1995).

1/

Prepared by Carmen M. Reinhart.

4/

See, for instance, Eichengreen, Tobin, and Wyplosz (1995); these authors point out that, in order to be effective, the tax would have to apply to all jurisdictions and the rate would have to be equalized across markets. See also Reinhart and Smith (1995) for a recent review of this literature.

2/

As Qian and Varangis (1994) discuss in a recent review of this literature, some studies have found supporting evidence while others have not.

3/

Exceptions are Frankel (1988a and 1988b).

1/

Official capital flows consist of changes in U.S. official reserve assets and changes in foreign official assets in the United States; the statistical discrepancy represents the sum of credits and debits in the balance of payments statement with the sign reversed. See U.S. Department of Commerce (1990) for details.

2/

The capital account, which attempts to measure transactions in financial assets between residents and nonresidents, may be particularly vulnerable to measurement error. Hence, any results and inference drawn from this data must be interpreted with some caution.

3/

These trends also became evident in other industrial countries, see Turner (1991).

4/

Volatility is measured as the mean absolute quarterly change over a rolling five-year period. All capital flow measures are expressed as a percent of GDP; these ratios are stationary series with well-defined variances.

1/

Examining the time series properties of various components of the capital account for five industrial countries (including the United States) and five developing countries, Claessens, Dooley, and Warner (1994) arrive at a similar conclusion.

2/

For a discussion of the advantages of using this measure of volatility see Anderson and Grier (1992).

3/

For the details of the construction of this index, see Board of Governors of the Federal Reserve (1978).

1/

The volatility measure was regressed against a constant and a time trend and the coefficients estimated using the Generalized Method of Moments (GMM) to allow for the presence of a moving average process in the error term as well as for more general forms of heteroskedasticity. In all cases, the time trend was positive and statistically significant. However, using comparable daily data through 1989 and nonparametric tests, Anderson and Grier (1992) do not find significant evidence of higher exchange rate volatility during the post-1979 period.

2/

The absolute value of the various components of capital flows is the measure of volatility used.

3/

Impulse responses, not shown in this paper, illustrate how an increase in exchange rate volatility raises the volatility of capital flows.

1/

See, for instance, Dominguez and Frankel (1993).

1/

While this series does include holdings by foreign official institutions other than central banks, those amounts account for a relatively small proportion of the total. The series excludes deposits and U.S. Treasury securities held for international and regional organizations.

1/

See also Wrightson Associates (March 31, 1995).

United States: Background Papers
Author: International Monetary Fund
  • View in gallery

    UNITED STATES EXTERNAL ACCOUNT INDICATORS

    (Four-quarter moving sum as a percent of GDP)

  • View in gallery

    UNITED STATES COMPONENTS OF PRIVATE CAPITAL FLOWS

    (Four-quarter moving sum as a percent of GDP)

  • View in gallery

    UNITED STATES VOLATILITY OF NET CAPITAL FLOWS

    (Mean absolute quarterly deviation as a percent of GDP)

  • View in gallery

    UNITED STATES EXCHANGE RATE VOLATILITY

    (Moving average of mean absolute daily percent change)

  • View in gallery

    UNITED STATES INDICATORS OF INTEREST RATE VOLATILITY

    (Moving average of mean absolute daily deviation)

  • View in gallery

    UNITED STATES FOREIGN OFFICIAL ASSETS AND THE DOLLAR