This Selected Issues paper estimates an equilibrium path for South Africa’s real effective exchange rate. The paper briefly describes the dynamics of the real exchange rate and its determinants. It investigates the presence of a long-term relationship between the real exchange rate and certain explanatory variables, estimates the speed at which the real exchange rate converges toward its equilibrium level, and derives measures for the equilibrium real exchange rate. The paper also examines the real money demand, consumer prices, and the real exchange rate in South Africa.

Abstract

This Selected Issues paper estimates an equilibrium path for South Africa’s real effective exchange rate. The paper briefly describes the dynamics of the real exchange rate and its determinants. It investigates the presence of a long-term relationship between the real exchange rate and certain explanatory variables, estimates the speed at which the real exchange rate converges toward its equilibrium level, and derives measures for the equilibrium real exchange rate. The paper also examines the real money demand, consumer prices, and the real exchange rate in South Africa.

VIII. Sovereign Risk Spreads Under Inflation-Targeting83

144. Sovereign risk spreads have become an important and widely used indicator for assessing macroeconomic conditions and the external vulnerability of emerging market countries. Understanding the determinants of sovereign risk spreads is also an important prerequisite for designing and implementing economic policies aimed at reducing debt service payments and smoothing the path of public expenditure by sustaining or enhancing access to capital markets. In addition, risk spreads are a key determinant of long-term interest rates. Therefore, the information these spreads provide about financial markets' perceptions of economic policy can be beneficial for future economic policy decisions.

145. During past episodes of emerging market crisis in South Africa and elsewhere, sharp currency depreciations have generally been accompanied by rising sovereign risk spreads.84 In contrast, the sharp depreciation of the rand in late 2001 was accompanied by a narrowing of South African risk spreads. Hence, economic and financial market developments during the most recent depreciation raise questions about the usefulness of sovereign risk spreads as vulnerability indicators and their interpretation. Answering these questions requires an analysis of the determinants of South African risk spreads.

146. This section assesses the usefulness of sovereign risk spreads as a vulnerability indicator after the introduction of inflation targeting in South Africa in February 2000 and describes the implications for economic policy. An assessment of the determinants of sovereign risk spreads suggests that spreads reflect the performance of monetary policy vis-à-vis inflation targets. It also suggests that, with a credible commitment of the South African Reserve Bank (SARB) to its inflation target, rand-denominated spreads may, in fact, become a better indicator of vulnerability than U.S. dollar-denominated spreads.

147. After a brief review of the literature, the section compares the currency depreciations of 1998 and 2001 with regard to sovereign risk spreads behavior and the different macroeconomic responses. The section then takes a closer look at movements in sovereign risk spreads after the introduction of inflation targeting and empirically investigates the determinants of sovereign risk spreads in inflation-targeting countries, including South Africa. The last part of the section suggests implications for economic policy.

A. Brief Review of the Literature

148. Most of the literature on the presence and determinants of sovereign risk spreads emerged during the past decade, motivated by financial crises in emerging market countries and in the European Monetary System (EMS). Previous research was largely focused on risk premia within currency unions,85 due to both methodological problems in comparing risk premia across currencies and the relatively low level of foreign bond financing by emerging market countries. Alesina and others (1992) were among the first to present an analysis aimed at a truly international comparison of sovereign risk spreads. The study relates levels of public debt to sovereign risk premia through the emergence of a “confidence crisis” and finds some empirical evidence in support of its hypotheses among OECD countries, but the comparability of results for individual countries remains limited due to the measurement of sovereign risk spreads across national currencies.

149. The more recent availability of reliable time-series data on foreign-currency bond yields of emerging market countries has led to a larger number of studies on the determinants of sovereign risk spreads of these countries. Arora and Cerisola (2001) examine the sovereign risk spreads in several emerging market countries. They conclude that country-specific variables, such as net (or gross) foreign assets, public external debt, and fiscal deficits, explain a significant proportion of sovereign risk spread volatility. Studies by Eichengreen and Mody (1998), Kamin and von Kleist (1999), and Min (1998) obtain similar results.

150. Arora and Cerisola (2001) also find the stance and predictability of U.S. monetary policy to be an important determinant of sovereign risk spreads in emerging market countries. Their findings indicate that the level of U.S. interest rates has a direct positive effect on sovereign bond spreads. In contrast, earlier analyses by Dooley, Fernandez-Arias, and Kletzer (1996) and by Calvo, Leiderman, and Reinhart (1996) found a significant negative impact of industrial-country interest rates on sovereign risk spreads in emerging market countries, while Kamin and von Kleist (1999) found no significant impact at all.

151. Kamin and von Kleist (1999) find important regional differences in sovereign risk spreads of emerging market countries, even after controlling for risk and maturity. They include credit ratings to explain risk spreads on both bonds and loans and obtain significant results. However, other empirical results on the value of credit ratings in explaining sovereign risk spreads are mixed. While findings by Cantor and Packer (1996) provide empirical support for credit ratings assigned by both Standard & Poor’s and Moody’s as a determinant of sovereign risk spreads, other studies reached different conclusions. The infrequent change in sovereign credit ratings suggests that their value is limited for explaining changes in sovereign risk spreads on the basis of monthly data.

B. Spread Behavior During the Currency Depreciations of 1998 and 2001

152. Spreads are most commonly and accurately measured as the difference in yields between U.S. dollar-denominated South African government bonds and U.S. Treasury bonds of similar maturity. The spread reflects different risk factors, including credit risk, portfolio risk,86 and illiquidity risk. Since a debtor’s share in the bond market and the liquidity of secondary market trading generally remain stable over shorter periods, the market’s valuation of the default risk largely determines movements in spreads.87 South Africa has issued four U.S. dollar-denominated bonds, with different maturities, that can be used to construct sovereign risk spreads. This analysis uses the bond maturing in 2017, as credit risk increases with time and the secondary market for this bond is sufficiently liquid.

153. South African spreads moved in opposite directions during the sharp depreciations of 1998 and 2001. They increased by almost 400 basis points between end-April and end-August 1998, while the rand depreciated by 28 percent in nominal terms against the U.S. dollar. In contrast, the rand depreciated by 26 percent against the U.S. dollar between end-September and end-December 2001, but U.S. dollar-denominated South African bond spreads narrowed by about 40 basis points (Figures VIII.1 and VIII.2). The two currency depreciations were also associated with very different macroeconomic outcomes.88

Figure VIII.1.
Figure VIII.1.

Nominal Rand/U.S. dollar Exchange Rate

(In percent change from beginning of year)

Citation: IMF Staff Country Reports 2003, 018; 10.5089/9781451840995.002.A008

Figure VIII.2.
Figure VIII.2.

Sovereign Risk Spreads

(In basis point change from beginning of year)

Citation: IMF Staff Country Reports 2003, 018; 10.5089/9781451840995.002.A008

154. The fall in sovereign risk spreads during the 2001 currency depreciation reflects low and decreasing external vulnerability. Lower external vulnerability during 2001, compared with 1998, largely stemmed from the implementation of sound macroeconomic policies, in particular a significant reduction in the net open forward position (NOFP)—of more than two-thirds since December 2000—and strong fiscal performance.89 During the 1998 currency depreciation, the high and increasing NOFP was a significant factor behind the increase in external vulnerability and sovereign risk spreads.90

155. The shift in monetary policy from supporting exchange rate stability to committing to an inflation target has allowed for the significant reduction in the NOFP and the related decrease in external vulnerability. During the 1998 currency depreciation, the SARB used sterilized intervention to support the rand, leading to an increase in the NOFP from US$13 billion in April 1998 to US$23 billion in October 1998. Since committing itself to inflation as the primary target of monetary policy in February 2000 in the context of its inflation-targeting strategy, the SARB has refrained from supporting the rand. This change in policy has enabled the SARB to significantly reduce the NOFP from US$9 billion at end-March 2001 to less than US$3 billion at end-May 2002.

156. After the adoption of an inflation-targeting policy, changing risk perceptions may have been translated to a larger extent into exchange rate fluctuations and to a lesser extent into sovereign risk spread movements, suggesting a decline in the usefulness of U.S. dollar-denominated sovereign risk spreads as a vulnerability indicator. Evidence from average monthly volatilities, which are calculated as standard deviations, shows a significant increase in exchange rate volatility by 78 percent from the 1998 depreciation to the 2001 episode, while the volatility of sovereign bond spreads declined by 7 percent (Table VIII.1).

Table VIII. 1.

Developments in Exchange Rate and Sovereign Risk Spread Volatility

article image
Sources: Datastream; and IMF staff calculations.

157. In sum, the decline in sovereign risk spreads during the currency depreciation of 2001—and in their volatility relative to the 1998 episode—suggests not only reduced vulnerability, but also a somewhat lower reliability of spreads as a vulnerability indicator after the introduction of a credible inflation targeting policy. This is because changing risk perceptions generally get reflected less in sovereign risk spreads when monetary policy becomes less discretionary, as binding monetary policy rules deprive governments of the option of inflationary financing during episodes of financial distress.91 If inflation is continuously kept at low rates within a defined target range, dampened inflation expectations will mitigate downward pressures on the nominal exchange rate that would otherwise raise sovereign risk spreads.

C. A Closer Look at the Sovereign Risk Premium

158. Movements in sovereign risk spreads of emerging market countries are highly correlated with changes in global risk aversion (Figure VIII.3). J.P. Morgan Chase’s liquidity and credit premia index (LCPI) provides a comprehensive quantification of global risk aversion. It captures not only credit spreads, but also the liquidity premia demanded in U.S. financial markets,92 which are considered to be an important indicator of risk appetite. It is, therefore, somewhat broader-based than other measures, such as, for example, an emerging market bond index.93 Figure VIII.3 shows developments in the LCPI and in sovereign risk spreads between January 2000 and April 2002. Spreads are shown for South Africa and for a group of emerging market countries with ratings similar to that for South Africa.94

Figure VIII.3.
Figure VIII.3.

Global Risk Aversion and Sovereign Risk Spreads

(January 2000–April 2002)

Citation: IMF Staff Country Reports 2003, 018; 10.5089/9781451840995.002.A008

159. South Africa’s risk spread was higher than the average risk spread for the group of countries in the same rating category for the entire period shown in Figure VIII.3, and the difference between the two spreads varied considerably, ranging from 7 basis points in February 2000 to 122 basis points in November 2000. The difference between the spreads increased in particular at times of rising global risk aversion, except for the latest rise in the LCPI, during which the gap actually narrowed.

160. The credibility of South Africa’s monetary policy—or, more precisely, the achievement of the inflation target—seems to be among the main forces driving movements in the gap between the spreads of South Africa and the peer group of countries (Figure VIII.4). Figure VIII.4 shows the evolution of the gap between the spreads and the deviation of the actual inflation outcome (CPIX) from the upper end of the target range of 3–6 percent for 2002.95 Both series show considerable co-movements over the period from February 2000 (when inflation targeting was introduced) until April 2002. The high correlation of more than0.7 between the inflation gap and the gap in sovereign risk spreads suggests that South Africa may be paying a premium—relative to similarly rated countries—on its U.S. dollar-denominated debt, as long as its monetary policy has not yet successfully implemented the new regime and actual inflation has not fallen into the defined target range.

Figure VIII.4.
Figure VIII.4.

Inflation Performance and the Difference in Sovereign Risk Spreads, January 2000–April 2002

(In percentage points)

Citation: IMF Staff Country Reports 2003, 018; 10.5089/9781451840995.002.A008

161. As U.S. dollar-denominated spreads declined under the operation of the inflation-targeting regime, rand-denominated spreads may contain more information about sovereign default risks and external vulnerability. The assumption of a higher importance of rand-denominated spreads under inflation targeting stems largely from the loss of access to inflationary finance and, therefore, a higher risk of outright default.96 Indeed, the average yield on South Africa’s long-term rand-denominated government bonds increased by as much as 68 basis points relative to the yield of rand-denominated World Bank bonds between September 2000 and April 2002 (Figure VIII.5).97

Figure VIII.5.
Figure VIII.5.

Rand-Denominated Sovereign Risk Spreads

(Basis point change since September 2000)

Citation: IMF Staff Country Reports 2003, 018; 10.5089/9781451840995.002.A008

162. Until January 2002, the World Bank Group paid a higher yield than the South African government on rand-denominated bonds, despite a significantly better credit rating for rand-denominated debt (World Bank: Aaa/AAA; South Africa: A2/A-). While South Africa may be better known to investors in the market for rand-denominated bonds than the World Bank and its bonds may have more liquid secondary markets, South Africa’s relatively low financing cost may be explained by taking into account the greater possibility to resort to inflationary finance under the previous monetary policy regime of money-supply rules. With the implementation of the inflation-targeting framework, the relative yield spread has widened in favor of the World Bank, thereby better reflecting the credit ratings of the two debtors (Figure VIII.5).

D. Results on the Determinants of Sovereign Risk Spreads from Econometric Analysis

163. The econometric analysis uses panel data on four inflation-targeting countries98 to conduct pooled regression analysis on the determinants of sovereign risk spreads, using monthly data on the LCPI, the gap between actual inflation and the targeted inflation rate, the ratio of reserves to imports, total external debt as a percentage of GDP, net foreign assets as a percentage of GDP, the fiscal balance as a percentage of GDP, and the credit rating assigned by rating agencies as explanatory variables. While the LCPI is used to proxy global risk aversion, the credit rating is supposed to capture the sum of country-specific risk components. The estimation period ranges from February 2000, when inflation targeting was introduced by the SARB, to December 2001.

164. The summary of estimation results shows that sovereign risk spreads under inflation targeting are driven by the gap between the actual inflation and the targeted rate of inflation, changes in global risk aversion, the ratio of reserves to imports, and credit ratings (Table VIII.2).99 The results suggest that each percentage point by which actual inflation exceeds the target range is reflected in a rise in the sovereign risk premium of 13–15 basis points on average across the sample. Also, a one-month increase in the import coverage of international reserves lowers the spread considerably, by between 43 and 88 basis points, depending on the specification of the model. In contrast, a rise in global risk aversion by 10 LCPI index points leads to an increase in the risk premium by only about 1–2 basis points. Overall, the results suggest that the performance of monetary policy vis-à-vis its inflation target is among the most important factors driving foreign-currency-denominated sovereign risk spreads.

Table VIII.2.

Determinants of Sovereign Bond Spreads under Inflation Targeting

article image
Source: IMF staff estimates.Explanation: ***, **, and * indicate statistically significant results at the 1,5, and 10 percent confidence level respectively; t-statistics are reported in parentheses.

165. Not surprisingly, many conventional determinants of sovereign risk spreads, such as external debt, net foreign assets, or the fiscal deficit as percentages of GDP have no explanatory power for sovereign risk spreads. The failure of these variables to influence U.S. dollar-denominated risk spreads over the sample period confirms the decline in usefulness of foreign-currency spreads as a vulnerability indicator under inflation targeting. At the same time, the R-squared values reported in Table VIII.2 suggest that fluctuations in the explanatory variables capture 77–90 percent of the volatility in sovereign risk spreads.

166. Changes in global risk aversion, as measured by the LCPI, are likely to affect mostly the sovereign risk spreads of the somewhat lower-rated emerging market countries in the sample, such as South Africa and Thailand. Pooled regression analysis allows to estimate individual coefficients for all countries in the sample. The estimation results are shown in the third column of Table VIII.2. They suggest that changes in global risk aversion had a statistically significant impact on sovereign risk spreads in South Africa and Thailand, but not in New Zealand and Poland. The results indicate that a rise in global risk aversion by 10 LCPI index points would increase South Africa’s risk spreads by about 28 basis points.100

E. Implications for Economic Policy

167. Poor performance under an inflation-targeting monetary policy framework increases the cost of official external borrowing through a rise in U.S. dollar-denominated sovereign bond spreads. The rise in spreads stems from a higher risk of currency depreciation, which corresponds to the option of providing inflationary finance in a situation of financial distress to avoid sovereign default. If the inflation target is met, the sovereign risk premium seems to partly shift from U.S. dollar-denominated spreads to rand-denominated spreads, thereby increasing the cost of domestic borrowing. For public debt management, this may suggest that it is desirable to correspondingly shift from domestic- to foreign-currency borrowing to keep the overall debt service at the lowest possible levels.

168. While U.S. dollar-denominated spreads still reflect external vulnerability to some extent, they have also become an indicator for monetary policy performance under inflation targeting. Meanwhile, most conventional vulnerability indicators, such as net foreign assets in percent of GDP, the fiscal balance in percent of GDP, and the ratio total external debt to GDP, seem empirically less meaningful in explaining U.S. dollar-denominated spreads under inflation targeting. However, the empirical results suggest that the reserves-to-imports ratio remains an important determinant of sovereign spreads.

169. The considerable disinflation from the introduction of inflation targeting until October 2001—although this may have led to higher spreads on rand-denominated government bonds—has also successfully guided inflation expectations in the South African economy, which, in turn, has induced a decline in long-term real interest rates. The benefits from growth-enhancing effects of this decline overcompensate for the rise in borrowing costs the government may face in the domestic debt market.

170. Some of these effects have been partly reversed since, starting in September 2001, the depreciation of the rand led to higher inflation starting in December 2001. But the results of this study should encourage the South African authorities to continue their policy of disinflation under the inflation-targeting framework and counter the adverse inflation effects of the recent currency depreciation.

References

  • Alesina, A., M. DeBroeck, A. Prati and G. Tabellini, 1992, “Default Risk on Government Debt in OECD Countries,Economic Policy, No. 15, p. 42763.

    • Search Google Scholar
    • Export Citation
  • Arora, V., and M. Cerisola, 2001, “How Does U.S. Monetary Policy Influence Sovereign Spreads in Emerging Markets?,Staff Papers, Vol. 48, No. 3, pp. 47498, International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Calvo, G., L. Leiderman, and C. Reinhart, 1996, “Inflows of Capital to Developing Countries in the 1990s,Journal of Economic Perspectives, Vol. 10, No. 2, pp. 12339.

    • Search Google Scholar
    • Export Citation
  • Cantor, R., and F. Packer, 1996, “Determinants and Impact of Sovereign Credit Ratings,Federal Reserve Bank of New York, Economic Policy Review.

    • Search Google Scholar
    • Export Citation
  • Cottarelli, C., and M. Mecagni, 1990, “The Risk Premium on Italian Government Debt,Staff Papers, Vol. 37, No. 4, pp. 86580, International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Dooley, M., E. Fernandez-Arias, and K. Kletzer, 1996, “Is the Debt Crisis History? Recent Private Capital Inflows to Developing Countries,World Bank Economic Review, Vol. 10, pp. 2750.

    • Search Google Scholar
    • Export Citation
  • Eichengreen, B., and A. Mody, 1998, “What Explains Changing Spreads on Emerging Market Debt: Fundamentals or Market Sentiment?,NBER Working Paper No. 6408, National Bureau of Economic Research.

    • Search Google Scholar
    • Export Citation
  • Goldstein, M., and C. Woglom, 1992, “Market-Based Fiscal Discipline in Monetary Unions: Evidence from the U.S. Municipal Bond Market,” in Canzoneri, M.B., V. Grilli, and P.R. Masson (eds.), Establishing a Central Bank: Issues in Europe and Lessons from the U.S., pp. 22860.

    • Search Google Scholar
    • Export Citation
  • Greene, W.H., 1997, Econometric Analysis, 3rd edition, Englewood Cliffs, NJ.

  • International Monetary Fund, 2001, Emerging Market Financing, Quarterly Report on Developments and Prospects November 2001.

  • Jonsson, G., 2001, “The Risk Premium in South African Long-Term Interest Rates,” in South Africa—Recent Economic Developments, pp. 2851, SM/01/79, International Monetary Fund.

    • Search Google Scholar
    • Export Citation
  • Kamin, S., and K. von Kleist, 1999, “The Evolution and Determinants of Emerging Market Credit Spreads in the 1990s,Working Paper 68, Bank for International Settlements.

    • Search Google Scholar
    • Export Citation
  • Min, H., 1998, “Determinants of Emerging Market Bond Spreads: Do Economic Fundamentals Matter?,Working Paper 1899, World Bank Development Research.

    • Search Google Scholar
    • Export Citation
  • Vocke, M., 1999, “Default Risk Premia on Sovereign Debt An Analysis on Default Risk and Devaluation Risk for Selected OECD Countries,Ph.D. Dissertation.

    • Search Google Scholar
    • Export Citation

APPENDIX I: Data Description

Data on sovereign risk spreads for each country were constructed on the basis of sovereign bond yield data obtained from Datastream. Country-specific data were based on information provided by national authorities. Several data series were available on a monthly basis, but some were available only on a quarterly basis, and a few only on an annual basis. Quarterly and annual data were converted to a monthly basis using a cubic spline interpolation.

Data definitions are as follows:

article image

APPENDIX II: Econometric Model and Methodology

The general form of the estimated model can be written as:

(1)yu=αi+χitlβi+εit

where yit is the dependent variable, ϵi is the individual effect, which is taken to be constant over time t and specific to the individual cross-sectional units i, and xit and βi are k-vectors of non-constant regressors and parameters for i=1,2,…, N cross-sectional units. Each cross-section unit is observed for dated periods t = 1,2,…, T.

Heterogeneity over the cross-section is common in panel data analysis and suggests the application of fixed or random effects approaches.101 The panel estimation uses a weighted least squares regression technique with estimated cross-section weights. This is done with a Generalized Least Squares (GLS) regression that uses estimated cross-section residual variances. The use of cross-section weights assumes the presence of cross-section heteroskedasticity in the data. Indeed, we cannot exclude the possibility of heteroskedasticity over the cross-section of our panel data.

In our analysis, we assume the residuals to be cross-section heteroskedastic and contemporaneously uncorrelated. In consequence, the residual covariance matrix can be written as:

(2)Ω=[σ12IT000σ22IT000σN2IT]

Any contemporaneous correlation of the residuals would indicate a misspecification of the estimated model since the residuals systematically pick up effects that are supposed to be captured by one of the explanatory variables.

To obtain cross-section specific weights, covariances σi2 are estimated from a regular pooled OLS regression. These estimated variances are computed as:

(3)σi2=t=1Ti(yityit)2/Ti,

where yit are the OLS fitted values for the dependent variable in the pooled estimation of the model.

Since heteroskedasticity may be present to some degree, White’s heteroskedasticity consistent covariance estimates are computed for all pooled specifications of the model. The White covariance matrix is based on a variance estimator that can be written as:

(4)var(b)=NTNTK(XlX)1(ijuit2xitxit|)(X|X)1

where K is the total number of estimated parameters. While this variance estimator is robust to heteroskedasticity within each cross-section, it does not account for the possibility of contemporaneous correlation across cross-sections.

The relatively short time-series dimension of the panel data used in this study largely excludes the possibility that autocorrelation of the residuals leads to spurious regression results.

83

Prepared by Matthias Vocke.

84

Currency depreciations increase the government’s stock of foreign-currency-denominated debt in domestic currency terms, which raises the default risk. In consequence, buyers of foreign-currency-denominated sovereign bonds will expect higher yields, which are reflected in higher sovereign spreads.

86

The increase in expected returns that is associated with an increase in a debtor’s share in global debt markets and bond portfolios and the resulting rebalancing of optimally-diversified portfolios is referred to as “portfolio risk”.

87

To exclude the effects of short-term distortions in secondary market prices of bonds, the analysis uses monthly averages to calculate spreads.

88

See Box 2 in the Staff Report for the 2002 Article IV Consultation (SM/02/176) for a further comparison of the currency depreciations.

89

The SARB’s credible announcement that the NOFP would be reduced to zero by March 2003 has further improved financial markets’ perceptions regarding South Africa’s external vulnerability.

90

Jonsson (2001) provides empirical evidence in support of this view.

91

See Vocke (1999) for empirical evidence.

92

These liquidity premia are calculated as the spread between the yields of U.S. Treasury bonds and U.S. swap rates.

93

See IMF (2001), p.5, for a detailed discussion of the LCPI.

94

South Africa is rated Baa2 by Moody’s and BBB- by Standard &Poor’s for its foreign-currency-denominated debt. The benchmark has been calculated as the unweighted average of sovereign yield spreads for Chile (Baal; A−), Malaysia (Baa2; BBB), Mexico (Baa3, BB+), and Korea (Baa2; BBB+) for similar maturities.

95

Actual CPIX inflation (the consumer price index, excluding interest on mortgage bonds) is lagging by one month in Figure VIII.4 to take into account the delay in the publication of inflation numbers.

96

Furthermore, South Africa’s official external debt currently amounts to only about 8 percent of GDP, compared with official domestic debt of more than 30 percent of GDP.

97

Data on World Bank rand-denominated yields before September 2000 are unavailable through Datastream.

98

These countries include South Africa, New Zealand, Poland, and Thailand. Many other countries, such as Korea, Mexico, and Hungary have not been included in the panel, as these countries have moved to inflation targeting too recently or are just about to introduce it. For some other countries, such as Chile or the Czech Republic, no appropriate data on U.S. dollar-denominated spreads are available.

99

The coefficients of other explanatory variables were not statistically significant and are not reported in Table VIII.2.

100

The strong response of South African spreads to changes in global risk aversion can partly be attributed to the outstanding depth and liquidity of South African financial markets relative to those in other emerging market countries, inducing emerging market investors to trade South African assets first for liquidity considerations.

101

See Greene (1997), chapter 14, for a detailed discussion of these empirical approaches.

South Africa: Selected Issues
Author: International Monetary Fund
  • View in gallery

    Nominal Rand/U.S. dollar Exchange Rate

    (In percent change from beginning of year)

  • View in gallery

    Sovereign Risk Spreads

    (In basis point change from beginning of year)

  • View in gallery

    Global Risk Aversion and Sovereign Risk Spreads

    (January 2000–April 2002)

  • View in gallery

    Inflation Performance and the Difference in Sovereign Risk Spreads, January 2000–April 2002

    (In percentage points)

  • View in gallery

    Rand-Denominated Sovereign Risk Spreads

    (Basis point change since September 2000)