Information about Sub-Saharan Africa África subsahariana
Journal Issue

Angola: Selected Issues and Statistical Appendix

International Monetary Fund
Published Date:
April 2005
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details

III. Inflation and the “Hard-Kwanza” Policy30

A. Introduction

62. Angola has made important progress in reducing inflation since 1999, after two decades of high- or hyper-inflation. Between 1999 and the peace agreement of April, 2002, consumer price inflation fell from over 300 percent to around 100 percent, around which level it oscillated for over a year. Inflation then again decelerated sharply, following the adoption of a stabilization program in September, 2003. By December 2004, the 12-month inflation rate had fallen to 31 percent. 31

63. This note attempts to briefly characterize the policies adopted since September 2003, which have many of the characteristics of an exchange-rate-based stabilization program. It assesses its mechanisms and effects, explores possible costs, and advances some statistical/econometric analysis for updating previous work.

64. The note is organized as follows. Section B reviews the main policies implemented since September 2003. Section C summarizes recent developments on inflation, money and exchange rate. Section D and E present preliminary results that help to explain the path of inflation and the main variables potentially driving it. Section F discusses some policy issues. Section G advances some conclusions and remarks.

B. The “Hard Kwanza Exchange Rate Policy”

65. A major anti-inflationary initiative was put in place in Angola in September 2003 following a number of administrative measures affecting monetary policy implementation earlier in the year. No official account of the aims or mechanisms of this initiative has been issued, but a number of inferences can be drawn from associated announcements and developments:

  • The government specified in the annual budget a target reduction for inflation to 20 percent by the end of 2004 (compared with an actual 77 percent rate at end-2003).
  • An intermediate objective of stabilizing the exchange rate, although without a specific target rate, would also appear to have been set; indeed, the initiative became known as the “hard-kwanza” policy in light of the virtual cessation of currency depreciation which accompanied its implementation.32
  • Increased emphasis was placed on controlling the liquidity of the banking system on a weekly basis, using operations in government or central bank securities or sales of foreign currency.
  • The policy was to be underpinned by reducing the reliance of the central government on domestic finance, either through cuts in the fiscal deficit or increased foreign financing. In the event, the only major fiscal measures taken were to reduce the cost of fuel subsidies by raising retail prices towards (rising) market levels. However, higher oil revenues substantially reduced the fiscal deficit.

66. Evidence of the importance attached to the exchange rate and banks’ liquidity includes aggressive sales of dollars by the National Bank of Angola (BNA). The practice seems to have been to intervene heavily in the foreign exchange market whenever either (i) the depreciation of the nominal exchange rate accelerated; or (ii) there was a widening in the spread between the formal and the informal foreign exchange rate; or (iii) both. Staff estimates that, in the year from September 2003, the BNA sold about US$1.2 billion more foreign currency than required for normal financing activities. These sales occurred despite the low level of Angola’s foreign currency reserves and a rise in the real exchange rate, which itself probably in part reflected an autonomous upward shift in the equilibrium real exchange rate arising from higher actual and expected real oil prices. The sales were financed mainly by heavy recourse to oil-backed commercial borrowing and further accumulation of external debt payment arrears.

67. The policy change was initiated in September 2003 by a number of administrative measures. These included: (i) the unification of the official and the informal foreign exchange markets, (ii) the revisiting of the operational guidelines of the official foreign exchange auction, and, notably, (iii) increases in the central bank’s weekly sales of foreign exchange through the auction. The BNA also resumed domestic open market operations, using very short term instruments (at 28 and 63 days) at increased nominal interest rates (in excess of 50 percent), and conducted government securities operations of slightly longer maturities (91 and 182 days).

68. Previous measures taken in 2003 included changes in banks’ minimum reserve requirements and regulations on banks’ foreign exchange positions. Reserve requirements on foreign currency deposits were also amended to require deposits at the central bank to be made in kwanzas rather than foreign currency, causing the banks’ needs for reserve money to vary with the exchange rate.

69. The changes in banks’ minimum reserve requirements were effected in two steps. In March 2003, reserve requirements on foreign currency deposits were increased from 5 percent to 10 percent, and those on domestic currency were reduced from 30 percent to 10 percent. This equalization of the rates probably reduced liquidity at the margin since foreign currency deposits accounted for more than 70 percent of total deposits. Reserve requirements on both domestic and foreign currency deposits were then raised to 15 percent in July, 2003. The increase was reinforced by the imposition of a 100 percent reserve requirement on all government deposits held with commercial banks.

70. The changes in regulations on commercial banks’ foreign exchange positions require banks to ensure that, on a daily basis, their open position does not exceed the equivalent of 15 percent of their capital. Previously, the limit on open positions was established in absolute values (US$500,000), with the exception that up to US$2 million could be used to preserve the value of the banks’ capital.

71. On the fiscal side, the “hard kwanza” policy package has been characterized by efforts to improve fiscal control, and since May 2004, to reduce the structural fiscal deficit. Benefiting from high oil prices, the fiscal deficit has declined from 7.9 percent of GDP in 2003 to 4.6 percent in 2004.

72. In terms of financing, the period since September 2003 has been characterized by: (i) a significant reduction in BNA credit to the central government; (ii) increasing use of domestic debt instruments; and (iii) increasing reliance on foreign sources. While in 2003, foreign and domestic financing amounted to 3.5 and 2.2 percent of GDP respectively, in 2004 they equaled 6.0 and -5.1 percent respectively.

C. Recent Developments

73. Since the start of the “hard kwanza” policy in September 2003, when 12-month consumer price inflation was 95 percent, inflation has declined rapidly. By December, 2004, it had fallen on a 12-month basis to 31 percent. At the same time, the rate at which the exchange rate has depreciated has declined sharply, from around 70 percent at an annual rate to less than 10 percent, while monetary developments to October were characterized by a continuous reduction in money growth. Reserve money and broad money M3 grew by 160 and 125 percent, respectively, in the twelve months to September 2003, but their growth slowed to 25 and 50 percent in the twelve months to October 2004. One consequence of these developments, which may in part reflect the impact of high oil prices, has been a sharp real appreciation of the exchange rate since September 2003.

74. Figure III.1 presents the path of inflation (inf), nominal depreciation (dep), currency growth (dcurr), reserve money growth (drm), and growth in broad money M3 (dm3) on a 12-month basis.

Figure III.1.Angola: Inflation, Depreciation, and Money Growth

Sources: BNA, IMF staff.

Note: inf = inflation, dcurr = currency in circulation growth, drm = reserve money growth, dm3 = M3 growth, dep = nominal depreciation.

D. Empirical Analysis: Money, Exchange Rate, and Prices

75. Table III.1 presents some basic statistics for inflation, depreciation of nominal exchange rate and three measures of money growth (currency in circulation, reserve money and broad money). The sample covers the period September 2002 – October 2004, and the variables are presented in terms of 12-month changes in the relevant levels. Basic statistics are calculated for the five variables for the whole period, and for the period pre and post September 2003, which marks the beginning of the current stabilization approach.

Table III.1.Angola: Inflation, Money Growth and Depreciation
Sep 02 - Oct 04
Standard Deviation29.936.945.145.440.4
Sep 02 - Sep 03
Standard Deviation6.36.619.540.010.5
Oct 03 - Oct 04
Standard Deviation19.219.129.542.222.0
Sources: BNA and IMF staff.Note: inf inflation, dcurr = currency in circulation growth, drm = reserve money growth, dm3 = M3 growth, dep = nominal depreciation.
Sources: BNA and IMF staff.Note: inf inflation, dcurr = currency in circulation growth, drm = reserve money growth, dm3 = M3 growth, dep = nominal depreciation.

76. Inflation, nominal depreciation and the three measures of money growth all significantly declined, on average, from September 2003. In particular, the depreciation of the kwanza declined, from an average of 86 percent during the first period, to 19 percent during the second period. Interestingly, though, the volatility of all the variables has increased, in particular those of inflation and depreciation.

77. Further insight may be gained when computing simple correlations among these variables. Table III.2 presents the correlations for the two sub periods.33

Table III.2.Angola: Simple Correlations
SubPeriod: September 2002 - September 2003

SubPeriod: October 2003 - October 2004
Sources: BNA, IMF staff.Note: inf inflation, dcurr = currency in circulation growth drm = reserve money growth, dm3 = M3 growth, dep = nominal depreciation.
Sources: BNA, IMF staff.Note: inf inflation, dcurr = currency in circulation growth drm = reserve money growth, dm3 = M3 growth, dep = nominal depreciation.

78. The correlation of inflation with growth in all monetary aggregates and depreciation decreases from one period to the other. The most significant declines are observed in the correlations between inflation and (i) depreciation and (ii) M3 growth (which include dollar-denominated deposits). It is expected that since the implementation of the “hard kwanza” policy, given that the exchange rate has been stabilized by means of sizeable intervention, the remaining inflation has to be explained by other variables. While also declining, inflation is still strongly correlated to growth in currency in circulation and reserve money (which may be more strongly influenced in the first period by the inclusion of foreign currency deposits held as required reserves).

79. Figure III.2 depicts the trends of the five series for the whole sample period (derived from the Hodrick-Prescott filter). Although converging, all monetary aggregates are still growing faster than prices. Only the nominal exchange rate has been increasing, on a trend basis, at a much lower pace than prices.

Figure III.2.Angola: Trends of Inflation, Depreciation, and Money Growth

Sources: BNA, IMF staff.

Note: inf = inflation, dcurr = currency in circulation growth, drm = reserve money growth, dm3 = M3 growth, dep = nominal depreciation.

80. The implications of this simple analysis are that the shift in policy stance after September 2003 succeeded in breaking, at least temporarily, the inflationary process that has long affected Angola. A tighter monetary stance, aggressive foreign exchange intervention, and a shift in expectations about the authorities’ resolve, supported by efforts to reduce the fiscal deficit, against the background of strengthening oil prices, stabilized the exchange rate and limited the rate of increase of prices.

E. Sources and Uses of Base Money

81. In order to gain some further insights about the main sources of money creation, Table III.3 provides a breakdown of sources and uses of base money for the periods August 2002-August 2003, and September 2003-September 200434.

Table III.3.Angola: Sources and Uses of Base Money 1/(Billions of Kz.)
Aug. ʹ02 - Aug. ʹ03Sep. ʹ03 - Sep. ʹ04
Base money32.9117.51
Currency in circulation20.6310.07
Loans and bills21.880.41
Other claims-0.32-0.94
Domestic deposits-3.96-9.82
Government foreign currency deposits2.088.39
Central bank operations1.119.00
Central bank operational surplust/deficit1.129.59
Central bank reg, transit, non financial assets2.4514.68
Central bank bills-5.110.43
Foreign exchange11.8110.34
Other foreign assets-2.303.57
Banks foreign currency deposits1.390.22
Other claims and liabilities0.310.13
Claims on financ.corp, priv sector (-) deposits0.340.13
Liabilities to sonangol-0.030.00
Source: BNA

In domestic currency

Source: BNA

In domestic currency

82. The numbers suggest some interesting developments since the implementation of the “hard-kwanza” policy package in September 2003. First, money creation declined sharply subsequent to the policy shift, decreasing from Kz. 32.91 billions in the first period to Kz. 17.51 billions in the second period. Second, the sources of money creation changed significantly. Credit to the central government was the main source of money expansion during the first period, but it fell in the second period, reflecting in part the central government’s heavy foreign currency borrowing, which enabled it to build its foreign currency deposits at the BNA. At the same time, the central bank’s operating deficit significantly increased in importance as a source of money expansion. Third, there has been a decrease of money base expansion coming from the net accumulation of reserves, despite the government’s foreign currency borrowing, reflecting sizeable interventions in the foreign exchange market to stabilize the kwanza.

F. Policy Issues35

83. Whether by design, or as a by-product of the way the policy has been implemented, and the increase in oil prices, the most significant feature of the intensified anti-inflationary strategy adopted since September 2003 has been the virtual stabilization of the nominal exchange rate. This has been a major factor in limiting price increases for tradable goods. However, growth in the monetary aggregates, while slowing, has been faster than would be consistent with achievement of the inflation objective, and the real exchange rate has appreciated significantly, implying that the policy has been much less effective in reducing the inflation rate of non-tradables. There have also been financial costs incurred as a result of sizeable and expensive foreign exchange intervention, which has increased Angola’s external liabilities.

84. The policy package seems to have focused primarily on the monetary aspects of the inflationary phenomenon, while placing relatively less weight on fiscal consolidation. Implicit in the strategy seems to have been the view that the substitution of debt issue for taxation and/or money creation would reduce the inflationary impact of fiscal deficits, as private agents heavily discount future tax liabilities—including the inflation tax—to repay the increased debt. However, there are clear limitations to this process. Specifically, prospective high interest payments—derived from costly foreign borrowing and/or expected domestic financing of future deficits—are likely to keep inflation expectations high, and might even bring inflation forward. Theoretical and empirical research suggests that large prospective deficits affect today’s economic conditions and expectations, and can, in some cases, explain balance of payments crises.36

85. As suggested by a number of authors, the international experience shows that successful stabilization programs are characterized, regardless of the monetary arrangement, by a significant fiscal adjustment (See for instance, Calvo and Vegh, 1999).

G. Conclusions

86. Angola has made significant progress towards macroeconomic stabilization, which is a notable achievement after nearly three decades of high inflation. The exchange rate has stabilized and inflationary expectations seem to have been substantially reduced. With high oil prices permitting fiscal consolidation, further reductions in inflation – towards single figure rates – should be within reach.

87. The achievement of low and stable inflation is a central ingredient for developing the economy’s potential. However, to continue to exercise anti-inflationary policy in the manner applied under the “hard kwanza” policy may impose high costs on the economy. Under the “hard kwanza” policy, the rise in the real exchange rate (part of which may have reflected the rise in oil prices) reduced the competitiveness of the non-oil economy, while the sizeable foreign exchange intervention was expensive and increased Angola’s external liabilities.

88. To allow greater flexibility for the exchange rate in the future, given the low level of foreign exchange reserves, a more fully-defined money-base approach may offer a better way forward. On the fiscal side, credible and sustainable fiscal adjustment, coupled by a change in the composition of deficit financing, would be key to the stabilization process.


    Alvesson, Magnus, and AlfredoTorrez, July2003, “Central Bank Operations and Macroeconomic Stabilization in Angola”, Angola: Selected Issues and Statistical Appendix (International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Burnside, Craig, MartinEichenbaum, and SergioRebelo, 1998, “Prospective Deficits and the Asian Currency Crisis”, NBER Working Paper No. 6758 (NBER).

    • Search Google Scholar
    • Export Citation

    Calvo, Guillermo, and CarlosVegh, 1999, “Inflation Stabilization and BOP Crises in Developing Countries”, NBER Working Paper No. 6925 (NBER).

    • Search Google Scholar
    • Export Citation

    Gasha, Jose Giancarlo, 2003, “A Note on Inflation”, Angola: Selected Issues and Statistical Appendix (International Monetary Fund).

    Gasha, Jose Giancarlo, and GonzaloPastor, 2004, “Angola’s Fragile Stabilization”, IMF Working Paper No. 04/83 (International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    Woodford, Michael, 2001, “Fiscal Requirements for Price Stability”, Journal of Money, Credit and Banking, Vol. 33.


Prepared by Jose Giancarlo Gasha.


See Gasha (2003) for a fuller account of Angola’s history of inflation and associated policies.


See Gasha and Pastor (2004) for a detailed account of events and measures.


The series for computing these correlations have been filtered using a Hodrick-Prescott filter (Hodrick and Prescott, 1997), and are thus stationary.


For a more extensive discussion on policy alternatives including a money-based program with sizable fiscal consolidation, see Gasha and Pastor (2004).

Other Resources Citing This Publication