Angola
Selected Issues and Statistical Appendix
Author: International Monetary Fund
Publication Date:
26 Oct 2007
eISBN:
9781451800562
Language:
English
Keywords:
ISCR; CR; exchange rate; monetary policy; base money; U.S. dollar; monetary policy implementation; inflation rate; dollar currency holding; Monetary base
Download PDF (1.9 MB)

This Selected Issues paper and Statistical Appendix on Angola underlie monetary policy framework. Angola has adopted an anti-inflation policy that has led to a sharp decline in inflation. To institute a monetary policy framework, a nominal anchor or constraint on the value of domestic currency must be established. Additional work to fine-tune the measures of currency in circulation, conduct more sophisticated tests to assess the relationship between inflation and the monetary aggregates, and determine how to incorporate the currency measure in monetary operations is needed.

Abstract

This Selected Issues paper and Statistical Appendix on Angola underlie monetary policy framework. Angola has adopted an anti-inflation policy that has led to a sharp decline in inflation. To institute a monetary policy framework, a nominal anchor or constraint on the value of domestic currency must be established. Additional work to fine-tune the measures of currency in circulation, conduct more sophisticated tests to assess the relationship between inflation and the monetary aggregates, and determine how to incorporate the currency measure in monetary operations is needed.

I. Options to Strengthen the Monetary Framework in Angola1

A. Introduction

1. Since September 2003, Angola has adopted an anti-inflation policy that has led to a sharp decline in inflation. Inflation dropped from about 100 percent in mid-2003 to about 12 percent at year-end 2006 (Text Figure 1). The exchange rate has also been stable since a step appreciation in late 2005. Still, despite attempts to tighten monetary policy, the inflation targets were missed in 2005-06 (Text Table 1).

Text Figure 1.
Text Figure 1.

Money, Inflation, and Exchange Rates, March 2003—December 2006

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

Text Table 1.

Monetary Policy Implementation, 2005-06

article image
Source: Angolan authorities.

In April 2005 the BNA revised its working assumptions on account of higher oil revenues. Targets for NIR, and overall government balance were revised. Inflation was revised upwards to a 6 percentage point around 15 percent and base money projection to 43.75 percent.

2. This paper examines Angola’s recent experience in lowering inflation and its options going forward in choosing the appropriate nominal anchor, given its highly dollarized economy, and continued remonetization. It finds that

  • The use of a monetary aggregate anchor has helped Angola rein in inflation, despite unpredictable demand for monetary balances as inflation rates have fallen and the structure of the economy and the financial system has changed.

  • To further reduce inflation, monetary policy could target the narrow measure of money, which in Angola appears more correlated with inflation.

  • Once the monetary aggregates are corrected to include foreign currency notes in circulation, the relationship between inflation and money is better explained.

  • Monetary aggregate targeting should be supplemented by other indicators, and policymakers should target ranges instead of absolute levels of monetary aggregates.

  • Addressing infrastructure limitations in the markets and institutional shortcomings would enhance monetary policy.

B. Recent Developments

3. The 12-month inflation rate fell to 31 percent in December 2004 from around 100 percent in 2003, after the government began financing its fiscal deficit with oil-backed loans (Text Figure 2). These loans also supplied the Banco Nacional de Angola (BNA) with foreign exchange to stabilize the exchange rate.2 Nonetheless, even though the inflation rate declined rapidly, it overshot the authorities’ targets in 2004, 2005, and 2006 and inflation remains above the average for other oil exporters in sub-Saharan Africa.

Text Figure 2.
Text Figure 2.

Sources of Change in Base Money, 2001–06

(Percent)

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

4. Increased oil revenues have helped stabilize the nominal exchange rate (Text Figure 3). After depreciating modestly in 2004, the kwanza appreciated in nominal terms in both 2005 and 2006; it then stabilized at about the level it reached when the “hard-kwanza policy” began in 2003.3 The real effective exchange rate index rose by a cumulative 50 percent between 2003 and 2006.

Text Figure 3.
Text Figure 3.

Exchange Rate and Reserves, January 2005–December 2006

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

5. After slowing sharply in 2003-04, monetary aggregate growth rose in 2005 then slowed again in 2006 (Text Figure 4). This pattern reflected policy actions by 1 the BNA. The bank took a less active stance in mopping up the increase in liquidity resulting from a draw-down of government deposits and matured treasury bills and BNA securities. The BNA did not issue sufficient bills because of balance sheet pressures and the government was reluctant to issue treasury bills in the absence of a financing need. In 2006, however, the BNA moved to sterilize liquidity by selling foreign exchange and BNA securities, causing base money growth to slow to just 1 percent.

Text Figure 4.
Text Figure 4.

Money Growth and Inflation, 2001–06

(Percent)

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

6. There have been considerable variations across the different monetary aggregates in recent years. The growth of broad money (M2 and M3) was about 60 percent in 2006 after following the same path as base money through 2005. This reflected a shift in the components: deposits (both domestic and foreign currency) rose sharply, causing the money multiplier (M3/base money) to also spike between 2003 and 2006. Narrow money has consistently grown more slowly than broad money—M1 grew about 10 percentage points less than, and currency by about half as much as, broad money. Velocity fell in the same period as remonetization continued, while inflation fell.4

7. Interest rates have dropped dramatically. Yields on three-month treasury bills fell along with inflation, from about 80 percent in 2003 to just above 6 percent at end-2006, (Text Figure 5). Commercial bank interest rates, however, declined less steeply than rates on government securities. Lending rates in domestic currency, meanwhile, plunged, to just over 19 percent in 2006 (from about 100 percent in December 2003), and stayed positive in real terms through the period. Yields on kwanza term deposits, by contrast, lagged inflation.

Text Figure 5.
Text Figure 5.

Selected Interested Rates, 1999–2006

(Percent)

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

8. The relationship between inflation and the monetary aggregates has weakened as dollarization has increased and remonetization has continued. As shown in Text Table 2, the correlation between inflation, the exchange rate, and monetary aggregates decreased from one period to the next. In addition to inflation and the exchange rate, inflation, M1, and currency in circulation appear strongly linked.5 Granger causality tests confirm that the exchange rate depreciation and the growth in the monetary aggregates each appear to Granger-cause inflation (Text Table 3) and that exchange rate depreciation affects monetary growth, though the link between exchange rate depreciation and monetary growth appears weaker.

Text Table 2.

Correlations, Dec. 2000-March 2007

(12-month rates of Change)

article image
Text Table 3.

Causality: Money, Exchange Rate and Inflation

article image

9. There remain risks to the current moderate inflation environment. Until 2003, the Angolan economy was characterized by persistent inflation induced by large fiscal and quasi-fiscal deficits.6 The recent moderate inflation has been helped by buoyant oil prices and a reversal of these fortunes could jeopardize macroeconomic stability if the authorities do not maintain sound fiscal and monetary policies.

C. BNA’s Monetary Policy Framework

10. Angola’s monetary policy is conducted under a monetary aggregate framework; its declared objective is to maintain price stability. However, the exchange rate has also recently become an objective.7 Base money (currency plus bank reserves) is used as the operating target. The monetary program sets the percentage variation target for base money and specifies the sterilization instrument to be used. The extent of monetary operations is based on government’s cash flow. (The formulation and execution of monetary policy and the instruments used to manage liquidity are discussed below). Sterilization has mainly occurred through foreign exchange sales, though BNA bills have also been issued recently to slow appreciation in the exchange rate, making the exchange rate level a monetary policy objective.8

Process for the formulation and execution of monetary policy

11. Three bodies deal with monetary policy formulation and implementation: the Board of Directors of the BNA, the Liquidity Committee, and the Market Committee of the BNA. The members of the board of directors appointed by the council of ministers are part of the BNA’s administration. The Liquidity Committee is made up of staff from the BNA and the Ministry of Finance (MOF). BNA representatives comprise the board members responsible for managing the bank’s technical departments that formulate and execute monetary policy, the directors of the departments of studies and statistics (DEE), foreign reserves management, payment systems, security markets, and accounting. The MOF’s members comprise the vice minister of finance, the director of the treasury, and the national director of taxes. The committee coordinates the design of monetary policy and its execution by the BNA and the MOF. The Market Committee comprises the governor of the BNA, other BNA board members, the directors of the technical departments responsible for formulating and implementing monetary policy,9 the director of the Department of Banking Supervision (DBS), and the director of the Audit Department. The MOF does not participate. Following discussions between the board members and BNA officers, a decision on monetary policy intervention is made.

12. The Department of Studies and Statistics of the BNA, with input from the BNA’s technical areas, drafts a monetary program based on the government’s two-year plan. The monetary program is drafted and revised in the first quarter of the year and is then approved by the BNA board of directors.

13. The minister of finance and the governor coordinate the approved monetary program. After the final document is submitted to the government’s economic team for approval, it is sent back to the BNA so it can be implemented.10 Monetary policy implementation is reviewed daily, weekly, monthly, and quarterly at different levels of the BNA. The government follows up by having the MOF participate in one of the weekly meetings of the BNA’s policy-execution bodies.

Instruments for liquidity management

14. The monetary program sets the percentage variation target for base money without specifying which instruments should be used to achieve it. Liquidity projections are prepared every week, and the sterilization instrument to be used and to what degree is decided in light of the cash-flow program and management of the government.

15. The main instrument used is intervention in the foreign exchange market through sales of foreign exchange to authorized dealers. These operations are based on decisions about the required amount of domestic currency sterilization made weekly by the Liquidity Committee and the Market Committee. The Department of Foreign Exchange Management conducts daily auctions by evenly distributing the foreign exchange to be sold for the week over five days.

16. The BNA, for monetary policy purposes, started issuing BNA bills with 28-day maturities in late 1999; the Ministry of Finance launched its first issue of domestic securities—treasury bonds and bills—in July 2003. To broaden the range of money market instruments available and to reduce their potential costs, the BNA in mid-2003 started issuing securities in 30-, 90-, and 180-day maturities. The treasury bonds were issued primarily to clear arrears with domestic suppliers; treasury bills, denominated in domestic currency and issued in 91- and 182-day maturities were short-term financing instruments. Treasury bills were meant to gradually replace short-term central bank bills, reinforce monetary policy management, provide the government with a less inflationary financing instrument than central bank credit, and deepen domestic securities markets.

17. The BNA in March 2004 established procedures for purchasing/repurchasing securities between banks. Transactions must be at least one day, and the BNA may set a maximum period. The value of repurchases cannot be higher than five times the selling bank’s own funds. Transactions are recorded at the BNA Central Securities Depository and are settled at the banks’ reserve accounts at the BNA.

18. The BNA also intervenes in the Interbank Money Market (MMI) using securities to manage liquidity. The BNA regulates the MMI, authorizes the institutions that may participate, and authorizes all bank transaction at the MMI. Transactions must be more than Kz 100,000 and be for a specified period of up to one year. Settlement may take place up to two working days after the transaction.

19. Deposits in kwanza and in foreign currency are subject to a 15 percent reserve requirement, which can be constituted in cash and with up to 50 percent of it in central bank or treasury bills. Since 2005 the reserve requirement ratio has not changed. Central government deposits at the Banco de Poupanca e Credito (BPC) must meet a 100 percent reserve requirement; deposits at the BPC’s branches in Angola’s provinces must meet a 50 percent reserve requirement. Reserves required for dollar-denominated deposits must be held in kwanza.

20. The BNA rediscount rate was reduced to 18 percent from 95 percent in December 2005. The rediscount facility is rarely used in Angola and is thus more of a signal than an instrument in managing liquidity.

D. Selecting a Nominal Anchor and Options for Angola

21. To institute a monetary policy framework, a nominal anchor or constraint on the value of domestic currency must be established and intermediate and operating targets to guide short-term liquidity adjustments must be set. Central banks can achieve anti-inflation credibility by committing to fix the domestic currency to a nominal magnitude, such as the exchange rate, the money supply, or the inflation rate. This section discusses the main features and requirements of different monetary policy frameworks.

Fixed exchange rate

22. Under a fixed exchange rate regime, the exchange rate of the currency is fixed to another currency or unit that behaves in a desired way. The quantity of money is then determined by, or adjusted according to, the public’s demand for it at the value that has been fixed by the exchange rate. This regime has obvious advantages—it is easy to administer and does not require knowledge of velocity, which is particularly difficult to estimate when the economy is undergoing a transition, as in Angola. It is also the quickest way to establish faith in the stability of the currency’s value.11 As a nominal target, the exchange rate is more transparent than money and can be readily monitored. A number of countries have successfully used exchange rate targeting to reduce inflation—for example, Argentina under a currency board arrangement in 1991 and Estonia, Latvia, and Lithuania under exchange rate pegs. Inflation in the franc zone countries of Africa, whose currencies were pegged to the French franc and now to the euro, has consistently been lower than in neighboring countries.

23. Fixing the exchange rate, however, has its costs. A fixed exchange rate can only be maintained if it is backed by sound fiscal policy. Government borrowing should be limited to the level that can be raised from the public. In the tradable and nontradable goods market, the exchange rate can anchor only a subset of prices. Under an exchange rate peg, scaled-up foreign inflows would temporarily result in higher inflation as higher spending pushes up the prices of nontradable goods, to which the central bank’s foreign exchange reserves and the money supply can respond endogenously (IMF 2007). A fixed exchange rate also undercuts monetary policy independence, because the country can no longer respond to shocks independent of those hitting the anchor country. For instance, swings in oil prices, if not handled through a stable spending pattern, could compromise the macroeconomic stabilization required under a fixed exchange rate regime. Fluctuations in the value of the particular currency to which the home currency is pegged can also produce needless volatility in the country’s international competitiveness. When large foreign exchange interventions are undertaken to maintain an exchange rate peg, net foreign assets may be volatile and difficult to predict, making it difficult to manage liquidity.

24. A fixed exchange rate leaves countries open to speculative attacks on their currencies as buyers can count on one-way bets. It also lowers the perceived risk for foreign investors, thus encouraging greater capital flows. If bank supervision is inadequate, loan losses are likely, causing bank balance sheets to deteriorate. Exchange rate targets can also mask expansionary policies, as in Thailand before the 1998 currency crises. Finally, fixing the exchange rate may not be enough to change expectations—a peg that is not supported by credible policies would collapse.

Monetary aggregate framework

25. In a monetary aggregate framework, the relationship between money demand, nominal income, and interest rates is assumed to be sufficiently stable. The quantitative monetary aggregate with the strongest relationship to the ultimate target, such as inflation, may thus be a suitable intermediate target. An example is the money supply. In economies with underdeveloped money markets and without instruments for interest rate targeting, the instrument could be the monetary base. Interest rates work better as operating targets when the link between the monetary aggregate, inflation, and short-term interest rates are easily determined. Both industrialized and developing countries have adopted monetary aggregate targeting successfully. In the mid-1970s the Bundesbank began using monetary-targeting to maintain low inflation. The United States used a number of monetary aggregates as intermediate targets in the 1960s and 1970s.

26. If there is a strong and reliable relationship between inflation and the monetary aggregate (such as M3 or M2), the central bank can accurately project the value of the money multiplier. The money multiplier is a function of the ratio of currency to deposits and the ratio of bank reserves to deposits. An increase in the ratio of currency to deposits reduces the reserve component of a given base money and thus reduces the money multiplier. An increase in the ratio of bank reserves to deposits similarly reduces the multiplier. The currency-to-deposits ratio reflects the public’s preferences for one form of payment over another (a choice that is influenced by the relative convenience and opportunity costs of holding money). To set monetary policy instruments appropriately, the central bank should be able to estimate the likely behavior of this ratio.

27. Monetary targeting allows the central bank to adjust its monetary policy to cope with domestic shocks. A flexible exchange rate arrangement under this framework can help the economy adjust to terms of trade changes, thus helping to maintain macroeconomic stability. This is particularly important in Angola because the price of oil fluctuates, and the current account will have to adjust in the medium term once oil production begins to fall. A flexible exchange rate regime facilitates a relative price adjustment while allowing monetary policy to be geared toward controlling domestic inflation. In Angola, where goods consumed have a high import content, the appreciation of the kwanza could help stabilize or lower inflation. Flexible exchange rate regimes also appear to be associated with less inflation and output variability than exchange rate pegs (IMF 2006).

28. A monetary targeting policy can be easily communicated to the general public as long as figures for monetary aggregates are announced. The public can then determine if the central bank is achieving its target, and the market has an immediate signal of the monetary policy stance.

29. An obvious drawback to a regime that fixes money growth is that fluctuations in money demand or in the behavior of the banking system weaken the relationship between the monetary aggregate and inflation. Thus, if velocity is not stable, the monetary aggregate might not reflect the desired outcome for the ultimate goal (e.g., inflation).

Inflation targeting

30. An inflation-targeting framework explicitly targets inflation. The target is a publicly announced, numeric range. Inflation targeting in practice involves adjusting monetary policy instruments as new information becomes available so as to meet the stated target. Monetary policy, which is guided by inflation forecasts, is conducted in a market-based, flexible, and transparent manner. This framework has been used successfully by several industrialized countries, such as Canada, the United Kingdom, and New Zealand, and is being adopted by many emerging market countries in which the relationship between inflation and monetary aggregates has weakened.

31. Inflation targeting enables monetary policy to focus on domestic factors and to respond to shocks to the domestic economy. Unstable velocity is not an issue because the relationship between money and inflation does not have to be stable. Inflation targeting, which is easy to understand, is also very transparent. The more open and precise is the central bank’s inflation objective, the more confident the public will be that the bank will meet the target. A precise, well-publicized target also fosters accountability because it is easier for the public to determine whether the inflation objective is consistently being met.

32. The following conditions help ensure that an inflation targeting framework can be successfully implemented: (1) The central bank must have the mandate to pursue an inflation target and must inform the public about the monetary policy framework; (2) monetary policy should not be dominated by fiscal priorities; (3) the external position should be strong and sustainable; (4) initial inflation should be low enough to ensure a reasonable degree of monetary control; (5) the financial sector should be stable and the financial markets well developed; and (6) the central bank should have the tools to implement monetary policy. Overall, successful inflation targeting works best if a country has strong operational capacity at the central bank; monetary policy credibility, backed by adequate political support; a fairly developed financial sector; strong public sector financial management; and consistent fiscal policy. Low dollarization of financial liabilities and limited vulnerability to sharp changes in capital flows and international investor confidence are also important.12

Options for Angola

33. Angola, however, does not yet have a deep enough financial sector, adequate instruments, or sufficient credibility to adopt a fixed exchange rate. Angola’s economy is not adequately resilient and its financial system is not strong enough to withstand speculative attacks. By contrast, a flexible exchange rate would allow the economy to adjust better to domestic and external shocks. A number of countries have dealt successfully with a surge in flows under a flexible exchange rate arrangement, with little impact on the exchange rate; others have allowed the exchange rate to appreciate.

34. Angola does not have enough reliable data to forecast inflation, therefore making inflation targeting difficult. The lagged relationship between monetary policy and inflation also suggests that information about the monetary policy stance would not reach the public in a timely way. In terms of transmission, because the financial sector is underdeveloped, the interest rate and credit are channels weak. Interest rates, while market determined, do not respond well to monetary policy signals, and government-owned bank dominates the credit channel. The high level of dollarization worsens these limitations.

35. Monetary targeting in Angola should be adapted to incorporate some aspects of the other frameworks. Monetary aggregate targeting has been used effectively to reduce inflation in Angola. Monetary targeting has in most cases been relegated to the background once the relationship between monetary aggregates and inflation or nominal income breaks down. If the financial environment in Angola is not expected to change much in the medium term, one must ask, “If it ain’t broke, why fix it?” Countries that have given up on monetary targeting have sometimes switched among different monetary aggregates, such as M1, M2, and base money. Similarly, the choice of the intermediate target in Angola should be monitored continuously to ensure that it is still relevant, including switching from one aggregate to another. Given the instability of velocity, the authorities should not rigidly uphold their target range but allow it to be under shot and over shot at times. The inflation goal should be allowed to vary over time as long as it converges to a long-run goal. The objectives of monetary policy, including a numeric inflation goal, should be clearly stated. And to make monetary policy more transparent and the central bank more accountable, monetary policy should be communicated to the public.

E. Improving the Monetary Framework in Angola

Absence of reliable data

36. To be effective, a monetary aggregate framework requires adequate data. We specify the following functional relationship for the demand for money:

  • (M/P) = f(Y, OC),

  • where13

  • M = demand for nominal money balances (M3 or M2); and

  • P = the price level (CPI);

  • Y = scale variable (income, wealth, or expenditure, in real terms—national income); and

  • OC = expected rates of return (interest rate).

  • Angola lacks reliable statistics on several aspects of economic activity, including the determinants of demand for the monetary aggregates, such as nominal GDP and prices. The problems are on both sides of the equation. Broad money (M2 or M3) does not capture all money, as discussed in the next section. Before May 1999, administrative decisions distorted key prices, such as exchange rates. The absence of reliable time series data has prevented the authorities from analyzing in-depth the main characteristics of the macroeconomy and the financial system.

37. The Consumer Price Index (CPI) covers only Luanda, Angola’s capital city; the non-oil sector data in the national accounts are unreliable; and the shallow financial sector constrains interest rates.14 To better explain the relationship between inflation and the money supply, the authorities should consider extending the CPI at least to regional centers, using the 2000/01 Household Expenditure Survey. Efforts should also be made to obtain more reliable data on the non-oil sector, in particular on construction, manufacturing, and commerce and trade.

Design of a monetary aggregate framework in the presence of dollarization

38. The Angolan economy is highly dollarized—the U.S. dollar, along with the kwanza, is accepted as means of payment, in most business. Private sector bank deposits in foreign currency were 74 percent of total deposits at year-end 2006, up from 66 percent at year-end 2005 (Text Figure 6). This measure, however, relates foreign currency deposits within the domestic banking system (including domestic deposits and foreign currency) to a monetary aggregate, predominantly M2, without accounting for the stock of foreign cash in circulation; it thus underestimates the actual degree of dollarization (see discussion below). With the dollar as legal tender, there is more scope for substituting assets, which reduces the effectiveness of monetary policy to affect aggregate demand through the interest rates and exchange rate channel.

Text Figure 6.
Text Figure 6.

Foreign Currency as Share of Total Deposits, 1999–2006

(Percent)

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

39. The strategy of Angola’s monetary aggregate framework is to calculate the base money level that would be consistent with the money supply target, derived using the following formulation:

M=m(B)(1)
m=(c+1)/(c+r),(2)

  • where

  • M = money supply;

  • B = base money;

  • m = money multiplier;

  • r = the ratio of bank reserves to deposits; and

  • c = the ratio of currency outside banks to deposits held at banks by the public.

40. The currency-to-deposits ratio reflects the public’s preferences, as influenced by convenience, return, and value, for one form of payment medium over another, and is thus not easily controlled by the central bank. The central bank does, however, have considerable influence over the bank reserves-to-deposits ratio. As shown in Text Figure 7, both the currency ratio and the reserve ratio have trended downward, although the currency ratio did recover somewhat in the last quarter of 2006. A decrease in the ratios implies a higher multiplier. The recent decline in the cash ratio, which may reflect shifts from domestic to foreign currency holdings, also distorts projections of the multiplier (Text Figure 8).

Text Figure 7.
Text Figure 7.

Components of Multiplier, March 2003–December 2006

(Percent)

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

Text Figure 8.
Text Figure 8.

Cash Ratio and Dollarization, January 1999–December 2006

(Percent)

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

41. Evidence from other countries, particularly in Latin America, reveals that dollarization has continued even after the macroeconomy stabilized. The BNA’s monetary aggregate framework should therefore focus not only on total money supply, but also on its composition—both foreign and domestic-denominated components. The use of broader aggregates would better inform inflation projections and expected exchange rate development than the current arrangement.15 Adding foreign currency in use outside the banking system would also better steer monetary policy implementation, as it would stabilize the currency-to-deposits ratio and would allow the BNA to more accurately project the multiplier. Indeed, Oomes and Ohnosorge (2005), who estimate U.S. currency holdings in Russia, find that including foreign currency holdings in the definition of money improves the stability of the money demand function. With some caveats, they conclude that in highly dollarized economies estimates of foreign cash holdings can better explain the relationship between money growth and inflation.16

42. There are three possible sources of information on foreign cash holdings of the Angolan public: data on bank shipments of U.S. currency to Angola are available from domestic banks in Angola or the U.S. Treasury; the net sales of foreign currency by authorized banks less net withdrawals from foreign currency deposits; and Currency and Monetary Instrument Reports (CMIRs) collected from travelers by the U.S. Customs Services. The U.S. Treasury, which typically estimates U.S. dollars in circulation based on currency shipments, CMIRs, and trips to selected countries, estimates that about 60 percent of the US$770 billion currency in circulation is held abroad. The U.S. Treasury estimates that Russia, Argentina, and South Africa, respectively, hold US$80 billion, US$50 billion, and US$2 billion in U.S. currency notes. 17

43. We estimate U.S. currency in circulation in Angola for December 1999-December 2006 using data from domestic banks on monthly imports of U.S. dollar notes in 2000-06. The bank data show note imports of just over US$7 billion during that period. Angola receives a significant share of dollar notes shipped to sub-Saharan Africa and notes shipped there have trended upward—in 2002-06 shipments almost tripled.18

44. Our estimates suggest that U.S. notes in circulation exceed both the current Ml (domestic currency in circulation plus all types of demand deposits) and total foreign currency deposits (Text Figures 9a-d). 19 This finding is not surprising for several reasons: (1) consumers and businesses hold dollars during periods of political and economic uncertainty to hedge against inflation and calamity; (2) the illicit diamond trade both before and after the war was conducted in dollars; (3) cross-border trade since the start of the war has been conducted in dollars; and (4) the increase in foreign currency deposits, mainly from expatriates, would naturally have a counterpart in currency notes.

Text Figure 9a.
Text Figure 9a.

Adjusted Currency and M1, FX deposits

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

Text Figure 9b.
Text Figure 9b.

Adjusted Money, Inflation, and Exchange Rates, March 2003–December 2006

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

Text Figure 9c.
Text Figure 9c.

Adjusted Narrow Money, Inflation, and Exchange Rates

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

Text Figure 9d.
Text Figure 9d.

12-month Changes in Broad Money

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

45. Our estimates are supported by Kamin and Erricsson (2003), who estimate Argentine holdings of U.S. currency based on an assumption about the stock at year-end 1987 and annual inflows since then based on CMIR data. U.S. currency holdings in Argentina, by their estimates, were more than double the holdings of dollar deposits in domestic banks; dollar currency holdings were also almost as large as all dollar deposits and domestic currency holdings combined.

46. In comparing our measure of dollarization with the standard measure, we find a steadier drop in dollarization since mid-2005 (Text Figure 10). While our measure shows more dollarization, it too trends downward, and the gap from the standard measure narrows from about 25 percent in 2000 to about 15 percent in 2006. The steadier drop in dollarization after mid-2005 found by our measure is more in line with inflation.

Text Figure 10.
Text Figure 10.

Measures of Dollarization

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

47. The integration of monthly imports of foreign currency notes into liquidity forecasts would greatly facilitate monetary operations. Our estimates suggest that the path for velocity and the money multiplier was somewhat smoother than the standard measures imply (Text Table 4 and Text Figure 11). According to simple correlations of inflation, the exchange rate, and the monetary aggregates, adjusted for foreign currency notes in circulation, the addition of foreign currency notes tends to strengthen the correlation between inflation and money (Text Table 5). The correlation between inflation and narrow money is especially strong. Thus, the Angolan authorities could enhance monetary policy by continuing to monitor the various monetary aggregates and adjusting accordingly the balance of local and foreign currencies—both deposits and currency notes in circulation.

Text Table 4.

Velocity and Multiplier

article image
Text Figure 11.
Text Figure 11.

Adjusted Money Multiplier, Dec. 2000–Dec. 2006

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

Text Table 5.

Correlations, Dec. 2003–March 2007

(12-month rates of Change)

article image

Adjusting for remonetization in money demand

48. The monetary aggregate framework hinges on a stable relationship between money and national income (that is, a stable velocity). This relationship (between money and national income), however, appears to change with remonetization as consumer confidence improves after a period of high inflation. This may be why monetary aggregates in Angola have continued to rise since 2003. In 2006 the 12-month real growth rate of base money decreased almost 10 percent, while M3 and M2 grew about 60 percent, reflecting a sharp shift in the money multiplier (Text Figures 12 and 13).20 In any case, money velocity in Angola has not been stable, partly because remonetization is ongoing and data on non-oil GDP are unreliable, making it difficult to project velocity.21

Text Figure 12.
Text Figure 12.

Money Multiplier: M2/Base Money, December 1999–December 2006

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

Text Figure 13.
Text Figure 13.

Real Money Balances, 1999–2006

(Percent)

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

49. According to De Broeck and others (1997), velocity has a tendency to increase early in the implementation of stabilization programs.22 After stabilization, velocity starts to gradually decrease and real money balances typically exhibit a U-shaped pattern over the entire stabilization program. After declining because of high inflation before a program begins, real money balances rise as progress is made in stabilizing the program. This pattern is observed in Angola (Text Figure 14) and in Ghana, Tanzania, and Uganda (Text Table 6).

Text Figure 14.
Text Figure 14.

Growth in Real Money Balances, March 2003–December 2006

(Percent)

Citation: IMF Staff Country Reports 2007, 355; 10.5089/9781451800562.002.A001

Table 6.

Inflation and Velocity, 2000–06

(Percent)

article image
Sources: Regional Economic Outlook and IMF country reports.

50. Analyzing velocity up to and throughout the stabilization effort does not help predict its outcome (De Broek and others 1997). Velocity is expected to decrease as inflation declines and income increases as real money balances are rebuilt. Other factors, however, affect velocity, including reforms in the financial sector (e.g., improvements in market efficiency and the payments system), which tend to accelerate monetization and prompt velocity to fall, as would real exchange rate appreciation. Adjustments are thus required to reflect changes in real money balances. The money supply in Angola could be projected on the basis of changes in real money balances (thus, allowing for remonetization) with due regard to the implied velocity.23 Base money could then be estimated using its links with the components of the broader monetary aggregates, after adjusting it for foreign cash holdings of the public, as discussed above.

51. Given these complications, as well as the uncertainty of growth in the non-oil sector, the authorities would need to target a range for money growth.24 Adjustments made should include which monetary aggregate to target, in line with shifts in the demand for money, reflecting substitution between local and foreign currency, cash and deposits, and more accurate measures of national income, supplemented by other indicators and informed judgments of the BNA.

F. Market Infrastructure and Institutional Shortcomings

52. The most effective transmission channels for monetary policy are availability of credit and interest rates. An efficient financial system is therefore vital. The interest rate channel typically operates through the money market instruments and longer-term bonds as they respond to a contraction or expansion of the economy. However, given Angola’s shallow financial markets and limited competition, it cannot support this kind of policy channel.25

53. The BNA’s monetary policy actions would be enhanced by an efficient interbank market. Although the BNA intervenes in the interbank market using securities, there is a limited secondary market for these securities because banks are reluctant to deal with other banks, partly owing to excess liquidity. This further inhibits the development of an interbank market. By adopting measures to facilitate the development of efficient money markets, the authorities could eventually move to a system where interest rates are an operating target. To develop a securities market, the BNA would need to issue longer-term securities and should develop its capacity to sell foreign exchange directly in the interbank market. Payments system improvements could reduce the cost of holding kwanza cash and deposits and help de-dollarize the economy.

54. The policy response to a surge in foreign exchange inflows can also affect interest rates, private sector activity, and the fiscal repercussions of liquidity management. The issuance of treasury or central bank bills reduces the money supply but can lead to rising interest rates. This action is also costly and difficult to execute in a thin financial market. The sheer magnitude of BNA bills (the equivalent of US$1 billion were sold in 2006 alone) are cause for concern. The government since mid-2006 has ceased to issue securities because it does not need financing. Open-market operations must therefore be carried out using central bank bills, a requirement that has increased the BNA’s operational costs and negatively impacted its interventions on the money market, though negative real interest rates have offset that somewhat.26 Attractive rates for these securities also tend to steer banks away from providing credit to the private sector.

55. Liquidity management should be improved. Foreign exchange sales should account for the bulk of the sterilization of liquidity; short-dated BNA bills can be used to fine-tune liquidity and longer-dated bills to allow base money to be approximately met while the financial sector develops. Fine-tuning reserve requirements would reduce liquidity and smooth its daily fluctuations. Such measures would include reconstituting reserves on foreign currency deposits, eliminating government securities as eligible assets to constitute reserves, no longer deducting vault cash, and allowing reserve holdings to be averaged over the maintenance period. To change the reserve requirements on foreign currency deposits to be held in domestic currency, one-off sale of foreign exchange could be followed by increases in foreign currency deposits to dampen the effect on the exchange rate. Repurchase agreements between banks would also obviate the need for the central bank to intervene frequently. These measures would tend to have a positive impact on BNA’s balance sheet.

56. The lack of autonomy of the central bank impedes the credibility of monetary policy and its implementation. The central bank law calls for it to collaborate closely with the Ministry of Finance on the annual financial programming exercise. The Ministry of Finance is generally more concerned about how an exchange rate appreciation would affect its revenues. The BNA, by contrast, favors a flexible exchange rate but faces cost constraint in issuing bills and the dilemma that an appreciating exchange rate leads to exchange losses. The challenge is thus for the BNA to focus on its key activities while keep its other costs down. The government, meanwhile, must support the costs of monetary policy implementation by promptly recapitalizing the BNA.

G. Conclusions

57. Persistent dollarization, remonetization, and an undeveloped financial market create instability in Angola’s monetary aggregates and hamper monetary policy. The current monetary framework, which is anchored to base money targeting and has helped to rein in inflation, should stand; but it should be adjusted to incorporate all measures of dollarization and be supplemented by other indicators. The exchange rate should be more flexible to allow for it to appreciate nominally in the event that inflationary pressures arise. Additional work to fine-tune the measures of currency in circulation, conduct more sophisticated tests to assess the relationship between inflation and the monetary aggregates, and determine how to incorporate the currency measure in monetary operations is needed. Developing more reliable and timely data would also help policymakers better test the stability of money demand in Angola.

References

  • Balino, Tomas, and Lorena Zamaloa, eds., 1997, Instruments of Monetary Management: Issues and Country Experiences (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Berg, Andrew, and Eduado Borensztein, 2000, “The Choice of Exchange Rate Regime and Monetary Target in Highly Dollarized Economies,” IMF Working Paper 00/29 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • De Broek, Mark, Kornelia Krajnyak, and Henri Lorie, 1997, “Explaining and Forecasting the Velocity of Money in Transition Economies, with Special Reference to the Baltics, Russia, and other Countries of the Former Soviet Union,” IMF Working Paper 97/108 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Gasha, Jose Giancarlo, 2005, “Inflation and the Hard Kwanza Policy,” Angola: Selected Issues and Statistical Appendix (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Gasha, Jose Giancarlo,, and Gonzalo Pastor, 2004, “Angola’s Fragile Stabilization,” IMF Working Paper 04/83 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Gupta, Sanjeev, Robert Powell, and Yongzheng Yang, 2006, “Macroeconomic Challenges of Scaling Up Aid to Africa: A Checklist for Practitioners” (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund, 2005, www.imf.org, Macroeconomics of Managing Increased Aid Inflows: Experiences of Low-Income Countries and Policy Implications (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund, 2006, Inflation Targeting and the IMF, www.imf.org, (Washington: International Monetary Fund).

  • International Monetary Fund, International Monetary Fund, Aid Inflows—The Role of the Fund and Operational Issues for Program Design (forthcoming paper to the Executive Board).

    • Search Google Scholar
    • Export Citation

  • Katz, Menachem, Ulrich Bartsch, Harinder Malothra, and Milan Cuc, 2004, “Lifting the Oil Curse: Improving Petroleum Management in Sub-Saharan Africa” (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Laurens, Bernard, eds., 2005, “Monetary Policy Implementation at Different Stages of Market Development,” IMF Occasional Paper 244 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Mishkin, Frederic, 1999, “International Experiences with Different Monetary Regimes,” Journal of Monetary Economics, Vol. 43, pp. 579– 605

    • Search Google Scholar
    • Export Citation

  • Oomes, Nienke, and Franziska Ohnsorge, 2005, “Money Demand and Inflation in Dollarized Economies: The Case of Russia,” IMF Working Paper 05/44 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Kamin, Steven B., and Neil R. Ericsson, 2003, “Dollarization in Post-Inflationary Argentina,” Journal of International Money and Finance, Vol. 22, No. 2, pp. 185– 211.

    • Search Google Scholar
    • Export Citation

  • U.S. Department of Treasury, 2006, “The Use and Counterfeiting of U.S. Currency Abroad,” September.

1

Prepared by Alexander Kyei.

2

Gasha (2003 and 2005) found a strong relationship between the exchange rate and inflation, but this correlation decreased from one period to another.

3

In September 2003 the Angolan government adopted an anti-inflationary initiative which was known as the “hard-kwanza” policy.

4

Velocity is measured by non-oil GDP as a multiple of end-of-period broad money.

5

These findings are in line with Gasha (2005).

6

See Gasha and Pastor (2004) for more on inflation and stabilization in Angola.

7

The Authorities have argued that the relative stability in the exchange rate has been a consequence rather than a target of their policies.

8

After a sudden step appreciation in late-2005, the exchange rate was stable in 2006. Buoyant oil revenues have kept the kwanza under pressure, but the currency appreciated in nominal terms by only 0.6 percent in 2006.

9

These are the same directors that participate in the Liquidity Committee.

10

Besides the government’s economic team, the Institutional Committee for the Follow-up of the Economic and Social Plan of the government meets to oversee and follow up on government policies and programs, including those of the BNA.

11

A currency board is the simplest monetary system with an externally fixed value—it is simple to administer and has the highest credibility. Therefore, under this arrangement monetary operations are relatively passive—domestic interest rates adjust in line with those of the reserve currency country.

12

See IMF (2006) for more on inflation targeting.

13

The measures used for Angola are in parenthesis.

14

A March 2005 technical assistance report indicates that data sources for the national accounts continue to have serious shortcomings in the basic statistical sources, whether derived from surveys (very low response rates) or those derived from administrative sources (low standards and/or insufficient level of detail). The World Bank is providing assistance in this area.

15

Berg and Borensztein (2000) find that broad monetary aggregates that include foreign currency deposits are more informative than those that do not. However, unlike Kamin and Ericsson (2003) they fail to find that adding a measure of dollar currency in circulation helps predict price levels.

16

This is based on estimates of the Central Bank of Russia and Moscow’s Bureau of Economic Analysis. They also conclude that an acceleration in domestic currency growth should not necessarily be inflationary to the extent that it reflects de-dollarization; neither should a slowdown in domestic money growth imply lower inflation if it reflects a slowing of de-dollarization or renewed dollarization.

18

Foreign currency deposits grew more than tenfold in 2002–06.

19

The revised monetary aggregates are denoted as M3+, M2+, M1+, and Cu+.

20

The composition of broad money also changes with an increase in the share of domestic bank deposits compared with cash and foreign currency deposits.

21

In the early part of Ghana’s stabilization program of the 1980s, IMF programs targeted a declining velocity as inflation was brought under control. Staff projections for Angola have targeted increases in real money balances (M3, M2); however, the recent drop in base money necessitates a review of these projections.

22

This is based on data from the Baltics, Russia, and other countries of the former Soviet Union.

23

While we have estimated U.S. dollar currency in circulation in the previous section, lack of reliable time series data on the non-oil sector does not make it possible to estimate money demand at this time.

24

Historical evolution of the target may provide a basis for setting the target range width.

25

The financial sector comprises 12 commercial banks, four insurance companies, two pension funds, and eight foreign exchange bureaus. During the 2000-04 period, commercial bank assets represented about 25 percent of GDP. With strong real growth in GDP, this figure fell to 19 percent in 2005. The two state-owned commercial banks and the two largest private commercial banks held 74 percent of deposits.

26

The government should bear the cost of implementing monetary policy, as price stability is a public good that benefits the whole economy.

References

  • Alexius, Annika, 2005, “Productivity Shocks and Real Exchange Rates, Journal of Monetary Economics, 52, pp. 555– 66.

  • Balassa, Bela, 1964, “The Purchasing Power Parity Doctrine: A Reappraisal,” Journal of Political Economy, 72, pp. 584– 96.

  • Chang, Tsangyao, Hsu-Ling Chang, Hsiao-Ping Chu, and Chi-Wei Su, 2006, “Does PPP Hold in African Countries? Further Evidence Based on a Highly Dynamic Non-Linear (Logistic) Unit Root Test,” Applied Economics, 38, pp. 2453– 59.

    • Search Google Scholar
    • Export Citation

  • Chudik, Alexander and Joannes Mongardini, 2007, “In Search of Equilibrium: Estimating Equilibrium Real Exchange Rates in Sub-Saharan African Countries,” International Monetary Fund, Working Papers, WP/07/90, Washington.

    • Search Google Scholar
    • Export Citation

  • Drine, Imed and Christophe Rault, 2005, “Can the Balassa-Samuelson Theory Explain Long-Run Real Exchange Rate Movements in OECD Countries?,” Applied Financial Economics, 15, pp. 519– 30.

    • Search Google Scholar
    • Export Citation

  • Edwards, Sebastian, 1989, Real Exchange Rates, Devaluation and Adjustment, Cambridge, Massachusetts, MIT Press.

  • Edwards, Sebastian, and Miguel Savastano, 2000, “Exchange Rates in Emerging Economies: What Do We Know? What Do We Need to Know?,” in Economic Policy Reform: The Second Stage, edited by Anne Krueger, Chicago, University of Chicago, Press.

    • Search Google Scholar
    • Export Citation

  • Froot, Kenneth A. and Kenneth Rogoff, 1995, “Perspectives on PPP and Long-Run Real Exchange Rates,” in Handbook of International Economics, Vol.3, edited by Gene M. Grossman and Kenneth Rogoff, Amsterdam, Elsevier Science Publishers, pp. 1647– 88.

    • Search Google Scholar
    • Export Citation

  • Gelbard, Enrique, and Jun Nagayasu, 2004, “Determinants of Angola’s Real Exchange Rate,” Development Economics, Vol. XLII, pp. 392– 404.

    • Search Google Scholar
    • Export Citation

  • Hinkle, Lawrence E. and Peter J. Montiel, 1999, “Exchange Rate Misalignment: Concepts and Measurement for Developing Countries, “Oxford, Oxford University Press.

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund, 2006, Methodology for CGER Exchange Rate Assessments,” Washington.

  • Johansen, Soren, and Katarina Juselius, 1988, “Hypothesis Testing for Cointegration Vectors: With An Application to the Demand for Money in Denmark and Finland,” Discussion Paper 88, University of Copenhagen, Institute of Economics (Denmark).

    • Search Google Scholar
    • Export Citation

  • Rogoff, Kenneth, 1996, “The Purchasing Power Puzzle,” Journal of Economic Literature, Vol. 34, pp. 647– 68.

  • Samuelson, Paul A., 1964, “Theoretical Notes on Trade Problems,” Review of Economics and Statistics, 46, pp. 145– 54.

  • Sargan, John D. and Alok Bhargava, 1983a, “Testing Residuals from Least Squares Regression for Being Generated by the Gaussian Random Walk, “Econometrica, 51, January, pp. 153– 74.

    • Search Google Scholar
    • Export Citation

  • Sargan, John D. and Alok Bhargava, 1983b, “Maximum Likelihood Estimation of Regression Models With First-Order Moving Average Errors When the Root Lies on the Unit Circle, “ Econometrica, 51, May, pp. 799– 820.

    • Search Google Scholar
    • Export Citation

Annex 1—Description of Key Variables

article image
Source: World Economic Outlook.

Appendix Tables

Appendix Table 1.

Angola: Correlation Matrix, 1980–2006

article image
Sources: Angolan authorities and IMF staff estimates.
Appendix Table 2.

Angola: Covariance Matrix, 1980–2006

article image
Sources: Angolan authorities, and IMF staff estimates.
Appendix Table 3.

Angola: Estimated Persistency of RER, 1981–2006

(Dependent variable is LnRER)

article image
Source: IMF staff estimates.
Appendix Table 4.

Residual Normality Tests

(Orthogonalization: Cholesky (Lutkepohl))

article image
Source: IMF staff estimates.
Appendix Table 5.

Residual Covariance Matrix

article image
Source: IMF staff estimates.
Appendix Table 6.

Residual Correlation Matrix

article image
Source: IMF Staff estimates.

Statistical Appendix

Table 1a.

Angola: Basic Data, 2002–06

article image
Sources: Angolan authorities and IMF staff estimates.

As a percentage of broad money at the beginning of the period.

Increase = appreciation.

Excludes late interest.

Includes government deposits in overseas accounts.

Scheduled debt service as percent of exports of goods and services.

Table 1b.

Angola: Basic Data

article image
Sources: Angolan authorities, Poverty Reduction Strategy, 2004; World Bank, World Development Indicators, 2006; and United Nations Development Program, Human Development Report, 2006.

Human development report, 2006. Percent of population with sustainable access to an improved water source.

Table 2.

Angola: Gross Domestic Product by Sector of Activity, 2002–06

article image
Sources: Angolan authorities and IMF staff estimates.

Liquefied petroleum gas.

Table 3.

Angola: Oil Production by Oil Field, 2002–06

article image
Source: Ministry of Petroleum and IMF staff estimates.

At year’s end.

Table 4.

Angola: Oil Balance, 2002–06

article image
Sources: Angolan authorities and IMF staff estimates.

As reported in balance of payments. Other sources differ slightly.

Includes pipeline losses and field consumption, as well as any discrepancies.

As reported in balance of payments; excludes natural gas liquids.

Table 5.

Angola: Mining Production, 2002–06

article image
Sources: Ministry of Petroleum, Endiama, and IMF staff estimates.