Information about Asia and the Pacific Asia y el Pacífico
Journal Issue

Indonesia: Selected Issues

International Monetary Fund
Published Date:
September 1997
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details

III. Implications for Monetary Policy of Large Capital Flows1

A. Introduction and Summary

1. Macroeconomic policy management in Indonesia has been complicated in recent years by large capital inflows. The authorities have faced a dilemma between maintaining monetary stability to contain inflation and limiting exchange rate appreciation to maintain external competitiveness. In practice, monetary policy has leaned more toward resisting an exchange rate appreciation and, as a result, money growth has tended to exceed the pre-announced target ranges. Relatively high domestic interest rates have compounded the policy dilemma by further increasing capital flows. With the prospect of sustained capital inflows over the medium term, reflecting the strength of fundamentals, the authorities will find it difficult to exercise effective monetary control while limiting real exchange rate appreciation.

2. This paper provides a preliminary assessment of medium-term options for monetary policy in the context of sustained large capital flows. The empirical relation between monetary policy instruments, monetary aggregates, financial market indicators and inflation suggests that monetary aggregates alone do not provide adequate indications of inflation pressures. The analysis in this paper suggests that—over the medium term—a monetary policy framework based on a wider set of financial and real sector variables and permitting greater exchange rate flexibility would constitute a more viable disinflation strategy. The paper also presents a preliminary analysis of the feasibility of moving to an inflation targeting approach. The explicit target range for inflation would need to be relatively wide because of the volatility of consumer prices, especially food prices.

3. As background, the paper examines the external and domestic factors that may have contributed to capital flows. The large share of foreign direct investment suggests that capital flows have been driven largely by the strength of domestic fundamentals and are likely to be sustained at relatively high levels (Section B). While the macroeconomic effects of capital flows have been similar in many emerging market countries (Section C), a review of experience in Indonesia suggests that, while the policy mix has helped to maintain stability, there has been excessive reliance on monetary policy, which in turn has been constrained by exchange rate objectives (Section D). The final part of the paper evaluates the medium-term options for monetary policy, including the monitoring of a wider set of real and financial indicators, and the feasibility of an inflation targeting approach (Section E).

B. External and Domestic Factors Influencing Capital Flows

4. The surge in private capital inflows during the 1990s contrasts sharply with the experience of previous decades, when official flows—particularly to the government sector—accounted for a much larger proportion of aggregate capital inflows. Moreover, the bulk of private capital flows in earlier periods comprised bank lending, often to public entities, whereas recent flows have been dominated by foreign direct investment. Portfolio investment flows, which were very modest until recently, have also increased. By contrast, net long-term borrowing by government and public enterprises has declined markedly in recent years, reflecting in part, the repayment of debt accumulated earlier.

External factors

5. External developments, particularly the cyclical downturn in economic activity and the associated decline in interest rates in the large industrial countries during the early 1990s, played an important role in the initial surge in capital flows to many emerging market countries. Indonesia was not at first a major recipient of these flows. In fact, there appears to have been no inverse relation between U.S. interest rates and equity prices on the Jakarta Stock Exchange, at that time, equity prices in Indonesia fell in line with the decline in U.S. interest rates (Chart 1). Over the more recent period, however, stock price movements in Indonesia have been more closely related to changes in industrial country interest rates and developments in other emerging markets.

Chart III.1Indonesia: Stock Prices In Emerging Markets And United States Interest Rates, 1990-96

Sources: IFC, Emerging Markets database; and IMF, International Financial Statistics.

6. A number of empirical studies that have attempted to distinguish between domestic and external influences suggest that external factors have played a greater role in Latin America than in Asia (see Calvo, Leiderman, and Reinhart (1993), Chuhan et al (1993), and Fernandez-Arias (1993)). These studies suggest that for some countries in Latin America, external factors typically explained about 50 percent of the variation in capital flows, while amongst Asian emerging market countries, including Indonesia, external factors accounted for about 30 percent. To some extent these differences reflect the greater share of foreign direct investment in aggregate net capital flows in Asia, typically attracted by longer-term considerations. In the aftermath of the Mexican financial crisis, contagion effects among Asian countries were relatively shortlived and quickly contained. Indeed, net private capital flows to Asia in 1995 were higher than in 1994.

7. The recent pattern of capital flows seems to reflect a longer-term trend toward globalization and international diversification of industrial country investments. Current trends in the world economy suggest that developing countries, and especially the rapidly growing emerging market countries in Asia, will account for an increasing proportion of world output. Many of these countries, such as China, India, and Indonesia, represent some of the largest markets in the developing world. This should provide strong incentives to investors in industrial countries to diversify their portfolios, which at present are underweighted in developing country investments relative to the size of the emerging stock markets.2 These markets, although more volatile, have yielded relatively high returns and because of their low correlation with financial markets in industrial countries, provide scope for portfolio diversification.3

Domestic factors

8. In common with many other countries in the region, the strength of the domestic fundamentals in Indonesia, have been instrumental in attracting capital flows. Prudent fiscal policy has enabled Indonesia to maintain macroeconomic stability over long periods which, in turn, has helped to sustain external confidence. Structural reforms, especially trade and financial liberalization, including fewer restrictions on acquisition of assets by foreign residents, has paved the way for an expansion of domestic equity markets. The promotion of private sector activity—as in many other countries in the region—has resulted in a strong expansion of exports and intra-regional trade and sustained rapid output growth.

9. During 1990-96, foreign direct investment flows accounted for about one-third of net private capital flows, and over the last two years this proportion has risen to about one half (Chart 2). Although the relatively low share of portfolio investment flows may partly be due to the comparatively small size of the Indonesian stock market and the absence of a well-developed bond market, the share of foreign direct investment flows is similar to that in many Asian emerging market countries. Other investment flows, including bank lending to private and official entities generally have declined in importance but have been subject to larger fluctuations. Some of these flows, particularly to the banking sector, are also primarily attracted by high short-term interest rates, including on central bank securities used to conduct open market operations.

Chart III.2Indonesia: Composition Of Capital Flows, 1990-96

(In billions of U.S. dollars)

Source: IMF, World Economic Outlook database.

10. The liberalization of trade and investment regimes improved the climate for foreign direct investment. Foreign investment approvals (in terms of value) have increased rapidly since 1989 (Table 1). Recent investment projects are widespread across sectors, but the largest increases in terms of value are concentrated in resource based manufacturing (chemicals, basic metallic industries and paper products), services (hotels, power supply), and infrastructure (including privatization in the electricity and gas sector). While Asian countries still represent the largest group of foreign direct investment to Indonesia—primarily reflecting inflows from Japan—there have been large increases from Australia, Germany, the United Kingdom, and the United States.

Table III.1.Indonesia: Foreign Direct Investment Approvals by Sector, 1991-96 1/(In billions of U.S. dollars)
Mining and quarrying0.
Source: Data provided by the Indonesian authorities.

Includes new projects and extensions of existing projects, but excludes investment in petroleum, banking, insurance, and leasing sectors. Components may not add to totals owing to rounding.

Source: Data provided by the Indonesian authorities.

Includes new projects and extensions of existing projects, but excludes investment in petroleum, banking, insurance, and leasing sectors. Components may not add to totals owing to rounding.

C. Macroeconomic Effects of Capital Inflows

11. It is useful to compare macroeconomic developments in Indonesia with those in other emerging market countries that have also experienced sustained inflows during the 1990s, when examining the impact of large-scale capital inflows and evaluating the policy response (Table 2). In most cases, inflows have tended to reduce domestic interest rates, raise aggregate expenditures, increase inflationary pressures, and widen current account deficits in the recipient countries. These consequences are attributable in part to exchange market policies that have attempted—with varying degrees of success—to limit the nominal appreciation of domestic currencies because of concern about the adverse effects of real exchange rate appreciation on competitiveness.

Table III.2.Selected Emerging Market Countries: Macroeconomic Indicators 1/(Annual averages, in percent of GDP unless otherwise noted)
Selected Asian emerging market countries
Net private capital flows1.53.5
Current account balance-0.8-2.6
Central government fiscal balance-5.1-2.9
Private consumption61.559.7
Private saving 2/21.921.4
Private investment17.621.2
Total saving24.826.3
Total investment25.529.0
Real GDP 2/7.35.9
Consumer prices (median) 3/6.08.6
Real effective exchange rate 3/-5.7-1.0
Selected Latin American emerging market countries
Net private capital flows0.33.3
Current account balance-0.6-2.3
Central government fiscal balance-4.1-0.5
Private consumption65.167.6
Private saving16.914.3
Private investment16.416.2
Total saving20.718.3
Total investment21.120.5
Real GDP 3/2.22.9
Consumer prices (median) 3/80.623.3
Real effective exchange rate 3/2.62.4
Net private capital flows2.14.5
Current account balance-3.0-2.6
Central government fiscal balance-1.1-0.1
Private consumption60.057.0
Private saving 2/23.223.5
Private investment24.626.0
Total saving25.730.3
Total investment29.433.0
Real GDP 3/6.97.8
Consumer prices (median) 3/7.58.7
Real effective exchange rate 3/-10.80.4
Source: IMF World Economic Outlook database.

Comprises Argentina, Brazil, Chile, Colombia, Hong Kong, India, Israel, Korea, Malaysia, Mexico, Pakistan, Peru, the Philippines, South Africa, Thailand, Turkey, and Venezuela. China, Singapore, and Taiwan Province of China are excluded because of their large current account surpluses in the 1991-95 period.

For Indonesia, the data reported are for 1987-90.

Annual percent change.

Source: IMF World Economic Outlook database.

Comprises Argentina, Brazil, Chile, Colombia, Hong Kong, India, Israel, Korea, Malaysia, Mexico, Pakistan, Peru, the Philippines, South Africa, Thailand, Turkey, and Venezuela. China, Singapore, and Taiwan Province of China are excluded because of their large current account surpluses in the 1991-95 period.

For Indonesia, the data reported are for 1987-90.

Annual percent change.

12. Despite facing similar policy dilemmas, there have been marked differences in macroeconomic performance across the emerging market countries, especially between the major recipient countries in Asia and Latin America. While increased private capital flows have often been associated with widening current account deficits—in Indonesia the current account deficit has also increased—the domestic counterpart is of larger current account deficits have developed differently. In Indonesia, as in other Asian emerging market countries, the ratio of private consumption expenditure to GDP declined by about 3 percentage points between the second-half of the 1980s and the first-half of the 1990s, while the ratio of private investment expenditure to GDP increased by just under 3 percentage points. In most Latin American countries, by contrast, the share of consumption expenditure has increased, while private investment expenditure remained broadly constant.

13. These differences in the division between consumption and investment expenditure mirror markedly different national saving patterns. Countries characterized by relatively high public sector saving have managed to sustain high private saving rates, whereas private saving rates have remained relatively low in countries with weak fiscal positions. Even if public sector dissaving has a positive effect on private saving—for example, through Ricardian effects, whereby the private sector anticipates future tax liabilities—the offset is likely to be partial. Bascand and Razin (1997) find evidence that fiscal consolidation from the mid-1980s in Indonesia did lead to some increase in national saving: the decline in private saving did not fully offset the higher public saving. Moreover, as they emphasize, when fiscal consolidation is undertaken in periods when debt levels are high—as in recent years—an increase in public saving should lead to a broadly equivalent increase in national saving.

14. An important concern for policymakers is whether capital flows complement domestic saving and enhance longer-term growth prospects. In countries where capital flows primarily finance consumption and substitute for domestic saving, policymakers need to be particularly vigilant about the sustainability of capital flows and current deficits. Moreover, while foreign resources can improve growth prospects, all countries need to ensure that financial stability is preserved in order to reduce the risk of disruptive changes in financial market sentiment that would necessitate severe macroeconomic policy adjustment.

D. Macroeconomic Policy Responses to Capital Flows

15. The appropriate policy response to large-scale capital inflows depends on a number of factors, particularly as to whether they are largely driven by external factors or by domestic fundamentals. The appropriate policy response may also depend on the composition of the inflows—whether they comprise largely direct investment or other flows that are more volatile or debt creating—and on the capacity of the domestic banking system to intermediate large foreign inflows.4 Improvements in the fiscal position, including when central government budgets are in broad balance and where initial current account deficits emanate from the private sector, would help to reduce demand pressures and the risk of sudden reversals of capital flows.5 If capital inflows are associated with an autonomous increase in the demand for money (perhaps owing to rapid financial market development), the expansion of the domestic money supply will not be inflationary; in such cases the inflows should be accommodated.

16. The potential inflationary impact of capital inflows will be influenced importantly by the exchange rate regime and the extent to which the monetary authorities sterilize their effect on the domestic money supply. Under a flexible exchange rate regime, inflows are likely to lead to an appreciation of the real exchange rate and should not give rise to sustained inflationary pressure. Under a fixed exchange rate regime, inflows will tend to increase the money supply and domestic demand, and raise domestic prices, thereby also leading to a real exchange rate appreciation. Thus, resisting nominal exchange rate appreciations in the face of sustained capital inflows may not ultimately help to limit real exchange rate appreciation.

Fiscal policy

17. The dilemma for fiscal policy in Indonesia has been that, despite the long track record of prudent fiscal management, capital flows have periodically resulted in overheating. The overall fiscal position—which improved markedly following the consolidation measures in the 1980s—has been maintained in broad balance since the early 1990s. Nevertheless, the current account deficit has widened to almost 4 percent of GDP over the last two years.

18. Fiscal policy has been used only moderately for countercyclical purposes, in part because available instruments are not sufficiently flexible for short-run macroeconomic management. However, monetary tightening—the main macroeconomic response to demand pressures stemming from capital flows—has not reduced domestic absorption sufficiently to prevent the recent widening of the current account. This suggests that further fiscal tightening would be a more appropriate policy response. It would act to contain the current account deficit and help to bring down interest rates nearer to those in international capital markets and in other countries in the region. Lower interest rates would also help to limit short-term capital flows and thereby reduce risks for the banking system.

Exchange rate policy

19. During the period of large capital flows, the authorities have generally managed to target the real exchange rate within a relatively narrow range by depreciating the nominal exchange rate vis-à-vis the U.S. dollar (and thereby the nominal effective exchange rate) to broadly offset inflation differentials between the two countries (Chart 3). The nominal rupiah/U.S. dollar exchange rate depreciated by an average of about 0.35 percent per month between 1989 and 1995 while, over the same period, the ratio of the consumer price index in Indonesia to the U.S. index increased by a similar amount. During 1996, the real effective exchange rate appreciated, in part reflecting the widening of the exchange intervention band, but also owing to the increase in the U.S. dollar/Japanese yen exchange rate.

Chart III.3Indonesia: Exchange Rates And Relative Inflation RATES, 1988-96

Sources: IMF, International Financial Statistics and Information Notice System.

20. The resistance to nominal exchange rate appreciation has limited progress on reducing inflation. During much of the 1990s, inflation in Indonesia—averaging about 8 percent—has been higher than in other ASEAN countries. There is widespread consensus, both in the theoretical literature and from experience in other emerging market countries, that attempts to target real exchange rates in the face of sustained increases in capital flows lead to higher inflation (see, for example, Calvo, Reinhart, and Vegh (1995)). For Indonesia, Siregar (1996) finds strong evidence that a stable real exchange rate has been the main objective of monetary and exchange rate policy and that the depreciation of the nominal exchange rate has resulted in higher inflation.

21. Resisting upward pressures on the nominal exchange rate will probably not help to limit real exchange rate appreciation over the medium term. In fact, exchange market pressures have strengthened further since mid-1995 as a result of sustained capital inflows. Allowing the real exchange rate to appreciate via nominal exchange rate appreciation (rather than through an increase in domestic prices) has a number of advantages. First, it helps to contain inflation by limiting exchange rate pass-through to domestic prices and avoids the potentially high output costs of disinflation policies in future. Second, it insulates the money supply, domestic credit, and more generally, the banking system from capital inflow. Finally, allowing the nominal exchange to fluctuate in a wider range introduces uncertainty and increases perceived exchange risk which would help to deter speculative flows.6

22. Despite the drawbacks of nominal exchange rate depreciation and real exchange targeting, an important question for policymakers in Indonesia is whether a real exchange rate appreciation is warranted. Evaluating the appropriate level of the real exchange rate is notoriously difficult, and calculations of equilibrium exchange rates are typically subject to wide margins of uncertainty. Thus, it is often difficult to make judgments about exchange rate policy on the basis of estimates from macroeconomic models, particularly when the degree of deviation from the “equilibrium” exchange rate is relatively small.7 There are, however, other considerations that can help in this assessment.

23. In countries where capital flows are attracted primarily by the high returns associated with rapid industrialization and sustained productivity gains, a tendency for the real exchange rate to appreciate is more likely to be an equilibrating response to improvements in profitability. In some cases, exchange market pressures reflect expectations that the real exchange rate will appreciate in the medium to longer run, in line with higher profitability in the traded goods sector. Performance of export sectors can therefore be an important guide to the underlying reasons for the inflows and their sustainability. Profitability would improve if relative unit labor costs—often a good measure of competitiveness of traded goods—rise less than the general price level in the economy.

24. The assessment of Indonesia’s export performance over the long term outlined in Chapter I suggests that export growth has been determined mainly by structural reform measures, particularly the rising capital-labor ratio associated with higher investment, including foreign investment and trade liberalization that have boosted supply. The world demand for Indonesia’s manufactured goods does not appear to be strongly affected by changes in relative prices. These results suggest that further real exchange rate appreciation would not hold back export growth.

Monetary policy

25. The monetary policy response to capital flows has been determined largely by Indonesia’s exchange rate policy. While the authorities have periodically engaged in sterilized intervention, monetary tightening has been effected mainly through increases in interest rates. This approach has not yielded monetary stability. Over the last three years, money growth has consistently exceeded pre-announced targets. Interest rate increases have recently been supplemented by strengthened direct measures, including reserve requirements and moral suasion to limit money and credit growth (Table 3).

Table III.3.Indonesia: Monetary and Credit Growth, 1994/95-1996/97(Annual percent change)
Broad Money (M2)Private Sector Credit
Source: Data provided by the Indonesian authorities.
Source: Data provided by the Indonesian authorities.

26. Sterilization of foreign inflows can help to contain domestic credit expansion, but financial markets in most emerging market countries are not large enough to sterilize sustained capital flows. Sterilization can be helpful as an initial response, in part because it can be implemented relatively quickly and allows policymakers time to assess whether the inflows are likely to be transitory or not. Even in the short term, sterilization can entail significant costs for the monetary authorities, especially if it takes the form of open market operations—as in Indonesia—that lead to an increase in domestic interest rates. The approximate cost of sterilization will be determined by the difference between interest rates paid on domestic securities issued by the central bank and interest income earned on foreign exchange reserves.

27. Use of direct controls to restrict bank lending has been motivated largely by the desire to limit interest rate increases and thereby deter short-term capital flows. Experience in recent years suggests that this policy has had only limited success. Moreover, such policies may create incentives for banks to by-pass controls and engage in off-balance sheet lending. The recent expansion in the commercial paper market appears to be partly due to increased pressure on banks by Bank Indonesia to restrict lending. To circumvent these directives, banks purchase commercial paper for sale to depositors and this form of intermediation of funds between depositor and borrower does not show up in bank balance sheets.

E. Medium-Term Monetary Policy Options

28. Empirical evidence on the relation between money growth, economic activity, and inflation in Indonesia raises the question whether the authorities should focus exclusively on monetary aggregates to guide monetary policy actions. An important prerequisite for operating such a policy framework is a stable and predictable demand for money because only then can the monetary authorities have a reasonable degree of confidence that, if actual money growth is above target, there is likely to be upward pressure on prices and the need for policy actions to tighten monetary conditions. If money demand behavior is not predictable, the authorities will face the difficulty of not knowing whether money growth reflects an underlying shift in the private sector’s desire to hold money balances or whether money holding is only temporarily above longer-term needs.

29. Empirical estimates of money demand equations in Indonesia do not provide strong evidence of stable relationships over the long term. Recent estimates by Dekle and Pradhan (1997) for the major ASEAN countries suggest that narrow or broad money growth in Indonesia cannot be explained by changes in real income and interest rates. These findings are attributed to the rapid changes in financial markets that have substantially increased the extent of financial deepening (as measured by the ratio of M2 to GDP or by the ratio of private sector credit to GDP). These results imply that money growth rates will typically be poor predictors of future inflation and output trends.

Monitoring a broader set of indicators

30. In the context of substantial changes in financial markets and large capital flows, the dilemma for policymakers is to decide on the extent that actions should be constrained by pre-announced targets. If money growth rates cannot be relied on exclusively to gauge inflationary pressures, it would be preferable to downgrade the emphasis on monetary targets and base policy actions on a wider set of indicators. This approach was adopted in a number of industrial countries, such as Canada, the United Kingdom, and the United States following financial deregulation and advances in transaction technologies. Reducing the emphasis on monetary targets in Indonesia may also be appropriate for maintaining credibility. If money targets are announced but policy actions are not seen to be based on money growth because specific episodes of rapid money growth are judged not to indicate inflation pressures, there is a risk that credibility of policies could be undermined.

31. Whether monetary policy in Indonesia would be better guided by a broader set of indicators depends on identifying variables that could be used as leading indicators of inflation. A preliminary assessment is that, while broad money does provide information on future inflation, a number of other variables also contain leading indicator properties for inflation. Indeed, the rate of depreciation of the nominal exchange rate vis-à-vis the U.S. dollar and the nominal effective exchange rate, and the nominal interest rate differential relative to the United States have stronger predictive power for inflation in Indonesia. Among real sector variables, the output gap (the difference between actual and trend output)—which essentially acts as a proxy for demand conditions relative to capacity output—has less predictive power than monetary and financial variables, although it is still useful in assessing inflationary pressures (Table 4). 8 These results, however, suggest only that the indicator variables contain information about movements in future inflation beyond that already contained in current and past values of inflation itself.

Table III.4.Indonesia: Qualitative Results of Financial and Real Variables for Inflation 1/
VariableLeading Indicator PropertiesExplaining Forecast Error

Variance of Inflation
(Bivariate Granger Causality)(Bivariate Variance Decomposition)
Monetary baseMediumMedium
Narrow money (M1)MediumMedium
Broad money (M2)MediumMedium
Indonesia/U.S. interest
Rupiah/US$ exchange rateMediumStrong
Nominal effective exchange rateStrongStrong
Output gapWeakMedium
Oil price inflationWeakWeak
Commodity price inflationWeakWeak
Source: Staff calculations.

These qualitative indications are based on the significance levels of each variable in the Granger Causality tests. “Strong” indicates that the variable is significant at the 95 percent significance level, “medium” denotes the variable was significant at the 90 percent level, and “weak” indicates significance level of less than 90 percent.

Source: Staff calculations.

These qualitative indications are based on the significance levels of each variable in the Granger Causality tests. “Strong” indicates that the variable is significant at the 95 percent significance level, “medium” denotes the variable was significant at the 90 percent level, and “weak” indicates significance level of less than 90 percent.

32. Although this analysis supports moving away from a monetary target framework, the authorities will also need to ensure that, in the absence of an explicit intermediate target, credibility of the central bank’s resolve to maintain low inflation is sustained. When the assessment of monetary conditions is based on a range of indicators, there is always a risk that policy inaction could be seen as a weakening in the anti-inflation stance. To minimize these risks, the monetary policy process needs to be more transparent so that market participants are fully informed about the rationale for policy decisions.

Feasibility of an inflation targeting framework

33. When monetary policy assessments are based on a range of indicators, or when policy is framed around intermediate targets, there may be a tendency for the policy framework to lack an explicit forward-looking element. In a number of countries, the formulation of an explicit medium-term objective centered on inflation targets has helped to fill an important gap following the abandonment of strict monetary targets and, in some cases exchange rate targets.9 A key advantage of the inflation targeting approach is that it forces the authorities to base policy actions on their forward-looking assessment of inflation, thereby helping private agents to gauge the policy stance.10

34. The adoption of inflation targets is not, however, costless because forward-looking assessments are typically subject to wide uncertainty. Forecasting errors in inflation can be relatively large, giving rise to a trade-off between flexibility and credibility of policies. To ensure that targets are met, central banks may set a relatively wide target range, but this may not enhance credibility. In Indonesia, a prerequisite for moving to an inflation targeting framework is to model the inflation process and to examine the magnitude and source of forecast errors. If, for example, the economy is subject to frequent supply shocks, or major components of the consumer price index are volatile, actual inflation may deviate widely from the target range. In such circumstances, it may be appropriate for monetary authorities not to tighten monetary conditions, but this could result in loss of credibility.

35. A preliminary assessment indicates that, although quarterly inflation rates can be forecast reasonably well, there is a wide margin of uncertainty. As a first step, the variables that contain good leading indicator properties (identified above) are used to estimate quarterly inflation. This exercise—which is reported in the companion paper on export performance—confirms that many of the indicator variables are useful to forecast inflation. The inflation forecasting equation—based on money, exchange rates, and past inflation—tracks actual inflation quite closely when it is used to forecast inflation from late 1993 onward (Chart 4). However, confidence intervals around one year ahead of inflation projections are rather large.

Chart III.4Indonesia: Forecast Of Inflation, 1986-96

(In percent)

Sources: Data provided by the Indonesian authorities; and Fund staff estimates.

36. In some countries that have adopted inflation targets, forecast uncertainty has been reduced by omitting volatile components from the price index. Such an option does not appear to be feasible in Indonesia. A decomposition of the variance of the CPI in Indonesia indicates that the volatility of inflation is mainly due to the volatility of food prices (Chart 5). Although in a statistical sense volatility would be lower if food was excluded from the index, an inflation measure would be of limited economic interest because of the relatively large weight of food in the consumer price index. A more disaggregated decomposition of the CPI would help to identify specific food items that account for the high volatility of food prices and help determine whether it would be possible to reduce aggregate volatility of the CPI by excluding individual food items.

Chart III.5Indonesia: Variability Of Inflation, 1992-97

(In percent)

Sources: Data provided by the Indonesian authorities; and Fund staff estimates.

Policy requirements for inflation targeting

37. Monetary policy can be based on an inflation targeting framework if the authorities accelerate real sector reforms, particularly structural and trade liberalization reforms. The volatility of food prices reflects in part the variability of supply conditions, but much of it can be alleviated by increasing the role of the private sector in the production, distribution, and marketing of agricultural commodities. Although public sector involvement in the agricultural sector is based on the objective of stabilizing prices to offset variations in supply conditions, state intervention goes beyond genuine areas of market failure. It has also not been effective in reducing the volatility of food prices. With the development of forward markets in a number of commodities and the greater capacity of the private sector to bear the inherent risks, private provision in the agricultural sector could shield consumers from the underlying volatility in production levels and thereby reduce volatility of prices.

38. Further, the impact of variations in domestic supply conditions on prices could also be reduced by liberalizing imports of major agricultural commodities, such as rice and soyabeans. In periods when domestic supplies are adversely affected by unfavorable weather conditions, imports of agricultural commodities could make up for the shortfall and limit price fluctuations. Moreover, trade liberalization in these areas would enhance efficiency by increasing competition. This would provide incentives for domestic producers to manage inventories more efficiently and avoid the need for large changes in prices.

39. Although the evidence presented in this paper does not support moving—in the short term—to an inflation targeting framework of the type adopted in a number of industrial countries, policy effectiveness would be enhanced by following a number of key ingredients of the targeting approach. Indeed, much of the success of inflation targeting in other countries stems from the greater consistency and transparency that has been associated with the forward-looking approach. In Indonesia as well, the authorities have framed the broad policy objectives around the medium-term development plan in which lower inflation is a key target. Nevertheless, it is important to emphasize that market participants may not always assign sufficient credibility to this goal, especially when monetary targets continue to be announced but policy actions appear to be guided by exchange rate concerns. The authorities could base monetary policy decisions more exclusively around a medium-term inflation target, which would provide the system with a strong nominal anchor and, over time as distortions in the real sector are reduced through structural and trade reforms, move to a narrower target.

APPENDIX: Indonesia: Leading Indicators of Inflation

1. Unit root tests to determine the time series properties of the data

Results of the Augmented Dickey-Fuller tests on the levels of all series indicate that most of the series are I(1), although it is unclear whether CPI is I(2). The Augmented Dickey-Fuller test on the first difference of the series, however, suggests that the rate of inflation is stationary.

2. Granger Causality tests

A variable is defined to contain leading indicator properties for inflation, if it adds information about future movements in inflation beyond that which is already contained in past inflation. Formally, Granger Causality tests for the hypothesis that β1 =…=βp =0 in the following regression:

where x is the final target variable and y is the indicator variable.

  • The set of target variables are
  • CPI inflation (denoted INFL in the tables)
  • WPI inflation (denoted WPIG in the tables)
  • WPI (excluding oil) inflation (denoted WPIXOILG in the tables)
  • The set of indicator variables examined are:
  • the rate of deprecation of the nominal exchange rate against the US dollar
  • M0 growth
  • M1 growth
  • M2 growth
  • the rate of deprecation of the nominal effective exchange rate (NEER)
  • the output gap (deviation of actual output from trend output)
  • oil price inflation
  • commodity price inflation
  • Indonesia/U.S. nominal interest rate differential

The marginal significance levels (p-values) for the bivariate Granger Causality tests for lags 1 to 8 indicate that the rate of depreciation of the nominal exchange rate (against the U.S. dollar, and against a basket of currencies—the nominal effective rate) has strong predictive power for future CPI inflation and moderate power for predicting WPI and WPIXOLIG inflation. Interest rate differentials have strong predictive power for inflation at short lags, particularly for oil price inflation.

3. Variance decompositions

Forecast error variance decompositions—calculated using the Choleski decomposition—for bivariate vector autoregressions defined on the target variable and the indicator variable are consistent with the conclusions of the Granger Causality tests: the rate of depreciation of the exchange rate has the strongest predictive power for inflation. The NEER accounts for about 12 percent of the error variance in CPI inflation after a lag of 2 years. The maximum impact of the exchange rate seems to be around 6 to 8 quarters. The interest rate differential also contains some predictive power, accounting for nearly 10 percent of the error variance in inflation after 24 quarters.

4. Variability of inflation

An autoregressive conditional hetroskedasticity (ARCH) model is used to estimate the time varying volatility of CPI inflation and of the major components of the CPI, food, housing, clothing, and all other items. The volatility in inflation is highest in 1993. The source of the volatility during this period was the volatility in food price inflation reflecting shortages due to the low agricultural production caused by floods. In fact, food prices are the major source of volatility in most periods. Over the full sample, volatility of food price inflation accounted for 40 percent of the volatility in overall inflation, followed by clothing, which accounted for 27 percent of the total volatility.


    BascandGeoffrey and AssafRazin (1997) “A Framework for Fiscal Sustainability, with Application to Indonesiapaper prepared for Jakarta ConferenceNovember1996forthcoming IMF working paper.

    • Search Google Scholar
    • Export Citation

    CalvoGuillermo A.LeonardoHeiderman and CarmenReinhart (1993) “Capital Inflows and Real Exchange Rate Appreciation in Latin America: The Role of External FactorsIMF Staff Papers Vol. 40 (March) pp. 10851.

    • Crossref
    • Search Google Scholar
    • Export Citation

    CalvoGuillermo A.CarmeaReinhart and CarlosVegh (1995) “Targeting the Real Exchange Rate: Theory and EvidenceJournal of Development Economics Vol 47 pp. 97133.

    • Crossref
    • Search Google Scholar
    • Export Citation

    ChuhanPunamStijnClaessens and NlanduMamingi (1993) “Equity and Bond Flows to Latin America and Asia: The Role of External and Domestic Factorsunpublished (Washington, D.C. World BankMay).

    • Search Google Scholar
    • Export Citation

    DekleRobert and MahmoodPradhan (1997) “Financial Liberalization and Money Demand in ASEAN Countries: Implications for Monetary PolicyIMF Working Paper WP/97/36March.

    • Crossref
    • Search Google Scholar
    • Export Citation

    Fernandez-AriasEduardo (1993) “The New Wave of Private Capital Inflows: Push or Pull?unpublished (Washington, D.C. World BankNovember).

    • Search Google Scholar
    • Export Citation

    HaldaneAndrew G. (ed.) (1995) Targeting Inflation a conference of central banks on the use of inflation targets organizedbythe Bank of EnglandMarch9101995.

    • Search Google Scholar
    • Export Citation

    HaqueNadeem UlDonaldMathieson and SunilSharma (1996) “Capital inflows to Developing and Transition Countries—Identifying causes and Formulating Appropriate Policy ResponsesAnnex IV in World Economic OutlookOctober1996 pp. 148153.

    • Search Google Scholar
    • Export Citation

    HellerPeter (1997) “Fiscal Policy Management in an Open Capital ReginieIMF Working Paper WP/97/20January.

    International Monetary Fund (1995) International Capital Markets Report: Developments Prospects and Policy Issues (August).

    International Finance Corporation (1996) Emerging Stock Markets Factbook (Washington, D.CMay).

    MontielPeter J. (1996) “Exchange Rate Policy and Macroeconomic Management in ASEAN Countriespaper prepared for Jakarta ConferenceNovember1996forthcomingIMF working paper.

    • Search Google Scholar
    • Export Citation

    SiregarReza Yamora (1996) “Real Exchange Rate Targeting and Inflation in Indonesia: Theory and Empirical EvidencePacific Basin Working Paper Series No. PB96-07December1996Center for Pacific Basin Monetary and Economic Studies Economic Research DepartmentFederal Reserve Bank of San Francisco.

    • Search Google Scholar
    • Export Citation
1The author of this chapter is Mahmood Pradhan.
2For a detailed analysis of the extent of international diversification of institutional investment portfolios, especially the limited holdings of emerging market securities, see the background papers in the IMF’s International Capital Markets Report,1995.
3Recent estimates indicate that the correlation between the return on the Standard and Poor’s index of the top 500 companies listed in the New York Stock Exchange and the return on the IFC composite index for emerging market countries has been less than 0.4 over the period 1990-95. See International Finance Corporation, Emerging Stock Markets Factbook,1996 (Washington, D.C., May 1996). Over the longer term, however, as stock markets deepen in the emerging market countries and become more closely integrated with those in industrial countries, short-term volatility in industrial country markets is likely to be transmitted faster to the emerging stock markets.
4For a schematic assessment of how to identify causes of capital inflows and the appropriate macroeconomic policy responses, see Haque, Mathieson, and Sharma (1996).
5Heller (1997) provides a detailed discussion of the rationale for a more conservative fiscal stance in countries faced with large and volatile capital flows.
6This argument has less force if more flexibility of the nominal rate leads to greater volatility of the real exchange rate, which could have adverse effects on the traded goods sector, especially in the absence of adequate financial instruments to hedge against such uncertainly. Real exchange rate volatility, however, is determined mainly by the macroeconomic environment, particularly the stability of macroeconomic policies.
7See Montiel (1996) for an analysis of equilibrium exchange rates for ASEAN countries.
8The estimation procedure and details of the extent to which movements in the indicator variable help to explain the variance of inflation are outlined in the Appendix.
9There may be a strong case for exchange rate targets in countries that lack credibility and have a history of relatively high inflation, although this has to be weighed against the greater difficulties of managing capital inflows and other real shocks. In any case, the high inflation argument does not apply to Indonesia.
10For an extensive analysis of inflation targeting experience in a number of industrial countries and a review of the key issues, see Haldane (1995).

Other Resources Citing This Publication