98206 I NDONESI A CURRENT ACCOUNT ASSESSMENT MILAN NEDELJKOVIC GONZALO VARELA Trade and Competitiveness MICHELE SAVINI ZANGRANDI Macroeconomics and Fiscal Policy Management Preface Emerging economies, including Indonesia, witnessed a marked reduction in their current account balances in the aftermath of the 2008-09 global financial crisis, driven by a combination of stronger domestic demand relative to developed countries and international interest rates at historically low levels, which allowed emerging and developing countries to finance consumption and investment by borrowing from abroad relatively cheaply. Growing current account deficits caused financing concerns in a number of countries, especially given the perspective of raising global interest rates following normalization of monetary policy in advanced economies. Appropriate policy responses depend on the structure of the economy and on the underlying drivers of current account dynamics. Understanding these drivers, in conjunction with external developments, is therefore crucial. The current account balance can be seen in two different and informative ways. First, the current account balance can expressed as the difference between the value of imports and exports of goods and services plus net factor payments and net transfers. As the trade balance tends to be the largest component of the current account, a current account deficit generally implies a trade deficit, although this need not be the case as countries also earn and pay income and send and receive transfers, most commonly in the form of interest and dividends and remittances. Second, the current account balance can also expressed as the difference between national saving and investment (both public and private). As such, the current account balance indicates the change in a country’s position as international net debtor or creditor. Although there is no strictly preferable current account position, certain positions are more consistent than others with a country’s stage of development. Developing countries tend to be net borrowers because, with less abundant capital, their economies offer higher returns to investment relative to those in developed economies. Particularly large current account deficits, however, may indicate that the economy is overheating, and may make the economy vulnerable to shocks. In this context, the World Bank’s Trade and Competitiveness Global Practice developed a toolkit to better understand the underlying drivers of current account imbalances, the financing structure of such imbalances, as well as their medium-term sustainability. The key objective of this toolkit is to provide policymakers with the deeper understanding of the issue which is necessary to formulate effective policy responses. The toolkit applies an innovative methodology that constitutes an improvement to existing frameworks used to examine external imbalances. The approach is country- focused. Rather than relying on cross-country panel estimations, the models are country-specific, which increases the policy-relevance of its conclusions. The “model- averaging” methodology used, in addition, addresses the frequently overlooked problem of uncertainty with respect to the true model that best describes the determination of the current account balance. 1 1 In this context, model averaging means that different combinations of the potential current account determinants are used as models. The final estimate is obtained by averaging across all the estimated regressions using a suitable criterion to select model weights. This report looks at the case of Indonesia through the lens of this toolkit. It aims to inform policy by shedding light on the drivers of its current account dynamics and the sustainability of the sustained deficits that have emerged since 2011. In addition, this report gauges the sustainability of Indonesia’s current account balance under benign and adverse external sector developments and in a scenario in which the Government embarked in a large infrastructure development program This study is a joint effort of the World Bank’s Trade and Competitiveness Global Practice and the Macroeconomics & Fiscal Policy Management Global Practice. The study was prepared by a team led by Gonzalo Varela (Trade Economist) and comprised of Milan Nedeljkovic (Consultant), and Michele Savini Zangrandi (Macro Economist). The team wishes to acknowledge the extensive comments received from Ashley Taylor (senior economist), Alex Sienaert (country economist), Elitza Mileva (economist), and the support received from Esteban Rojas (research analyst), Magda Adriani (research analyst), Fitria Fitrani (research analyst) and editing by Arsianti. This report is a product of the staff of the International Bank for Reconstruction and Development/The World Bank, supported by funding from the Australian government under the Support for Enhanced Macroeconomic and Fiscal Policy Analysis (SEMEFPA) program. The findings, interpretations, and conclusions expressed in this report do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent, or the Australian government. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The cut-off date for the data in this report was 2013 for the current account determinants analysis that requires annual data, and 2014Q2 for the rest of the analysis based on quarterly data. For more information about the World Bank and its activities in Indonesia please visit: www.worldbank.org/Indonesia. For questions about this report, please contact Gonzalo Varela (gvarela@worldbank.org) or Alex Sienaert (asienaert@worldbank.org). THE WORLD BANK | BANK DUN IA Table of contents EXECUTIVE SUMMARY ........................................................................................................ I 1. Background .................................................................................................................................... I 2.Determinants of the current account in Indonesia ....................................................................... II a. External shocks ................................................................................................................................................. III b. Domestic policies .............................................................................................................................................. III c. Integration in the global marketplace ................................................................................................................ III d. Stage of development and demographics .......................................................................................................... IV 3.Policy discussion .......................................................................................................................... IV A. STRUCTURE OF THE REPORT ................................................................................... 1 B. CONTEXT .........................................................................................................................2 1. Recent developments in Indonesia’s current account ................................................................... 3 a. Putting the recent decline in Indonesia’s current account balance in context ......................................................5 b. Financing of Indonesia’s current account ............................................................................................................7 c. Export competitiveness ...................................................................................................................................... 11 C. WHAT IS BEHIND INDONESIA’S CURRENT ACCOUNT DEFICIT? .................. 18 1. Determinants of Indonesia’s current account............................................................................. 20 2. Contributions to Indonesia’s current account dynamics ............................................................ 25 D. ARE RECENT SHIFTS IN INDONESIA’S CURRENT ACCOUNT STRUCTURAL OR CYCLICAL? ............................................................................................................... 27 1. Determinants of investment in Indonesia ................................................................................... 29 a. Structural component of investment ................................................................................................................... 30 b. Cyclical component of investment ...................................................................................................................... 30 2. Determinants of saving in Indonesia ...........................................................................................31 a. Structural component of saving .......................................................................................................................... 32 b. Cyclical component of saving ............................................................................................................................. 32 3. Structural and cyclical components of the current account balance .......................................... 32 E. HOW FAR CAN THE CURRENT ACCOUNT AND THE NET INTERNATIONAL INVESTMENT POSITION (NIIP) DEVIATE FROM THEIR PATH? .................... 34 REFERENCES ....................................................................................................................... 41 APPENDIX I........................................................................................................................... 45 LIST OF FIGURES Figure 1: Framework of the current account imbalances toolkit...................................................... 1 Figure 2: Services imports are a function of goods exports .............................................................. 4 Figure 3: Equity income payments are linked to growth ................................................................. 4 Figure 4: Outflows from income payments on FDI equity are the flipside of inflows from FDI .... 4 Figure 5: The deterioration in Indonesia’s current account balance…............................................ 5 Figure 6: …started much later than in the other “fragile five” countries ......................................... 5 Figure 7: Indonesia’s trade deficit remains narrower than its “fragile five” peers… ....................... 6 Figure 8: …and its decline over time has not been simply a product of swings in commodity prices ................................................................................................................................................. 6 Figure 9: Investment outpaced savings in Indonesia around 2011 giving rise to a current account deficit ................................................................................................................................................ 6 Figure 10: The composition of the deficit differs across the ”fragile five” ....................................... 6 Figure 11: Current account balance, private and public saving and investment in 2003 and in 2013 ........................................................................................................................................................... 7 Figure 12: Indonesia’s current account financing ............................................................................ 9 Figure 13: The decline in Indonesia’s basic balance is a relatively recent phenomenon ................. 9 Figure 14: The rise in Indonesia’s gross external financing needs… ..............................................10 Figure 15: …has been driven by the recent current account deficit ................................................10 Figure 16: 2000-2013 Exports/GDP .................................................................................................12 Figure 17: 2000-2013 Imports/GDP .................................................................................................12 Figure 18: Merchandise Export Values............................................................................................13 Figure 19: Merchandise exports and imports since 2011 – monthly series ......................................14 Figure 20: Growth of merchandise exports and imports .................................................................14 Figure 21: The share of the top 5 export destinations decreased….................................................14 Figure 22: …While the overall number of export destinations increased slowly.............................14 Figure 23: The concentration of exports in the top 5 products increased…. ..................................15 Figure 24: …while the overall number of export products declined ...............................................15 Figure 25: Export growth has been concentrated in existing products in old markets ..................15 Figure 26: Indonesia’s export sophistication has been declining…................................................16 Figure 27: …while quality of exports has increased slowly .............................................................16 Figure 28: Indonesia’s export survival is amongst the lower in the group… ..................................17 Figure 29: …with some products performing better than others ....................................................17 Figure 30: Actual and predicted current account balances ............................................................ 20 Figure 31: Relationship between FDI and current account deficits ................................................21 Figure 32: Contributions to the current account deficit ................................................................. 26 Figure 33: Components of aggregate investment ........................................................................... 29 Figure 34: Components of aggregate saving ...................................................................................31 Figure 35: Components of current account balance ....................................................................... 33 Figure 36: Sensitivity of NIIP simulation to different rates of return relative to baseline.............. 35 Figure 37: Simulation crude oil price.............................................................................................. 36 Figure 38: Simulation: non-oil commodity prices .......................................................................... 36 Figure 39: Simulation ViX/VxO ..................................................................................................... 37 Figure 40: Simulation terms of trade .............................................................................................. 37 Figure 41: Simulated current account relative to baseline .............................................................. 38 Figure 42: Simulated NIIP position relative to baseline ................................................................ 38 Figure 43: Links between current account deficits, credit ratings and net foreign liabilities ........ 38 Figure 44: Net foreign assets as percentage of GDP for the fragile five (2012 and 2013) ............... 39 Figure 45: Equity and debt structure of Indonesia’s net foreign liabilities .................................... 39 Figure 46: Gross foreign liabilities as percentage of GDP for the fragile five (2013) ...................... 40 LIST OF TABLES Table 1: Prospective determinants of the current account balance in Indonesia............................18 Table 2: Estimated coefficients of current acount balance models................................................ 24 Table 3: Determinants of structural and cyclical investment and saving....................................... 28 Table 4: Determinants of aggregate investment ............................................................................ 29 Table 5: Determinants of aggregate saving .....................................................................................31 Table 6: Projections and sources for relevant variables .................................................................. 35 Table 7: Simulation scenarios ......................................................................................................... 36 Table 8: Dependent Variable is the Trade Balance ........................................................................ 48 LIST OF BOXES Box 1: What does the current account balance show? ...................................................................... 2 Box 2: Accounting relationships in Indonesia’s balance of payments ............................................. 4 Box 3: Current account deficits, reversals and sudden stops: how costly? ......................................10 Box 4: Conceptual framework for the analysis of determinants of the current account balance ....18 Box 5: Methodology for the estimation of dynamics and determinants of saving and investment 27 Box 6: Determinants of structural and cyclical investment and saving ......................................... 28 Box 7: Simulating Indonesia’s current account balance and its net international investment position ........................................................................................................................................... 34 Box 8: Is the level and the composition of Indonesia’s net foreign liabilities problematic? .......... 38 Indonesia Current Account Assessment Executive summary 1. Background Rapid GDP growth, Indonesia’s current account fell into deficit in late 2011 for the first time since the weak external 1997/8 Asian Financial Crisis. The current account balance subsequently dropped to demand and -2.78 percent of GDP in 2012, to -3.27 percent of GDP in 2013, and -2.97 percent declining commodity of GDP for the first three quarters of 2014, with most of the deterioration happening prices pushed the in the balance in goods and services trade, which averaged 3.1 percent of GDP in the current account into period 2000-2007, and fell to 1.8 percent during 2008-2013. The goods and services deficit in late 2011 trade balance dropped to -0.2 percent of GDP in 2012, further to -0.7 in 2013, and recovered mildly to -0.4 percent of GDP during the first three quarters of 2014. These changes over the 2011-2014 period were associated with rapid GDP growth, weak external demand and declining commodity prices. Current account Prior to the 1998 crisis, Indonesia experienced three decades of robust economic deficits are not a growth and reduction in trade barriers, leading to an increase in integration in the novelty in Indonesia, global economy. The rapid convergence required significantly higher levels of and not unusual for investment. Indeed, starting at 28.5 percent of GDP in 1980, investment increased to emerging and 37.2 percent of GDP by 1997. Constrained by low domestic saving, the increase in developing countries investment required and attracted foreign financing, which contributed to current account deficits averaging 3 percent of GDP annually. 2 The foreign financing instruments consisted mainly of short-term loans and portfolio inflows. As the Asian financial crisis unfolded in 1998, external financing suddenly dried up and the current account reverted to a surplus, alongside a collapse in the value of the rupiah. Since then, policy reforms aimed at reducing external debt and building domestic buffers against sudden stops in financial flows. During 1998-2011, Indonesia recorded current account surpluses, averaging 2.44 percent of GDP. These current account surpluses contributed to a substantial reduction in the stock of net foreign liabilities which dropped from 70.2 percent of GDP in 2001, to 38 percent of GDP in 2011. The recent shift into The recent shift into deficit has been based on two structural developments and a deficit reflects both combination of adverse exogenous and policy shocks. First, since the early 2000s structural Indonesia’s economy has grown faster than its major trading partners. Second, while developments and investment growth has been generally strong, gross national saving have stagnated policy factors around 32 percent of GDP since 2009, dropping to 30 percent in 2013. In 2011, Indonesia was hit by a major terms of trade shock as the end of a decade-long commodity “super-cycle” hit exports prices, leading to a sizable drop in the value of Indonesia’s commodity exports. The shock was compounded, until mid-2014, by sustained high oil prices relative to non-oil commodities and strong domestic demand for heavily subsidized fuel, as the country recorded a growing oil and gas trade deficit, having become a net oil importer in 2004. Finally, the implementation of the partial ban on raw mineral exports in January 2014 reduced revenues from this source. Structurally, this is At a high level, the reduction in the current account balance in the last 10 years is due the result of very to a combination of very strong private investment and public sector dissaving. 2 World Bank, July 2012, “Indonesia Economic Quarterly: Rising to Present and Future Challenges”. THE WORLD BANK | BANK DUN IA I Indonesia Current Account Assessment strong private Between 2003 and 2013, the current account balance decreased by 6.7 percentage investment and points of GDP, shifting from a surplus of 3.5 percent of GDP in 2003, to a deficit of public sector 3.3 percent of GDP in 2013. Over the period, public saving dropped from 4.3 percent dissaving of GDP to 2.6 against a virtually unchanged public investment share, while private saving increased from 24.7 to 27.8 percent of GDP, well short of the strong increase in private investment from 21.9 to 30 percent of GDP. Financing turned to As the current account turned negative, financing has come to rely increasingly on portfolio and other portfolio and “other” investments (as opposed to foreign direct investment), investments, exposing Indonesia to short-term financing constraints. Foreign direct investment to exposing Indonesia Indonesia surged over 2010-2013, yet remained under 2 percent of GDP, and was to short-term hence insufficient on its own to finance the current account deficit. Global appetite constraints for emerging market economy (EME) assets has helped maintain strong portfolio inflows into Indonesia, particularly into its higher-yielding government bonds, which since early 2012 have provided the most stable source of portfolio investment. The current account Although total external debt-to-GDP remains moderate in Indonesia, at 9.3 percent however is only a of GDP as of end 2013 and 11.6 percent of GDP as of the Q1-Q3 2014 period, the portion of overall level of the external debt stock and amortization requirements has grown strongly, external exposures particularly for the private sector, from 2011 onwards. Of the estimated total gross and… external financing needs in 2013 of USD 83 billion, the current account deficit accounted for only 35 percent, while private and public sector debt amortizations accounted for 50 percent and 15 percent respectively. Short-term external funding pressures could therefore emerge not only from a reduction in the availability of external financing relative to the current account deficit, but also external debt amortizations. …investing While Indonesia’s current account deficit is not necessarily an inherent concern, it efficiently, does mean that the country is building up liabilities to the rest of the world. maintaining reliable Understanding the drivers of the trend in the current account balance, and what these financing and may reveal about the economy, is crucial for constructive policy responses. In building buffers Indonesia’s case, the evidence suggests that a current account deficit is “normal” in the sense that investment needs exceed domestic savings. As long as investments generate sufficient growth, meeting the resulting financing costs should not be a concern. This points to the need to minimize unnecessary dis-saving (for example, by reducing wasteful public spending on fuel subsidies, as intended by recent reforms), raise the efficiency of investments, while managing short-term financing risks by maintaining reliable external financing, and building policy buffers against volatility in international markets. 2. Determinants of the current account in Indonesia Current account The analysis presented in this report suggests that Indonesia’s recent current account determinants include deficit results from the interaction of short, medium and long run factors that can be external shocks, grouped into four blocks: external shocks, domestic policies, international integration, domestic policies, and stage of development and demographics. Each of these blocks is discussed below. integration, and Based on the analysis in the report, a structural current account balance between -3 development stage and 1 percent of GDP is to be expected. THE WORLD BANK | BANK DUN IA II Indonesia Current Account Assessment a. External shocks Indonesia’s faster Indonesia’s robust growth relative to its main trading partners has contributed to a growth than that of worsening in the current account balance by almost three quarters of a percentage its trading partners… point of GDP, over the 2009-2014 period. Indonesia’s GDP growth has on average exceeded that of its main trading partners by 2.25 percentage points per year. With domestic growth faster than main trading partners” growth, import demand outstripped export demand, leading to a gradual worsening of the trade balance particularly since the global financial crisis in 2008/9. Indonesia’s export structure is highly exposed to high income economies, most of which have struggled to emerge decisively from the global financial crisis. …compounded by The fall in the prices of Indonesia’s export commodities led to a decline in the current the fall in the prices account balance of one quarter of a percentage point of GDP since 2012. The price of Indonesia’s export index of Indonesia’s 6 main commodity exports, accounting for approximately 50 commodities… percent of export revenues, dropped by 28 percent from early 2011 to mid-2014. …and the sizable Since the year 2000, the Rupiah appreciated by almost 32 percent in real terms, with appreciation of the an appreciation of ten-percent over the previous three years tending to reduce the Rupiah, all put current account balance by almost half of a percentage point of GDP. Real exchange external pressure on rate (REER) appreciations encourage imports and inhibit export competitiveness, the current account weighting negatively on the trade balance. In this context, the depreciation of the balance Rupiah since August 2013 has acted as a shock absorber for the economy and helped to sustain the current account adjustment to date. b. Domestic policies Domestically, the The fuel subsidy contributed to a widening oil and gas deficit, reducing the current fuel subsidy spurred account balance by 0.81 percent of GDP for every percentage point of GDP increase domestic demand, in the subsidy bill. The fuel subsidy scheme contributed to a widening oil and gas contributing to a deficit at a time of high oil and gas prices by artificially suppressing the domestic retail widening oil and gas price of fuel, while also contributing to public sector dissaving and hence to deficit… Indonesia’s overall domestic saving deficit. The reduction in budgeted fuel subsidy spending by 2 percent of GDP following reforms enacted in January 2015 mark a major change in government policy, and are expected to help rebalance the current account. …while credit Credit expanded by 2.38 percent per annum since 2010 contributing to a decline in expansion fuelled the current account balance by about one-tenth of a percentage point every year. By domestic demand the same token, recent tightness in the domestic credit market is expected to alleviate and imports pressure on the current account balance by putting a brake on domestic demand. c. Integration in the global marketplace Indonesia’s Over the last two decades Indonesia has increased its degree of integration into the integration into the global marketplace. Indonesia’s export and import orientations today are well above global marketplace the world’s average, greater than China and India and substantially above those of has increased the Brazil, Turkey and South Africa. 3 Increased trade openness, measured by the ratio of imports and exports to GDP, is systematically associated with increases in the current 3 Controlling for the size of the internal market, the level of development, and physical characteristics (such as being landlocked, or islands). THE WORLD BANK | BANK DUN IA III Indonesia Current Account Assessment current account account balance. This implies that further trade integration in the global marketplace balance… relative to its trading partners should have a beneficial effect on the current account. …and strong FDI Every extra 1 percent of GDP in FDI inflows is associated with a 0.17 percent of flows attracted self- GDP reduction in the current account balance in the short run. This is the case as financing imports FDI inflows increase demand for imports of intermediates and capital goods and that helped the profit repatriations by more than they increase exports or induce import substitution. 4 economy grow Such imports are self-financing in the sense that they are the result of investment inflows. Over the long-term, international evidence suggests that increased FDI spurs economic growth. d. Stage of development and demographics In the economic In its process of economic catching up, Indonesia has experienced remarkable growth catch-up, investment in investment, contributing to real GDP growth averaging 5.4 percent per annum grew faster than since 2000. As domestic saving fell short of financing this investment relied, and is saving… likely to continue to rely, on foreign financing. As such, it is to be expected for Indonesia to run a moderate current account deficit in the medium term. …saving have been The population dependency ratio, by reducing national saving, has also been an low due to important structural contributor to the current account deficit. Going forward, stands Indonesia’s to benefit from a “demographic dividend”, as the current dependent population (changing) reaches working age, leading to higher domestic saving, and hence to an increased demographic current account balance. However, fertility rates in Indonesia do remain high, which structure is likely to constrain the future reduction in the contribution of demographic factors to the current account deficit. 3. Policy discussion Indonesia’s current The deterioration of the current account balance since 2011 has put sustainability in account balance is the spotlight for policymakers. This report finds Indonesia’s structural current consistent with the account balance to be in the range of -3 to 1 percent of GDP. Consequently, while structure of its Indonesia’s current account deficit must be monitored closely, at current levels it is economy… not symptomatic of economic overheating and does not warrant short-term policy actions. Rather, policymakers should continue to focus generally on sound macroeconomic management, and on tackling the long-run determinants of the deficit, such as enhancing public investment efficiency, private sector competitiveness, and deepening the financial sector so as to mobilize more domestic savings and reduce financing risks. Simulations show Simulations of the path of the current account and net international investment that an infrastructure position (NIIP) in the face of benign or adverse external conditions show relatively push should be small deviations from the baseline projections. In the event that the government sustainable from an embarked on a major infrastructure development program, simulations show that the external balance current account balance would decrease initially under the pressure of increased perspective… imports and domestic demand but set itself on an increasing path thereafter, on the back of increased trade integration and export competitiveness. Higher expected returns to the domestic economy would also attract complementary foreign funding to finance the temporary decrease in the current account balance. 4 A similar result is obtained when examining the link between lagged FDI and the trade balance, suggesting that the effect of FDI on net exports is negative even before adding profit repatriation. THE WORLD BANK | BANK DUN IA IV Indonesia Current Account Assessment Macroeconomic Exchange rate flexibility has proved to be an effective shock absorber and is helping management will be the adjustment of the trade balance to weaker commodity export prices. This has been crucial to reduce aided by Bank Indonesia’s more “hands off” approach to managing the currency since vulnerability to mid-2013, and its measures to support liquidity in the domestic foreign exchange external shocks… market and to attract portfolio inflows. …as more efficient Although the overall fiscal stance appears prudent at a macro level, the composition government and quality of spending also matter. Public investment levels remain low by spending can help international standards, and adds rigidities to the process of adjustment following spur growth shocks. Increasing public investment is a high priority for the government, which has doubled the allocation of capital spending in the revised 2015 budget. Execution of this increased infrastructure budget might cause a deterioration in the current account balance and NIIP in the short run (as shown in the simulations in the last section of this paper), but improved infrastructure can also attract investment flows and provide a boost to international competitiveness. Integration in global Integration in global markets is key for competitiveness, productivity, and helps markets can generate manage external balances. Indonesia is highly integrated in global markets. A exports, jobs and supporting trade regime helps domestic firms grow in the region and beyond, and investment… ultimately provide more and better quality jobs to Indonesians. Supporting integration will require continuous effort in reducing barriers to trade while improving competitiveness of Indonesian firms will entail addressing issues of infrastructure, cost of logistics, and availability of services inputs such as access to finance, and a skilled labor force. 5 …although targeted Indonesia’s predominantly low-skilled labor force can benefit from increased trade policies, will be openness, but targeted policies will be needed for Indonesian labor to benefit fully. needed for Active policies to stimulate innovation and human capital accumulation help ensure Indonesian labor to that greater integration translates into technology diffusion and learning; increased benefit fully… efficiency, diversification and sophistication of the production structure, which will reduce the vulnerability to commodity price shocks and the volatile nature of trade flows. …and to increase FDI is the external funding source least prone to sudden stops or outflows, and FDI, the least generates important positive spillovers through technology transfers to domestic outflow-prone source firms. FDI is associated with increased imports, and the possibility of a temporary of finance… deterioration in the trade balance (as found in this paper), as it also tends to crowd in further domestic investment, increasing aggregate demand. Increased import levels, however, tend to be self-financing and contribute to economic growth in a virtuous cycle in which high FDI generates import demand and economic growth, which in turns attracts more FDI and lowers dependency on more volatile and potentially more expensive portfolio inflows. …which might come Although resilient, FDI has been underpinned by three factors that may come under under short-term pressure. First, the large resource base, whose attractiveness has suffered in recent pressure future due years due to moderation in commodity prices. Second, the large and growing domestic to a number of market, though constrained by needed policy responses since 2013 to support factors… adjustment to the commodity price shock and mitigate external financing risks. Third, 5 See Duggan, Rahardja and Varela, 2013 and Cattaneo, Gereffi, Miroudot and Taglioni, 2013 for evidence of how reliable energy and services inputs help increase firms’ sophistication and competitiveness in global value chains. THE WORLD BANK | BANK DUN IA V Indonesia Current Account Assessment Indonesia’s potential as a regional production hub in Asia, which has been undercut by regulatory uncertainties and skills and infrastructure gaps. To boost Indonesia’s attractiveness as an FDI recipient, improvement in the investment climate is needed. 6 …including policy The introduction of a partial ban on exports of unprocessed minerals added to uncertainty, as pressures on the current account in the short term, while bringing uncertain long run exemplified by the benefits, and at a potentially high cost for the wider business environment. With the introduction of a intention of increasing the value of commodity exports, the government of Indonesia partial ban on raw imposed since the beginning of 2014 imposed a ban on the export of certain raw minerals minerals. 7 In addition to causing a loss in export revenues and weighing on an already weakening current account, the lengthy and unclear process leading to the introduction of the revised regulation increased uncertainty in the mining sector, weakening Indonesia’s mining and related industry investment climate. From a long run perspective, this could prove to be a larger burden for the prospects of moving up the value chain in natural resource-intensive production. Finally, financial In the medium term, reducing pressure from the current account deficit can be deepening would achieved also via increasing the portion of domestic financing for investment. reduce dependency Increasing the saving rate in Indonesia will require measures to deepen the financial on foreign funding markets, improve access to finance and improve confidence in the domestic financial and banking sector. 6 Indonesia currently ranks 120 out of 189 economies in the World Bank Group’s Doing Business 2014 rankings, below the regional average and peer countries’ performance such as China, Thailand and Malaysia. 7 For a discussion of the near-term impact of the ban on unprocessed minerals and its expected outcome, see World Bank, 2013, “Indonesia Economic Quarterly: Slower Growth, High Risks”. THE WORLD BANK | BANK DUN IA VI Indonesia Current Account Assessment A. Structure of the report This report is structured as follows. First, it provides an overall description of trends and developments in Indonesia’s current account over the last twenty five years. Second, it presents an econometric analysis of the determinants of the current account, identifying the contributions of each. In doing so, this report discusses the scope for policy action. Third, taking a more structural approach, it analyses the underlying drivers of the current account balance: saving and investment. It separates their structural and cyclical components and identifies the determinants of each. Finally, it simulates the possible path of the current account and NIIP under a number of external developments and policy scenarios. Figure 1: Framework of the current account imbalances toolkit Source: Authors’ elaboration based on Cusolito, A. & M. Nedeljkovic (2013) THE WORLD BANK | BANK DUN IA 1 Indonesia Current Account Assessment B. CONTEXT Box 1: What does the current account balance show? From a balance of payments perspective, the current account comprises three subcomponents: the trade balance, the income balance and the transfer balance. 8 The trade balance records all transactions with the rest of the world related to the exchange of goods and services. The income balance records net interest and dividend payments, while the last component, the transfer balance reflects net payments (that do not correspond to purchases of any good, service, or asset) received or sent to the rest of the world, among which, the most notable corresponds to remittances received from nationals working abroad, and sent by foreigners working domestically. Current Account = Trade Balance + Income Balance + Current Transfers Balance From a national accounts perspective, the current account results from the difference between what an economy produces and what it consumes and invests in a given period (or alternatively, the difference between saving and investment). When countries consume and invest more than what they produce, they need to borrow from the rest of the world to finance that gap. The current account deficit reflects that amount that a country borrows from the rest of the world to finance the investment and consumption in excess of its level of production. Current Account Balance= [Output (GDP) – Consumption] – Investment Or Saving – Investment If Saving > Investment  The economy is a net lender of the rest of the world If Saving < Investment  The economy is a net borrower of the rest of the world The saving and investments can be decomposed further into the portions that are attributable to the public and the private sectors. Thus, the current account balance can be expressed as the sum of the private sector’s surplus (private saving minus private investment, Sp-Ip) and the government’s surplus (tax revenues minus government expenditures, T-G). This is why current account deficits and fiscal deficits are often referred as the “twin deficits”. Given the private sector balance (Sp-Ip), increases in fiscal deficits co-move with increases in current account deficits. Current Account Balance = (Savingp - Investmentp) + (Tax – Government Expenditures) From the relationship shown above, it is possible to see how a positive or a negative current account balance is not a bad thing per se, and it depends on the country’s specific circumstances. For example, if a relatively poor country in its path to growth, and constrained by low levels of domestic saving, finances its high investment rates through foreign capital, it might run into a current account deficit. This deficit may be inevitable for economic development, that requires firms building up capital stocks, and often the public sector investing in, for example, infrastructure upgrading. 9 Conversely, current account deficits can also arise from booming private or public consumption. When governments spend more than what they collect in taxes, or the private sector consumes more than it produces, countries can run into current account deficits. These deficits, however, are unlikely associated to faster future growth, and will generate concerns about their sustainability. These simple examples stress the importance of identifying the underlying sources of the current account deficit: is it mainly driven by fiscal deficits, by low private saving rates, or by high private investment rates? Source: Authors’ elaboration. 8 Under the new IMF guidelines (Balance of Payments Manual 6, BPM6), to which Indonesia adhered in 2014, the current account comprises the following components: the “trade balance”, the “primary income balance” (equivalent to the income balance under BPM5), and the “secondary balance” (equivalent to the transfer balance under BPM5). Given that the analysis in this report focuses on the period in which Indonesia followed the BPM5, this structure is used to describe trends and interpret results. For more on the specific changes that the conversion from BPM5 to BPM6 entails, see IMF, 2013, “Sixth Edition of the IMF’s Balance of Payments and International Investment Position (BPM6)”. 9 Recent IMF analysis showed effects of public infrastructure investments on growth. In the context of industrial economies, the IMF finds that when the investment is financed by borrowing, an increase in public investment equivalent to 1 percent of GDP boosts GDP by 0.9 points in the first year and 2.9 percent in the fourth. The effects are bigger in slow-growing economies when rates are low and competition for loans is subdued. Under those conditions, the boost in public investment leads to a 1.5 percent increase in GDP in the first year and 3 percent in the fourth (Chapter 3 in IMF, 2014) “Is it time for an infrastructure push? The macroeconomic effects of public investment”). THE WORLD BANK | BANK DUN IA 2 Indonesia Current Account Assessment 1. Recent developments in Indonesia’s current account Indonesia’s current Indonesia’s current account balance turned negative towards the end of 2011, and account balance remains negative to date, initiating the longest streak of negative current account turned negative records since 1998. Current account deficits are not a novelty in Indonesia, as the towards the end of country maintained a negative current account balance, averaging -2.7 percent of 2011… GDP between 1986 and end-1997. This reversed in the aftermath of the 1998 crisis up until 2011, when Indonesia recorded positive, though generally modest, current account balances averaging 1.7 percent of GDP from 2005 until the global financial crisis, and 1.3 percent until the end-2011 swing into deficit. …due to two First, since the early 2000s Indonesia’s economy has tended to grow faster than its structural major trading partners. Domestic demand outstripped external demand and import developments… growth has been faster than export growth, particularly since the global financial crisis in 2008/9. Second, while investment growth has been generally strong, gross national saving have stagnated around 32 percent of GDP since 2009, dropping to 30 percent in 2013. These dynamics have led to a rise, in dollar terms, of Indonesia’s net foreign liabilities, the returns to which have pushed the income account deeper into deficit, from about USD -3 billion in the 2005-2009 period to USD -6 billion thereafter, adding a structural drag on the current account balance. …and a combination Indonesia’s once ample goods trade surplus suffered from a number of shocks since of adverse 2011. First, in 2011, a major terms of trade shock following what has been described exogenous and as the end of a decade-long commodities “super-cycle” hit exports. The price index policy shocks… of Indonesia’s 6 main commodity exports, accounting for just below 50 percent of export revenues, dropped by 28 percent from February 2011 to July 2014, and has subsequently continued falling. Second, Indonesia’s trade balance has been affected by a growing oil and gas trade deficit since becoming a net oil importer in 2004, when domestic demand outstripped declining exports. Oil exports have been on a long-declining trend, due to lack of investment and new discoveries in the sector. Until the drop in crude oil prices over the second half of 2014, this was aggravated by sustained high oil prices relative to non-oil commodities, and strong domestic demand for heavily subsidized fuel. Third, the implementation of the mineral export ban in January 2014 caused most of the affected raw mineral exports to grind to a halt. Although some of the blockages have been removed since August 2014, a number of mineral products continue to be affected, and the longer-term impacts of the policy on the commodity sector remain uncertain. THE WORLD BANK | BANK DUN IA 3 Indonesia Current Account Assessment Box 2: Accounting relationships in Indonesia’s balance of payments The services and income components of Indonesia’s Figure 2: Services imports are a function of goods current account are fairly stable and structurally negative. exports (year-on-year growth, percent) The services balance records all service related transactions 50 Goods exports between residents and non-residents. Service related transactions can be imports or exports of freight financial 40 or telecom services as well as tourism flows. In the last 3 30 years, services imports have outstripped exports by about USD 2.7 billion per quarter, a figure almost entirely 20 explained by freight imports, which averaged USD 2.3 10 billion over the same period. Although Indonesia is a large exporter – chiefly of bulky commodities – most freighting 0 services are handled by foreign operators, which results in -10 freighting transactions between residents and non-residents -20 for every export shipment. The correlation is clearly visible in Figure 2. -30 Services imports -40 The income balance, similarly, records all income related Mar-06 Mar-08 Mar-10 Mar-12 Mar-14 transactions – that is payments of interest, coupons or dividends between residents and non-residents. Outbound Source: CEIC; World Bank staff calculations income payments in the last 3 years outstripped inbound payments by about USD 6.7 billion – causing, together with the service deficit, a structural drag on the current account of about USD 9.5 billion per quarter. Approximately USD 4.3 billion of the income payments are equity-related payments of dividends and USD 1.2 billion debt-related payments of interest and coupons. On the equity side, the large stock of foreign investor equity holdings gives rise to dividends to non-residents when corporate profits are high, for example, due to high exports or strong domestic income growth (Figure 3). In addition, the balance of payments treatment of retained earnings pertaining to non-residents is to consider them as paid to non-residents (causing a current account outflow) and invested by non-residents as FDI (causing a financial account inflow), creating an accounting relationship between outflows due to payments on equity and FDI Inflows (Figure 4). Figure 3: Equity income payments are linked to Figure 4: Outflows from income payments on FDI growth equity are the flipside of inflows from FDI (year-on-year growth in nominal USD terms, percent) (billion, USD) 100 5.0 Income payments on equity Income payments on 80 4.5 equity 4qma 4.0 60 3.5 40 Nominal GDP yoy 3.0 20 2.5 0 2.0 1.5 -20 1.0 Inbound equity FDI -40 Goods exports 0.5 -60 0.0 Mar-06 Mar-08 Mar-10 Mar-12 Mar-14 Mar-06 Mar-08 Mar-10 Mar-12 Mar-14 Source: CEIC; World Bank staff calculations Source: CEIC; World Bank staff calculations THE WORLD BANK | BANK DUN IA 4 Indonesia Current Account Assessment Figure 5: The deterioration in Indonesia’s current Figure 6: …started much later than in the other account balance… “fragile five” countries (balances, USD bn) (balances, USD bn) Trade Balance Income Balance Transfer Balance 4 12 Indonesia 2 10 8 0 Brazil 6 -2 India 4 -4 2 -6 0 -8 -2 South Africa -4 -10 Turkey -6 -12 2000 2002 2004 2006 2008 2010 2012 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: WEO Source: WEO a. Putting the recent decline in Indonesia’s current account balance in context Comparing Indonesia, Brazil, India, South Africa and Turkey were characterized in 2013 as the Indonesia with the “fragile five” 10 due to their sizable current account deficits, the financing of which rest of the “fragile depended on volatile capital flows. Comparing Indonesia to its “fragile” peers is five”… informative and helps put magnitudes into perspective. …Reveals However, there are sizable differences between the “fragile five”, and Indonesia’s Indonesia’s position position may not be as fragile as suggested, although it deserves careful attention. may not be as Indonesia’s current account balance deteriorated rapidly since 2011, but in 2013, the “fragile” as deficit was still lower relative to GDP than in Brazil, South Africa and Turkey. Indeed, suggested, although Indonesia only started registering current account deficits in 2012, while Turkey, it deserves careful South Africa, India and Brazil have been accumulating current account deficits since attention as early as 2002 (Figure 6). 11 Indonesia’s net stock of foreign liabilities, however, is the second largest of this group of countries, reaching 42.4 percent of GDP in 2012, after Turkey’s that reached 53.4 percent of GDP in 2013. 12 Trade balances had Trade balances have deteriorated across the “fragile five” since 2000 despite strong been falling since commodity prices through 2011, with weak commodities pushing them further down before the subsequently (excluding India in 2013, Figure 7). As Figure 8, shows Indonesia’s trade commodities price balance in goods declined steadily over a decade that showed volatile yet raising falls since 2011 commodity prices, pointing to the presence other (non-commodity price) factors affecting the trade balance. 10 The term was coined by market analysts at Morgan Stanley in the summer of 2013. 11 Current account balances turned negative in 2002 in Turkey, 2003 in South Africa, 2004 in India and 2007 in Brazil. 12 In Brazil, net foreign liabilities reached 36.7 percent of GDP in 2013, in India 17.5 percent and for South Africa, they reached 4.1 percent in the same year (source: IMF). THE WORLD BANK | BANK DUN IA 5 Indonesia Current Account Assessment Figure 7: Indonesia’s trade deficit remains narrower Figure 8: …and its decline over time has not been than its “fragile five” peers… simply a product of swings in commodity prices (balances, percent of GDP, USD bn) (lhs: balance, percent of GDP; rhs percentage change, percentage points) 10 20 90% 8 Trade balance in goods Growth in commodity prices 6 15 70% Trade balance in goods Indonesia 4 Trade balance 50% 2 10 Brazil 30% 0 -2 South 5 10% -4 India Africa -10% -6 0 -8 -30% Turkey Growth commodity prices -10 -5 -50% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: WEO Source: WEO and World Bank Through the saving- As the current account balance reflects the difference between saving and investment investment lens, in an economy (see Box 1), it is important to understand whether the current account Indonesia’s current deficit in Indonesia is associated with high investment, low saving, or a combination account deficit of both. Although in 2013 the investment rate fell by more than one percentage point results from a (from 34.7 percent of GDP in 2012 to 33.6 percent of GDP in 2013), the decline of sustained increase in saving rates was greater than that (from 32 percent of GDP to 30.4 percent of GDP investment, and a over the same period, Figure 9). Looking across the so-called “fragile five”, it is clear decline in saving... that very different saving-investment combinations have given raise to similar current account deficits (Figure 10); Indonesia and India show the highest rates both of investment and saving (above 30 percent), while Brazil, Turkey’s, and South Africa’s are in the 15-20 percent of GDP range. Figure 9: Investment outpaced savings in Indonesia Figure 10: The composition of the deficit differs across around 2011 giving rise to a current account deficit the ”fragile five” (percent of GDP, percentage points) (percent of GDP, percentage points) 36 Investment (-) Saving (+) Current account balance Investment, Saving 40 34 30 32 Gross national saving 20 (percent of GDP) 30 10 28 0 26 -10 24 -20 Gross capital formation 22 (percent of GDP) -30 20 -40 2000 2002 2004 2006 2008 2010 2012 Brazil India Indonesia South Turkey Africa Source: WEO Source: World Bank staff calculations based on WEO ..in particular, by Data from the IMF World Economic Outlook decomposes consumption, investment strong private and saving into private and public components, providing an insightful breakdown of investment and THE WORLD BANK | BANK DUN IA 6 Indonesia Current Account Assessment public sector the current account. 13 Comparing the year 2003, in which the current account balance dissaving… reached 3.5 percent of GDP, and 2013, with a current account balance of -3.3 percent of GDP is revealing. The current account deteriorated by 6.7 percentage points over the period, with public saving dropping from 4.3 percent of GDP to 2.6 against virtually unchanged public investment, and private savings increased from 24.7 to 27.8 percent to GDP, although failing to offset an increase in private investment from 21.9 to 30 percent of GDP (Figure 11). …while the In Turkey, the increase in the current account deficit is mainly explained by an underlying drivers of increase in the private sector deficit that was only partially offset by a contraction in current account the public sector deficit. Indeed, the gap between private saving and private deficits in the investment shifted from a surplus of 7.9 percent of GDP in 2003 to a deficit of 7.3 remaining “fragile percent of GDP in 2013. Over the same period, the public sector gap shrank five” are quite substantially from a 10.4 percent deficit to a 0.57 percent deficit. In Brazil, the private diverse sector surplus contracted from 5.7 percent of GDP in 2003 to 0.8 percent of GDP in 2013, while the public sector continued to run a deficit (4.9 percent in 2003 and 4.4 percent in 2013). In India, the private sector’s investment expansion contributed to increase the deficit despite the contraction in the public sector deficit. The 7.5 percentage point increase in private investment rates observed in the period 2003- 2013 was not matched by increases in private saving, leading to a contraction in the private sector surplus from 7.7 in 2003 to 1.9 in 2013. During the same period, the public sector reduced its deficit from 5.4 percent of GDP in 2003 to 3.9 percent of GDP. In South Africa, both private sector and the public sector deficit expansions contributed to increase the current account deficit (Figure 11). Figure 11: Current account balance, private and public saving and investment in 2003 and in 2013 (percent of GDP, percentage points) 30 Saving (+) Investment (-) Balance 20 10 0 -10 -20 -30 2003 2013 2003 2013 2003 2013 2003 2013 2003 2013 2003 2013 2003 2013 2003 2013 2003 2013 2003 2013 Private Public Private Public Private Public Private Public Private Public TUR ZAF BRA IND IDN Source: World Bank staff calculations based on WEO b. Financing of Indonesia’s current account When a country runs When a country runs a current account deficit or, equivalently, when it invests more a current account than it saves, it must borrow from the rest of the world (or use its own foreign deficit, it borrows exchange reserves) to finance that deficit. Conversely, when it runs a current account 13 To break down consumption, investment, and saving into public and private components, the WEO data combine data from National Accounts and Fiscal Accounts. THE WORLD BANK | BANK DUN IA 7 Indonesia Current Account Assessment from the rest of the surplus, it will either lend its excess saving to the rest of the world, potentially in part world… by accumulating foreign exchange reserves. Indonesia’s recent current account deficits make it a net borrower in international markets. …and the In the two decades following 1980, “Other” investments – largely constituting composition of currency, deposits and bank loan inflows, provided the bulk of current account Indonesia’s current financing. These turned around sharply around in the aftermath of the Asian Financial account financing Crisis, with large outflows of both on the assets and liabilities side and remained a has changed over the small but very volatile component of financing since. Foreign direct investment years provided a steadily increasing share of financing up to the Asia Financial Crisis, collapsing with the crisis and recovering in the subsequent five years. During the mid- 2013 mini-crisis, described in more detail below, the basic balance (the sum of the current account and FDI) turned negative, indicating that the financing of the current account depends on portfolio investment. Portfolio investment only started playing a role in financing Indonesia’s current account deficit in the last decade, prominently through increased foreign purchases of domestic debt securities. The mid-2013 Indonesia’s ability to meet its external financing needs was challenged in May-August volatility is an 2013. During this period global risk appetite towards EMEs fell as investors priced in example of how US monetary policy normalization (“tapering” of the Fed’s quantitative easing external conditions program beginning in late 2013 and anticipation of continued normalization of can cause concerns accommodative monetary policy). This coincided with the widening of the quarterly around current current account deficit to a record USD 10.1 billion (4.5 percent of GDP). During account financing this episode, Indonesian equities fell by 24 percent from peak (20 May) to trough (27 August), Rupiah-denominated government bond yields rose by almost 400bp to touch levels last seen in early 2011, and the Rupiah depreciated by 26 percent against the US Dollar. Indonesia’s position Bank Indonesia’s response in tightening credit conditions contributed to a domestic improved since the demand slowdown while facilitating, together with a weaker Rupiah, the adjustment 2013 mini-crisis, but of the trade balance. Higher domestic interest rates increased the attractiveness of the adjustment is portfolio assets, encouraged the return of capital flows and helped rebuild reserve still incomplete… assets. To date, reserves have been rebuilt to more comfortable levels, but the current account adjustment remains incomplete – mostly driven by import compression, while non-commodity exports have failed to compensate for the ongoing weakness of commodity exports. …the current Current account financing should be considered in the context of overall gross account however is external financing needs, that is current account financing and total short term debt only a portion of liabilities. Total external debt-to-GDP, the sum of the current account balance and overall external external debt amortizations, which is a measure of external exposure, remains exposures… moderate in Indonesia, at 11.6 percent of GDP as of the Q1-Q3 2014 period, but the level of the external debt stock and associated amortization requirements has grown strongly, particularly for the private sector, for 2011 onwards. Overall, Indonesia’s gross external financing needs have risen significantly over recent years, to finance the current account deficit, and to meet external debt amortizations. Indeed, of the estimated total gross external financing needs in 2013 of USD 83 billion, the current account deficit accounted for only 35 percent, while private and public sector debt amortizations accounted for 50 percent and 15 percent respectively. Short-term external funding pressures could therefore emerge not only from a shift in the availability of external financing in general but also from a rise in financing needs as THE WORLD BANK | BANK DUN IA 8 Indonesia Current Account Assessment a result of not only of a widening of the current account deficit but also if there were to be a rise in external debt amortizations. …but the greatest Comparing the evolution of gross external financing needs across the “fragile five” source of change for it is evident that the most important development in Indonesia is the emergence of Indonesia relative to a current account deficit (Figure 15). Indonesia gross external exposures relative to the “fragile five” has GDP are well below the “fragile five” average, which are driven by the much larger been the swing into external exposures of Turkey and South Africa, but are still larger then Brazil’s and current account India’s. While the bulk of gross external financing needs still come from debt deficit… amortization, the big driver of the increase in gross external financing needs for Indonesia since 2010 has been the emergence of a current account deficit. ..stressing the The country’s current account deficit, while not at unsustainable levels, has added to importance of gross external financing needs, and stresses the importance of securing a reliable securing a reliable mix of external funding, including FDI. As less outflow-prone FDI has fallen short mix of external of covering the current account deficit, Indonesia has remained dependent on more funding, including volatile portfolio capital flows. Encouraging the sustained growth of typically less FDI volatile FDI inflows will require more steps to support investment in general, including by increasing productivity and enhancing the overall business environment. Figure 12: Indonesia’s current account financing Figure 13: The decline in Indonesia’s basic balance is (balances, USD bn) a relatively recent phenomenon (balances, USD bn) Capital account FDI 6 Portfolio Other investment Errors and omissions Current account deficit 4 Indonesia 4 3 2 Brazil 2 0 India 1 -2 0 -4 -1 -6 Turkey -2 -8 South Africa -3 -10 1980 - 1989 1990 - 1999 2000 - 2009 2010 - 2013 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source: IMF WEO Note: Basic Balance is the difference between the current account balance and Net Direct Investment Source: IMF WEO THE WORLD BANK | BANK DUN IA 9 Indonesia Current Account Assessment Figure 14: The rise in Indonesia’s gross external Figure 15: …has been driven by the recent current financing needs… account deficit (USD bn) (USD bn) Private external debt amortizations Public external debt amortizations Current account deficit Current account deficit Short term debt by remaining maturity 100 Total gross external financing requirements 2013 gross external financing needs avg 30 80 25 20 60 15 40 10 5 20 0 0 -5 2007 2010 2013 2007 2010 2013 2007 2010 2013 2007 2010 2013 2007 2010 2013 -20 2009 2010 2011 2012 2013 2014 to Turkey South Brazil India Indonesia Q3 Africa Source: CEIC; World Bank staff calculations Source: IMF WEO Box 3: Current account deficits, reversals and sudden stops: how costly? Large current account deficits do not intrinsically represent an economic imbalance, but can rather be consistent with a country stage of development. However, large current account deficits tend to increase a country’s vulnerability to sudden stops in foreign financing. When external financing dries up, these large deficits need to be reduced abruptly. This phenomenon, often referred to as “current account reversals”, can be disruptive and costly for an economy and has received a lot of attention in academic and policy circles following the balance of payments and currency crises of the 1990s and early 2000s in developing economies. Sudden stops of external financing happen in response to changes in economic fundamentals or in market sentiment. They typically induce a sharp depreciation of the real exchange rate, and a reduction in imports, leading to the reversal of the current account to a surplus position (or to a substantial reduction of the deficit). The probability of a reversal, in turn, has been associated with high current account deficits, high external debt and service costs, a rapid rate of growth of domestic credit and low levels of international reserves. 14 Indonesia’s experience in 1997/98 provides an example of how disruptive current account reversals induced by sudden stops can be. Indonesia grew robustly during the period 1990-1996, led by private sector investment, with fiscal accounts in surplus, and strong inflows of FDI, and more volatile short term capital flows that helped financing an apparently manageable current account deficit. Bank Indonesia set the value of the rupiah against a basket of currencies, and intervened in the market around that central rate, which depreciated gradually so as to stabilize the real exchange rate. A band was introduced in 1994 to allow some flexibility, and its width was progressively widened over time. This buoyant economy and ample foreign investment encouraged domestic firms to borrow directly offshore and through banks onshore in foreign currency, without hedging against foreign exchange risk, probably taking for granted that Bank Indonesia would defend the weaker edge of the band, at least long enough for them to hedge their foreign exchange exposure. 15 A number of shocks starting in the mid-1990s led to a progressive deterioration in market sentiment: the Mexican peso crisis, a slowdown in exports, more adverse terms of trade, and, finally, the collapse of the Thai baht. This was combined with a number of vulnerabilities that the Indonesian economy exhibited: growing overinvestment, financial fragility – rapid growth in risky lending, and the aforementioned currency mismatches in corporate and banks” balance sheets. 16 Sporadic speculative pressure occurred, and the collapse of the 14 Edwards, 2004b. 15 See page 87 in Kenward, 1999, “Assessing Vulnerability of Financial Crisis: Evidence from Indonesia”, Bulletin of Indonesian Economic Studies, vol. 35 No. 3. Pp.71-95. 16 See Moreno, 2008, “Experiences with Current Account Deficits in Southeast Asia” in Current Account and External Financing, eds: Cowan, K., S. Edwards, and R. Valdes. THE WORLD BANK | BANK DUN IA 10 Indonesia Current Account Assessment baht in July 1997 triggered sharp depreciations of the Indonesian rupiah – putting pressure on firms with dollar debt, and creating gloomy growth prospects. FDI and portfolio inflows dried up, and turned into outflows in the first quarter of 1998. The current account deficit of 2.2 percent of GDP in 1997 reversed dramatically to a 4.1 percent surplus. Real GDP declined by 13 percent in 1998. Investment fell faster than GDP, with the Investment/GDP ratio declining by 3 percentage points from 1997 to 1998, and falling a further 5 percentage points in 1999. CPI inflation escalated up to 58.5 percent in 1998, relative to a moderate 6.15 percent in 1997, with food prices soaring, affecting the poor disproportionately. The literature shows many episodes of these phenomena occurring in the developing world, apart from Indonesia. Calvo, Izquierdo, and Talvi (2006) looked at episodes of sudden stops and current account reversals in a number of countries. The authors found that output in Mexico declined by 4.8 percent between 1981 and 1983 and by 6.2 percent between 1994 and 1995 (both periods corresponding to sudden stops and current account reversals), while in South Korea output declined by 6.9 percent between 1997 and 1998, and in Thailand by 11.7 percent between 1996 and 1998. During these episodes, asset prices, such as the exchange rate and equity prices, also tend to experience large falls, wiping out wealth. 17 The exchange rate regime in place also matters for the size of the real effects of these reversals induced by sudden stops of capital flows. Flexible exchange rates tend to act as shock absorbers. For example, through gradual depreciations, economies maintain export competitiveness and discourage offshore borrowing, which allows countries to accommodate better this type of external shock, and avoid the costly currency crises that are often times associated with reversals and sudden stops. 18 Indeed, Komárek and Melecky (2005) looked at the impact of these shocks on economic growth and found that the largest negative effects are associated with the combination of reversals and currency crises (that is, when the impact of the sudden stop on the real exchange rate is large). The effects on current account reversals differ across sectors. Craighead and Hineline (2013), for example, look at the effect of reversals by sector for a sample of industrial economies, finding that it matters how investment-intensive the sector is, and whether it is tradable. Investment-related sectors, such as construction, mechanical engineering and shipbuilding suffer significant contractions, while other sectors, particularly those related to tradable commodities, and particularly those that are also not heavily dependent on investment, may even benefit. These dynamics are consistent with real exchange rate depreciations and credit crunches accompanying reversals. Debt-dollarization makes current account reversals and sudden stops even more costly. When firms hold unhedged dollar-debt, as it was the case in Indonesia in 1997/98, their net worth falls since the sudden stop leads to a sharp depreciation of the real exchange rate. This often means that firms struggle to service their debts, which may have systemic consequences on the payments system. Public finances can also be severely hit by sudden stops. To the extent that in developing countries and large portion of public debt is denominated in dollars, sudden stops often increase the burden of public debt repayment, expressed in domestic currency. To maintain servicing of that debt, the authorities will have to run higher primary surpluses, with negative implications for aggregate demand and economic activity. Source: Authors’ elaboration c. Export competitiveness Export Open economies like Indonesia are exposed to shocks in the global marketplace that competitiveness is affect their ability to earn foreign exchange through exporting and thus their current key to reducing account position. Diversification of destination and product dimensions helps to vulnerability reduce these vulnerabilities. 17 Empirical studies also suggest that the output effects of sudden stops vary substantially by country, depending on a number of factors including openness and exchange rate misalignments prior to the sudden stops. When exchange rates are aligned with equilibrium levels prior to the sudden stop, then the induced effect of the shock is not as large, and nor are the real effects on the economy. For more on the effects of sudden stops, see for example Milesi-Ferreti and Razin, 2000 or Freund and Warnock, 2007. 18 See Edwards, 2004. Also Edwards, 2004a and Mello et al, 2011, look at the effect of current account reversals on growth and find economically and statistically significant negative effects. THE WORLD BANK | BANK DUN IA 11 Indonesia Current Account Assessment Indonesia is an During the period 2000-2013, Indonesia’s exports and imports of goods and services active participant in as a share of GDP averaged 30.2 and 26.3 respectively. 19 These indicators are slightly international trade below the average that other countries with similar or even lower levels of markets… development exhibit, such as Vietnam (63.5 and 70.2 respectively). However, assessing integration into the global marketplace by just looking at export and import ratios to GDP can be misleading. Many factors determine a country’s opportunities to participate in international trade markets, apart from its relative level of development. These include, for example, the size of its internal market, which affect domestic firms” incentives to trade, or whether these countries are landlocked or islands, which affect transportation costs. Figure 16 and Figure 17 show country rankings of export and import orientation indices that factor in these elements. These indices are zero when the actual ratio of export to GDP matches what would be predicted given the aforementioned characteristics (i.e. if exports to GDP are at the average level of those countries with similar characteristics). 20 When exports to GDP are above average, then the index is positive (and vice versa). Indonesia’s export orientation is well above the average, greater than China and India, and substantially above those of Brazil, Turkey and South Africa. With respect to import orientation, Indonesia is just above average, close to the scores of India, and China. Figure 16: 2000-2013 Exports/GDP Figure 17: 2000-2013 Imports/GDP 60 60 50 50 40 40 THA 30 30 IDN 20 20 CHN CHN IND 10 IND 10 IDN 0 0 -10 -10 ZAF -20 ZAF TUR TUR -20 BRA -30 BRA -30 -40 -40 -50 -50 Source: World Bank staff calculations based on WDI Source: World Bank staff calculations based on WDI 19 Exports, imports and GDP are measured in nominal terms. 20 To calculate the export and import orientation index, average export to GDP ratios over the period 2000-2013 are regressed against averages over the same period as for the log of real GDP, the log of population, a landlockedness dummy and an island dummy. THE WORLD BANK | BANK DUN IA 12 Indonesia Current Account Assessment …but growth in The value of Indonesia’s Figure 18: Merchandise Export Values exports has lagged merchandise exports (index base=100 in 2005) some of its peers… (measured in current USD) IDN BRA IND ZAF grew at an average annual TUR CHN VNM MYS rate of 11.2 percent during 450 the period 2002-2013. In 400 this respect, Indonesia’s 350 performance lags behind 300 that of the other so-called 250 “fragile five” emerging 200 economies such as India, 150 with export growth rates of 100 19 percent, Turkey, with 50 14.4 percent, Brazil, with 0 13.4 percent, and South Africa with 13.3 percent, and also behind other Source: UN Comtrade countries in the region such as China, with 19 percent and Vietnam with 21 percent growth per annum. Since 2011, Indonesia’s export growth slowed down substantially, on the back of sluggish export demand and decreasing prices for its key commodities. Since 2011, Indonesian exports of merchandise fell at a rate of 5.3 percent, while those of India, for example, grew at a rate of 7.9 percent (Figure 18 and Figure 19). …and import Considering the most recent 2012-2014 period, and examine both merchandise deceleration followed export and import trends we observe, first, that export deceleration was followed, export deceleration with a lag, by import deceleration. This pattern is predictable. Previous World Bank analysis shows both a short and long run relationship between imports and exports. 21 Indeed, a large portion of the slowdown in import growth observed during 2012 is explained by the contraction in exports (while the demand for imported inputs to export production is estimated to have contributed almost all import growth in 2013). As a consequence, during 2012-2014, the merchandise trade balance remained relatively unchanged despite the reduction in import growth. 21 World Bank, March 2014,“Indonesia Economic Quarterly: Investment in Flux”. THE WORLD BANK | BANK DUN IA 13 Indonesia Current Account Assessment Figure 19: Merchandise exports and imports since 2011 Figure 20: Growth of merchandise exports and imports – monthly series (growth rates, percent) (USD bn) Exports Imports Trade balance 30% Exports Growth Imports Growth 20 4 20% 18 Merchandise exports, imports 3 Merchandise trade balance 16 10% 14 2 12 1 0% 10 8 0 -10% 6 -1 4 -20% -2 2 0 -3 -30% May-11 May-12 May-13 May-14 May-12 May-13 May-14 Source: GTIS Source: GTIS The last 10 years saw In 2002/03, Indonesia’s five largest export destinations (Japan, USA, Singapore, a decline in the Korea, Rep., and China) accounted for 58 percent of total exports. In 2012/13, the concentration of largest five (with India now replacing Korea) accounted for 53 percent. The decline exports in the largest in the concentration of export earnings is substantial, although the level of five destinations, concentration remains relatively high (Figure 21). Despite more than 50 percent of reducing export earnings being concentrated in five export destinations, Indonesian exporters vulnerabilities to sold into 177 markets in 2012/13. This number is relatively large (Figure 22), and country-specific higher than those reached by Brazilian, Indian, South African, Turkish and shocks… Vietnamese exporters, although lower than those reached by Chinese or Malaysian ones. Figure 21: The share of the top 5 export destinations Figure 22: …While the overall number of export decreased… destinations increased slowly (percent, percent) (count) 70% 2002/03 2004/05 2006/07 2002/03 2004/05 2006/07 2008/09 2010/11 2012/13 2008/09 2010/11 2012/13 220 Percentage of exports explained by Number of countries reached by 60% 200 180 50% 160 top 5 markets 140 exporters 40% 120 30% 100 80 20% 60 10% 40 20 0% 0 IDN BRA IND ZAF TUR CHN VNM MYS IDN BRA IND ZAF TUR CHN VNM MYS Source: UN Comtrade Source: UN Comtrade …but exports The product diversification trend is opposite to that observed above with respect to became more destinations. Indonesia has increased its exposure to its 5 most important export concentrated along products (natural gas, coal, petroleum oils, palm oil and rubber in 2011-2013), from THE WORLD BANK | BANK DUN IA 14 Indonesia Current Account Assessment the product 25.8 percent to 30.3 percent of total exports (an increase is also observed in India and dimension, Brazil) (Figure 23). This concentration is the expected consequence of increasing increasing risk of commodity prices. They have an immediate effect on the values exported of these product-specific commodities, and a further effect that takes longer to materialize, via volumes. As shocks commodity prices increase, resources are reallocated away from other activities and into the production of that commodity. The process of concentration is also evident when looking at the trends in number of export product varieties, which fell, from 3,297 in 2002/03 to 3,146 in 2012/13. Among the group of comparators, Indonesia shows one of the lowest product scopes (Figure 24). Figure 23: The concentration of exports in the top 5 Figure 24: …while the overall number of export products increased…. products declined (percent, percent) (count, units) 2002/03 2004/05 2006/07 2002/03 2004/05 2006/07 2008/09 2010/11 2012/13 2008/09 2010/11 2012/13 40% 5000 Number of exported products Percentage of exports explained by 35% 4500 30% 4000 3500 top 5 products 25% 3000 20% 2500 15% 2000 1500 10% 1000 5% 500 0% 0 IDN BRA IND ZAF TUR CHN VNM MYS IDN BRA IND ZAF TUR CHN VNM MYS Source: UN Comtrade Source: UN Comtrade As observed in most When we decompose Figure 25: Export growth has been concentrated in large countries, most Indonesian export growth existing products in old markets export growth in into a portion from selling (percent, percent) Indonesia consists of more of the same products 2002-2006 2007-2009 2010-2013 more exports of the to the same destinations 140% same products to the (intensive margin), and Percentage of export growth 120% same markets selling new products to 100% new markets, new 80% products to old markets, 60% or the same existing 40% products to new markets 20% (extensive margin), we 0% observe that more than 95 -20% percent of export growth -40% Old New New Old is accounted for by the Products in Products in Products in Products in intensive margin. This is Old Markets New Old Markets New not surprising for a large Markets Markets economy such as Source: UN Comtrade Indonesia; when a country already trades a wide variety of products, with numerous other countries, diversifying is costly. This is because it typically implies overcoming technological or input-related constraints that prevented firms from producing these products in the first place, or market access barriers that prevented firms from penetrating those THE WORLD BANK | BANK DUN IA 15 Indonesia Current Account Assessment markets earlier. However, the pattern observed in Figure 25 is a confirmation of what was revealed by Figure 23 and Figure 24. Diversification along the product dimension decreased from low to null when comparing the pre-commodity boom period 2002-2006, with the commodity boom and the post-boom periods 2007- 2009 and 2010-2013. Even the contribution of diversification along the destination dimension to export growth has fallen over the years. Sophistication The sophistication of Indonesia’s export basket, as measured by Hausmann, Hwang appears to be and Rodrik (2006)”s EXPY indicator has been declining even before the increases in declining… commodity prices. 22 Comparing the export sophistication trends with other natural resource abundant countries, such as Brazil and South Africa, suggests that the decrease is not due only to rising commodity prices. For these two latter countries, the downward trend in sophistication starts in the mid-2000s – the period of the commodity price boom - while for Indonesia, the negative trend pre-dates the boom period and is more or less continuous (Figure 26). Figure 26: Indonesia’s export sophistication has been Figure 27: …while quality of exports has increased declining… slowly (index) (index) BRA CHN IDN IND BRA CHN IDN IND MYS TUR VNM ZAF Export Sophistication Index (EXPY) MYS TUR VNM ZAF 24000 1 Export Quality Index (Frontier=1) 23000 22000 0.9 21000 20000 0.8 19000 18000 0.7 17000 16000 15000 0.6 Source: WDI, UN Comtrade Source: IMF-WTO Database …but quality, An indicator of quality based on the prices that export products obtain in international although still low, is markets reveal that Indonesia’s export quality is low, both when compared to the increasing “fragile five” club, and when compared to other East Asian economies (Figure 27). Until 2008, the average export quality was below that of India, Brazil, China, Malaysia, Turkey, South Africa, and Korea. Despite the low export quality level, Indonesia’s 22 The EXPY indicator estimates the sophistication of products on the basis of income levels of countries that produce (and export) these products. If a product, say, internal combustion engines, is largely produced by rich countries, that product would be revealed to be “rich” and sophisticated. This outcome-based measure of sophistication for each product, called PRODY, is a weighted average of the per capita GDP of countries producing that good. PRODY values of all products that a country exports are then weighted by the products’ share in the country’s total export basket, and summed to derive a country’s level of GDP per capita as inferred from the sophistication of its export basket. Called EXPY, this measures the income content of a country’s export basket, and it is the indicator used in this report. The indicator is regarded as a more inclusive measure of sophistication than intensity in technology or R&D, and it is a predictor of a country’s future growth rates. Indeed, countries with high EXPY tend to have higher growth rates in the future – countries “become” what they export by converging to the income level implied by their export baskets. THE WORLD BANK | BANK DUN IA 16 Indonesia Current Account Assessment export bundle has been systematically upgrading in quality since 2000, now being slightly above of that of Brazil. 23 Export survival is The chances of Indonesian export flows that start in a given year to remain active amongst the lower in past the first year is among the lowest, together with that of South Africa, among the the group comparator group. For countries to achieve fast export growth and diversification, both successful entry into export markets and survival of export flows are crucial. In fact, the literature finds exporting to be an extremely risky activity, and particularly so in less advanced countries. In Indonesia, the chances that a specific product exported to a given destination remains active for at least one year is 51 percent, similar to those exhibited by South African (50 percent), Vietnamese (51 percent), Malaysian (52 percent) or Indian (52 percent) products, and substantially below those displayed by Brazilian (56.4 percent), Turkish (59 percent) or Chinese (65 percent) products. The probability of Indonesian export flows persisting for five years or more is the lowest in the group (21 percent). 24 Among different types of products exported, Indonesia’s footwear, rubber, and wood products are those with the highest chances of remaining active for longer. More sophisticated products, like machinery and equipment and transport equipment show the lowest survival rates. International empirical evidence shows that networks and productivity are crucial for export survival. Firms that are more productive tend to be more resilient to negative shocks, and to maintain their exports for longer. In addition, networks have been found to positively affect export survival by increasing the circulation of information related to compliance with standards required to export specific products or to specific destinations. 25 Figure 28: Indonesia’s export survival is amongst the Figure 29: …with some products performing better lower in the group… than others (probability) (probability) BRA CHN IDN IND Vegetable Mineral MYS TUR VNM ZAF Chemicals Plastic, Rubber 1.00 Wood Textiles, Clothing 1.00 Footwear Machinery Transportation 0.75 0.80 Survival probability Probability 0.60 0.50 0.40 0.25 0.20 0.00 0.00 0 1 2 3 4 5 6 7 8 9 10 11 12 0 1 2 3 4 5 6 7 8 9 10 11 12 Years active Years active Source: WDI, UN Comtrade Source: WDI, UN Comtrade 23 The export quality index is obtained from the IMF-WTO database on export quality and is based on unit values obtained by a given country for a given product relative to the unit value obtained by the country that exports that product at the highest unit value (the quality “frontier”). The quality frontier is normalized to unity. 24 The 95 percent confidence interval for Indonesia’s export survival probability is (0.504, 0.508). This implies that Indonesia’s survival probabilities are statistically significantly lower than those of India, Brazil, Turkey, China, Malaysia and Vietnam (and only marginally greater than those of South Africa). 25 For more on links between export survival and productivity, see Martuscelli, A. and G. Varela, 2015. For more on networks and survival, see Cadot, O. et al, 2013. THE WORLD BANK | BANK DUN IA 17 Indonesia Current Account Assessment C. What is behind Indonesia’s current account deficit? This section analyses the determinants of the current account deficit over the period 1988-2013. First, it discusses the estimated coefficients for a number of determinants (see Box 4, and Cusolito and Nedeljkovic [2013] for more on the methodology). Second, it combines information on the estimated sensitivities with the evolution of these determinants to calculate their contributions to the dynamics of the current account deficit. Box 4: Conceptual framework for the analysis of determinants of the current account balance We begin with the notion, developed in Box 1, that the current account balance reveals the difference between national saving (private and public) and national investments (private and public). To identify the underlying drivers of the current account balance, we relate that balance with the drivers of national saving, investments, government deficits and trade competitiveness factors that may affect the current account directly (with indirect effects on saving and investment): CA( X CA ) = S P ( X S ) − I P ( X I ) + S G − I G where XS are private consumption/saving determinants, XI denotes factors that affect private investment and XCA denotes factors that may influence the current account directly (for example, the export/import determinants, past FDI inflows). Assuming exogeneity of the trade drivers XCA the current account balance is defined as: CA = g ( X S , X I , S G , I G , X CA ) and the function g(•) is assumed linear. The prospective determinants are classified according to their sensitivity to policy decisions, and the sign of their expected effect on the current account balance (Table 1). Table 1: Prospective determinants of the current account balance in Indonesia expected effect on the current account balance Positive Negative Ambiguous SENSITIVITY TO POLICY DECISIONS Government expenditures, Sensitive to policy in the short run REER, credit, energy subsidies Sensitive to policy in the medium run Openness, FDI Sensitive to policy in the long run Relative income GDP growth Growth of main trading External partners, Commodity Oil prices Global uncertainty prices, terms of trade Lagged current Others Dependency ratio NIIP account/GDP Source: Authors” elaboration based on Cusolito, A. and M. Nedeljkovic (2013) The dependent variable in the analysis is the ratio of the current account balance to GDP. The explanatory variables and their relationship with the current account balance are described below: • Lagged current account balance: At an annual frequency, current account balances tend to show high persistence, associated with habit formation in consumption and saving, or agglomeration effects in investment. • Lagged Net International Investment Position (NIIP): net foreign assets can affect the current account balance in two ways. First, a large stock of foreign liabilities will require a country to run current account surpluses to pay them off. Second, the country will still pay interest on those liabilities, generating outbound income payments and thus tilting the current account towards a deficit. THE WORLD BANK | BANK DUN IA 18 Indonesia Current Account Assessment • Oil prices: for a net oil importer like Indonesia, oil prices directly affect the oil import bill, thus the current account. As Indonesia turned into a net oil importer only since 2002, oil prices are interacted with a dummy variable that takes value 1 from 2002 onwards, and zero otherwise. • Commodity prices: a positive terms of trade (ToT) shock can improve the current account via increased saving due to larger current income relative to permanent income (the Harberger-Laursen-Metzler effect). On the other hand, ToT shocks can also affect the optimal capital stock and change investment plans, leading to bigger current account deficits. The greater the persistence of the shock, the more dominant is the investment effect. • Government expenditures: rising government expenditures increase aggregate demand and depress national saving. Increased demand will be partly satisfied by additional imports (Ahmed 1996). The financing strategy of these expenditures (taxes versus debt) may matter if current generations do not internalize future government debt repayments (Obstfeld and Rogoff 1996). • Openness: trade openness has ambiguous effects on the current account balance. Less open economies may import less, which may reduce the current account deficit. However, the same countries may have difficulties servicing external liabilities, resulting in higher debt service costs and a greater current account imbalance. On the other hand, greater openness typically allows countries to undertake more investment and to finance the resulting current account deficits with capital flows from abroad. International trade is also a conduit for the transfer of technology, leading in the long run to economic development, thereby improving the current account balance. • Real Effective Exchange Rates (REER): REER appreciations induce an expenditure switching effect away from domestic goods and into foreign goods, for a given level of expenditure, which increases the current account deficit, all else equal. • Foreign Direct Investment (FDI): FDI has ambiguous effects on private domestic investment and the current account. It can crowd out domestic investment when local and offshore firms compete for scarce domestic resources (e.g. labor or finance). FDI may also generate local spillovers that “crowd in” domestic investment. Gross FDI may also worsen the current account, depending on the balance of the import content and the amount of profits repatriated on one hand, and the export orientation of multinationals as well as the increase in their production that may substitute imports for domestic consumption. • Relative GDP growth: the effect on the current account balance depends on agents” expectations about the implications of growth for future income. If agents consider it permanent, then saving rates could decline, increasing the current account deficit. If instead it is perceived as temporary, saving will increase and the current account balance will increase. GDP growth is computed relative to Indonesia’s main trading partners. Faster trading partners” growth boosts demand for exports, increasing the current account balance. • Credit to the Private Sector: proxies financial deepening, and aggregate demand. Relaxed borrowing constraints can reduce private saving (increase current account deficits). Also, if it reduces transaction costs and improves risk management, credit may encourage private saving (decrease current account deficits). • Relative income: developing economies are expected to run current account deficits as they accumulate capital goods. Eventually, the country will be sufficiently developed to pay its debts by running current account surpluses. • Global uncertainty (proxied with VXO index 26): an indicator of global uncertainty in financial markets, increases in the VXO index are expected to increase saving and reduce investments, according to the buffer stock theory. Moreover, higher global financial market volatility is associated with lower availability of foreign financing for emerging markets, thereby contributing positively to the current account balance. • Energy subsidies: a subcomponent of government expenditures, energy subsidies are expected to also increase aggregate demand. The negative effect of energy subsidies on the current account balance is expected to be larger than that of general government expenditures, as these subsidies directly affect the consumption of imported goods (such as oil). Some of the factors proposed as potential determinants are in fact jointly determined with the current account balance notably including the real exchange rate, for example. To deal with this “simultaneity” problem, we instrument variables that are likely to be endogenously determined with the current account balance. The instruments consist of lagged values of the variable in question, which are predetermined at time t. 27 Pre-determination, however, does not necessarily imply “exogeneity” if anticipation effects are present. For these reasons, a caveat is necessary: the results presented here should be read as conditional associations between variables, rather than as strict causal relationships. Source: Authors” elaboration based on “Toolkit for the Analysis of Current Account Imbalances” (Cusolito, A. and M. Nedeljkovic) 26 The VXO index is the Chicago Board Options Exchange Market Volatility Index based on S&P 100 (OEX) options. Given a longer available time history the index is preferred to the more conventional VIX index (computed based on traded S&P 500 [SPX] options). 27 This methodological choice is based on data availability (both in terms of the scope of potential additional explanatory variables, and on the time period available). THE WORLD BANK | BANK DUN IA 19 Indonesia Current Account Assessment 1. Determinants of Indonesia’s current account The model we use to Results of the model in Figure 30 are summarized in Table 2 below. The model explain the current performs generally well, with the difference between the annual or three year averages account dynamics of the implied current account balance and the observed balance remaining relatively fits the data well… small throughout the period (Figure 30). The only exception is the current account balance improvement over 2005-06 which is only partially explained. The specification captures the most recent deterioration of the current account balance, although its magnitude is slightly under-predicted. Figure 30: Actual and predicted current account balances (percent of GDP) 10% Actual CA Predicted CA 8% Actual and predicted current 6% account balance 4% 2% 0% -2% -4% -6% Source: World Bank staff calculations A reduction in the A one percentage of GDP reduction in the energy subsidy spend increases the current energy subsidy bill account balance by 0.81 percentage points of GDP. Among the group of variables supports the current that are, to some extent, under the control of the policy maker in the short run, account balance Indonesia’s current account balance is highly sensitive to a specific portion of government expenditure: energy subsidies. It is likely that recent fuel subsidy reforms will help reduce the current account deficit, by reducing the demand for oil, which accounts for a sizable portion of imports. The dynamics of all other government expenditures were found to be not statistically significant in the main specification, although the result may be due to the small variation in expenditure as a percentage of GDP in the non-fuel subsidy portion of government expenditure. The results from the second specification (last three columns in Table 2), instead, show the expected negative sign of the government expenditure coefficient, suggesting that increases in expenditures are associated with decreases in the current account balance. A depreciation of the A ten-percent average currency depreciation over the previous three years is estimated REER reduces the to increase the current account balance by almost half of a percentage point of GDP. current account Real exchange rate (REER) depreciations make domestic goods relatively cheaper balance than foreign goods, inducing an expenditure switching effect away from foreign into domestic production. As the structure of the regression indicates, however, the effect takes time to materialize. 28 The modest REER depreciation since 2013, therefore, is expected to be supportive of Indonesia’s external balancing. 28 The magnitude of the effect doubles if instead of looking at the relationship between the real effective exchange rate and the current account balance, we focus on the links between the bilateral real exchange rate of the rupiah/dollar and the current account balance. THE WORLD BANK | BANK DUN IA 20 Indonesia Current Account Assessment An expansion in the A one percent of GDP expansion of the credit stock decreases the current account credit stock balance by 0.05 percent of GDP. The rapid expansion of credit in the years through decreases the current 2012 is thus estimated to have pushed down the current account balance; for example account balance in 2013 by 0.13 percent of GDP. Slower credit growth since 2013 should help alleviate current account pressures. Trade integration Indonesia’s trade integration in the global marketplace relative to its trading partners, into the global as measured by the ratio of imports and exports to GDP, has been systematically marketplace associated with increases in the current account balance, after controlling for increases the current commodity prices. The effect of openness on the current account balance estimated account balance for Indonesia is in line with the findings for both developed and developing countries reported in Gruber and Kamin (2007) and Chinn and Ito (2007). This implies that further trade integration in the global marketplace relative to its trading partners should have a beneficial effect on the current account. Foreign direct An average increase in FDI inflows by 1 percentage point of GDP during the previous investment attracted three years leads to a decrease in the current account balance by 0.17 percent of GDP. complementary, self- This implies that FDI inflows increase demand for imports of intermediates and financing imports capital goods and profit repatriations by more than they increase exports or induce import substitution. A similar result is obtained when examining the link between lagged FDI and the trade balance, suggesting that the effect of FDI on net-exports is negative even before adding profit repatriations. There is ample The benefits from FDI Figure 31: Relationship between FDI and current evidence that should not be solely account deficits Indonesia has gained assessed by its impact on from increased FDI the current account, but inflows by the welfare gains that it brings the population. For example, Blalock and Gertler (2008) find evidence of FDI spillovers to local suppliers. Examining Indonesian manufacturing establishments, the authors show that local firms that supply foreign entrants gain Source: Authors” elaboration productivity as they face greater competition, which results in lower prices of inputs, not only for those foreign entrants, but also for domestic firms operating in the same market, which also benefit. Duggan, Rahardja and Varela (2013) find that reducing barriers to FDI in the services sector leads to increases in the productivity of manufacturers that use services inputs more intensively, with gains larger when restrictions are reduced in transport and telecommunications sectors. Indonesia’s faster Real GDP growth in excess to trading partners by 1 percent, leads to a decline in the growth than its current account balance by 0.3 percent of GDP. As domestic demand, and trading partners” consequently import demand, grows faster than external demand for exports, the trade balance deteriorates, putting pressure on the current account. In this sense, THE WORLD BANK | BANK DUN IA 21 Indonesia Current Account Assessment decreases the current Indonesia’s long span of above-trading-partners”-growth has contributed to the account balance reduction in the current account balance. Increasing relative Relative income, measured by the ratio of real GDP per capita in PPP terms to trading income has partners, has a positive impact on the current account balance. This result is expected, increased the current since due to diminishing marginal returns to capital accumulation, countries” need to account balance import capital declines as their average income increases. However, as Indonesia’s relative income per head remains low, the positive contribution of income convergence to the current account balance has been small over the sample period. Demographic Demographic changes, measured by the average five year dependency ratio, have a change supports an negative sign of small magnitude, which is expected given the longer horizon over increase in the which the demographic trends tend to materialize. The overall impact on the current current account account balance of demographic change in Indonesia is positive, in line with the balance decline in the dependency ratio. Commodity price Changes in the non-fuel commodity price index have a positive contemporaneous changes are impact on the current account balance. In addition, a positive sign of the coefficient important in the for the three year average terms of trade gains signals a higher positive impact of the short and medium commodity price changes on domestic saving related to investment in the medium run. run, thus contributing to the current account improvements beyond the direct trade impact. 29 The sign of the effect also implies the perceived transitory character of the past terms of trade movements. 30 Global Higher global uncertainty, measured by the change in the VXO index has a positive macroeconomic (but not statistically significant) effect on the current account balance. This is expected uncertainty increases and in line with the positive impact of the uncertainty on saving and negative on the current account investment due to precautionary behavior of the risk-averse household and the balance asymmetric nature of firm’s adjustment costs. Moreover, higher global uncertainty is associated with lower global liquidity and lower availability of foreign capital for financing current account deficit. Current account The current account balance in a given year is highly dependent on its previous levels. persistence matters This can be related to habit formation in consumption and saving or agglomeration effects in investment, which suggests a certain degree of inertia in the current account following its reversal. 31 The size of the persistence effect is in line with the literature on other developing and emerging markets (0.37-0.59). 32 It is however higher relative to some advanced emerging markets (e.g. Turkey or Poland, where this analysis has also been conducted) implying a slower current account adjustment after transitory shocks in Indonesia. The size of the past The size of the past NIIP is an important determinant of the future current account NIIP is an important balances and the estimated coefficient (0.009) is slightly lower than those typically determinant of the found in the literature for other developing and emerging markets (0.02-0.04), but future current well above that of some advanced emerging markets where this toolkit was account balances implemented (Turkey, Poland). Both the current account balance and the NIIP are defined in nominal terms, hence a positive coefficient implies that increases in the 29 The Harberger-Laursen-Metzler effect, due to Harberger, 1950 and Laursen and Metzler, 1950. 30 An alternative model presented in Table 2 incorporates oil prices, and reveals a small but negative effect of oil prices changes on the current account balance. 31 The result also provides further evidence on the stationarity of the current account balance. 32 See Table 1 and Table A2 in Beidas-Strom and Cashin, 2011 for overview of the literature. THE WORLD BANK | BANK DUN IA 22 Indonesia Current Account Assessment NIIP (reductions in net liabilities) result in increases in the current account balance, and vice versa. The positive coefficient suggests two channels of influence on the current account. First, the accumulated level of the NIIP has not presented a strong external borrowing constraint for most of the past period. In order to control for strong decline of the NIIP during the 1998 crises we include the interaction dummy for the period 1998-2001. The estimated coefficient for the interaction term is negative, implying a reduced impact of the NIIP on the current account deficit for high levels of the NIIP. Second, the NIIP accumulation also leads to income outflows in the current account, related to profit repatriations and the interest repayments on debt instruments. This is in line with the difference in the magnitude of the coefficients for the level and the interaction term (implying an additional impact beyond the borrowing constraint) and the increasing contribution of the NIIP to the current account balance over the most recent period which we discuss below. THE WORLD BANK | BANK DUN IA 23 Indonesia Current Account Assessment Table 2: Estimated coefficients of current acount balance models (dependent variable is the current account balance in percent of GDP) CI 90 CI 90 CI 90 CI 90 Standard Standardized percent percent Standard Standardized percent percent Variable Coefficient Coefficient error coefficient Lower Upper error coefficient Lower Upper bound bound bound bound Credit in percent of GDP (change) -0.0453* 0.0204 -0.1190 -0.0800 -0.0106 -0.0456* 0.0199 -0.1198 -0.0818 -0.0094 REER (ln change, 3 year average) -0.044 0.0220 -0.1260 -0.0895 0.0006 -0.05* 0.0227 -0.1427 -0.0985 -0.0021 WEO non-fuel commodity price index 0.048* 0.0180 0.1749 0.0183 0.0778 0.06* 0.0301 0.2192 0.0097 0.1108 (ln change) Relative GDP growth -0.3184* 0.0645 -0.3789 -0.4302 -0.2066 -0.26* 0.0862 -0.3094 -0.4131 -0.1069 Dependency ratio (5 year average) -0.013 0.0042 -0.0300 -0.0368 0.0106 -0.011* 0.0045 -0.0251 -0.0199 -0.0021 FDI in percent of GDP (3 year average) -0.1658* 0.0219 -0.0790 -0.2842 -0.0474 -0.1526* 0.0152 -0.0727 -0.1898 -0.1154 Relative openness 0.0155* 0.0038 0.1080 0.0012 0.0299 0.0174* 0.0045 0.1210 0.0093 0.0255 Relative income 0.02 0.0068 0.0146 -0.0127 0.0535 0.0176* 0.0063 0.0126 0.0044 0.0308 VXO index (ln change) 0.009* 0.0041 0.0799 0.0011 0.0168 0.003* 0.0010 0.0275 0.0015 0.0047 Government expenditure in percent of 0 0 0 0 0 -0.018* 0.0078 -0.0081 -0.0310 -0.0052 GDP (change) 3 year average terms of trade (ln 0.1058* 0.0260 0.1713 0.0518 0.1598 0.116* 0.0391 0.1882 0.0516 0.1809 change) Lagged current account 0.48* 0.0470 0.4760 0.3897 0.5699 0.5134* 0.0428 0.5093 0.4409 0.5859 NIIP in percent of GDP (beginning of 0.009* 0.0019 0.0769 0.0001 0.0179 0.0146* 0.0025 0.1247 0.0100 0.0192 period) NIIP*Crises dummy -0.004 0.0013 -0.0582 -0.0118 0.0031 -0.0112* 0.0011 -0.1505 -0.0138 -0.0086 Energy subsidy in percent of GDP -0.5847* 0.1524 -0.1975 -0.9120 -0.2574 (change) * Post 2002 dummy Oil prices (change) * Post 2002 dummy -0.022* 0.0119 -0.1173 -0.0415 -0.0030 Energy subsidy in percent of GDP -0.23* 0.0603 -0.0907 -0.4018 -0.0568 (change) Oil prices (change) -0.012* 0.0052 -0.0947 -0.0205 -0.0042 Note: *indicates significance at 90 percent confidence level Source: World Bank staff calculations 24 THE WORLD BANK | BANK DUN IA Indonesia Current Account Assessment 2. Contributions to Indonesia’s current account dynamics What have been the In addition to understanding the determinants of Indonesia’s current account, and key contributing the current accounts sensitivity to each determinant, as described in the previous factors to current section, this section explores the contribution of each determinant to current account account dynamics? dynamics. 33 REER appreciation REER appreciation over 2010-11 had a mild, lagged negative contribution to the contributed to the buildup of the current account deficit in 2012. The subsequent real depreciation of decrease in the the rupiah since 2012, however, has helped to partially offset the deficit during 2013. current account Further REER depreciation, though modest, should support further current account balance adjustment, though with a lag as its effect takes time to impact on agents” decisions. Indonesia’s trade As Indonesia integrated into the global economy faster than its main trading partners integration over the period 1988-1995, the contribution of integration to increase the current contributed account balance also increased, to add up to 1.5 percentage points to the current positively account balance. During the past decade, however, this contribution decreased substantially to below 1 percentage point, since, relative to its peers, Indonesia is becoming less integrated into the global marketplace. FDI inflows had a FDI inflows have made a mild negative contribution to the current account balance, narrow negative by generating more import demand and dividend payments than they generate contribution through exports in the short run. However, FDI is generally associated with increased self-financing productivity in the domestic economy, increased diversification and sophistication of imports the production and export bundles, thus reducing vulnerabilities associated with current account deficits, as well as benefiting the economy beyond its direct, short- term impact on the current account balance. Stronger growth than Periods of strong growth of the Indonesian economy relative to the rest of the world in trading partners have boosted demand for imports beyond the increased demand for Indonesia’s contributed to a exports. Higher relative GDP growth made a substantial negative contribution to the decrease in the current account balance during 1990-92 and 1993-95, when Indonesia’s growth was current account stronger than its main trading partners. A combination of continued solid economic balance growth and a sluggish world economy caused an increasingly large negative contribution of relative GDP growth during recent years. In 2011-2013, for example, this variable added, on average, 0.8 percent of GDP to the current account deficit. Rising relative The stage of the economic development, measured by relative income, has had a income made a slowly increasing positive effect over the entire period of analysis. This suggests that positive contribution the income convergence process is still at an early stage in Indonesia and its to the current importance is expected to increase in the future. account balance 33 The distinction between the sensitivity and contribution of a given determinant is important. The former illustrates by how much the current account changes given a small change in one of the determinants. The latter combines information on the sensitivity to a given determinant with the changes in such a determinant over the period, and illustrates how much of the current account deficit observed during a given period is attributable to changes in determinants. Take the case of credit. Indonesia’s current account is sensitive to credit growth, which increases the deficit, all else constant. However, the contribution of credit growth to the current account deficit of Indonesia has fluctuated over the period of analysis. Credit contractions led to improvements in the current account balance over the 1998-2001 crises period, while credit expansions helped increasing the expanding current account deficit over the most recent period through 2013, given by the relative changes in the contributions illustrated in Figure 32. THE WORLD BANK | BANK DUN IA 25 Indonesia Current Account Assessment The dependency The surge in commodity prices had a strong and positive contribution over 2005- ratio, reducing 2011. In this respect, it has been argued that Indonesia will benefit from a national saving, “demographic dividend”, as the current dependent population reaches working age. contributed to the Reaping the “demographic dividend” benefits related to the current account require decrease in the declining fertility rates, so that dependency ratios decrease (that is, the working age current account population grows at a faster rate than the overall population). This, all else equal is balance expected to increase national saving, and with it, the current account balance. The surge in The surge in commodity prices had a strong and positive contribution over 2005- commodity prices 2011. In particular, the commodity price boom of the period 2005-2007, added two had a strong and thirds of a percentage point of GDP to the current account balance during the period. positive contribution The contribution shrank by a factor of ten toward the period 2011-2013, when over 2005-2011. commodity prices only contributed a 0.07 percent of GDP to the current account balance. The average Finally, the average contribution of the accumulated net international investment contribution of the position to the current account balance increased over the post 2008 period, although accumulated net the size of its contribution is still lower relative to the pre-1998 period. The finding international reflects a strong buildup of foreign gross liabilities, which increased from USD 228 investment position billion in 2008 to USD 560 billion in 2013, including an increase in the FDI stock to the current from USD 72 billion to USD 251 billion. These developments imply further financial account balance integration in addition to higher potential growth due to foreign investment, with the increased over the consequent decrease in the income balance over the same period (from -2.77 to -3.13 post 2008 period percent of GDP). As the share of Rupiah-denominated liabilities (mostly government bonds and equities) increases, exchange rate movements will have a lesser effect on the income balance. Figure 32: Contributions to the current account deficit (percent of GDP) 6% d.energ*oil_b d.commodities 4% ToT_ch_3Y 3Y fdi 2% d_vix rel. income 0% 3Y d.reer CA lag -2% rel. opennes 5Y dependancy -4% rel. gdp credit -6% NIIP Actual CA Predicted CA Note: findings of model version where fuel subsidy bill is interacted with 2002 step dummy variable Source: World Bank staff calculations THE WORLD BANK | BANK DUN IA 26 Indonesia Current Account Assessment D. Are recent shifts in Indonesia’s current account structural or cyclical? Adding a structural The analysis of the current account determinants presented so far identified key versus cyclical drivers of the current account but did not distinguish either their origin (national perspective saving or national investment changes), nor the cyclical or structural nature of their impact. This section investigates the cyclical and structural determinants of the saving and investment components of the current account, estimating an approximation to the cyclical and structural dynamics of the current account itself, which is helpful to assess whether the current account is in line with its fundamental value. Box 5: Methodology for the estimation of dynamics and determinants of saving and investment To analyze the persistence of the variables and the influence of the corresponding drivers we assume that the net national saving and net national investment (seasonally adjusted in percentage of GDP, Y) are the sum of its structural, cyclical and irregular components: Yt = Yt S + Yt C + ε t Further, in order to identify the individual components of the variable Y we use additional information embedded in the structural and cyclical determinants of the variable (discussed in Box 6) and specify the following law of motion for the components: −1 + β X t + ε t Yt C = θYt C C C Yt S = ϕYt −1 + δX t + ε t S S S where , ɛ are structural and cyclical shocks which are assumed to be zero mean, normally distributed, serially and cross uncorrelated. and , denote variables which affect the structural and the cyclical component of the variable Y, respectively. By using additional information embedded in the structural and cyclical determinants within the state space, the estimates � should be closer to the underlying economic movements in the variable in comparison to � , purely statistical measures of � . The choice of whether a variable impacts the structural or temporary segment � , of the variable Y is made a priori and is based on the results in the existing literature or statistical properties of the drivers. For example, variables which may have both a long-run and temporary impact (e.g. commodity price index) are decomposed into slow-moving (3 year rolling average) and cyclical components (current or lagged change) and the two are used as separate drivers of the two components of saving (investment). Both components and the parameters are estimated jointly in an extended state space specification. A Kalman filter with diffuse initialization is used for estimation of the state space. The application of the Kalman filter requires further assumptions on some parameters of the model. Following the literature on estimation of the natural rate of unemployment, we assume two “signal to noise” ratios – a ratio of the variance of the structural component to the variance of irregular component of 0.15 and a ratio of the variance of the temporary or cyclical component to the variance of irregular component of 0.4. 34 Alternative specifications of the ratio do not materially change the results. Due to the relatively small length of the sample and a large number of potential determinants, different specifications were estimated and the final specification is that which minimizes the AIC and BIC criterion. 34 The ratio allows imposing a constraint on the smoothness of the component. Lower values of the ratio imply lower volatility of the components. THE WORLD BANK | BANK DUN IA 27 Indonesia Current Account Assessment Following the estimation of the structural and cyclical component of the saving and investment component of the current account, the structural and cyclical components of the current account itself are calculated as the difference. Quarterly data over the Q1 1992-Q3 2014 period is collected from various sources. The sample size and the choice of variables below reflect data availability, as discussed further in Box 6. Source: Authors” elaboration Box 6: Determinants of structural and cyclical investment and saving Table 3: Determinants of structural and cyclical investment and saving Structural Cyclical Investment • Two-year average growth of output per • Change in the real lending rate 36 as a proxy for the worker as a proxy for productivity growth 35 user cost of capital • Three-year average level of the ICRG index • Lagged credit expansion (seasonally adjusted) as a composite measure of quality of measured by the change in credit to GDP ratio institutions, the business environment and • Lagged real exchange rate change, lagged change political and macroeconomic risks in the commodity price index • Three-year average FDI inflows (in • Yield on a 10-year US government bond, as an percentage of GDP) indicator of international liquidity. • Three-year average change of the commodity price index • Three-year average house price index as to capture construction investment trends, which have been positive during the period • Two year average growth of the GDP per • Change in real deposit rate, GDP per capita growth Saving as a proxy for short term income increases capita as a proxy for income effects; • Three-year average level of the ICRG • Previous credit expansion measured by the change economy index as a composite measure of in the (seasonally adjusted) credit to GDP ratio the macroeconomic conditions 37; • Previous change in the commodity price index • Demographic changes (dependency ratio) • GDP per capita growth as a proxy for a short term • Two -year average change of the income increases, commodity price index as a proxy for the • Change in VXO index as a measure for global terms of trade effects uncertainty • Previous one year and three month return on stock price index • Change in inflation volatility as a proxy for short- term macro uncertainty Data for a number of potentially important determinants of structural investment and saving were unfortunately not available or not compatible with the frequency and time period of the analysis. For structural investment, data were not available on labor cost dynamics (average wages or real wages), labor market policies (ratio of minimum to average wage, gross unemployment replacement rate) and the business and product market environment variables were not available for the period of study. For structural saving, data were unavailable on social security programs (public social or health spending), labor market policies (ratio of minimum to average wage, gross unemployment replacement rate) and urbanization. In the cyclical investment and saving components, some of the variables entered the specification with lags in order to capture the time-varying effects in investment and saving decisions. The specification also includes a dummy variable for the quarters with significant changes in the investment to GDP ratio during the 1998 crisis (1998 Q1-Q2), and in the saving to GDP ratio (1998 Q1-Q3). Source: Authors’ elaboration 35 Total factor productivity data were not available at quarterly frequency. Data on total employment are interpolated (using cubic splines) to quarterly periodicity for construction of the output per worker series. The results are similar using a three year average growth. 36 A measure of expected inflation is constructed using Consensus forecast data after 1995 and the past three year rolling average inflation for the 1992-1995 period. 37 Note that in comparison to the analysis of investment determinants we use ICRG index which focuses on the state of the economy rather than a broad index which includes quality of the institutions and the business environment as the latter have less impact on saving decisions. THE WORLD BANK | BANK DUN IA 28 Indonesia Current Account Assessment 1. Determinants of investment in Indonesia Structural Figure 33 shows the breakdown of investment in Indonesia into its cyclical and investment declined structural components. The structural component of investment took a declining path at the onset of the at the onset of the 1998 crisis, which continued for the subsequent four years. The 1998 crisis, decline in structural investment was driven by heightened macroeconomic recovering only in uncertainty, and falls in economic activity and FDI inflows. The structural component 2003 and stabilizing picked up around the year 2003, driven by favorable commodity prices and an at around 29 percent improved macroeconomic environment, encouraging the return of FDI inflows. The of GDP in 2010 contribution of productivity to structural investment became stronger after 2008, while investment stabilized at a level of around 29 percent of GDP from 2010 onwards. Cyclical investment Cyclical dynamics exhibit a Figure 33: Components of aggregate investment appears to have similar behavior. The (percentage of GDP) peaked in 2012 cyclical component of Investment Structural Cyclical investment fell sharply 35 during and after the 1998 30 crisis. It subsequently 25 recovered, with a positive 20 cycle from 2002-05, and 15 period of relative decline in 10 2005-07. Since the year 5 2007 the movements in the 0 cyclical component have been more positive, -5 peaking around 3 percent -10 1992 Q4 1993 Q4 1994 Q4 1995 Q4 1996 Q4 1997 Q4 1998 Q4 1999 Q4 2000 Q4 2001 Q4 2002 Q4 2003 Q4 2004 Q4 2005 Q4 2006 Q4 2007 Q4 2008 Q4 2009 Q4 2010 Q4 2011 Q4 2012 Q4 2013 Q4 of GDP in 2012. The last quarters of 2013 saw a slowdown in the cyclical Source: Authors’ estimates component following a slowdown in commodity prices and credit growth, with a minor rebound in the first two quarters of 2014. Table 4: Determinants of aggregate investment Variable Coefficient Standard Error structural Lagged value 0.9083 0.0394 GDP growth per worker (2Y average) 0.2477 0.1381 ICRG (log, 3Y average) 0.3031 0.1535 FDI (in percent of GDP, 3Y average, 1Y lag) 0.1291 0.0542 CommPixog (3Y average change) 0.0542 0.0220 House price index (3Y average change, 1Y lag) 0.0394 0.0212 cyclical Lagged value 0.9119 0.0341 Real exchange rate (change, 3Q lag) -0.0200 0.0103 Yield on 10 year US government bond (change, 4Q lag) -0.0033 0.0026 CommPixog (change, 4Q lag) 0.0129 0.0064 Credit (in percent of GDP, change, seasonally adjusted 1Q lag) 0.1322 0.0288 Real lending rate (change) -0.6216 0.0947 Dummy (1998Q1,Q2) 0.0389 0.0145 Log likelihood -98.803 AIC 225.607 BIC 259.471 Source: Authors’ estimates THE WORLD BANK | BANK DUN IA 29 Indonesia Current Account Assessment a. Structural component of investment Productivity growth Productivity growth is found to be a determinant of the structural component of boosts structural investment. Rising medium-term productivity growth increases firms” profit investment… expectations and consequently investment. …as does Macroeconomic stability is positively associated with structural investment. Reduced macroeconomic macroeconomic and political risks and improvements in structural characteristics of stability… the economy (as captured by the ICRG indicator) lead to higher investment. This result is in line with that previously reported by the IMF (2012). 38 …favorable Due to the large size of the commodity sector, commodity prices play a structural role commodity and in investment. Commodity price changes over the medium-term tend to impact the housing prices… investment behavior of firms, reflecting the large importance of the commodity sector for the overall dynamics of investment, and of economic activity in general. The structural component of investment in Indonesia is also positively associated with medium run trends in house prices. …and past FDI Lagged FDI inflows are positively associated with the structural component of inflows help crowd- investment rates. An increase in the average of the previous three years of FDI in domestic inflows by 1 percent of GDP, increases gross investment rates by 0.13 percent of investment GDP. This may be indicative of “crowding in” effects of past FDI on national investment, given that the model controls for output growth and commodity prices. This helps explain the finding of the previous section suggesting that increases in FDI flows into Indonesia are associated with subsequent increases in the current account deficit. As FDI increases crowd in further domestic investment, this increases the needs for foreign financing, and therefore increases the current account deficit. The result may also be capturing a “momentum” effect that operates beyond commodity price effects or booms in economic activity. For example, if investment promotion efforts are clustered together in a particular time period, then past FDI booms may be correlated with current booms in investment (and the converse is also true). b. Cyclical component of investment Lower cost of A fall in the lending rate lowers the user cost of capital and leads to higher capital, and credit investment. 39 The positive estimated impact of the credit expansion that took place expansions, boost through 2013 is in line with the experience of other emerging markets (consistent cyclical with a still shallow financial market and the prevalence of bank credit as the source investment… of external finance for firms). 40 …as do real currency Increases in the cyclical component of investment are positively associated with real depreciations… exchange rate depreciation. A negative sign for the real exchange rate change suggests that improvements in price competitiveness lead to export-led investment that tends to dominate the increased cost of imported capital and intermediate goods. Real exchange rate depreciations also encourage investment to substitute away from more expensive imports. 38 The IMF explored the impact of two specific sources of macroeconomic volatility on aggregate investment: interest rates and exchange rates. They found a negative correlation between the two volatility indicators and aggregate investment. These two variables are included under the cyclical component of investment 39 IMF (2012) reports a similar result on the growth of aggregate investment. 40 For example, in Turkey and Georgia, two countries in which this toolkit was also piloted, revealed similar responses of the cyclical component of investment to credit expansions. In Turkey, the estimated coefficient was 0.1388, while in Georgia it was estimated at 0.0345. THE WORLD BANK | BANK DUN IA 30 Indonesia Current Account Assessment …upswings in global Commodity price gains are associated with increases in the cyclical component of commodity prices… investment, with a one year lag. Increases in the index of Indonesia’s main 6 commodity prices excluding oil and gas tend to increase investment over the cycle. Higher commodity prices incentivize investments in the commodity-producing sector as its relative profitability increases. In addition, they also impact on the availability of investable funds, thus relaxing credit constraints that may be impeding investment in that sector or elsewhere. ...and lower cost of Lower yields on US bonds are associated with a lagged increase in cyclical investment. international This finding is consistent with reallocation of international portfolios in search for financing higher yields. 2. Determinants of saving in Indonesia The structural The estimated cyclical and Figure 34: Components of aggregate saving component of saving structural components of (percentage of GDP) has declined since saving are shown in Figure Saving Structural Cyclical 2009… 34. The estimated 40 structural component 35 dropped from 26 percent 30 of GDP at the onset of the 25 1998 crisis to slightly above 20 21 percent in 2000. It 15 fluctuated over 2000-02 10 and increased continuously 5 thereafter, until 2009, 0 driven by improvements in -5 all key determinants. The 1992 Q4 1993 Q4 1994 Q4 1995 Q4 1996 Q4 1997 Q4 1998 Q4 1999 Q4 2000 Q4 2001 Q4 2002 Q4 2003 Q4 2004 Q4 2005 Q4 2006 Q4 2007 Q4 2008 Q4 2009 Q4 2010 Q4 2011 Q4 2012 Q4 2013 Q4 post-2009 decline in saving was a consequence of both structural and cyclical Source: Authors” estimates factors, driven on the structural side by declining GDP growth and the extended fall in commodity prices. ..together with the The cyclical behavior over the post crises periods implies relatively long cycles of five cyclical component years: a positive cycle following the 1998 crisis, negative cycle over 2004-08, and a positive cycle post 2008, which is showing signs of moderation. Table 5: Determinants of aggregate saving Variable Coefficient Standard Error structural Lagged value 0.8901 0.0401 CommPixog (2Y average change, 1Y lag) 0.0502 0.0226 GDP per capita growth (2Y average) 0.2685 0.2137 ICRG (log, 3Y average, 1Y lag) 0.0125 0.0058 Dependency ratio (3Y average, 1Y lag) -0.0318 0.0211 cyclical Lagged value 0.8151 0.1216 Credit (in percent of GDP, change, seasonally adjusted 1Q lag) -0.0011 0.0005 Three month return on stock market index 0.0203 0.0092 CommPixog (change, 1Q lag) -0.0192 0.0147 GDP per capita growth (2Q lag) 0.0189 0.1202 Dummy (1998Q1,Q2) 0.0815 0.0159 Log likelihood -159.490 AIC 342.990 BIC 372.720 Source: Authors’ estimates THE WORLD BANK | BANK DUN IA 31 Indonesia Current Account Assessment a. Structural component of saving Increasing real An increase in real income (GDP per capita) over the medium term is associated income per capita with increases in the structural component of saving. The positive sign of the adds to the structural relationship is consistent with the empirical evidence for developing countries component of (Carroll and Weil 1994) that in growth-accelerating episodes, higher growth tends to saving… precede and statistically cause higher saving, while the opposite tends to be the case in developed countries. …together with Improvements in the overall state of the economy increase the structural component economic stability… of saving. The positive impact of the indicator of economic stability (proxied by ICRG) on saving rates contrasts with the results typically found in the literature. Generally, as uncertainty increases households and firms tend to save more for precautionary reasons, and vice versa. However, in the case of Indonesia, the result is the opposite. It is possible that if agents lost confidence in the banking sector and financial markets as saving channels during the 1998 crises, subsequent reductions in economic uncertainty led to an increase in confidence in the banking and financial system, and therefore to increased saving rates. …favorable Commodity price gains over the medium horizon support savings-formation. Two- commodity prices… year averages of commodity price gains have a positive impact on the structural component of saving, consistent with the gains being perceived as temporary (as discussed in the current account determinants section). …and a decline in Structural saving is positively affected by a reduction in the proportion of the the dependency ratio dependent population. The three-year average dependency ratio has the expected negative sign, in line with standard life-cycle models. As the ratio of dependents to the working population declines, saving will tend to increase. b. Cyclical component of saving Reduced by credit Greater access to credit reduces saving over the cycle, by softening household and expansions, and corporate liquidity constraints. Equity returns and the growth of real GDP per capita, boosted by GDP conversely, have a positive impact on the cyclical component of saving through growth and stock income effects and increases in profitability, although the latter variable is not market returns… statistically significant. …and reduced by Commodity price gains lead to lower cyclical saving, suggesting feedback from higher short term income from commodities to increases in consumption. The estimated effects of commodity price commodity price changes on the cyclical and on the structural component of saving gains go in opposite directions. It is possible that if agents face liquidity constraints, then even if they perceive that the revenues originating in a commodity price boom will be temporary, they may consume a portion of it in the short run, to bring their consumption levels closer to the desired level. 3. Structural and cyclical components of the current account balance The structural The reduction in the current account balance, particularly since 2011, reflects a component of the deterioration of both cyclical and structural factors, suggesting a structural current current account account balance of almost -1 percent of GDP, with a 90 percent confidence interval balance is estimated of ±2 percentage points of GDP (i.e. between -3 and 1 percent of GDP). As Figure to be at -1 percent of 35 shows, the fall in the medium term component is driven by the strong increase THE WORLD BANK | BANK DUN IA 32 Indonesia Current Account Assessment GDP, ±2 percentage (catch-up) in investment after 2005, driven by improved growth prospects, points of GDP commodity price gains, and a slowdown in saving since 2009. Further, the strong cyclical expansion of investment and the recent slowdown in saving contributed significantly to the worsening of the cyclical component of the current account. Credit expansion Credit expansions lead to both a decrease in the cyclical component of saving and an decreased increase in the cyclical component of investment, thus leading to a higher current (increased) cyclical account deficit in the short-run. This result is consistent with the findings on current savings (investment) account determinants reported in Section 2. The effect of The effects of commodity prices on the current account balance are more complex. commodity prices Commodity price increases lead directly to increases in the trade balance and thus on gains differs in the the current account balance. Their effects on investment and saving differ when we long and the short focus on the short run (cyclical components) or long run (structural component). In run the short run, commodity price gains lead to both a decrease in the cyclical component of saving and an increase in the cyclical component of investment, thus leading to a short run increase in the current account deficit. Over the long run, commodity price increases induce an increase in saving that predominates over the induced positive effect on investment, leading to an overall positive impact on the current account. Increased real The positive impact of real income growth on saving suggests that the positive impact income affects the of relative income on the current account balance is working through increased current account via saving. The negative association between lagged FDI inflows and the current account saving, while FDI described in Section 2 is likely associated with the fact that FDI inflows tend to crowd- crowds in domestic in further domestic investment, thus increasing the foreign financing needs of the investment economy and increasing the current account deficit, at least in the short-run. Figure 35: Components of current account balance (percentage of GDP) 90% CI 10 Structural component of current account Cyclical component of current account 8 Current Account 6 4 2 0 -2 -4 -6 -8 1994 Q2 1996 Q2 1998 Q2 2000 Q2 2002 Q2 2004 Q2 2006 Q2 2008 Q2 2010 Q2 2012 Q2 2014Q2 Source: Authors’ estimates THE WORLD BANK | BANK DUN IA 33 Indonesia Current Account Assessment E. How far can the Current Account and the Net International Investment Position (NIIP) deviate from their path? This section uses the estimates described above to simulate the deviations of the current account and NIIP from a baseline projection 41, which can be thought of as no major change, under three alternative scenarios: a “benign” world, an “adverse” world, and the case in which the government embarks in a large infrastructure investment program. Box 7: Simulating Indonesia’s current account balance and its net international investment position The current account balance and NIIP are simulated over the medium term using projections of the determinants of the current account from the IMF”s World Economic Outlook (WEO), and the World Bank Indonesia country office, together with assumptions on rates of return and relative valuation effects of foreign assets and liabilities. This simulation is akin to a static “what if” analysis rather than a forecast and it should not be read as such. First, estimated parameters from the model are kept constant over a relatively long horizon, while a more sophisticated, general equilibrium analysis might allow these to change. Second, the underlying model is descriptive in nature, hence built to provide an accurate representation of the determinants of the current account balance, rather than to minimize forecast errors. Third, projections of some key determinants, such as GDP and commodity prices, have continued to shift since the time that this report was prepared (in late 2014). The methodology applied does not impose steady-state assumptions on the economy as these do not typically hold for emerging countries (Cusolito and Nedeljkovic (2013)), but rather focuses on the reasoning that the current account deficit may be sustained as long as international investors are willing to finance it. This depends, ultimately, on the accumulated level of the NIIP. The methodology consists of following steps: 1. Simulations of the current account balance are generated from 2014 onwards using the coefficients from the current account determinants equation and the projected values of determinants. 2. The NIIP stock is calculated using the current account projections described above, the stock of foreign assets and liabilities in the previous year, projections of capital flows and calibrated rates of return on foreign equity assets and liabilities. 3. The steps above are repeated recursively to generate the paths of the current account and the NIIP until 2019, and the analysis is repeated for different scenarios. 41 A brief description of the simulation methodology is provided in Box 6 and a more detailed discussion is relegated to the Appendix. THE WORLD BANK | BANK DUN IA 34 Indonesia Current Account Assessment Figure 36: Sensitivity of NIIP simulation to different No capital gains on reserves, other investment assets and rates of return relative to baseline liabilities, portfolio debt assets and liabilities and the debt component of FDI are assumed, since their returns enter 2014 2015 2016 2017 2018 2019 the current account directly. The growth rate of prices on 0 portfolio equity liabilities and assets is set at 5.5 percent and 7 percent respectively. This assumption is based on -1 Liabilities: 5.5 the past 5-year average total return on the MSCI percent, Indonesia and MSCI World indices, adjusted for nominal Assets: 5.5 -2 GDP growth. percent Liabilities: As the projected rate of returns to assets and liabilities are -3 6.5 percent, tilted in favor of Indonesia it is important to gauge to Liabilites: 7 Assets: 5.5 percent what extent this assumption drives the results. The -4 percent, assumption has little impact on the dynamics of the Assets: 5.5 Liabilities: current account, changing it by a maximum of 0.1 percent 7 percent, -5 percent of GDP per year for a 1.5 percentage point Assets: 7 increase in liability returns or a 1.5 percentage point -6 decrease in asset returns. The NIIP dynamics on the other hand are moderately sensitive to the net returns. Source: World Bank staff calculations Figure 36 shows the pattern of the simulated NIIP under alternative assumptions on the rates of returns on assets and liabilities. Assuming an equal rate of return on assets and liabilities, leads to a more negative NIIP by at most 0.7 percent of GDP per year. Finally, The NIIP dynamics are mechanically sensitive to the assumed rate of return on the equity liabilities, increasing the annual rate of return by 100 basis points worsens the NIIP position by a cumulative 4 percentage points of GDP by 2019. The growth rate of FDI prices is calibrated following Tille (2008), assuming a nominal interest rate of 7.5 percent on liabilities and 5 percent on assets; and the ratio of reinvested to distributed earnings on FDI to be equal to 0.5, in line with the average yield on long-run government bonds and the observed income outflows, respectively. The nominal GDP growth for the 2014-2019 is consistent with the World Bank Indonesia country office Medium Term Fiscal Framework (MTFF) allowing for (realistic) possibility that the economy has not reached the steady-state and hence a constant nominal GDP growth rate. All additional projections and sources are reported in Table 6 below. Table 6: Projections and sources for relevant variables Unit 2014 2015 2016 2017 2018 2019 Source Real GDP growth pc 5.40 5.60 6.00 6.00 6.00 6.00 WEO Nominal GDP growth pc -0.88 10.39 11.09 5.24 8.67 8.72 WBIDN Nominal credit growth pc 17.00 15.00 15.00 15.00 15.00 15.00 WBIDN Government expenditure pp 0.20 0.20 0.20 0.20 0.20 0.20 WBIDN (change) Dependency ratio p 0.529 0.526 0.524 0.521 0.519 0.516 Authors Real GDP growth of major pc 3.75 3.80 3.80 3.80 3.80 3.80 WBIDN trading partners* REER appreciation pc -2.00 -2.00 1.00 1.00 1.00 1.00 WBIDN Crude oil price USD 96.2 53.2 56.3 60.8 65 69.4 WB Non-fuel commodity prices pc -3.54 -4.01 -1.00 -0.45 -0.48 -0.65 WEO Terms of trade pc -0.88 -0.01 -2.50 -1.32 -1.57 -0.74 WEO Relative trade openness pGDP 0.88 0.82 -0.04 -0.06 -0.12 -0.15 WEO (change) Relative income p 16.2 16.6 17 17.4 17.8 18.2 WEO/WBIDN Gross FDI inflows pGDP 2.12 2.10 2.03 2.05 1.98 2.00 WEO VxO index 13.30 14.64 15.98 17.32 18.66 20.00 Authors * These are: Australia, China, Germany, India, Italy, Japan, Korea, Malaysia, Philippines, Singapore, Taiwan, Thailand, US Note: pc = percentage change, p = percent, pGDP= percent of GDP. WEO = IMF World Economic Outlook, WB = World Bank commodity outlook, WBIDN = World Bank Indonesia country office Source: Authors’ elaboration THE WORLD BANK | BANK DUN IA 35 Indonesia Current Account Assessment The current account Under the scenario of benign external conditions, oil prices continue declining relative balance and NIIP to the World Bank’s commodity outlook projections, falling an additional 20 percent. are simulated under Non-oil commodity prices, on the other hand, recover somewhat from the recent a “benign” scenario decline, gaining 15 percent. Indonesia’s terms of trade thus improve consistently and the global risk indicator (VxO) falls by 15 percent relative to the baseline, indicating favorable global risk appetite and financing conditions. …an “adverse” The adverse world scenario is designed as the opposite of the benign scenario scenario… described above. In this world, oil prices recover faster than anticipated, non-oil commodity prices decline faster than anticipated, Indonesia’s terms of trade deteriorate and the global risk indicator spikes by 15 percent relative to baseline, along the lines of the increase in international market volatility experienced during the mid- 2013 mini-crisis. … and a “policy The policy push scenario keeps external conditions unchanged relative to the baseline push” scenario and simulates the impact of a large increase in infrastructure investment from the government. Under this scenario, infrastructure investment increases, generating a 0.5 percentage point increase in GDP growth. Increased domestic demand spurs credit growth, which increases by 1 percentage point relative to baseline. Increased investment (and an improved capital stock) attracts additional FDI, which increases by 0.7 percentage points, corresponding to an additional USD 2 billion per quarter; the import content of FDI increases trade openness by a total of 0.4 percent of GDP. The three scenarios are summarized in Table 7 below and the evolution of external conditions for the first two scenarios are depicted in Figure 36 to Figure 39. Table 7: Simulation scenarios Variable Benign world Adverse world Policy push: Oil price 20 percent lower 20 percent higher -- Non-oil commodities 15 percent higher 15 percent higher -- Terms of trade Stronger* Weaker* -- VxO 15 percent lower 15 percent higher, -- Credit growth -- -- 1pp higher Relative GDP growth -- -- 0.5pp higher FDI in percent of GDP -- -- 0.7pp higher Relative openness -- -- 0.4 percent of GDP Note: * as implied by relative prices of crude oil and non-oil commodities. Values relative to baseline. Figure 37: Simulation crude oil price Figure 38: Simulation: non-oil commodity prices (index: 2013 = 100) (index: 2013 = 100) Benign scenario Adverse scenario Benign scenario Adverse scenario Baseline scenario Baseline scenario 120 105 100 100 80 95 60 90 40 20 85 0 80 2013 2014 2015 2016 2017 2018 2019 2013 2014 2015 2016 2017 2018 2019 Source: World Bank staff calculations Source: World Bank staff calculations THE WORLD BANK | BANK DUN IA 36 Indonesia Current Account Assessment Figure 39: Simulation ViX/VxO Figure 40: Simulation terms of trade (index: 2013 = 100) (index: 2013 = 100) Benign scenario Adverse scenario Benign scenario Adverse scenario Baseine scenario Baseline scenario 120 105 110 100 100 90 95 80 70 90 2013 2014 2015 2016 2017 2018 2019 2013 2014 2015 2016 2017 2018 2019 Source: World Bank staff calculations Source: World Bank staff calculations Under the baseline, In the baseline scenario, detailed underlying projections for which are provided in the current account Box 7, the current account deficit is expected to narrow over the medium run and, in balance is expected the absence of additional policy actions, to stabilize around -2.5 percent of GDP. This to remain mildly is well within the confidence interval of the structural estimate presented in the negative previous section. While consistent with other projections of the structural current account 42, it must be stressed that these are “raw” model estimates from a descriptive model that is not updated or optimized for forward projections. What is most informative is deviations that different scenarios generate from the projected baseline, to which we now turn. In a benign scenario Under the benign scenario, the current account balance could increase quickly by the current account about 0.6 percent of GDP relative to the baseline and decline slowly thereafter, balance and NIIP settling at approximately 0.4 percent of GDP above the baseline (i.e. a narrower increase vs. the current account deficit than projected in the baseline). A higher current account baseline balance than in the baseline would mechanically lead to a lower stock of net foreign liabilities by approximately 2.2 percent of GDP relative to baseline by 2019. In an adverse Under the adverse scenario, the current account balance could decrease slowly relative scenario, conversely, to the baseline, settling 0.4 percent of GDP below the baseline (i.e. a larger current they could decrease account deficit than projected in the baseline). A lower current account balance than to still sustainable in the baseline would lead to a larger stock of foreign liabilities, by approximately 1.7 levels… percent of GDP by 2019. Despite the relatively significant deterioration in the external conditions envisaged by this scenario, the pattern of the current account and NIIP is not radically different. Under the “policy Under the “policy push” scenario, in which the government, embarks successfully on push” scenario, the a large infrastructure development program, the current account balance would current account decrease in the near term under the push of stronger domestic demand and imports, balance would and increase gradually thereafter, reaping the benefits from increased trade integration decrease first, then and export competitiveness. Under this scenario, the stock of foreign liabilities would recover slowly, but be higher by 2.9 percent relative to baseline, but possibly reduce in the future on the increased potential back of a continued increase in the current account balance. In this scenario, investor 42 See in particular the IMF’s external balance assessment for Indonesia THE WORLD BANK | BANK DUN IA 37 Indonesia Current Account Assessment future returns are perceptions of the lower current account balance and increased stock of foreign likely to attract liabilities would likely be quite different from that of the “adverse” scenario, as heavy foreign financing infrastructure investment would be expected to increase the productive capacity (and hence returns) to Indonesia’s economy, supporting foreign investment. Figure 41: Simulated current account relative to Figure 42: Simulated NIIP position relative to baseline baseline Benign scenario Benign scenario Adverse scenario Adverse scenario Policy push Policy push 0.80 3.00 0.60 2.00 0.40 1.00 0.20 0.00 0.00 -0.20 -1.00 -0.40 -2.00 -0.60 -0.80 -3.00 2014 2015 2016 2017 2018 2019 2014 2015 2016 2017 2018 2019 Source: World Bank staff calculations Source: World Bank staff calculations Box 8: Is the level and the composition of Indonesia’s net foreign liabilities problematic? Net foreign liabilities (or analogously, the net international Figure 43: Links between current account deficits, investment position [NIIP]) of a country link today’s credit ratings and net foreign liabilities current account deficit to the deficit that future generations will face. A present current account deficit builds into the stock of net foreign liabilities, generating a stream of debt servicing and amortization payments that will fall on future generations. The accumulation of a large and negative NIIP can pose a risk to the sustainability of a large and persistent current account deficit. Accumulated foreign liabilities lead directly to a worsening in future current account balances through larger income outflows while increasing the perception of country risk, reducing the attractiveness to foreign investors. In a recent cross-country study of external crises over the period 1970-2011, Catao and Note: Bar chart labels refer to export volumes in million tons Milesi-Ferretti (2013) show that the risk of a balance of (MT) Source: Indonesian Ministry of Energy and Mineral Resources; payments crisis is related to the stock of net foreign World Bank staff calculations liabilities, and particularly so when these exceed 50 percent of GDP or their historical average by more than 20 percent. Equity liabilities have historically been less crisis-prone. Liabilities can take the form of debt or equity, and the prevalence of one type of liability affects the probability of a balance of payments crisis. Debt liabilities have tended to introduce currency mismatches (in cases where most debt is foreign currency-denominated) or maturity mismatches (in cases where a large part of the debt is short term). Equity liabilities, being national-currency-denominated residual claims are less likely to introduce such risks (Fernandez Arias and Hausmann (2001)). Increasing Within equity, FDI is commonly seen as the most desirable form of foreign liability. FDI is widely acknowledged to be related to technological and knowledge transfers, increased competition, and job creation and upgrading. In THE WORLD BANK | BANK DUN IA 38 Indonesia Current Account Assessment addition, FDI is a less-outflow-prone source of foreign financing compared to portfolio investment. This is because FDI involves a long-term commitment. 43 FDI is also less subject to capital reversals and contagion that affect other flows, since the presence of large, fixed, illiquid assets makes rapid disinvestments more difficult than the sale of stock holdings (Fernandez Arias and Hausmann (2001)). Empirical tests find supporting evidence that the composition of external liabilities affects the risk of external crises for developing countries; in particular, non-FDI liabilities are more crisis prone then FDI. In a recent study, Catao and Milesi Ferretti (2013) find that the composition of the NIIP matters for vulnerability. Increased debt liabilities are associated with higher risks of external crises, the results for non-FDI equity liabilities are weaker, and when foreign liabilities take the form of FDI, increases in these liabilities are not associated with increases in the risk of external crises. Indonesia’s net foreign liabilities increased from 41.15 percent of GDP in 2012 to 42.67 percent of GDP in 2013. This increase is the result of a current account deficit of 2.8 percent of GDP, and of valuation gains of net foreign assets relative to GDP, and will imply increases in the associated debt servicing, interest payments and profit repatriation in years to come. Comparatively, and as a share of GDP, Indonesia’s net foreign liabilities are the second highest, after Turkey, among the group of the “fragile five” (Figure 44). Figure 44: Net foreign assets as percentage of GDP for Figure 45: Equity and debt structure of Indonesia’s net the fragile five (2012 and 2013) foreign liabilities (NIIP as percentage of GDP) South 2010 2011 2012 2013 ** Brazil Indonesia India Turkey Africa 0% 0% -10% Percentage of Debt vs Equity Net -20% -10% -30% Foreign Liabilities -20% -40% -50% -30% -60% -70% -40% -80% -50% -90% 2012 2013 -100% Debt Equity -60% Source: World Bank staff calculations based on Central Bank data of Source: World Bank staff calculations based on Central Bank data of respective countries respective countries 43 This “bolted down” hypothesis has been disputed. Hausmann and Fernandez Arias (2000) argue that FDI is not a physical asset of a firm, but only one of its liabilities. The firm has other liabilities and assets, and some assets may be “bolted down”, but “financial claims backed by “bolted down” physical capital can easily fly away” (page 36). Foreign investors could hedge their earnings and protect the value of their assets, or speculate, by borrowing in domestic currency and pledging physical capital as collateral. Put differently, outflows may be generated under an account other than FDI. The volatility of FDI thus provides limited information on the overall volatility of the financial account. THE WORLD BANK | BANK DUN IA 39 Indonesia Current Account Assessment About 60 percent of Indonesia’s net liabilities are Figure 46: Gross foreign liabilities as percentage of accounted for by equity-based instruments. This GDP for the fragile five (2013) proportion has not changed substantially over the last 120% four years (Figure 45). Figure 46 focuses instead on Other Investment gross liabilities, showing the different subcomponents Gross Internat. Liabilities as 100% Portfolio Investment as a percentage of GDP for 2013, and comparing with 80% Direct Investment percent of GDP the other “fragile five” countries. 60% In Indonesia, FDI liabilities account for a sizable share 40% of gross liabilities. 45 percent of gross foreign liabilities are FDI-related, portfolio investment accounts for 29 20% percent, while the remaining 26 percent is accounted 0% for by other investment-related liabilities. Among the fragile five, the FDI share in Indonesia is only lower than that in Brazil (48 percent), but higher than that in India (31 percent), in Turkey (24 percent) and in South Africa (33 percent). Source: World Bank staff calculations based on Central Bank data of respective countries Source: Authors” elaboration based on Hausmann and Fernandez Arias (2000), Fernandez Arias and Hausmann (2001), and Catao and Milesi-Ferretti (2013). THE WORLD BANK | BANK DUN IA 40 Indonesia Current Account Assessment References Albuquerque, R., N. Loayza, and L. Serven (2005). “World Market Integration through the Lens of Foreign Direct Investment”. Journal of International Economics, vol. 66, No. 2, pp. 267–295. Beidas-Strom, S., and P. Cashin (2011). “Are Middle Eastern Current Account Imbalances Excessive?”. IMF Working Paper WP/11/195. Blalock, G. and P. Gertler (2009). “How firm capabilities affect who benefits from foreign technology”. Journal of Development Economics, vol. 90, No. 2, pp. 192-199. Bluedorn, J. and D. Leigh (2011). "Revisiting the Twin Deficits Hypothesis: The Effect of Fiscal Consolidation on the Current Account". IMF Economic Review, Palgrave Macmillan, vol. 59(4), pages 582-602, November. Calderon, C., A. Chong, and N. Loayza (2002). “Determinants of Current Account Deficits in Developing Countries.” Contributions to Macroeconomics 2, Article 2. Calvo, G., A. Izquierdo, and L.Mejia (2008). “Systemic Sudden Stops: The Relevance of Balance-Sheet Effects and Financial Integration”. NBER Working Paper 14026. Calvo, G. A., A. Izquierdo and E. Talvi (2003). “Sudden Stops, the Real Exchange Rate, and Fiscal Sustainability: Argentina”s Lessons”. NBER Working Paper No. 9828. Calvo, G. A., A. Izquierdo and E. Talvi (2006). “Phoenix Miracles in Emerging Markets: Recovering without Credit from Systemic Financial Crises”. NBER Working Paper No. 12101 Calvo, G., L. Leiderman, and C. Reinhart (1993). “Capital Inflows and the Real Exchange Rate Appreciation in Latin America: The Role of External Factors”. IMF Staff Papers vol. 40, No. 1, pp. 108-151. Campa, J.M., and A. Gavilan (2011). “Current Accounts in the Euro Area: An Intertemporal Approach”. Journal of International Money and Finance 30, 205–228. Carroll, C.D and D.N. Weil (1994). “Saving and Growth: A Reinterpretation”. Carnegie-Rocherster Conference Series on Public Policy, vol. 40, No. 1, pp. 133-192. Cattaneo, O., G. Gereffi, S. Miroudot and D. Taglioni (2013). “Joining, Upgrading and Being Competitive in Global Value Chains”. Policy Research Working Paper No. 6406, World Bank. Catao, L. and G.M. Milesi-Ferretti (2013). “External Liabilities and Crises”. IMF Working Paper WP/13/113. International Monetary Fund, Washington, DC. Chinn, M. and E. Prasad (2003). “Medium-term Determinants of Current Accounts in Industrial and Developing Countries: an Empirical Exploration”. Journal of International Economics, Vol. 59, pp. 47-76. Chinn, M.D., and H. Ito (2007). “Current Account Balances, Financial Development and Institutions: Assaying the World”s Saving Glut”. Journal of International Money and Finance 26, 546-569. Chuhan, P., C. Claessens and N. Mamingi (1998). “Equity and Bond Flows to Latin America and Asia: The Role of Global and Country Factors”. Journal of Development Economics. Vol. 55, No. 2, pp. 439-463. THE WORLD BANK | BANK DUN IA 41 Indonesia Current Account Assessment Claessens, S., M.Dooley, and A. Warner (1995). “Portfolio Capital Flows: Hot or Cold?” World Bank Economic Review, vol. 9, No. 1, pp. 153-174. Craighead, W. D. and D. Hineline, (2013). "As the Current Account Turns: Disaggregating the Effects of Current Account Reversals in Industrial Countries". World economy, 36 (12), p. 1516. Cusolito, A. and M. Nedeljkovic (2013). “Toolkit for the Analysis of Current Account Imbalances”, International Trade Department, World Bank, DC. de Mello, L., P. C. Padoan and L. Rousová (2011). "The Growth Effects of Current Account Reversals: The Role of Macroeconomic Policies". OECD Economics Department Working Papers, No. 871, OECD Publishing. Debelle, G. and H. Faruque (1996). “What Determines the Current Account? A Cross-Sectional and Panel Approach”. IMF working Paper No. 58. IMF, Washington, DC. Edwards, S. (2004a) “Financial Openness, Sudden Stops and Current Account Reversals”. NBER Working Paper No. 10277. Edwards, S. (2004b). “Thirty Years of Current Account Imbalances, Current Account Reversals, and Sudden Stops”. NBER Working Paper No. 10276. Edwards, S. (2007). “Crises and Growth: A Latin American Perspective”. NBER Working Paper 13019. National Bureau of Economic Research, Cambridge, MA. Engel, C., and J. Rogers (2006). “The U.S. Current Account Deficit and the Expected Share of World Output”. Journal of Monetary Economics 53, 1063-1093. Fernandez-Arias, E. (1996). “The New Wave of Capital Inflows: Push or Pull?” Journal of Development Economics 48, 389-418. Fernandez-Arias, E. and R. Hausmann (2001). “Is foreign direct investment a safer form of financing?” Emerging Markets Review, Elsevier, vol. 2(1), pp 34-49. Forbes, K. and F. Warnock (2012). "Debt- and Equity-Led Capital Flow Episodes". NBER Working Papers 18329. Frankel, J. and A. Rose (1996). “Currency Crashes in Emerging Markets: An Empirical Treatment”. Journal of International Economics. Vol. 41, pp. 351-366. Fratzscher, M. (2012). “Capital Flows, Push versus Pull Factors and the Global Financial Crisis”. Journal of International Economics, vol. 88, No. 2, pp. 341-356. Froot, K., P. O”Connell, and M. Seasholes (2001). “The Portfolio Flows of International Investors”. Journal of Financial Economics, vol. 59, pp. 151-193. Freund, Caroline and Frank Warnock. “Current Account Deficits in Industrial Countries: The Bigger They Are, the Harder They Fall?”, In Richard H. Clarida, ed., “G7 Current Account Imbalances: Sustainability and Adjustment”. Chicago: University of Chicago Press, 2007. Ghosh, A., J. Kim, M. Qureshi, and J. Zalduendo (2012). “Surges”. IMF Working Papers 12/22. IMF, Washington, DC. THE WORLD BANK | BANK DUN IA 42 Indonesia Current Account Assessment Gourinchas, P.O, and M.Obstfeld (2012). “Stories of the Twentieth Century for the Twenty First”. American Economic Journal of Macroeconomics, Vol. 4, No. 1, pp. 226-265. Griffin, J., and J.S. Martin (2003). “Momentum Investing and Business Cycle Risk: Evidence from Pole to Pole”. The Journal of Finance, vol. 58, No. 6, pp. 2515-2547. Gruber, J.W., and S. B. Kamin (2007). “Explaining the Global Pattern of Current Account Imbalances”. Journal of International Money and Finance, 26, 500-522. Haddad, M., and M. Nedeljkovic (2012). “Turkey: Managing the Current Account in Period of Volatility”. World Bank note. Halac, M. and S. Schmukler (2004). “Distributional Effects of Crises: The Financial Channel”. Economia, Fall 2004. Hansen, B., and Racine, J. (2012). “Jackknife Model Averaging”. Journal of Econometrics 167, 38-46. Hausmann, R. and E. Fernandez Arias (2001). “Is Foreign Direct Investment a Safer From of Financing?” Emerging Markets Review, vol. 2, 1, pp. 34-48. Hausmann, R., Klinger and Lawrence (2007). “Examining Beneficiation”. Center for International Development, Kennedy School of Government, Harvard University. Jorda, O., M. Schularick and A.M.Taylor (2011). “Financial Crises, Credit booms, and External Imbalances: 140 Years of Lessons”. IMF Economic Review 59: 340-378. Kamin, S.B., J. Schindler and S. Samuel (2007). “The contribution of domestic and external factors to emerging market currency crises: an early warning systems approach”. International Journal of Finance and Economics, Vol. 12, No. 3, pp. 317-336. Kenward, L. (1999). “Assessing Vulnerability to Financial Crisis: Evidence from Indonesia”. Bulletin of Indonesian Economic Studies, vol. 35, No. 3, pp. 71-95. Komárek, Luboš and Melecky, Martin (2005). “Currency crises, current account reversals and growth: the compounded effect for emerging markets”. Working Paper. Coventry: University of Warwick, Department of Economics. (Warwick economic research papers). Kwiatkowski, D., P.C. S. Phillips, P. Schmidt, and Y. Shin (1992). “Testing the Null Hypothesis of Stationarity against the Alternative of a Unit Root” Journal of Econometrics, Vol. 54: 159-178. Levy Yeyati, E., S.Schmukler and N. Van Horen (2008). “Emerging market Liquidity and Crises”. Journal of the European Economic Association, vol. 6, No. 2, pp. 668-682. Milesi-Ferretti, G.M. and A. Razin (1996). “Sustainability of Persistent Current Account Deficits”. NBER Working Paper 5467. National Bureau of Economic Research, Cambridge, MA. Milesi-Ferretti, G. M. and Razin, A. (2000). “Current account reversals and currency crisis: empirical regularities”, in Krugman P. (ed.), Currency Crises, Chicago IL: University of Chicago Press, pp. 285–323. Moreno, R. (2008). “Experiences with Current Account Deficits in Southeast Asia”, in Current Account and External Financing, edited by K. Cowan, S. Edwards and R. Valdes. Obstfeld, M. and K. Rogoff (1996). “Foundations of International Macroeconomics”. Cambridge, MA: MIT Press. THE WORLD BANK | BANK DUN IA 43 Indonesia Current Account Assessment Otto, G. (1992). “Testing a Present-Value Model of the Current Account Evidence from U.S. and Canadian Times Series”. Journal of International Money and Finance 11, 414-430. Pesaran, M. H., and R. Smith (1995). “Estimating long-run relationships from dynamic heterogeneous panels”. Journal of Econometrics 68, 79-113. Prasad, E. (2014). “The Dollar Trap: How the US Dollar Tightened its Grip on Global Finance”. Princeton University Press. Reinhart, C. and Reinhart, V. (2009). “Capital Flow Bonanzas: An Encompassing View of the Past and Present”. In NBER International Seminar on Macroeconomics 2008, pp. 9-62. Chicago: University of Chicago Press. Robertson, D., and J. Symons (1992). “Some Strange Properties of Panel Data Estimators”. Journal of Applied Econometrics 7, 175-189. Rodrik, D. (2008). “The Real Exchange Rate and Economic Growth”. Brooking Papers on Economic Activity, 2, pp. 365-412. Rodrik, D. and A. Velasco (1999). “Short-Term Capital Flows”. NBER Working Paper 7364. Sachs, J. (1981). “The Current Account and Macroeconomic Adjustment in the 1970s”. Brooking Papers on Economic Activity 1: 201-268. Sarno, L. and M.P.Tayor (1999). “Hot Money, Accounting Labels and the Permanence of Capital Flows to Developing Countries: An Empirical Investigation”. Journal of Development Economics, 59(2): 337-364. Sheffrin, S.M., and W. T. Woo (1990). “Present Value Tests of an Intertemporal Model of the Current Account”. Journal of International Economics 29, 237-253. Tille, C. (2008). “Composition of International Assets and the Long-run Current Account”. Economic Notes 37, 283–313. Varela, G. (2011). “A framework to analyze the Impact of Exchange Rate Uncertainty on Output Decisions”. Sussex Working Paper Series. World Bank (July, 2011). “Indonesia Economic Quarterly: 2008 Again?”. World Bank (July, 2012). “Indonesia Economic Quarterly: Rising to Present and Future Challenges”. World Bank (October, 2013). “Indonesia Economic Quarterly: Continuing Adjustment”. World Bank (December, 2013). “Indonesia Economic Quarterly: Slower Growth, High Risks”. World Bank (March, 2014). “Indonesia Economic Quarterly: Investment in Flux”. THE WORLD BANK | BANK DUN IA 44 Indonesia Current Account Assessment APPENDIX I Conceptual framework and methodology for the estimation of current account determinants By focusing on the underlying determinants of saving and investments, the specification used in this report is partially related to the inter-temporal approach to the current account (Sachs, 1981, Obstfeld and Rogoff, 1996). 44 Our specification below, however, constitutes an augmented version of these models. This is because its aim is to capture as many potential influences on the current account as possible, to disentangle the drivers of the deficit, rather than discriminating between the competing intertemporal models by choosing a limited number of variables in the reduced form that correspond to a particular model in empirical exercise. CAt = α + β1CAt −1 + β 2 NIIPt −1 + β 3GovExpt + β 4 FuelSubsidyBillt + β 5 NonFuelsCommPt + β 6 g Re lIncomet + + β 7 REERt + β 8 FinDeept + β 9 Opennesst + β10 Re lg GDPt + β11 FDI t + β12VXOt + + β13 DepRatiot + β14ToTt + β15 NIIP * CrisisD + β16 FuelSubsidyBill * Post 2002 D + ε t CAt = α + β1CAt −1 + β 2 NIIPt −1 + β 3GovExpt + β 4Oil Pr icest + β 5 NonFuelsCommPt + β 6 g Re lIncomet + + β 7 REERt + β 8 FinDeept + β 9Opennesst + β10 Re lg GDPt + β11 FDI t + β12VXOt + + β13 DepRatiot + β14ToTt + β15 NIIP * CrisisD + β16Oil Pr ices * Post 2002 D + ε t Where current accountt is the current account balance in t, NIIPt-1 is the net international investment position in t-1, GovExp are government expenditures (as a share of GDP), FuelSubsidyBill is the cost of fuel subsidies as a share of GDP, NonFuelsCommP is the weighted average of non-fuel commodity prices (in changes), RelIncome is GDP per capita in PPP prices, expressed in deviations from the average of Indonesia’s main trading partners, REER is the real exchange rate (in changes), FinDeep is a proxy of financial deepening (credit to the private sector as a share of GDP, in changes), Openness is the ratio of trade/GDP in deviations from the average of Indonesia’s main trading partners, RelgGDP is relative GDP growth of Indonesia with respect to that of its main trading partners, FDI are foreign direct investment inflows (as a share of GDP), VXO is an index of global uncertainty, DepRatio is the dependency ratio, ToT are the terms of trade (in changes), NIIP*CrisisD is the interaction of the NIIP with a Crisis Dummy for 1998, FuelSubsidyBill*Post2002D is the interaction of the cost of fuel subsidies as a share of GDP and a Dummy variable that takes value one from 2002 onwards, and epsilon is an error term. An alternative specification uses oil prices instead of the fuel subsidy bill. These two models are also estimated on the trade balance, instead of on the current account balance. Results are reported in Table 8. 44 The inter-temporal models treat current account balances as an outcome of consumption and investment decisions made over a long-term horizon under forward looking expectations. More concretely, with an infinitively lived representative agent who smoothens consumption by lending or borrowing abroad, the standard inter-temporal model implies that the movements in the current account should reflect factors that affect country’s underlying saving and investment positions. Assuming free capital movements, the current account will absorb any temporary shocks to the net national cash flow thus allowing domestic agents to smooth their consumption over time. Although theoretically rigorous, inter-temporal models tend to exhibit a poor empirical fit – while the model- predicted and the actual series were usually positively correlated, the latter series were substantially more volatile (see, inter alia, Sheffrin and Woo (1990), Otto (1992) for early applications and Campa and Gavlan, 2011 for other references). In addition, with the focus on the long-run, the models are of limited applicability for assessing the current account sustainability, at least from the short to medium run perspective. THE WORLD BANK | BANK DUN IA 45 Indonesia Current Account Assessment The reduced-form specification is related to a large empirical literature, which uses various econometric techniques to identify the relationships between the current account and a set of macro and socio-economic variables (see Debelle and Faruqee, 1996, Calderon, et al., 2002, Chinn and Prasad, 2003, for early applications in this expanding literature). The aim behind the specification (2) is to capture as many potential influences on the current account as possible using a large number of variables in the empirical analysis. 45 Since annual data are used, estimation of the single-country determinants puts relatively strong limits on the number of variables to be included in standard OLS regression, even in the first step of general-to- specific exercise. Along the lines of Haddad and Nedeljkovic (2012), a preferable solution for confronting parametric model uncertainty and limited number of observations is model averaging. Model averaging in the present context means that the different combinations of the potential current account determinants (not all at once) are used as particular models and each combination is estimated by OLS. The final estimate is obtained by averaging across all the estimated regressions using a suitable criterion to select individual model’s weights. We use the recently proposed Jackknife Model Averaging (JMA) estimator for non-nested and heteroscedastic models (Hansen and Racine, 2012) where the weights are chosen by minimizing a leave- one-out cross-validation criterion (see Appendix for a detailed description). In this way, a large number of variables can be included, whilst, at the same time, only single country estimation is performed (thus eliminating a heterogeneity-induced bias in the resulting parameter estimates of panel studies, Robertson and Symons, 1992, Pesaran and Smith, 1995). Methodology for the estimation of dynamics and determinants of saving and investment Quarterly data over 1992Q1-2014Q2 period are collected from various sources. The sample size and the choice of variables above reflect the data availability. Yt = Yts +Ytc + αDt + εt Further, in order to identify the individual components of the flows we use additional information embedded in the structural and cyclical determinants of the saving and investment and specify the following law of motion for the components: −1 + β X t + ε t Yt C = θYt C C C Yt S = ϕYt −1 + δX t + ε t S S S where , ɛ are structural and cyclical shocks which are assumed to be zero mean, normally distributed, serially and cross-equation uncorrelated. and , denote variables which affect the structural and the cyclical component of the variable Y, respectively. The components of the flows and the parameters are estimated jointly in the extended state space specification. Kalman filter with diffuse initialization is used for estimation of the state space. Application of the Kalman filter requires further assumptions on some 45 By focusing on the underlying determinants of saving and investments the specification is partially related to the intertemporal approach to the current account (Sachs, 1981, Obstfeld and Rogoff, 1996). The intertemporal model treats current account as an outcome of consumption and investment decisions made over a long-term horizon under forward looking expectations. More concretely, with an infinitely lived representative agent who smooths consumption by lending or borrowing abroad, the standard intertemporal model implies that the movements in the current account reflect factors that affect country’s underlying saving and investment positions. Assuming free capital movements, the current account will absorb any temporary shocks to the net national cash flow thus allowing domestic agents to smooth their consumption over time. Although theoretically rigorous, intertemporal models tend to exhibit a poor empirical fit – while the model-predicted and the actual series were usually positively correlated, the latter series were substantially more volatile (see, inter alia, Sheffrin and Woo (1990), Otto (1992) for early applications and Campa and Gavlan, 2011 for other references). In addition, with the focus on the long- run, the models are of limited applicability for assessing the current account sustainability, at least from the short to medium run perspective. Moreover, rather than choosing a limited number of variables in the reduced form that correspond to a particular model, in the empirical exercise the focus is on a broad set of potential determinants. THE WORLD BANK | BANK DUN IA 46 Indonesia Current Account Assessment parameters of the model. Following the literature on estimation of the natural rate of unemployment, we assume two “signal to noise ratio” – ratio of the variance of the structural component to the variance of irregular component – of 0.15 and ratio of the variance of the temporary or cyclical component to the variance of irregular component – of 0.4. 46 Alternative specifications of the ratio do not change the results. Due to a relatively small length of the sample and a large number of potential determinants different specifications were estimated and the final specification is the one which minimizes the AIC and BIC criterion. Methodology for the external sustainability analysis Here we ask, given the identified influences of macro-variables on the current account and the current stock (end-2013) of foreign assets and liabilities of the country, what will be the future paths of NIIP under different scenarios? This methodology does not impose any steady-state assumption on the evolution of the economy—which might not hold for an emerging country; it represents the other side of the sustainability coin—the current account may be sustainable as long as the foreigners are willing to finance it, which is ultimately connected to the accumulated level of NIIP. Moreover, given the experience from the recent crises, market perceptions of what constitutes a sustainable level of NIIP (often proxied by external debt) may change in the short-run, which signals the importance of focusing on the front part of the horizon. The procedure consists of four steps. First, we use the estimated coefficients from the current account determinants analysis and the projected values of determinants to generate projections of the current account balance at the end of 2014. Given: (i) the current account projection from step 1, (ii) the stock of foreign assets and liabilities in 2013, (iii) projections of the relative movements in capital inflows and outflows and calibrated rates of return on foreign equity assets and liabilities for 2014; generate the stock of the NIIP at the end of 2014. Then, using the calculated stock of NIIP at the end of 2014, steps 1-2 are recursively repeated generating the paths of the current account and the NIIP until 2018. Different projections of the selected variables are alternated to obtain the range of potential scenarios and assess the risk implications. The composition of the net foreign position of Indonesia matters for this analysis and it is explicitly taken into account by introducing the valuation gains associated with that composition, which involves introducing the rates of return on foreign equity assets and liabilities in each year (Tille, 2008). Incorporating valuation effects is key. The case of the USA provides a concrete example. The USA has a big external balance-sheet, with its foreign liabilities exceeding its overseas assets. Although this may seem worrisome, this statement conceals the fact that while USA”s foreign assets are unusually lucrative; its liabilities are liquid, safe and low-yielding. 47 Over time, these return differentials will contribute to a self-correction of the net foreign liabilities of the USA. A simple, abstract example illustrates the importance of net foreign liabilities” composition and valuation effects. Consider a country with the balanced trade and constant GDP. Suppose the country holds equity assets amounting to 100 percent of GDP, has bond liabilities of the equal relative amount and let both types of assets earn the same annual return of 5 percent of GDP. The current account deficit of 5 percent of GDP then does not deteriorate the NIIP position of the country. The key idea here is the difference in recording and valuation of returns on different types of assets. While the returns on bonds take the form of the interest payments, which enter the current account, the gains on equity do not enter the current account, but offset the interest outflows in the NIIP. Similarly, a country that is in the growing phase may see its external position deteriorating more than what the accumulated current account deficits show due to equity price inflation relative to its debtors. In Indonesia, at the end of 2013 FDI equities constitute a sizable portion of total liabilities in its net international investment 46 The ratio allows imposing a constraint on the smoothness of the component. Lower values of the ratio imply smaller volatility of the components. 47 Prasad, E., 2014, “The Dollar Trap”. THE WORLD BANK | BANK DUN IA 47 Indonesia Current Account Assessment position (while the rest is accounted by debt). This fact highlights the importance of taking into account the composition of the net foreign position and differences in returns in evaluation of the external position. Table 8: Dependent Variable is the Trade Balance Variable Coefficient Standard error Coefficient Standard error Credit in percent of GDP (change) -0.03122 0.02202 -0.0242 0.0241 3 year average REER (ln change) -0.16301 0.02580 -0.1758 0.0325 WEO non-fuel commodity price index (ln 0.04969 0.02212 0.0508 0.0309 change) Relative GDP growth -0.18453 0.06457 -0.1478 0.0841 Dependency ratio (5 year average) -0.01220 0.00512 -0.0106 0.0054 FDI in percent of GDP (3 year average) -0.67039 0.06623 -0.5416 0.0364 Relative openness 0.10917 0.00563 0.1117 0.0050 Relative income 0.0209 0.0035 0 0 VXO index (ln change) 0.0032 0.0017 0.0036 0.0023 Government expenditure in percent of GDP 0 0 0 0 (change) 3 year average terms of trade (ln change) 0.10640 0.00458 0.1019 0.0080 Lagged trade balance 0.66092 0.05533 0.7953 0.0575 NIIP in percent of GDP (beginning of period) 0 0 0 0 NIIP*Crises dummy 0 0 0 0 Energy subsidy in percent of GDP (change) * -0.4650 0.1777 Post 2002 dummy Oil prices (change) * Post 2002 dummy -0.0088 0.0117 Energy subsidy in percent of GDP (change) 0.0837 0.1190 -0.0120 0.0053 Oil prices (change) THE WORLD BANK | BANK DUN IA 48