Indonesia’s external borrowing spreads increased by more than 1000 bps from mid-2007 to late 2008, before subsiding in recent months. The large increase in spreads prompted questions about whether the spreads adequately reflect the improvements in fundamentals made over the past few years. This Selected Issues paper examines the determinants of Indonesia’s spreads, and finds that fundamentals can explain both the level of and the increase in spreads. It uses a cross-country panel regression model of emerging market sovereign spreads to yield valuable insights into the pattern of Indonesia’s sovereign spreads.

Abstract

Indonesia’s external borrowing spreads increased by more than 1000 bps from mid-2007 to late 2008, before subsiding in recent months. The large increase in spreads prompted questions about whether the spreads adequately reflect the improvements in fundamentals made over the past few years. This Selected Issues paper examines the determinants of Indonesia’s spreads, and finds that fundamentals can explain both the level of and the increase in spreads. It uses a cross-country panel regression model of emerging market sovereign spreads to yield valuable insights into the pattern of Indonesia’s sovereign spreads.

I. Explaining Indonesia’S Sovereign Spreads1

Indonesia’s external borrowing spreads increased by over 1,000 bps from mid 2007 to late 2008, before subsiding in recent months. The large increase in spreads, especially compared with Indonesia’s peers, has prompted questions about whether the spreads adequately reflect the improvements in fundamentals made over the past few years. This paper examines the determinants of Indonesia’s spreads, and finds that fundamentals can explain both the level of and the increase in spreads. Spreads could be lowered through further improvements in the policy framework, such as lowering the level and volatility of inflation and increasing the buffers to meet external financing needs.

A. Introduction

1. Indonesia’s external borrowing costs increased sharply during the global financial crisis in late 2008. The EMBI spread and the CDS spreads co-moved closely, falling from 2004 to reach their lowest levels in mid 2007. With the onset of the global crisis in mid 2008, global risk aversion rose, and Indonesia’s spreads began to rise. In late 2008, coinciding with the turmoil in international financial markets, Indonesia’s spreads increased to nearly 1,200 bps, over 1,000 bps higher than the pre-crisis levels and higher than its peers.

Figure I.1.
Figure I.1.

EMBI Global Spreads

(In basis points)

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

Figure I.2.
Figure I.2.

CDS Spreads

(In basis points)

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

Source: Bloomberg L.P.

2. This increase in spreads comes against the backdrop of generally improved domestic fundamentals in recent years. Over the past five years, public and external debt ratios have been halved to about 30 percent of GDP, primary fiscal surpluses have averaged 1½ percent of GDP per year, the external current account has been in surplus, and the financial sector has been strengthened with relatively large capital buffers, low nonperforming loans and improved provisioning. These improvements contributed to a decline in Indonesia’s spreads through mid 2007, along with those of other emerging markets, as global financial markets were benign.

3. The improved fundamentals and declining spreads—in line with other emerging markets prior to the crisis—raises the question as to why Indonesia’s spreads have underperformed in the most recent period. In particular, is the widening of Indonesia’s spreads warranted by the fundamentals? And what, if anything, can be done to lower the spreads?

4. To answer these questions, a cross-country panel regression model of emerging market sovereign spreads is used. This model was developed recently at the IMF by Hartelius (2006) and Hartelius, Kashiwase, and Kodres (2008) and explains much of the movements in spreads across countries. The next section describes the model and its application to Indonesia and some comparator countries. Then, factors outside the model that potentially impact spreads are discussed, and policy conclusions are drawn.

B. Applying the Spreads Model to Indonesia

5. The spreads model incorporates both external financial factors as well as domestic macroeconomic, financial and political variables. The cross-country panel is based on 33 emerging markets, using monthly data spanning January 1998 to January 2009, with fixed effects. The dependant variable is the logarithm of the country’s EMBI spread.2

  • The external financial factors are: (i) VIX index (as a measure of global risk aversion), (ii) yield on 3-month Fed funds futures (as a measure of the U.S. monetary policy stance and international liquidity), and (iii) volatility of the Fed funds futures market (to measure the extent of uncertainty about future U.S. monetary policy).

  • The domestic factors comprise macroeconomic, financial, and political risk ratings from the PRS Group’s International Country Risk Guide (2009), aimed at measuring the strength of fundamentals and, consequently, risks. The economic risk rating assesses current economic strengths and weaknesses, and includes real GDP growth, annual inflation, overall fiscal balance as a share of GDP and the external current account to GDP. The financial risk rating measures the ability to finance external obligations, and includes external debt as a share of GDP, international reserves coverage of imports, and a measure of exchange rate stability. The political variable measures political risk, such as government stability, socioeconomic conditions such as unemployment and consumer confidence, conflict, corruption, law and order, and quality of the bureaucracy.

6. The model fits Indonesia’s spreads well, tracing both the declining trend through mid-2007 as well as the subsequent upturn. From 2004 to mid 2007, benign external financial factors contributed about one-quarter of the reduction in spreads; improving domestic macroeconomic and financial factors contributed over 30 percent, while lower political risk also contributed about 30 percent to the reduction in spreads from over 400 bps in mid 2004 to a historic low of about 136 bps in June 2007. Subsequently, the deteriorating external financial environment contributed significantly to the increase in Indonesia’s spread. From January 2008 to January 2009, external factors accounted for over 50 percent of the increase. The remainder can be accounted for by some deterioration in economic and financial indicators—the rupiah depreciated sharply, quarterly growth decelerated, and the current account swung into deficit—accentuated by relatively weaker initial conditions.

Figure I.3.
Figure I.3.

Indonesia: EMBI Spreads

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

7. The close fit of the model suggests that there is no puzzle behind the large increase in Indonesia’s spreads. Rather, the increase reflects how the worsening of the external financial environment can severely impact Indonesia. Continued improvements in Indonesia’s policy framework—associated with lowering domestic macroeconomic, financial, and political risks—would help in further reducing spreads. Some of these factors are described in further detail below.

8. A decomposition of the different factors across countries sheds further insights. Consider countries that experienced relatively small increases in spreads during the market turmoil of 2008-09, which is the type of response that could be considered desirable. Unlike for Indonesia, where the increase from January 2008 to January 2009 was nearly 450 bps, China, Chile, Egypt, Malaysia, and Poland experienced increases in spreads of only 100–250 bps. What sets these countries apart from Indonesia is a smaller contribution of external factors, as well as smaller contributions of macroeconomic, financial, and political risks.3 These contributions are related to lower initial levels of spreads and to better risk indicators among the domestic risk factors.

Figure I.4.
Figure I.4.

Change in Spreads by Contributing Factors January 2008-January 2009

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

9. While Indonesia’s economic risk ratings improved, the increase was not as strong when compared to “top performing” countries or others that might be considered its peers. This relatively weaker performance along the economic indicators—growth, inflation, fiscal balance, and current account balance—can largely be attributed to a weak inflation performance. Indonesia’s inflation over the past 5 years has averaged close to 9 percent, and the volatility of inflation has also been relatively high. Strengthening the monetary policy framework to lower inflation levels and volatility would not only contribute to improved welfare but also facilitate financial deepening, including development of the markets for longer-term financial instruments. If the improvement in Indonesia’s economic risk rating had been at the top end of the range figured below, then the spreads could have been as much as 145 bps lower.

Figure I.5.
Figure I.5.

Economic Risk Ratings

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

10. The improvement in Indonesia’s financial indicators has been relatively stronger, placing the economy in a better position at the start of the global crisis. The sharp reduction in external debt in particular and the buildup in international reserves has created a useful buffer. In terms of ratings, however, the improvement in these indicators relative to other countries matters. Therefore, further building up of buffers—whether in the form of multilateral, regional, or bilateral insurance arrangements, which tend to be relatively less expensive, or alternatively the accumulation of further international reserves, which could be more expensive—would add to the strength of the external position. Deepening markets and the resilience of the economy, including through greater inflation stability, would also contribute to greater exchange rate stability.

Figure I.6.
Figure I.6.

Financial Risk Ratings

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

11. Indonesia’s political risk indicators have improved more than several other countries, but its level leaves room for further improvement. Improvements were noteworthy in the dimensions of corruption, investment profile, and socio-economic conditions. Continued improvement in these dimensions of the political risk ratings, as well as in law and order and bureaucratic quality, would further lower political and policy risks. Even though the political risk indicators improved substantially in Indonesia, because the level of the rating was still lower than some comparators, the increase in the spread associated with this factor was higher.

Figure I.7.
Figure I.7.

Political Risk Ratings

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

12. To summarize, the model yields valuable insights into the pattern of Indonesia’s sovereign spreads. It fits well the decline in Indonesia’s spreads from 2004 to mid 2007 and captures the subsequent rise. It also fits the pattern of spreads of several other countries. As such, there is no puzzle regarding the recent increase in Indonesia’s spreads. The model accurately captures the quantitative impact on spreads of the worsening in global financial markets since mid 2007. This suggests that continued strengthening of Indonesia’s policy framework, including with regard to price stability but also building additional international reserves buffers, would further bolster the credibility of policies and the economic and financial fundamentals.

C. Beyond the Spreads Model: Discussion of Additional Factors

13. While the model explains the spreads in most countries well, there are a few countries for which it does not account very well for the most recent period. This set of countries includes the Philippines, which has a similar credit rating as Indonesia and is viewed by some in the region as a particularly important comparator. In the Philippines’ case and the other cases, the model predicted a sharper increase in spreads than actually occurred, suggesting that relevant, additional explanatory factors are missing from the analysis.

Figure I.8.
Figure I.8.

Change in Spreads by Contributing Factors

(January 2008-January 2009)

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

Figure I.9.
Figure I.9.

Philippines: EMBI Spreads

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

Figure I.10.
Figure I.10.

Brazil: EMBI Spreads

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

Figure I.11.
Figure I.11.

Turkey: EMBI Spreads

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

14. This section discusses some additional factors that could help account for the behavior of spreads in these countries relative to Indonesia’s. Four factors are considered: (i) corporate sector vulnerabilities; (ii) liquidity in financial markets; (iii) a spending bias in fiscal policy; and (iv) behavior in past crises. Each is considered in turn.

15. Corporate sector vulnerabilities have been somewhat higher in Indonesia. Although corporate sector leverage is low in Indonesia, estimates of default probabilities inferred from asset prices are a bit higher. Using Moody’s KMV tool, one quarter of Indonesian corporates reached 1-year default probabilities of more than 10 percent, higher than some comparators, during late 2008 and early 2009. The asset price behavior was exacerbated by the default of a large conglomerate on its obligations and by general concerns about corporate governance and the external exposure of the corporate sector. Furthermore, the vulnerability to a shock to profits—as measured by the estimated impact on impaired debts—has also been somewhat higher in Indonesia. This reflects the larger number of firms in Indonesia close to the distress level where cash flows are insufficient to cover the interest on debt at the end of 2007. Further strengthening corporate sector health and corporate governance, including through better collection and dissemination of data, could contribute towards lowering risks as well as the perceptions of risks in Indonesia.

Figure I.12.
Figure I.12.

75th Percentile EDFs of Corporates

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

Source: Moody’s KMV.
Figure I.13.
Figure I.13.

Asia: Share of Impaired Debt of Firms with Interest Cover Ratios Less than One

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

Sources: Worldscope; and IMF staff estimates.1/ Assuming a 15 percent decline in profits.

16. During the peak of the crisis, liquidity in Indonesia’s financial markets fell sharply. Transactions reportedly fell from $50–75 million daily in the CDS markets prior to the crisis to $20–30 million daily during late 2008 and early 2009. This limited liquidity exacerbated the price action. Financial deepening would help develop the liquidity of the markets for sovereign and corporate paper, though this is a longer-term agenda. In the near term, building adequate foreign exchange buffers and using these buffers to inject foreign currency liquidity in the markets during times of stress could lower spreads.

Figure I.14.
Figure I.14.

Spending Bias Index 1/

Citation: IMF Staff Country Reports 2009, 231; 10.5089/9781451818451.002.A001

Source: Baldacci, 2009.1/ Based on fixed effect residuals.

17. The spending bias of fiscal policy is greater in Indonesia, potentially exacerbating volatility and raising spreads. As noted in Baldacci (2009), Indonesia’s spending bias is higher than in several other emerging markets. Lowering this bias, while undertaking structural fiscal reforms to enhance budgetary flexibility and public finance management and anchoring fiscal policy in a medium-term framework consistent with public debt sustainability, will contribute to the stabilization role of fiscal policy, lower volatility, and could potentially contribute to lower spreads.

18. Finally, Indonesia has a history of volatility and distress during periods of global stress. This volatility can be captured by the changes in Indonesia’s sovereign ratings during these periods, which in the past decade were related to cases of debt restructuring. The Philippines and Turkey have similar credit ratings to Indonesia, but do not have the same history of external loan restructuring and volatility. Several Latin American countries and South Africa are currently investment grade—higher than Indonesia’s rating—and also do not have a history of loan restructuring. More consistent efforts at transparency and communication could help convey the substantial improvements in Indonesia’s fundamentals in recent years, while limiting any lingering investor concerns that are rooted in the past.

D. Conclusions

19. The decline in Indonesia’s spreads from 2004 to mid 2007 and the subsequent increase in spreads can be accounted for quantitatively. The benign external financing conditions in the first period, together with improvements in political and financial fundamentals, resulted in a sharp decline in sovereign spreads from over 400 bps in mid 2004 to less than 140 bps in mid 2007. The subsequent turmoil in the global financial markets led to a fairly sharp increase in Indonesia’s spreads.

20. Continued improvements in Indonesia’s policy framework would help lower spreads. In particular, achieving and maintaining low inflation would increase domestic welfare, enhance the credibility of the monetary policy framework, and contribute to financial deepening and lower spreads. Similarly, building adequate buffers to meet external financing needs—whether through multilateral, regional or bilateral cooperative arrangements, which are less expensive, or through relatively more expensive self insurance—would further contribute to lower spreads. These efforts could be bolstered by undertaking structural fiscal reforms to enhance budget flexibility and anchoring fiscal policy in a medium-term framework to facilitate a greater stabilization role.

21. Additional steps could be taken to lower spreads. More sustained efforts at gathering and disseminating information on the health of the corporate sector as well as on the fundamental improvements achieved in recent years could assuage investors’ concerns, especially during periods of global stress. In the medium term, financial deepening would help improve the liquidity of sovereign and corporate bonds further limiting spreads volatility.

References

  • Baldacci, Emanuele, 2009, “Neither Sailing Against the Wind, Nor Going with the Flow: Cyclicality of Fiscal Policy in Indonesia,” Indonesia: Selected Issues, forthcoming (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hartelius, Kristian, 2006, “Main Drivers of Emerging Market Bond Spreads: Fundamentals or External Factors?,” Global Financial Stability Report, April, pp. 28– 31 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hartelius, Kristian, Kenichiro Kashiwase, Laura E. Kodres, 2008, “Emerging Market Spread Compression: Is it Real or is it Liquidity?IMF Working Paper 08/10 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • The PRS Group, Inc., 2009, International Country Risk Guide. Available via the Internet: www.prsgroup.com

1

Prepared by Rishi Goyal and Marta Ruiz—Arranz.

2

For Indonesia, the EMBI series are available from mid 2004 onwards.

3

Since the model is estimated in logarithms, the change in spreads depends on the initial level of spreads, or Δ EMBI = EMBI × Δ log(EMBI). Therefore, countries with higher initial level of spreads will, for the same external factors, have a higher contributing factor associated with that factor.

Indonesia: Selected Issues
Author: International Monetary Fund