The recovery from the East Asian crisis has seemed remarkably quick. With the exception of Indonesia, economies have now bottomed out and East Asia is on the road to recovery. The restructuring of bank and corporate debt is, however, not completed. The experience of banking crises elsewhere has shown that emerging markets take an average of three years to return to normal growth. Given the magnitude of the crisis, the level of corporate distress and the continued lack of good governance in these countries, the recovery of banking and corporate sectors, and therefore of the East Asian economies, will take even longer (Claessens et al. 1999). Indonesia’s banking crisis was the worst, and therefore its recovery can be expected to take the longest.
Bank restructuring needs to be the cornerstone of Indonesia’s economic recovery. International experience shows that only with the successful resolution of the banking sector and the construction of a sound financial sector have economies been able to emerge from crisis and reduce their vulnerability to future crises. To understand why the currency and monetary shock hit Indonesia’s banking sector and economy so hard, it is important to understand the vulnerable state of the banking sector before the crisis. Initial interventions to stabilise and rehabilitate the banking system actually deepened the financial crisis and, in combination with the exchange rate and interest rate shocks, led to the systemic banking crisis. This paper explains why Indonesia’s economy and banking sector were hit so hard and looks at the measures to deal with the banking crisis, the lessons learned, and the huge challenges Indonesia faces to restructure and strengthen its banking system.
II. The economic boom and build-up of vulnerabilities
Three main factors contributed to the vulnerability of the banking sector before the crisis. First, after comprehensive reforms in 1988, a rapid expansion of the banking sector took place without the necessary strengthening of prudential regulations and central bank supervision. Second, the high concentration of ownership in the banking sector had led to weak corporate governance of banks. Third, the economic boom and increased international financial integration in the 1980s amplified the structural vulnerability of Indonesia’s financial system.
A. Rapid Expansion of the Banking Sector and Improper Sequencing
As an oil exporter, Indonesia responded to declining oil prices in the early 1980s by deregulating its financial sector to direct domestic savings into developing new industries and to increase the efficiency and competitiveness of the sector. Reforms in 1983 removed credit ceilings, reduced the number of credit categories financed by liquidity credit (which previously had come from oil revenues), removed controls on state bank deposit and lending rates and ended subsidies on state bank deposit rates. As a result real interest rates became positive and time deposits increased dramatically, the ratio of M2 to GDP rose from 18 percent in 1982 to 30 percent by 1988, and the share of private domestic banks in total bank assets increased from 12 percent to 26 percent over the same period.
The dramatic decline in oil prices in 1986 brought macroeconomic adjustments and further structural and financial sector reforms. In October 1988 most of the entry barriers and various restrictions that favoured certain types of banks were removed. The main reforms were:
Open entry for joint ventures (since 1969 the sector had been closed to foreign banks), with a minimum capital requirement of 50 billion rupiah (US$28 million) and maximum foreign ownership of 85 percent. Open entry for domestic banks (since 1977 new entry had been prevented) with minimum capital requirement of 10 billion rupiah (US$6 million). Sound domestic banks were permitted to trade in foreign exchange.
Rules on branching were substantially relaxed. Foreign banks were allowed to open one branch in six other major cities (since 1967 foreign banks had only been allowed two branches in Jakarta).
The government stipulated that 50 percent of foreign bank lending should be to export- oriented businesses (although this requirement was neither monitored nor policed) and 20 percent of domestic lending should be to small and medium-sized companies.
State-owned enterprises were no longer required to deposit all their funds in state banks and could place up to 50 percent of their funds in private banks.
The reserve requirement was reduced from 15 percent of demand deposits and 10 percent of savings and time deposits, to 2 percent of all deposit liabilities.
Legal lending limits were established for loans to a single borrower and to groups of borrowers. In March 1989 bank capital was defined and it was stipulated that banks could not invest in stocks. The ceiling on foreign borrowing was replaced by a net open position of 25 percent of equity.
Banks were allowed to issue shares, and the tax exemption allowed for interest on time deposits was removed to equalise the treatment of interest payments and dividends.
Within a few years of the reforms there was a dramatic increase in the number of banks and branches, money supply and credit. Between 1988 and 1991, the number of new banks entering the system increased from 101 to 182. The range of new products and services also increased. Various types of saving schemes tied to lottery prices and gifts were introduced, and savings deposits (not including government programs) increased from 605 billion rupiah in 1988 to 9,064 billion rupiah by June 1992.
Poor-quality assets and low capital hindered the growth of state banks, but private banks expanded rapidly and by 1994 had begun to overtake the state banks in terms of loans, deposits (private banks were already ahead in 1992) and total assets (Table 1). Private bank branches quadrupled in number between 1988 and 1991 from 559 to 2,639. The government attempted to strengthen the state banks by announcing plans for mergers and privatisation, but only BNI, the largest of the state banks, went public (in late 1996) and no meaningful progress on mergers took place before the crisis.
The rapid increase in liquidity owing to the reduction in reserve requirements and the growth in money supply led to rising inflation in the early 1990s. The monetary authorities tightened monetary policy and responded to growing concern about the rapid expansion of the banking sector by improving prudential regulations.
The regulations announced in February 1991 included a comprehensive capital, asset, management, equity, and liquidity (CAMEL) quantitative-rating system. The system stipulated necessary qualifications of bank owners and managers, a schedule to meet Bank for International Settlement (BIS) capital adequacy requirement (CAR) of 8 percent on risk- weighted assets by 1993, stricter information and reporting requirements, and tougher limits on lending within a corporate group or to one individual group. A new banking law was passed in 1992, allowing sanctions to be imposed on bank owners, managers, and commissioners for violations of laws and regulations related to bank management. Foreigners were allowed to purchase bank shares, and state banks became limited liability companies to allow them more autonomy as private corporations. In October 1992, as part of the government’s desire to limit the number of banks, the capital required to set up a domestic bank was increased fivefold and was doubled for setting up a joint-venture bank.
Despite these new regulations and further improvements in prudential supervision, weaknesses still existed in the legal and regulatory framework especially with regard to loan classification and loan provisions. The number of new banks continued to increase after 1991. An even more serious problem was the lack of ability to enforce prudential regulations because of the weak capacity and capability of central bank supervisors, widespread corruption, and political interference from bank owners who were close to the centre of power. Violations of prudential regulations were not properly penalised and non-compliance was widespread, as revealed by the audits of banks undertaken after the crisis.
B. Weak Corporate Governance
Although prudential requirements and regulations such as legal lending limits were introduced to improve corporate governance in the banking sector, governance was weak and there was little incentive for banks to be cautious in corporate lending. There were three main reasons for this problem.
The first is that despite the rise in the number of banks and the increase in the issuance of bank shares in the capital market, the banking sector remained highly concentrated, as did the ownership of banks. Following the 1988 reforms, the number of private banks doubled to reach 164 in 1996, but just ten banks owned 68 percent of total bank assets. Bank Central Asia was the largest private bank and largest bank in Indonesia before the crisis, even surpassing the largest state bank, Bank Negara Indonesia (BNI). Bank Danamon was the seventh largest bank and Bank International Indonesia was the ninth largest bank. The other large banks were owned by the state. The top ten private banks and the six state banks together accounted for 75 percent of total bank assets. The majority of bank shares were still in the hands of the original owners, and this concentrated shareholding had created information asymmetries between the majority shareholders and the minority shareholders, investors and creditors. Despite legal limits on lending to affiliated firms within the same group or to just one group, there was gross violation of these limits. The top ten private banks were linked to politically powerful business conglomerates (Table 2). Both Bank Central Asia (BCA) and Bank Umum Nasional (BUN) had shareholders who were linked to former President Suharto. Bank Duta was controlled by Suharto foundations and held the funds of Badan Urusan Logistik Negara (Bulog), the state logistics agency.
|Bank||Total assets||Part of a group||Strategic focus||Status|
|1.||Bank Central Asia (BCA)||Rp 36.1 tn||Salim div. group||All market segments||Taken over|
|2.||Bank Danamon||Rp 22.0 tn||Danamon div. group||Retail-commercial||Taken over and merged|
|3.||BII||Rp 17.7 tn||Sinar Mas div. group||Retail-commercial||Recapitalized|
|4.||BDNI||Rp 16.7 tn||Gajah Tunggal div. group||Retail-commercial||Closed|
|5.||LTPPO||Rp 10.2 tn||Lippo div. group||Retail-commercial||Recapitalized|
|6.||Bank Bali||Rp 8.0 tn||Bali financial group||Retail-commercial||Taken over|
|7.||Bank Niaga||Rp 7.9 tn||Hasjim div. group||Corporate-consumer||Taken over|
|8.||Bank Umum Nasional||Rp 7.1 tn||Bob Hasan-Ongko div. group||Retail-commercial||Closed|
|9.||Panin||Rp 5.4 tn||Panin financial group||Retail-commercial|
|10.||Bank Duta||Rp 5.3 tn||Berdikari div. group||Retail-commercial||Taken over|
|Total bank loans||150.3||188.9||234.6||292.9|
|Total property loans||21.7||33.2||42.8||58.9|
|Property loans/total loans||14.5||17.6||18.3||20.1|
A number of the large banks, including BNI, did issue shares to the public. However, the state or the main owner retained most of the shares and therefore a great deal of control over the bank. Other mechanisms to impose good governance on owners and managers, such as central bank supervision and credit-rating agencies, were not effective because of a lack of enforcement and the existence of information asymmetries.
Second, state-owned banks and some private banks were under political pressure to direct lending to particular sectors or groups without proper evaluation of the loans. This was most prevalent during the oil boom, but remained the case to a lesser extent in the post-1988 reform period.
A third reason is the implicit guarantees given to certain groups and state-related banks or corporations. Poor corporate governance, combined with a concentration of ownership by those close to the centre of power, led to imprudent lending because it was believed banks were ‘too big to fail’ or were connected to powerful groups, and would be bailed out by the government. There was no effective exit mechanism for failed banks, and well-connected insolvent banks were allowed to remain open. Only one bank was closed between 1988 and 1997, with the other problem banks being bailed out directly by the government or by corporations that the government or President Suharto persuaded to step in. This created a situation of moral hazard that contributed to the risky behaviour of banks and excess capacity in the sector, which in turn were central to the structural vulnerability of the banking sector before the crisis. The argument that a bank was too big or too important (because of political connections) to fail was very common (see Box 1 for examples).
Box 1.Too Big or Too Important to Fail
The state-owned development bank Bappindo had been having problems for many years. In the late 1980s, Bappindo was found to be holding a large number of nonperforming loans and to have been involved in serious corruption to obtain credit for large projects. Instead of closing down or dramatically restructuring the bank, the government allowed the bank to continue to operate. Bank officers but not managers were punished for the corruption, and one businessman who escaped from prison was never prosecuted.
Bank Duta, a private domestic bank, had experienced large foreign exchange losses from currency speculation. Despite these losses the bank went public, using fraudulent financial statements. Because the bank held the deposits of Bulog and the Suharto family foundations, it was rescued by a contribution to a bail out package from an Indonesian conglomerate. The manager of the treasury division was jailed and the management of the bank was changed.
Indocement and a CRMI, a cold rolling mill for processing steel, were both part of the Salim group and were rescued by the government coming in as a shareholder.
The government gave implicit guarantees to firms by providing captive markets or favouring firms through special policies and directed lending (often involving state banks and/or central bank liquidity credits). One of the most blatant examples was Indonesia’s national car, the Timor, and the domestic clove trading monopoly, that were both linked to the former president’s youngest son, Tommy. The Timor company was allowed to import parts and components duty free, and later to fully import cars duty free from Kia in Korea. The argument was that local content requirements would be met within three years. Not only was Timor given duty-free status, it had a captive market in that it stocked the police fleet and civil servants received incentives to purchase the vehicle. Banks, including state banks, were also asked to loan money to the venture. Another example is the clove trading monopoly that had control of purchasing cloves from farmers for resale to cigarette manufacturers, and was given low- interest credit directly from the central bank. Later on it ran into many difficulties and had to be dissolved.
C. Macro-financial Linkages2
The build-up in vulnerability before the crisis was created by the reinforcing dynamics between capital inflows, macroeconomic policies and weak financial and corporate sector institutions.
Capital inflows and policy responses
Indonesia’s reform program since the mid-1980s has aimed to move the economy away from a dependence on oil exports toward higher-growth sectors. In response private capital has surged into Indonesia and there has been a progressive integration with world financial markets. Capital inflows validated and exacerbated domestic macroeconomic cycles, leading to overheating of the domestic economy in 1990–91.
Aiming to counteract the inflationary pressures, the Indonesian government chose a macroeconomic policy mix that actually encouraged further capital to flow into the country. Because fiscal policy was not countercylical, most of the burden for responding to excess demand pressures fell on monetary policy. Monetary policy was ineffective, however, as the amount of liquidity taken out of the system by sales of Bank Indonesia certificates or Sertifikat Bank Indonesia (SBIs)3 began to match the amount of interest paid out. Furthermore, Indonesia’s open capital account meant that banks and non-banks were able to borrow offshore, thus increasing liquidity and offsetting the government’s attempts to tighten monetary policy.
In the early 1990s, the monetary authorities attempted to curb these leakages of liquidity by imposing offshore borrowing limits on banks, state-related lenders and, eventually, non-bank financial institutions. Even when monetary policy eased, interest rates on lending did not fall immediately because of the lag effect and because banks had increased provisioning for problem loans. By 1991 at least two banks were known to be carrying problem loans. The government responded very differently in each case. Bank Duta, which held the funds of Bulog and the Suharto foundations, and was saved by an injection of private capital. Bank Summa, on the other hand, was liquidated. The increase in problem loans was an early warning sign of the vulnerabilities that existed in the banking sector.
In the mid-1990s, inflation rose, especially in the non-tradable sector, and the current account deficit widened. Another bout of overheating occurred in 1994–96, after a second surge of private capital inflows. Large private capital inflows further added to the vulnerability of the banking system.
The monetary authorities continued to try to curb overheating by maintaining a relatively tight monetary stance so that interest rates remained high, increasing reserve requirements, using moral suasion and, in 1996-97, finally imposing limits on property lending. In late 1995 the central bank banned the issuance of commercial paper by finance companies, which triggered a massive switch to offshore borrowing. Finance companies alone borrowed US$5.1 billion in 1996, more than one-quarter of total corporate debt issuance in the year, and up from only about US$819 million in 1995. In 1996 the central bank stipulated that commercial paper traded by banks must be rated, effectively requiring all companies issuing commercial paper to be rated.
The policy mix adopted encouraged capital inflows. High domestic interest rates led to an increase in deposits from non-residents or from Indonesians based overseas. Although in the early 1990s, banks took out sizeable external loans, limits on offshore borrowing had curbed their activities. However, private corporations were not restricted from borrowing abroad, and with high domestic interest rates and the predictable downward movement in the exchange rate, companies borrowed abroad without hedging and deposited the funds in the domestic banking system for activities that were predominantly rupiah based. The figure shows the extent to which private capital flows to the banking sector -through borrowing by banks and through the growth of the deposit base - have supported the growth in domestic credit to the private sector.
Figure.The Growth of Credit to the Private Sector and the Contribution of Foreign Capital
Sources: Bank Indonesia and World Bank estimates.
The dependence of the private sector on the domestic banking system for finance suggests that the bank lending is likely to be an important transmission mechanism in Indonesia. Although difficult to obtain, empirical evidence does suggest that economic activity in Indonesia is more sensitive to changes in domestic credit than to changes in the money supply.4
Although Indonesian firms have had greater access to overseas funds, either through syndicated loans or through the issuance of international equity, bonds and commercial paper, such access remains limited to the larger firms. Most firms, particularly small and medium-sized enterprises, continue to rely primarily on the domestic financial system. Furthermore, despite greater development of the financial sector, most firms turn to domestic banks for external finance. Between 1990 and 1996, banks intermediated around 40 percent of private non-bank investment on average. Direct finance from abroad funded 7 percent of investment, the government financed 10 percent and the remaining 43 percent was funded by the domestic non-bank sector. Net new lending from banks totalled 58.3 trillion rupiah in 1996, compared with the new issuance of IPOs (initial public offerings) and rights of 14.6 trillion rupiah, and new bond issuance of 8.6 trillion rupiah (Ghosh et al. 1999).
Financial integration and overheating
Flush with liquidity, domestic banks rapidly expanded credit, particularly to risky sectors such as property. Banks were also keen to lend to consumers to capture shares of the domestic retail market. Consumer borrowing went into consumption and speculation in property and stocks, especially up until 1995 while interest rates were still low. This added to the inflation in the real estate and stock markets. Attracted by high prices, banks increased lending to the property sector. Property lending increased from 6 percent of GDP in 1993 to 16 percent of GDP in 1996.
The rapid expansion of credit to the property sector fed through into the demand for and supply of commercial and residential property. Between 1991 and 1997, office occupancy rates in Jakarta were around 90 percent and property expansion was based on the continuation of this demand. Capacity in the residential property market can be gauged by the number of property development licenses issued by the National Land Registry Agency (BPN). By August 1997, 57,600 hectares of the 121,629 hectares licensed for residential use in the Greater Jakarta area had been acquired and cleared, but less than 13,600 hectares (about 11 percent of the total licensed land) had been developed. The remaining 44,000 hectares had been acquired but not yet developed, making property developers susceptible to an economic downturn, and potentially leaving banks with between 8 trillion rupiah and 16 trillion rupiah worth of nonperforming loans. The amount of mortgage loans almost tripled between 1993 and 1996 from 6 trillion rupiah to 16 trillion rupiah. Similar excess capacity and over investment was seen in infrastructure projects such as power generation, and some manufacturing sectors, such as automobiles.
The liberalisation of the banking sector after 1988 had increased the presence of foreign banks. It was hoped that foreign competition would transfer technology through technical assistance or the movement of foreign personnel to local banks, but it is unclear whether this occurred. Two indicators of competitiveness – bank net interest and operating margins – showed no real decline, partly because foreign competition took away prime borrowers from local banks and increased the risk and costs of competing domestically.
Foreign banks have mainly focused on the corporate sector, particularly on the business of multinational companies operating in Indonesia and the top domestic corporations, which had more conservative and stricter credit-risk profiles. There was intense competition for business in this market. The top Indonesian corporations, with the help of foreign investment banks and commercial banks, were able to tap capital markets (both foreign and domestic) through the issuance of equity or debt instruments (short-term commercial paper and long- term bonds). Stocks issued through the capital markets grew from 27.6 trillion rupiah at the end of 1991 to 152.2 trillion rupiah by the end of 1995. Bonds issued grew from 2.2 trillion rupiah to 5.3 trillion rupiah over the same period. With high interest rates at home, the top firms enjoyed the benefits of offshore funding, where the risk premium charged was lower and was declining as the reputation of these firms increased. The spread on Eurobonds for large companies such as Astra narrowed from an average of 250-300 basis points in the late 1980s to around 150-200 basis points in the 1990s.
Domestic banks tried to avoid competing directly with the foreign banks and gradually shifted their focus to what they often called ‘the middle market’ – of second-tier corporations, small and medium-sized businesses, and consumers. Most of the main private banks began to focus on the retail market in the mid-1990s. These loans entailed greater risk, reflecting the lack of information and transparency in these sectors and the higher levels of nonperforming loans. Interest spreads on a retail loan are normally 2–3 percentage points higher than on a corporate loan. Before the crisis average corporate lending rates were 19–20 percent, collaterised consumer loan rates were 22–23 percent, and unsecured consumer loans, such as credit card debts, carried interest rates of approximately 30 percent.
As loans involved a greater risk and the quality and quantity of individual or sectoral credit information was limited, a great deal of effort was put into establishing effective credit assessments. Lending was based mainly on the value of specific collateral, such as a business property, house or car, which could be quickly sold. When lending opportunities reached a saturation point, banks looked to expand their market geographically or to launch new financial products. Although most banks continued to focus on the larger cities, many ventured into smaller towns, where there was little or no previous experience of bank borrowing.
Net interest margins were kept high and operating margins low by the risk profile of domestic banks, as they shifted into new markets and locations and made efforts to go public, which demanded investment in technology, human resources, branch networks and improvements to bank management.
Despite their higher risk profile, most domestic banks continued to have inadequate provision for bad debts. It was not uncommon for nonperforming loans to be higher than provisioning for these loans and this practice was reinforced by Bank Indonesia, which allowed banks to deduct the value of loan collateral from provisioning needs. Most domestic banks were still in the early phase of entering the retail sector and expanding their networks when the crisis hit.
Financial integration and corporate debt
As there were no limits on private sector borrowing, but domestic interest rates were high and exchange rate movements seemed predictable, large Indonesian companies increased their short-term unhedged external liabilities in the years before the crisis.
Table 5 shows the nature and extent of private sector debt in Indonesia. The data only became available after the crisis, as a result of mandatory reporting to Indonesia’s External Debt Committee. Private sector debt was very large, with the main debtors being corporations rather than banks. Commercial lending rather than securities was the main debt instrument, and most lending was short term. Industrial corporations, including state-owned enterprises, had an outstanding foreign debt of US$65.3 billion in 1998. Of the banking and corporate sectors’ total US$82.0 billion debt, US$72 billion or 88 percent was borrowed from banks and the remaining was in the form of marketable securities (commercial paper, medium-term notes and floating rate notes).
|Bank + corporate||71,952||10,026||81,978|
Detailed data from PEFINDO, Indonesia’s credit-rating agency, which tracked the cross- border debt of about 400 Indonesian companies, including banks, show several interesting trends. First, it appears foreign borrowing increased rapidly between 1994 and 1997, before ending abruptly when the crisis started. The increased lending was driven by the lower cost of offshore borrowing and the issuance of the central bank’s Yankee bonds in 1996, which provided a benchmark for Indonesian entities. Interest rates at home were high because of the tight monetary policy and because of the inefficiency of domestic financial markets and the high transaction costs of domestic borrowing.
Table 6 compares the issuance costs of domestic debts (bonds, commercial paper and bank loans) with that of overseas loans in 1995. The most attractive instrument, in terms of the issuance cost, had been commercial paper (CP), before the central bank tightened regulation of this instrument in early 1996. The CP issuance process was relatively simple and cheap and there was flexibility to roll over the debt. The estimated issuance cost for CP was estimated at about 50-100 basis points (bps) compared with 259 bps for domestic bonds, 179 bps for medium-term notes, 200 bps for bank loans and 75 bps for offshore borrowing. In addition, before 1996 banks were able to use investment in CPs to circumvent the legal lending limit. The CP market has boomed, with capitalisation almost doubling every year between 1989 and 1996. A substantial proportion of the CPs is held by foreign investors.
|Risk, premium||100–200 bps||100–200 bps||200–300 bps||300–400 bps||150–200 bps|
|SWAP (p. a.)||n.a.||n.a.||n.a.||n.a.||700–900 bps|
|Processing time (months)||6–8||1–2||1–2||1–3||2–3|
|Annualised issuance costa||259 bps||50–100 bps||179 bps||200 bps||75 bps|
A large proportion (80 percent) of corporate debt was short term and was issued to companies oriented to domestic markets rather than to export markets that would have provided the natural hedge of dollar earnings. The five most active cross-border borrowers during the period were banking and finance companies, followed by infrastructure, property, and pulp and paper companies. All except the last group of companies were in the non- tradable sector. The bias toward non-tradable goods was caused by the commissioning of large-scale private infrastructure projects (e.g., power, telecommunications, toll roads and water utilities) and large investments in cement, chemicals, auto assembly, and auto parts and components. Rising real wages combined with other cyclical factors such as the downturn in the Japanese economy and the depreciation of the yen, caused a slowdown in export growth. In the absence of a natural dollar hedge, the majority of cross-border borrowing (approximately 80 percent) was unhedged; that is, not covered by future currency swap contracts.
III. The bust: evolution of the crisis and the collapse of the banking sector5
The exchange rate and interest rate shock that triggered the crisis in July 1997 had a dramatic effect on the balance sheets of banks and highly leveraged corporations. As a result of the exchange rate collapse, banks had to repay in full their liabilities in foreign currency when loans extended in foreign currency were not fully repaid. The liabilities of banks increased sharply owing to their exposure to interest rates, while corporate distress affected the value of bank assets. In the following months, a combination of a crisis of confidence, which was related to initial policy miscalculations, further rounds of exchange rate and interest rate rises, hyperinflation (inflation reached close to 80 percent in 1998) and the contraction of the real economy (-14 percent GDP contraction in 1998), worsened the distress in the banking and corporate sectors. A limited banking crisis quickly became a systemic banking crisis.
A. Conventional Wisdom on Bank Restructuring
There are various options for undertaking bank restructuring, all of which entail trade-offs between speed of restructuring, fiscal costs, incentives for bank performance and confidence in the banking system (Claessens 1998). A bail out is the fastest option, but entails the highest cost, the greatest disincentive for bank performance and financial discipline, and does not increase confidence in the banking system. It is not surprising that this option was not adopted in East Asia. In Indonesia, particularly, the severity of the crisis meant bail outs did not make sense. The other extreme of closing down non-viable banks and paying off creditors and depositors would also have been speedy, would have sent a strong signal about financial discipline and would have involved a relatively low fiscal cost (depending on the extent of non-viable institutions). However, it would have had a dire effect on the confidence of the banking system. Most East Asian governments, including the Indonesian government, have selectively closed non-viable banks, facilitated mergers of banks and recapitalized distressed banks with the plan to sell these at a later date. The experience of other bank crises, including elsewhere in East Asia,6 suggests there are some key principles that should be followed in bank restructuring:
only viable institutions should remain in operation;
costs of restructuring should be transparent and the burden on taxpayers should be minimised;
restructuring should work toward establishing good governance by allocating losses to existing shareholders, creditors and perhaps large depositors;
measures introduced and implemented should preserve the incentives for new private capital and impose discipline on bank borrowers; and
restructuring needs to take place at a sufficient pace to restore channels of credit while maintaining confidence in the banking system.
B. Failed Stabilization Efforts: From a Limited Banking Crisis to a Systemic Banking Crisis (August 1997 – December 1997)
Indonesia’s initial response to the flotation of the Thai baht on 2 July 1997 was to widen its exchange rate band from 8 percent to 12 percent. This led to an immediate 7 percent depreciation of the rupiah to 2,600 per US dollar and the first wave of capital flows out of the country, probably from international mutual funds and hedge funds. As pressure built up and it became obvious that intervention to defend the band would be too costly, the rupiah was floated on 14 August and monetary policy was tightened considerably. Overnight call rates increased to a very high 81 percent and SBI rates rose from 12 percent to 30 percent. However, the rupiah continued to weaken and depreciated even further through the remainder of August and the first week of October, when it was around 3,000 to the US dollar, as corporate borrowers with large and unhedged external debt tried to cover their positions.
The rupiah depreciation, rise in interest rates and the problems beginning to be experienced by over leveraged borrowers, led to the first round of support for the banking system. In September the government announced a stance of fiscal austerity, including postponing projects planned by the president’s children, and released a plan to restructure the banking sector, including closures of unsound banks. The plan lacked clarity and credibility, and the rupiah continued to weaken. The direction of policy was further confused with the decision to loosen monetary policy, and SBI rates fell to about 20 percent. The conflicting signals about monetary policy encouraged further capital outflows.
On 8 October the Indonesian government announced that it would ask the IMF for assistance. Uncertainty about the president’s commitment to this move caused the rupiah to further weaken to around Rp3,500. The first IMF letter of intent (LOI) was announced on 1 November 1997. The initial response to what was then perceived to be a limited banking crisis affecting only a small number of the weakest banks, including the state banks, was in accordance with conventional wisdom. That is, in order to stabilise the financial system and prevent capital flight and a breakdown of the payments system, non-viable or unsound banks should be closed without causing a loss of confidence. In order to protect depositors and to maintain the confidence of creditors, a scheme for guaranteeing deposits was introduced.
The first LOI involved a bank restructuring program that was quite comprehensive, although at the time of the announcement this was not made clear to the public. The plan was to immediately close sixteen small and deeply insolvent banks (with a market share of 2.5 percent). Protection was limited to small depositors - those with deposits of up to 20 million rupiah (around US$6,000) - who accounted for 90 percent of the depositors in the banking system. It was thought that decisive government action would improve confidence in the banking system, but failed to do so in the absence of a comprehensive deposit insurance system. The banking reforms also included steps to intensify the supervision of a number of the largest private banks, rehabilitation and surveillance plans for a number of smaller private banks, and mergers of state-owned banks.
In the following two weeks there was initially a positive response to the first LOI and the rupiah strengthened to around Rp 3,000 to the US dollar. Confidence then began to falter. The bank closures were not well planned or executed. The details of the first IMF LOI were never made public; the only information provided was a summary of the main reforms announced by the minister of finance and the governor of the Central Bank. Before the announcement of the first LOI, it was widely expected that banks would be closed, including those known to be weak such as Bank Pacific, Bappindo and a number of others. However, the arbitrary choice of banks to be closed and the unclear criteria used led to speculation that more banks would close, especially since the names of the other thirty-four banks to be rehabilitated were not announced. The deposit guarantee of 20 million rupiah was seen as insufficient, and domestic investors transferred deposits from private banks to state banks in a flight from quality to safety. Many also transferred funds to foreign banks or exchanged rupiah for dollars and repatriated their funds. This was the beginning of the flight of domestic capital, which worsened rapidly in December as the crisis of confidence deepened.
Toward the end of November and into December several events worsened the crisis of confidence, including another round of external contagion in response to Korea’s deepening crisis. In addition, one of the closed banks was owned by a son of the president, who challenged the minister of finance about the closure. The bank was allowed to be resurrected under a different name. This was the first indication that the president was not going to adhere to the IMF’s reforms. The deposit runs accelerated further with rumours of further bank closures, the illness of the president and rumours of the death of Sudono Salim, the head of the largest business conglomerate and owner of the largest bank, Bank Central Asia. The flight to safety intensified. By mid-December 1997, 154 banks (approximately 50 percent of the banking system) had experienced a run on their deposits. During December 1997 Bank Indonesia’s liquidity support to banks increased from 13 trillion rupiah to 31 trillion rupiah or 5 percent of GDP. Most of this liquidity was being funnelled abroad (Lindgren et al. 1999).
C. Second Round of Stabilization and Second Shock (January 1998 - May 1998)
The crisis of confidence worsened in January 1998, with the rupiah plummeting to 15,000 to the US dollar at the end of January. The president’s lack of commitment to the reforms was evident, with the announcement of an unrealistic budget in early January and indications that he did not intend to fulfil the second LOI signed in mid-January. Credibility in the management of the country and economic policy was dealt a further blow when as the president made it clear that his choice of vice-president would be the controversial B.J. Habibie. When four of the central bank’s directors were fired, it became clear that the technocrats were no longer in charge of economic policymaking. The worsening crisis of confidence tipped the banking sector into a full-fledged systemic crisis. By the end of January liquidity support from Bank Indonesia exceeded 60 trillion rupiah. The increase in liquidity support to the banking system immediately entered into circulation, further adding to the growth of the money supply as people withdrew their deposits from banks. Fears of hyperinflation began to emerge as a result of the increase in money supply and the exchange rate depreciation.
To restore confidence, the government (with IMF approval) announced another bank restructuring package on 26 January 1998. The package had three components. First, all bank liabilities were guaranteed by the government. The guarantee covered both on- and off- balance-sheet obligations, with subsequent automatic extensions every six months unless a change was announced. Derivative transactions (other than currency swaps) and bank liabilities to affiliated parties and those who held 10 percent or more of the bank’s shares were excluded from this guarantee. Second, the Indonesian Bank Restructuring Agency (IBRA) was set up to supervise and restructure the banking sector. The IBRA’s mandate was to close, merge or take over and recapitalize troubled banks. Recapitalized banks would eventually be sold. The IBRA was also given the tasks of recovering the bad loans belonging to banks that had been taken over or closed, and of monitoring and selling corporate assets pledged or transferred from former bank owners as collateral for emergency central bank liquidity credits. The IBRA was expected to complete these tasks in five years, after which time the institution would no longer exist. Third, the government finally realised that recovery and bank restructuring could not be achieved without corporate restructuring.
Box 2.Policy Response to the Crisis: Failed Stabilization Efforts
(end of period)
|11/97||16 banks closed||3,648||20.0||44.8|
|1/98||Blanket guarantees given, IBRA established||10,375||20.0||43.2|
|2/98||54 banks transferred to IBRA control||8,750||22.0||50.8|
|4/98||10 banks frozen, 3 taken over||7,500||46.4||50.3|
|5/98||Suharto resigns, Massive deposit run,|
Liquidity support increases.
The immediate effect on confidence was relatively positive, and the rupiah strengthened to below Rp 10,000. By February 1998, fifty-four banks (36.4 percent of banking sector) that had borrowed heavily from Bank Indonesia (more than 200 percent of their capital) and had a CAR of less than 5 percent were placed under IBRA supervision. This included the four state-owned banks (BAPINDO, Bank Bumi Daya, BDNI and Bank Exim), which together accounted for one-quarter of the liabilities of the banking sector, and fifty private and regional banks. However, continued uncertainties regarding the implementation of the IMF reforms, including the president’s statements in February about introducing a currency board system, the replacement of the head of the IBRA and political upheaval leading up to the presidential selection in March further undermined confidence. Deposit runs continued and credit lines to domestic banks were withdrawn. Further liquidity was needed.
In early March Bank Indonesia’s liquidity support was unified into a single facility, with interest rates only slightly above market rates as the focus shifted from using high interest rates to deter irresponsible usage of liquidity credits, to non-market sanctions. If a bank’s borrowings were outstanding for more than one week, the central bank would inspect the bank and report whether its activities should be restricted or whether it should be put under IBRA control. Meanwhile, the monetary authorities hiked up interest rates, with SBI rates doubling to 45 percent, and therefore the interest rates on liquidity support remained high.
In early April the IBRA announced its first major action – the takeover of seven large banks. These banks had borrowings of more than 2 trillion rupiah (15.6 percent of banking liabilities) and received over 72 percent of the liquidity support provided by Bank Indonesia. One of the seven was a state bank (Bank EXIM) and the other six were the major private banks. Seven smaller banks (0.4 percent of the banking system) that had borrowed more than 500 percent of their capital were also closed. Because of the previous experience of bank closures, great effort was made to create a smooth transition by ensuring that the deposits of these banks were directly transferred to a designated state bank, Bank Negara Indonesia, on the weekend of the closure. Public announcements were made about the criteria for closure and takeover. The actions were well received by the market and there were only sporadic bank runs.
However, the banking system was hit by another big shock in the weeks before the May riots that led to the resignation of President Suharto. The rupiah, which had stabilised at around Rp 10,000 between February and April, depreciated again to above Rp 10,000. There was a serious loss of confidence by both domestic and foreign investors. In the week of the riots and following the riots, there were massive deposit runs on Indonesia’s largest bank, Bank Central Asia (BCA), which accounted for 12 percent of the banking system. The BCA was majority owned by the Salim group, which was close to President Suharto (30 percent of BCA’s shares were held by two of the president’s children). The central bank and two state banks injected liquidity of around 30 trillion rupiah into BCA over the week following 16 May. On 29 May the BCA was taken over by the IBRA, shareholders’ rights were suspended and the management was changed. This stemmed the runs on the BCA.
Interest rates once again climbed, with SBIs fetching an interest rate of 70 percent and deposit rates reaching 60–70 percent as banks sought to maximise their liquidity to protect against potential deposit runs. Inflation reached 50 percent. The negative spread experienced by the banking sector increased substantially during this period, further damaging its capital base. In October 1998 the private national banks and the seven state banks had negative equity, leaving a major portion of the banking sector technically insolvent. GDP contracted by nearly 14 percent in 1998 and bank nonperforming loans reached 75 percent of total loans.
IV. Rehabilitation and recapitalization of the banking sector (June 1998 - December 1999)
After the resignation of President Suharto in May 1998, President Habibie took the reins and a few months of uncertainty followed. The exchange rate remained weak and interest rates stayed high. There was little progress in bank restructuring as efforts were focused on auditing the banks and preparing for recapitalization.
The new government faced the tasks of completing the selection of viable and non-viable banks, dealing with non-viable banks and recapitalising the remaining viable banks. To achieve these tasks, a clear criterion of viability was needed, which should have been linked to the terms of the restructuring of operations by imposing costs on existing owners (e.g., diluting shareholding, forcing consolidation, changing ownership/management). Proper prudential oversight was also needed. Unfortunately the blanket guarantee meant that the cost of recapitalization was the government’s responsibility. The cost was determined by the government’s ability to resolve value-impaired assets by restructuring nonperforming loans (restructuring, rescheduling, sale and swap), and selling off assets and banks. Although the government had handed over responsibility for restructuring and the sale of assets to the IBRA, it could not resist interfering with the process.
The audits undertaken on the banks taken over revealed the complexity and magnitude of the banking crisis. In June 1998 the audit of the six private banks taken over in April 1998 revealed that on average nonperforming loans were 55 percent of total loans (90 percent in one large bank), that most loans were to affiliates and that banks were deeply insolvent. On 21 August the assets of three of these six banks were frozen (Bank Umum Nasional, BDNI and Bank Modern) and their deposits were transferred to designated state banks. Bank Danamon would be recapitalized by the government and act as a bridge bank for further mergers with other banks. PDFCI and Bank Tiara were given a final opportunity to be recapitalized by their owners or they would either be closed or merged with Bank Danamon. In early August 1998 audits of the remaining banks revealed their weak situation and therefore the weakness of the whole banking system.
Under the government’s guarantee program, many interbank loans were categorised as ineligible. Group-affiliated banks were affected most by this restriction. The IBRA took action against ten former owners of taken-over banks who were deemed to have violated their legal requirements. They were asked to pay back the liquidity support obtained from Bank Indonesia and the amount of affiliated lending. By late September several of these owners had pledged assets that they had valued at 200 trillion rupiah, as well as about 1 trillion rupiah in cash. The IBRA’s advisors valued the assets at 92. 8 trillion rupiah and tentative settlement as reached. A protracted debate took place as to how much up-front cash that owners should provide. Suggestions of transferring the ownership of assets, including the possibility of giving some shares to cooperatives, sparked political controversy. In the end it was agreed that obligations should be settled within four years, with 27 percent to be realised in the first year.
Box 3.Rehabilitation and Recapitalization of Banks
|9/98||Audit and recapitalization.||10,700||68.8||10.8|
|10/98||4 state banks merge into Mandiri.||7,550||59.7||(28.5)|
|3/99||38 banks closed,|
9 recapitalized, 7 taken over.
|5/99||11 private banks, 12 regional|
development banks recapitalized.
Wahid becomes president.
B. Recapitalization and the Cost of Bank Restructuring
With bank equity becoming negative, the IBRA launched its recapitalization program in September 1998. The objective was to recapitalize viable banks and share the burden of restructuring between the government and the private sector. The government’s contribution would be in the form of bonds, while the owners would contribute cash. After three years the government would commission an independent valuation of the bank and the owners would be able to reacquire their share in the bank by repaying the government’s contribution. The owners had the first option to buy back the government’s share, but the government could sell it to other investors after a specified period. To encourage owners to inject new capital, the government allowed them to retain management control. Category-5 loans (bad loans) or those already written off were transferred at zero price to the IBRA’s Asset Management Unit. The proceeds from the resale of these loans would be used to buy back the government’s shares, giving the possibility of returns on the government’s capital injection and reducing the amount owners would have to pay to reacquire the bank.
Banks were categorised into three groups based on an audit by international accounting firms. ‘Category A’ banks had a CAR of above 4 percent and were exempt from the program and could resume operations. ‘Category B’ banks had a CAR of between 4 percent and −25 percent and were candidates for the program provided that their owners could inject 20 percent of the new capital required to attain a CAR of 4 percent. Banks with a CAR of less than −25 percent were put in ‘category C’ and their owners were given time to inject sufficient equity to push them into a higher category that would make them eligible for recapitalization. Category B and C banks whose owners could not add sufficient capital would be taken over by the IBRA or closed.
The recapitalization program experienced various delays.The deadline of 26 February 1999 for announcing the categorisation of banks was delayed and some banks that should have been closed ended up being taken over by the government. These glitches damaged confidence and the rupiah weakened again to Rp 10,000. Finally, in mid-March the government announced that of 73 of the 140 category A banks did not need government support, nine category B banks (10 percent of the banking system) were eligible for the recapitalization program, thirty-eight banks (5 percent of the banking sector) would be closed and seven banks (2 percent of banking system) would be taken over by the IBRA.
The owners and managers of the category A banks were reviewed to see whether they were fit and proper, and one-third did not pass this test. The managers and commissioners who did not pass were replaced and the owners who failed were given 90 days to divest their shares.
The nine category B banks were given five weeks to add additional capital, and seven met the 20 April deadline. The other two, Bank Bali and Bank Niaga, experienced problems. Bank Bali was in the middle of negotiations with Standard Chartered when a corruption scandal broke out. Bank Niaga’s major shareholder failed to provide sufficient capital and the bank was taken over by the IBRA.
Of the thirteen banks taken over by the IBRA, nine were merged with Danamon, while BCA, Niaga and Bali were recapitalized. The larger recapitalized banks were BII (affiliated with the Sinar Mas Group), Lippo (Lippo Group) and Universal (Astra Group).
Four smaller banks were also recapitalized: Bukopin (Cooperative Bank), Prima Ekspress, Arta Media and Patriot.
The IBRA negotiated performance contracts and memorandums of understanding with the owners and managers of the eight banks to be recapitalized, obtaining ordinary stock and allowing owners to take management control. The estimation of the amount needed for recapitalization had been done in September 1998 but the economy and political situation had not improved by May, and therefore leading up to the elections, the rupiah was still weak. The updated audits indicated that the amount needed for recapitalization would be almost double what was originally predicted.
Box 4.Summary of Bank Restructuring
|No. of banks|
|No. of banks|
|>4%||−25% to 4%||<−25%|
|7||–||–||7||4 merged to 1||5|
|1 new (BEI),|
|est. Rp 290 tn (US$36 bn)|
|Mandiri||Rp 178 tn|
|BNI||Rp 70 m|
|BRI||Rp 29 tn|
|BTN||Rp 14 tn|
|Regional development banks|
|cost Rp 1.2 tn (US$2 bn)|
|Private national banks|
|7 recapitalized||cost Rp 17.8 tn (US$2 bn)|
|13 BTO (4 recapitalized|
9 merged into 1)
|cost Rp 130 tn (US$16 bn)|
The restructuring of the state banks has been much slower than planned. Even though all the state banks and some of the regional development banks fell under category C, they were all recapitalized after restructuring and mergers had taken place. The four state banks – Bank EXIM, BDN, BBD and Bapindo – merged to become Bank Mandiri in September 1998. The corporate business segment of BRI was also merged into Bank Mandiri, and BRI changed its focus to small business lending. The nonperforming loans of the four banks were transferred to the IBRA’s Asset Management Unit. The management of Bank Mandiri was entrusted to professionals, with technical assistance provided by Deutsche Bank. Half of the staff of the four banks was retrenched and branches were closed. Bank Mandiri has been recapitalized and is being prepared for privatisation.
The remaining three state banks (BNI, BTN and BRI) have been recapitalized after submitting restructuring plans and changing their management. The president’s nominated candidate for the position of president-director of BRI was rejected by Bank Indonesia in an ongoing tug of war between the president and the central bank. As well as the costs of recapitalization, the government also bore the costs of the guarantee program and shareholder settlement agreement. The bill for recapitalization amounted to 439 trillion rupiah (Box 5). Government bonds issued to repay the interbank borrowings of the forty-eight banks closed totalled 53.8 trillion rupiah. BCA shareholder settlement, in connection with the IBRA’s takeover of the Salim group’s loans in excess of the BCA’s legal lending limit, reached 60.9 trillion rupiah. About 114 trillion rupiah of liquidity credits are still outstanding from several banks. Total government bonds outstanding are therefore around 659.5 trillion rupiah.
V. Lessons from the restructuring program
The audits of the banks and the results of the bank restructuring program have revealed several important lessons.
The banks that survived the crisis were largely those that were not trading in foreign exchange or had only limited foreign currency transactions compared with their rupiah activity. This is despite the fact that some of the banks had weak management and credit practices. The large banks (both state and private) that provided foreign currency services and were exposed to considerable foreign exchange risk were badly hit by the crisis. This is not surprising given the volatility of exchange rates during the crisis. The extent of the damage caused by the crisis was often directly correlated with the level of foreign currency loans a bank had on its books. Panin Bank was an exception, given its higher CAR and its quick response to reduce its outstanding US dollar loans. Foreign banks were able to absorb these losses, given their worldwide and diversified earnings.
Foreign currency loans were the first to experience problems. Bank customers, attracted by lower interest rates and the steady 4–5 percent annual depreciation of the rupiah against the US dollar, had borrowed heavily. Most of these borrowers earned in rupiah. When the rupiah dropped in value by 70–80 percent, the cost of the loans rose fourfold.
Banks that had lent to their affiliates beyond the legal lending limit were also badly hit. Loans to affiliates were often not adequately scrutinised, collateralised, documented or appropriately priced and monitored. Practices to overcome the legal lending limit included loan swaps between banks and interbank placements to banks that would on lend to affiliates of the funding bank. As problem loans mounted, many banks were stuck with interbank borrowings that could not be repaid and were classed as ineligible for the government’s guarantee program.
The international audits of the banks under IBRA control show evidence that concentration of ownership is correlated with unsoundness. The audits show gross violation of the legal lending limit. It is estimated that an average of 50 percent of total lending of these banks was to their own group. The legal lending limit was 35 percent of equity and, assuming a CAR of 8 percent, it can be estimated that intragroup lending was nearly 20 times more than the legal lending limit.7
VI. Present conditions and remaining challenges
In January 2000 Bank Putera was closed, in May the IBRA sold 22 percent of its stake in BCA, and then Bank Danamon, which had been taken over by the IBRA, was merged with the other eight banks that had been taken over, in a move that amounted to the closure of these banks. Finally, the recapitalization of the remaining private and state banks was completed in several tranches. Despite the progress in rehabilitating and recapitalising the banking system, many challenges and problems remain. The banking sector is still dominated by the state banks, which have been recapitalized but are weak. The banking system remains fragile bccause of the weaknesses of the banks, the lack of progress in corporate debt restructuring, the limited economic recovery and continued political uncertainties.
The state now dominates the banking sector, either through state banks or because it has taken over or recapitalized private banks. Nearly 85 percent of the banking sector’s third- party liabilities are owned by the government, with thirteen banks taken over, the IBRA controlling 80 percent of the seven recapitalized banks, and four state banks remaining in operation. Although the state banks have been recapitalized and their managements restructured, these institutions still have many problems because of the political pressures they face.
The banking system also includes the former large private banks that were taken over, merged and recapitalized: BCA, Bank Niaga, Bank Bali and the ten merged banks under Danamon. Only BCA has undergone divestment, with 22.5 percent of its shares sold to the public for 0.9 trillion rupiah in May 2000. Plans are underway for divestment of the other banks, but the market is weak and there are many uncertainties surrounding the prospects for further divestment. There are also sixty-three smaller categories A private banks that have not been under IBRA control, twenty-six regional development banks and fifty joint-venture banks. The top four foreign banks are Citibank, Standard Chartered, ABN Amro and the Hong Kong Shanghai Bank.
The banking sector faces three main problems. First, even after recapitalization, with the exception of a few banks, capital adequacy standards are still low at close to or below the 4 percent minimum. New loans would easily lower CARs, as risk-weighted assets would rise while capital would stay more or less the same. Recapitalization pushed up the CAR by increasing the assets side of the balance sheet with government bonds, but there is no real cash to increase loans unless the bonds are sold. If the economy recovers and loan demand increases, banks will face the problem of having to liquidate government bonds in the secondary market to create funds for issuing loans. Government bonds still trade at a discount, which if too large, would reduce the CAR. The amount of recapitalization needed was underestimated, as it was to cover losses up to March 1999, but the bonds were not issued until May and for the remainder for 1999 and 2000, when the losses had gone up.
Second, earnings are still low, as reflected by very low interest margins. The majority of banks assets are government bonds, which have low yields (12–13 percent), while deposit rates are slowly rising with the weakening of the rupiah. Interest margins are often too low to cover operational costs.
Third, nonperforming loans are still high, even after the category-5 loans were transferred to the IBRA’s Asset Management Unit. The slow economic recovery has meant that corporations have not significantly improved their debt service capabilities. Another round of losses is probable.
With the economic recovery likely to remain slow, given the current political situation, bank earnings are not expected to be sufficient even to maintain the already low CAR levels. Nonperforming loans are still high and even category A banks are showing earnings fatigue. If CARs decline a second round of recapitalization may be needed. The question is where is this capital going to come from? The government budget is already spread thin and there are few public resources available.
This then raises the question of whether private investors (both foreign and local) would be prepared to buy into an Indonesian bank. Judging from the limited interest private investors had in the BCA, the failure of Bank Bali’s negotiations with Standard Chartered Bank, and the likely strong domestic reaction against foreign investment in the banking sector, divestment would have to be made very attractive to entice investors.
We predict that the likelihood of foreign investors entering the banking sector is small. Although foreign banks have discussed the possibility of buying into a local bank, their interest has been low. This has also been the case in other East Asian economies. Unlike in the Latin American banking crises, foreign investors will not be the source of new capital or better governance, management and expertise. Those foreign banks that do invest are more likely to look at less-risky medium-sized local banks, which have limited branch networks, or banks with a customer composition that closely resembles their own (i.e., the top end of the corporate and consumer market). Data on branches per bank in Table 7 compared by ownership type shows that even when foreign banks were allowed to extend their branch network to six other major commercial cities apart from Jakarta, they still opted for an average of two branches. These considerations narrow the target for foreign takeover to less than a handful of banks.
|July 31– Sept. 31||State banks||Private banks||Regional|
|Foreign and joint-|
|No. of Banks||7||5||160||83||27||26||44||39|
|Assets (Rp tn)||152.6||458.7||237.9||331.9||12.7||23.2||44.2||113.9|
|Loans (Rp tn)||120.0||98.6||179.8||73.6||8.2||9.3||32.1||67.5|
|Deposits (Rp tn)||68.6||237.7||146.0||228.0||8.0||17.6||6.9||25.2|
|Capital (Rp tn)||14.3||10.0||22.7||16.8||1.1||2.2||5.4||6.2|
Investment in Indonesian banks may attract financial investors who bring capital and a new management team into restructure a bank with the hope of reselling at substantial capital gain. However, the major issue remains as to how to make banks more attractive for investors and also what other conditions are needed for investors to want to come in.
VII. Priorities in Banking Reform and Restructuring
The priorities for bank restructuring are to complete the much-needed separation of nonviable from viable banks, to recoup losses (through asset-management and sales strategies, and state divestiture of banks), to implement new rules and regulations, and to develop an incentive-based system for the consolidation of banks.
A. Completing Restructuring: Developing Core Banks
The commercial banking sector remains weak and undercapitalised, more than 100 banks still exist (with low franchise values) and the state accounts for 85 percent of the liabilities of the banking sector. The state banks still hold a large amount of nonperforming loans, which remain undercapitalised and could increase. The situation is not likely to improve given the uncertainties that continue to plague economic recovery and corporate debt restructuring. Another round of purging nonperforming loans and recapitalising banks will probably be required, followed by a further consolidation of private and state banks to establish a smaller number of sound core banks.
There are a number of justifications for developing a smaller number of core banks. First, given the limited number of qualified and experienced employees in the banking sector, fewer banks would allow expertise to be consolidated. The IBRA has had to resort to placing foreign managers in the banks under its control and has used foreign advisers and consultants. There is a limit to the ability of foreign staff to meet the shortfall in domestic personnel, given the complexities of operating in Indonesia. Second, fewer banks would lessen the supervisory and monitoring tasks of the central bank (and the independent supervisory agency in the near future). Third, economies of scale could be achieved, given the high fixed cost of developing bank technology. Fourth, consolidation is likely to improve performance and profitability, add to the franchise value of the remaining banks and attract private investment in the banking system.
Consolidation should not be based on deciding the number of ‘ideal banks’ and picking winners with less than objective criteria. An incentive-based framework should be put in place to ensure risk-appropriate behaviour and good governance by the owners, managers, and supervisors of banks. A possible path for consolidation could be as follows.
Further mergers of state banks could be undertaken, with the nonperforming loans transferred to the IBRA’s Asset Management Unit or to a separate subsidiary of the IBRA set up for the nonperforming loans of state banks. The management of the newly merged state bank should be changed and top Indonesian personnel installed. Principles of transparency, disclosure, independence and proper credit evaluation for loans (with no political interference) should be followed, with an outside directorship or statutory body overseeing the bank. Public capital injections could be linked to a change in management, or capital could also be raised in the capital market. The injection of public funds and other steps to increase the franchise value of banks will hopefully attract the interest of private investors.
Another round of mergers and consolidations should be encouraged in the private banking sector. The few core private banks that emerge would, along with the two or three remaining state banks, form the backbone of the banking system. The remaining sixty-three smaller private banks that are not under IBRA control should also be encouraged to merge and consolidate to perhaps twenty or thirty banks, which would be regarded as second-tier or community banks focused on a different market segment. The consolidation of the private banks should be based on the following incentive-based framework.
First, international experience has shown that incentives are required to ensure that core banks are financially strong and behave in a risk-appropriate way. For instance, permission to trade in foreign exchange could be dependent on a high CAR (e.g., 15 percent). The higher capital requirement would encourage further consolidation in the sector, show that owners and managers are seriously committed to the banking business, and protect banks from unfair and imprudent competitors (Bossone and Promisel 1998). Additional incentives such as tax relief for bank mergers can also be provided.
Second, to ensure there is pressure on bank managers to follow principles of good governance, foreign exchange banks should be publicly listed. Measures of soundness should be published by the Central Bank and made accessible to the public. To ensure appropriate behaviour of the banking supervisors, the banks should be rated by both international and domestic rating agencies. A similar approach was adopted in Chile, where in addition to central bank risk ratings and valuations, two independent private accountancy firms must audit the banks every year and publish their findings. The central bank publishes ratings based on capital requirements and the quality of the banks’ assets.
Third, given the barriers to good governance from the concentration of ownership in the banking sector, excessive affiliate or group lending and the fact that banks finance their affiliated businesses, diversification of ownership will be important. As mentioned already, diversification of ownership through increased foreign ownership is likely to be limited. In addition, wider ownership may not provide effective oversight unless prudential regulations are adequately enforced (World Bank 2000). The divestment of government shares in the banks through the capital markets or to financial investors would be another avenue for achieving diversification. The funds raised can then be used for recapitalization. Given the past problems of excessive violation of the legal lending limit by business groups, the share of financial institutions that can be owned by business groups and the percentage of single ownership should be limited (e.g., to less than 49 percent).
Fourth, banks with foreign exchange licenses must have the capacity to manage risk. This implies strong and proper criteria for evaluating whether bank owners and managers are ‘fit and proper’. Bank Indonesia is currently implementing such criteria.
Fifth, foreign banks can bring in capacity, expertise and skilled staff, and can also introduce better governance and a more efficient corporate culture.
B. Strengthening Prudential Regulation and Supervision
The IMF reforms stipulated a comprehensive set of changes to prudential regulations. It is important to ensure that prudential regulations and bank supervision follow market- compatible principles (Bossone and Promisel 1998). Priorities for ensuring risk-appropriate behaviour would be as follows.
First, temporary limits are needed on asset growth and the growth of risky assets such as real estate to ensure risk diversification, as well as smooth and reasonable growth.
Second, capital requirements for banks in developing countries may need to be higher since they are operating in riskier environments. In the United States, for example, small community banks have higher capital requirements than major banks because their portfolios are not as diversified (Bossone and Promisel 1998). This requirement could lead, however, to perverse behaviour such as disintermediation, the booking of loans in offshore subsidiaries, and excessive investment in government bonds, which bear lower returns. These risks must be incorporated in any decision to strengthen capital requirements.
Third, prudential regulations could reward prudent and honest behaviour with positive incentives such as allowing lower CARs and less regulation or intervention for those institutions deemed to be well managed.
Fourth, sanctions for misconduct should be implemented and strictly enforced, whether concerning bank owners, managers or supervisors. In Chile, for instance, if the capital requirement is not met, a bank will be closed unless the uninsured creditors and supervisors agree to restructure the bank.
Finally, given the experience of the East Asian crisis, there must be exit mechanisms in place to allow orderly bank closures.
The problem faced by Indonesia and many other developing countries is that while incentive- based rules can be designed, poor enforcement of rules and underdeveloped legal and supervisory infrastructure can prevent their proper implementation. Governments and international institutions have been attempting to address this problem, but institution and capacity building will take time. In the meantime market participants can establish institutions that reduce risks, encourage risk-appropriate behaviour and avoid corruption. Such institutions could include private credit-rating agencies, independent corporate governance bodies and consumer or minority shareholder groups.
C. Incentive-Based Safety Nets
Safety nets are necessary to reduce the risk of a systemic crisis, but in their design and implementation the need to protect consumers have to be balanced against the need to minimise moral hazard and link the cost of protection to the risk. A blanket guarantee of deposits was perhaps the appropriate response for Indonesia given the crisis of confidence in January 1998, but the scheme needs to be replaced by one that is suitable for normal conditions and does not create moral hazard. To make the safety net for investors more credible, the deposit insurance scheme should not be a blanket guarantee and should only cover deposits (demand, time and savings) up to a certain maximum.
Experience in other countries has shown that financial institutions do take risks if they know they are protected by some kind of deposit insurance (Caprio and Klingebiel 1996a, b). This was the experience in Argentina in 1990 and Chile in the mid-1980s. If the deposit insurance scheme is extensive and the government is expected to fully protect depositors, bank managers will be less concerned about how their actions affect depositors, and depositors will not be as prudent in their choice of banks.
It is not easy to introduce safety nets that retain incentives and minimise moral hazard. In redesigning a deposit insurance scheme for Indonesia, incentives could be given to better performing banks by linking the annual premium payments to their risk profile. The risk profile could be measured by the level and quality of bank capital and the bank’s credit rating. This way the deposit insurance scheme is self-funded by the banking sector and is less of a burden on the government. The Chilean deposit insurance scheme introduced after the crisis provides a good example of such a scheme - partial coverage gives private debt holders an incentive to monitor banks and punish inappropriate behaviour.
D. Political Economy: State Divestiture of Assets and Banks
The most difficult problem facing countries such as Indonesia is that there are political and social constraints on instituting rapid restructuring and reforms that will strengthen the financial sector. As indicated above, the ownership of banks and major corporations now largely rests in state hands. Restructuring involves the redistribution of wealth and control directly through the ownership of assets and liabilities, and indirectly through taxation, wage and employment adjustments (Claessens 1998). A clear consensus has not emerged in Indonesia with regard to the division of ownership and control in the banking sector between the state and the private sector, between domestic and foreign companies, and between large companies and small and medium-sized companies. This division is likely to be politicised, given the predominantly Chinese ownership of banks and businesses. Until these issues are resolved, the restructuring progress is likely to continue to be slow and be plagued by problems and government interventions.
The Indonesian banking crisis offers a number of policy lessons on avoiding or minimising the build-up of vulnerabilities during integration with international financial markets. Financial liberalisation needs to be preceded or accompanied by a strengthening of institutions and prudential regulations. These regulations must be enforced, with sanctions in place for non-compliance. If financial integration takes place when exchange rate regimes are not flexible, prudential supervision of foreign currency exposures and risks or, at the very minimum, monitoring of the exposures is needed so that there is awareness if vulnerabilities become crucial. Policymakers must be aware of and be able to manage financial linkages in the macroeconomy that can exacerbate economic cycles. Concentration of bank ownership in Indonesia made it difficult to monitor behaviour and led to gross violations of prudential regulations. This implies a need to reduce single ownership and substantially improve prudential regulations, the qualifications of owners and managers, and corporate governance norms and regulations to strengthen information disclosure. Finally, moral hazard is more likely when there are no clear exit mechanisms and when banks are bailed out because they are ‘too big or too important’ to fail.
Indonesia’s responses to the financial and banking crisis show that liquidity support and lender-of-last-resort facilities need to be designed in a way that does not lead to misuse and is accountable. Furthermore, failure to sterilise liquidity fuelled inflation and capital outflows, further weakening the rupiah. Indonesia also had a shaky political situation, which added to the crisis of confidence in the banking sector. The usual relationship between capital flows and interest rates broke down and high interest rates could not stem the outflow of capital.
Indonesia’s experience underlines the importance of ensuring that closures of non-viable institutions are accompanied by clearly explained criteria for closure, consistency in implementation and a well-defined deposit guarantee scheme. The deposit guarantee scheme must be prepared in advance so that it is clear to depositors that they will get their money back or be able to transfer to quality banks (IMF 1999). If there is a massive crisis of confidence, a comprehensive deposit guarantee will be needed; however, it is debatable in Indonesia’s case whether the guarantee should have been extended to all liabilities of banks.
What are the lessons so far from bank restructuring in Indonesia? Recapitalization was necessary, but the selection of viability was questionable, including the lack of uniform treatment of state and private banks. The recapitalization program did not seem to be linked to a serious restructuring program, and as such the need for a second round of recapitalization has emerged. Thus, recapitalization alone is not sufficient to attract private capital unless there is confidence in the restructuring program.
Political interference in the reforms has been and continues to be a major problem leading to delays and inconsistencies. It is clear that restructuring cannot proceed without the full commitment of the government to support the agencies involved. The IBRA needs to be given sufficient independence to operate, be protected from lawsuits and have the means to attract the necessary expertise.
The difficulty of valuing nonperforming loans and other value-impaired bank assets during changing economic circumstances remains the most challenging task of the restructuring program. Yet accurate and realistic valuations are the key to reducing the fiscal burden of bank restructuring. Nonperforming loans need to be properly valued to avoid bailing out existing shareholders and undermining private sector recapitalization, and to encourage good governance of banks. The responsibility for asset disposal has been given to the IBRA, but there has been no consensus over the strategy of asset sales, especially with regard to the speed of disposal, or how to conduct the divestiture of state ownership in banks or assets.
With hindsight, the policy lessons that can be drawn from the build-up of vulnerabilities before the crisis and the management of the crisis are clear. It is important that the same mistakes are not repeated. The establishment of a sound banking sector that is part of a developed financial sector is going to take time, and will require substantial public resources and significant changes in institutions, regulations and the behaviour of the key participants.
Although the policy lessons and possible way forward may be evident, Indonesia faces serious fiscal constraints because of the magnitude of the distress in the banking and corporate sectors and the size of its external debt. Furthermore, Indonesia has the weakest institutional framework of the crisis-affected countries for resolving banking and corporate sector problems (Claessens 1998). While economic recovery is likely to continue to be slow, it is important that momentum of reform is strengthened and maintained.
BossoneB. and L.Promisel1998Strengthening Financial Systems in Developing Countries: The Case for Incentive-Based Financial Sector Reforms (Washington: World Bank).
CaprioG. and KlingebielD.1996a “Bank Insolvencies: Cross-Country Experience” Policy Research Working Paper No. 1620 (Washington: World Bank).
CaprioG. and KlingebielD.1996b “Bank Insolvency: Bad Luck, Bad Policy, or Bad Banking” in Annual World Bank Conference on Development Economics edited by BrunoM. and PleskovicB. (Washington: World Bank).
ClaessensS.1998Systemic Bank and Corporate Restructuring: Experiences and Lessons for East Asia (Washington: World Bank).
ClaessensS.DjankovS. and KlingebielD.1999 “Financial Restructuring in East Asia: Halfway There?” Financial Sector Discussion Paper 3 (Washington: World BankSeptember).
EnochC.2000 “Interventions in Banks During Banking Crises: The experience of Indonesia” IMF Policy Discussion Paper (Washington: International Monetary Fund).
GhoshS.M.Pangestu and M.Habir1999 “Macro-financial linkages and the build-up of vulnerabilities” country paper for an Asian Development Bank project on financial integrationunpublished.
LindgrenC.-J.T.J.T.BalinoC.EnochA.-M.GuideM.Quintyn and L.Teo1999Financial Sector Crisis and Restructuring: Lessons from AsiaIMF Occasional Paper No. 188 (Washington: International Monetary Fund).
World Bank1998East Asia: Road to Recovery (Washington: World Bank) Chapter 3.
Mari Pangestu, a consultant at the IMF’s Regional Office for Asia and the Pacific when this paper was prepared, is Director at the Centre for Strategic and International Studies, Jakarta. Manggi Habir, was President Director of PEFINDO Credit Rating Agency, Jakarta, when this paper was written; currently he is studying for a Master in Public Policy degree at the Kennedy School of Government, Harvard University. An earlier version of this paper was presented at the “Financial Markets and Policies in East Asia” conference, organized by the Australian National University and IMF Regional Office for Asia and the Pacific, held in Canberra on September 4-5, 2000.
See Ghosh et al. (1999) for more detailed discussions.
Including the famous Sumarlin shock that resulted in 8 trillion rupiah of state-bank deposits being converted into Bank Indonesia certificates (SBIs) and interest rates more than doubling.
The empirical evidence (Ghosh et al. 1999) on private consumption, for example, finds that the coefficient on domestic credit is statistically significant at the 12 percent level, while that on money is not significant at all.
Factual information taken from Lindgren et al. (1998), Appendix 1; World Bank (1998, 1999); and press releases.
Estimates based on interviews with IBRA officials.