The East Asian Crisis and Micro Finance

Transcription

1 The East Asian Crisis and Micro Finance The Experience of Bank Rakyat Indonesia Through July, 2000 Siti Sundari Nasution (Mrs.) IVP Director SBU Micro Banking BANK RAKYAT INDONESIA Jakarta, Indonesia September 11, 2000

2 The East Asian Crisis and Micro Finance The Experience of Bank Rakyat Indonesia Through July 2000 Executive Summary The Bank Rakyat Indonesia (BRI) Unit system is recognized as one of the largest and most successful micro-finance institutions in the world. Indonesia was more drastically affected by the East Asian monetary crisis which began in mid-1997 than were other countries in the area. It is therefore worthwhile looking at the BRI experience during the crisis - not only the experience in micro enterprise credit, but also in small, medium and corporate credit and in savings mobilization. Impact of the Crisis During the crisis, micro enterprises financed by BRI, proved more resilient than larger enterprises. BRI micro-finance borrowers have continued to pay back over 97% of loan installments which have fallen due. This is a better record than the small, medium and corporate customers of BRI have attained. From January 1998 to the end of March, 1999, BRI s there was a slow decline of about 3.5% in the total outstanding micro credit; in previous years there was normally an increase in outstanding from year to year. This appears to have been the result of slower demand from borrowers who felt uncertain about their business activities in the face of inflation of roughly 80% per year which prevailed from January to September Such inflation had not been experienced in Indonesia since the 1960's. Borrowing began to increase again in early April 1999, about eight months after prices began to stabilize. The average size loan increased as borrowers apparently tried to return to their previous levels of real turnover. In the sixteen months from the beginning of April 1999 to the end of July 2000, BRI s micro credit portfolio has grown by 50%. BRI as a whole experienced a rapid increase in savings (passbook accounts, giro and time deposits) during the worst of the crisis in 1998; mainly the result of a flight to safety as several private banks were closed down or taken over by the government s Bank Reconstruction Agency. Savings in the BRI Units, the micro finance arm of BRI, grew at about the same rate as total savings in BRI, though there was a shift toward time deposits when interest rates on these were very high and then back to passbook savings as the time deposit rates returned to levels close to that of savings. i

3 Reasons for Strong Micro Enterprise Repayment Performance The following factors appear to have been important in the excellent performance of BRI s micro enterprise borrowers during the crisis: First, the micro enterprise loans are all installment loans adjusted to the borrower s cash flow. As borrowers pay back the installments, they normally reinvest profit to maintain the level of business which they had reached with the loan proceeds. As a result, most of these micro enterprise owners have built up their equity and lowered their loan leverage, thereby making their enterprises less vulnerable financially to external disturbances such as a sudden increase in interest rates caused by a shift in monetary policy or a sudden decrease in demand from customers who have missed a crop or become unemployed. Second, the micro enterprises are more likely to be engaged in purchase and sale of domestically-produced essentials for which demand is more stable than it is for less essential goods and services and goods with higher imported content which are more often traded by small, medium and large enterprises. As their real income declined during the recession, and prices of goods with high imported content rose, consumers moved down market to meet their daily needs, substituting the domestically-produced goods handled by micro enterprises for the goods handled by larger enterprises. Third, the rural sector was less affected by the monetary crisis than the urban areas, but more affected by a severe drought that struck Indonesia at about the same time. The effect of the drought on farmers demand for other micro enterprise services and products was largely overcome by two reasonably good rice crops after the drought. Areas which raise export crops experienced something of a boom because of the devaluation of the rupiah. Finally, the micro borrowers appear to value their access to credit and savings services very highly. They were reluctant to break their banking relationship with the BRI Units even if they had to squeeze consumption to make their loan repayments on time. BRI was careful to keep the micro credit window open so that those who paid on time were able to borrow again if their enterprise justifies it. BRI had learned from an earlier economic downturn when this window was partially closed for fear that savings were in decline; the result was an increase in the number of borrowers who did not pay on time. ii

4 The East Asian Crisis and Micro Finance The Experience of Bank Rakyat Indonesia (BRI) Through July, 2000 Contents I. The organization of BRI into Micro, Retail and Corporate Strategic Business Units (SBUs). II. III. IV. The effect on micro enterprises in Indonesia of monetary crisis and the drought which coincided with and compounded the crisis, both directly and through the monetary policies pursued under IMF agreements. The contrasting experience of the BRI s three lending SBUs during the crisis. Reasons for strong micro enterprise repayment performance Charts Chart 1 Chart 2 Chart 3 Chart 4 Chart 5 Chart 6 Chart 7 Inflation and 1 Month SBI Rates. SBI, Time Deposit, and SIMPEDES Interest Rates. Effective Rates on KUPEDES and SME Loans. KUPEDES Productivity and Units Cost of Funds. BRI Unit System Loss Ratios. Savings & Credit at BRI Units (in Rupiahs). Savings & Credit at BRI Units (Number of Accounts). Note: Usually a paper such as this would state some information such as outstanding loan volume or total savings in both rupiahs and U.S. dollar terms, so that readers could compare these with micro finance operations which they know in other countries. However, during the crisis period, the exchange rate has been extremely volatile, and the meaning of changes in the dollar value of these volumes is unclear. For example, the dollar value of the Units KUPEDES loan portfolio has swung from $ 1.7 billion in June 1997, before the crisis, to $ 308 million in June 1998, to $ 724 million at the end of June BRI Unit customers deal almost exclusively in rupiahs and during the period, as explained in the text and shown in charts 6 and 7, there has been almost no change in the rupiah loan volume or the number of KUPEDES borrowers. Therefore, this paper states all values in rupiahs and not in dollars. 1

5 I. Background on Bank Rakyat Indonesia (BRI) BRI is a state owned bank. It is organized into four Strategic Business Units (SBUs): SBU Micro Banking is in charge of 3,694 Units. A Unit is a small bank office with four to eleven staff. The Units also maintain 319 cash posts which are open three to five days a week to receive and pay out savings and to receive loan repayments. Each Unit is a separate profit/loss center with its own balance sheet and profit and loss statement. The Units make loans of Rp 25 thousand to Rp 25 million. There were million micro loans Account at the end of July The Units also provide savings, giro and time deposits accounts; there were 25.1 million accounts in the Units at the end of July SBU Retail Banking is in charge of the 323 branches of BRI. The branches provide full banking services. They make small business loans up to Rp 3 billion, loans against salaries and pensions, and program loans (channeling of loans for government - such as those to co-operatives or to the government agency which provides agricultural price support). The branches also provide savings, giro and time deposit accounts. The branches have recently introduced a simple on-line facility for holders of savings accounts. SBU Corporate makes loans of more than Rp 3 billion, including dollar loans. SBU Treasury and Investment handles treasury functions including maintaining of BRI primary reserves in Bank Indonesia, purchase of Bank Indonesia certificates (SBIs), interbank lending and borrowing, and controls BRI financial subsidiaries. 2

6 II. The Monetary Crisis The monetary crisis in Indonesia began (or was signaled) in mid 1997 when Bank Indonesia, the central bank, decided that its foreign reserves were insufficient to support the exchange rate, at that time Rp 2,450 = $1. By mid 1998, the exchange rate had plunged to Rp 14,900 = $1. By the end of June 1999, the crisis appeared to have ended. Present predictions from Bank Indonesia are for a growth rate of 4% to 5% for the year 2000, a stable exchange rate between Rp 7,500 and Rp 8,500 to the dollar, and an inflation rate of less than 10%. II.A. There were at least three causes of the crisis which are relevant to the situation of BRI: 1. There was a severe drought in 1997/98 and one entire crop of rice was lost. This resulted in both rising prices and a shortage of demand from farmers for the products and services of micro and other enterprises. 2. For many years Indonesia has maintained an open capital account. In recent years, the rupiah had been allowed to depreciate about 5% a year. High rupiah interest rates compared to rates in neighboring countries plus the seeming certainty of the rate of depreciation brought in large amounts of short-term money. Once there was any loss of confidence, this short-term money could and did move out very quickly, triggering Bank Indonesia s decision to allow the market to set the exchange rate. 3. The large Indonesian business conglomerates borrowed dollars abroad, running up a short-term foreign exchange debt more than twice the total foreign debt of the Government, most of which consists of long term concessional loans. Bank Indonesia was unable to keep track of this private foreign debt, which was more than twice the country s foreign exchange reserves, and was therefore uninformed of possible short term demands against the country s foreign exchange reserves if these loans were not rolled over. A large amount of this debt was used for real estate investment, most of which was not complete, much less sold or rented, at the time the crisis began. Even had the funds been used for financing relatively quick return activities, these borrowers still could not have repaid at the new exchange rate; most are now technically insolvent. 3

7 Unlike the Latin American debt crisis in the 1980's and the Mexican crisis in 1991, government over-spending was not one of the root causes of the Indonesian crisis. Including the proceeds of long-term foreign aid loans as income, and loan repayment as expenditure, the budget had been balanced or in surplus since the late 1960's. II.B. After Indonesia began to receive assistance from the IMF, monetary policy for several months concentrated on raising interest rates to support the exchange rate and to control inflation. Since the Government of Indonesia has not issued bonds that were traded in the market, the main monetary policy instrument of the central bank is the Sertificat Bank Indonesia (SBI). The rate for a one-month SBI was raised to over 70% in July and August of Apparently this policy was considered a success; the one-month SBI rate has been lowered rapidly beginning in October By the end of June 1999, this rate had been lowered to 13.86% and has been fairly stable since. It was at 13.53% at the end of August (See Chart 1 Inflation and 1 Month SBI Rates ). At the start of the crisis, BI gave an unlimited guarantee of all deposits in all banks. The result was provision of enormous BI liquidity credits to many banks to pay off their depositors. There were runs on several private banks; the two largest were taken over by the Indonesian Bank Restructuring Agency (IBRA). Between November 1, 1997 and March 13, 1999, sixty-one banks were closed and eleven were nationalized to be consolidated before resale to the original owners or to the market. Four of the seven state banks were merged. All four surviving state banks have been wholly or partially have been recapitalized through the issuance of Government bonds. Two thirds of the BRI recapitalization bonds were received by the end of July 2000 and the other third are expected at the end of September. II.C. The exchange rate moved from Rp 2,450 = $1 before the crisis to over Rp 14,900 = $1 at the time of the worst panic. In June 1999, it improved to Rp 7,433 = $1 following relatively peaceful elections in May. It deteriorates each time there is an outbreak of violence anywhere in the country, the appearance of political uncertainty, or a delay of the release of IMF funds. Basically it has moved between Rp 6,500 and Rp 8,500 = $1 most of the time since mid-june, At the end of August 2000 the exchange rate was Rp 8,310 = $1. II.D. BRI along with other Indonesian-owned banks raised their time deposit rates in line with the SBI rate until May 1998, when they peaked at 60% for onemonth time deposits. Because of the large inflow of deposits, BRI set time deposit rates well below the SBI rate until October From October, 1998, BRI has kept rates just below the SBI rate as this has moved down. As of the end of August, 2000, BRI was paying 10.5 % per annum on a one-month time deposit; 4

8 while the one month SBI rate is %. Top rates on SIMPEDES and other passbook savings in BRI moved from 13% before the crisis to a peak of 20% and have now declined to 9% at the end of August (See Chart 2 SBI, Time Deposit, and SIMPEDES Interest Rates.) II.E. Minimum interest rates on most small, medium, and large loans doubled from a range of 17% to 19% per annum in mid 1997 to a high of 36% from September 1998 through May, They have since fallen steadily, reaching a level of 19% to 26% by March, 2000, and have remained steady since that time. Rates on SME loans apply almost immediately, as the loan agreements provide for adjustable rates. The interest rate on KUPEDES, the Units micro credit instrument, was raised from 31.72% before the crisis to 45% at the 1st of September, Since then the rate has been lowered in steps all the way back to 31.72% from the 1st of September, 1999 and have remained at that level. (See Chart 3 Effective Rates on KUPEDES and SME Loans.) New KUPEDES interest rates apply only to new loans, while new rates on savings apply immediately to all existing savings accounts or to time deposits taken out or renewed after the rate change. Thus there is a lag in the increase in actual credit interest rates and a squeezing of the spread as interest rates move up and an increase in spread as interest rates move down. (See Chart 4 KUPEDES Productivity and Units Cost of Funds.) 5

9 III. Credit Repayment Experience of BRI During the Monetary Crisis Credit repayment experience of the Micro, Retail and Corporate SBUs of BRI has not been uniform. III.A. SBU Corporate 1. Dollar Loans - Because of the drastic devaluation, the rupiah value of SBU Corporate dollar loans ballooned from about Rp 3.5 trillion to about Rp 16 trillion. In the balance sheet of BRI, this meant that the capital adequacy declined sharply and one or two loans were more than the legal lending limits (20% of the bank s equity). As a result, the Bank Indonesia rating of BRI went down. The rupiah value of these loans has again declined as the Rupiah has moved from Rp 14,900 = $1 to a level of between Rp 7,500 to Rp 8,500 = $1. Many of these loans were for hotels, which were good foreign exchange earners before the tourist trade declined. These hotels are still experiencing low occupancy rates, though these have improved since June, Even if there had not also been an increase in interest rates on these loans, it is extremely unlikely that they could have been repaid at Rp 7,500 = $1, a near tripling of their previous rupiah value, much less at the Rp 14,900 = $1 rate which was the lowest point of the rupiah. At present, most are non-performing. 2. Rupiah Loans - Performance on SBU Corporate s rupiah loans varies according to whether the corporate customer had also borrowed abroad in dollars, whether the loan proceeds were put into office construction or other activities with a delayed return, and whether the customer s business sold goods or services for which demand has decreased drastically. Most of the corporate portfolio has been written off and transferred to the Asset Management Unit of the Indonesian Bank Reconstruction Agency (IBRA). 6

10 III.B SBU Retail Retail lending includes several different categories of loan and the performance has varied among these categories. 1. Small and Medium Business Loans - The small and medium business loans of SBU Retail are all rupiah loans. Most were on a line of credit basis rather than paying back in regular installments which include principal. The figures indicate that these have performed better than the large corporate loans, but about 41% had missed at least one interest payment and were therefore classified. These borrowers faced a combination of sharply declining demand, doubling of interest rates, and, in some cases more than a tripling of the rupiah price of imported inputs. As interest rates have moved back to pre-crisis levels, repayment performance has improved. After transfer to IBRA of loans considered as unlikely to be repaid and repayment of some loans as economic conditions improved, the percentage of the remaining retail business portfolio which is classified has fallen to 12.34% as of the end of July, Line of credit loans in this portfolio are being replaced by loans requiring regular payments. The portfolio has recently begun to expand again. 2. Consumer Loans - An important business for SBU Retail is making installment loans to employees and pensioners with repayment through payroll deductions or deduction from pensions which are paid through BRI. These are classified as consumer loans. Most of the payroll deduction loans are to government employees. Repayment has held up very well although real wages have dropped. The same is true of loans repaid through deduction from pensions paid through BRI. Repayment on these loans is almost as good as those for micro enterprise credit in the Units. There were some problems with the relatively small number of payroll deduction loans to employees of private businesses where employees have been laid off. 3. Program Loans - BRI channels loans for various government programs. In the past these were financed by liquidity credits from Bank Indonesia, and Bank Indonesia and/or the Government carried most of the 7

11 risk. Interest rates for the liquidity credits financing the programs, as well as interest rates to the final borrowers, were set by BI or the Government. The performance of program loans varies according to the objectives of the program and the quality of the intermediary between BRI and the final borrower. The drought reduced repayment of credit to farmers passed through the Government sponsored co-operatives. On the other hand, it increased the size of credit passed by BRI to BULOG, the government logistical agency, which had to import much more rice than usual to control prices; loans to BULOG are 100% guaranteed by Government. Following the drought, Government pushed a large amount of credit through the cooperatives for farmers with a 95% or 100% guarantee from government. There was clearly a need for increased farm credit after the farmers savings had been drained away by the drought. However, the increase was very large, an increase of over thirty times from the previous year s total. This credit was due back from the cooperatives on March 31, Only 30% has been repaid by July. 8

12 III.C. SBU Micro The BRI Units have a single credit instrument, General Rural Credit (KUPEDES). All KUPEDES loans require repayment in installments. Over 30 repayment schedules are available to fit the cash flows of borrowers with various kinds of enterprise such as agriculture or fish production, trading of various kinds, handicraft and small industrial production, or services. KUPEDES is also available for employees whose repayment is from salary deduction; most of these loans have been used for home improvement, enterprises of family members, or education of children. 1. KUPEDES Repayment Experience to end of July 2000 During the drought and monetary crisis, KUPEDES borrowers have continued to pay back more than 97.5 % of everything that has fallen due. Each month the MIS of SBU Micro produces three Loss Ratios which measure the total amount overdue, including everything which has been written off, divided by the total amount which has fallen due. The Short Term Loss Ratio (STLR) is for the particular month. It tends to rise each year at the major Muslim holiday of Lebaran, which was in early January this year and in mid-january in In March, April, May, and again in September and October 1999, the STLR was below 1%, but increased in November and December 1999 as it usually does just before Lebaran. It remained above 2% during January 2000, but then dropped to 0.28% in February and 1.19% in July. The Long Term Loss Ratio (LTLR) shows results since KUPEDES lending began in February It is of course not sensitive to movements within a single month period. The LTLR stood at 1.95% at the end of July The Twelve Month Loss Ratio (TMLR) shows the same ratio for the past twelve months, and is the most useful in detecting changes in trends. The TMLR was at 1.72% at the end of December 1999 and has declined to 1.35% at the end of July Interestingly, this ratio has declined fairly steadily during the crisis and has been below 2% from November 1998 through July (See Chart 5 BRI Unit System Loss Ratios ) 9

13 2. Change in Size of the KUPEDES Portfolio Total KUPEDES outstanding reached a peak in January 1998, when Lebaran occurred at the end of the month. A peak in outstanding in the 30 days or so month before Lebaran followed by a decline thereafter is normal. Traders borrow to stock up before Lebaran when much of the whole year s purchase of consumer items such as clothing occurs; they usually do not borrow just after Lebaran. From this peak in January, 1998, outstanding was relatively stable for a year and a half, not surpassing the previous peak until June 1999; in previous years there was normally an increase in outstanding from year to year. In the sixteen months from the beginning of April 1999 to the end of July 2000, BRI s micro credit portfolio has grown by 50%. The initial leveling off of outstanding was not the result of any policy instruction from BRI policy makers. BRI did not restrict KUPEDES lending during this crisis, in contrast to their actions in a somewhat similar, though milder situation, in The different reaction is partly because of the high inflow of savings during 1998 and 1999, as will be discussed below. The small decline and leveling off of KUPEDES outstanding during the early part of 1998 was not caused by a rise in the interest rate. BRI was deliberately slow in raising KUPEDES rates as the cost of funds increased; instead they accepted decreased interest spreads because of the psychological importance to the borrowers of stability. The rate for all new KUPEDES loans was raised from 1.5% to 2.2% per month flat (on the original balance) only from September 1st, It was cut back to 1.95% per month flat on April 1st, 1999 and then to 1.65% flat on July 1st, It returned to 1.5 % flat on September 1st, 1999 and has remained there through September The main reason for the lack of an increase in the portfolio during this period appears to have been slower demand from borrowers who felt uncertain about their business activities in the face of inflation of roughly 80% a year which prevailed from January to September Such inflation had not been experienced in Indonesia since the 1960's. Unit loan officers and managers reported that the reaction of some of their good repeat borrowers was to postpone borrowing again and instead to run down their stocks and place some of their capital in savings at the Unit. Borrowing began to increase again in April 1999, about six months after 10

14 prices began to stabilize and borrowers began to try to return to their previous real turnover level. 11

15 3. Savings in the Units During the worst of the monetary crisis, there was a very rapid increase in savings, both in the BRI Units and in the rest of BRI. In the Units, total savings increased from Rp 8.3 trillion at the end of October 1997 to Rp 17.7 trillion at the end of June 1999, and then declined slightly to the end of December During 2000 total savings have increased again, reaching Rp 18.5 trillion at the end of July. The large increase in savings through mid-1999 was partly a flight to safety, even though all deposits at all banks had been guaranteed by Bank Indonesia. Many private banks experienced runs and were closed or taken over by the bank reconstruction agency (IBRA). Even with the BI guarantee, savers appeared to want to avoid the more complicated procedures which would impede their withdrawal of their savings when a bank is being restructured or closed. There was also a change of composition in savings at the Units. In response to monetary policy, the rate on a one month time deposit in BRI jumped from 19% in January 1998 to 57% in September of that year. The top rate paid for SIMPEDES, the main Unit passbook type savings instrument, has moved up from 16% to 20% during the same period. While the rupiah total of passbook savings remained relatively stable, time deposits grew from 12.6% of savings in the Units in February 1998 to 40.5% in April As time deposit interest rates fell after that time, there has been a shift from time deposits to passbook savings accounts. Time deposits constituted 17.8 % of total rupiah savings in the Units at the end of July Interestingly, the rupiah value of passbook savings continued to grow through August 1998, fell slightly through January 1999 and then began to rise again. From January 1999 to July 2000 passbook savings increased from Rp 9.6 trillion to Rp 15.0 trillion. (See Chart 6 Savings & Credit at BRI Units (in Rupiahs).) The number of passbook savings accounts grew throughout the period, with the exception of one month, December From 18.0 million accounts at the end of 1997, the number of passbook savings accounts grew to 24.7 million at the end of July (See Chart 7 Savings & Credit at BRI Units (Number of Accounts).) 12

16 IV. Reasons for Strong Micro Enterprise Repayment Performance There appear to be at least four factors which explain the excellent repayment performance of BRI s micro enterprise borrowers, as compared to small, medium and large enterprise borrowers: First, the micro enterprise loans are all installment loans adjusted to the borrower s cash flow. As borrowers pay back the installments, they normally reinvest profit to maintain the level of business which they had reached with the loan proceeds. As a result, most of these micro enterprise owners have built up their equity and lowered their loan leverage, thereby making their enterprises less vulnerable financially to external disturbances such as a sudden increase in interest rates caused by a shift in monetary policy or a sudden decrease in demand from customers who have missed a crop or become unemployed. Second, the micro enterprises are more likely to be engaged in purchase and sale of domestically-produced essentials for which demand is more stable than it is for less essential goods and services and goods with higher imported content which are more often traded by small, medium and large enterprises. As their real income declined during the recession, and prices of goods with high imported content have up, consumers moved down market to meet their daily needs, substituting the domestically-produced goods handled by micro enterprises for the goods handled by larger enterprises. Third, the rural sector was less affected by the monetary crisis than the urban areas, but more affected by a severe drought that struck Indonesia at about the same time. The effect of the drought on farmers demand for other micro enterprise services and products was largely overcome by two reasonably good rice crops after the drought. Areas which raise export crops have experienced something of a boom because of the devaluation of the rupiah. Finally, the micro borrowers appear to value their access to credit and savings services very highly. They were reluctant to break their banking relationship with the BRI Units even if they had to squeeze consumption to make their loan repayments on time. BRI was careful to keep the micro credit window open so that those who pay on time were able to borrow again if their enterprise justified it. BRI had learned from an earlier experience when this window was partially closed for fear that savings were going to decline; the result was an increase in the number of borrowers who did not pay on time. This mistake was not repeated by BRI during the 1997 to 1999 crisis. 13