China - Third and Fourth Rural Credit Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 19519 PERFORMANCE AUDIT REPORT CHINA THIRD RURAL CREDIT PROJECT (CREDIT 1871 - CIIA) FOURTH RURAL CREDIT PROJECT (LOAN 3265-CHA/CREDIT 2182 - CHA) Operations Evaluation Department Sector and Thematic Evaluations Group This document has a restricted distribution and may be used by recipients only in the performance their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Currency Equivalents (annual averages) Currency Unit = Yuan (Y) 1987 US$1.00 Y 3.7 1988 US$1.00 Y 3.7 1989 US$1.00 Y 4.7 1990 US$1.00 Y 5.2 1994 US$1.00 Y 8.5 1996 US$1.00 Y 8.3 Abbreviations and Acronyms ABC Agricultural Bank of China ADB Asian Development Bank ADBC Agricultural Development Bank of China DCA Development Credit Agreement ERR Economic rate of return FRR Financial rate of return FSTAP Financial Sector Technical Assistance Project FX Foreign exchange IBRD International Bank for Reconstruction and Development ICR Implementation Completion Report ID Institutional development IDA International Development Association MOF Ministry of Finance OED Operations Evaluation Department OD Operational Directive OP Operational Policy PAR Performance Audit Report PBC People's Bank of China RCC Rural Credit Cooperative RC I Rural Credit I Project (CR 1462-CHA) RC II Rural Credit II Project (CR 1642-CHA) RC III Rural Credit III Project (CR 1871-CHA) RC IV Rural Credit IV Project (CR 2182-CHA; LN 3265-CHA) SAR Staff Appraisal Report TA Technical assistance TEG Technical Expert Group Fiscal Year July 16 - July 15 Director-General, Operations Evaluation : Mr. Robert Picciotto Director, Operations Evaluation Dept. : Ms. Elizabeth McAllister Manager, Sector and Thematic Evaluations Group Mr. Gregory Ingram Task Manager : Mr. Madhur Gautam FOR OFFICIAL USE ONLY The World Bank Washington, D.C. 20433 U.S.A. Office of the Director-General Operations Evaluation MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report - China: Third Rural Credit Project (Credit 1871-CIIA) and Fourth Rural Credit Project (Loan 3265-CIA/Credit 2182-CHIA) Attached is a combined Performance Audit Report prepared by the Operations Evaluations Department (OED) for the above two projects. The Third Rural Credit project (RCIII) was supported by Credit 1871-CHA) for US$170 million. The Fourth Rural Credit project (RCIV) was supported by Loan 3265-CHA and Credit 2182-CHA in the amounts of US$75 million and US$200 million, respectively. RCe was approved on January 26, 1998, and closed on June 30, 1994, 12 months later than planned. RCIV was approved on October 30, 1990, and closed on December 31, 1996. Both projects were fully disbursed. While both projects aimed to stimulate growth and diversification of agriculture by expanding the availability of medium- and long-term credit for profitable investments, the scope of their institutional development objectives was different. RCIII sought to increase the operational efficiency of the Agricultural Bank of China (ABC) through institutional development (ID). RCIV sought to promote the efficiency of rural financial intermediation through the introduction of market-oriented policies, and the development of institutions concerned with rural savings mobilization and lending. The ICR for RCIV, however, refers to a third objective reform of the broader financial sector policies. The audit disagrees based on agreements reached at negotiations and the project loan documents: the project was prepared and appraised as a real sector project and its ID objectives were limited to the rural financial sector. The projects substantially achieved their physical objectives, but the ID objectives were only partially achieved. Implementation of the physical components was smooth and the projects disbursed well, financing investments that were both financially and economically profitable. Supervision missions were generally able to monitor and effectively deal with implementation issues. ABC was responsive. Supervision, however, was more concerned with physical implementation and afforded low priority to loan covenants concerned with institutional development. This was particularly true for RCIV. One issue to which the ICR for RCIV assigns significant importance is the apparent violation of a loan covenant on the lending rate for project funds from the Ministry of Finance to ABC. The rate charged (5.4%) was lower than the stipulated IBRD rate. However, the audit concludes that the supervision missions tacitly agreed to the lower rate. On the substantive side, the action does not appear to have had any major distortionary impact on the Chinese financial markets, either by injecting subsidized funds into the system or by mis-allocating resources within the system. This document has a restricted distribution and may be used by recipients only In the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. 2 In other instances, however, the supervision missions were deficient in dealing with loan covenants. For example, both projects specifically included requirements for positive real interest rates, yet none of the supervision missions noted the impact of high inflation in several implementation years. In capacity building, both projects made substantial progress in improving the skills of ABC staff in sub- project appraisal techniques, which is likely to have a significant payoff as ABC gets ready for commercialization. On the other hand, the studies program, designed to improve ABC's operations and policies, had little impact; transfer of technology was limited; the technical assistance was underutilized, primarily due to the reluctance of ABC to use foreign consultants; and progress on the development of RCCs was limited. The true financial position of ABC or its portfolio quality is unknown for lack of reliable data conforming to international banking standards. The financial impact of RCIII is likely to be neutral on ABC (although it will be negative on the government), while that of RCIV on ABC is uncertain (neutral on the government). Based on available data, the quality of the project sub-portfolios appears to be reasonably good. The projects' impact on the financial management of ABC is mixed. Some ID and financial policy objectives were either beyond ABC's control (interest rate rationalization, provisioning practices) or were not advisable (recapitalization). Others were either reduced in scope or made redundant by being taken over by another project. Other aspects, aimed specifically at upgrading ABC's financial management and operations, were not achieved. In this respect, consistent violation of some loan agreements, and the neglect of others, is a notable shortcoming of RCIV. The audit concurs with the ICR in rating the outcome of RCII as satisfactory. However, for RCIV, the audit disagrees with the ICR rating of unsatisfactory and rates it as marginally satisfactory. The primary reason for the difference in the rating of RCIV is that the ICR places greater weight on the non-achievement of financial sector policy objectives and the inefficiency in financial intermediation implied by the subsidy on RCIV funds from MOF to ABC. In the view of the audit, the project did not have any broad financial sector policy objectives and it does not appear that there was either a significant subsidy element or any major distortions as a result of the MOF lending rate to ABC. Yet, in light of the modest achievement of the relevant institutional development objectives, the audit does not consider the outcome of RCIV to be fully satisfactory. The audit agrees with the ICRs' ratings for institutional development impact (modest for both), sustainability (likely for RCIII and uncertain for RCIV) and borrower performance (satisfactory for RCHI and unsatisfactory for RCIV). On Bank performance, the audit agrees with the satisfactory rating for RCIII, but disagrees with the ICR's satisfactory rating for RCIV and instead rates Bank performance as unsatisfactory. In RCIV, Bank performance in identification and preparation assistance was satisfactory. However, for both appraisal and supervision, it was unsatisfactory. The appraisal was marked by a prolonged debate on the role of the project as a vehicle for reform of the entire financial sector. Eventually, financial covenants included in the loan agreements were not entirely consistent with the concurrent dialogue with the government on financial sector reforms. Supervision was infrequent, had a high turnover of staff, and attached a relatively lower priority to policy and institutional development objectives. As a result, a number of covenant violations were either overlooked or not pursued. A lack of action on key financial covenants signaled to the borrower that loan agreements were not to be taken seriously. 3 The lessons of broad applicability from these operations are: (a) Either the projects should be kept simple enough so as not to over extend the supervision capacity of the Bank or the borrower, or, once a project is designed in a particular way and accepted by the Bank, the resources to supervise it must be made available. (b) Supervision missions must be well versed with and should diligently pursue all aspects of project objectives, design and legal agreements. (c) For effective technology transfer, it is not sufficient to impose procurement guidelines and provide access to international markets, but knowledge of best practices and technical advice must be made available to investors and lenders. (d) Commercial financial institutions are likely to focus on more profitable large-scale and urban projects. The impact of this on rural poverty reduction is an important unanswered question. (e) It is difficult to introduce commercial banking only in one sub-sector, without regard to the overall sectoral and institutional environment. (f) It is critical to maintain an accessible institutional memory, especially in times of high staff turnover. Attachment C 2  Contents Principal Ratings.....................................................m Preface ............................................................ v 1. Introduction......................................................1 2. The Project ......................................................2 Objectives ................................................... 2 Components ..........................2....... ...............2 Costs ...................................................... 3 Evaluation of Project Objectives and Design .................. .........4 3. Implementation ................................................... 6 RCIH...................................................... 6 RCIV......................................................6 4. Outcomes and Findings ................................... ...........9 Physical Achievements..........................................9 Additionality.................................................10 Capacity Building.......................... ...................10 Financial Impact..............................................11 Profitability.....................................................12 Financial Management.........................................12 Portfolio Quality................................ ...................12 Impact of Inflation.............................................13 Commercialization.............................................14 ICR Quality................................................. 14 5. Ratings......................................................... 15 Outcome................................................... 15 Institutional Development....................................... 15 Sustainability................................................ 15 Bank Performance............................................. 16 Borrower Performance.......................................... 16 6. Lessons... ......................................................... 17 Annex A. Basic Data ............. ....................................................................... 19 Annex B. Interest Rates and Inflation ........ . ................ ........... 23 Annex C. Borrower Comments ................................... .......24  Ill Principal Ratings Rural Credit Project I ICR Audit Outcome Satisfactory Satisfactory Sustainability Likely Likely Institutional Development Modest Modest Bank Performance Satisfactory Satisfactory Borrower Performance Satisfactory Satisfactory Rural Credit Project IV ICR Audit Outcome Unsatisfactory Marginally Satisfactory Sustainability Uncertain Uncertain Institutional Development Partial Modest Bank Performance Satisfactory Unsatisfactory Borrower Performance Deficient Unsatisfactory iv Key Staff Responsible Rural Credit Project III Task Manager Division Chief/ Department/Country Sector Manager Director Appraisal Ramesh Deshpande Joseph Goldberg Shahid Javed Burki Midterm Ramesh Deshpande Joseph Goldberg Shahid Javed Burki Completion Rick Scobey Joseph Goldberg Nicholas Hope Rural Credit Project IV Task Manager Division Chief! Department/Country Sector Manager Director Appraisal Ramesh Deshpande Joseph Goldberg Shahid Javed Burki Midterm Hoonae Kim Joseph Goldberg Shahid Javed Burki Completion Kay Hill/ Rick Scobey* Geoffrey Fox Yukon Huang *The ICR was prepared by Ramesh Deshpande, who was also the team leader for the ICR mission. V Preface This is a combined Performance Audit Report (PAR) for the Third and Fourth Rural Credit projects in China. The Third Rural Credit Project (Credit 1871 -CHA) was approved on January 26, 1988 for SDR 123.8 million ($170 million) and made effective on June 17, 1988. The credit closed on June 30, 1994, 12 months beyond the date originally envisaged, June 30, 1993. It was fully disbursed, and the last disbursement took place on July 7, 1994. The Fourth Rural Credit Project was supported by Loan 3265-CHA in the amount of $75 million equivalent and Credit 2182-CHA in the amount of SDR 143.7 million ($200 million equivalent). The project was approved on October 30, 1990, and made effective on March 22, 1991. The loan and credit were closed on December 31, 1996. These were fully disbursed, and the last disbursement took place on February 5, 1997. The audit is based on the information contained in the Implementation Completion Reports (ICRs) for the two projects, the Staff Appraisal Reports (SAR), the President's reports, the credit and loan documents, a study of the projects' correspondence files, and discussions with government and borrowing agency officials, and Bank staff. An OED mission visited China in December 1998 to discuss the implementation and development effectiveness of the two projects with the relevant officials and sub-borrowers. The cooperation and assistance provided by the Government of China, headquarters staff of the Agricultural Bank of China, and the staff of the Jiangxi and Guizhou provincial and county branches is gratefully acknowledged. The ICRs provide an adequate account of the implementation and the development effectiveness of the projects, with some shortcomings. The PAR concurs with the ICR for RCIII on the principal ratings for project but differs with the ICR for RCIV. Following standard OED procedures, copies of the draft PAR were sent to the relevant government authorities for comments. The response(s) is attached as Annex C.  1. Introduction 1.1 Since the initiation of policy reforms in 1979, Chinese agriculture has experienced rapid growth and diversification. With the rapid increase in farm production, the government's agriculture sector strategy began to focus on production of higher-value goods and agro-processing activities. A concurrent effort to restructure and reform the financial sector, including the rural financial sector, was undertaken to better serve the emerging market-oriented and decentralized economy.' 1.2 The Agricultural Bank of China (ABC) was reestablished in 1979 to take over the rural banking functions from the central bank, the People's Bank of China (PBC). Ever since, efforts have been underway to transform the bank from being a fiscal agent to the government to a more competitive and independent financial intermediary. At present, ABC and the Rural Credit Cooperatives (RCC) are the main rural financial institutions responsible for deposit mobilization and provision of credit. 1.3 The World Bank has actively supported the government's strategies in both the agricultural and financial sectors. The Bank Group started working with ABC in 1981. By 1988, when Rural Credit III (RCIII) became effective, the Bank had successfully completed one IDA-assisted Rural Credit I project and another, Rural Credit II, was under implementation. When Rural Credit IV (RCIV) was approved, RCIII was still mid-way through implementation. Successive projects have targeted different provinces for the support of investment opportunities and for institutional development. 1.4 RCIII and RCIV were designed against the background of ongoing financial sector reforms. At the time RCIV was approved, the Bank was engaged in a substantive dialogue with the Government of China on reforming the financial sector on the basis of the just-completed Financial Sector Review (FSR).2 The actual process of financial sector reforms has, however, been uncertain and progress, particularly on policy issues, has been slow. Both RCIH and RCIV were conceived and implemented when ABC had no control over the pricing of credit or deposits, and credit allocation was governed by quotas set by the annual credit plan sanctioned by the State Development Planning Commission. 1.5 During the implementation of RCIV, the government decided to separate the commercial banking and "policy" related operations of ABC, as recommended by the FSR. The commercial functions were given to ABC, which retained the name, and policy functions were given to the new Agricultural Development Bank of China (ADBC). At the same time, the government transferred the supervision of RCCs from ABC to PBC, thereby ending a generally unhealthy and unhappy relationship between the two organizations. 1. As pointed out to the audit mission on several occasions, China has a "socialist market economy". 2. The World Bank: "China, Financial Sector Review: financial Policies and Institutional Development," Report No. 8415- CHA, June 29, 1990. 3. The interest rates are still set by PBC, and while the credit plan has given way to management by asset-liability ratios in 1998, ABC still does not have adequate autonomy in key policy areas, including credit allocation, and continues to be subject to interference by central or local governments. 2 2. The Project Objectives 2.1 Rural Credit III (RCIII) sought to (a) stimulate growth and diversification of agriculture by expanding the availability of medium- and long-term credit, and (b) increase ABC's operational efficiency through institutional development. Rural Credit IV (RCIV) sought to (a) stimulate growth and diversification of agriculture by financing profitable investments by farmers, collectives, state farms, and enterprises, and (b) promote the efficiency of rural financial intermediation through the introduction of sound, market-oriented policies, and development of the main institutions concerned with rural savings mobilization and lending. 2.2 While the real sector objectives were similar in both projects, the scope of the intended institutional development (ID) was different. RCIII's ID efforts were limited to improving the operational efficiency of ABC. RCIV aimed at the broader development of the rural financial sector in addition to the strengthening of ABC. It was to support the recommendations of the FSR that applied to rural financial institutions, ABC, and the Rural Credit Cooperatives (RCCs), with a view to improving their management and rural financial intermediation in general. 2.3 The ICR for RCIV, however, maintains an unstated third objective: reform of the general financial sector policies. However, the project was prepared and appraised as a real sector operation and was not intended to be a vehicle for reform of the entire financial sector. The Bank's intention to this effect was communicated to the borrower in a post-appraisal letter, and was reiterated in the SAR Yellow Cover Review Meeting.5 Nevertheless, with a view to placing the project within a consistent and comprehensive reform framework, negotiations were postponed six months until after a meeting of the Senior Vice President Operations with the Government of China on the financial sector reform process. At the discussions, the government agreed to a technical assistance project covering the entire financial sector to pursue the conclusions and recommendations of the FSR. 2.4 In light of these developments, RCIV was negotiated with the objective of supporting some of the recommendations of the FSR, but only as they applied to the rural financial sector and ABC's institutional development. Accordingly, the conditionality included the loan documents applied to the rationalization of the ABC's policies. The reference to the financial sector in general is made only with respect to the scope of the midterm review, which was to review the progress on the reforms in the overall financial sector. It is unfortunate, thus, that despite the clarity of the legal documents and agreements reached at negotiations, the ICR considers policy reform for the financial sector as the overriding objective of RCIV. The audit disagrees with this assessment. Components 2.5 RCIII was to (a) finance long-term profitable investments by households, cooperatives, collective enterprises, state-owned enterprises, and state-collective partnerships through ABC's 4. A letter of December 26, 1989, to the President of ABC clearly stated that the project was not a financial sector operation. 5. The minutes record that most comments were based "...on the vain hope that the project could be used as a structural adjustment loan or even as a sector adjustment operation." 3 provincial branches in Guizhou, Sichuan, Yunnan, Anhui, Hubei, and Henan;6 (b) provide technical assistance for staff training and ABC studies to support its systems reform activities, and (c) support a pilot program to introduce long-term lending to selected RCCs in four provinces. 2.6 RCIV was to finance long-term profitable investments by farmers, collectives, state farms, and enterprises through ABC's provincial branches in Inner Mongolia, and the provinces of Shaanxi, Jilin, Liaoning, Jiangxi, and Guangdong (new provinces) as well as the Guangxi Zhhuang Autonomous Region and the provinces of Fujian, Hunan, and Hubei (provinces covered under previous projects).7 ABC had the discretion to use project funds in other areas, including provinces covered by the preceding three Rural Credit projects, with prior approval by the Bank. The project's institutional development component consisted of technical assistance to upgrade ABC's financial management, develop long-term lending strategies and programs, review ABC's lending policies, upgrade ABC's management information systems (MISs), provide office technology and equipment to ABC, and to train ABC's senior staff, trainers, and project staff. 2.7 The government also agreed to conduct a study of rural investment and finance and to review progress on the financial sector reforms during the project's midterm review. These components, however, were partially taken over by the Financial Sector Technical Assistance (FSTA) project. Costs 2.8 For RCIII, total cost was estimated at US$340.0 million, of which US$4.0 million was for technical assistance. The IDA credit (US$170.0 million) was to finance about 50 percent of sub-project costs, including all foreign exchange (FX) and 35 percent of local costs. The remaining costs were to be financed by ABC (20 percent) and sub-borrowers (30 percent). 2.9 Actual costs were 97 percent higher in Renminbi (local currency), and 18.3 percent higher in FX at US$ 402.5 million. The difference is due to inflation, which was high in 1988 and 1989, fluctuations in the SDR exchange rate, and the devaluation of the Renminbi by 57 percent from 1988 to 1994. With these developments, the argument in the ICR that rising costs due to inflation forced ABC to finance a smaller number of more expensive sub-projects is not valid. It is not clear why the project financed only 61 percent of sub-project investments instead of the planned 70 percent (40.7 percent by Bank funds).' 2.10 For RCIV, total project cost was estimated at US$550.0 million (Y 2,596 in local currency), including US$3.0 million for technical assistance. IBRD (US$75.0 million) and IDA (US$200.0 million) were to finance about 50 percent of sub-project costs, including all FX costs and 36 percent of 6. The Credit Program provided funds for the development of tree crops, including rubber, tea, coffee, gallnut and fruit; livestock, including pig, sheep, goat, poultry and cattle including pasture development; aquaculture; and the establishment of processing facilities, including those for fruit, oilseed, rubber, tea, animal and aquatic products, and feed mills. 7. The Credit Program sought to provide funds for the establishment and rehabilitation of orchards; citrus post-harvest facilities; rehabilitation of green tea production; silk production (including mulberry planting); vegetable, mushroom, fungi, ginseng, and bamboo production; irrigation works; pasture-based and feed-based livestock production; fisheries; agroprocessing; slaughterhouses; cold storage; and miscellaneous other facilities supporting production. The program included $10 million for lending through RCCs. 8. The data on investments by province, expressed in local currency, show that in all provinces except Hubei actual investments were much higher than anticipated. There was, however, a significant shift in the allocation across provinces. This is because some of the sub-projects appraised before effectiveness had to be cancelled owing to difficulties in obtaining the necessary equity contributions and because of changes in market conditions. 4 local costs. The remaining costs were to be financed by ABC (20 percent) and sub-borrowers (30 percent). 2.11 Actual cost in US dollars increased by 17.8 percent to US$ 648 million and in local currency by 54 percent. The difference is again due to inflation (from 1993 to 1995) and the devaluation of the Renminbi from 1992 to 1997 by 56 percent. Hence, local currency equivalent funds grew substantially, enabling the project to finance higher costs or more sub-projects. Evaluation ofProject Objectives and Design 2.12 The ICRs correctly rate the agricultural growth and diversification objectives for both projects as appropriate. Regarding institutional development, the ICR on RCIII notes that the objectives were slightly unrealistic with respect to the pace of financial sector reform and that the original terms of reference for the policy studies were too ambitious. Similarly, the ICR for RCIV states the project's financial sector objectives seem in retrospect to have been too ambitious. While the audit agrees with this assessment, its rationale is different. 2.13 A distinction needs to be drawn between objectives that pertained to (a) the institutional capacity building of ABC, (b) the policies that were expected to apply to ABC, ostensibly to accelerate the bank's commercialization and deemed to be in its control, and (c) the broader financial sector agenda. As noted earlier, the third objective is was never an explicit objective of the project. 2.14 The design of the institutional capacity building objectives of the projects should be considered strategically, as a natural evolution in an inevitably long and slow process. They were intended to help ABC become a more efficient and effective institution in the elusive post-reform era. These objectives too were appropriate. 2.15 Finally, the financial objectives of RCIV that pertained to ABC, while generally consistent with the long-term view envisaged in FSR, were inconsistent with its recommendations for the short term. The FSR concluded that interest rate liberalization was neither feasible nor desirable in light of the continued role of credit planning as an allocation mechanism, the lack of competition between lending institutions, and the fact that enterprises were subject to price controls. Under these circumstances, the financial objectives of RCIV were over-optimistic about the pace of reforms, even though such optimism was not shared by the Bank's own FSR. 2.16 As it turned out, some of the objectives of reforming of ABC's financial policies were premature and ignored the institutional and political reality. It was unrealistic to expect that ABC could set its own policies without the commensurate managerial and political autonomy to do so (this also applies to the studies included in RCIII and RCIV). In particular, it was naYve to expect that ABC could charge higher lending rates than those set by PBC. In principle, ABC had the discretion to vary PBC rates by up to 20 percent, but in practice it did not have the will or the autonomy to do so; and in fact does not have that ability even now. Furthermore, it was unrealistic to expect ABC to charge a rate higher than other financial institutions (e.g., for financing of agro-processing projects) with whom it competes for customers (rates for similar investments are set uniformly for all banks by PBC). ABC did not have the will to charge higher rates in agriculture, as PBC rates are reportedly well-known to the borrowers, and local governments continue to exercise considerable control on the lending by local ABC branches. 5 2.17 The ICR prominently states that the project was one of the first operations using guidelines contained in OD 8.30.9 However, the audit does not see compliance with the requirements in the OD guidelines. The eligibility criteria for financial intermediaries on managerial independence and on financial solvency, as confirmed by independent external auditors, were not met.'0 Project design regarding the lending program followed the traditional credit line approach. The technical aspects of the sub-projects in the indicative investment program were appraised in great detail in both projects. 2.18 The Bank's overarching goals of poverty alleviation and protecting the environment were mentioned in both projects, although they were not explicitly stated as objectives. The project designs included poverty areas, but neither set targets nor included indicators to monitor progress. In ABC's view, as expressed to the audit mission at various times, even the limited poverty focus of the projects was a design flaw. Making small loans in remote and poor areas increases ABC's administrative cost and forces it to finance riskier and less profitable sub-projects, leading to a decline in the quality of its loan portfolio. ABC prefers to make loans at its own discretion, without any sectoral or geographical restrictions. 2.19 Both SARs concluded that the projects did not face any unusual or major risks. While it is easier to pass judgment in retrospect, the analysis was at best incomplete. Neither of the SARs considered inflation or FX risks or their financial implications for ABC or the government. Most significantly, the RCIV fails to go into the consequences of lack of government and ABC commitment to carry out the policy reforms only in the rural sector without affecting the conditions in other sectors. 9. OD 8.30 was issued only in February 1992; the ICR apparently refers to its precursor, the report by the special task force on financial sector operations of 1989. 10. An independent confirmation of (i) a positive return on equity with realistic provisions for potential and identified losses, (ii) a minimum capital ratio, (iii) appropriate conditions on rescheduling of loans, or (iv) a collection rate that avoids the erosion of capital, was not provided. 6 3. Implementation R CIII 3.1 The implementation of RCIII was generally free of major events. Initially, disbursements were much faster than projected at appraisal, the reason being that a large number of sub-projects had already been appraised and were awaiting approval at the time of project effectiveness. Disbursements slowed toward the end of the project, necessitating an extension of the closing date to June 1994. The reason for the slowdown was the callback of funds for one sub-project that was financed without IDA approval, despite being above the free-limit for approval. It subsequently took ABC some time to identify substitute investments. 3.2 During the implementation of RCIII supervision missions noted various shortcomings and brought them to ABC's attention. Most of these were effectively addressed, but some persisted until the end of the project, such as inadequate staff for project supervision. Another issue was the quality of sub- project appraisal, with shortcomings related to weak market and cash flow analysis, and the inadequate consideration of working capital requirements. Although training improved the quality of appraisals toward the end of the project, the importance of market analysis was not given adequate attention by ABC branches. The issue of environmental impacts of agro-processing plants were also regularly identified, but these were generally dealt with on an individual basis. 3.3 Throughout implementation there were temporary violations of loan covenants on on-lending interest rates and spreads. Such violations were closely monitored by supervision missions and remedial action was usually taken promptly. The issue of inflation and negative real interest rates during certain periods of implementation, however, was never raised, despite the requirements in the Development Credit Agreement (DCA). 3.3 Lack of working capital, especially in the early sub-projects, or its misuse (e.g., repayment of loan) deprived the sub-projects, especially in agro-processing, of resources for raw materials, resulting in reduced capacity utilization, operational difficulties, and even failed sub-projects. This situation was made worse by having borrowers finance more than the permissible 30 percent of the investments. Another problem was the mismatch of loan maturities and sub-project cash flow. 3.4 The pilot component designed to reach small farmers was to lend funds through selected RCCs in four provinces. ABC, however, showed little commitment to the pilot. It was, nevertheless, belatedly implemented when it became clear that the funds could not be used for other purposes. Despite ABC's visible reluctance to deal with RCCs, however, a similar, larger component was included in RCIV. 3.5 The most serious event during implementation was the approval of a sub-project that was above the free-limit but for which no approval was sought from the Bank. This covenant violation prompted a strong response by the Bank, which requested a refund of the sub-project loan amount of $4.7 million to the Special Account. In response, ABC called a conference of managers from all participating provinces to emphasize the need to follow the covenant stipulations strictly. RCIV 3.6 By the time RCIV became effective, the Ministry of Finance (MOF) had decided that it would not bear the FX risk for projects involving financial institutions. Hence, all RCIV funds were on-lent by 7 MOF to ABC in FX. ABC, thus, never had a chance to exercise the option provided by the project for it to borrow project funds in local currency and avoid the incidence of FX risk. 3.7 As in RCIII, early sub-project appraisals resulted in rapid initial disbursements under RCIV. An early supervision mission detected a violation of the interest rate covenants, with the spread being less than the stipulated 2.5 percent (after tax). Throughout implementation, ABC did not use its discretion to vary (by up to 20 percent) the interest rates it charged its clients from those prescribed by PBC. 3.8 While the lending rates to ultimate borrowers were discussed by some missions, the interest rate from MOF to ABC was never mentioned in any supervision report. The ICR, however, takes a particularly strong exception to an apparent violation of the related covenant. The rate applied was 5.4 percent, which was less than the prevailing IBRD rate. Because of the importance the ICR assigns this action, because of its implications for the efficiency of financial intermediation, and since it is more serious because it involves the MOF in addition to ABC, the issue warrants careful consideration. 3.9 Formally, section 3.01 (b)(ii)(B) of the DCA specifies that for funds on-lent in FX, ABC should pay to the MOF interest "...at a rate not less than the prevailing IBRD Loan rate, or such other market- based rate as shall be acceptable to the Association, the Borrower and ABC." The question, therefore, is whether the lack of mention of the covenant violation by supervision missions showed a lack of due diligence or voluntary disregard. According to the statements ABC made to the audit mission, both ABC and the supervision missions were fully aware of the on-lending terms as well as the conditions in the DCA. However, supervision missions never raised it as an issue, and ABC assumed that the interest rate was acceptable to the Bank." The audit, thus, presumes that the lower interest rate was acceptable to the borrower and the Bank. 3.10 While the arrangement might not be considered a formal violation of the loan agreement, the substantive issue is whether the interest rate violated the spirit of the project's ID objectives. In this respect, the audit considers whether the 5.4 percent rate was a significant deviation from a market-based rate and whether the action resulted in any major distortion of the financial markets. 3.11 The rate charged to ABC was similar to the rate other banks had to pay for long-term FX funds in the international market and hence may conceivably have been considered a market-based rate by MOF.12 On the issue of how the figure of 5.4 percent was arrived at, ABC noted that the rate was based on the rates for FX loans set by the Bank of China, the only Bank dealing with FX transactions at the start of RCIV. The Bank of China in turn based its rates on the prevailing LIBOR rate. The rates on ABC's direct commercial borrowings from foreign sources were also LIBOR-based." It should be noted, however, that RCIV funds were long-term funds and LIBOR rates quoted are for one-year maturities and may therefore not be directly comparable. Nevertheless, the (simple) average LIBOR 11. The audit gets this impression as well from the End of Mission document of the May/June 1994 mission. The document cites a case in Inner Mongolia where an interest rate of 6.1 percent was charged for a FX loan, "well below international levels and this provided a spread to ABC of only 0.55 percent." It is apparent that the mission assumed that ABC would receive the funds at 5.55 percent. 12. The term "conceivably" is used since no authoritative word on this was obtained from the MOF. The audit mission was unable to meet with any person from the MOF who was associated with the project or project-related decisions. 13. Average annual LIBOR rates for one year loans, quoted in Chinese financial statistics, are as follows: 6.29 percent in 1991, 4.2 percent in 1992, 3.64 percent in 1993, 5.59 percent in 1994, 6.24 percent in 1995, and 5.63 percent in the first half of 1996. 8 rate for the period 1991 to 1996 is 5.41 percent.14 While the precise answer to how the MOF-ABC interest rate was established is unknown, the 5.4 percent rate used by the MOF is probably not far from the true market rate in China (which was indeed lower than the prevailing IBRD rate). 3.12 On the distortionary effect of this action on the Chinese financial markets, according to ABC, the rate it paid for project funds and the rates it charged its sub-borrowers were the same as those for other financial intermediaries for similar maturities and similar purposes. Finally, the volume of funds provided by RCIV as a proportion of ABC's overall portfolio was very small."s Hence, it is unlikely that the action had any major distortionary effects. 3.13 In general, however, covenant violations were not adequately dealt with. On several occasions violations were inconsistently noted or simply ignored. For example, while almost all supervision missions dealt with the 2.5 percent spread for ABC, none of the mission reports, with the exception of the last one in May 1996, dealt with the covenant that interest rates had to be positive in real terms- although it was the very next item in the DCA (section 3.02 (b) (ii)). The result was that in the years 1993 to 1995, when on-lending rates were highly negative, it was not brought to ABC's notice. 3.14 Similarly, ABC was to present a recapitalization plan by December 31, 1991, and keep its net worth at not less than 5 percent of its risk-weighted assets. Although the supervision mission always came back to this issue, the plan was never submitted. Subsequently, however, as the ICR notes, both the IMF and the Bank are of the view that recapitalization without first addressing more fundamental issues is not a good strategy. This of course points to the inappropriateness of the condition to begin with. 3.15 In April 1992 it was still intended to conduct the midterm review as agreed at negotiations, i.e., to include an assessment of the progress on financial sector reforms. Eventually, the FSTA project was to review the progress annually, making the RCIV requirement in that respect redundant; accordingly, it was dropped. Similarly, the Rural Finance Study was reduced in scope to eliminate elements that were to be covered by the FSTA project. However, these changes were made without formally modifying the loan documents. As a result, the continuation of redundant covenants appears to have led to confusion at project closing about its revised objectives. 3.16 In 1994, the government decided, as a result of its dialogue with the Bank on financial sector reforms, to split ABC into a commercial bank and a policy bank. ABC would retain its name and would become the commercial bank. A new policy bank would be created to handle all policy lending, essentially subsidized credit, including that for poverty alleviation, minorities, and grain purchases. ABC retained the World Bank portfolio. 14. The ICR, on the other hand, states that the 5.4 percent rate was the weighted average cost of the IDA/IBRD blend. However, the weighted average between IDA and Bank on US$200 million for the credit and USS75 for the loan would certainly have been much lower than 5.4 percent. 15. The peak loan portfolio generated by Bank funds was reached in 1989 for RCIII and in 1995 for RCIV. In these years, the share of project funded loans in ABC's total portfolio was less than one half of one percent. The actual share of Bank/IDA funds was 0.312 percent and 0,267 percent, respectively. 9 4. Outcomes and Findings 4.1 RCII and RCIV substantially achieved their physical objectives and the ICRs present an accurate and detailed description of the main achievements. However, in both projects, the institutional development objectives were only partially achieved. As noted earlier, neither of the projects had any objectives related to the reform of the entire financial sector. There were, under RCIV, some objectives related to the rural financial sector. These were achieved only to a limited extent. 4.2 The ICRs for both projects mention poverty reduction in passing with no substantive assessment of the projects' achievements. ABC maintains that the projects had significant impacts on poverty, primarily through job creation in designated poverty counties. The actual outcomes, however, are unclear for lack of evidence. Physical Achievements 4.3 In both projects the physical achievements were substantial. RCIII financed 850 sub-projects compared to 1,200 expected at appraisal. The reduction was the result of financing larger agro- processing sub-projects and a corresponding decline in the smaller sub-projects. Under RCIV, 1,780 sub-projects were financed (no number of sub-projects was given in the SAR). While some of the sub- projects faced problems, in relatively few cases have sub-projects folded. 16 4.4 Investments in agro-processing were much higher than estimated at appraisal. The reasons for the shift under RCIII, as noted in the ICR, were changes in market conditions, which led to investors moving into higher-value production offered by agro-processing investments. The increase in agro- processing in RCIV was due to the replacement of low-return sub-projects by those promising higher returns. In provinces that had already participated in earlier projects, there was a greater demand for processing sub-projects to meet the increase in basic production. ABC management indicated that in both projects, business sense and practicality also played a role. The design of the projects to reach small farmers and remote areas was complicated, costly, and entailed higher risks. The provincial branches, thus, decided to substitute larger and potentially more profitable projects. 4.5 For RCIII, the overall FRR was calculated by the ICR at 31 percent, significantly higher than the 20 percent anticipated at appraisal, while the ERR was 20 percent against the 24 percent expected at appraisal. The comparable figures for RCIV are 22 percent compared to 24 percent estimated at appraisal for the FRR, and 24 percent versus 27 percent for the ERR. Overall, then, both projects financed profitable and economically beneficial projects. The audit did not attempt to replicate the complex and enormous task, already undertaken for the ICRs, of recalculating the projects' rates of return. The audit mission did, however, calculate the FRRs for the sub-projects it visited, and was satisfied that the analysis done by the ICRs was indeed of high quality. Discussions with sub-project managers and ABC supervising staff also revealed that the assumptions and figures used were plausible and appropriate. 16. In such cases, ABC was forced to foreclose on the investments and in most cases was able to enforce repayment of outstanding dues. 17. For RCIII, agro-processing's share was 59 percent of total investments, up from the 39 percent projected at appraisal, and for RCIV it was 53 percent against 31 percent planned. 10 Additionality 4.6 A major issue that was identified at the ICR review stage was that of the additionality of Bank funds, i.e., would ABC have financed the sub-projects without the availability of Bank/IDA funds? To appreciate the issue, it is necessary to consider the context and rationale for the two projects. 4.7 The primary rationale for both projects was the strengthening of medium- and long-term (development) lending and savings mobilization. The RCIII SAR does not give details on the composition of ABC's deposits (short-term versus long-term), nor does it give a breakdown of lending other than indicating that loans are mostly short-term. The SAR for RCIV provides better information. Total resources of Y 319 billion at the end of 1988 included deposits of Y 171 billion. The deposits included Y 44 billion in time deposits, of which Y 16.5 billion were term-deposits for three years or more. The share of long-term deposits increased, totaling Y 30.9 billion, by June 1989. Medium- and long-term loans at the same time amounted to Y 28.7 billion, or 65 percent of time deposits; the balance of time deposits was used to make short-term loans. 4.8 Through RCIII, the IDA funds added Y 810 million to ABC's long-term resources, and RCIV provided Y 1.7 billion over six years. These figures show the relatively small proportion of funds provided by the two projects in ABC's total lending. However, the data also show that ABC had access to long-term resources and thus the question of the additionality of bank funds becomes important. The argument that ABC did not have enough long-term resources and that Bank/IDA funds were needed to build up a portfolio of term loans is incorrect. Also, the suggestion that the projects assisted ABC in the term-transformation of short-term deposits into long-term loans (ostensibly through the 20 percent of share investment costs provided by ABC) is inaccurate. The best that can be said is that the projects helped reduce the pervasive use of costly long-term deposits for short-term loans, which are less profitable than term loans. 4.9 Central to the role of the Bank/IDA funds is the credit policy that was in force in China when the projects were implemented. ABC had to follow a credit plan prepared by PBC and approved by the government, which specified the sectoral and provincial quotas for "fixed asset investments."" These quotas had to be met with the available resources of ABC, regardless of whether they were short-term or long-term. Bank funds, however, were outside the credit plan and thus helped ABC increase its effective long-term lending quotas. ABC headquarters allocated Bank funds, in addition to the quotas, among its provincial branches. The FX that ABC got for RCIV was allocated to provinces that could use it for their FX business in exchange for Renminbi resources (generated through additional deposit mobilization). ABC then allocated the local currency funds to the other participating provinces. It appears, therefore, that Bank funds were additional to ABC's lending program and that the sub-projects thus financed would otherwise not have been permitted under the existing credit system in China. Capacity Building 4.10 Following their predecessors, RCIII and RCIV supported the development of staff skills.'9 Although gaps remain in certain areas, e.g., the economic analysis and cash flow considerations, the projects succeeded in substantially upgrading the skills of ABC staff in sub-project appraisal. The staff at all levels were anxious to learn and willingly accepted the new techniques. On several occasions during the audit's field visits, ABC staff noted that the primary benefit of the projects was not the funds 18. Starting with 1998, the credit plan was given up and replaced by asset/liability management to be exercised by the banks. 19. Each project covered new provinces, and hence was complementary in its efforts. 11 but their training. The techniques are now standard procedures for ABC's non-project lending. This capacity-building effort, although not complete by any means, is significant considering ABC started as a government agency and is now much better placed as a prospective commercial entity. 4.11 The outcome of the studies program was not satisfactory. The limited impact it had on ABC operations and policy resulted from a lack of commitment by ABC to address substantive issues and its reluctance to use foreign consultants. This is unfortunate considering that significant changes were expected in the banking sector, and considering ABC's desire to become a fully commercial bank. The Bank could have made more efforts to convince ABC management of the strategic importance of the studies and the benefits of experienced consultants. This is noteworthy since on several occasions during project preparation, concerns were expressed about the need to ensure timely start-up and implementation of the institutional development and training components. 4.12 Another area of missed opportunity was transfer of technology, although this is by now a familiar outcome following the earlier projects. As noted by the ICRs, the design standards in many sub-projects were poor, which contributed to poor operating performance. ABC, however, views this issue differently. For cost reasons, but mainly for reasons of employing labor-intensive technologies in accordance with the priorities of the provincial governments, the design of facilities was kept simple and based on local technology. Local consultants, if any, were employed, and ABC's lack of qualified technical staff meant that it had to rely mainly on the local Planning Commissions and technical bureaus for technical appraisal. Technical expert groups were created toward the end of RCIII and gained more importance during implementation of RCIV. Technology transfer took place only where local technology was not available. Judging by the small proportion of FX loans, the impact was quite limited. 4.13 In general, ABC is reluctant to use loan funds for foreign technical assistance (TA), especially on IBRD terms. ABC has been disappointed in the past, when expensive consultants have produced reports but have left without any measurable benefits to ABC. This outcome reflects a lack of commitment of the targeted departments of ABC to work closely with the consultants. It also reflects a lack of supervision (from ABC management and the Bank) to ensure maximum benefit from the TA. Thus, if technical consultants only meet with the International Banking Department instead of the operational departments, the impact will likely be limited. 4.14 Both projects incorporated components for medium- and long-term lending through, and the institutional development of, RCCs in an effort to reach a larger number of small farmers and contribute to poverty alleviation more directly. After initial problems, because of ABC's reluctance to deal with RCCs, through the persistence of supervision missions, the component was successfully completed in both projects. However, the impact of the projects on the institutional development of RCCs was limited. In 1996, the supervision of RCC affairs was taken from ABC and given to the State Council's Coordinating Group for Rural Reform housed in PBC. ABC gladly accepted this change, but it lost Y 170 billion of RCC's relatively cheap deposits. Financial Impact 4.15 ABC's true financial position and the quality of its portfolio are unknown due to a lack of reliable data conforming to acceptable international banking standards. This remains an area of concern, especially if ABC is seriously contemplating commercialization. The data available to the audit are 20. RCC deposits were replaced by long-term PBC funds at a higher cost. 12 those provided by ABC. They are based on existing local accounting standards and without the benefit of an independent portfolio audit. The issues of ID are thus addressed on the basis of the limited information available and are at best indicative. 4.16 Profitability. The profit figures in the ICR for RCIII, without knowing the basis for their calculation, are useful only as an indicator of the reasons given for the profits: adequate spread between cost of funds and lending rates, good collection, and low operational cost. These reasons hold for RCIV as well, although with some differences. The spread on Bank funds under RCIII varied between 1.5 percent and 4 percent as against the required 2 percent after taxes. Under RCIV, the required 2.5 percent was achieved only in 1995. ABC applied the interest rates prescribed by PBC and decided not to approach PBC for the approval of higher rates or to exercise its discretion to vary the prescribed rates by up to 20 percent. ABC maintains that the rural community, in particular the farming households, would not be able to afford higher rates, nor would they be willing to pay higher rates since PBC rates are common knowledge. Thus, given the choice of charging "unrealistically" high rates or ensuring smooth implementation of RCIV, ABC chose the latter.21 4.17 Based on available data, repayment rates appear to be satisfactory (discussed further below). ABC's operational costs also appear to be well contained, based on the data provided to the audit mission. 4.18 Financial Management. The ICR on RCIV comments that ABC, like other traditional institutions disbursing directed and subsidized credit, was indifferent to savings mobilization. This characterization is rather harsh since ABC has traditionally had a substantial deposit base, particularly in rural areas. Also, as discussed earlier, it is not clear what the subsidy element was in the on-lending rate from MOF to ABC or how it adversely affected the efficiency of ABC's financial intermediation. 4.19 The rationalization of the interest rate structure, as discussed earlier, is a broader sectoral issue and beyond ABC's control. Nevertheless, ABC failed to take advantage of the opportunity it had (to vary PBC rates by up to 20 percent) to increase its interest spread and at least attempt to abide by the agreements in the loan documents. At the same time, the Bank failed to take decisive action to ensure that the spread was maintained. Nor were any efforts made to rectify the negative lending rates. One aspect that did not receive the desired attention in the ICR, but is critical for financial management, is the improvement of the MIS, which was intended to be done with the help of foreign consultants. However, no progress was made under the Bank projects on upgrading the MIS, and ABC is currently using grant funds from ADB for this purpose. 4.20 Portfolio Quality. As noted, the quality of ABC's total portfolio is unknown. The impact of the projects on the overall portfolio is also unknown. As noted earlier, however, project funds were a small proportion of ABC's total portfolio. ABC informed the audit mission that its overall portfolio is of a better quality than the sub-project portfolios.22 On the other hand, it is also likely that applying the sub- project appraisal and monitoring methodology to the non-project portfolio, that a positive impact is likely. 21. Higher rates were considered "unrealistic" by ABC in terms of being higher than the prescribed PBC rates. Financially, of course, given the high FRRs of the sub-projects financed, the argument is not tenable. 22. The reason is that under the projects, ABC had to follow the selection criteria set by the Bank, which were geared toward poverty alleviation, remote rural areas, and specific sectors. Besides raising operating costs, ABC felt that this led to a riskier mix of sub-projects and compromised the portfolio quality to some extent. 13 4.21 Turning to the quality of the sub-portfolio, the audit attempted an indicative analysis of the portfolio of sub-loans under RCIII and RCIV based on information obtained from ABC headquarters and various field and branch offices in two provinces (Jiangxi for RCIV and Guizhou for RCIII). An ABC-wide analysis of the age-wise breakdown of overdues was not possible since the details are not available at ABC headquarters for lack of a proper MIS. Thus, the audit had to rely on the classification of loans as prescribed by PBC, using the former Chinese standards of classification.3 Somewhat more detailed information with age-wise breakdown of overdues was available in the two provinces visited by the audit mission. 4.22 A review of the available data confirmed ABC's assertion that overdues for RCIII are not high.24 The loss due to unrecoverable loans is estimated at about 3 percent of the total project loans disbursed. If the data present an accurate picture, this would be a satisfactory result. For RCIV, the situation is somewhat different. Disbursements were completed only in 1996, and repayments will carry on for another five to eight years. Nevertheless, the audit estimates that losses under RCIV may be higher than for the RCIII sub-portfolio, possibly as high as 6 percent of the amount disbursed. This is still an acceptable result. If the accumulated provisions of 1 percent p.a. on the term loans disbursed under the project, as allowed by the MOF, are made available, they should be sufficient to cover the losses incurred on the project funded term loans. It should, however, be noted that the provisioning rules apply to ABC's entire portfolio of term loans. Without the knowledge of the quality of ABC's portfolio not related to the projects, the audit is unable to determine whether or not the overall level of provisioning is sufficient to cover ABC's total loan loss.25 4.23 Impact ofInflation. For RCIII, the impact on the financial position of ABC is expected to be neutral since almost all of the projects to which it on-lent funds were obtained from MOF in local currency. The financial impact on MOF, however, will be negative, since the interest income generated by the credit would be insufficient to make up for the erosion of capital.26 The value of the recovered principal is likely to be substantially lower than the amount financed by the project. The situation under RCIV, however, is unclear. All RCIV funds were on-lent by MOF to ABC in FX. The bulk of ABC's sub-loans were in local currency, with the FX used for apparently more profitable FX business. Clearly, for sub-loans made in FX ($21.9 million, or 8 percent of total), the impact of local inflation on ABC would be zero. However, for local currency loans, the repayments of principal would represent only about 65.5 percent of the value of disbursements. 4.24 It is apparent that the loss will have to be recovered by ABC by other means, such as positive lending interest rates, negative deposit rates, or profits from the use of FX. Insofar as the swap of FX for local currency was financed through additional deposits, given that the deposit rates were less than 23. A new classification system following more stringent standards has been introduced by PBC but is not yet compulsory. 24. As per ABC policy, rescheduling is permitted only once. The policy is apparently strictly followed, as witnessed by the audit mission on one of the field visits. In the case of a kiwi plantation, a second rescheduling would have been advisable but could not be done because of the standing rules. The rule was adhered to despite the fact that ABC had made the mistake in the original loan agreement of not matching the repayments with the project's cashflow. How strictly this is followed in general, or whether it is consistently applied to all types of borrowers, however, is unknown. 25. In the drive for the commercialization of ABC within the reform of the entire banking system, the reform of the state- owned enterprise sector is envisaged which would include writing off bad debts and establishing a new provisioning system. 26. Inflation was high in 1988 (18.5 percent), 1989 (17.8 percent), 1993 (13.2 percent), 1994 (21.7 percent), and 1995 (14.8 percent). Assuming inflation at -2.5 percent in 1998, -2 percent in 1999, 0.5 percent in 2000, 2 percent in 2001, and 4.5 percent thereafter, the value of the principal due to MOF (over 20 years with a 5-year grace period) would be reduced by approximately 38 percent. 14 lending rates in absolute terms (see annex B), the net impact on ABC would likely be negligible. Further, the ICR states that ABC generated significant profits in its FX business. While ABC confirmed this to the audit mission, it did not provide any supportive data. Hence, it is not known how profitable the FX business was. In sum, considering the various factors, the audit concludes that RCIV probably did not have a major negative impact on ABC's financial position. 4.25 Commercialization. The ICR on RCIV prominently describes ABC's progress toward commercialization. While the developments are significant in the evolution of ABC and the financial system in China, the progress on most fronts cannot be attributed to the project. For example, the separation of policy lending and RCCs from ABC had a beneficial impact on the quality of ABC's portfolio but was not an outcome of RCIV. 4.26 Although the commercialization of ABC is not directly dependent on project outcomes, from the institutional perspective it is worth noting some recent events that are likely to affect ABC's progress in commercialization. The government appears committed to transforming ABC into a commercial bank; asset-liability management has replaced the old credit-plan as a mechanism for financial management; new loan classification standards have been introduced; and accounting procedures are expected to be upgraded to international standards shortly. Key areas where uncertainty remains are ABC's autonomy in terms of operational and financial policies, including provisioning, write-offs and the ability to independently determine interest rates. It is assumed that as part of the commercialization drive, these issues will eventually be addressed. 4.27 ICR quality. Overall the ICRs for both projects provide a reasonably accurate description of the physical achievements of the progress and implementation experience. The quality of the economic analysis, and the effort involved in the enormous task, is also impressive. There are, however, some shortcomings. 4.28 Neither ICR presents a reasonable plan for future actions agreed with the borrower as required by OP 13.55. Both ICRs, especially for RCIV, include discussions for potential future projects, but do not discuss how the respective projects' outcomes are to be maintained or improved. In this respect, given the shortcomings noted in the ICRs, it would have been reasonable to expect an action plan to (a) achieve and maintain positive lending rates, (b) suggest a change in the lending rate from MOF to ABC for the remainder of the time till full repayment, (c) make arrangements with RCCs and PBC for the administration of loan portfolio transferred to PBC as the new supervising agency, and (d) make arrangements regarding the administration of the loan portfolio and the use of recycled funds for similar purposes until the repayment to MOF is due. 4.29 Another shortcoming in the ICR for RCIV is the absence of the ICR mission's Aide Memoire. This was apparently left out because of difference of opinion on the conclusions reached and the project ratings between the Aide Memoire and the final version of the ICR. On the ratings, there are internal inconsistencies in the ICR relating to the project's original objectives, the developments during implementation that led to a change in the scope and objectives of the project, the project's actual achievements, and the conclusions arrived at by the ICR with respect to project outcome ratings. 15 5. Ratings Outcome 5.1 Both RCIII and RCIV substantially achieved their real sector objectives. As for institutional development, RCIII made substantial progress toward achieving its main objectives. RCIV, which according to the ICR, "had higher aims" and included sub-sectoral policy objectives in addition to ABC's institutional development, is more tricky to rate. 5.2 A number of the institutional development and sector policy objectives of RCIV were either reduced in scope or made redundant by explicitly being taken over by the FSTA project. Other objectives were either not advisable according to the Bank's own financial sector strategy (recapitalization) or were not feasible (rationalizing the interest rate structure, change in provisioning practices). At the same time, the studies aimed at improving ABC's operations had limited impact, and the project objectives specifically aimed at upgrading ABC's financial management were not achieved. In this respect, the consistent violation of some loan agreements, and the neglect of others, is a notable shortcoming of the project. These violations should have led to either a suspension of disbursements to enforce compliance or to an amendment of the DCA if the borrower, ABC, and the Bank agreed that the covenants were unrealistic. 5.3 The audit concurs with the ICR in rating the outcome of RCIII as satisfactory. However, for RCIV, the audit disagrees with the ICR rating of unsatisfactory and rates project outcome marginally satisfactory. The primary reason for the difference in rating of RCIV is that the ICR places greater weight on the non-achievement of financial sector policy objectives. The project, in the view of the audit, never had such objectives; they were clearly not incorporated in the legal agreements. Yet, in light of the modest achievement of its stated and relevant institutional development objectives, the audit does not consider the outcome of RCIV as fully satisfactory. Institutional Development 5.4 The audit agrees with the ICRs that the achievement of institutional development objectives by both RCIII and RCIV was modest. RCIII had both positive elements (staff training program, upgrading of training colleges, successful sub-project management, collection performance, impact on ABC's profitability) and negative elements or those that were only partially achieved (partially negative interest rates, poor outcome of the studies program, limited technology transfer, and little impact on RCC development). In RCIV, as in RCIII, while staff training was reasonably successful, the achievement of other institutional development objectives was limited. Sustainability 5.5 For RCIII, sustainability is rated likely while for RCIV it is rated uncertain; both ratings are the same as in the ICRs for the two projects. For both projects, the sustainability of the real sector benefits is likely. The difference between the two projects is in the institutional sustainability aspect. The sustainability of ABC as a financial institution depends on its autonomy in business decision making and the progress on financial sector reforms. Although recent events are encouraging, experience with financial sector reforms in China suggests that the process is likely to be uncertain. In terms of their contribution to ABC's financial condition, RCIII is rated as positive and RCIV as neutral. The FX risk under RCIII was borne by the government and under RCIV by ABC. Without knowing the 16 full consequences of the FX business on ABC in financial terms, the sustainability of RCIV outcomes is uncertain. For the government, of course, the impact of the two projects would be reversed. Bank Performance 5.6 Bank performance is rated as satisfactory for RCIII but unsatisfactory for RCIV. In RCIII, Bank performance at all stages is rated as satisfactory (rather than the highly satisfactory rating given in the ICR for supervision). For RCIV, the identification and preparation assistance by the Bank are rated as satisfactory, but appraisal and supervision as unsatisfactory. The appraisal was marked by a prolonged debate on the role of the project as a vehicle for reform of the entire financial sector. Eventually, financial covenants included in the loan agreements were not entirely consistent with the concurrent dialogue with the government on financial sector reforms. Supervision of the project is rated unsatisfactory because of the low frequency of supervision missions, a high turnover of task managers and a relatively lower priority attached to policy and institutional development objectives relative to the technical aspects of sub-projects. Partly because of these problems, supervision missions were deficient in either overlooking or not pursuing some covenant violations. A notable shortcoming was the failure of the Bank to insist, in line with the loan agreements, on at least some increase in interest rates within the 20 percent variance allowed by PBC. This approach, as well as the lack of action on some other covenants noted above, signaled to the borrower that the individual legal covenants of the loan agreements were not to be taken seriously. Borrower Performance 5.7 Borrower performance is rated as satisfactory for RCIII but unsatisfactory for RCIV. The audit concurs with the ICR's assessment of borrower performance at various stages with the following qualifications. The audit considers the non-compliance with loan covenants as more serious in RCIV. While the deficiencies in the Bank's supervision efforts in not detecting or not pursuing covenant violations may have contributed to this unsatisfactory performance, the audit considers covenant compliance to be primarily the borrower's responsibility. The audit also notes that while the impact of institutional development was limited-due partly to the non-integration of RCIV activities with the main stream credit department, as noted by the ICR-the integration was not covenanted because the Bank's efforts to make this part of the project failed even before negotiations. 17 6. Lessons 6.1 For both RCIII and RCIV, the frequency of supervision was insufficient for projects of their size and complexity. In this regard, the conclusion of the ICR for RCIV that projects must be kept simple enough that both Bank and implementing agency resources are not stretched beyond capacity to supervise implementation is a valid lesson. At the same time, though, once a project is designed in a particular way and the Bank has accepted and approved it, the resources to adequately supervise its implementation must be made available. 6.2 It is surprising and disturbing that both projects had covenants on positive real interest rates and yet no supervision mission paid attention to the impact of inflation. Two possible factors influenced this neglect: One is that the supervision missions were preoccupied with the physical implementation of sub- projects. The other is the lack of attention to detail in filling out form 590, particularly the Status of Legal Covenants. The lesson here is that supervision missions, especially in times of high staff turnover, should be well versed in all aspects of project objectives, design, and binding legal agreements. 6.3 The lack of technology transfer is an old problem, already identified by the PAR for RCI in 1992.27 The projects could have improved on the situation by either insisting on the availability of technical staff within ABC or providing training for the staff of the design bureaus or technical expert groups to update their technical knowledge. A general lesson to be drawn from this is that it is not sufficient to impose procurement guidelines and provide access to international markets; knowledge of best practices and technical advice must also be made available to investors and lenders. 6.4 Both projects essentially paid lip service to the Bank's overarching goal of poverty reduction. Specific targets or key indicators for poverty alleviation should have been included to monitor the projects' impact. Related to this is a notable shift in the attitude and portfolio composition of ABC away from the rural smallholder investments and toward larger agro-processing and other investments. This finding confirms the general perception that commercial financial institutions are likely to focus on more profitable large-scale and urban projects. The impact of this trend on poverty reduction is an important unanswered question. 6.5 The lessons contained in the ICR on RCIII are important and relevant. In particular the reference to the tight phasing of four projects in seven years confirms what is said above. To say later that many of the lessons were incorporated into the preparation of RCIV is somewhat contradictory and does not reflect what has actually happened. 6.6 The ICR for RCIV notes the inadequacy of a financial intermediary loan to achieve reforms for the whole financial sector. While true in general, this lesson is not relevant for the project. A more appropriate lesson is that it is difficult to introduce commercial banking only in one sub-sector, without regard to the overall sectoral or institutional environment. 27. The comment from the PAR still holds: "It is difficult to see how a modernization drive through technology transfer. ..can succeed if technically qualified staff are not available, domestic supplies are sometimes rudimentary and lack efficiency, and procurement arrangements do not require competition in world markets." 18 6.7 Another lesson from the PAR of RCI still applies. The filing system for project correspondence files continues to be a poor state. The files for RCIII are particularly "thin"; the situation is somewhat better for RCIV but far from satisfactory. While this may be due to the various reorganizations over time, the importance of maintaining institutional memory especially with a high turnover of task managers cannot be overstated. 19 Annex A Basic Data Sheet THIRD RURAL CREDIT PROJECT (CREDIT 1871-CHA) Key Project Data (amounts in US$ million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 340.0 Credit amount 170.0 163.9 96.428 Cancellation - - 0 Date physical components completed 12/31/92 6/30/94 Economic rate of return (percent) 20 20 83.34 Cumulative Estimated and Actual Disbursements FY88 FY89 FY90 FY91 FY92 FY93 FY94 FY95 Appraisal estimate (US$M) 14.0 30.0 70.0 120.0 160.0 170.0 Actual (US$M) 13.6 91.9 125.4 148.7 147.8 158.1 163.8 163.9 Actual as % of appraisal 97.1 306.3 179.1 123.9 92.4 93.0 Date of final disbursement: July 7, 1994 Project Dates Original Actual Identification 09/10/86 09/10/86 Preparation 12/13/86 12/13/86 Appraisal 04/15/87 05/07/87 Negotiations 10/01/87 12/11/87 Board presentation 11/17/87 01/26/88 Signing 01/03/88 02/29/88 Effectiveness 04/01/88 06/17/88 Project completion 12/31/92 06/30/94 Closing date 06/30/93 06/30/94 28. The credit was fully disbursed; the difference reflects SDR/US$ fluctuation. 20 Staff Inputs (staff weeks) Planned Revised Actual Weeks $ Weeks $ Weeks $'000 Through appraisal na na na na 104.2 227.8 Appraisal-board na na na na 82.9 165.2 Board-effectiveness na na na na 6.4 18.5 Supervision na na na na 136.9 312.9 Completion na na na na 11.0 20.2 Total 341.4 744.7 Mission Data Date No. of Staff days in Specializations Performance rating' Types of (month/year) persons field Represented' problems Implem. Development Status Objectives Through 09/86 Appraisal 11/86 10 35 FS/EC/AGS/AQS/LS/PS/ APS/HRD Appraisal 04/87 8 25 FSIEC/AGS/LSIAPS/ HRDS 09/87 1 14 FS Supervision 1 07/88 6 30 FS/EC/AGS/APS HS NR Supervision 2 03/89 3 10 FS/EC/AGS/APS HS NR Supervision 3 10/89 1 7 FS HS NR Supervision 4 11/90 7 20 FS/EC/AGS/APS/AQS HU NR Supervision 5 11/91 7 20 FS/EC/AGS/APS/AQS U NR PP/TA/EI/PM/TRG Supervision 6 11/92 6 15 FS/EC/AGSIAPStLS U NR PPITA/El/PM Supervision 7 08/93 2 15 ENV/APS U NR TA Supervision 8 05/94 1 7 FS S NR Completion 11/94 3 14 FS/EC/APS S S 1: AGS: Agricultural Specialist; APS: Agroprocessing Specialist; AQS: Aquacultural Specialist; EC: Economist; ENV: Environment Specialist; FA: Financial Analyst; HRD: Human Resource Development Specialist; LS: Livestock Specialist; PS: Procurement Specialist. 2: HS: Highly satisfactory; S: Satisfactory; U: Unsatisfactory; HU: Highly unsatisfactory. 3: El: Environmental Issues; PM: Project management; PP: Procurement Progress; TA: Technical. 21 FOURTH RURAL CREDIT PROJECT (CREDIT 1871-CHA) Key Project Data (amounts in US$ million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 550 64% 117.8 Credit amount 75 75 0 Cancellation - - 0 Date physical components completed 06/30/95 12/31/96 - Economic rate of retum (percent) 27 24 88.9 Cumulative Estimated and Actual Disbursements FY88 FY89 FY90 FY91 FY92 FY93 FY94 FY95 Appraisal estimate (US$M) 30.0 80.0 150.0 220.0 275.0 170.0 Actual (US$M) 31.4 112.3 172.5 220.8 259.2 158.1 268.8 274.6 Actual as % of appraisal 104.8 140.4 115.0 100.4 94.3 93.0 Date of final disbursement: February 5, 1997 Project Dates Original Actual Identification 04/88 10/03/88 Preparation - 08/08/89 Appraisal 11/89 12/11/89 Negotiations 04/90 09/10/90 Board presentation - 10/30/90 Signing - 12/10/90 Effectiveness 12/90 02/22/91 Project completion 06/30/98 12/31/96 Closing date 12/31/95 12/31196 Staff Inputs (staff weeks) Planned Revised Actual Weeks $ Weeks $ Weeks $1000 Preparation to appraisal - - - 175.5 488.3 Appraisal-board - - - 70.8 209.4 Board-effectiveness - - - - 14.4 47.3 Supervision - - - 130.0 446.3 Completion - - - 22.5 75.0 Total - - - 413.2 1271.3 22 Mission Data Date No. of Staff days in Specializations Performance rating2 Types of (monthlyear) persons field Represented' problems Implem. Development Status Objectives Identification 10188 7 26 CS/EC/AWS/LS/AGS/ APS/FA Preparation 04/89 4 32 CS/EC/APS/FA Preappraisal 09189 10 35 CSIFAJECIAPS/AQS/ AGS/MS/HS/LS Appraisal 12/89 8 30 CS/HS/FA/ECIAPS/AGS/ LS/IDS Negotiation 09/90 4 CS/EC/LW/DS Board Approval 10/90 Signing 12/90 Effective 03/91 Supervision 1 05/91 5 21 CS/ENIGD/APS/LS 1 1 LC/PP Supervision 2 04/92 4 17 EC/CS/APS/AQS 2 1 OS/LC/PP/SP Supervision 3 09/92 9 23 EC/CS/APS/AQS/GS/ HS NR LS/MS/IE/AGS Supervision 4 11/93 4 22 APS/CS/AQS/OO 1 1 OS/LC/CF/SP Supervision 5 06/94 4 13 CS/APS/AQS/OO S S CF/SP/ES Supervision 6 05/95 4 13 CS/APS/AQS/OO S S SP Supervision 7 05/96 1 3 EC S S LC Completion 03/97 4 20 CS/EC/APS/AQS S S 1: AGS: Agricultural Specialist; APS: Agroprocessing Specialist; AQS: Aquacultural Specialist; CS: Credit Specialist; DS: Disbursement Specialist; EC: Economist; EN: Engineer; FA: Financial Analyst; GS: Grassland (Rangeland) Specialist; HS: Horticulture specialist; IDS: Institutional Development Specialist; IE: Irrigation Engineer; LS: Livestock Specialist; LW: Lawyer; MS: Marketing Specialist; 00: Operations Officer. 2: 1: Highly satisfactory; 2: Satisfactory; S: Satisfactory. 3: CF: Counterpart Funds; El: Environmental Issues; LC: Legal Covenants; OS: Overall Status; PP: Procurement Progress; SP: Studies Progress. 23 Annex B Interest Rates and Inflation Year 1988 1989 1990 1990 1991 1992 1993 1993 1994 1995 1.995 1996 1996 1997 1998 1998 Effective date Sep. 1 Feb. 1 Apr. 15 Aug. 21 Apr. 21 Jan. 1 May 15 Jul. 11 Jan. 1 Jan. 1 Jul. 1 May 1 Aug. 23 Oct. 23 Mar. 25 Jul. 1 Inflation 18.5 17.8 2.2 2.2 2.9 5.5 13.2 13.2 21.7 14.8 14.8 6.1 6.1 0.8 -2.6 -2.6 Rates for Time Deposits 1 Year 8.64 11.34 10.08 8.64 7.56 7.56 9.18 10.98 10.98 10.98 10.98 9.18 7.47 5.67 5.22 4.77 3 Years 9.72 13.14 11.88 10.08 8.28 8.28 10.80 12.24 12.24 12.24 12.24 10.80 8.28 6.21 6.21 4.95 5 Years - - 13.68 11.52 9.00 9.00 12.06 13.86 13.86 13.86 13.86 12.06 9.00 6.66 6.66 5.22 Rates for Term Lending 1-3 Years 9.90 12.78 10.80 10.08 9.00 9.00 10.80 12.24 12.24 12.96 13.50 13.14 10.90 9.36 9.00 9.00 3-5 Years 10.80 14.40 11.52 10.80 9.54 9.54 12.50 13.86 13.86 14.58 15.12 14.94 11.70 9.90 9.72 9.72 5+ Years 13.32 19.26 11.88 11.16 9.72 9.72 12.24 14.04 14.04 14.76 15.30 15.12 12.42 10.53 10.35 10.35 Notes Figures in bold represent negative real interest rates. 24 Annex C The Agricultural Bank of Ckina Attn: Mr. Gregory K. Ingram Date: June 10, 1999 Manager Sector and Thematic Evaluations Group Operations Evaluation Department From: Hua Rouming Fax No.: 86-10-68416011 Manager Foreign Funds Division International Department Agricultural Bank of China Pages: 1 Dear Mr. Ingram Re: Response to Draft Performance Audit Repor For RCIII and RCIV Projects of China This is referred to reflection of the Draft Performance Audit Report for the Third Rural Credit Project and the Fourth Rural Credit Project. We have carefully red the report and noticed that there are some changes in the Ratings compared with the conclusion of Completion Report for RCIV Project. We deem that the Report generally reflected the facts and we have no objection to the comments. However both the Ministry of Finance and ABC maintain the same position, as it was, toward the Bank's conclusion of Completion Report for RCIV Project that the design of the Project was over ambitious regarding to the project's financial sector reform. And what we can learn from the project is the objective and monitoring indicator designing should be more relevant. Yours incerely, Hua uoming Internati al Department Agricultural Bank of China