Moldova - Pre-export Guarantee Facility Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No: 17905 IMPLEMENTATION COMPLETION REPORT MOLDOVA PRE-EXPORT GUARANTEE FACILITY (Loan No. 3851-MD) May 11, 1998 Private/Financial Sector Development H:ungary, Czech Republic, Moldova, Slovak Republic, Slovenia Country Unit Europe and Central Asia Region 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.  CURRENCY EQUIVALENTS (as of January 23, 1998) Currency Unit Leu I Leu US$ 0.2132 US$ 1 = 4.6895 AVERAGE EXCHANGE RATES 1995 1996 1997 4.5 4.56 4.6 WEIGHTS AND MEASURES Metric System MOLDOVA 'S FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS IBRD - International Bank for Reconstruction and Development IDA - International Development Association MIGA - Multilateral Investment Guarantee Agency IFC - International Finance Corporation EBRD - European Bank for Reconstruction and Development EU - European Union GAU - Guarantee Administration Unit MOF - Ministry of Finance PGF - Pre-Export Guarantee Facility FSU - Former Soviet Union Vice President: Johannes Linn Director: Roger Grawe Sector Leader: Gerhard Pohl Responsible Staff; Onno Ruahl  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT MOLDOVA PRE-EXPORT GUARANTEE FACILITY (Loan No. 3851-MD) Contents Preface ......................................................................... Evaluation Summary ........................................................... ii Part I. Project Implementation Assessment ...........................................................1 Part II. Statistical Tables ........................................................... 10 TABLE 1: SUMMARY OF ASSESSMENTS .................................................................. 10 TABLE 2: RELATED BANK LOANS/CREDITS .................................................................. 12 TABLE 3: PROJECT TIMETABLE ................................................................... 12 TABLE 4: LOAN/CREDIT DISBURSEMENTS: CUMULATIVE ESTIMATED AND ACTUAL ....................................... 13 TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION .................................................................. 1 3 TABLE 6: STATUS OF LEGAL COVENANTS ...................................... . 14 TABLE 7: BANK RESOURCES: STAFF INPUTS ...................................... 14 TABLE 8: BANK RESOURCES: MISSIONS ...................................................................... 15 Appendix: A. Borrower contribution to the ICR 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.  IMPLEMENTATION COMPLETION REPORT MOLDOVA PRE-EXPORT GUARANTEE FACILITY (Loan No. 3851-MD) Preface This is the Implementation Completion Report (ICR) for the Pre-Export Guarantee Facility in Moldova, for which Loan 3851-MD in the amount of US$ 30 million equivalent was approved on March 9, 1995 and made effective on October 30, 1995. Given the lack of demand for guarantees, the contingent loan was closed on May 17, 1997, compared with the original closing date of December 31, 2003, at the request of the borrower. No disbursements were made as the contingent loan served the purpose of backing up guarantees that were to be issued by the Guarantee Administration Unit, and no guarantees were called or issued. The ICR was prepared by Marie-Sophie Tar and Onno Riuhl, ECSPF, of the Europe and Central Asia Region and reviewed by Marc Blanc and Kathleen McCollom. The Borrower's contribution is included as an appendix to the ICR. Preparation of this ICR was begun during the Bank's final supervision mission, in May 1997. It is based on material in the project file. The borrower's contribution to the ICR includes evaluations of the project and comments on the ICR from the Ministry of Finance, the Ministry of Economy and Reforms, and the National Bank of Moldova.  - ii - IMPLEMEENTATION COMPLETION REPORT MOLDOVA PRE-EXPORT GUARANTEE FACILITY (3851-MD) Evaluation Summary Introduction 1. The Pre-Export Guarantee Facility was prepared in the context of an IMF supported stabilization program and a structural adjustment program with Bank assistance. It was designed to address the lack of working capital finance which resulted from a tight domestic credit policy and a capital base which had been seriously eroded by inflation. Furthermore, the perceived government performance risk inherent in doing business in Moldova was inhibiting foreign traders and input suppliers from providing capital inputs directly to local enterprises. The agriculture and industry sectors were suffering in this environment as even potentially viable Moldovan agricultural and industrial enterprises with good export prospects were having difficulty in obtaining working capital. Project Objectives 2. The principal objective of the Pre-Export Guarantee Facility (PGF) was to attract foreign private fmance for pre-export transactions in a manner that would allow the market to select the most viable activities and bear the full commercial risk of each transaction. By providing credible guarantees against risks associated with government performance and war and civil disturbanceforce majeure, the PGF was aimed at mitigating risks that were inhibiting private input suppliers, trading companies, and commercial lenders from financing the provision of inputs for commercially viable production to Moldovan enterprises. 3. The PGF was designed in parallel with the Structural Adjustment Loan that was approved in December 1994. The PGF was aimed to support the structural reform program by: (i) enabling the private sector to act as a screening mechanism, identifying and channeling private resources to commercially viable enterprises and transactions; (ii) helping restore employment and viable production for export in the short term, while reforms were taking hold; (iii) helping Moldovan enterprises build links with foreign partners, diversifying their sources of supply and improving their access to markets; (iv) being sufficiently flexible to support more sophisticated financial instruments for the financing of pre-export transactions, as they were introduced in the Moldovan market; (v) promoting a transparent business climate in which the rule of law would prevail. Implementation Experience and Results 4. Project implementation was unsatisfactory as no guarantees were issued over a period of 22 months, from effectiveness in October 1995 to the closing date in July 1997. The loan was canceled at the request of the borrower, as guarantee activity was not expected to pick up. This project was the first of its kind, and, as such, in some ways an experiment.  - iii - 5. Several impediments to the issuance of guarantees emerged during implementation. The main problem was lack of demand for guarantees under the facility. Some factors which affected demand were specific to the design of the facility, such as pricing of guarantees, while others were specific to Moldova, such as the provision of full Govermment guarantees to selected firms. 6. Initially, effectiveness delays resulted in a detrimental time lag between marketing efforts which had sparked interest in the guarantee facility among input suppliers and trading companies, and availability of guarantees which came only later, once the facility was fully operational in October 1995. As a result, demand for guarantees was not as high as it could have been in the early stages of the project, and momentum was lost. 7. The project was designed to offer guarantees for projects in both the agriculture and industry sectors. Transactions in the processing and light industry sectors suffered from a number of specific problems in the legal environment for business, as well as business practices in Moldova. In particular, it emerged that difficulties in assigning the benefits of a specific export contract to the bank which finances a transaction, and uncertainty regarding the ability to enforce such an assignment were an impediment to concluding deals. In addition, as most Moldovan enterprises conduct their export business on an undocumented basis, transactions cannot be financed by a commercial bank and cannot attract outside investment. Such transactions could not benefit from the PGF. Thus, although applications were received, the underlying transactions did not materialize as a result of problems which arose regarding the commercial aspects of the deals envisaged. 8. Different problems prevented the GAU from issuing guarantees in the agriculture sector. When the project was being prepared, it was expected that liberalization of the agriculture sector would be fast paced. Demonopolization of input supply and distribution did not occur, however, limiting the opportunities for foreign suppliers to enter the market through an emerging private distribution network. Foreign suppliers that did business in Moldova worked with Cereale and Fertilitatea, two large state- owned enterprises, and were offered full Government guarantees. The impact of the Government guarantees was twofold. First, these guarantees appeared to be more beneficial than those issued by the GAU, as they covered both commercial and political risks, and could be less expensive. Second, as some foreign firms had bad experiences with Government guarantees which had to be called and were not awarded, but had to go to arbitration, certain companies became wary of doing business in Moldova, regardless of the availability of guarantees. Demand for GAU guarantees suffered as a result. 9. In the initial implementation period, guarantees were priced too high relative to demand. All guarantees were 3% per annum of the underlying transaction amount, regardless of the type of transaction, and its maturity. A premium rate schedule was introduced in the Summer of 1996 to address this issue. 10. The institutional set-up of the project was not conducive to promoting the GAU and the guarantee facility within the Government. As foreign firms were accustomed to working with the Government and state-owned enterprises, this set-up also limited the GAU's links to foreign input suppliers. The GAU effectively operated as part of the Department of Foreign Economic Relations of the Ministry of Economy, which led to a marginal involvement of the Ministries of Agriculture, Industry, and Finance in the implementation of the project. Not only were valuable contacts lost as a result, but it appeared that business contacts heard contradictory stories from different Government officials on the availability of Government and PGF guarantees.  -iv - 11. The GAU's work was further affected by funding problems which led to the non-payment of salaries of GAU staff, and resulted in temporary demotivation of the staff. Although this problem was solved after some time, it contributed considerably to the loss of momentum in project implementation. Summary of Findings, Future Operations, and Key Lessons Learned 12. The experiences gained during implementation of the Pre-Export Guarantee Facility Project were useful in understanding under what circumstances such facilities could achieve the stated objectives, and which steps should be taken during project preparation to sound the market, tailor the product to the needs of the private sector, and establish the Borrower's ownership of the project. 13. The fact that the legal and business environment in Moldova had not undergone sufficient reform to allow foreign suppliers to enter the market without the comfort of commercial risk guarantees was a significant barrier to the success of the PGF. Indeed, in the later stages of the project, the Government requested that the Bank widen the coverage of the guarantees issued by the GAU to include commercial risks. The Bank felt that an important objective of the PGF was to unbundle political and commercial risks, so that the Government would be exposed only to those risks which it could control, i.e. the political risks, and did not agree to restructuring the project to allow for coverage of commercial risks. The experience gained in Moldova clearly demonstrates that Bank guarantees cannot overcome structural problems which exist in transition economies. Rather Bank guarantees can provide guarantee-holders with the comfort that, if structural reforms are reversed, or governments with little experience in running a market economy take specific actions which interfere with business transactions, their resulting losses will be covered. Preparation of similar projects should therefore involve an in depth analysis of the private sector's perception of risks involved in doing business in a specific country, in order to determine whether guarantees against political risks are sufficient to tilt the balance in favor of doing business there. An initial market survey to identify interested companies and financial institutions should be done. In the case of Moldova, a survey was conducted in March-April 1997, and the results showed a low level of interest. One of the conclusions of the survey was that the perceived high commercial risk for Moldovan enterprises indicated the need for further reforms in Moldova's private sector. 14. A survey can also be useful to improve the design of a guarantee facility by tailoring it to the specific needs of the private sector in a given country. It is important to incorporate a degree of flexibility in a political risk guarantee facility, and be responsive to market signals both during project preparation and implementation. In Moldova, it appeared that the guarantee rate that had initially been set at 3% for all transactions regardless of the type of transaction or its maturity was too high. The rates were adjusted in light of this information. Even though this change did not directly lead to the issuance of guarantees, it is useful to note that an important dimension of this kind of project is its degree of responsiveness and flexibility in terms of the types of transactions which can be covered, the fluctuations in demand, and the perceived risks. For example, a later project, similar to the Moldova PGF, the Bosnia and Herzegovina Emergency Industrial Re-start Project, included the imposition of a United Nations embargo as a covered risk, to adapt the guarantee to the conditions in the country. 15. The ownership of the Borrower is crucial to the success of a political risk guarantee facility project. In the case of Moldova, only one part of the Government was fully on board and supportive of the PGF. As stated above, the Ministry of Agriculture was offering unconditional sovereign guarantees, which were in fact competing with guarantees offered by the GAU. It was determined that the link between the GAU and the Ministry of Economy should be severed, and that, instead, the GAU should work more closely with the Ministry of Finance with specific liaisons to the Ministries of Industry and Agriculture. In terms of future projects, it is important to keep in mind institutional issues when  establishing the agency that will manage the guarantee facility so as to maximize its exposure both within and outside of government circles and at the same time ensure its independence from the government in terms of day to day management. The independence of the guaranteeing agency is central to the project's success as prospective guarantee holders must feel that the agency can issue guarantees, and review and award claims, if they arise, in an impartial manner.  IMPLEMENTATION COMPLETION REPORT MOLDOVA PRE-EXPORT GUARANTEE FACILITY (3851-MD) Part I. Project Implementation Assessment INTRODUCTION 1. At the time the Pre-Export Guarantee Facility was under preparation, Moldova was experiencing a wrenching economic contraction, with a cumulative decline in GDP since 1990 at over 60 percent, and a sharp fall in living standards. A macroeconomic stabilization program had achieved some significant results with inflation down to monthly rates of under 2 percent, and a fiscal deficit narrowed to 8 percent of GDP for the first part of 1994. However, adjustment in the real economy was occurring at a slow pace. State-owned enterprises continued to predominate, and claimed the larger share of credit, crowding out the, nascent private sector. The Government was under pressure to lend from the budget for seasonal agricultural finance, as banks, affected by tight credit ceilings, would not extend credit for these purposes. In addition, foreign traders and input suppliers were reluctant to provide working capital inputs directly to local enterprises due to the perceived government performance risk inherent in doing business in Moldova. Overall, the inefficient allocation of credit and productive resources in both banks and enterprises was directly threatening the stabilization program. 2. The Pre-Export Guarantee Facility was the Bank's fourth loan to Moldova. It was prepared in parallel with a Structural Adjustment Loan (US$ 60 million; Ln. 3815-MD), approved by the Board on December 8, 1994. The SAL focused on privatization, hardening the budget constraint on enterprises, creating a competitive environment, and better targeting the social safety net within fiscal constraints. The Pre-Export Guarantee Facility was designed to address the specific problem of access to credit for the productive sector in Moldova. 3. The project was not successful in achieving its objectives for a number of reasons which are spelled out in this report. The project was the first of its kind and has served as a useful learning process for future projects. This report identifies the main reasons for the unsatisfactory outcome of the project and presents suggestions for improvements in subsequent projects. PROJECT OBJECTIVES 4. The principal objective of the Pre-Export Guarantee Facility (PGF) was to attract foreign private finance for pre-export transactions, in order to increase the level of productive activity in Moldova, which was bounded by the lack of working capital available in the economy. Restraints on the pace of domestic credit creation limited the local banking system's provision of working capital for sound, commercially viable transactions. Furthermore, linkages with traditional trading markets had been broken, and the PGF was aimed at fostering new partnerships between Moldovan and foreign companies, and assisting Moldovan enterprises to diversify their sources of supply and improve their access to markets. 5. The PGF was structured to allow the market to select the most viable activities and bear the full commercial risk of each transaction. By providing credible guarantees against risks associated with  -2 - government performance and war and civil disturbanceforce majeure, the PGF was designed to mitigate risks that were inhibiting private input suppliers, trading companies, and commercial lenders from financing the provision of inputs for commercially viable production to Moldovan enterprises. 6. The PGF was designed in parallel with the Structural Adjustment Loan that was approved in December 1994. The PGF was aimed to support the structural reform program by playing a role in the transition to a market driven economy and improving the business environment in Moldova, both for local and foreign enterprises. These goals would be achieved by enabling the private sector to act as a screening mechanism, by identifying and channeling private resources to commercially viable enterprises and transactions. Guarantees were to be issued for deals where a Moldovan and a foreign company had already identified a transaction, and were to be granted on a first-come, first-served basis. Thus, the selection of covered transactions would have been market driven. The Guarantee Facility was designed to be sufficiently flexible to support more sophisticated financial instruments for the financing of pre-export transactions, as they were introduced in the Moldovan market, thereby playing a part in the development of the banking and financial sector. It was also envisioned that the guarantee mechanism would promote a transparent business climate in which the rule of law would prevail. It created disincentives for the Government to intervene in private sector transactions', and incorporated acceptable dispute resolution mechanisms in case claims occurred (international arbitration). The PGF was further aimed at helping structural reform by facilitating the transition period through its contribution to employment and viable production for export, while reforms were taking hold. EVALUATION OF OBJECTIVES 7. The objectives of the project were adapted to the country's needs at the time the project was developed. Moldova's economic development and transition would have greatly benefited from an increase in productive activity and commercial links with new markets and foreign companies, at a time when both production and employment were plummeting. Purchasers in the FSU Republics, the traditional market for Moldovan products, were often unable to pay for goods, and when payment was made, long delays on transferring the revenues to Moldova via the inter-bank system further eroded the working capital base of local enterprises. 8. The perceived political risk of doing business in Moldova was inhibiting western foreign input suppliers from providing working capital inputs directly to local Moldovan enterprises. Companies that had done business in other FSU countries had faced situations where the Government revoked export licenses or imposed retroactive changes in rules and regulations that prevented them from being paid for the inputs which they had already delivered to the local enterprise. These precedents were used to evaluate Moldovan risks, and severely restricted the flow of working capital to Moldova. Discussions with commercial lenders, input suppliers, and trading companies indicated that effective mitigation of political risk would significantly increase the volume of commercially viable pre-export transactions with Moldovan enterprises. 9. Most export credit agencies (ECAs) were off cover for Moldova, which made the case for a Bank-supported guarantee program compelling. Without Bank involvement, the resumption of short- term, commercial banking flows was highly unlikely, and would continue to restrain growth in productive activity and employment. To provide cover, ECAs required full counter-guarantees from the Government, including both commercial and political risks. This guarantee structure would not have Indeed, if the Government were to act in a way that resulted in a claim under the Guarantee Facility, this would result in a direct financial loss for Moldova, as the Bank would have a claim on the Government.  -3 - been desirable, had the Government been deemed creditworthy, since, from a policy standpoint, the government should not assume commercial risks. 10. The PGF was not intended to and could not address all of the barriers limiting foreign entities from taking financial exposure in Moldova, and concluding deals. Discussions with private sector players had seemed to indicate that the PGF would be sufficient to enable a critical mass of commercial transactions. During implementation, however, it appeared, that mitigating political risk was not enough to overcome other risks and barriers which stood in the way of commercial deals. The problems that were encountered are discussed in detail below. ACHIEVEMENT OF OBJECTIVES 11. The objectives of the project were not achieved. No guarantees were issued, and the project was closed in July 1997, while the original closing date was December 31, 2003. Although no guarantees were issued, the GAU was active in promoting and developing several potential transactions involving foreign comrnmercial banks and input suppliers. Over the implementation period, the GAU made three guarantee offers for deals that did not go through for a variety of reasons. Examples of deals that were envisaged include a working capital advance to a winery, a loan for importing oil drilling equipment, and a barter transaction involving the exchange of oil and electricity for cement. It is important to note that established export credit agencies see many potential transactions for each deal that materializes (a 1 to 10 success rate is not unusual). Thus, in a small economy like Moldova, where fewer opportunities for viable transactions exist, it could be expected that it would take time before a successful deal came through. 12. As mentioned earlier, the Moldova Pre-Export Guarantee Facility was the first project of its kind. Several problems led to the unsatisfactory outcome of the project. Many lessons were learned during project implementation which have subsequently been used when designing similar projects in other countries, including Bosnia and Herzegovina, Ukraine, and Albania. The following sections of this report will explain why the objectives were not achieved and will focus on project design, preparation, and implementation. MAJOR FACTORS AFFECTING THE PROJECT Demand 13. The main problem which arose was that demand for guarantees was not as high as expected. Several factors contributed to the lack of demand, some of which were independent of the project. 14. The PGF covered political risks, which were perceived to be high, and constitute a significant barrier to increased foreign investment in Moldova. However, commercial risks were also perceived to be high. Despite the fact that many foreign banks, input suppliers and trading companies had indicated that political risk cover would be sufficient to result in a number of transactions, their assertions did not translate into business deals, often because of the perceived commercial risks, which could not be covered. The results of the survey which was conducted in March and April 1997 indicated that 65% of the companies which responded were mostly concerned about non-performance of a potential Moldovan partner. 15. The participation of local banks, confirming the local partner's payment obligation, could have been a solution to this problem. However, local banks viewed pre-export transactions as risky and less  -4- profitable than their traditional lending activities and other investment opportunities in the local market, such as Government treasury bills. For example, a letter of credit confirmation facility was put in place by ING Bank through the Moldovan Agroindbank which was never used. The facility was designed so that ING Bank would confirm Agroindbank L/Cs, as long as the underlying transaction benefited from a GAU guarantee. Agroindbank did not make use of the facility, partly because it felt that returns on domestic lending were more favorable to its shareholders. Furthermore, Agroindbank was accustomed to receiving cash collateral from its clients involved in import transactions, and was not prepared to start taking commercial risk on these clients. Thus, neither local banks nor foreign input suppliers were prepared to take the risk of the Moldovan counterpart, and deals did not go ahead. 16. The general business and legal environment in Moldova constituted a significant limitation to the success of the project. A major impediment to the conclusion of commercially financed deals was that it was very difficult to assign the benefits of a specific export contract to a bank that would be financing the transaction. If one succeeded in assigning these benefits, the enforcement of such an assignment was highly uncertain. Transactions in the processing and light industry sectors were particularly handicapped by this weakness of the country's legal system. Indeed, the first type of transaction that foreign banks tend to consider when entering a new market is the so-called "packing credit" where the local enterprise receives a working capital loan from a bank to fulfill a confirmed export order. To limit the commercial risk to the performance risk under a specific transaction, the benefits of the export order are assigned to the financing bank. 17. The business practices of Moldovan enterprises further inhibited PGF-type transactions. It became apparent that most Moldovan enterprises conducted their business on an undocumented basis. This meant that they would export goods based on the expectation that their invoice would get paid, without a letter of credit or equivalent payment document. This was especially true for exports going to the CIS markets, which represented the bulk of exports. Business carried out on this basis could not attract outside investment, could not be financed by a commercial bank, and therefore could not benefit from the PGF. 18. Moldova's agriculture sector had been an important source of exports, and at the time of project preparation, it was assumed that input financing for the agriculture sector would be the PGF's main market. It was also expected that the liberalization of input supply and distribution in the agriculture sector would progress rapidly, creating opportunities for foreign suppliers to enter the market through an emerging private distribution network. This did not happen. Agriculture input financing continued to be dominated by unconditional Government guarantees for transactions through two large state-owned enterprises, Cereale and Fertilitatea. The impact of the Government guarantees was twofold. First, these guarantees appeared to be more beneficial than those issued by the GAU, as they covered both commercial and political risks, and could be less expensive. For example, a full Government guarantee costing a once only charge of 5% in respect to a five year loan period was offered to a company interested in purchasing equipment from Italian suppliers, while the GAU would have charged 3% per annum, and would have covered political risks only. Second, as some foreign firms had bad experiences with Government guarantees which had to be called and were not awarded, but had to go to arbitration, certain companies became wary of doing business in Moldova, regardless of the availability of guarantees. 19. The question of cost was another issue that limited the success of the PGF. Initially, the cost of guarantees was set at 3% per annum of the value of the underlying transaction. The price was the same for all transactions, regardless of maturity. It quickly became apparent that charging the same rate for transactions of different types and different maturities was not justified. Certain types of transactions  -5 - were inherently more risky than others, and, for similar transactions, those with longer maturities should be priced higher to account for the additional risk. A premium rate schedule was developed within six months of effectiveness to fine-tune the PGF. The 3% per annum rate remained in place for crop financing (sale of agricultural inputs in exchange for the following season's crop), and, for other types of transactions, the rate varied from 1.35 % per annum (up to 90 days credit period with a local commercial bank guarantee) to 3% per annum (up to 3 years credit period without a commercial bank guarantee). Lower rates for transactions involving local bank guarantees were introduced to encourage the involvement of the Moldovan banks (see para. 15 above). 20. Initially, effectiveness delays resulted in a detrimental time lag between marketing efforts during project preparation which had sparked interest in the guarantee facility among input suppliers and trading companies, and availability of guarantees which came only later, once the facility was fully operational. As a result, demand for guarantees was not as high as it could have been in the early stages of the project, and momentum was somewhat lost. This delay, combined with the high premium rate, may have played a role in dampening the initial interest demonstrated by input suppliers, trading companies and commercial banks during project preparation. 21. The small size of the Moldovan economy was an additional factor which limited demand for the PGF. As stated above, the GAU issued three guarantee offers and was actively involved in promoting and developing a number of transactions. However, the small size of the market led to take-off problems for the GAU. In fact, the Moldovan economy may not have been large enough to support a full-fledged guarantee agency, such as the GAU, especially given that there was no specific political risk that stood out as a critical barrier for increased foreign involvement, such as war risk for example. Institutional Issues 22. The institutional set-up of the project was not conducive to promoting the GAU and the guarantee facility within the Government. As foreign firms were accustomed to working with the Government and state-owned enterprises, this set-up also limited the GAU's links to foreign input suppliers. The GAU effectively operated as part of the Department of Foreign Economic Relations of the Ministry of Economy, which led to a marginal involvement of the Ministries of Agriculture, Industry, and Finance in the implementation of the project. Not only were valuable contacts lost as a result, but it appeared that business contacts heard contradictory stories from different Government officials on the availability of Government and PGF guarantees. 23. The GAU's work was further affected by funding problems which led to the non-payment of salaries of GAU staff, and resulted in temporary demotivation of the staff. Factors not generally subject to government control 24. Many factors described above were not subject to Government control. The most significant ones include the cost of the facility, the size of the economy, and the perception of high commercial risks. The fact that Moldova's legal and business environment were not yet sufficiently developed to provide comfort to foreign companies and facilitate commercial transactions, regardless of the availability of political risk cover, can also come under this heading, at least in the short-run. Moldova was undergoing a transition process which takes time, and it could not be expected that all the deficiencies would be resolved at once. Thus, the introduction of the PGF may have come too early in Moldova.  - 6- Factors generally subject to government control 25. Certain government actions had a negative impact on the implementation phase of the project. The main one, flagged above, is the issuance of unconditional Government guarantees for short-term transactions in the agricultural sector, which could have been one of the PGF's main markets. One of the policy objectives of the PGF was to end the Government's role in taking commercial risk on these types of transactions. Under the PGF, the Government would only have taken political risk, which are the risks that the Government is best able to manage and control. It appeared that the Ministry of Agriculture did not support this objective and continued to offer full guarantees to prospective input suppliers. The coordination among different ministries discussed under "Institutional Issues" exacerbated this problem, as the Ministry of Economy, that the GAU operated under, was not successful in marketing the GAU and the guarantee facility to other important ministries that were in direct contact with companies, such as the Ministry of Finance, Agriculture, and Industry. These ministries did not refer prospective transactions to the GAU, and did not inform private companies about the PGF. 26. The lack of reform in the agriculture sector was a further impediment to the success of the PGF. The Bank worked with the Government on this issue under the auspices of other Bank loans and technical assistance programs, but progress was not rapid enough to increase the private sector's participation in input supply and distribution to a level that would have created business for the PGF. Factors generally subject to implementing agency control 27. The Guarantee Administration Unit was responsible for marketing the PGF. It did a solid marketing job within Moldova which resulted in a high level of credibility of the facility in the Moldovan market. The GAU's marketing efforts outside Moldova were not as successful, however. The fact that the PGF did not get enough exposure abroad may have added to the difficulty of attracting foreign traders and input suppliers to the Moldovan market. Better marketing outside of Moldova may have facilitated deals between Moldovan enterprises and foreign partners, as more foreign companies may have considered doing business in Moldova had they been aware of the availability and benefits of the PGF. 28. It should also be noted that the GAU's work and results were affected by the funding problems mentioned above, which led to the non-payment of salaries of GAU staff, and resulted in temporary demotivation of the staff. PROJECT SUSTAINABILITY 29. As the project did not achieve its objective for the reasons described above, it is not meaningful to assess the probability of maintaining the projects achievements. BANK PERFORMANCE 30. The Moldova Pre-Export Guarantee Facility was the first project of its kind. Its identification and preparation were therefore managed differently from most Bank projects. Continuous dialogue among various Bank staff, involving a wide range of departments and specialties, was an integral part of project preparation. The concept of the project had to be developed and agreed upon. The project's structure was modified during the preparation process in order to create an instrument that could both effectively guarantee commercial transactions against political risks, and be in compliance with the Bank's internal requirements. The Bank's performance must be judged on the basis that this project was in some ways an experiment, and was meant to be a learning process for both the Borrower and the Bank.  -7 - 31. Despite the fact that the project team conducted discussions with a wide number of input suppliers, trading companies, and commercial banks, the demand for the PGF was incorrectly assessed. The preparation of the project could have benefited from a more in-depth study of the market, and of competing programs and instruments. In Moldova, a survey could have helped to identify that, although political risks represented an impediment to productive activity, commercial risks were a significant concern of foreign input suppliers and banks. A survey was conducted, but only in March-April 1997, a year and a half after project effectiveness. A better assessment of demand could have led to a smaller contingent loan to back up GAU guarantees which would have reduced the Government's commitment fees, and thus reduced the overall cost of the project. Had the commitment fees been lower from the start of the project, the Government may have been willing to maintain the facility for a longer time, thereby allowing the GAU to overcome initial difficulties, such as high premium rates, and lack of cooperation among ministries. 32. The fact that other agencies offered competing loan programs for private sector transactions should have been taken into account by the Bank when assessing the match between supply of and demand for guarantees, although these programs may have been developed after the PGF. However, the situation which arose points to a more general observation, regarding the need for better coordination and communication between multilateral and bilateral institutions. 33. The Bank could also have done more research on private risk insurers, and the pricing of their instruments. Indeed, the fact that the GAU only offered one premium rate, which was relatively high, was not in line with general practice in the risk insurance industry. The premium rates were adjusted in the Summer of 1996. The major marketing efforts of the PGF occurred before this adjustment, however, which could have had a negative impact on demand, and general interest in the facility. 34. The Bank's performance during supervision was good. Supervision was thorough from both a reporting and substantive standpoint. Project implementation was adequately reported, and implementation problems were identified and assessed. The Bank worked closely with the implementing agency in order to resolve the issues that have been discussed previously. Technical assistance to the GAU encountered difficulties early on, but they were resolved, and the GAU benefited from the presence of a foreign consultant on site for the duration of the project, who played an important part in identifying the barriers to increased demand for the guarantee. Solutions to some of these problems were proposed and put in place. The solutions, such as adjusting the premium rate schedule, may have been insufficient to overcome other impediments that were external to the project's design and implementation, but they addressed important issues and provided for a learning process that has been useful for future projects. BORROWER PERFORMANCE 35. The Borrower was responsible for certain institutional problems that hindered the GAU's performance. During project preparation, Bank staff and Government representatives in the relevant ministries (Agriculture, Industry, Economy, Finance) discussed and agreed on the project's design, goals, and institutional set-up. During project implementation, however, the GAU did not receive the support that it needed from the Government. The Ministry of Agriculture's unconditional sovereign guarantees to input suppliers severely restricted the GAU's ability to become active in the agricultural sector which had been identified as one of the most promising areas for the guarantee facility during project preparation. Neither the Ministry of Industry nor the Ministry Agriculture referred prospective private sector companies interested in doing business in Moldova to the GAU. Furthermore, it appeared that the Department of Foreign Economic Relations within the Ministry of Economy did not have enough clout within the Government to build up support for the GAU.  -8 - ASSESSMENT OF OUTCOME 36. Despite the fact the outcome of Moldova Pre-Export Guarantee Project was unsatisfactory, it was the first project of its kind, and provided a useful learning experience. The many lessons learned have enabled the Bank to identify, prepare and implement projects with similar goals and structure in other client countries, avoiding the many pitfalls that prevented the success of the PGF in Moldova. The following section summarizes the lessons of experience, and provides recommendations for future projects. KEY LESSONS LEARNED AND RECOMMENDATIONS FOR FUTURE OPERATIONS 37. During identification and preparation, it is crucial to assess the demand for guarantees against political risk. The approach to determining demand has to be free of any bias, which is quite difficult. As mentioned earlier, a survey to obtain initial feedback from potential users can be helpful. A survey, however, must be followed up with detailed discussions with respondents, both to get a more in depth understanding of the critical barriers to commercial transactions in a country, and, if there seems to be sufficient demand, to ask the private sector for its input regarding the design of the guarantee facility. If the facility is designed without taking into account the needs of potential users, demand may be low, not only in the case that there is little or no demand for political risk cover per se, but for other reasons such as the facility is not credible in the eyes of investors, it does not cover the most important risks, it is too expensive, it is not easily accessible, etc. Thus, many different aspects of a political risk guarantee facility need to be analyzed and discussed with potential users during the identification and preparation phase of this type of project. 38. The dialogue with the private sector is a two-way dialogue. Political risk guarantee facilities backed up by Bank funds are innovative and not well known in the market. It is therefore necessary to advertise the project to potential users in order to ensure a high enough level of demand. The guarantee agency which is established to administer the facility should have staff that is capable of marketing the facility effectively. The preparation phase should be used to establish a network of contacts with private sector companies, banks, chambers of commerce, etc., so that the facility is well known in relevant circles when it becomes operational. Ideally, a pipeline of applications of guarantees should be built during project preparation. 39. The institutional set-up of the guaranteeing agency is another important component of project preparation. As was pointed out above, the GAU in Moldova was not attached to the most appropriate ministry(ies) in terms of getting governmental support and exposure both within and outside government circles. The process of setting up the guaranteeing agency should be used to create a consensus within the government regarding the role of the agency, its mandate, and its level of independence. Bank staff should engage the government in a discussion to identify a suitable organizational structure and operational rules for the guaranteeing agency. 40. The commitment of the Government to economic reform should also be tested before creating a political risk guarantee facility. This comes down to assessing the Government's ownership of the project. The facility covers guarantee holders against government interference in their business transactions. If the Government interferes in business transactions, guarantees may be called, which would both create a direct Government obligation to the World Bank, and hurt the reputation of both the country as a place to do business. The ultimate goal of a guarantee facility is to build confidence in the government's commitment to maintain the rules of the game that apply to commercial transactions, so that a political risk cover is no longer needed to encourage private sector transactions. Thus, if it is  -9- determined, that the Government is likely to continue interfering in private business, it is preferable to withhold from offering political risk cover. In the case of Moldova, the fact that the Government did not liberalize the agriculture sector as quickly as planned, and continued to offer unconditional sovereign guarantees was a sign that it was not ready to let go of one of the most important sectors in the country's economy.  -10- Part II. Statistical Tables Table 1: Summary of Assessment Table 2: Related Bank Loans/Credits Table 3: Project Timetable Table 4: Loan/Credit Disbursements: Cumulative Estimated and Actual Table 5: Key Indicators for Project Implementation Table 6: Status of Legal Covenants Table 7: Bank Resources: Staff Inputs Table 8: Bank Resources: Missions Table 1: Summary of Assessments A. Achievement of Objectives Substantial Partial Neglible Not applcable Macro Policies 5 5 Sector Policies E E Cl Financial Objectives [ E a Institutional Development E E El Physical Objectives O 5 al Poverty Reduction [ [ ,l Gender Issues O E El Other Social Objectives E E El Environmental Objectives E E E Public Sector Management a E El Private Sector Development E E El Other (specify) n o E E (Continued)  B. Project Sustainability Likely Unlikely Uncertain (ef (if) (if) C. Bank Performance satisfactory Satisfactory Deficient (V) (V) V/) Identification i a: Preparation Assistance n 5 Appraisal VI 'f Supervision n O Highly D. Borrower Performance satisfactory Satisfactory Deficient (i) (f) (f) Preparation a i O Implementation E D Covenant Compliance i E Operation (if applicable) if ighly E. Assessment of Outcome satisfactory Satisfactory Unsatisfactory unsatisfactory (i/) (i) () () El El El  - 12 - Table 2: Related Bank Loans/Credits Loan/credit title Purpose Year of Status Approval Preceding operations 1. Structural Adjustment Loan Promote a recovery in sustainable, non- 1994 inflationary growth by stimulating competition throughout the economy, increasing the responsiveness of economic agents to the new economic environment, while maintaining a minimum level of social protection Table 3: Project Timetable Steps in Project Cycle Date Planned Date Actual Latest Estimate Identification (Executive Project Summary) Preparation Jun. 18 - Jul.2, 1994 Jun. 18 - Jul. 2, 1994 Appraisal Nov. 15 - Nov. 27, 1994 Nov. 26 - Dec. 10, 1994 Negotiations February 1-2, 1995 February 1-2, 1995 Letter of Development Policy (if applicable) n/a n/a Board Presentation March 1, 1995 March 9, 1995 Signing April 19, 1995 April 19, 1995 Effectiveness April 1, 1995 October30, 1995 Midterm review (if applicable) n/a n/a Project Completion December 31,2003 May 17, 1997 Loan Closing December 31, 2003 May 17, 1997  - 13 - Table 4: Loan/Credit Disbursements: Cumulative Estimated and Actual (US$ thousands) FY96 FY97 FY98 FY99 FYOO Appraisal estimate 0 0 0 0 0 Actual 0 0 0 0 0 Actual as % of estimate 100 100 100 100 100 Date of final disbursement n/a Table 5: Key Indicators for Project Implementation Estimated Actual L Key implementation indicators in SAR 1. Issuance of Guarantees n/a 0 II. Other indicators - Development Objectives 1. Amount of coverage issued, not n/a 0 leading to any claims under the guarantee 2. Increase in short to medium term n/a 0 import-export transactions financed by non-barter trade finance mechanisms  - 14- Table 6: Status of Legal Covenants Moldova Pre-Export Guarantee Facility Covenant Present Original Revised Description of Covenant Comments Agreement Section Type Status Fulfillment Fulfillment Date Date Project 4.01 1 C n/a n/a (a) The GAU shall maintain records and accounts n/a Agreement adequate to reflect in accordance with sound banking practices its operations and financial condition, including records and separate accounts for the Project. (b) The GAU shall: (i) have the records referred to in Section 4.01 of the Agreement, its accounts and financial statements (balance sheets, statements of income and expenses and related statements) and records and accounts for the Guarantee Account each fiscal year audited, in accordance with appropriate auditing principles consistently applied, by independent auditors acceptable to the Bank; (ii) furnish to the Bank as soon as available, but in any case no later than six month after the end of each such year: (A) certified copies of said financial statements for such year as so audited and (B) the report of such audit by said auditors of such scope and in such detail as the Bank shall have reasonably requested; and (iii) fumish to the Bank such other information conceming said records, accounts and financial statements and the audit thereof, as the Bank shall from time to time reasonable request. Covenant types: Present Status: 1. = Accounts/audits 8. = Indigenous people C = covenant complied with 2. = Financial performance/revenue generation from 9. = Monitoring, review, and reporting CD = complied with after delay beneficiaries 10. = Project implementation not covered CP = complied with partially 3. = Flow and utilization of project funds by categories 1-9 NC = not complied with 4. = Counterpart funding 11. = Sectoral or cross-sectoral budgetary 5. = Management aspects of the project or executing or other resource allocation agency .12. = Sectoral or cross-sectoral policy/ 6. = Environmental covenants regulatory/institutional action 7. = Involuntary resettlement 13. = Other  - 15 - Table 7: Bank Resources: Staff Inputs Planned Revised Actual Stage of project cycle Weeks US$ Weeks US$ Weeks US$ Preparation to appraisal 30.0 100.0 30.0 100.0 27.9 95.0 Appraisal 10.0 40.0 10.0 40.0 10.9 43.8 Negotiations through Board 8.0 18.0 8.0 18.0 7.2 16.7 approval Suapervision 40.0 124.2 23.6 64.8 29.1 84.3 Completion 8.0 15.3 8.0 15.3 8.0 15.3 Total 48 139.5 31.6 80.1 77.6 243.6 Table 8: Bank Resources: Missions Perfornance Rating2 Types of Stage of project cycle Month/ No. of Days in Specialization' Implem. Developm. Problems3 Year Persons Field status objectives Through appraisal 6/94 2 11 G 7/94 2 10 AB Appraisal through Board 11/94 5 10 G, L, EC approval Supervision 5/95 1 5 G S S S 7/95 1 5 G S S S, E 9/95 1 4 G S S S 2/96 1 5 G S S S, D, R 6/96 1 8 G U S S, D, G,R 10/96 1 14 G U U D, G, R 2/97 1 4 G U U D, G, R Completion 5/97 1 3 G U U D, G, R Total 79 1 - Key to Specialized staff skills: G- 2 - Key to Performance Ratings: S- 3 - Key to Types of Problems: S- GAU Guarantees; AB- Agribusiness; L- Legal; Satisfactory; U- Unsatisfactory salaries and staff; E- Effectiveness delays; EC- Export Credit D- Lack of Demand;; G- Government guarantees; R- Lack of reform in agriculture;  Appendix A The Ministry of Finance of the Republic of Moldova 2012, 7 Cosmonautilor St., Chisinau, Moldova No. 03/4-2/3-220/700 April, 13, 1998 Ref: 2207-242 April, 6, 1998 The Government of the Republic of Moldova Dear Sirs: In conformity with the Instruction of the Government of the Republic of Moldova no.2207-242 from April, 6, 1998 the Ministry of Finance has examined the Implementation Completion Report on the Pre-export Guarantee Facility, which was elaborated by the World Bank, and considers that the analysis of the failure of the Project is ample and complete. At the same time, we would like to point out the following. One of the main causes of the Project not being viable, was a wrong determination of the objectives at the preparation stage of the Project. Thus, the Facility was oriented to guarantee only political risks for Westem working capital provided for the production of goods with export potential. At the same time, the Report does not stress that, while the Project was being developed and the credit agreement was signed, there did not exist any specific political risk which would constitute a serious barrier to the increase of foreign investments, both in the country and the region. The country's political stabilization was also implicitly shown in 1995 by the successful results of the macroeconomic stabilization policy, which indicated the Republic's orientation towards reforms and creation of a market economy. Thus, the Project was not viable from the date of its constitution, and the deficiencies caused by the management of the Project and other conditions, which were defined in the report as barriers to the implementation of the Project, played a secondary role during the implementation of the Project. One of the arguments specified as a barrier to Project implementation was its institutional structure, i.e., the establishment of an Agency within the framework of the Department of Foreign Economic Relations that led to the lack of direct connections with other governmental bodies, such as the Ministry of Finance, the Ministry of Agriculture and the Ministry of Industry which had to participate in the implementation of the Project, even though the establishment of this Agency under any of these ministries would have produced the same effect. From this point of view, an optimal option would have been to establish an independent and non-governmental agency. In case of project viability, this structure would provide more confidence among investors.  Regarding the lack of governmental support to the Project, it should be mentioned that optimal coinditions were created for the activity of the Project and the need for additional support just indicates that the Project was not viable. It should also be pointed out that the failure of the Project imposed additional costs, which were not compensated by any positive economic effects. Respectfully yours, V. Chitan Minister  The Ministry of Economy and Reforms of the Republic of Moldova 2033, 1 Piata Marii Adunarii Nationale., Chisinau, Moldova No. 09-1333 April, 10, 1998 Ref: 2207-242 April, 6, 1998 The Government of the Republic of Moldova Dear Sirs: In conformity with the Instruction of the Government of the Republic of Moldova no.2207-242 from April, 6, 1998 on the Completion Report on the Pre-export Guarantee Facility presented by the World Bank, we would like to point out the following. The Ministry of Economy and Reforms generally agrees to consider the conclusions presented in the Report that refer to the implementation process and results of the project, and are impartial and well justified. We would like to point out that according to the Table (page 10) the majority of evaluation criteria is satisfactory but for the 8th criterion, which is inferior to the satisfactory level. In our view, this situation results from the impact of a range of objective factors: * in Moldova this Project was the first of its kind and, thus, an experimental one; * the conditions and the possibilities of the Project were new and difficult for local economic agents to assimilate; * political risks are not a priority in international transactions; * the implementation of the Project was based also on a range of reforms to be taken in different economic sectors (agriculture), the realization of which was delayed. We also consider that both the foreign personnel and the local staff of the Implementation Agency did their best to perform their functions, although the results show that the Project could have been more effective if the Agency had a mandate to provide guarantees for commercial risks. We would like to mention that this problem was discussed several times while the Agency was active, and that the suggestion to cover commercial risks was not accepted by the World Bank. At the same time, we consider that the proposal for the Agency to provide commercial risk guarantees was presented relatively late. Such proposals should have been discussed at the initial stage of preparation and negotiation of the Project, and were not implemented since the World Bank did not accept them.  In our opinion, this might be considered as the main deficiency of the Project. Under such conditions the suspension of the Project demanded by the Government was totally justified. To conclude, we would like to mention our agreement with the modest assessment given by the World Bank of the results of the Project, which points out that the Project did not achieve its objectives. Respectfully yours, D. Braghis Vice-minister  National Bank of Moldova 2006, 7 Renasterii Ave., Chisinau, Moldova Nr.10-01 125/18 March 13, 1998 We hereby deliver you the notification of the National Bank referring to the Final Report on the Pre- Export Guarantee Project (PGP) implementation. The Pre-Export Guarantee Facility was experimented for the first time in the Republic of Moldova without (i) conducting a thorough demand survey for guarantees against political risks; (ii) establishing favorable guarantee fees; and (iii) developing an efficient marketing system for this facility, and so the Project did not achieve its objectives, and had an impact on the budget which bore the costs related to the commitment fee until the facility cancellation date. As correctly mentioned in the submitted Report, commercial risk coverage is considered to have priority compared to political risk coverage. In our opinion, war risk, military conflict, as well as any restrictions on goods export and import imposed by the Government and the conversion and repatriation of currency by foreign investors covered by the PGP, are quite void of significance in the Republic of Moldova. This facility would be appropriate in the case of Bosnia and Herzegovina, referred to in the Report, where an embargo imposed by the UN was included among the political risks covered by the facility. This facility was approved in 1994 when the EBRD credit lines aimed at supporting small and medium size enterprises and providing micro-credit were not yet in place. These credit lines have proven to be more viable. Also, during the last years the amount of external private credits under nonresidents' contracts to increase working capital has gone up significantly. The credit lines provided by correspondent banks of Moldovan commercial banks for the management of letters of credit and collections have become a frequent practice. We consider that this facility had an indirect favorable impact on the improvement of access to credit for private enterprise, the enhancement of access to Western markets, and export promotion. Also, the PGP, being a new tool of the World Bank, contributed to the attraction of potential foreign investors to Moldova and the improvement of the country's image on international markets. We totally agree with the opinion that the Government should not provide guarantees covering commercial risks. Commercial risk guarantees must be provided exclusively by commercial banks. Along with the development of the financial sector, service sector expansion and the increase in competition among financial institutions, the banks will offer various guarantee tools against commercial risks at advantageous conditions and fees. Vice Governor Veronica Bacalu