financeTHA-ADB

Completion Report

Project Number: 32437 Loan Number: 1735 December 2005

Thailand: Restructuring of Specialized Financial Institutions

CURRENCY EQUIVALENTS

Currency Unit – baht (B)

At Appraisal At Present (14 November 1999) (26 October 2005)

B1.00 = $0.0257 $0.0245 $1.00 = B38.875 B40.810

ABBREVIATIONS ADB – Asian Development Bank BAAC – Bank for Agriculture and Agricultural Cooperation BOT – Bank of Thailand FPO – Fiscal Policy Office (Ministry of Finance) GDP – gross domestic product IFCT – Industrial Finance Corporation of Thailand IMF – International Monetary Fund MOF – Ministry of Finance NPL – nonperforming loans PSA – public service account RRP – Report and Recommendation of the President SFI – specialized financial institution SICGC – Small Industry Credit Guarantee Corporation SIFC – Small Industry Finance Corporation SME – small- and medium-sized enterprise TA – technical assistance TOR – terms of reference

NOTES

(i) The fiscal year (FY) of the Government ends on 30 September.

(ii) In this report, "$" refers to US dollars.

Vice President L. Jin, Operations Group 1 Director General R. Nag, Mekong Department Director R. Boumphrey, Governance, Finance, and Trade Division,

Mekong Department Team Leader J. Ahmed, Advisor, Mekong Department Team Members J. Ahmed, Advisor, Mekong Department M. Ventura, Associate Project Analyst, Mekong Department

CONTENTS

Page

BASIC DATA ii

I. PROJECT DESCRIPTION 1 II. EVALUATION OF DESIGN AND IMPLEMENTATION 1

A. Relevance of Design and Formulation 1 B. TA Loan and Grant Outputs 5 C. TA Loan and Grant Costs 7 D. Disbursements 7 E. Implementation Arrangements 7 F. Conditions and Covenants 8 G. Consultant Recruitment and Procurement 8 H. Performance of Consultants 8 I. Performance of the Asian Development Bank 8

III. EVALUATION OF PERFORMANCE 8 A. Relevance 8 B. Efficacy in Achievement of Purpose 9 C. Efficiency in Achievement of Outputs and Purposes 9 D. Preliminary Assessment of Sustainability 9 E. Environmental, Sociocultural, and Other Impacts 10

IV. OVERALL ASSESSMENT AND RECOMMENDATIONS 10 A. Overall Assessment 10 B. Lessons Learned 10 C. Recommendations 12

APPENDIXES 1. Estimated and Actual Project Costs 13 2. Status of Compliance with Loan Covenants 14

BASIC DATA A. Loan and Grant Identification 1. Country 2. Loan Number 3. Grant Number 4. Project Title 5. Borrower 6. Executing Agency 7. Amount of Loan 8. Amount of Grant 7. Project Completion Report Number

Thailand 1735 3355 Restructuring of Specialized Financial Institutions Kingdom of Thailand Ministry of Finance $4.5 million $3.0 million PCR: THA 929

B. TA Loan and Grant Data 1. Fact-Finding and Appraisal – Date Started – Date Completed 2. Loan Negotiations – Date Started – Date Completed 3. Date of Board Approval 4. Date of Loan Agreement 5. Date of Loan Effectiveness – In Loan Agreement – Actual – Number of Extensions 6. Loan Closing Date – In Loan Agreement – Actual – Number of Extensions 7. Terms of Loan – Interest Rate – Maturity (years) – Grace Period (years) 8. Grant -- Date of Effectiveness -- Date Completed

18 March 1999 10 April 1999 5 August 1999 24 November 1999 21 December 1999 12 May 2000 10 August 2000 10 August 2000 0 31 December 2002 11 March 2004 1 OCR LIBOR (converted on 15 December 2002) 15 3 19 September 2000 31 December 2005

OCR = Ordinary Capital Resources, LIBOR = London Interbank Offered Rate.

iii

9. Loan Disbursements a. Dates Initial Disbursement

11 April 2001

Final Disbursement

4 June 2003

Time Interval

26 months

Effective Date

10 August 2000

Original Closing Date

31 December 2002

Time Interval

28 months

b. Amount ($ million) Category

Original

Allocation

Last Revised

Allocation

Amount

Canceled

Net Amount

Available

Amount

Disbursed

Undisbursed

Balance Consultant 3.92 3.92 0.00 3.92 3.92 0.00 Manuals, seminar, training, and equipment

0.06

0.06

0.06

0.00

0.00

0.00

Unallocated 0.52 0.52 0.52 0.00 0.00 0.00 Total 4.50 4.50 0.58 3.92 3.92 0.00

10. Local Costs (Loan and Grant)a - Amount ($) 392,926 - Percent of Local Costs 10.8% - Percent of Project Costs 3.9% 11. Technical Assistance Grant Disbursementsb Category

Original

Allocation

Last Revised

Allocation

Contracts

Amount

Disbursed

Amount

Uncommited

Undisbursed

Balance Consultant 2,183,000 2,863,107 2,837,502 2,701,450 25,605 161,657 Equipment - 37,200 37,182 36,279 18 921 Seminars and Training

314,000

94,322

80,133

65,405

14,189

21,576

Contract Negotiation

6,000

2,090

2,090

2,090

0

0

Contingency 497,000 3,281 868 - 2,413 3,281 Total 3,000,000 3,000,000 2,957,775 2,805,224 42,224 187,435

C. Project Data (Loan and Grant)

1. Project Cost ($ million)

Cost RRP Estimate Actual

Foreign Exchange Cost 6.08 6.49 Local Currency Cost 4.68 4.51 Total 10.76 11.00 a Asian Development Bank (ADB) financing only. b ADB financing only.

iv

2. Financing Plan ($ million) Cost RRP Estimate Actual Implementation Costs Borrower-Financed 3.26 4.12a ADB-Financed 7.50 6.88 Total 10.76 11.00

ADB = Asian Development Bank, RRP = Report and Recommendation of the President.

3. Cost Breakdown by Project Component ($ million)

Component RRP Estimateb Actual Loan

IFCT, SIFC, SICGC BAAC 6.44 6.29

Grant Rationalization and Corporate Governance of SFIs 4.32 4.71 SME Financing Strategy

Total 10.76 11.00 BAAC = Bank for Agriculture and Agricultural Cooperation, IFCT = Industrial Finance Corporation of Thailand, RRP = Report and Recommendation of the President, SFI = Specialized Financial Institution, SICGC = Small Industry Credit Guarantee Corporation, SIFC = Small Industry Finance Corporation, SME = small and medium-sized enterprise.

4. Project Schedule

Item RRP Estimate Actual Date of Contract with Consultants Date of Award Ernst & Young April 2000 26 April 2001 Development Alternatives Inc. April 2000 23 May 2001 Sanwa Research Institute & Consulting Corp. April 2000 5 December 2000 Aries Group, Ltd. April 2000 13 January 2001 IFG Development Initiatives (C. Lin)c 17 June 2002 IFG Development Initiatives (G. Palmer)c 18 July 2002 IFG Development Initiatives (T. Reid)d 29 May 2003 UK National Audit Office (J.A. Cant)d 4 July 2003 Prasert Chokchaikasemsukd 21 July 2003 Completion of Work Ernst & Young October 2000 28 March 2003 Development Alternatives Inc. October 2000 18 November 2002 Sanwa Research Institute & Consulting Corp. October 2000 20 September 2001 Aries Group, Ltd. April 2002 30 April 2002 IFG Development Initiatives (C. Lin)c 31 December 2004 IFG Development Initiatives (G. Palmer)c 31 December 2004 IFG Development Initiatives (T. Reid)d 31 December 2004 UK National Audit Office (J.A. Cant)d 31 July 2003 Prasert Chokchaikasemsukd 31 January 2005 RRP = Report and Recommendation of the President. a Estimated. b Includes ADB and Government financing. c Continued the work of previous consultant (Aries Group). d Additional experts that required a minor change in implementation.

v

5. Project Performance Report Ratings

Ratings Implementation Period

Development Objectives

Implementation Progress

From 21 Dec 1999 to 31 Dec 1999 Satisfactory Satisfactory From 1 Jan 2000 to 28 Feb 2001 Satisfactory Satisfactory From 1 Mar 2001 to 31 May 2001 Satisfactory Highly Satisfactory From 1 Jun 2001 to 28 Feb 2002 Satisfactory Satisfactory From 1 Mar 2002 to 31 Jul 2004 Satisfactory Highly Satisfactory D. Data on Asian Development Bank Missions

Name of Missiona

Date

No. of Persons

No. of Person-Days

Specialization of Membersb

Fact-Finding-cum-Appraisal 4–9 Apr 1999 5 30 a, b, c Special Loan Administration 1 27 Sep 2000 1 1 c Special Loan Administration 2 14–15 Dec 2000 2 4 a, c Special Loan Administration 3 18–20 Dec 2000 1 3 d Special Loan Administration 4 24–26 Jan 2001 1 3 c Special Loan Administration 5 14–16 Feb. 2001 2 6 c, e Review 1 29 Mar–

4 Apr 2001 1 7 c

Inception 9–11 May 2001 1 3 e Review 2 1–5 Jun 2001 1 5 a Review 3 19–20 Jul 2001 2 2 a, c Special Loan Administration 6 8–9 Nov 2001 1 2 c Review 4 28 Feb 2002 1 1 c Review 5 25–28 Mar 2002 1 4 c Special Loan Administration 7 26–28 Jun 2002 1 3 c Review 6 30 Sep–4 Oct

2002 2 5 c

Review 7 6 Jun 2003 1 1 c Consultation 26–27 Jun 2003 1 2 c Project Completion Reviewc 17–21 Oct 2005 3 15 c a Between 22 November 2003 and 26 March 2005, the Extended Mission to Thailand administered the Project in

Thailand. b a - capital markets specialist, b - financial analyst, c - economist, d - consulting services specialist, e - control officer c J. Ahmed, Advisor, prepared the project completion report.

I. PROJECT DESCRIPTION 1. In December 1999, the Asian Development Bank (ADB) approved a technical assistance (TA) loan of $4.5 million from its ordinary capital resources for the restructuring of specialized financial institutions (SFI). At the same time, ADB approved a TA grant of $3.0 million, financed from the Asian Currency Crisis Support Facility, funded by the Government of Japan, to (i) develop strategies for financing small- and medium-sized enterprises (SME), and (ii) rationalize and enhance the corporate governance of SFIs. 2. Under an International Monetary Fund (IMF)-sponsored stabilization and structural adjustment program, supported by a $17.2 billion assistance package from ADB and other donor agencies, the Government of Thailand pursued an agenda to contain the effects of the 1997 Asian financial crisis. The aim was to put the economy back on a sustainable growth path. The TA loan and grant were in line with these objectives, and sought to improve the effectiveness of SFIs. The Project provided information on the SFIs, as well as governance improvement and capacity building, which the Government found very valuable. 3. The original purpose of the TA loan and grant, which were part of ADB’s assistance to Thailand for recovery from the Asian crisis, was to help the Government promote a more resilient financial sector. In particular, the TA loan was to assist the Government in devising appropriate restructuring plans, with institutional strengthening as needed, for the four selected SFIs: Bank for Agriculture and Agriculture Cooperatives (BAAC), Industrial Finance Corporation of Thailand (IFCT), Small Industry Finance Corporation (SIFC), and Small Industry Credit Guarantee Corporation (SICGC).1 4. The TA grant had two main objectives. First, it aimed to develop appropriate institutional mechanisms and strategies for SME financing. Second, the grant sought to rationalize the Government's role in the financial system and the SFIs, and to establish an appropriate governance structure for SFIs. 5. With Thailand's international financial position strengthened, the Government prepaid the TA loan on 27 May 2004 as part of its strategy of prudent external debt management.

II. EVALUATION OF DESIGN AND IMPLEMENTATION A. Relevance of Design and Formulation 6. Thailand’s fiscal management since 1997 has been of a high standard. Risks have been contained and considerable progress has been achieved in fiscal consolidation. At the same time, wide ranging and deepening reforms have been instituted in the civil service, public enterprises, and in budget management and formulation. The progress towards modern governance structures in the public sector and the withdrawal, or lessening, of public sector interventions in economic activities are impressive. 7. This review of the financial and governance arrangements of the SFIs should be seen against this background, as well as the improved governance arrangements emerging from Thailand’s 1997 Constitution. For the first time, the 1997 Constitution provides for accountability

1 ADB. 1999. Report and Recommendation of the President to the Board of Directors on the Proposed Technical

Assistance Loan and Grant to the Kingdom of Thailand for Restructuring of Specialized Financial Institutions. Manila (page 5).

2

of the state to the people, and for greater citizen participation in public policy at all levels.2 These rights are provided within the context of a free market economy that requires the Government to protect consumers from direct and indirect monopolies. Additionally, the state has agreed not to enter into businesses that compete with the private sector, except where necessary for the benefit of national security or public interest, or the provision of a public utility.3 8. The SFI Project was conceived in the midst of a crisis. It began implementation in the context of widening governance reforms, as well as responsible and skilled fiscal policy management. Against this backdrop, the SFIs posed a policy challenge because of their dual mandate: (i) to discharge specific public policy programs (directed lending); and (ii) to ensure that these programs are effectively managed, appropriately supervised, and commercially sustainable. While this challenge had appeared impossible to meet—in Thailand and elsewhere—the wider adoption of the innovative governance structures, the accounting framework, and the cooperative regulatory oversight developed by the Government under this Project can be a way to do so. 9. The SFIs were viewed, especially by the new Government,4 as valuable institutions for fighting rural poverty and promoting the development of SMEs. In this context, the Government’s three specific challenges were to (i) manage the SFIs’ short-term contributions to poverty reduction through a series of measures that required them to expand lending and other operations to vulnerable groups and communities, often on concessional terms; (ii) privatize or close selected SFIs as appropriate, and implement the required institutional and other changes to enhance the capacity of SFIs that remained in the public sector to meet their public service obligations; and (iii) contain SFI claims on the budget, while introducing measures that improved transparency and effectiveness in SFI operations, management, monitoring, and supervision. These broad challenges mirror those faced in other countries with similar institutions.5 10. The SFIs were a diverse group that included new, relatively weak institutions, such as SICGC, as well as established and renowned institutions, such as BAAC. Reflecting this diversity, IFCT was expected to make the transition to a deposit-taking institution, requiring corresponding changes in its operational and governance structures. Indeed, it was merged midway through the Project with a major commercial bank. SIFCT faced the challenge of transitioning to a bank, handicapped by its limited branch network, limited skills, and high level of nonperforming loans (NPLs) from the expansion of its operations in 1999 and 2000. SIFCT became the Small and Medium Enterprise Development Bank under adverse conditions. For SICGC, enhancing credit risk management skills was essential to protect its capital under these difficult lending conditions. For BAAC, the challenge was to consolidate its progress towards the

2 Article 6 of the 1997 Constitution states: “Every person has right, as provided in law, to participate in the

administrative process that has or may have impact on his or her rights or freedom.” 3 Article 87 of the 1997 Constitution states: “The state shall promote free market economy, ensure fair competition,

protect consumers, and prevent direct and indirect monopoly. The state shall repeal and shall not promulgate any law or order controlling business not in accordance with economic necessity. The state must not enter into business against private sector except where necessary for the benefit of national security, public interest, or the provision of public utility.”

4 Prime Minister Thaksin Shinawatra assumed office on 9 February 2001. 5 Until recently, Australia had a large SFI sector. While the Government has withdrawn largely from deposit-taking

and most lending operations in Australia, important public policies similar to Thailand’s remain, such as support for home ownership. However, this now is supported largely through grants, not lending, in Australia. See Vince FitzGerald. 2002. Accounting for the Public Interest in SFIs. Thailand Australia Capacity Building Facility.

3

primary goal of commercializing operations and profitability within its large rural branch network, while insulating these functions from its “program” or directed credits. 11. Overall, the Project was well conceived to respond to the challenges that the SFIs and the Government faced. A key concern was the possibility that large quasi-fiscal deficits and contingent liabilities would emerge on the SFIs’ balance sheets.6 That this did not occur is a measure of the continuity and commitment to the overall goals noted above. Nevertheless, as implementation began, the Government and ADB agreed that the passage of time, as well as the changed economic and policy circumstances, justified modifications to the weight of effort and allocation of resources to the Project’s components. These modifications did not alter materially the project objectives or scope, and were reasonable responses to the changing needs of the SFIs and the Government. 1. TA Loan Component 12. The TA loan provided for a financial audit of the SFIs’ operations to address their potential vulnerabilities. This operational audit was to establish the foundation for strengthening the finances and operations of these institutions, as necessary. This Project, therefore, was highly relevant to Thailand's recovery from the Asian crisis. It was intended to strengthen the resilience of the financial sector by addressing any potential problems in the SFI subsector, while ensuring SFI governance would be enhanced over the longer term. This TA loan was the largest ever provided to Thailand, reflecting the seriousness of the issues being addressed. The loan component of the Project was implemented effectively as designed. 13. More broadly, the TA loan component aimed to diagnose the organizational and institutional structure, operational performance, and weaknesses in financial solvency through an

(i) assessment of loan quality; (ii) assessment of institutional capacity, principally credit risk management,

Information Technology and Management Information System support systems, and operations efficiency; and

(iii) evaluation of the financial soundness of the SFIs. 14. Based on this diagnosis, the Project developed

(i) assessments of each SFI with respect to portfolio quality, institutional capacity, and financial soundness;

(ii) detailed restructuring and improvement (capacity building) plans; and (iii) training in loan restructuring, risk management, and other key areas, as

appropriate. 15. However, the weight given to each of these areas varied to reflect the needs and capacities of the institutions. 6 While the issue of contingent liabilities remains, especially in relation to SME Development Bank’s lending to SMEs,

their overall magnitude for all SFIs under worst case assumptions are 3%–3.5% of GDP. This is small relative to the Government guarantees on the debt of publicly owned enterprises (about 12% of GDP, though the public enterprises have positive net worth, so the figure of 12% overstates their fiscal risk). SFI contingent liabilities are insignificant compared with those associated with the blanket deposit guarantee issued in the aftermath of the 1997 Asian financial crisis.

4

2. TA Grant Component 16. The TA grant had two components. The first involved upgrading financial services to SMEs. This component was designed to be implemented in parallel with a World Bank program that addressed corporate debt restructuring. The intention of both programs was to accelerate the access to bank credit, thereby promoting an early economic recovery through increased investments. The Asian crisis had hit SMEs hard, and they were having great difficulties accessing bank credit. At the same time, SMEs accounted for 96% of registered enterprises and 70% of all industrial employment in the country. The inclusion of the development of an SME financing strategy under the TA grant was intended to enhance SME access to bank credit, and correspondingly to support an early industrial recovery. The SME financing strategy, based upon the experiences in many other countries, was developed effectively and presented to the Government in September 2002. 17. The other major component of the TA grant involved the difficult issue of governance of SFIs, which required comprehensive reforms. As in all state-owned enterprises, policy making needed to be separated from regulation—and both, in turn, from the provision of financial services. For the SFIs, the Ministry of Finance was overseeing policy making and regulation, although the Bank of Thailand has been delegated to undertake SFI inspections. In addition, the management of SFIs’ public sector obligations needed to be reformed, e.g., the debt forgiveness programs for small borrowers who had loans through BAAC or SME Bank. Often such public service obligations, where funded, had been compensated through capital injections. More transparency was needed in committing to such public sector obligations and funding them. 18. In this respect, the key difficulty for the Government in monitoring and evaluating the performance of the SFIs was their mixture of commercial and policy activities. Previously, identifying which business stream—commercial or policy—contributed what results to an SFI’s performance was impossible. SFIs that sustained sizeable costs or losses carrying out policy programs usually needed to seek additional capital to restore their lost equity. Generally, such requests were ad hoc, occurring when the SFI encountered a liquidity shortage or a deficiency in capital adequacy. There was no organized methodology for assessing which part of an SFI’s business caused the losses, or for estimating what future costs or losses might be sustained. 19. The motives for distinguishing or segregating the two business streams were to

(i) enable a clearer measurement of the SFI’s past performance, without the uncertainty created by mixing commercial and policy results;

(ii) enable more reliable budgeting for the SFI’s future performance; (iii) allow the Government to determine the future funding needs of the SFI more

accurately; and (iv) assist the Government in assessing the SFI’s long-term viability more accurately.

20. The PSA system, therefore, was designed to distinguish the policy accounts of the SFI from its normal commercial accounts. This was accomplished by segregating the policy accounts under a public service account, and then allocating the costs related to the policy accounts to the PSA, as far as practical. 21. In addition, the development of corporate governance in the SFIs faced implementation difficulties arising from its design. The consulting contract for the development of the corporate governance component was terminated in March 2002 due to nondelivery of key components,

5

namely a corporate governance framework and the associated monitoring and surveillance systems. While a consulting contract failed, a fundamental difficulty arose from the design of the process. ADB staff and relevant Government officials conceptualized the required interactive process to include (i) review of existing materials and development of the broad terms of reference (TORs) for the design of a governance implementation plan; (ii) detailed design of the scope of the implementation plans, and an assessment of the resources needed for their implementation; and (iii) implementation of all designed systems in the four beneficiary SFIs (paras. 27-31). The expected outputs reflected the original objectives and scope of the TA grant. Their phased implementation required a high level of staff inputs. This resulted in a sharper focus on key expected project outputs: (i) a system of PSA;(ii) a monitoring and reporting system for Fiscal Policy Office (FPO) and SFIs; (iii) a set of corporate governance codes and implementation guidelines; and (iv) capacity building in the SFIs, FPO, and other government agencies. Almost all SFIs, not just the four original targets, received capacity building support under the TA grant. 22. ADB recruited individual experts to address these corporate governance TORs, which were augmented by major input of staff resources to coordinate those inputs and ensure full Government involvement and ownership. This considerable input of staff time and resources, and the extensive dialogue with all of the agencies involved in the Project, was necessary to win acceptance and ownership of the TA outputs, given the initial failure to meet Government and ADB expectations under the contract. These changes reflected the need to select highly qualified individual consultants to ensure leading edge governance inputs, while providing an overall framework for policy dialogue. This continuity in the policy dialogue, which entailed soliciting the views of relevant Government officials and ensuring effective incorporation of their feedback, was necessary to ensure Government ownership. B. TA Loan and Grant Outputs

1. TA Loan Component: Financial and Operational Audits

23. The objective of the TA loan component was to help the Government devise appropriate restructuring plans, including any required institutional strengthening, for the four SFIs: BAAC, IFCT, SIFC, and SICGC. The TA loan funded a financial audit and a diagnostic review of each of these SFIs, based on international accounting and reporting requirements. This provided the informational base for designing the required financial and operational enhancements, including governance reforms.

24. The Ministry of Finance had established a set of broad principles and good governance practices for the SFIs to follow. These included (i) full transparency in accounting and reporting; (ii) use of external auditors and International Accounting Standards; (iii) full compliance with Bank of Thailand regulations for activities that compete with commercial banks; and (iv) capital adequacy, liquidity, and solvency assessed in accordance with the prudential standards for commercial banks. An equivalent set of principles was in place for risk management, fiscal treatment, role and scope of activities, recapitalization, and supervision. Based on these financial and operational principles, the financial and operational audits were completed satisfactorily for the four SFIs. These audits provided sufficient information to undertake the remedial actions, as required relative to meeting the board objectives that had been set by the Ministry of Finance. 25. The loan-financed consultants conducted financial and operational audits of the four selected SFIs: BAAC; IFCT, which subsequently was taken over by Thai Military Bank; SIFC,

6

which subsequently became the SME Development Bank; and SICGC. The consultants also proposed measures for institutional strengthening, as required. The consultants’ reports generally were well-implemented. However, while the consultants reflected international best practices in their reports, they often did not adapt those practices effectively to the local conditions in Thailand. In addition, the project design incorporated very little help or guidance for adjusting those international best practices and implementing them as appropriately modified. This is a key lesson that needs to be reflected in future ADB projects, such as in the Greater Mekong Subregion countries, where ADB might undertake similar projects in the future, possibly in partnership with Thailand. 26. The updating of these financial and operational audit results revealed one major problem at the SME Development Bank: an NPL level of more than 21%. The review mission viewed these NPL results as indicative of a banking institution that needs to strengthen its credit approval procedures. Given the Ministry of Finance’s maximum target for NPLs of 15%—in itself a generous allowance—SME Development Bank required immediate corrective action before its balance sheet deteriorates further. At a minimum, a major shift toward a market-based credit culture was required for its lending committees. Further, monthly monitoring of new NPLs was essential for all loan sizes and by all clusters. Another concern was the increasing loan sizes and the market perception of potential political influence, however ill founded, in lending operations. 2. TA Grant Component: SMEs and SFI Governance 27. The TA grant had two major components: (i) develop appropriate mechanisms and strategies for SME financing; and (ii) rationalize the Government’s role in the financial system and the SFIs, and establish an appropriate governance framework for SFIs. 28. The SME study provided a useful set of recommendations that were based on approaches used successfully in other countries. Perhaps the most enduring has been the need to support the borrowing entrepreneurs with technical skills and marketing support. 29. Under the SFI governance reform component, the Fiscal Policy Office of the Ministry of Finance embarked on a program to restructure and strengthen the corporate governance of the SFIs. As part of these efforts, the costs associated with PSA were to be segregated, and a sound reporting and monitoring system for SFIs was to be established within FPO. Applied training for FPO officials on the technical and policy sides supported the governance system that emerged from the inputs of Government officials, individual governance experts financed under the TA grant, and ADB staff. As a further consequence of these efforts, the informal unit in FPO that was responsible for supervising the SFIs was upgraded to a formal unit in 2002 to better address these governance issues. The governance reform program has been increasing the efficiency and transparency of SFIs through improved corporate governance and rationalization. 30. Under this component, the consultants developed a generic code of corporate governance applicable to all SFIs. In addition, they provided model articles of association and implementing guidelines, so that the governance code and articles of association could be adapted to the individual institutions and then approved by the Government. Besides the recommendations for improving SFIs’ corporate governance, enhanced reporting systems – including a spread sheet broad monitoring system – were instituted to provide the necessary regulatory oversight. Subsequently, the corporate governance code and articles of association

7

were tailored for two SFIs—SME Development Bank and for BAAC—for implementation on a pilot basis. 31. Perhaps the single most important component of these efforts to improve the governance of SFIs were the PSAs, which were developed to reflect the public service obligations of these SFIs. Compensating these SFIs for those public service obligations activities would allow for the rest of these SFI activities to be supervised and assessed as market-based banking operations in accordance with the general banking guidelines of the Bank of Thailand. While the SFIs (notably BAAC and SME Development Bank under this Project) developed these PSAs internally, the Ministry of Finance still has to formally endorse this approach and incorporate PSAs uniformly in the budgetary process. However, PSA approaches have been used to recognize the required compensation to the SFIs for the debt forgiveness programs approved by the Government. Earlier, Ministry of Finance adoption was indicated. C. TA Loan and Grant Costs 32. The cost of the TA loan at appraisal was $6.44 million equivalent, comprising foreign exchange costs of $3.68 million and local currency costs of $2.76 million equivalent (Appendix 1). While consulting costs were roughly in line with the appraisal estimates, loan cancellations totaled $0.58 million, or 12.8% of the loan of $4.5 million. 33. The cost of the TA grant at appraisal was $4.32 million equivalent, comprising foreign exchange costs of $2.40 million and local currency costs of $1.92 million equivalent. While the actual costs appear to be higher than the original estimates, the TA grant account had an undisbursed balance of $187,468 as of September 30, 2005. D. Disbursements 34. Disbursements under the TA loan were delayed because of contract disputes (para. 36). Disbursements under the TA grant were delayed substantially due to the design reformulation of the corporate governance component (paras. 21–22). E. Implementation Arrangements 35. The FPO of the Ministry of Finance was the Executing Agency for the TA loan and the TA grant. Throughout implementation, FPO consulted and interacted closely with all the agencies concerned. FPO and its staff demonstrated a high level of ownership of the Project. Further, FPO’s staff made major conceptual and practical contributions to the development of key outputs. These included a balance of accounts for the PSAs for SFIs, as well as the design, implementation, and refinement of a monitoring system. The establishment of an SFI policy section within FPO, a reflection of the importance given to the subject, strengthened FPO’s capacity to manage the Project. 36. The loan consultants experienced contractual and payment difficulties, reflecting a separation of responsibilities within the Ministry of Finance. The SFI section supervised the Project, while FPO’s project implementation unit handled contract and payment issues. Oversight for payment issues was particularly sensitive due to the high-level policy stressing the importance of reducing foreign exchange expenditures, especially those to foreign consultants. Under these circumstances the implementation of the loan Project and the satisfactory

8

resolution of the payment issues reflected the Government’s and FPO’s strong commitment and ownership. 37. The implementation of the corporate governance component under the TA grant was redesigned (paras. 21–22). F. Conditions and Covenants 38. All TA loan covenants have been complied with. (Appendix 2) G. Consultant Recruitment and Procurement 39. Consultant recruitment and procurement were in accordance with the expectations at appraisal. The Ministry of Finance and ADB did not differ on consultant recruitment and procurement. However, the decision to delegate the selection and recruitment of loan consultants to ADB created unnecessary principal-agent problems. This kind of arrangement should not be replicated in other medium-income developing member countries. H. Performance of Consultants 40. The consultants generally performed at a high level. The initial contract for the corporate governance component, which did not incorporate effectively the changing needs of the FPO that emerged during the study, was the only exception. The redesign effectively resolved this issue (paras. 21–22). I. Performance of the Asian Development Bank 41. ADB performed to a high standard, which was in line with expectations. The only exception was the consulting arrangements for the corporate governance component, which initially were designed inappropriately. The complex corporate governance issues required seeking feedback from relevant Government officials with potential redefinitions of approaches and objectives. Given these complexities, reform activity could not be supported effectively or managed under a fully specified, a priori contract with a consulting firm. However, once that approach failed to meet the needs of the FPO, the redesigned approach utilizing individual consultants and ADB staff for coordination was very effective. This approach provided the needed continuity in the policy dialogue, while incorporating the feedback from the Ministry of Finance officials. The performance of ADB was, on balance, satisfactory.

III. EVALUATION OF PERFORMANCE A. Relevance 42. The TA loan and grant were designed to contribute to Thailand’s recovery from the Asian crisis by making the financial sector more resilient through governance reforms, and by accelerating the access of SMEs to bank credit. The Government increasingly saw these SFIs as effective institutions for fighting rural poverty and promoting the development of SMEs. As such, the outputs provided under the TA loan and grant were useful in helping the Government ensure the effectiveness of these SFIs in undertaking their public service obligations.

9

B. Efficacy in Achievement of Purpose 43. The objectives of the TA loan and grant were achieved effectively, given the redesign of the implementation of the governance component under the TA grant. While the original purposes of the TA loan and grant—eliminating the perceived governance risks associated with publicly owned SFIs—were not achieved fully, those risks were managed effectively through governance reforms. The resulting development benefits have contributed substantially to poverty reduction and SME development. The TA loan component provided the information on the financial and the operational viability of the SFIs that allowed the appropriate corrective actions to be implemented. The TA grant component provided enhanced strategies for the SFIs to address SME needs. Perhaps most importantly, the grant designed governance enhancements in cooperation with FPO for the SFIs themselves. The establishment of an SFI policy section within FPO also has strengthened supervisory capacity and created continuity in policy formulation and governance reform implementation. The project outputs contributed to the operational policies and structure of SME Development Bank, to implementation of new governance mechanisms in selected SFIs, and to better monitoring of SFI performance by FPO. C. Efficiency in Achievement of Outputs and Purposes

44. The process was highly efficient, except for the corporate governance component under the TA grant. The other components were delivered within strict budgets and to a high technical standard. In fact, once the corporate governance component was redesigned to meet the needs of the Government by contracting leading edge governance specialists, it achieved the highest levels of technical standards. ADB staff directly supported these individual consultants, taking responsibility for the policy dialogue to ensure that the needs of the Government were reflected fully in the governance reform proposals. While this process involved substantially higher direct ADB inputs, this revised implementation process was highly efficient in achieving the indicated objectives. 45. The Project’s noteworthy outputs included the development of a system of accounts for separating commercial from directed lending. The PSAs that were developed enable public enterprises—financial and nonfinancial—to be held accountable or compensated transparently for the tasks and objectives for which they assume responsibility. A reporting system was added, and performance indicators for the SFIs were developed. The design of a monitoring system for SFIs was an innovative feature of the Project, and linking this to the PSA was highly noteworthy. Most importantly, FPO staff played a key role in conceptualizing and implementing the PSA and monitoring systems. In these areas, FPO now has the capacity to develop further refinements of its own choosing. 46. The Project also developed a detailed set of corporate governance codes and implementation guidelines initially for BAAC, though they were extended to some of the other SFIs. These outputs were of very high quality. These will be implemented in phases, contingent on further refinement of the overall framework of reform of public sector enterprises. D. Preliminary Assessment of Sustainability 47. The governance reform program appears to be sustainable. The temporary unit within the FPO of the Ministry of Finance to manage the SFIs has been upgraded to a formal unit—an indication of the increased importance now given to the governance of the SFIs. In addition, a number of measures in each of the SFIs have been implemented to improve their financial

10

resilience and governance on a sustainable basis. Their articles of association, which will institutionalize their governance reforms, are being finalized for adoption. The sustainability of the TA loan and grant is rated satisfactory. E. Environmental, Sociocultural, and Other Impacts 48. Environmental impacts were not anticipated, and none has arisen. The social impacts were to be achieved by improving SME access to bank credit, thereby promoting the economic recovery. However, the impacts have gone well beyond this limited objective. SIFC has been converted into an effective SME bank that should have longer-term socioeconomic benefits. Given its unrivaled outreach, BAAC has been a key institution in the rural poverty reduction programs of the Government. The Government is proud of its progress in reducing poverty in Thailand. Its Millennium Development Goals have been fully met well in advance of the 2015 target date. The impact of the TA loan and grant is rated satisfactory.

IV. OVERALL ASSESSMENT AND RECOMMENDATIONS

A. Overall Assessment 49. The TA loan and grant were implemented largely as conceived. Implementation was slowest for the corporate governance codes developed for the SFIs. However, this has been a difficult area internationally. Thailand has been a leader in formulating appropriate governance reforms, including the PSAs, which make the public service obligations income- and balance sheet-neutral. This paves the way for the Bank of Thailand to exercise regulatory oversight on an equal footing with commercial banks. A more resilient financial sector also has been established, as the governance reforms have made the SFIs more effective, though they are still publicly owned. The corporate governance component under the TA grant encountered implementation difficulties. As noted above, a contract with a consulting firm was not suitable for this component due to the complexities of the subject matter, the potentially variable outputs, and the changing political environment. This component was redesigned to utilize individual consultants with leading edge knowledge, supported by ADB staff playing a leadership role in soliciting feedback from Ministry of Finance officials and incorporating those changing needs. Once redesigned, this component was implemented successfully. Thus, the TA loan and grant are rated successful. B. Lessons Learned

1. The National Policy Making Context 50. The intensification of reforms since the 1997 crisis can be seen today in the context of a process of modernization and reform that is longstanding and internally driven. While the ramifications of the crisis remain, the crisis itself did not disturb the long-term trend of policy making for social and economic development in Thailand. ADB’s sustained experience in Thailand in the aftermath of the crisis highlights the importance of understanding the national contexts for policy making. This context throws up a set of issues that can be anticipated, though rarely, if ever, predicted. 51. The recent economics literature has stressed that governments are not monolithic entities. The complex interaction between national stakeholders frequently involves conflicts of interest that introduce uncertainty into the final outcomes of reform processes. The different interests of the stakeholders give rise to common agency problems. Thus, coordination issues

11

become important and cannot be taken for granted.7 The principal implication for project design is the recognition of two factors. First, the domestic political process will determine outcomes. Second, project design will need to be adjusted during implementation to reflect this process.8 52. In fact, the project design did not explicitly recognize this issue, implicitly relying on FPO to resolve any conflicts of interests seamlessly for the common good. In practice, the willingness to implement the recommendations and the outputs of the Project was hampered at times by these considerations. Generally, however, the differences were resolved by the steering committee, or through negotiations, persuasion, and sometimes FPO fiat. Overall, the presence of a strong and skilled bureaucracy, and FPO’s leadership, helped to promote stability in outcomes and continuity of policy support even through a change in Government in early 2001. 53. Achieving results took longer than anticipated in the Report and Recommendation of the President, reflecting the need to adjust to new circumstances and resolve issues through the policy making process. Correspondingly, greater effort was required to clarify and reprioritize objectives within the scope of the project design, and achieve consensus on the weight of effort and allocation of resources for specific project components. Thus, the development of the corporate governance codes, the PSA, and monitoring systems required nearly 3 years of phased work. The objectives and TORs for each phase had to be refined in the light of the achievements and difficulties encountered in the previous phase. This process helped strengthen the Project’s continued relevance, and ensured Government ownership. It also imposed high administration and staff costs on ADB. 54. Some issues required greater effort than others to reach consensus and resolution. This was most relevant to the corporate governance codes and articles of association for the SFIs. The State Enterprises Policy Office, a separate department within the Ministry of Finance, was responsible for this set of issues, not the FPO. As a result, the FPO could not guarantee that its support for this component of the Project would produce the same degree of implementation and reform as in other project components. Partly for this reason, and partly because high level consensus needed to be reached on this issue, the codes of best practice have yet to be implemented in full. At the same time, greater coordination is occurring between the relevant agencies on this and related issues. 55. The complexity of national contexts for economic policy making is highlighted by another example from the Project. The National Assembly is reviewing legislation incorporating the project recommendation to convert BAAC into a rural bank with strengthened governance structures involving a board chairman other than the minister of finance. Some Assembly members have advocated the retention of the minister as chairman of the board. Thus, the principles for insulating management from political influences have been articulated and clarified, and generally are known to all. Still, this issue can be resolved only through the political process itself, not through fiat.

7 Multiple agent and multiple principal interactions, such as are common in economic policy making, are especially

prevalent in the public sector. See, for example, Dixit, Avinash. 2002. Incentives and Organizations in the Public Sector: An Interpretive Review. Journal of Human Resources vol. XXXVII(4): pages 696–727.

8 Policy choices are endogenous as the authorities have “multiple objectives and are subject to competing influences.” See Ivanova, Anna et al. 2001. What Determines the Failure of Fund-Supported Programs? IMF. Working Paper. Policy Development and Review Department (page 6). The relevance of political economy issues to project design is also discussed in ADB. 2005. Policy Reform in Thailand and the Asian Development Bank’s Agriculture Sector Program Loan; and Abonyi, George. 2005. ERD Working Paper Series No. 71. Mandaluyong City.

12

56. The issue of country ownership is thus central. While ADB and other multilateral development banks are not passive agents of finance, the “legitimate political institutions of the country should determine the nation’s economic structure and the nature of its institutions.”9 The ownership of this Project resided with the Government of Thailand. ADB discharged its legitimate role by providing support through technical judgments on economic policies and institutional reforms, and by monitoring policy impacts. The technical judgments provided by the ADB enabled the Government to make a more informed choice from the policy options available. In this way, ADB supported—but did not supplant—the higher obligation of the Government to determine, in conjunction with its relevant stakeholders, the right choices for Thailand in the light of its own national political process.10

2. The Counterfactual: Ownership and Additionality

57. The project cost in financial terms and the high level of ADB staff input for administration raises a fundamental counterfactual issue: what would have been different if the Project had not been undertaken? The issues were sufficiently relevant and important that the Government would have undertaken the reforms envisaged under the Project. Although more time and effort would have been needed, it is highly probable the Government would have achieved its objectives. This underscores the salience of Government commitment and ownership. Further, where the Government takes ownership, ADB’s role—no matter how highly valued—will involve less “additionality.” A trade-off is inevitable.

3. Other Lessons

58. Finally, complex issues such as governance reforms cannot be promoted effectively with consulting firm support, as the outputs cannot be specified fully in advance. ADB staff involvement is needed in conjunction with individual consultants to adjust the programs during implementation to meet the needs of the Government effectively. C. Recommendations 59. In middle-income countries such as Thailand, a partnership framework is a more suitable approach to providing development assistance. Clearly, alternative modalities are available to achieve the same development objectives. Since the most important determinant of success is Government ownership, development assistance must be designed to assist the Government through its chosen modalities. 60. ADB should remain vigilant about Thailand’s changing economic circumstances, and adapt its modalities to the country’s needs. At this juncture, a development partnership with Thailand is clearly required. This implies that ADB staff will need to be involved fully in the innovative design and delivery of development banking and capital market services. The client’s needs must be viewed with the utmost importance, and programs and projects must be adjusted to meet the changing needs of the Government.

9 Feldstein, Martin. 1998. Refocusing the IMF. Foreign Affairs. March–April. New York City. 10 Government ownership of this Project and of the broader reform agenda meant that some issues within the overall

scope of the Project did not need to be addressed. One example was the establishment of a centralized asset management agency to handle the NPLs of public financial institutions. The Government set up and operated without external assistance the Thai Asset Management Corporation for this purpose.

Appendix 1

13

ESTIMATED AND ACTUAL PROJECT COSTS

($ million) Appraisal Estimate Actual Component FX LC Total FX LCa Total A. Loan: SFIs Restructuring 1. ADB Financing a. Consultants 3.21 0.71 3.92 3.92 0.00 3.92 b. Manuals and seminars 0.05 0.01 0.06 0.00 0.00 0.00 c. Contingency 0.42 0.10 0.52 0.00 0.00 0.00 Subtotal (1) 3.68 0.82 4.50 3.92 0.00 3.92 2. Government Financing

a. Office, secretarial, and related services 0.00 0.20 0.20 0.00 0.30 0.30

b. Training facilities 0.00 0.03 0.03 0.00 0.12 0.12 c. Local counterpart staff 0.00 0.31 0.31 0.00 0.45 0.45 d. Preparatory audit and restructuring- related expenses 0.00 1.10 1.10 0.00 1.50 1.50 e. Contingency 0.00 0.30 0.30 0.00 0.00 0.00 Subtotal (2) 0.00 1.94 1.94 0.00 2.37 2.37 Total (A) 3.68 2.76 6.44 3.92 2.37 6.29 B. Grant: Corporate Governance for SFIs and Strategies for SME Financingb 1. ADB Financing a. Consultants 1.74 0.44 2.18 2.46 0.38 2.84 b. Seminars, training, and workshops 0.28 0.03 0.31 0.06 0.02 0.08 c. Contract negotiation relating to EA 0.01 0.00 0.01 0.00 0.00 0.00 d. Equipment 0.00 0.00 0.00 0.04 0.00 0.04 e. Contingency 0.37 0.13 0.50 0.00 0.00 0.00 Subtotal (1) 2.40 0.60 3.00 2.56 0.40 2.96 2. Government Financing

a. Office, secretarial, and related services 0.00 0.15 0.15 0.00 0.20 0.20

b. Training Facilities 0.00 0.02 0.02 0.00 0.25 0.25 c. Local counterpart staff 0.00 0.96 0.96 0.00 1.30 1.30 d. Contingency 0.00 0.19 0.19 0.00 0.00 0.00 Subtotal (2) 0.00 1.32 1.32 0.00 1.75 1.75 Total (B) 2.40 1.92 4.32 2.56 2.14 4.71 Grand Total (A+B) 6.08 4.68 10.76 6.49 4.51 11.00 of which ADB Financing 6.08 1.42 7.50 6.49 0.39 6.88 Government Financing 0.00 3.26 3.26 0.00 4.12 4.12 FX = foreign exchange, LC = local currency, SFI = specialized financial institution, SME = small- and medium-sized enterprise. a Estimated. b TA grant portion is still active; actual amounts represent funds committed. Sources: Loan Financial Information System, Technical Assistance Information System, and Project Financial Reports.

Appendix 2

14

STATUS OF COMPLIANCE WITH LOAN COVENANTS

Covenant

Reference in Loan

Agreement

Status of Compliance 1. The Borrower shall make available to Ministry of Finance, promptly as needed, the funds, facilities, services and other resources which are required, in addition to the proceeds of the loan, for the carrying out of the technical assistance and for the operation and maintenance of the technical assistance facilities. 2. The Borrower shall, or shall cause MOF to, ensure that competent and qualified consultants, acceptable to the Borrower and the ADB, are employed to an extent and upon terms and conditions satisfactory to the Borrower and the ADB. 3. The Borrower shall ensure that the activities of MOF and its other departments and agencies with respect to the carrying out of the technical assistance and operation of the technical assistance facilities are conducted and coordinated in accordance with sound administrative policies and procedures. 4. The Borrower shall (i) maintain, and cause MOF to maintain, separate accounts for the technical assistance; (ii) have such accounts and related financial statements audited annually, in accordance with appropriate auditing standards consistently applied, by independent auditors whose qualifications, experience and terms of reference are acceptable to the ADB; (iii) furnish to the ADB, as soon as available but in any event not later than six (6) months after the end of each related fiscal year, certified copies of such audited accounts and financial statements and the report of the auditors relating thereto (including the auditors’ opinion on the use of the loan proceeds and compliance with the covenants of this Loan Agreement, as well as on the use of the procedures for imprest account and statement of expenditures), all in the English language; and (iv) furnish to the ADB such other information concerning such accounts and financial statements and the audit thereof as the ADB shall from time to time reasonably request.

5.Other Covenants

Section 4.02 Section 4.03 (a) Section 4.04 Section 4.06 (b) Article IV and Schedule 6 thereto.

Complied with Complied with Complied with Complied with Complied with