Document of The World Bank Report No: ICR00003533 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IBRD-79020 IBRD-79030 IBRD-79040) ON LOANS IN THE AMOUNT OF USD100 MILLION TO TÜRKİYE KALKINMA BANKASI A.Ş., IN THE AMOUNT OF USD200 MILLION TO T.C. ZIRAAT BANKASI A.Ş., AND IN THE AMOUNT OF USD200 MILLION TO T. VAKIFLAR BANKASI T.A.O. WITH THE GUARANTEE OF THE REPUBLIC OF TURKEY FOR THE SECOND ACCESS TO FINANCE FOR SMALL AND MEDIUM ENTERPRISES PROJECT August 5, 2015 Finance and Markets Turkey Country Unit Europe and Central Asia Region CURRENCY EQUIVALENTS (Exchange Rate Effective March 31, 2015) Currency Unit= New Turkish Lira (TRY) TRY 1.00 = US$ 0.382 US$ 1.00 = TRY 2.616 FISCAL YEAR ABBREVIATIONS AND ACRONYMS AF Additional Financing BRSA Bankacılık Düzenleme ve Denetleme Kurumu, Banking Regulation and Supervision Agency CBRT Central Bank Republic of Turkey CPS Country Partnership Strategy D/E Debt-to-Equity DSCR Debt service coverage EFIL Export Finance Intermediation Loan EIB European Investment Bank Eximbank Turkiye Ihracat Kredi Bankasi A.S. GDP Gross Domestic Product IBRD International Bank For Reconstruction And Development ICR Implementation Completion Report IFI International Financial Institution IEG Independent Evaluation Group IL Investment Loan ISR Implementation Status And Results IT Information Technology M&E Monitoring And Evaluation PAD Project Appraisal Document PDO Project Development Objective PFI Participating Financial Institution PIU Project Implementation Unit SME Small And Medium Enterprise TL Turkish lira TSKB Turkiye Sinai Kalkinma Bankasi TUIK Türkiye İstatistik Kurumu, Turkish Statistical Institute WCL Working capital loan Senior Global Practice Director: Gloria M. Grandolini Practice Manager: Aurora Ferrari Project Team Leader: Alper Ahmet Oguz ICR Team Leader: Ilias Skamnelos TURKEY SECOND ACCESS TO FINANCE FOR SMALL AND MEDIUM ENTERPRISES PROJECT CONTENTS Data Sheet A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs H. Restructuring I. Disbursement Graph 1. Project Context, Development Objectives and Design ............................................... 1 2. Key Factors Affecting Implementation and Outcomes .............................................. 4 3. Assessment of Outcomes .......................................................................................... 13 4. Assessment of Risk to Development Outcome......................................................... 20 5. Assessment of Bank and Borrower Performance ..................................................... 21 6. Lessons Learned ....................................................................................................... 23 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners .......... 25 Annex 1. Project Costs and Financing .......................................................................... 26 Annex 2. Outputs by Component ................................................................................. 27 Annex 3. Economic and Financial Analysis ................................................................. 35 Annex 4. Bank Lending and Implementation Support/Supervision Processes ............ 37 Annex 5. Beneficiary Survey Results ........................................................................... 38 Annex 6. Stakeholder Workshop Report and Results................................................... 39 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR ..................... 40 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders ....................... 46 Annex 9. List of Supporting Documents ...................................................................... 47 MAP .............................................................................................................................. 48 A. Basic Information Second Turkey Access to Country: Turkey Project Name: Finance for Small and Medium Enterprises Project IBRD-79020,IBRD- Project ID: P118308 L/C/TF Number(s): 79030,IBRD-79040 ICR Date: 07/09/2015 ICR Type: Core ICR Lending Instrument: FIL Borrower: REPUBLIC OF TURKEY Original Total Commitment: USD 500.00M Disbursed Amount: USD 500.00M Revised Amount: USD 500.00M Environmental Category: F Implementing Agencies: Turkiye Kalkinma Bankasi (TKB) Vakiflar Bankasi (Vakif Bank) T.C. Ziraat Bankasi A.S. (Ziraat Bank) Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 11/24/2009 Effectiveness: 08/12/2010 Appraisal: 04/05/2010 Restructuring(s): 09/16/2014 Approval: 06/15/2010 Mid-term Review: Closing: 09/30/2014 03/31/2015 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Satisfactory Risk to Development Outcome: Moderate Bank Performance: Satisfactory Borrower Performance: Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Satisfactory Government: Satisfactory Implementing Quality of Supervision: Satisfactory Satisfactory Agency/Agencies: Overall Bank Overall Borrower Satisfactory Satisfactory Performance: Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Indicators Rating Performance (if any) Potential Problem Quality at Entry Project at any time No None (QEA): (Yes/No): Problem Project at any Quality of No None time (Yes/No): Supervision (QSA): DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Banking 5 5 General finance sector 15 15 General industry and trade sector 10 10 SME Finance 70 70 Theme Code (as % of total Bank financing) Micro, Small and Medium Enterprise support 67 67 Other Private Sector Development 33 33 E. Bank Staff Positions At ICR At Approval Vice President: Cyril E Muller Philippe H. Le Houerou Country Director: Martin Raiser Ulrich Zachau Practice Manager/Manager: Aurora Ferrari Lalit Raina Project Team Leader: Alper Ahmet Oguz Carlos Pinerua ICR Team Leader: Ilias Skamnelos ICR Primary Author: Aminata Ndiaye F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) The project's main development objective is to broaden and deepen the access of Turkish small and medium enterprises to medium- and long-term finance, with a view to ultimately contributing to an expansion of productive activities and job creation. Revised Project Development Objectives (as approved by original approving authority) N/A (a) PDO Indicator(s) Original Target Formally Actual Value Values (from Revised Achieved at Indicator Baseline Value approval Target Completion or documents) Values Target Years Indicator 1 : Number of SME beneficiaries Value (Quantitative or n/a 200 746 Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 Comments (incl. % Fully achieved achievement) Indicator 2 : Increase in employment in beneficiary SMEs Value (Quantitative or n/a >0 6,918 Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 Comments (incl. % Fully achieved achievement) Indicator 3 : Increase in sales in beneficiary SMEs Value (Quantitative or n/a >0 TL2.1 billion Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 Fully achieved. TL2.1 billion is equivalent to USD900 million using end- Comments (incl. % 2014 exchange rate. Due to exchange rate differentials over the life of the achievement) project, project performance cannot be reported in U.S. dollars. Indicator 4 : Share of finance provided to underserved regions Value (Quantitative or n/a 25% 17% 20% Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 03/31/2015 Comments (incl. % Fully achieved achievement) Indicator 5 : Quality of project financed portfolio ‒ NPL ratio Value (Quantitative or n/a 7% 0.7% Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 Comments (incl. % Fully achieved achievement) Indicator 6 : Number of participating financial institutions Value (Quantitative or n/a 5 6 Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 Comments (incl. % Fully achieved achievement) (b) Intermediate Outcome Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised approval Completion or Target Values documents) Target Years Indicator 1 : Volume of Bank support Value (Quantitative n/a 100% of loan 104% or Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 Fully achieved. Due to exchange rate differentials over the life of the Comments (incl. % project, a total of USD518.5 million was disbursed by the Borrowers to achievement) SME beneficiaries Indicator 2 : Borrower- Compliance with prudential norms Value (Quantitative compliant compliant compliant or Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 Comments (incl. % Fully achieved achievement) Indicator 3 : PFI - Compliance with prudential norms Value (Quantitative compliant compliant compliant or Qualitative) Date achieved 06/15/2010 09/30/2014 03/31/2015 Comments (incl. % Fully achieved achievement) G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 01/04/2011 Satisfactory Satisfactory 79.50 2 10/05/2011 Satisfactory Satisfactory 270.94 3 05/01/2012 Satisfactory Satisfactory 356.46 4 01/02/2013 Satisfactory Satisfactory 413.75 5 10/28/2013 Satisfactory Satisfactory 460.03 6 07/05/2014 Satisfactory Moderately Satisfactory 460.03 7 12/30/2014 Satisfactory Moderately Satisfactory 472.52 8 03/31/2015 Satisfactory Moderately Satisfactory 498.75 H. Restructuring (if any) Board ISR Ratings at Amount Restructuring Approved Restructuring Disbursed at Reason for Restructuring & Date(s) PDO Restructuring Key Changes Made DO IP Change in USD millions (i) extend the project closing by six months to end-March, 2015; (ii) revise the disbursement requirement for priority regions from 25 to 15 percent of total disbursements; (iii) cancel legal covenants regarding financial requirements for priority regions; 09/16/2014 N S MS 472.52 (iv) revise the target value of the PDO level indicator “Share of finance provided to priority regions” in accordance with point (ii) above, and; (v) modify the disbursement estimates due to the extension of the closing date in accordance to point (i) above. I. Disbursement Profile The Second Access to Finance for Small and Medium Enterprises (SME II) Project fully disbursed a total of USD500 million after a six-month project extension. The original project was approved by the World Bank in June 2010, and comprised of three loans: (i) USD100 million to Turkiye Kalkinma Bankasi (Kalkinma Bank); (ii) USD200 million to Ziraat Bankasi (Ziraat Bank), and; (iii) USD200 million to Vakiflar Bankasi (Vakif Bank). Kalkinma Bank, acting as on- lender (also known as wholesale or tier two lender), fully disbursed in September 2012. Ziraat Bank and Vakif Bank, both direct lenders (also known as retail or tier one lenders), fully disbursed at project closing in March 2015, following a six-month project extension from December, 2014. In total, the project provided access to medium and long-term finance to 746 small and medium enterprises (SMEs) for working capital and investment purposes, in Euros, US Dollars and Turkish Lira. 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal Turkey sustained strong economic growth between 2002 and 2007, translated into significant poverty reduction. After the 2001 banking crisis that led to sharp recession and restructuring of the financial sector, economic growth averaged 6.8 percent and stayed above 4.5 percent during this time period. Trade volumes increased significantly and continuously, with trade accounting for about 45 percent of GDP in 2008, up from 38 percent in 2002. Poverty levels were substantially reduced, falling from 30 to 8.4 percent, between 2002 and 2007. Investment levels remained modest, however, and growth was increasingly driven by domestic demand. The Turkish economy was significantly affected by the global financial crisis, although it turned to rapid recovery by 2010. GDP growth declined from 4.7 percent in 2007 to 0.7 percent in 2008 and negative 4.8 percent in 2009. Real private investment dramatically dropped from 30 percent per annum until 2004, to just 2.6 in 2007 and negative 22.5 in 2009. Total exports plunged 23 percent between 2008 and 2009 to just over USD100 billion. Unemployment rose from 10 to 14 percent from 2002 to 2009. Macroeconomic volatility impacted the financial sector, which was still recovering from the 2001 crisis, and led to increased difficulties for enterprises in accessing investment finance. The authorities’ response to the crisis was broad-based - it included monetary policy easing, employment measures, and targeted fiscal stimulus. The preparation of the SME II Project was undertaken as Turkey begun its recovery, with annual GDP growth rebounding to 9.2 percent in 2010. Access to finance for Small and Medium Enterprises (SMEs) tightened in the aftermath of the crisis, with evidence pointing to significant inequality in the provision of credit. Between December, 2007, and December, 2009, the SME share in total loans declined by about 7 percentage points to around 23 percent, and the share of SME loans as a percentage of corporate credit dropped from 53 percent to around 45 percent. SMEs were perceived as higher risk, often lacking the necessary collateral to access funding and, in particular, longer-term finance. In 2010, 99.9 percent of Turkish businesses were SMEs, which accounted for 76 percent of employment, 64 percent of total turnover and 50 percent of investments in tangible goods. The World Bank’s Enterprise Survey of 2008 identified access to finance as the single most severe obstacle. The survey also found that the loan application rejection rate for SMEs was significantly higher than that of larger corporates. This inequality was more acute in the traditionally underserved regions - regions classified as priority accounted for just 8 percent of total bank loans as of end-2008. This prompted an SME-friendly policy response, including a request to the World Bank for SME support in the form of a credit line intermediated by state-owned banks. In the wake of 1 the crisis, the Turkish authorities passed legislation earmarking USD1 billion to the Credit Guarantee Fund for SMEs (prior to which, the fund had a modest paid-in capital of only USD55 million). In addition, funding for the Small and Medium Industry Development Organization (KOSGEB) almost doubled to support credit subsidies and technical services. Turkish authorities also asked the World Bank for a project that was expected to increase the ability of banks to provide longer term maturities to SMEs, and promote investments and business growth. Project funds were also earmarked for financing SMEs operating in provinces with low credit activities (in Eastern and Southeastern regions). The government aimed to ensure that SME credit and productivity gaps between advanced and lagging regions would not widen during the post-crisis period. Demand for affordable longer-term funding was strong, given the prevailing high interest rates (policy rates of more than 16 percent) and short loan maturities (with 60 percent of corporate loans having a maturity of three years or less). Moreover, the project would help the financial sector further develop its investment lending business by demonstrating that term lending can be a viable business proposition, while building the necessarily skills at Participating Financial Intermediaries (PFIs) to appraise term loans. The credit line operation was consistent with the FY08–11 Country Partnership Strategy (CPS) and part of a comprehensive package of analytical, advisory, and financial sectoral support. The top priorities of the FY08–11 CPS included improving the business climate, financial markets stability, and filling the gap in accessing credit facilities. Moreover, SMEs were included as a core pillar of the "Increasing Competitiveness" axis of development in the Turkish Government’s Ninth Development Plan (2007–13), with financial system improvements as a key driver. Finally, preliminary results of the 2010 Investment Climate Assessment provided evidence that access to finance was consistently associated with the ability of firms to generate employment and grow in size. The credit line also complemented existing credit facilities provided by the World Bank and other international organizations, including the Access to Finance for SME project (SME I). The SME II project sought to build on the success of the SME I project, which was approved in 2006 and had been scaled up twice with additional financing. A total amount of USD700 million was being intermediated by a state-owned bank, Halkbank, and a private development bank that enjoys the guarantee of the Turkish Treasury, Türkiye Sinai Kalkinma Bankasi (TSKB). The SME I project was being successfully implemented at the time of SME II approval: 450 firms operating in developed and underserved regions obtained investment and working capital loans with an average maturity of over 5 years. At the time of SME II project approval, the World Bank’s Fourth Export Finance Intermediation Loan (EFIL IV) project was also being implemented by Eximbank and TSKB, along with several credit lines provided by other International Financial Institutions (IFIs) and bilateral development agencies - the European Investment Bank (EIB), the European Bank of Reconstruction and Development (EBRD), the German Development Bank (KfW), and the French Development Agency (AFD). 1.2 Original Project Development Objectives (PDO) and Key Indicators (as approved) The project’s main development objective is to broaden and deepen the access of Turkish small and medium enterprises to medium- and long-term finance, with a view to ultimately contributing to an expansion of productive activities and job creation. To measure progress towards the achievement of PDOs, the following performance indicators were approved:  Number of SME beneficiaries financed under the project;  Increase or preservation of employment and sales in the beneficiary SMEs; 2  Share of finance provided to the underserved regions under the project;  Quality of portfolio financed under the project;  Number of financial institutions participating under the project. Other key indicators for measuring project outcomes were:  The amount of credit line disbursed;  The financial performance of Borrowers and PFIs. The PAD anticipated that additional information would be collected on: (i) SMEs (size, sector, geographical location, and exports), and (ii) loans (size, interest rates, and maturity). The Bank would also monitor overall trends in access to finance for SMEs, as well as the financial performance of participating banks and PFIs. 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification N/A 1.4 Main Beneficiaries, The primary beneficiaries for this project, as identified in the PAD, were SMEs. SMEs were defined as firms with fewer than 250 employees and an annual turnover of less than USD20 million. This definition was in line with Turkish legislation and the Ministry of Industry and Trade regulations. Private SMEs (private ownership of more than 50 percent) operating in all regions and sectors were eligible for financing under the project. The project also sought to make financing available to SMEs in underserved regions (as identified in the Official Gazette). There were no sector restrictions or quotas. The secondary project beneficiaries were the six financial institutions - three participating banks and three PFIs. First, the project’s requirements for PFIs to comply with banking and leasing regulations and financial covenants was expected to help PFIs build capacity in terms of risk management, procurement, and financial management and environmental safeguards. PFIs would gain valuable experience working with IFI procedures, and ultimately be able to attract new credit lines financed by other IFIs active in Turkey. Second, all institutions’ access to medium- and long-term funding would help them expand business in medium- and long-term lending and financial leasing activities, while reducing the maturity mismatches on their balance sheet. Third, it was anticipated that the three participating banks, which were selected for their strong management and implementation capacity, would learn from the project’s specific lending requirements, in terms of credit assessment tools. The project specifically required financial institutions to rely on firms’ financial condition for their lending decisions (assessed through cash flows, the debt-to-equity and debt service coverage ratios). Financial institutions did not usually rely on such credit assessment tools, instead focusing on collateral. As such, the project was expected to lead to changes in SME financing, especially in the maturity of financial products offered to SMEs and the collateral requirements. 1.5 Original Components (as approved) The project built on the design of the predecessor SME I project. The project comprised of three components, one for each borrower: Component 1 (USD100 million) consisted of credit-line to Kalkinma Bank, as the borrower and implementing agency. Kalkinma Bank, as on-lender, passed funds to PFIs through Subsidiary 3 Finance Agreements for further on-lending to eligible SMEs. There was no pre-allocation between PFIs. Component 2 (USD200 million) consisted of a credit line to Ziraat Bank, which operated as a direct lender to SMEs. Component 3 (USD200 million) consisted of a credit line to Vakif Bank, which operated as a direct lender to SMEs. Loans to Kalkinma Bank, Ziraat Bank and Vakif Bank were provided by the World Bank, with a Government of Turkey guarantee. 1.6 Revised Components N/A 1.7 Other significant changes With regards to Component 1, Kalkinma Bank had one reallocation of Loan proceeds between PFIs. At the beginning of the project, Kalkinma Bank had selected two banks ‒Garanti Bank and Deniz bank‒ to on-lend project funds to SMEs. Within the first year, Kalkinma Bank reallocated part of its line of credit from Garanti Bank to Garanti Leasing. Table 1: Allocation of Kalkinma Bank credit line Initial Final (in USD millions) allocation allocation Garanti Bank 65 40 Deniz Bank 35 35 Garanti Leasing - 25 Total 100 100 Project requirements were revised in September, 2014, affecting Components 2 and 3. A project restructuring was undertaken to remove financial covenants for SMEs operating in underserved regions (i.e. the requirements related to the debt-to-equity and the debt service coverage ratios). The minimum amount to be disbursed to SMEs in these regions was also reduced from USD100 million to USD60 million. 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry The SME II project built on significant World Bank experience with credit lines in Turkey. Between 1999 and 2010, four export finance intermediation projects (the EFIL series) and the predecessor SME finance project were approved by the World Bank, totaling over USD2 billion. The Government of Turkey had also signed several lines of credit with other IFIs and bilateral development agencies. The accumulated knowledge helped inform the SME II project design in increasing the new line’s developmental impact. The project’s preparation and design were based on the following lessons learned from previous credit line operations: (i) Consistency with Bank and Government priorities. The project’s design was aligned with the FY08-11 CPS and reflected the Turkish government’s strategic priorities. (ii) Flexible design. 4  Currency and interest rate: The World Bank financing included an option for the borrowers to convert the original loans into an “approved currency”, and from variable to fixed interest rate (and vice-versa). Lending in local currency was also allowed.  Sub-lending requirements: First, the project had no restrictions on eligible sectors. Experience from previous credit lines showed that imposing any type of restrictions on credit lines was counter-effective, because lending would not be demand-driven. Second, loans could be denominated in any currency (USD, EUR or TL). There were no rigid pre- allocations per PFIs (for Kalkinma Bank), nor per sector or per type of loans (investment or working capital). The project however did provide for a minimum percentage of loans to finance SMEs in underserved regions. This requirement was driven by the Turkish Treasury, as applicable to all credit line operations at the time. Sub-lending was expected to be demand-driven, as long as demand form underserved regions exceeded the regional targets that have been imposed. This optimistic scenario did not materialize during project implementation, and lead to significant delays. The SME I ICR explicitly pointed to the inefficiency of regional lending quotas, and to Halkbank’s difficulty in achieving this lending target towards project closing. (iii) Inclusion of leasing in credit lines. Leasing companies were included in the SME II project, as an enhancement compared to the SME I project.1 The inclusion of leasing companies as financial intermediaries built on the successful experience under the EFIL series, which had helped: (a) reach smaller firms that are more accessible by leasing companies; and (b) assist the development of the leasing sector, and thereby deepening and diversifying the financial sector. (iv) Quality of participating financial institutions. Kalkinma Bank, Ziraat Bank and Vakif Bank were all selected on the basis of sound bank analysis and audited financial performance data. The World Bank preparation team conducted detailed assessments, after which the three banks were deemed to be well capitalized and profitable, with a sound loan portfolio. All participating banks are government-owned, with strong management capacity. Kalkinma Bank was rated by Fitch ratings and had previous experience working with IFIs, including the World Bank. Ziraat Bank and Vakif Bank were also rated by several agencies at the time of approval, including by Moody’s. They have also been selected given their large network, strong focus on SMEs and geographically diversified lending practices. (v) Decentralized decision-making and market-oriented lending. Two important lessons, incorporated from previous credit lines, were the use of established market practices for the selection of PFIs and SMEs/sub-projects. For example, with Kalkinma Bank selecting and carrying the credit risk of PFIs, and PFIs selecting and carrying the credit risk of SMEs, the choices of the participants in the project was made by the entities best capable of identifying good financial intermediaries and eligible enterprises. Qualitative criteria set out in the project, and monitored during supervision, helped ensure that the choices were made on a sound basis. (vi) Lower transaction costs for PFIs, thanks to business-oriented project requirements. In line with simplification trends initiated in the predecessor EFIL series and the SME I project, the SME II project reduced project requirements for PFIs, closely aligning these with business practices, in order to reduce the transaction costs. Earlier ICRs had highlighted that onerous environmental, procurement, and other project requirements tend to lead to funds being allocated more restrictively, 1 When the SME I project was originally approved, it was expected that leasing companies would participate in the project through TSKB’s credit line. When TSKB requested to operate as a direct lender, instead of on-lender, leasing companies were, by consequence, excluded. 5 towards larger transactions, and to firms with better access to finance. This meant, for example: (a) use of procurement requirements that are suitable for private sector borrowers to make the project more client-friendly and ensure its expeditious implementation; and (b) combining the Borrower and Implementing Agency functions in one and the same entity for higher quality and expeditious project implementation. (vii) Clear identification of risks and associated mitigation measures. The project’s risks included: (i) macroeconomic risk, due to the global economic slowdown; (ii) credit risk of SMEs; and (iii) implementation risks ‒ with this being Ziraat Bank’s and Vakif Bank’s first operations with the World Bank, and Kalkinma Bank’s first operation with the World Bank as an on-lender. Macroeconomic risks were mitigated by economy’s demonstrated resilience to external shocks. At the project level, the Bank team had strong confidence in the three Participating banks’ institutional, operational and outreach capacity. Furthermore, with regards to Kalkinma Bank, road shows were proactively organized ahead of project approval to market the loan, and written expressions of interest from four potential PFIs had been received before project approval. The SME II project differed from the previous SME I operation in three ways:  Financial intermediaries: the SME II project involved three borrowers, all of which were state-owned banks, while a private development bank had participated in the SME I project (TSKB). In addition, the SME II project allowed for the coexistence of the on-lending and direct lending approaches in an identical operating environment (i.e. sub-lending criteria and project requirements). The SME I project involved only two banks, both acting as direct lenders.  Underserved regions: Both the list of targeted regions and the amount earmarked for these regions differed substantially between the SME I and SME II projects. First, the list of underserved regions as defined in the Official Gazette dated October 17, although longer than in SME I, consisted of a greater number of less developed provinces 2 The number of provinces that had a strong entrepreneurial base with sufficient capacity to meet all project requirements was lower. Three large and dynamic provinces (Hatay, Kayseri and Gaziantep) were considered “priority” in SME I, but excluded from the list in SME II. The table in Annex 2 presents a comparison of target regions, and highlights the absorption capacity of these target regions (through two indicators - total exports and per capital gross value added). SMEs in Gaziantep were more likely to request and receive WB financing in the SME I project, than SMEs in Elazig at the time of SME II. Second, the SME I project assigned lending targets to Halkbank only, not TSKB. For every new loan agreement with Halkbank, differentiated lending targets had been set ‒ 25 percent in 2007, 20 percent in 2008, and 30 percent in 2009. As a result, while the combined target for Halkbank was 25 percent over the lifetime of the SME I project, the disaggregated picture shows that lending requirements need to take into account the overall dynamics of the economy. The regional lending quota was indeed lowered in 2008 at the onset of the global financial crisis. In 2009, the 30 percent lending quota for underserved regions, set as the economy was recovering, was still difficult for Halkbank to fill. The SME I project operated under different circumstances when it came to underserved regions. The SME I project (approved in June 2006 and closed in April 2012) was the precursor to several projects in Turkey that explicitly targeted underserved regions (both financed by the World Bank and other development partners). The SME II project preparation discussions begun with how 2 The list of underserved regions is determined by a decree each year attached to the budget law and published in the Official Gazette. 6 much should be earmarked for the underserved regions, rather than if earmarking funds was appropriate. At preparation, the project envisaged a 30 percent underserved regions target and, by the time of approval, the project required that at least 25 percent of each of the three lines of credit be disbursed in priority regions. The priority regions were identified in the Official Gazette dated October 17, 2009, and including provinces located in South Anatolia, East Anatolia, and selected ones in the Mediterranean, Central Anatolia and Black Sea regions of Turkey. The 25 percent requirement was set by the Turkish Treasury at the time of project approval for all IFI loans that were backed by a Turkish government guarantee. This ambitious target was set without an assessment of demand and capacity in these regions, leading to saturation from diverse IFI loans under blanket regional requirements.  Financial covenants: In SME II, the minimum required DSCR was 1.2, higher than the 1.1 required in SME I in 2006 and the EFIL IV project approved in 2008 (see next section for implications). 2.2 Implementation The project became effective in August 2010, just two months after Board approval. Kalkinma Bank’s credit line disbursed in 2012, two years ahead of the original project closing date. The remaining two banks were able to fully disburse six months after the closing date, after project restructuring. A midterm review of the project was not undertaken. Factors that affected implementation outcomes can be organized into six categories. A. Relevant country and sector context The Turkish economy recovered rapidly from the global financial crisis, although some sectors were more profoundly affected than others. SMEs, which played and continue playing an important role in the economy, were the first casualties of the financing pressures that had surfaced during project implementation. Overall, economic developments include: The Central Bank’s monetary policy: In the aftermath of the global financial crisis, the expansionary monetary policies adopted in the U.S. and the EU translated into the availability of abundant and low-cost short-term foreign financing, which led to a rapid credit expansion and appreciation pressure on the Turkish lira. In an effort to contain the accumulated macro-financial risks and rising external imbalances, the Turkish Central Bank (CBRT) launched a new policy in November 2010, which included the use of reserve requirements and an interest rate corridor to regulate lending in the economy. Over the life of the SME II project, the policy rates and related interest rate corridor fluctuated intensively. Figure 1: Central Bank policy rates and interest rate corridor Source: CBRT 7 Interest rate and exchange rate pressure: the Turkish Lira has continuously depreciated against the US dollar since November 2010. The excessive volatility affected the willingness of SMEs to borrow in US dollar, even to take advantage of the slightly lower rates. This ultimately reduced the attractiveness of the credit line. Participating banks and PFIs decided to lend in Turkish Liras and even in Euros, although the loan received was in US dollars. For example, the share of local currency lending for Kalkinma Bank and Vakif Bank was 70 and 53 percent respectively. During the ICR mission, most SMEs (even those sourcing materials from or selling abroad) expressed some hesitation in borrowing in foreign-currency, especially in US dollars due to the high volatility in the foreign-exchange market. On the other hand, the spike of the Central Bank’s monetary policy rates translated into more expensive commercial loans, with the SME segment being no exception. The project thus helped finance SMEs at a competitive rate. Figure 2a: Exchange rate USD/TRY during project implementation Figure 2b. Weighted average interest rates for commercial bank loans TRY | USD * please note the different scales Source: CBRT (Jan 2010 to May 2015) B. Financial covenants Similarly to previous credit line operations implemented in Turkey since 1999, the SME II project filtered the group of eligible private SMEs based on their financial condition. In an effort to stimulate cash-flow based financing, as opposed to the traditional collateral-based financing, the project required banks to assess the eligibility of SMEs based on the debt-to-equity ratio (maximum D/E ratio of 80:20) and debt-service coverage ratio (minimum DSCR of 1.2) based on financial statements provided. The minimum requirements set for the SME II project were however more restrictive than those of previous credit lines. This was primarily the result of the World Bank team’s desire to be more prudent in incorporating new borrowers into the World Bank’s lines of credit. Nevertheless, SMEs were required to be less leveraged and more liquid than EFIL IV exporters and SME I beneficiaries, despite the context of the global economic slowdown. 8 Table 2: comparison of financial covenants applicable to lines of credit in Turkey Amount D/E ratio DSCR ratio Project Name Year Closing (USD M) (minimum) (minimum) SME I 6/2006 650.0 4/2012 80:20 1:1 EFIL IV 5/2008 900.0 12/2014 85:15 1.1:1 SME II 6/2010 500.0 3/2015 80:20 1.2:1 SME III 6/2013 300.0 12/2017 80:20 1.2:1 Innovative Finance 7/2014 250 12/2018 85:15 1.1:1 Overall, there were three main implications for the SME II project: (i) Financial covenants did not account for challenges specific to the underdeveloped regions. The likelihood that financial statements accurately reflect business activities is higher in developed regions. Firms in underserved regions are often not able to present comprehensive financial statements, given higher levels of informality. The underreporting of commercial activity often includes sales, profit and employment. (ii) Financial covenants may be a crude tool compared to existing financial intermediary practices. Turkish banks have, over time, developed credit assessment methodologies to reflect the country specific context and the SME segment in particular, placing only partial importance on financial covenants such as the D/E and DSCR in the case of SMEs. Under the banks’ own credit policies, branches collect standard information on firm ownership, financial position and activities to assess creditworthiness frequently based on credit scoring methodologies. With the SME II project, participating financial institutions reported that the financial covenants would often be stringent and exclude loans that their methodology judged as viable. Discussions with other development partners highlight the uniqueness of World Bank lines of credit in that respect. Other development partners undertake due diligence of public and private financial institutions’ internal credit procedures at project preparation stage and avoid imposing specific financial constraints. (iii) Financial covenants may also be a crude tool for the specific assessment of high growth firms. A company with a high D/E ratio may be pursuing an aggressive growth strategy to propel itself among the top firms within the sector, region, or the country. The World Bank financial covenants would, however, exclude the inclusion of such high growth companies. To address such issues, the project could have been more flexible in deploying these covenants. Possible courses of actions could have included: (i) dropping the financial covenants for all firms in underserved regions, and possibly relaxing them for firms in developed regions, depending on firm size or sub-loan size; (ii) allowing financial intermediary exceptions if a bank’s credit appraisal policies were assessed and found appropriate and adequate (similar to other development partners); (iii) allowing sub-loan exceptions when justified e.g. when banks provided additional information influencing the interpretation of the financial ratios, such as the a firm belonging to a wider group in solid standing; or (iv) relaxing financial covenants for firms that apply for investment or working capital loans within 12 months of obtaining an investment loan (if well justified). C. Lending in underserved regions The regional lending requirement applied to all participating banks, independently of the absorption capacity of the regions, and stalled implementation. For the SME I project, only 9 Halkbank was required to meet specific lending targets for priority regions3 . The rationale for including all banks was that they were all state-owned banks. It was expected that Kalkinma Bank, the public development bank, would partner with PFIs that have strong outreach capacity. With regards to Ziraat Bank and Vakif Bank, their branch network and lending patterns under their own resources was satisfying in terms of their capacity to reach the underserved regions. Ultimately, the five institutions (Ziraat Bank, Vakif Bank, Garanti Bank, Garanti Leasing and Deniz Bank) intermediating funds to SMEs had to compete to provide funds to often the same SMEs in these areas. This was mostly due to the narrow pool of eligible SMEs, as defined by the project requirements. In other words, the number of SMEs operating in these regions that were both willing and able to receive financing through the project was low, compared with the total number of SMEs operating in the regions. The criteria that restricted the SME pool were: (a) ability to provide financial statements, including cash-flow statements that reflect the company’s activities; (b) compliance with D/E ratio requirements; (c) compliance with DSCR requirements; (d) ability to provide collateral acceptable in quality and quantity by the banks; (e) ability and willingness to provide additional documentation required for compliance with safeguards and fiduciary requirements that are not required by other IFI loans; (f) willingness to receive funding in foreign currency. As a result, the SME II project started to underperform in terms of disbursements in 2013, as Ziraat Bank and Vakif Bank (cumulatively responsible for four-fifths of the project amount) experienced difficulties in meeting the lending target for underserved regions. For example, in October 2013, Vakif Bank had fully disbursed the amount for SMEs in developed regions but only disbursed 10 percent of the loan amount in underserved regions. By the time of restructuring in September 2014, it had only reached 15 percent in underserved regions despite significant efforts. The lending target set by the government at the time of project approval was revised downwards in subsequent projects approved during SME II’s project implementation. Based on the feedback received from the borrowing state banks, the Turkish Treasury reduced its regional disbursement requirement from 25 to 15 percent of project funds. The lower regional targets were already in effect at the time of the SME III project’s approval in June 2013. E. Collateral requirements Collateral requirements were a major impediment to project implementation in underserved regions. The high collateral requirements from banks were formally identified as a constraint to implementation in implementation mission Aide Memoires. This was confirmed during the project completion mission. SMEs and banks reported that collateral-to-loan value requirements are much higher in the underserved regions, while land valuation is much lower. This is due to higher perceived credit risk and the more difficult execution of the collateral. Under a EIF credit guarantee program (that underwrote loans of a substantially lower amount - up to TL 50,000), banks did not require collateral and implementation was rapid. F. Ziraat Bank’s reorganization Ziraat Bank’s strong disbursement performance during the first year of project implementation was reversed due to the bank’s comprehensive overhaul. The restructuring process was launched in end-2011 and covered several areas, including the bank’s central and regional organization structure, customer and branch segmentation, and portfolio management. The objective of the restructuring was to increase efficiency, achieve sustainable profitability and better 3 Halkbank had an overall target of 25 percent for priority regions under the SME 1 project, including additional financing (25% in 2007; 20% in 2008 and 30% in 2009). TSKB, as a private development bank, was not expected to lend to firms in underserved regions. 10 align products and services with client demand. The Bank’s new strategy had a clear focus on the SME segment, with the establishment of 150 new branches solely dedicated to corporate, commercial and the SME/Entrepreneur segments. The restructuring process was expected to be completed in early 2013 but encountered delays. It halted Ziraat Bank’s disbursements as the PIU team was affected. During the process, corporate, commercial and SME departments were moved to Istanbul, including the PIU4. As a result, while disbursements cumulated to USD107 million between September, 2010, and September, 2011, they slumped to just over USD13 million between September 2011, and June 2013. This can be attributed to the low attention due to conflicting priorities, but also to the need to establish collaboration mechanisms to facilitate the implementation of the loan. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization The project’s monitoring and Evaluation (M&E) framework was comprehensive and well- designed. The framework designed PDO indicators in line with the project’s objectives. The indicators were in line with the project’s activities and adequately measured the project’s impact (portfolio quality, geographical coverage, number of beneficiaries and PFIs). In an improvement from the SME I project, the M&E framework monitored the project’s impact on SMEs in terms of employment and revenues. However, two indicators that were monitored in the SME I project results framework were dropped in the SME II project - the maturity and the currency of subloans. As a result, these indicators were not reported in supervision documents. M&E arrangements were clear and well implemented. The PAD and the project’s operations manuals provided explicit M&E arrangements, as well as a clear definition of expectations in terms of reporting. Baseline and targets were provided for all indicators. The Operations Manuals designed appropriate reporting templates for consistency and reliability purposes. Indicators were also effectively monitored during project supervision. Almost all Aide Memoires and Implementation Status and Results Reports (ISRs) provided updated information for the results framework. As a result, compared with its predecessor SME I, the consistency and quality of data received for this ICR was satisfactory. The M&E framework also anticipated the undertaking of a survey of project beneficiaries, in order to better assess the impact of lines of credit operations. Such as survey aimed to reach SMEs beyond those financed by the SME II Project – such as beneficiaries of EFIL and SME I projects. The impact evaluation analysis was undertaken in 2011 and included data on SME I/II and EFIL III/IV clients. The study was useful in providing insights on the additionality of such credit lines. A beneficiary survey was not undertaken at project completion. 2.4 Safeguard and Fiduciary Compliance Procurement, financial management, and environmental safeguards requirements were detailed in the project Operations Manuals. Kalkinma Bank had a different role during project implementation compared to Ziraat Bank and Vakif Bank, given the different intermediation models. The project Operations Manuals provided strong guidance to Kalkinma Bank, Ziraat Bank, Vakif Bank, and the three PFIs. Each borrower established a PIU that was fully funded and staffed by each bank, and operated under the overall supervision of a Deputy General Manager. Implementation of fiduciary and environmental safeguards procedures was overall satisfactory. The well-developed and comprehensive Operations Manuals, and the proactive and 4 New staff was assigned to the PIU. 11 continuous support from the World Bank team contributed to effective safeguards and fiduciary compliance. The PIUs effectively undertook all project arrangements. The three participating banks ensured that their operating systems were fully aligned with the project’s accounting, supervision, and reporting requirements as per World Bank procedures. Moreover, PFIs selected by Kalkinma Bank had prior experience with World Bank requirements through their participation in EFIL operations. In addition, no disruption was experienced when Ziraat Bank’s PIU moved to Istanbul from Ankara. Finally, the World Bank specialists provided several trainings to participating banks and PFIs on fiduciary and environmental safeguards procedures. Compliance with World Bank standards was, at times, challenging - in particular for fiduciary requirements. Project implementation required intensive support from the relevant specialists due to the borrowers’ unfamiliarity with World Bank procedures and the necessary adjustment to their loan appraisal and monitoring practices that did not always match the project’s requirements. With regards to compliance with environmental safeguards requirements, the project implementation was satisfactory. The borrowers were mainly required to collect due-diligence documents (i.e. proof of compliance), as the majority of subloans financed working-capital or equipment leasing. Borrowers however had to overcome some challenges in order to comply with fiduciary requirements.  Procurement: Overall, all financial intermediaries followed the procedures described in the Operations Manuals. A few ineligible expenditures were identified during reviews and subsequently addressed. The participating banks and leasing company complained on the systematic requirement to justify their eligibility by keeping records of working-capital expenditures, which often comprise of several small-amount items. PFIs would rather have subborrowers keep their own records, and rely on audit reports to verify the eligibility of expenditures. In comparison, when using their own resources to provide working-capital loans, PFIs request only a procurement plan (or spreadsheet of planned transactions) at the time of the loan application. Procurement requirements were otherwise found to be generally in line with commercial practices.  Financial management: Financial arrangements were mostly satisfactory. Kalkinma Bank developed a web-based platform that was used by PFIs for both the application process and financial management purposes. Ziraat Bank and Vakif Bank each developed internal systems to improve loan monitoring by the PIU, and access to documentation for verification purposes. Financial reporting between PIUs and branches for the SME II loans improved in quality during project implementation. 2.5 Post-completion Operation/Next Phase An SME III project was approved three years into the implementation of the SME II project. SME III, the successor to the SME II project, was approved in June 2013 with a similar project design. The SME III project requires a minimum maturity of three years for subloans (compared to two years in SME II), and at least 15 percent disbursement in underserved regions. The list of underserved regions was updated to the Official Gazette dated October 18, 2012. As an innovative departure from the SME I and SME II projects, the project design incorporated requirements of 25 percent co-financing by the borrower and 100 percent reflows to SMEs. Three additional credit lines to Turkish banks were under implementation. Two energy efficiency projects, approved in May 2009 and March 2013, provide financing for renewable energy and energy efficiency investments‒ the latter is specifically targeting SMEs. Also, the Innovative Access to Finance project, which was approved in July 2014, provides funding to SMEs and exporters through participation banks (Islamic finance) and factoring companies – the first time 12 the World Bank explores these two market segments in lines of credit. These credit lines were provided to state and private banks with a guarantee from the Turkish Treasury for on-lending to target beneficiaries either directly or through a subsidiary PFI. In all these projects, the objective remains to provide medium-and long-term financing. Despite some notable progress over the past 10 years, the maturity of loans in the overall banking sector has only gradually increased. This is largely due to the unfavorable maturity structure of banks’ balance sheets. World Bank and other IFI loans have contributed to gradually shifting the maturity structure of banks’ balance sheets towards longer–term funding and lending. New approaches to increase the leveraging potential of World Bank financing should be explored. Credit lines have substantial leveraging potential, through co-financing and reflow requirements, but also the ability of participating banks to attract additional financing from other IFIs. Going forward, other approaches such as partial credit guarantees or risk sharing facilities could be considered. The former would help banks receive more funding on international markets, while the latter would provide more funding to Turkish firms and/or help address collateral requirements. The World Bank made efforts to deploy an IBRD Partial Credit Guarantee in Turkey to attract longer term finance from international lenders, however the project was dropped before Board approval at the request of the Turkish Treasury. 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation The project remains highly relevant to Turkey’s development priorities and its design reflected constraints to SME finance. The project was fully in line with the CPS priorities at preparation and during implementation, as well as the 2007–13 Government’s Development Plan. The project is also fully consistent with the FY12–16 CPS, as well as the Tenth National Development Plan (NDP) (2014–18). The project provided critical long-term funding in the aftermath of the global financial crisis. In the context of an unprecedented international liquidity crisis, the project provided for a quick disbursement instrument at reasonable cost. As a wholesale lender, Kalkinma Bank quickly allocated funds to PFIs, and fully disbursed its credit line by September 2012, two years ahead of schedule. Demand from PFIs can only attest to demand in the real economy. For SMEs, funding was provided at competitive rates. Ziraat Bank and Vakif Bank both used their large branch network to reach out to remote, eligible borrowers. Although financial reflows to SMEs were not required and despite the long maturity of SME II subloans, Kalkinma Bank reported reflows of USD18 million since September 2012. The project relied on multiple delivery models, as an improvement from the SME I project. First, there were three borrowers, acting as on-lenders and direct lenders. On-lenders are able to reach several financial intermediaries with the end-goal of widening impact to PFIs and a potentially larger number of SMEs. Direct lenders included in this project already had the privilege of a large and wide branch network, clientele, and product offering. Second, the decision to include state-owned banks was also fortuitous, given their risk tolerance and mandate to lend in underserved regions compared to private banks. Third, leasing companies were involved in the implementation of an SME finance project as effective relays for investment finance. These coexisting models in a project design that imposed the same requirements on all borrowers allowed for a comparison of lending patterns useful for future interventions. 13 Implementation and M&E arrangements were strong and well-defined. During project preparation, the World Bank team candidly assessed potential implementation risks and developed effective arrangements to mitigate these risks. In addition, the project’s implementation arrangements were fully described in the Operations Manuals. The M&E framework was comprehensive (including baselines and targets), fully integrated with implementation arrangements, and even led to the completion of an impact study useful for both SME and EFIL projects for the first time. The project also provided a learning opportunity for all participants. The financial intermediaries – all three participating banks and PFIs– in this project had no prior experience in World Bank credit line operations providing finance to SMEs. Moreover, only Kalkinma Bank and the PFIs on its credit line had been exposed to World Bank fiduciary and safeguards requirements. Kalkinma Bank, which had initial difficulties in implementing the project and a previous IFI credit line as an on-lender, learned effective ways to accelerate the implementation of future projects. Ziraat Bank built on its experience as retail lender in SME II, for the implementation of the SME III project, this time as a wholesale lender. For Ziraat this meant relying on the same PIU and pursuing proven practices to ensure compliance with fiduciary and environmental procedures. Project implementation was affected in the past two years by the ambitious lending targets for underserved regions. The project very ambitiously aimed to provide financing of two years or more for SMEs operating in underserved regions, under the assumption that these firms would have similar financial capabilities as those operating in developed regions (i.e. financial covenants). Moreover, it was assumed that the demand for project funds from eligible borrowers would more than cover the funds earmarked for these regions. Demand was however affected to some extent by the lower export volumes in the southeastern region due to the geopolitical crisis in neighboring countries. Disbursements in these regions thus progressed slowly until the restructuring was undertaken in September, 2014. Participating banks and PFIs struggled to find eligible borrowers that were interested in receiving financing during the implementation period. The project’s requirements may have adversely led to the selection of more sophisticated, thus less financially- constrained SMEs, especially in underserved regions. However, even for more sophisticated firms, the project provides a distinct advantage through the provision of longer-term finance (minimum maturity of 2 years). The project restructuring accelerated disbursements and allowed for a successful project completion. The restructuring aimed to better align project developmental objectives with the regions’ absorption capacity. The project’s closing date was extended by only six months, but the project fully disbursed within this short timeframe. The key factor that explains this successful outcome was the removal of financial covenants for SMEs operating in underserved regions, along with intensified marketing efforts by the two remaining borrowers. 3.2 Achievement of Project Development Objectives The project’s main development objective of (i) broadening and deepening the access of Turkish SMEs to medium- and long-term finance, and (ii) contributing to an expansion of productive activities and job creation, was fully achieved. All project development indicators exceeded their end-project targets. At closing, the project’s achievements were substantial. Broaden and deepen the access of Turkish SMEs to finance.  Financial intermediation: The project contributed to the development of financial intermediaries. The project relied on six financial institutions for implementation: five private and state banks, as well as a leasing company, were responsible for intermediating 14 funds to SMEs. These financial institutions had no prior exposure to World Bank standards for SME lines of credit and, at closing, valued this learning experience that was helpful in securing and administering other IFI credit lines. Furthermore, the credit risk of the project was low compared with the average for the SME sector that had an NPL ratio of 3.5 percent as of end-March 2015. The project’s NPL ratio was 0.5 percent in number and 0.7 percent in volume (below the 7 percent target). Finally, the project contributed to a healthy financial sector because all project participants were required to maintain strong financial health throughout the life of the project. Kalkinma Bank, Ziraat Bank, Vakif Bank, and the PFIs have all fully complied with the project’s financial covenants throughout project implementation.  Lending in underserved regions: At the time of project restructuring, 10.6 percent of total project funds had been disbursed in underserved regions (or around USD54 million). At project closing, a combined 20 percent (USD 102 million) had been disbursed by all three participating banks in underserved regions. This significant achievement can be disaggregated as follows: Kalkinma Bank disbursed 30 percent of its credit line (USD30 million) to SMEs on priority regions, Ziraat Bank 18 percent (USD34 million) and Vakif Bank 19 percent (USD38 million). The loan agreement did not specify the regional lending requirements per borrower in terms of number or disbursed amount of subloans5 . The indicator is considered fully achieved. The table below highlights the project’s achievements. Participating banks successfully disbursed 20 percent in underserved regions, twice the usual lending share dedicated to these regions by Turkish banks in general (see Annex 2). Table 3: Lending in underserved regions Number of Percentage of Percentage of Regional priority total loans total loans requirement6 provinces with (by number) (by amount) SME II clients Kalkinma Bank 25% 26 36% 30% Ziraat Bank 15% 17 18% 16% Vakif Bank 15% 24 30% 19% Total 17% 39 27% 20% Provision of medium- and long-term finance to SMEs. A total of 956 loans were provided to 746 SMEs operating throughout Turkey, for a total amount of USD519 million. The average loan size was USD620,000 (median USD404,000). The average size of the beneficiary SMEs was 56 employees (median 40 employees) with an annual turnover of TL13 million (median TL8.2 million) at the time of the loan application. The average maturity of SME II loans was four years (47 months) disaggregated as follows: 38 months for loans disbursed by Kalkinma Bank, 58 months for Ziraat Bank and 41 months for Vakif Bank. Investment loans (USD142 million, 28 percent of total) had an average maturity of 49 months while working capital loans (USD373 million, 72 percent of total) had an average maturity of 47 months. These loans were offered at competitive rates and reflected market conditions, as evidence by mirroring rate hikes in 2012 as per CBRT’s monetary policy. Finally, the spatial inequality that has been observed 5 “The Borrower shall (…) ensure that at least twenty-five percent (25%) of the aggregate of Sub-loans and Lease Financings shall be made (…) in the Priority Regions.” 6 The regional lending requirement for Kalkinma Bank was not revised downwards because the bank had already fully disbursed its credit line at the time of project restructuring. 15 in the overall banking system, is reduced in the SME II project: not only was a higher share of loans disbursed in underserved regions, but more importantly, in developed regions lending has been much less concentrated in Ankara, Gaziantep, Hatay and Istanbul which traditionally account for 57 percent of loans in Turkey7, compared with 29 percent in SME II. As a result, even within developed regions, Bursa, Izmir, Kayseri and Konya received more financing then usual as a percentage of total amounts, in order to finance growth in these regions. Figure 3a: SME II loan distribution by loan amount 3b: SME II loan distribution by maturity Contribute to an expansion of productive activities and job creation.  Sales growth: For beneficiary SMEs, the project contributed to increased revenues in the total amount of TL2.1 billion. Due to the depreciation of the lira over the life of the project, additional sales in Turkish Lira often result in lower sales in US dollar. For that reason, it is not possible to provide meaningful ratios of sales to disbursed amounts. Nevertheless, individual loan data provide useful insights. First, firms that received financing in 2010 were affected by the crisis and mostly sustained, rather than expanded, their activities. Only 15 percent of loans approved in 2010 were for investment purposes. Second, the project’s impact in terms of additional sales is largest for SMEs that received a loan in 2011. This suggests that the impact of loans is best measured at least three years after the loan was disbursed. Third, providing larger loans to larger firms seems to have the highest impact in terms of sales growth.  Employment preservation and generation: about half of SMEs that received financing before 2015 created new jobs8. Overall, close to 7,000 net jobs have been created by firms 7 Data as of December 31, 2013 provided by the Turkish Banking Association 8 We do not have official employment and sales data for close to 200 firms that received financing between January and April 2015. For the rest of the beneficiaries, 98 percent of employment data is available and 88 percent of sale data. 16 financed by the project9, after receiving an SME II loan. For firms that grew, the employee headcount increased by 16 percent with the project. For SMEs that received a loan only in 2015, an estimated 350 jobs will be created in 2015. Furthermore, the project helped sustain employment at 173 beneficiary SMEs (7,800 jobs). As a result, in total, over 35,000 jobs have been preserved by firms that generated or sustained employment levels after receiving the loan, i.e., for every US$75,000 disbursed to an SME, one (1) job was created and five (5) jobs were preserved. Finally, further analysis of individual loan data suggests that: (i) providing financing to smaller firms has the highest employment impact; and (ii) SME beneficiaries created more jobs than firms that did not receive financing. We also find that the employment and sales impact of the SME II project (firms that took loans in before 2014) is statistically higher than that of the SME I project (firms that took loans in 2007 and 2008). Further details are provided in Annexes 2 and 3. 3.3 Efficiency The project was highly efficient in generating sales and employment. Due to substantial exchange rate differentials, estimated sales multipliers (ratio of increased sales to loan amount) are necessarily biased and underestimated. 10 Nevertheless, the median sales growth for beneficiary SMEs was 14.5 percent after receiving the loan (average of 40 percent, after excluding outliers). Most firms (87 percent) were able to sustain if not increase their employment levels thanks to the project (median growth of 0 percent but average of 34 percent, even after excluding startups). Furthermore, the project allowed banks to provide financing to SMEs over a long period of time, thereby providing financing to accompany these firms to grow. Finally, the project accompanied 11 firms in graduating from SME status, with an average headcount growth rate of 57 percent (end- project employee headcount ranged from 255 to 545 employees). The design of the project helped reach firms with high growth potential in underserved regions (eastern and south eastern regions). The job creation rate was highest in these regions (see table 4). Sales generation capacity was highest in firms operating in more developed regions. This suggests that firms in underserved regions have more labor-intensive activities. Table 4: employment and sales by regions Investment Increase in sales Increase in employment loans (TL million) (% total New Avg. Expected Total per Total Avg. amount) jobs in 2015 USD1mln disbursed Aegean 14% 1,796 12 555 20 554.8 4.7 Black Sea 25% 220 2 83 11 108.9 5.4 Central Anatolia 26% 1,432 10 383 18 382.0 3.8 Eastern Anatolia 53% 698 21 90 48 154.5 9.1 Marmara 27% 1,625 7 316 11 730.7 4.6 Mediterranean 43% 483 8 86 15 86.5 2.7 Southeastern Anatolia 40% 664 15 27 20 124.9 5.7 Total 28% 6,918 9 1,540 16 2,142.3 4.6 9 8,549 gross jobs were created and 1,631 jobs were lost; hence a net creation of 6,918 jobs 10 Total sales increased in Turkish Lira by TL2.1 billion. However, when converting annual sales into US dollars at year-end rates, we find that several firms had lower US dollar sales after they received the loans (despite having higher Turkish lira sales). 17 3.4 Justification of Overall Outcome Rating Rating: ― Satisfactory The project met its development objectives, which remain highly relevant for Turkey. The project coincided with an increased focus on SME development and contributed to an increase of medium to long-term financing. The project contributed to job creation and investments in the Turkish economy, at a time of a global economic slowdown. The M&E framework was well designed and arrangements were helpful to provide more reliable information on the project’s outcomes (compared to the SME I project). However, the original regional lending target was met by only one borrower, and certain SMEs operating in underserved regions were de facto excluded from project funds due to the project’s strict financial covenants. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development The project had a strong, albeit indirect, impact on poverty through the beneficiaries. The project helped create 6,918 jobs and increased the revenues of participating SMEs by TL2.1 billion11 over the life of the project. An additional 1,540 jobs are expected to be created in 2015 by the project beneficiaries. In particular, the project benefitted 220 SMEs in underserved regions (i.e. 30 percent of the beneficiaries): 254 loans for a total amount of USD102 million (USD437,000 average) were provided and helped generate 1,540 jobs and TL440 million additional sales. The project thus helped SMEs mitigate the impact of the global economic crisis and the geopolitical crisis in neighboring countries (for regions in the Southeast). In Southeastern provinces alone, the project contributed to the peace dividend: 50 loans of a total amount of USD35 million (USD780,000 average) were provided and helped generate 664 jobs and TL118 million additional revenues (a third of the beneficiaries are exporting firms, whose activities were affected by the crisis in the Middle Eastern countries). Furthermore, during visits to project beneficiaries, the ICR team also confirmed that the project helped some SMEs finance the shift from trade towards productive activities and investments to increase firm productivity. Contrasted with poverty data at the sub-regional levels, the analysis of project disbursements highlights the broad impact of the project across the country. The project disbursed USD86 million (i.e. 17 percent of the project amount) in the four regions whose share of the population living below the USD5 PPP poverty line is greater than the national average. (b) Institutional Change/Strengthening The project helped strengthen credit appraisal procedures among participating banks and PFIs and ensure PFIs’ compliance with strict prudential standards (prudential ratios were monitored throughout the project). In addition, the five banks and one leasing companies that participated in the project experienced a significant learning phase, partly owing to their initial lack of familiarity with similar World Bank operations. Even though the three PFIs previously participated in EFIL operations, the project came with new challenges due to regional lending requirements, and the smaller borrower and transaction size. Furthermore, all six participating financial intermediaries have successfully demonstrated that medium and long-term lending is a viable business proposition, 11 This is equivalent to USD286 million in additional revenues, using year-end exchange rates over the life of the project. This is because the Turkish lira has significantly depreciated over the life of the project. 18 even in periods of economic downturn. Finally, PFIs built capacity to assess the environmental risks and impacts of borrowing firms. The project was also an attempt to revise the existing SME lending practices of financial intermediaries, by requiring the use of cash-flow statements and financial covenants during the credit appraisal process. These project requirements were expected to be more favorable to SMEs than existing credit assessment methodologies. Over the life of the project, these project requirements were however counter-effective in underserved regions, where the capability of firms was lower than in developed regions. The SME II project was Kalkinma Bank’s first experience with the World Bank as an on-lender. The bank developed a web-based IT system for smooth collaboration with PFIs and for monitoring purposes. In addition, Kalkinma Bank signed an agreement of USD80 million with the Council of Europe Development Bank (CEB) to on-lend to SMEs via leasing companies. The bank capitalized on its experience with the SME II project and developed “best practices” that it intends to apply to future on-lending credit lines. These include the systemic undertaking of roadshows before project approval to promote the credit line and ensure acceleration of project implementation; and the flexible allocation of funds between PFIs and use of commitment fees to avoid implementation delays and foster competition. Ziraat Bank and Vakif Bank were newcomers in World Bank operations, and experienced a significant learning curve that lead to the signing of multiple financing agreements with the World Bank but also other IFIs (see table below). Significantly, the SME II project was implemented while both banks were increasing their own SME lending resources. The number of SMEs financed by Ziraat Bank increased by 21 percent between 2010 and 2015. Vakif Bank’s SME loan portfolio almost quintupled over the same period. Table 5: Ziraat Bank and Vakif Bank – SME Credit Line with IFIs (2010-2015) PROJECT BANK YEAR IFI AMOUNT (millions) BENEFICIARY Vakif 2010 EIB SMEs EUR 75 Vakif 2010 EBRD SMEs and Midcaps USD 60 Vakif 2010 EBRD SMEs and Midcaps USD 44 Vakif 2011 EIB SMEs EUR 50 Ziraat 2011 EIF Guarantee for loans to €1.3bn disbursed to beneficiaries (implementation enterprises of less than 9 by Ziraat, leveraging €300mln started in 2013) employees partial guarantees Vakif 2012 EIB SMEs and Midcaps EUR 200 Vakif 2012 EBRD SMEs and Midcaps EUR 20 Ziraat 2012 EIB SME direct lending EUR 100 Vakif 2013 EIB Municipalities EUR 100 Vakif 2013 EIB SMEs and Midcaps EUR 50 Vakif 2013 EBRD SMEs and Midcaps USD 80 Vakif 2013 IBRD SMEs and Midcaps USD 67 Vakif 2013 KfW SMEs EUR 100 Ziraat 2013 EIB SMEs and Midcaps EUR 100 (1st tranche) Ziraat 2013 IBRD SME energy efficiency USD 67 Vakif 2014 EIB SMEs and Midcaps EUR 100 Vakif 2014 EBRD SMEs EUR 25 Vakif 2014 EBRD SMEs and Midcaps EUR 75 Ziraat 2014 EIB SMEs and Midcaps EUR 100 (2nd tranche) Ziraat 2014 KfW Agriculture SMEs EUR150 Ziraat 2014 AfD Agriculture SMEs EUR100 19 PROJECT BANK YEAR IFI AMOUNT (millions) BENEFICIARY Ziraat 2014 CEB SMEs (50% Ziraat EUR100 leasing, 50% Ziraat bank) Ziraat 2014 EIB Financing of IPARD EUR 100 projects Ziraat 2015 EIB SMEs and Midcaps EUR200 Ziraat 2015* EIF SMEs Ziraat Bank may also participate in the EIF COSME guarantee program, in which SME loans of up to €150,000 would be eligible. (c) Other Unintended Outcomes and Impacts (positive or negative) PFIs, through their participation in SME II loans, build infrastructure and capacity to work with other IFIs, and have a greater awareness of available sources of funding. During the ICR mission, all financial intermediaries insisted on their ability to leverage World Bank loan proceeds to obtain additional funding from other IFIs or the market. In addition, they demonstrated great sensibility to the need to improve their monitoring and evaluation systems. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops N/A 4. Assessment of Risk to Development Outcome Rating: Moderate Risks to the development outcomes relate to general country and financial sector risks and export performance (for Southeaster regions mostly), in addition to the project-specific risks. Country risks include the geopolitical context and turbulence in the financial markets. The International Monetary Fund, Fitch and Moody’s warned against foreign funding dependence and foreign exchange lending in the Turkish banking sector. Central bank data show that the short-term foreign debt of banks (i.e., maturing in less than two years) more than doubled over the past four years to USD94 billion, representing 40 percent of the total foreign debt of banks. These trends may affect the provision of medium to long-term funding to SMEs. Nonperforming loans in the banking sector have remained low despite the sharp depreciation of the Turkish lira since 2013, but could rise as NPLs are a trailing indicator. Project-specific risks are related to the future performance of Kalkinma Bank, Ziraat Bank and Vakif Bank, the remaining PFIs, and the end-beneficiaries. The financial performance of financial intermediaries has been monitored during supervision missions and, even during challenging times, all of them complied with project requirements. With regards to demand, there is sustained demand from SMEs for term finance. Kalkinma Bank was able to find additional borrowers for reflows (USD 18 million), which is a sustainable element of this type of operation. This has become a feature in the design of the follow-up SME III and Innovative Access to Finance projects. However, going forward, credit line operations may become more difficult to implement and their development outcome compromised, if project requirements naturally exclude the most financially constrained firms. 20 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory The project design was highly relevant to strategic priorities and responsive to requests of the Turkish government for fast assistance to the financial sector. The project’s risk assessment framework was comprehensive and candid, and mitigation measures were appropriate. M&E arrangements were well-written and very detailed. The Bank also developed reporting templates that helped ensure the availability of quality data at the time of the ICR. Baseline and end-project targets were provided for all indicators. Furthermore, to prevent implementation delays, the Bank team accompanied Kalkinma Bank in several roadshows with potential PFIs before the approval of the project. This was very effective in ensuring effectiveness just a month after signing despite having three borrowers, including one on-lender. The project design was largely based on the design and success of the predecessor SME I project. The regional lending requirement for the SME II project was not project-specific, but set by the Turkish authorities for all IFI credit lines. The financial covenants were determined by the Bank team. The financial covenants which SMEs had to comply with were an obstacle in achieving the regional lending target of 25 percent of disbursed loans on time. Given the lack of flexibility in the minimum lending requirement in underserved regions and the lower capacity of SMEs in these regions, it would have been preferable to set differentiated project requirements for SMEs operating in developed and priority regions. A more prudent approach allowing more flexibility would have helped satisfy the triple goal of: (i) providing long-term funding to finance their growth, (ii) ensuring that beneficiaries implement good practices (environmental and fiduciary), and (iii) reaching financially-constrained SMEs. Paradoxically, the most financially underserved SMEs may have been excluded by the project design even before the local banks’ credit appraisal, due to the Bank’s financial eligibility conditions. Several improved features of the loan built, nevertheless, on the development outcomes achieved in the SME I project. These include: (i) longer minimum maturity requirements‒ a minimum maturity of 2 years compared with 1 year in SME I; (ii) full flexibility for banks to manage the FX risk and on-lend in local currency, Euro or US dollar as needed, which ultimately lead to a higher share of lira-denominated SME loans; (iii) diversity of financial intermediaries with the inclusion of a leasing company, which helped reach smaller SMEs (the loan ticket size was smaller at USD620,000 compared with USD820,000), and (iv) inclusion of three borrowers that could not only substantially benefit from exposure to IFI procedures to raise funds with other IFIs, but also ensure greater spatial diversification (with 67 provinces in SME II, 51 of which are in eastern and central regions ‒ compared to 67 total provinces in SME I, and 44 in eastern and central regions ). The Bank’s quality at entry is thus rated satisfactory. (b) Quality of Supervision Rating: Satisfactory Project supervision was undertaken regularly and entailed field visits to borrowers, PFIs, and end- beneficiaries. Aide-memoires and ISRs following implementation support missions were submitted on time and provided updated information on economic developments in Turkey and their impact on EFIL IV. All financial intermediaries were visited for this ICR and all praised the Bank team’s effective collaboration and proactivity at times. The field-based presence of the TTL and the fiduciary and safeguards team was well appreciated. In addition, there was full continuity of the Bank team, as the TTL for the project was involved since the beginning of the project. ISRs were 21 comprehensive and ratings were in line with project performance. For future operations, it would be advisable that ISRs provide the number of beneficiaries, amount disbursed and loan maturity per borrower (in the component section) to have a better snapshot of individual borrowers’ performance. The project restructuring was delayed until the end of the project, because both the Bank and borrowers sought to achieve a very ambitious developmental outcome. In 2013, the Bank team worked intensively with Ziraat Bank and Vakif Bank to identify all measures that could be taken by both banks to accelerate disbursements in underserved regions, without undertaking a project restructuring. These included: greater use of regional offices, lending in local currency and at fixed rate for USD loans, as well the introduction of disbursement targets and incentives. When all alternatives failed to produce the expected results, discussions on project restructuring, initiated in May, 2014, materialized with a request from the two borrowers in August 2014. The Bank team swiftly proceeded with project restructuring in September 2014, after already intensifying implementation support for the project since April 2014. A project midterm review was not undertaken. Such a review would have been helpful to address implementation bottlenecks and undertake an earlier project restructuring, possibly being a missed opportunity to assess the impact of removing financial covenants for SMEs operating in underserved regions on the banks’ ability to achieve the 25 percent regional lending target. The satisfactory Supervision rating is justified by the successful completion of a large-scale project with three borrowers, the successful and rapid project restructuring and the considerable efforts extended by the Bank team in addressing implementation bottlenecks throughout supervision. (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory Based on the above, the overall rating for Bank performance is Satisfactory. 5.2 Borrower Performance (a) Government Performance Rating: Satisfactory The Government’s performance as a guarantor was Satisfactory. The Government demonstrated strong ownership of the project and commitment to its objectives. In addition, the Government supported the development of the leasing industry with the adoption in November 2012, of a specific law that recognizes the activity and allows for new products, such as operational leasing, and sale and lease back. Furthermore, in December 2011, the Government reduced the VAT for machinery and equipment leases from 18 to 1 percent, which stimulated productive investments. Although effective in enhancing the development outcomes of credit line operations in Turkey, regional lending targets were applied across the board to all IFI operations, regardless of eligibility criteria, financial instruments, and project requirements. In addition, regional lending requirements were set at 25 percent in 2010 and 15 percent in 2012, although a thorough assessment of the absorption capacity of underserved regions had not been undertaken for this project (taking into account project requirements and economic conditions). (b) Implementing Agency or Agencies Performance Rating: Satisfactory 22 The performance of Kalkinma Bank was highly satisfactory: the bank exceeded all project development outcomes, and upgraded its internal system to comply with monitoring and reporting requirements. Kalkinma Bank proceeded with a reallocation of funds between PFIs to accelerate project implementation and complied with all project requirements. Finally, the bank disbursed fully its credit line, two years ahead of schedule. The performance of Ziraat Bank was moderately satisfactory. Ziraat Bank complied with requirements throughout project implementation, and effectively staffed the PIU. Through prudent FX management, the bank managed to disburse USD218 million to SMEs, based on USD200 million funding. Also, Ziraat Bank provided longer term funding compared to the other two participating banks. Implementation delays for this project can be partly attributed to initial shortcomings in ensuring compliance with fiduciary standards, and the lengthy reorganization during which the bank disbursed only USD510,000 in 2012, after two years of outstanding performance. Finally, Ziraat bank’s disbursement rate in underserved regions was the lowest among the participating banks. The performance of Vakif Bank was satisfactory. Vakif Bank fully complied with all requirements throughout project implementation, and ensured good quality reporting. Vakif Bank undertook the foreign exchange risk and disbursed the majority of loan proceeds to SMEs in local currency. This flexibility enabled Vakif Bank to achieve the greatest spatial distribution of loans (with 45 provinces in total). In underserved regions, Vakif Bank provided loans to SMEs in 24 provinces located in underserved regions, compared with a maximum of 17 provinces by other financial intermediaries. Finally, the average loan amount in underserved regions was lower for Vakif Bank (USD 329,000) compared to other intermediaries (that had an average disbursed amount greater than USD 418,000). (c) Justification of Rating for Overall Borrower Performance Rating: Satisfactory Based on the above ratings of the government performance and of the implementing agencies, the overall rating for borrower performance is Satisfactory. 6. Lessons Learned Key lessons for ongoing and future operations are: The World Bank’s current prudential credit eligibility criteria need to be balanced against the World Bank’s efforts to reach smaller, financially underserved firms. Credit eligibility criteria (requirements on firm leverage and debt service coverage) had to be relaxed for SMEs operating in underserved regions in order to avoid excluding firms that would otherwise not be eligible to receive this longer-term financing. While the use of credit eligibility criteria is necessary to prudentially exclude non-viable or over-indebted firms, current indicators (D/E ratio and DSCR) are inadequate to appraise the financial health of firms in certain regions and countries due to the unavailability of reliable financial statements. The current approach to identifying eligible firms could be revised if the existing credit appraisal methodologies of PFIs is assessed to be satisfactory and able to provide reliable information on firms’ financial capacity. Such an assessment could be undertaken at project preparation, and lead to the removal of financial covenants. These financial covenants, along with project’s fiduciary and environmental requirements, not only increase the transaction costs for PFIs that, by consequence, find it more attractive to lend in larger amounts, but also lead to the financing of only larger and more sophisticated firms. 23 Lending quotas are inefficient alone. The performance of the project would have been highly satisfactory if lending targets had been removed. The desire to increase the project’s development impact impeded swift project implementation, because of the disconnect between eligibility criteria and disbursement targets in underserved regions. A more effective approach to lending in these underserved regions needs to be considered, in order to address other financing constraints in these regions beside unavailability of financing. Given the pre-existing commitment of PFIs intermediating the SME II project to lend to SMEs and reach out to clients in underserved areas with their own resources, the focus could shift to loan appraisal, collateral requirements and increasing SME capacity. In addition, incentives to increase lending in underserved regions, or traditionally less attractive/ riskier SMEs, could be considered (e.g. reducing reflows or co- financing requirements for PFIs based on their outreach to more financially-constrained firms). During the last year of the SME II project, but also the SME I project, participating financial intermediaries struggled to meet geographical lending requirements, particularly due to a large inflow of financing in the underserved regions by other IFIs. Given the strong demand for financing in these areas, better disbursement performance in the future would require increasing their absorption capacity. The experience of the SME II project and other LOCs in Turkey suggest that a wholesale lending model is preferable. An on-lending design increases the project’s overall impact. First, a wider range of financial intermediaries increases the number and diversity of potential project beneficiaries and may ultimately lead to funding of underserved firms. Second, the on-lending design enables new financial intermediaries to gain experience in working with international organizations, which will help them develop other sources of funding. Third, on-lending encourages healthy competition and causes less market distortion than direct lending, as the benefits are spread over a large share of the financial sector. Development banks are unable to reach out to smaller beneficiaries due to their business model and are most effective in an on-lending model. In the SME II project, a leasing company was introduced for the first time through Kalkinma Bank’s credit line, which led to a higher share of investment financing compared to working capital financing. Kalkinma Bank was also able to accelerate project implementation with the inclusion on a new PFI. However, given the intermediation costs, interest rates were slightly higher than that of other borrowers. On the other hand, direct lending would provide distinct advantages in crisis times (quick disbursement and counter-cyclical funding) and when lending in niche markets with targeted operations (where the private sector is absent or not able to fulfill the demand). Implementation support missions and midterm reviews provide opportunities for strengthening the project. In the SME II project, implementation support missions included regular visits to PFIs and field visits to branches and beneficiaries. In addition to providing insights for overall implementation bottlenecks, upfront training and capacity-building on fiduciary and environmental safeguards procedures is highly beneficial for stakeholders. Results achieved in the SME II project can be largely attributed to the intense supervision undertaken since 2013, including field visits. Bank teams should take advantage of midterm reviews to assess performance holistically and set expectations for end-project results. Project restructuring may be recommended before midterm review based on borrowers’ commitments, changing economic conditions and prior experience with similar operations. While project restructuring can be highly administrative and may put undesired attention on a project, quick restructuring is advisable to avoid losing momentum in project implementation and may yield higher-than-expected results. Alignment with business practices increases the project’s attractiveness. In Turkey, commercial practices have been assessed as acceptable with regard to World Bank procurement requirements. As a result, project procurement requirements should be similar to existing practices of PFIs for subloan preparation and record-keeping requirements. Given the recurrent complaints of PFIs and 24 subborrowers, a clear definition of working-capital expenditures and distinctions between procurement and non-procurement working-capital expenditures is warranted. The Innovative Access to Finance Project is already following this approach, which should be more widespread among credit line operations. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies See annex 7. (b) Cofinanciers N/A (c) Other partners and stakeholders N/A 25 Annex 1. Project Costs and Financing (a) Project Cost by Component (in USD Million equivalent) Actual/Latest Appraisal Estimate Percentage of Components Estimate (USD (USD millions) Appraisal millions) Total Baseline Cost 500.00 500.00 100% Physical Contingencies 0.00 0.00 Price Contingencies 0.00 0.00 Total Project Costs 500.00 500.00 100% 0.00 0.00 IBRD Front-end Fee 1.25 1.25 100% Total Financing Required 500.00 500.00 100% (b) Financing Appraisal Actual/Latest Type of Estimate Estimate Percentage of Source of Funds Cofinancing (USD (USD Appraisal millions) millions) Borrower 0.00 0.00 International Bank for Reconstruction 500.00 500.00 100% and Development 26 Annex 2. Outputs by Component Table A2.1: Results framework Baseline End- Target project PDOs Number of SME beneficiaries n/a 746 200 Increase in employment in beneficiary SMEs n/a 6,918 jobs >0 TL2.1 Increase in sales in beneficiary SMEs n/a >0 billion* Share of finance provided to underserved regions n/a 20% 17%** Quality of project financed portfolio ‒ NPL ratio n/a 0.7% 7% Number of participating financial institutions n/a 6 5 Intermediate outcomes Volume of Bank support n/a 104%*** 100% of loan **** Borrower- Compliance with prudential norms compliant compliant compliant PFI - Compliance with prudential norms**** compliant compliant compliant Notes: *TL 2.1 billion is equivalent to USD900 million using end-2014 exchange rate. However, for better accuracy, we calculated the sales differential based on the annual exchange rate for each of the firm’s reported sales before and after receiving the loan. We thus found that beneficiary SMEs increased their sales in equivalent dollar amount by USD286 million over the life of the project. For example, if a firm had TL40mln in sales prior to receiving the loan in end-2010 and increased its sales to TL55mln in end-2014, in U.S. dollar terms, that would mean sales of USD26mln (using end-2010 exchange rate) and USD21mln (using end-2014 rate). As a result, we would be reporting lower sales of USD5mln, as opposed to higher sales of TL15mln, only due to exchange rate differentials. The return on investment hence cannot be calculated in U.S. dollar terms **The original target was 25 percent. Following the restructuring for components 2 and 3, the target was dropped to 17 percent. *** Due to exchange rate differentials over the life of the project, a total of USD518.5 million was disbursed by the borrowers to SME beneficiaries. **** The compliance of borrowers and PFIs to prudential norms was systematically monitored during implementation. In addition, several additional indicators were reported in ISRs for each borrower. These indicators were not part of the results framework and served to provide additional information on borrowers’ financial and lending performance. Table A2.2: Additional information on borrowers Kalkinma Bank Ziraat Bank Vakif Bank Baseline Most recent Baseline Most recent Baseline Most recent Return on Assets/Equity 2% 1.3% 3.1% 1.8% 1.98% 1.11% Portfolio at Risk – SME 10.8% 3.81% 2.3% 2% 6.1% 3.7% Outstanding SME Loan 316 813 2,358 16,567 2,800 11,263 Portfolio (USD millions) # active loan accounts –SME 530 179 92,119 1,810,821 40,000 263,504 Volume of Bank Support: 0 100 0 218.1 0 200.4 Lines of Credit - SME A detailed analysis of the results requires going beyond the original results framework and reviewing individual subloan data through the lenses of the project’s objectives. We also review lending patterns in underserved regions 27 Table A2.3: Breakdown of Subloans (Kalkinma Bank) Number of Amount Sector Subloans USD (%) Count (%) Agriculture, forestry and fishing 910,317 0.9% 2 1.4% Basic Metals 1,732,503 1.7% 3 2.0% Construction 28,391,324 28.4% 36 24.3% Electrical Equipment 1,075,207 1.1% 2 1.4% Industrial Packaging 986,519 1.0% 1 0.7% Marketing 2,382,748 2.4% 1 0.7% Mining and Quarrying 2,989,859 3.0% 4 2.7% Pulp paper printings 557,605 0.6% 2 1.4% Textile 10,035,754 10.0% 15 10.1% Tourism 5,478,862 5.5% 10 6.8% Transportation and Storage 14,668,760 14.7% 17 11.5% Wholesale and Retail Trade 11,739,800 11.7% 22 14.9% Wood products 192,772 0.2% 1 0.7% Medical Equipment and Services 4,964,953 5.0% 7 4.7% Machinery and Equipment 6,266,589 6.3% 11 7.4% Food products 5,895,569 5.9% 10 6.8% Rubber and Plastic Products 1,730,858 1.7% 4 2.7% TOTAL 100,000,000 100% 148 100% Region USD (%) Count (%) Aegean Region 5,507,525 5.5% 9 6.1% Black Sea Region 6,669,358 6.7% 14 9.5% Central Anatolia Region 17,623,702 17.6% 26 17.6% Eastern Anatolia Region 5,484,593 5.5% 9 6.1% Marmara Region 47,700,087 47.7% 63 42.6% Mediterranean Region 5,426,005 5.4% 7 4.7% Southeastern Anatolia Region 11,588,730 11.6% 20 13.5% TOTAL 100,000,000 100% 148 100% Loan Size USD (%) Count (%) < USD250,000 5,208,858 5.2% 40 27.0% USD250,000 - USD500,000 13,084,807 13.1% 37 25.0% USD500,000 - USD1,000,000 28,968,318 29.0% 40 27.0% USD1,000,000 - USD2,000,000 30,224,344 30.2% 22 14.9% > USD2,000,000 22,513,673 22.5% 9 6.1% TOTAL 100,000,000 100% 148 100% Beneficiary size (# employees at application) USD (%) Count (%) <10 14,344,108 14.3% 30 20.3% 10-19 7,307,963 7.3% 15 10.1% 20-49 21,820,162 21.8% 37 25.0% 50-99 24,937,104 24.9% 31 20.9% 100-249 31,590,662 31.6% 35 23.6% TOTAL 100,000,000 100% 148 100% % Avg. loan Avg. Avg. Avg. firm Avg. % of % Inv. By PFI amount amount maturity firm D/E DSCR firm size loans Loans disbursed (USD) (months) ratio ratio (empl.) Banks 75% 72% 700,935 60% 37 61:39 2.55 71 Leasing 25% 28% 609,756 100% 41 57:43 3.02 37 28 Table A2.4: Breakdown of Subloans (Ziraat Bank) Number of Amount Sector Subloans USD (%) Count (%) Accomodation, tourism and food service 6,548,899 3.0% 4 1.1% Agriculture, forestry and fishing 9,493,914 4.4% 18 5.1% Basic Metals 18,867,873 8.7% 36 10.2% Chemicals 600,000 0.3% 2 0.6% Construction 16,611,423 7.6% 17 4.8% Food products 12,601,892 5.8% 18 5.1% furniture 1,520,000 0.7% 2 0.6% Glass products 150,000 0.1% 1 0.3% Information and communication 100,000 0.0% 1 0.3% Machinery and Equipment 32,934,625 15.1% 52 14.8% Medical Equipment and Services 2,201,900 1.0% 4 1.1% Mining and Quarrying 1,266,000 0.6% 3 0.9% Rubber and Plastic Products 9,026,135 4.1% 16 4.5% Services 7,634,285 3.5% 12 3.4% Textile 30,173,050 13.8% 33 9.4% Transportation and Storage 1,298,160 0.6% 4 1.1% Wholesale and Retail Trade 26,727,236 12.3% 57 16.2% Wood products 1,245,701 0.6% 3 0.9% Other manufacturing 33,998,445 15.6% 60 17.0% Other 5,079,231 2.3% 9 2.6% TOTAL 218,078,771 100% 352 100% Region USD (%) Count (%) Aegean Region 26,728,954 12.3% 35 9.9% Black Sea Region 8,448,748 3.9% 30 8.5% Central Anatolia Region 61,895,291 28.4% 102 29.0% Eastern Anatolia Region 8,233,299 3.8% 16 4.5% Marmara Region 80,088,245 36.7% 125 35.5% Mediterranean Region 20,702,857 9.5% 35 9.9% Southeastern Anatolia Region 11,981,377 5.5% 9 2.6% TOTAL 218,078,771 100% 352 100% Loan Size USD (%) Count (%) < USD250,000 16,490,309 7.6% 110 31.3% USD250,000 - USD500,000 31,980,180 14.7% 92 26.1% USD500,000 - USD1,000,000 54,246,382 24.9% 83 23.6% USD1,000,000 - USD2,000,000 58,866,711 27.0% 47 13.4% > USD2,000,000 56,495,189 25.9% 20 5.7% TOTAL 218,078,771 100% 352 100% Beneficiary size (# employees at application) USD (%) Count (%) <10 25,881,334 11.9% 54 15.3% 10-19 20,812,907 9.5% 49 13.9% 20-49 61,939,342 28.4% 118 33.5% 50-99 52,178,833 23.9% 66 18.8% 100-249 57,266,355 26.3% 65 18.5% TOTAL 218,078,771 100% 352 100% 29 Table A2.5: Breakdown of Subloans (Vakif Bank) Number of Amount Sector Subloans USD (%) Count (%) Accommodation, tourism and food service 5,043,759 2.5% 10 2.2% Agriculture, forestry and fishing 5,437,390 2.7% 9 2.0% Basic Metals 14,094,162 7.0% 26 5.7% Chemicals 12,503,667 6.2% 25 5.5% Construction 17,264,437 8.6% 48 10.5% Food products 4,367,910 2.2% 15 3.3% furniture 484,878 0.2% 3 0.7% Glass products 5,076,041 2.5% 7 1.5% Information and communication 756,116 0.4% 4 0.9% Machinery and Equipment 36,555,546 18.2% 83 18.2% Marketing 1,500,000 0.7% 1 0.2% Medical Equipment and Services 2,305,425 1.2% 4 0.9% Mining and Quarrying 7,388,651 3.7% 14 3.1% Pulp paper printings 2,606,622 1.3% 4 0.9% Rubber and Plastic Products 6,014,145 3.0% 13 2.9% Textile 27,721,078 13.8% 53 11.6% Transportation and Storage 4,015,078 2.0% 13 2.9% Water and waste management 309,579 0.2% 2 0.4% Wholesale and Retail Trade 40,546,440 20.2% 112 24.6% Wood products 4,765,569 2.4% 7 1.5% Other 1,693,974 0.8% 3 0.7% TOTAL 200,450,465 100% 456 100% Region USD (%) Count (%) Aegean Region 83,724,342 41.8% 157 34.4% Black Sea Region 13,204,883 6.6% 60 13.2% Central Anatolia Region 21,327,703 10.6% 53 11.6% Eastern Anatolia Region 3,297,473 1.6% 10 2.2% Marmara Region 55,222,597 27.5% 122 26.8% Mediterranean Region 12,123,767 6.0% 33 7.2% Southeastern Anatolia Region 11,549,700 5.8% 21 4.6% TOTAL 200,450,465 100% 456 100% Loan Size (total per borrower) USD (%) Count (%) < USD250,000 13,010,565 6.5% 117 25.7% USD250,000 - USD500,000 29,786,132 14.9% 100 21.9% USD500,000 - USD1,000,000 61,753,648 30.8% 129 28.3% USD1,000,000 - USD2,000,000 60,094,917 30.0% 79 17.3% > USD2,000,000 35,805,203 17.9% 31 6.8% TOTAL 200,450,465 100% 456 100% Beneficiary size (# employees at application) USD (%) Count (%) <10 9,752,768 4.9% 63 13.8% 10-19 17,980,726 9.0% 55 12.1% 20-49 58,808,479 29.3% 164 36.0% 50-99 51,259,606 25.6% 102 22.4% 100-249 62,648,886 31.3% 72 15.8% TOTAL 200,450,465 100% 456 100% 30 Given specific objectives of facilitating finance in priority regions we derive the following: 1. Underserved regions Table A2.6: list of underserved regions in the SME I and SME II projects Exports 2012 SME I SME II Per capita gross value added (GVA) (TL) USD’000 TRA11 Erzurum x x 45,859 TRA12 Erzincan x x 3,082 9,893 TRA13 Bayburt x x 348 TRA21 Ağrı x x 42,142 TRA22 Kars x x 3,241 6,708 TRA23 Iğdır x x 84,662 TRA24 Ardahan x x 1,881 TRB11 Malatya x x 282,729 TRB12 Elazığ x x 55,277 9,759 TRB13 Bingöl x x 7,350 TRB14 Tunceli x x - TRB21 Van x x 21,708 TRB22 Muş x x 13,988 5,894 TRB23 Bitlis x x 2,528 TRB24 Hakkari x x 362,993 TRC11 Adıyaman x x 104,826 TRC12 Gaziantep x 5,579,875 8,303 TRC13 Kilis x x 11,671 TRC21 Diyarbakır x x 109,818 7,179 TRC22 Şanlıurfa x x 198,959 TRC31 Mardin x x 948,394 TRC32 Batman x x 83,513 7,862 TRC33 Şırnak x x 1,018,062 TRC34 Siirt x x 6,226 TR222 Çanakkale x 169,713 15,013* with Balýkesir TR522 Karaman x 278,689 11,935* with Konya TR631 Hatay x 2,039,566 TR632 Kahramanmaraş x x 753,878 9,900 TR633 Osmaniye x x 117,871 TR711 Kırıkkale x 5,603 TR712 Aksaray x 62,720 TR713 Niğde x 62,647 11,882 TR714 Nevşehir x 46,912 TR715 Kırşehir x 203,331 TR721 Kayseri x 1,582,410 TR722 Sivas x x 93,295 11,192 TR723 Yozgat x x 9,611 TR811 Zonguldak x 321,158 TR812 Karabük x 167,424 14,313 TR813 Bartın x 14,116 TR821 Kastamonu x 30,383 TR822 Çankırı x 50,905 11,057 TR823 Sinop x 27,292 TR831 Samsun x x 422,749 TR832 Tokat x x 37,346 11,338 TR833 Çorum x x 166,369 TR834 Amasya x x 65,015 TR901 Trabzon x x 1,099,596 TR902 Ordu x x 334,142 TR903 Giresun x x 175,028 11,154 TR904 Rize x x 389,437 TR905 Artvin x x 77,594 TR906 Gümüshane x x 216 31 Table A2.7: Comparison between SME II and sector wide lending in underserved regions Provinces Banking sector-wide (as of 12/31/2013) % of total SME II loans % of total loans by amount by amount Aksaray 0.2% 0.2% Amasya 0.2% 0.1% Ardahan 0.0% 0.0% Artvin 0.1% 0.4% Batman 0.1% 0.3% Bitlis 0.1% 0.0% Çanakkale 0.4% 0.5% Çankırı 0.1% 0.5% Çorum 0.4% 1.3% Diyarbakır 0.6% 1.2% Elazığ 0.3% 0.9% Erzincan 0.1% 0.4% Erzurum 0.4% 0.3% Giresun 0.2% 0.1% Gümüşhane 0.1% 0.0% Iğdır 0.1% 0.4% Kahramanmaraş 0.7% 3.3% Karabük 0.2% 0.0% Karaman 0.2% 0.0% Kastamonu 0.2% 0.2% Kırşehir 0.2% 0.1% Malatya 0.4% 0.3% Mardin 0.3% 0.3% Muş 0.1% 0.1% Nevşehir 0.2% 0.2% Niğde 0.2% 0.6% Ordu 0.4% 0.4% Osmaniye 0.2% 0.7% Rize 0.3% 0.0% Samsun 1.0% 2.2% Şanlıurfa 0.5% 0.3% Siirt 0.1% 0.0% Sinop 0.1% 0.0% Şırnak 0.1% 0.5% Sivas 0.3% 0.2% Tokat 0.3% 0.7% Trabzon 0.6% 1.3% Van 0.2% 0.8% Yozgat 0.2% 0.6% Total 10.3% 19.7% The disbursement rate in underserved regions was 10 percent sector-wide in 2013, compared to 20 percent for the SME II project. In addition, the SME II project provided greater access to finance to 18 such provinces (highlighted), where the share of . These achievements are noteworthy. 32 Table A2.8a: SME II lending by region (loan size) %# D/E DSC Firm size Firm size loans ratio R before after SME Loan Average, for firms Maturity SME II II loan amount operating in: (months) loan (employee (USD) (employee s) s) priority regions 27% 39 436,882 0.59 4.00 41 47 < USD250,000 46% 37 118,822 0.58 5.9 25 29 USD250,000 - 25% 38 366,376 0.60 2.8 40 42 USD500,000 USD500,000 - 17% 46 736,894 0.57 3.5 71 81 USD1,000,000 USD1,000,000 - 7% 50 1,279,742 0.64 2.0 67 108 USD2,000,000 > USD2,000,000 4% 34 2,719,478 0.56 2.0 80 70 other regions 73% 51 691,526 0.59 4.01 63 72 < USD250,000 21% 45 145,139 0.59 3.8 37 42 USD250,000 - 24% 52 347,819 0.59 3.8 51 59 USD500,000 USD500,000 - 30% 53 692,890 0.59 5.2 69 81 USD1,000,000 USD1,000,000 - 18% 51 1,327,400 0.59 3.1 89 100 USD2,000,000 > USD2,000,000 7% 55 2,588,587 0.62 2.2 102 125 Table A2.8b: SME II lending by region (beneficiary) size) %# D/E DSC Firm size Firm size loans ratio R before after SME Loan Average, for firms Maturity SME II II loan amount operating in: (months) loan (employee (USD) (employee s) s) priority regions 27% 39 436,882 0.59 4.00 41 47 <10 22% 38 262,072 0.51 2.7 3 15 10-19 20% 40 334,358 0.62 3.2 13 14 20-49 33% 39 385,611 0.61 3.3 33 35 50-99 14% 36 708,317 0.63 2.6 69 66 100-249 11% 46 831,454 0.59 2.8 156 186 other regions 73% 51 691,526 0.59 4.01 63 72 <10 13% 50 444,553 0.61 3.7 5 8 10-19 10% 48 480,463 0.62 3.5 14 21 20-49 33% 52 608,056 0.58 4.3 34 41 50-99 23% 51 765,186 0.59 4.3 69 75 100-249 21% 51 984,001 0.59 3.7 152 173 33 Note: These trends were confirmed when using individual loan data for Kalkinma Bank which on- lent to SMEs via banks and leasing companies. There were no significant differences between these two types of financial intermediaries. 2. Maturity More than a fifth of the total SME II amount consisted of loans with a maturity greater than six years (primarily 7-year loans actually), as figure A2.1a shows. In underserved regions, the median loan maturity was three years. Figure A2.1a: Loan distribution by maturity Figure A2.1b: Loan distribution by maturity (number and volumes) for all regions (number and volumes) in priority regions only % total amount % loans (number) % total amount % loans (number) 30% 50% 25% 40% 20% 30% % total % total 15% 20% 10% 5% 10% 0% 0% 24 24˂≥36 36˂≥48 48˂≥60 60˂≥72 >72 24 24˂≥36 36˂≥48 48˂≥60 60˂≥72 >72 months months 3. Graduating SMEs Fifteen firms graduated from SME status (i.e. the number of employees is greater than 250) after receiving an SME II loan. Three of these firms operate in underserved regions. On average, firms received loans of USD1 million with a 5-year maturity. Firms in priority regions received investment loans, while those in other regions received working capital loans. 34 Annex 3. Economic and Financial Analysis The sample for our analysis consisted of the 746 project beneficiaries, which received loans between 2010 and 2015. Thanks to the quality of data provided by all five financial intermediaries, the sample may be considered as representative of the SME population eligible for financing. Since no data is available for 2015, the firms that received financing in 2015 cannot be evaluated, however they can be used as a natural control group for firms that received financing between 2010 and 2014. The differential analysis of the changes within the two groups is helpful to more confidently attribute outcomes achieved to the project, rather than overall economic developments and trends. We excluded outliers in both groups for better comparability. Preliminary evidence suggest from these comparisons suggest that, on average, SMEs that received financing before 2015 (treatment group) outperformed SMEs that did not receive loans until 2015 (the control group), in terms of employment and sales. The median growth rate of employment for both the treatment and control group was zero (i.e. no jobs were created). In contrast, the median sales growth rate was higher in the control group: there is thus less clear evidence on the impact of the loans in terms of sales. Figure A3.1: average and median growth rates (before and after the loan), employment and sales for all SME beneficiaries based on the year that they received the SME II loan. 60% 50% 40% 30% 20% 10% 0% -10% -20% Employment growth Sales growth Sales growth (avg.) (avg.) (median) 2010-2011 2012-2014 2015 Employment. As we run basic regressions (see below), we find some evidence that the loans had a greater employment impact on smaller firms and firms that received loans before 2015 (i.e. treatment group) and firms that received investment loans. Sales. Unsurprisingly, the impact is greater for larger firms, in terms of the additional volumes of sales generated. Evidence also suggest that firms that did not receive loans before 2015 (control group) were larger, and thus had a larger absolute value of sales generated than firms that did (treatment group). However, if we consider firm growth (i.e. percentage increase in sales) in practice, firms in the treatment group (loans before 2015) grew more quickly than firms in the control group. This preliminary finding is promising, as it suggests that SME II helped beneficiaries reach their growth potential. Underserved regions. With regards to priority regions, we find preliminary evidence that firms in these regions grow more quickly in terms of sales than firms in other regions. 35 Table A3.1: Linear regressions for employment and sales Employment Employment Sales increase Sales growth increase (log) growth (log) (percentage) (percentage) Employment at application (log) -0.51**** -0.11** 0.56*** 0.03 Loan amount (log) 0.02 -0.04 0.36*** -0.01 Maturity 0 0 -0.01*** -0.005*** Percentage of amount received 0.31* 0 0 0 as investment loan Firm has received the loan -0.05 -0.52*** 0.60*** -0.28** before 2015 Firm is in a priority region 0.19 -0.11 -0.29 0.21* Adjusted R-squared 0.18 0.02 0.28 0.01 Number of observations 281 658 396 518 ***,**,** indicate significance at the 1,5 and 10 percent significance levels respectively. Comparison SME I and SME II beneficiaries. For the purpose of the ICR, we also compare results achieved in the SME I and SME II project. We use data on firms that had participated in the SME I project which had been collected at the time of the impact evaluation. We thus have employment growth and sales growth figures between 2008 and 2009 for firms that had taken an SME I loan in 2007 or 2008. Although 2007 data is not available (these are thus not perfect before and after figure), we can use the available data as a proxy for performance after the program since the impact on sales and employment was most probably not immediately felt on the firm just a year after employment. For better comparability, we only compare the performance of SME I beneficiaries to those of SME II beneficiaries that received a loan before 2014. We find that the impact of the loans was larger for SME II beneficiaries than SME I beneficiaries. We are not able to disaggregate these findings between developed and priority regions or based on other criteria, due to the limited availability of data. Table A3.2: Comparison between SME I and SMEII in terms of employment and sales Employment Employment Sales increase Sales growth increase (log) growth (log) (percentage) (percentage) Employment at application (log) 0.51*** -0.21*** 0.63*** 0 Firm received an SME I loan -0.91*** -0.6*** -10.7*** -0.41** Adjusted R-squared 0.37 0.06 0.84 0.02 Number of observations 301 542 360 478 ***,**,** indicate significance at the 1,5 and 10 percent significance levels respectively. 36 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility/ Names Title Unit Specialty Lending Halil Agah Environmental Advisor CSSD Furuzan Bilir Operations Officer ECCU6 Vinod K. Goel Consultant, FIand SME Specialist GTCDR Ulker Karamullaoglu Program Assistant ECCU6 Isfandyar Zaman Khan Program Leader ECCU5 Hala Khattar Sr. Financial Officer (Treasury) BCFBD Hannah Koilpillai Sr. Finance Officer LOAFC Zeynep Lalik Senior Financial Management Specialist GGODR Alper Ahmet Oguz Senior Financial Sector Specialist GFMDR Ahmet Gurhan Ozdora Senior Operations Officer ECSEG Carlos Pinerua Country Manager ECCHR TTL Margaret Png Senior Legal counsel LEGEC Tatiana Segal Senior Operations Officer GTCDR Ibrahim Sirer Senior Procurement Specialist GGODR Supervision/ICR Esra Arikan Senior Environmental Specialist GENDR Ayse Seda Aroymak Financial Management Specialist Steen Byskov Senior Financial Economist GFMDR Salih Bugra Erdurmus Procurement Specialist GGODR Selma Karaman Program Assistant ECCU6 Ulker Karamullaoglu Program Assistant ECCU6 Isfandyar Zaman Khan Program Leader ECCU5 Aminata Ndiaye ET Consultant GFMDR ICR primary author Alper Ahmet Oguz Senior Financial Sector Specialist GFMDR TTL since 2012 Carlos Pinerua Country Manager ECCHR TTL until 2011 Ilias Skamnelos Senior Financial Sector Specialist GFMDR ICR TTL (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including No. of staff weeks travel and consultant costs) Lending FY 10 30.93 252.65 Total: 252.65 Supervision/ICR FY 11 24.21 127.87 FY 12 22.51 86.75 FY 13 27.50 71.05 FY 14 14.69 36.00 FY 15 17.89 65.76 Total: 387.43 37 Annex 5. Beneficiary Survey Results N/A 38 Annex 6. Stakeholder Workshop Report and Results N/A 39 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR Kaklkinma Bank Based on the studies and activities on wholesale banking between 2007-2009 within SELP-II (Small Enterprise Loan Program) , which is a EU financed pre-accession technical assistance program project, Kalkınma has espoused it as a new growing instrument. Before implementing IBRD SME-II project Kalkınma have two wholesale banking program financed by its own sources and one program financed by French Development Agency (AFD). After the signature of the loan agreement with IBRD for SME-II project Kalkınma has signed two new wholesale banking loan agreements with Council of Europe Development Bank (CEB) in August 2010 and December 2013. Intermediary financial institutions’ selection has been undertaken in two stages. The first is to determine the limits according to the scoring from the quarterly financial data of Intermediary financial institutions, prepared by the Financial Institutions Department of Kalkınma. The second is the conformity assessment of PFI apex project executed by the Business Development Department. Pre-defined six (6) potential financial institutions have been informed in order to open up negotiations by top management of Kalkınma, by taking into consideration the requirement that 25% of the loan shall be used for financing priority (less developed) regions for project development objectives and expectations of World Bank Officials as to include small and medium size financial institutions in to the program. In the framework of the above information and impressions from road shows, Board of Kalkınma has approved to negotiate loan agreements with Denizbank ve Türkiye Garanti Bankası in the amount of 35 million USD and 65 million USD respectively. The Subsidiary Loan Agreements (SLAs) with the Financial Intermediaries were submitted to WB no-objection in September 2010 and SLAs have been signed on 29th September 2010. Since expected progress has not been achieved and disbursement period for Garanti was over on 30 September 2011, upon the request of Garanti for extension of final disbursement request date to 31st March 2012 and transfer of 25 m USD loan limit to Garanti Finansal Kiralama A.Ş. (Garanti Leasing-GFK) negotiations with Garanti Leasing has been accelerated. Amendment Agreement with Garanti and Subsidiary Loan Agreement in the amount of 25 m USD with GFK have been signed on 15th December 2011 and 26th December 2011 respectively and WB no-objection has been received for both agreements. Project Implementation Unit (PIU): the PIU is located in the Business Development Department of Kalkınma. The overall administrative coordination of the SME II, and general information management, is the responsibility of the Kalkınma’s PIU. During the implementation period several visits to PFIs have been held by the WB and Kalkınma. After completion of the disbursements during WB mission some of the sub-borrower project sites have been visited. Furthermore Kalkinma has visited some of the sub-project sites individually during the post review stage. SME-II Implementation Results Regarding PDOs  GARANTI has completed its WB SME-II Loan in the amount of 40 million USD (100%), 54 SMEs have been financed (6 working capital, 49 investment -1 both), loan request period extension to 31 March 2012 has been applied 12,41 million USD in Priority Regions (31%)  DENIZBANK has completed the utilization of WB SME-II loan in the amount of 35 million USD (42%), 39 SMEs have been financed (32 working capital, 7 investment) -loan 40 request period extension to 31 July 2012 has been applied. 7.6 million USD sub-loans are from Priority regions (21,7%)  GARANTI LEASING has utilized 25 m USD, 41 SMEs have been financed (all investment), 7.1 million m USD in Priority Regions (28.4%) PFI NAME NUMBER EMPLOYMENT EMPLOYMENT INCREASE INCREASE OF SMEs PRESERVED CREATED IN SALES IN (*) (%) EXPORT (%) GARANTI BANK 54 3,781 45 25 50 DENIZBANK 39 3,365 188 21 11 GARANTI LEASING 41 1,530 1,031 40 159 TOTAL 134 8,676 1,264 (*) The comparison has been made between 2012 figures and 2010 or 2011 figures by paying attention rationality. Conclusion and Lessons Learned With the completion of the project implementation 134 SME projects from several different sectors have been financed and the considerable amount of increase in employment, export and sales revenues have been achieved when comparing the figures of 2010, 2011 and 2012. In conclusion although there are strict reporting requirements and financial covenant, the IBRD SME-II project is beneficial for all the stakeholders and it is instrumental for Türkiye Kalkınma Bankası for capacity building in wholesale banking activities. It has provided an opportunity of cash flow based loan implementation to be executed by the intermediary financial institutions. Although implementation has some difficulties regarding financial covenant, procurement and environmental policy of WB group the intermediary financial institutions were comfortable with the long-term funding facility and prestigious implementation opportunity of IBRD financed project execution. For the future project implementation a training including procurement procedures, environmental procedures and financial evaluation of the final beneficiaries should be better delivered for intermediary financial institutions staff with the accompany of the IBRD and PIU staff right after the loan agreements signed. Additional comments received on the Draft ICR Kalkınma disbursed fully its credit line two years ahead of schedule and the performance of Kalkınma has been evaluated as highly satisfactory by IBRD. Based on the studies and activities on wholesale banking between 2007-2009 within SELP-II (Small Enterprise Loan Program), which is an EU-financed pre-accession technical assistance program project, Kalkınma has espoused it as a new growing instrument. Before implementing IBRD SME-II project, Kalkınma has two wholesale banking programs financed by its own sources and one program financed by French Development Agency (AFD). After the signature of the loan agreement with IBRD for SME-II project, Kalkınma has signed two new wholesale banking loan agreements with Council of Europe Development Bank (CEB) in August 2010 and December 2013. Within the context of its completed and ongoing wholesale banking activities, Kalkınma provided over 800 SMEs with financing, and generated additional employment opportunities to more than 6,300 people through intermediary financial institutions. And the share of wholesale banking loans in Kalkınma’s total loans reached 11% by the end of 2014. 41 Thanks to the experience acquired from previous SME projects and IBRD SME-II project, Kalkınma continues to carry out new SME projects through its own sources as well as ones provided by international financial institutions. Moreover, Kalkınma is ready to take part, in the future, in all kinds of SME projects planned to be carried out over the county or on the regional basis. ZiraatBank SME-2 project has been successfully completed by Ziraat Bankası as of 31.03.2015 which was the Project deadline. All the related Marketing and Analysis Departments as well as Funding Management Department carried out the implementation of the Project in cooperation with each other. During the implementation period of five years (2010-2015), Ziraat Bankası experienced difficulties as well as benefiting from the advantages of the project. The most important advantages of the credit line was its grace period, long maturity and reasonable cost that provided both Ziraat Bankası and final beneficiary SMEs with the long term financing and investment opportunities. Besides, brand value of the World Bank combined with the well-established Ziraat Bankası name made the loans under the project more attractive for SMEs. The cooperation contributed to the perception of both the World Bank and Ziraat Bankası as long term and reliable finance providers for the SMEs. Comments & Recommendations We would recommend the World Bank to review its financial eligibility criteria and decrease or remove the DSCR and Debt to Equity Ratio requirements for prospective projects focusing on SMEs and micro enterprises market and rely on the intermediary institution’s, i.e. Ziraat Bankası’s own assessment system. In other words, we believe that it would be more convenient if implementing agency applies its own financial eligibility procedure based on its own experience and assessment criteria when assessing a firm’s financial risk and it would provide a much smoother implementation of the project. As the analysis procedures of the banks are strict enough to show a very low NPL ratio and is in line with the World Bank’s objectives, we would highly recommend that for the projects that will be implemented in the future. In addition, we think that easing of the documentation requirements of World Bank would have a big positive impact on the implementation of the projects. Rigid documentation and procurement rules of the World Bank cause the occurrence of a negative perception towards World Bank projects not only by our clients but also by the marketing people. According to the World Bank rules, our branches are obliged to fill the table for Statement of Expenditures invoice by invoice when utilizing a SME-2 loan. This seems normal in the first place. However, when it comes to practice, it proves to be very inefficient and sometimes unnecessary. For instance, when one of our branch makes a SME-2 loan to a client, that particular branch fills the SOE table with invoices that are subject to the loan, but, does not precisely match the total invoice amount with total loan utilization amount. In other words, sometimes our branches add more invoices to the SOE table than necessary. This shouldn’t be a problem in our point of view given that total invoice amount is not below total utilization amount. However, during World Bank audits, World Bank auditors pay so much attention to these sort of details, this makes our customers very sensitive when they are offered a new World Bank loan and even leads to an unwillingness to take World Bank loans. Although World Bank procurement guidelines cannot be changed, we recommend World Bank to adapt its documentation requirements for these kind of credit lines in order not to weaken marketing efforts of implementing agencies. 42 Vakif Bank 1. Assessment of the Project’s Objective, Implementation, and Operational Experience 1.1 Project’s Objective The objective of the project was to strengthen the access of small and medium enterprises to medium term finance. By supplying finance to SMEs in Turkey, the project aimed to help raise the sales and export volumes of these SMEs and reduce the credit gap between SMEs and large firms. Evaluation of the results indicated that objectives of the facility were met by Vakifbank. 1.2 Project’s Implementation The marketing of the facility was carried out by SME, Commercial and Corporate Marketing Departments jointly. SME Product Development Department also supported them in this respect. The implementation progressed very well both inside and outside of the underserved regions, however especially the success of the project inside of the underserved regions is highly related to the restructuring made in the midst of it, which should be particularly mentioned. In September 2014, a significant restructuring was made in the project which reduced the disbursement portion of underserved regions from %25 to %15 by the approval of Turkish Treasury. This restructuring also exempted the companies in the underserved regions from the Debt to Equity and Debt Service Coverage Ratio requirements. The data about the implementation of the project before and after the restructuring clearly indicate that the approach tried after the restructuring, especially exemption from some ratio requirements is much more appropriate for structures and preferences of the companies in the underserved regions. By the means of these vital restructuring, Vakıfbank finalized this esteemed World Bank facility on 30 March 2015. Thus, we would like to propose World Bank to remove or ease restricted conditions particularly in the underserved regions for the further projects. 1.3 Operational Experience The implementation of the facility was performed by Vakifbank PIU. The PIU was consisted of the following departments;  International Banking Department  SME Product Development Department  SME and Commercial Marketing Departments  Central Operation Department  Financial Analysis Department International Banking Department was responsible for facilitating coordination with World Bank, as well as disbursements, reporting to World Bank and Turkish Treasury under the project. SME Product Development Department was in charge of developing the product for the clients in accordance with the eligibility criteria of the facility. They also coordinated the marketing of the facility. SME, Commercial and Corporate Marketing Departments were responsible for pricing for the loan applications. Central Operation Department was in charge of receiving the loan applications and granting prior approvals considering the number of employees and turnover of the firm. Financial Analysis (Appraisal) Department was responsible for calculating the Debt Service Coverage Ratio and Debt to Equity Ratio of the clients having prior approval. 43 2. Assessment of the Outcome of the Project against the Agreed Objectives Under the facility Vakifbank financed 454 loans with the total amount of USD 200.03 Million. The average loan size was USD 440.600. The loans were made according to the criteria set up in the project document with the following details:  The total disbursement amount was USD 200.03 Million.  The number of SMEs financed by World Bank SME II project was 334.  The average maturity of the portfolio was 43.7 months.  Loans were made in TL, USD and EUR.  Loans were given in 7 geographic regions and 48 provinces in Turkey.  %19 of the loan portfolio was disbursed in underserved regions which was well above the %15 disbursement requirement for those regions.  The NPL ratio of the project is %0.2. The above results indicate that the overall objectives of the project were achieved. Distribution of loans according to sector, loan size and region is presented Table 1, 2 and 3 below. 3. Evaluation of the borrower’s own performance during the implementation of the project The implementation of this esteemed World Bank project suggest that needs and characteristics of the clients should be appraised very well in the structuring phase of a product. In this respect, the restructuring made in September 2014 allowed us to reach much more eligible customers in the underserved regions when they were exempted from the financial ratio requirements of DSCR and D/E. 4. Evaluation of the performance of World Bank during the implementation of the project The World Bank’s performance during the implementation was very satisfactory. During the implementation, we had a close relationship with World Bank team in Ankara, Washington and Zagrep. The issues relating the implementation of the project were smoothly resolved thanks to the World Bank Team’s valuable support. The only point we would like to raise is related to eligibility criteria of the project. The financial ratio requirements (DSCR and D/E) were binding for us in terms of reaching eligible customers especially in the underserved regions. More flexibility in those requirements would speed up the implementation of the project and the project would be finalized without the need of any restructuring. We would also like to note that the 25% disbursement requirement to the underserved regions set by Turkish Treasury was also challenging due to the lack of enough eligible clients for the project in those regions. Vakifbank does not have a negative bias against the underserved regions. Quite the contrary, we are very active in those regions with our various products. However, restrictive conditions such as financial ratio requirements or target of disbursement amounts create difficulties in terms of meeting the needs of the clients. Thus, we would kindly propose you to remove or ease the restrictive conditions in those regions for the further projects. We would like to emphasize that we will also be honored to collaborate with the World Bank in the future. We are of the opinion that, if our recommendations are taken into consideration in the further projects we could be much more successful in the implementation of the facilities.   44 Table 1 Sector Number Percentage MANUFACTURING INDUSTRY 200 44,1% WHOLESALE TRADE 137 30,2% CONSTRUCTION SECTOR 37 8,1% OTHER SOCIAL AND INDIVIDUAL 23 5,1% HEALTH 15 3,3% COMMUNICATION 13 2,9% MINING 10 2,2% AGRICULTURE 8 1,8% TOURISM 5 1,1% FISHING INDUSTRY 1 0,2% REAL ESTATE 2 0,4% ELECTRICITY 2 0,4% FINANCIAL INTERMEDIATION 1 0,2% Table 2 Loan Amount(USD) Number Percentage Amount Percentage 0-100.000 86 19% 4.925.214 USD 2% 100.000-250.000 104 23% 17.806.677 USD 9% 250.000-500.000 127 28% 46.168.577 USD 23% 500.000-1.000.000 95 21% 69.668.490 USD 35% 1.000.000-3.500.000 42 9% 61.468.442 USD 31% Table 3 Geographic Regions Number Percentage Amount Percentage AEGEAN REGION 156 34% 83.667.413 USD 42% SOUTHEAST ANATOLIAN REGION 18 4% 10.746.466 USD 5% MARMARA REGION 126 28% 56.159.110 USD 28% BLACK SEA REGION 69 15% 15.052.503 USD 8% CENTRAL ANATOLIA REGION 42 9% 18.982.076 USD 9% MEDITERRANEAN REGION 33 7% 12.127.583 USD 6% EASTERN ANATOLIA REGION 10 2% 3.302.252 USD 2% Additional comments received on the Draft ICR We strongly believe that, success of the project, is highly related to the restructuring made in the midst of it, which should be particularly mentioned. In september 2014, a significant restructuring was made in the project which reduced the portion of underservedr egion from 25% to 15% by the apporval of Turskih Treasury. This restructuring also exempted the companies in the underserved regions from the Deb to Equity and Debt Service Coverage Ratio requriements. The data about the implementation of the project before and after the restructuring clearly indicate that the approach tried after the restructuring, especially exemption from some ratio requirments is much more appropriate for structures and preferences of the companies in the underserved regions. By the means of this vital restructuring, VakifBank finalized this esteemed world Bank facility on 30 March 2015. Thus we would like to propose World Bank remove, or ease, restricted conditions particularly in the underserved regions for further projects. 45 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders N/A 46 Annex 9. List of Supporting Documents 1. Project Appraisal Document, Implementation Completion Report and IEG ICR review and available portfolio data: Access to Finance for SME Project (SME I) 2. Second Access to Finance for SME Project (SME II): i. Project Appraisal Document ii. Concept note review meeting: Paper, Decision, comments iii. ROC meeting minutes iv. Loan agreements, Guarantee agreement and Supplemental Letters v. Operational Manual vi. Restructuring papers – and amended agreements vii. Project Aide memoires, ISRs, Financial Management and Procurement Reports 3. Third Access to Finance for SME Project (SME III): Project Appraisal Document 4. Turkey: EFIL and SME Credit Line Projects- An Assessment of Outcomes (2011) 5. Turkey: Improving Conditions for SME Growth - Finance and Innovation (2011) 6. Turkey’s Transitions: Integration, Inclusion, Institutions Flagship Report (2014) 7. Country Partnership strategies 47 MAP 48