Document of The World Bank FOR OFFICIAL USE ONLY Report No: ICR00004625 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA 5292-NG) ON A CREDIT IN THE AMOUNT OF SDR199.5 MILLION (US$ 300 MILLION EQUIVALENT) TO THE Federal Ministry of Finance of Nigeria, Nigeria Mortgage Refinance Company FOR A Housing Finance Development Program October 31, 2019 Finance, Competitiveness And Innovation Global Practice Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective May 02, 2019) Currency Unit = Nigerian Naira US$1= 306.45 US$1= SDR 0.72 FISCAL YEAR July 1 - June 30 Regional Vice President: Hafez M. H. Ghanem Country Director: Shubham Chaudhuri Senior Global Practice Director: Alfonso Garcia Mora Practice Manager: Rashmi Shankar Task Team Leader(s): Andrej Popovic ICR Main Contributor: Wenye Dong ABBREVIATIONS AND ACRONYMS AMCON Asset Management Company of Nigeria CAR Capital Adequacy Ratio CBN Central Bank of Nigeria CMU Country Management Unit CPS Country Partnership Strategy CS Consulting Services DFID Department for International Development DLI Disbursement Linked Indicators DMB Deposit Money Banks E&S Environmental and Social EA Economic Analysis e-MARs Electronic Mortgage Assets Registry System ERR Economic Rates of Return ESMS Environmental and Social Management System ESOM Environmental and Social Operations Manual FMBN Federal Mortgage Bank of Nigeria FMBN Federal Mortgage Bank of Nigeria FMoF Federal Ministry of Finance FMoJ Federal Ministry of Justice FMoLHUD Federal Ministry of Land Housing and Urban Development FSAP Financial Sector Assessment Program FSS Financial System Strategy FX Foreign Exchange GIZ Gesellschaft fuer Internationale Zusammenarbeit HMF Housing Microfinance Product M&E Monitoring & Evaluation MBAN Mortgage Bank Association of Nigeria MFB Microfinance Banks MLIs Mortgage Lending Institutions MPR Monetary Policy Rate MTR Mid-Term Review MTR Mid-term review NBS National Bureau of Statistics NHF National Housing Fund NMGC Nigeria Mortgage Guarantee Company NMRC Nigeria Mortgage Refinance Company NPL Non-Performing loans NPL Non-Performing Loans NPV Net Present Value OFISD Office of Other Financial Institutions Department PAD Project Appraisal Document PAT Project Administration Team PLR Program Learning Report PMB Primary Mortgage Banks PMLs Primary Mortgage Lenders PPP Public Private Partnership ROA Return on Assets ROE Return on Equity SSC Strategic Steering Committee TA Technical Assistance UUS Universal Underwriting Standards UUS Uniform Underwriting Standard JMRC Jordan Mortgage Refinance Company MTR Mid-Term Review STEP Systematic Tracking of Exchanges in Procurement TOR Terms of References PPR Procurement post review WHT With-Holding Tax VAT Value-Added Tax ERGP Economic Recovery and Growth Plan TABLE OF CONTENTS DATA SHEET .......................................................................................................................... 1 I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES ....................................................... 5 A. CONTEXT AT APPRAISAL .........................................................................................................5 B. SIGNIFICANT CHANGES DURING IMPLEMENTATION (IF APPLICABLE) .....................................12 II. OUTCOME .................................................................................................................... 15 A. RELEVANCE OF PDOs ............................................................................................................15 B. ACHIEVEMENT OF PDOs (EFFICACY) ......................................................................................16 C. EFFICIENCY ...........................................................................................................................30 D. JUSTIFICATION OF OVERALL OUTCOME RATING ....................................................................32 E. OTHER OUTCOMES AND IMPACTS (IF ANY)............................................................................33 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................ 35 A. KEY FACTORS DURING PREPARATION ...................................................................................35 B. KEY FACTORS DURING IMPLEMENTATION .............................................................................36 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME .. 40 A. QUALITY OF MONITORING AND EVALUATION (M&E) ............................................................40 B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE .....................................................41 C. BANK PERFORMANCE ...........................................................................................................45 D. RISK TO DEVELOPMENT OUTCOME .......................................................................................47 V. LESSONS AND RECOMMENDATIONS ............................................................................. 48 ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS ........................................................... 51 ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION ......................... 58 ANNEX 3. PROJECT COST BY COMPONENT ........................................................................... 60 ANNEX 4. BORROWER’S COMMENTS TO THE ICR ................................................................. 61 ANNEX 5. SUPPORTING DOCUMENTS (IF ANY) ..................................................................... 62 ANNEX 6. DETAILS OF THE FOCUS GROUP STUDY SESSIONS ................................................. 64 ANNEX 7. IDA RESOURCE ALLOCATION ON DIFFERENT HOUSING FINANCE PROJECTS ........... 68 ANNEX 8. DESCRIPTIONS AND BACKGROUNDS OF VARIOUS INSTITUTIONS .......................... 71 The World Bank Housing Finance Development Program (P131973) DATA SHEET BASIC INFORMATION Product Information Project ID Project Name P131973 Housing Finance Development Program Country Financing Instrument Nigeria Investment Project Financing Original EA Category Revised EA Category Partial Assessment (B) Partial Assessment (B) Organizations Borrower Implementing Agency Federal Ministry of Finance of Nigeria, Nigeria Central Bank of Nigeria Mortgage Refinance Company Project Development Objective (PDO) Original PDO The project development objective is to increase access to housing finance by deepening primary and secondary mortgage markets in the Federal Republic of Nigeria. Page 1 The World Bank Housing Finance Development Program (P131973) FINANCING Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) World Bank Financing 300,000,000 198,783,487 178,980,005 IDA-52920 Total 300,000,000 198,783,487 178,980,005 Non-World Bank Financing 0 0 0 Borrower/Recipient 0 0 0 Total 0 0 0 Total Project Cost 300,000,000 198,783,487 178,980,005 KEY DATES Approval Effectiveness MTR Review Original Closing Actual Closing 26-Sep-2013 05-Mar-2014 21-Sep-2016 31-Dec-2018 31-Dec-2018 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 25-Apr-2018 130.13 Change in Disbursements Arrangements Other Change(s) 19-Dec-2018 179.20 Cancellation of Financing KEY RATINGS Outcome Bank Performance M&E Quality Moderately Unsatisfactory Moderately Satisfactory Negligible RATINGS OF PROJECT PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 10-Jan-2014 Satisfactory Satisfactory 0 Page 2 The World Bank Housing Finance Development Program (P131973) 02 20-Jul-2014 Satisfactory Satisfactory 22.00 03 01-Jan-2015 Satisfactory Satisfactory 22.00 04 22-Jun-2015 Satisfactory Satisfactory 122.14 05 29-Dec-2015 Satisfactory Moderately Satisfactory 122.14 06 30-Jun-2016 Satisfactory Moderately Satisfactory 122.14 07 04-Jan-2017 Satisfactory Moderately Satisfactory 122.30 08 21-Jul-2017 Moderately Satisfactory Moderately Unsatisfactory 122.63 Moderately 09 14-Feb-2018 Moderately Unsatisfactory 130.13 Unsatisfactory Moderately 10 27-Jun-2018 Moderately Unsatisfactory 130.13 Unsatisfactory Moderately 11 19-Dec-2018 Moderately Unsatisfactory 179.20 Unsatisfactory SECTORS AND THEMES Sectors Major Sector/Sector (%) Financial Sector 100 Banking Institutions 37 Capital Markets 28 Other Non-bank Financial Institutions 35 Themes Major Theme/ Theme (Level 2)/ Theme (Level 3) (%) Finance 100 Financial Stability 98 Financial Sector oversight and policy/banking 98 regulation & restructuring Financial Infrastructure and Access 2 Financial inclusion 2 Page 3 The World Bank Housing Finance Development Program (P131973) ADM STAFF Role At Approval At ICR Regional Vice President: Makhtar Diop Hafez M. H. Ghanem Country Director: Marie Francoise Marie-Nelly Shubham Chaudhuri Director: Gaiv M. Tata Elisabeth Huybens Practice Manager: Paul Noumba Um Rashmi Shankar Task Team Leader(s): Michael D. Wong Andrej Popovic ICR Contributing Author: Wenye Dong Page 4 The World Bank Housing Finance Development Program (P131973) I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES A. CONTEXT AT APPRAISAL Context Macroeconomic Context 1. As outlined in the Project Appraisal Document (PAD), at the time of project appraisal in 2013, Nigeria had experienced stable economic growth averaging 8 percent over the past decade and 6.56 percent in 2012 1 . Macroeconomic risks remained significant, but the short- and medium-term outlook was generally positive. Substantial tightening of the macroeconomic policy helped to reduce year on year inflation to 9 percent in 2013, after years of annual inflation above 10 percent 2. The economy was highly dependent on oil, which accounted for over 80 percent of imports and over 70 percent of consolidated budgetary revenues 3. The general government budget deficit was reduced from an estimated 5.7 percent in 2010 to 1.9 percent in 2012. The balance of payments had been primarily in surplus since Q4 of 2011, which enabled authorities to build gross monetary foreign reserves from US$32 billion to US$48 billion between September 2011 and May 2013. Risks from an oil price shock were substantial, because the fiscal reserve (Excess Crude Account) was under US$6 billion as of mid-2013. An unexpected decline in crude oil output had tightened the budgetary situation and increased the medium-term balance of payments risks. Sectoral and Institutional Context 2. The Nigerian financial system was growing fast and becoming increasingly integrated into regional and global financial systems 4. At the end of 2011, gross financial assets had grown to a level equivalent to 61 percent of GDP. Banks were the largest asset owners, cumulatively holding 79 percent of total assets, followed by pension funds (12 percent of the total assets) 5. The Financial Sector Assessment Program (FSAP) conducted at the time confirmed that the Nigerian banking system was well capitalized, liquid, and profitable. The capital adequacy ratio (CAR) grew from less than 2 percent in 2010 to an average of 18 percent in June 2012. As of June 2012, the non-performing loans (NPL) ratio fell to about 5 percent of gross loans, mainly due to the transfer of bad assets to the Asset Management Company of Nigeria (AMCON). Return on assets (ROA) for the industry was at 1.2 percent, while return on equity (ROE) was at 8.9 percent. 3. Housing finance was underdeveloped. Mortgage loans outstanding were estimated at around N100 billion (US$850 million) 6, which represented just 0.36 percent of Nigeria’s 2011 GDP. Primary Mortgage Lenders (PMLs, which included commercial banks and Primary Mortgage Banks) had been unable to deliver the finance necessary to allow for the required investment into the housing sector. As of September 2012, there were 101 PMLs operating at both State 1 National Bureau of Statistics, GDP Report 4th Quarter and Full Year 2012 (Washington DC). 2 www.macrotrends.net 3 Nigeria Developing Housing Finance Report 4 Nigeria FSAP Aide-Mémoire March 2013 5 ibid 6 Estimated based on interviews with lenders in 2009 (Making Finance Work for Nigeria, World Bank, 2009). Page 5 The World Bank Housing Finance Development Program (P131973) and Federal level. Many were operating as shell companies without any significant number of mortgage loans on their books. Housing finance was limited especially due to several impediments that are summarized below: - Maturity mismatch in PMLs. PMLs faced a lack of longer-term resources due to failure of the Federal Mortgage Bank of Nigeria (FMBN) 7 to fulfill its mandate of developing a secondary mortgage market. Commercial banks have avoided retail banking and mortgage loans due to the long-term funding challenge. Developing a source of long-term funds for mortgage lenders was essential for the market to grow. Further, given the volatility of inflation in recent years, long-term savings products that could be a source of long- term funding were very limited. Capital markets were not well connected to the mortgage market despite strong demand from PFAs to invest in longer term assets. - High interest rates made housing finance unaffordable. The Monetary Policy Rate (MPR) was set at 12 percent 8 at the time of project design, which resulted in a high base lending rate for the market. Further, the high cost of appraising mortgages, lack of reliable credit information, high probability of default, high loss in the event of default due to the difficulties in foreclosing on properties, lack of reliable property valuations, and lack of a liquid property market on which to resell foreclosed properties, etc., drove up financing costs that were passed on to the mortgage borrowers, which made housing finance unaffordable. - The lengthy and the expensive property registration and transfer process was a major blockage to the growth of a formal housing finance sector. Costs varied from state to state but ranged from 20 percent to as much as 30 percent of the value of the property 9, which made formal acquisition and titling unaffordable for most homeowners. Also, the process took from six months to two years, largely due to the Land Use Act 1978, which vested ownership of all land to the Governors of each state and required formal consent for every single transaction related to the statutory rights of occupancy of urban land. Some states however, had been able to improve process and use mandated authorities to speed up process. - Very few mortgages were “perfected” 10. Due to the cost and time it took to register charges on a property, this process was often done using shortcuts such as ‘equitable’ mortgages which were a promise to register if needed. This form of security was unlikely to be sufficient for developing a secondary mortgage market. - High construction costs and limited supply of new housing. It was estimated that construction of a 3- bedroom house in Nigeria costed US$50,000 at the time, compared to US$36,000 in South Africa and US$26,000 in India. The costs were high due to high costs of building materials, high costs of skilled labor, and the costs of associated serviced infrastructures 11 . Cement prices were about 30-40 percent higher compared to neighboring countries or to the world market prices due to epileptic power supply 12. 7 Please see Annex 8 for the description and background of FMBN. 8 CBN 9 Nigeria FSAP Aide-Mémoire, March 2013 10 ibid 11 Okonjo-Iweala, 2014 12 Ibem, 2011, Okonjo-Iweala, 2014 Page 6 The World Bank Housing Finance Development Program (P131973) - The Institutional Framework to support the Housing Sector at Federal level was performing poorly. At the time of project design, there were several government institutions and programs in the housing sector. All these agencies and parastatals had their roles in ensuring housing production and delivery. The role and function of some of these agencies (i.e. FMBN and National Housing Fund) needed to be comprehensively reviewed and reoriented13. 4. The impact of urbanization trends at the time showed that Nigeria’s housing demand was expanding rapidly. It was estimated that up to 700,000 housing units were needed annually to keep up with the growing demand, while the production capacity at the time was below 100,000. As of 2013, the overall accumulated deficit in housing was reported at about 17 million units 14. 5. The development of the housing finance market could provide a wide range of income opportunities through the construction sector and related industries, as evidenced in other countries. Research showed that in India, 1.5 direct and 8 indirect jobs were created for each housing unit. In South Africa the number was 5.62 jobs per housing unit 15. Relevance to the Country Partnership Framework 6. At the time of the project preparation, the project aligned well with the Country Partnership Strategy (CPS) for the period 2010-2013, which aimed at maintaining non-oil growth through diversifying financial markets. Developing housing finance system would deepen the banking system and create diversification within financial markets to new product lines 16, and hence promote non-oil growth. The CPS also outlined the Government’s Seven Point agenda, which called for sufficient affordable housing to meet the needs of the country’s large population and the pressure of rural-urban migration 17. Theory of Change (Results Chain) 7. The project was articulated around a series of activities and outputs designed to address the affordable housing finance gap in Nigeria. The theory of change is illustrated in Figure 1 below. Figure 1: Nigeria Housing Finance Project Results Chain 13 Please see Annex 8 for the description and background of FMBN and NHF. 14 Ministry of Lands, Housing and Urban Development (Lagos) 15 Viruly, “Thinking about the Social Housing Market”, Presentation at the 2011 affordable housing Conference; Tibaijuka, “Building Prosperity: The Centrality of Housing in Economic Development” (2009); McKinsey Presentation at the 5th Global Housing Finance Conference, 2012 16 Nigeria CPS 2010-2013, page vi, para 8 17 Nigeria CPS 2010-2013, page 17, box 5 Page 7 The World Bank Housing Finance Development Program (P131973) Page 8 The World Bank Housing Finance Development Program (P131973) 8. The results chain represents how the activities envisaged under the project components lead to the associated outputs and connect to the achievement of the outcomes. The results chain is explained in more details in the following paragraph. 9. As mentioned in earlier section, housing finance market in Nigeria was underdeveloped due to a few key impediments including maturity mismatch in PMLs, high interest rate, unfavorable legal & regulatory framework, high cost of housing construction and limited housing supplies, as well as poor institutional framework for the housing sector. The Bank was asked to bring solutions that could help improve access to housing finance, with the assumption that doing so would create induced responses in the supply side. The Bank considered possible interventions based on its expertise and experiences from other countries and decided to include the following components in the project design: i. Establish a mortgage liquidity facility: The establishment and operationalization of the Nigeria Mortgage Refinance Company (NMRC) 18 would help address the maturity mismatch problems experienced by the PMLs. It would provide liquidity support to PMLs through refinancing activities, which would allow PMLs to provide longer term funds at better terms and conditions to mortgage lenders. The creation of NMRC would also enable more PMLs to develop mortgage products, which would introduce competition, hence promote efficiency and affordability. ii. Establish a mortgage guarantee product: The establishment of a guarantee product would help improve the housing finance affordability issue raised in the earlier section, as the product would bear some of the credit risk for lower income borrowers. The guarantee would provide credit default loss protection to mortgage lenders. Project design included a technical assistance (TA) component to design the guarantee product. iii. Housing Microfinance Product (HMF): HMF is another product that was designed to reach lower income borrowers. Project design included a TA component to help microfinance banks (MFBs) to test new HMF products, develop new or improve emerging HMF products. It also aimed to provide risk capital to MFBs in form of soft loans and a line of credit to MFBs to roll out the tested HMF products. iv. Market development activities: These include TA and capacity building activities with key stakeholders to create awareness and provide institutional strengthening, as well as activities to strengthen legal and regulatory environment underpinning the housing finance sector (i.e. develop mortgage foreclosure law, etc.). 10. Supply side interventions 19 were considered at first, but the decision was made to focus mainly on the demand side 20 due to various factors. Detailed analysis using census data was conducted ahead of the project to explore the supply side issues and the quality of housing stock. A number of these issues, such as land reform and construction sector development, were already being addressed through other projects and development partners at the time. For example, The WB/DFID GEM Project was addressing land reform issues in several states. GIZ had supported Technical Assistance (TA) to provide capacity building in the construction value chain. It was assumed that these projects would contribute to Page 9 The World Bank Housing Finance Development Program (P131973) the long-term outcome of improved land and legal systems, as well as improved housing construction and job creation. The project design also assumed that the liquidity facility would provide secured long-term funding to the PMLs at market-based rates, which would incentivize them to develop mortgage products with better terms (i.e. lower interest rates and longer tenors). Mortgage products with better terms would then attract more mortgage applicants, and as the demand for mortgage increases, the supply side (developers) would respond by providing more housing units in the same price range. Finally, the project design assumed that any investment in construction would have a multiplier effect on the national output and employment of country that goes beyond the direct investment itself. Project design had focused on addressing demand side constraints, with the expectation of catalyst effects on construction sector output and job creation. 11. As for the PDO outcomes, the above-mentioned activities and outputs would translate to the PDO outcomes by creating a deepened primary and secondary mortgage market in Nigeria. The strengthened capacity and structure within PMLs intended to encourage more efficient issuance of mortgages, especially mortgages below N5 million. Increasing the number of mortgages refinanced by NMRC intended to encourage development of mortgage products with better loan terms (lower interest rates and longer tenors), by reducing liquidity risk. The creation of mortgage guarantee and HMF product would diversify the mortgage instruments introduced to the market, as well as further address credit risk for lower income groups. TA and capacity building provided to regulators and other relevant ministries intended to improve their supervision capacity, which would further strengthen the legal and regulatory environment. 12. In the long-run, the project intended to relieve both the demand and supply side impediments facing the housing market in Nigeria. On the demand side, the support to NMRC refinancing and the mortgage guarantee product intended to help increase access to finance by extending housing finance products to lower income groups, hence further deepening the mortgage market. The piloting of HMF product would introduce housing-related loans to micro borrowers (e.g. home improvement, incremental construction). On the supply side, the market would respond to the demand side increases by providing more housing units with the more affordable price range. The capacity building on the legal and regulatory environment intended to help improve the land and legal system in general, address the capacity issue in the current housing value chain, lower the cost of housing, create jobs, as well as create a transparent housing ecosystem in the market. Project Development Objectives (PDOs) 13. As per the Loan Agreement (Credit Nr. 5292-NG) dated December 18, 2013, the Project Development Objectives were: To increase access to housing finance by deepening the primary and secondary mortgage markets in the Federal Republic of Nigeria. 18 Please see Annex 8 for the description and backgrounds of NMRC. 19 Supply side refer to interventions that could affect affordable housing supply, including land administration, serviced infrastructure, developer capacity, construction materials, etc. 20 Demand side refer to interventions that could boost demand for affordable housing, including reduced mortgage interest rate, increased mortgage tenor and alternative mortgage products to enable purchase of affordable housing. Page 10 The World Bank Housing Finance Development Program (P131973) Key Expected Outcomes and Outcome Indicators 14. PDO Level Indicators: - Number of new mortgage loans provided - Number of new mortgages below N5 million - Proportion of mortgage debt outstanding refinanced by NMRC 15. Intermediate Level Indicators: - NMRC efficiency indicator – cost to income ratio - NMRC new bond issuance volume (year 3-5 are without sovereign guarantee) - Number of guarantees - 3-5 pilot MFBs develop/adapt/adopt housing microfinance loan products, access project credit line, and establish first cycle of microfinance lending programs - Number of HMF loans disbursed - Volume of HMF loans disbursed - Portfolio at risk for HMF pilot Components 16. The original components at project approval are summarized below: Part 1. Establishment of the Mortgage Refinance Company (US$250 million) 17. A line of credit of US$250 million would be disbursed to NMRC as tier 2 capital, which intended to help the set-up and operationalization of NMRC. NMRC will act as a bridge between mortgage lenders and the capital markets. It will issue standard corporate bonds into the capital markets and subsequently issue loans to mortgage lending institutions. Part 2. Establishment of a Mortgage Guarantee Product, Targeted at Lower Income Borrowers (US$25 million) 18. US$25 million equivalent intended to develop a guarantee product to bear some of the credit risk for lower income lenders. The guarantee product will provide credit default loss protection to mortgage lenders. The project will fund the product’s development costs, including testing and financial modeling, and launching a competitive process to select a mortgage guarantee agent to implement the project. Part 3. Housing Microfinance (US$15 million) 19. US$15 million intended to support the development and piloting of new or emerging formal HMF products, and to demonstrate a sustainable business case for these activities. The project will fund: a) tailored TA to selected MFBs to develop new and/or improve emerging HMF products; b) risk capital provided to MFBs in the form of limited soft loans intended to test the new HMF products; and c) a line of credit to MFBs to roll out the tested HMF products. Page 11 The World Bank Housing Finance Development Program (P131973) Part 4. Technical Assistance and Capacity Building (US$10 million) 20. US$10 million intended to support a technical assistance and capacity building program comprising following activities: a) capacity building support on housing mortgage products, including training on underwriting, collateral valuation, mortgage risk management, housing finance policy and delinquent loans management; b) collection, collation and dissemination of housing market information, including development of a housing price index system and mortgage market database; c) provision of technical assistance to NMRC on bond issuance; d) mortgage consumer protection and financial literacy; e) creation of the Mortgage Guarantee Scheme including feasibility study and implementation; f) housing microfinance product development, piloting, Housing Microfinance Provider staff training, consumer protection, impact evaluation, and other related ancillary support services necessary for establishing sustainable housing microfinance operations; and g) strengthening legal and regulatory environment underpinning the housing financing sector. B. SIGNIFICANT CHANGES DURING IMPLEMENTATION (IF APPLICABLE) Revised PDOs and Outcome Targets 21. PDO and outcome targets were not revised Revised PDO Indicators 22. PDO indicators were not revised Revised Components 23. Components were not revised. The components costs changed after the second Level II Restructuring with partial IDA cancellation. The revised component amounts are shown in Table 1 below: Table 1: Revised component amounts 21 Component Amount (before Amount (after cancellation, cancellation, USD) USD) 1. NMRC $250,000,000 $165,697,791 2. Mortgage Guarantee $25,000,000 $0 3. HMF Loan $15,000,000 $7,382,539 21Amounts calculated using the SDR USD exchange rate on Dec 17, 2018 (The date the Bank received the request for cancellation). Page 12 The World Bank Housing Finance Development Program (P131973) 4. TA and Project $10,000,000 $1,549,813 Management Total $300,000,000 $174,630,142 Other Changes 24. A Level II Project Restructuring was carried out during the project implementation phase. The restructuring revised the DLI targets in the remaining tranches (III, IV and V). In addition, the restructuring adjusted the amount of the DLIs for these tranches (shown in Table 2 below). Finally, as the original credit was committed in SDR, the restructuring made this correction as well, by transforming the maximum amount allocated to SDR (shown in table 3 below). Table 2: Revised DLI Amounts and Targets Tranche Maximum amount allocated DLI target: value of new bonds (US$ millions equivalent 22) and refinanced mortgages (Naira millions) III 50 [instead of 70] 18,930 IV 30 [instead of 20] 15,775 V 30 [instead of 20] 7,888 VI 20 7,888 Table 3: Revised DLI Amounts in SDR and Targets Tranche Maximum amount DLI target: value of new bonds and allocated refinanced mortgages SDR 23 (Naira m.) I 12,952,950 Compliance with Conditions of Effectiveness under Section 5.01 of the Financing Agreement II 71,083,310 Evidence that NMRC has carried out Due Diligence for refinancing or pre- financing transactions of mortgage loans of Participating Mortgage Lenders III 34,500,000 18,930 IV 20,800,000 15,775 V 20,800,000 7,888 VI 6,163,740 7,888 22 At project design, maximum amount allocated was only available in USD. 23 At project restructuring, maximum amount allocated was switched to SDR only. Page 13 The World Bank Housing Finance Development Program (P131973) Note: DLIs I and II have already been met and disbursed; the only change will include presenting the maximum amount allocated in SDR. 25. In December 2018, the Borrower had requested cancellation of undisbursed balance of SDR 73,202,610.03. Therefore, a second Level II Restructuring was done for the partial cancellation. Rationale for Changes and Their Implication on the Original Theory of Change First Level II Restructuring: 26. The DLI targets were expressed in dollars at the project design phase. However, the Naira had devaluated since project approval (from US$1 = N157.75 in June 2013 to US$1 = N305.7. official rate and approximately N360 in the convertible Investors and Exporters Forex (IEFX) window, which was at about 20 percent premium over the official rate). The restructuring brought the remaining tranches of DLI targets (tranches III, IV and V) to Naira, using the prevailing exchange rate at the time of project approval, to ensure that the DLI targets were not adversely affected by exchange rate fluctuations, which are considered exogenous factors beyond the control of the project. 27. In addition, the restructuring adjusted the amount of the DLIs for tranches III, IV and V based on concern that NMRC would receive a significantly higher amount than originally planned: US$70 million at the current official exchange rate (305) was higher than US$70 million at the historical exchange rate (157). The reduction on the NMRC component was requested by the Project Administration Team (PAT) 24 explicitly ahead of the project restructuring, although the Bank at the time had advised to leave the allocations unchanged as these resources were planned for NMRC by design and the purpose of the restructuring was to address the exchange rate fluctuations which had adversely affected the project targets. 28. Finally, as the original credit was committed in SDR, the restructuring made this correction as well, by transforming the maximum amount allocated to SDR. The DLI conversion was guided by the authorities’ request provided in US dollars as per the Table 1 above. Although the actual amount disbursed in dollars would depend on the SDR conversion rate at the time of disbursement, the amount would not exceed a total SDR allocation of SDR166,300,000, of which SDR12,952,950 and SDR71,083,310 had already been disbursed (upon achievement of DLIs I and II). The DLIs III- VI had been recalculated as of April 05, 2018 and expressed in SDR in Table 2 above. The restructuring required slight amendments to the disbursement letter. Second Level II Restructuring: 29. The project cancelled IDA funds in the amount of SDR 73,202,607 in December 2018, as any undisbursed amount would otherwise be lost to the Nigeria IDA envelope. Through the prompt cancellation of the funds before project closure, these funds could be recycled into the Nigeria IDA envelope to support other development initiatives. 24 Please see Annex 8 under Project Implementation Arrangements the description and backgrounds of PAT. Page 14 The World Bank Housing Finance Development Program (P131973) 30. These changes did not affect the original theory of change. II. OUTCOME A. RELEVANCE OF PDOs Assessment of Relevance of PDOs and Rating Rating: High 31. The project’s objectives remain highly relevant. 32. At Project Appraisal, the project’s objectives were in line with government’s Financial System Strategy (FSS) 2020, which aimed to transform Nigeria into the financial hub of Africa by becoming the safest and fastest growing financial system among emerging market countries. The project would contribute to these objectives by building a secondary mortgage market that addresses the maturity mismatch and access to long-term funding. The objectives were also aligned with the focus of the national "Transformation Agenda", which prioritized support to non-oil sectors that have strong job creation and multiplier potential, such as housing and agriculture. The increased availability of funds to the mortgage market would stimulate demand for house improvements and housing-related expenditure and support job creation in the housing market value chain. Finally, the project objectives were also in line with the World Bank Group’s Nigeria FY10- FY13 Country Partnership Strategy (CPS), especially related to its second theme: “Maintaining non-oil growth through the diversification within financial markets to new product lines such as trade finance, housing finance, insurance, small and medium enterprise finance and increased power generation and distribution, developing local fixed income capital markets and credit bureaus”. 33. At Project Completion, Nigeria still lacks a substantial mortgage market which is considered critical component of housing market development, therefore making the project objectives still highly relevant despite limited progress to this end. The project’s objectives are aligned with several recently released country strategy documents. A key outcome indicator of the FY14-FY17 Country Partnership Strategy (CPS) is the percentage of housing finance loans in the financial sector increased to 5 percent of the total loans. The CPS specifically identifies the Housing Finance Project supported by IDA and IFC as key contributor towards increasing the supply of long-term finance to the Nigerian housing market and enhancing access to finance for lower-income borrowers. The housing finance agenda also supports the current strategies of the National Economic Recovery and Growth Plan (ERGP) for the period 2017-2020. ERGP’s strategies include accelerating non-oil revenue generation, stimulate construction by building affordable housing, leverage private capital for housing, as well as boost job creation through constructing housing units. Finally, the project objectives are consistent with the WBG’s ambition to support structural reform agenda for diversified growth and jobs focuses specially on increasing the supply of long-term financing for the housing sector (given that the mortgage market remains Page 15 The World Bank Housing Finance Development Program (P131973) underdeveloped), expanding financing opportunities for SMEs, and strengthening the ability of Development Finance Institutions to mobilize private finance for key sectors of the economy. B. ACHIEVEMENT OF PDOs (EFFICACY) 34. The project did not have an updated results framework at project closing; the most recently available information was from 2017. The results indicators were submitted in April 2019, four months after project closure. Based on the indicators received, the project achieved one (1) of its three (3) PDO-level indicators. The indicator “Number of new mortgage loans provided” was underachieved by 70 percent. The indicator “Number of new mortgages below N5 million” was overachieved by 38 percent. The indicator “Proportion of mortgage debt outstanding refinanced by NMRC” was under-achieved by 35 percent. The list of the project’s PDO level indicators and achievement status is recorded in Table 4 below. 35. The Project achieved two (2) out of the seven (7) intermediate level indicators. The project achieved the indicator measuring the percentage of the housing microfinance pilot portfolio reported at risk: only 0.06 percent of the housing microfinance portfolios were reported at risk following project closure (compared to the upper limit of 2.5 percent). The project also achieved the indicator measuring the number of MFBs that participated in the HMF pilot: 7 MFBs participated in the HMF pilot program (compared to the target of 3-5 MFBs). Other intermediate level indicators were not met partly due to gaps in project administration. Importantly, the initial assessment of the HMF pilot indicated use of funds for ineligible activities by one MFB, none of which were addressed prior to project closing, ultimately resulting in declaration of funds disbursed by this MFB as ineligible for financing under the project. Following the pilot, since none of the technical, eligibility and fiduciary issues identified and recorded in the evaluation report had been addressed (see Annex 5), no subsequent HMF loans were disbursed after the pilot. The project did not achieve these indicators: the number and volume of HMF loans disbursed were 69 and 78 percent below the target respectively. The list of the project’s intermediate level indicators and achievement status is recorded in Table 4 below. 36. A summary of the achievement of PDO and intermediate level indicators is provided in Table 4. Page 16 The World Bank Housing Finance Development Program (P131973) Table 4: Summaries of Indicators and Achievement Indicators 25 Unit of Measure End End Achievement Target Actual Value Value PDO Level Number of new mortgage loans Number 50,000 14,978 -70% provided Number of new mortgages below Number 7,500 10,348 38% N5 million Proportion of mortgage debt Percentage 20 13 -35% outstanding refinanced by NMRC Intermediate Level Portfolio at risk for housing Percentage 2.5 0.06 98% microfinance (HMR) Pilot NMRC Efficiency Indicator – Cost Percentage 42 55 -31% to income Ratio NMRC New Bond issuance volume Number 50 19 -62% (year 3-5 are without guarantee) Billion Nigerian Naira Number of Guarantees Number 900 0 -100% 3-5 pilot of MFBs Number 5 7 29% develop/adapt/adopt housing finance loan products, access project credit line and establish first cycle of micro financing lending programs Number of HMF loans disbursed Number 2,860 889 -69% Volume (USD) of HFM loans Amount (Million 8 1.8 26 -78% disbursed USD) The next paragraphs assess the achievement of each expected outcome of the Project: i) Deepen primary mortgage markets in Nigeria; ii) Deepen secondary mortgage markets in Nigeria; and iii) Increase access to housing finance in Nigeria. 25 Results are submitted by PAT on April 17, 2017 and updated in April 2019. 26 Ibid, using exchange rate US1 = N305.8 Page 17 The World Bank Housing Finance Development Program (P131973) i) Deepen primary mortgage markets in Nigeria 37. The Primary mortgage market is usually defined as “The market where borrowers and mortgage originators come together to negotiate terms and effectuate mortgage transaction27.” For the purpose of this ICR, the deepening of the primary mortgage markets would primarily be assessed by: i) The number and volume of mortgages originated by the PMLs, including loans below N5 million; ii) The mortgage interest rate and tenor and their impact on affordability; iii) Results generated from focus group studies on NMRC interventions; iv) Percentage of primary mortgage lenders (PMLs) that became NMRC member banks; and v) Development of alternative mortgage financial products. 38. The number and volume of mortgages originated by the market: As shown in Figure 2 below, the Nigerian residential mortgage market size increased from N166.4 billion in 2013 to N206.6 billion in 2018 28, a growth of 24 percent. During the project period, the Primary Mortgage Lenders (PMLs) generated 14,978 new mortgages totaling N111 billion, which accounts for above 50 percent of the entire mortgage market. Average loan size was around N7.7 million (US$21,000). In terms of volume, only 7 percent of these new mortgages generated are for loans below N5 million, with average loan size of N772,192 (US$2,000). Table 5 below gives an overview of the accumulated number and volume of mortgages (including mortgages below N5 million) by year. Figure 2. Size of Nigerian Mortgage Market, 2013-2018 29 Page 18 The World Bank Housing Finance Development Program (P131973) Table 5: Accumulated Number and Volume of New Mortgages Facilitated by the Project, 2014-2018 30 2014 2015 2016 2017 2018 Total No. of new mortgages created 1,195 2,083 3,045 4,102 4,553 14,978 No. of new mortgages created (below N5 million) 1,572 1,554 1,990 2,948 2,325 10,389 31 Volume of new mortgages created (Billion Naira) 11.5 10.4 11.6 39.6 38.0 111.1 Volume of new mortgages created below N5 million (Naira) 3.0 1.0 1.4 1.4 1.3 8.1 39. There are currently 54 PMLs in Nigeria including 20 commercial banks and 34 Primary Mortgage Banks (PMBs). Only 5 commercial banks and 20 PMBs had joined NMRC as member banks. As of project closure, the total number of mortgages generated by the 25 member banks stood at 4,759 32, totaling N57 billion (51 percent of the new mortgages generated by the market, excluding commercial bank data from 2014-2016). Among these, about 3,043 mortgages totaling N4.7 billion were below N5 million (58 percent of the total new mortgages below N5 million generated by the market). 40. Among NMRC’s 25 member PMLs, however, only 4 commercial banks and 6 PMBs had done refinancing with NMRC during the project implementation period. These 10 PMLs had generated 1,580 mortgages throughout the project implementation period with a total volume of N39.4 billion (35 percent of the total new mortgage generated from the market). Among these, only 347 mortgages 33, or about 2 percent, were loans below N5 million. In other words, NMRC’s market share is approximately 35 percent of the new mortgages generated, or 19 percent of the entire Nigerian residential market. 41. Table 6 below shows the total number and volume of new mortgages generated by these 10 PMLs, together with the average loan size for each PML. The table demonstrates that the average loan size provided by these PMLS (N24million, or US$81,470) was much higher than the average loan size provided by the market (around N7.5 million, or US$24,474). Based on the analysis of the current interest rate, tenor and affordability index (illustrated by Table 7 in the 27 Investopedia 28 Data provided by the Project Administration Team, OFSID, CBN. These figures only include mortgages generated by the Primary Mortgage Banks, plus the Residential Mortgage Loans and Home Equity generated by Deposit Monetary Banks. 29 Data submitted by PAT in April 2019de 30 Year 2014-2016 only contain data from primary mortgage banks + 5 commercial banks including Stanbic IBTC, Sterling, Access, Heritage and SunTrust. 31 Note that the total no. of new mortgages below N million provided by NMRC (10,389) slightly different from the figure provided by the PAT (10,348). However, PAT did not provide breakdowns by year. Therefore, we are using breakdown figures provided by NMRC given that the difference in total is not substantial. 32 The number was based on 15 member PMBs and 5 commercial banks, as 5 member PMBs did not provide relevant data. 33 One Bank did not provide information on number of mortgages below N5 million, but there is an outstanding loan amount for loans below this threshold. Page 19 The World Bank Housing Finance Development Program (P131973) next section), these loans cater more to the population at top of the income pyramid based on the annual household income data in Figure 5. Table 6. Number, volume and average mortgage size generated by PMLs actively doing refinancing with NMRC No. of Average Loans size Average Loan size USD PML Loan Amount Mortgages (Naira) (US$1=N306.45 Commercial Bank 1 11,592,748,730 113 102,590,697 334,771 34 Commercial Bank 2 3,977,492,575 201 19,788,520 64,573 Commercial Bank 3 748,922,020 26 28,804,693 93,995 Commercial Bank 4 1,361,001,031 98 13,887,766 45,318 Primary Mortgage Bank 1 6,347,310,947 179 35,459,838 115,712 Primary Mortgage Bank 2 4,719,961,997 197 23,959,198 78,183 Primary Mortgage Bank 3 2,590,000,000 113 22,920,354 74,793 Primary Mortgage Bank 4 3,868,099,744 290 13,338,275 43,525 Primary Mortgage Bank 5 2,760,000,000 166 16,626,506 54,255 Primary Mortgage Bank 6 1,481,669,200 197 7,521,163 24,543 Total 39,447,206,244 1,580.00 24,966,586 81,470 42. The refinancing operations of NMRC appeared to have facilitated extension of tenors and reduction in mortgage cost, as reflected by movements on the interest rates and tenor of mortgages throughout the project phase. Before the project, access to housing finance was out of reach for most Nigerians due to unfavorable conditions and terms of the loans. Average lending rates were between 20 and 26 percent. The maturity varied from 4.5 to 15 years. Only subsidized loans obtained through the NHF scheme had a maturity of 30 years and an interest rate of 6 percent (although the maximum loan amount was N15 million). The operationalization of NMRC helped PMLs to provide long-term funds to the market, which encouraged lenders to lend more at lower rates and longer tenor by supplying liquidity. Average mortgage interest rates were reduced from 26 percent in 2015 to 18 percent in 2018 35. Average mortgage tenor, on the other hand, increased from 4.5 years in 2015 to 20 years in 2018 (as shown in Figure 4). Both increase in mortgage tenor and decrease in mortgage interest rates will have a positive impact on mortgage affordability, as explained in the next paragraph. Figure 4: Movements in mortgage interest rates and tenor from 2015 to 2018 34 Commercial Bank 1 is a tier 1 bank while the other three commercial banks are tier 3 Banks. Therefore, commercial bank 1 has a much larger balance sheet and loan portfolio than the other three banks. 35 Data provided by NMRC and through focus group studies Page 20 The World Bank Housing Finance Development Program (P131973) Movements in Mortgage Rate and Tenor 25 30% 26% 24% 22% 25% 20 20 20 20% 15 18% 15% 10 10% 5 7 5% 4.5 0 0% 2015 2016 2017 2018 Average Mortgage Tenor (years) Average Mortgage Rate 43. The improvement in mortgage terms (increase in maturity and decrease in interest rate) will enhance mortgage affordability. As illustrated in Table 7 below, for a given loan amount, an increase in the maturity from 120 to 240 months 36 (column one and two), and a decrease of interest rate from 26 percent to 18 percent, would decrease the monthly mortgage payment by N160,498, and reduce the qualifying monthly income by N534,995. The decrease in monthly payments (and reduced qualifying income) increases the number of potential households who can afford loans by approximately 1 million (see Figure 5 below). While the typical household who could afford the mortgage product at this stage is still at the top of the income pyramid, the movements in interest rates and tenor is helping mortgage affordability to reach lower income groups. Table 7 Mortgage affordability calculations – Different maturity Maturity, months 120 240 Loan amount, N N 20,000,000 N 20,000,000 Interest, per annum 26% 18% Monthly Payment, Naira N 469,161 N 308,662 Monthly Income Required in Naira N 1,563,869 N 1,028,874 Monthly Income Required in USD (1 USD = 306.45 N) $3,153,135 $2,771,781 Annual Income Required in USD $37,837,625 $33,261,373 Figure 5 Nigeria Annual Household Income US$ 36 Data provided by NMRC Page 21 The World Bank Housing Finance Development Program (P131973) Source: http://housingfinanceafrica.org/countries/nigeria/ 44. Results of focus group discussion - the impact on NMRC interventions: Focus group sessions with 4 Primary Mortgage Banks (PMBs) and 10 PMB beneficiaries were conducted during the ICR mission. It is important to note that the purpose of the focus group discussion was aimed at recording general impressions about the project. The beneficiaries interviewed all had existing mortgages with their respective PMBs, and they were notified by the banks that their mortgages were eligible for refinancing. Almost all PMB mortgage beneficiaries reported decreased rates and increased tenors at refinancing compared to their original mortgage. Beneficiaries also credited their PMBs for improved levels of transparency and literacy, as PMBs had launched more advocacy activities since they engage the refinancing activities with NMRC to ensure that their mortgage products were known and well understood to the mortgage borrowers. Beneficiaries also perceived that the processes of getting mortgages had been simplified and time to process mortgage had been reduced. PMBs found the Uniform Underwriting Standards (UUS) very valuable as it set an industry standard and provided PMBs with the underwriting tool kit, which increased their loan portfolio quality. PMBs also shared that the Refinancing provided by the Project had improved their liquidity situation. Most PMBs were able to develop new incentives to offer their loan officers after joining the refinancing scheme. Some PMBs managed to expand customer base since the launch of Refinancing. Details of the focus group sessions are included in Annex 6. 45. Development of alternative mortgage financial products: The project aimed to further support reach of mortgages down the income distribution by developing a mortgage guarantee product, as well as housing microfinance products. The mortgage guarantee would help otherwise qualified borrowers who had not accumulated enough cash to meet current down payment requirements, as well as bankable households with variable or self-employed income to successfully apply for mortgage loans. The housing microfinance products would enable the MFBs in the market to offer appropriate instruments to small loan applicants for home improvement and incremental construction. The development of the guarantee product would facilitate deepening of the primary mortgage market, while HMF would facilitate Page 22 The World Bank Housing Finance Development Program (P131973) financing for low income borrowers. Unfortunately, the development of these products experienced many challenges and did not achieve the expected results. 46. Mortgage Guarantee: The design of the mortgage guarantee component experienced numerous problems and delays. Since the start of the project, there were delays with commencement of this activity and subsequently in procuring the consultant to design the guarantee product. There were also issues with the first consultant who failed to deliver satisfactory outputs and had to be replaced by a consulting firm. The consulting firm made its first trip to Nigeria and conducted the assessment in June 2018. Since then there had been some significant progress, including: i) the Business and Investment plan for the Mortgage Guarantee Company developed and discussed with key stakeholders; ii) Draft Regulatory and Supervisory Framework for the Nigeria Mortgage Guarantee Company (NMGC) prepared by the CBN; and iii) the Board of the NMRC approved investment into the NMGC (for NMRC to be the anchor investor with planned N1.25 billion, or 25 percent equity stakes in the NMGC). Despite these achievements, the key output for the component, namely the operationalization of the mortgage guarantee product, remains to be completed. NMGC’s capital base also requires strengthening to offer highly rated services as well as to comply with prudential requirements. 47. Housing Microfinance: The HMF component was managed by the PAT. At project closure, seven MFBs had received TA services provided by two international consulting firms. TA services included training provided to MFB employees on marketing, technical, risk management and loan monitoring, those necessary to administer the HMF products. Following the belated implementation of training and as a result of initial delays with project preparation, the first pilot tranche of US$1.82 million was not disbursed to these MFBs until April to June 2018. An assessment on the utilization of the first tranche of the HMF loan was conducted as a condition for disbursement of the second tranche. The assessment report was shared by the PAT with the Bank in October 2018, less than three months before project closure. The assessment had concluded that 4 MFBs had fully utilized their first tranche loan; 2 MFBs were yet to meet 100 percent utilization of their first tranche loan at the time; and 1 MFB had ineligible expenditure and substantial governance issues. The assessment also identified several irregularities in the administration of the first tranche, including inconsistencies in loan agreements and incomplete E&S policy for some MFBs. Further, there was no identified assessment of the portfolio quality of the disbursed funds. Based on the assessment, the Bank task team called for remedial action of identified issues as well as additional information on the utilization of funds and compliance with project requirements (e.g. verification of MFB eligibility criteria, implementation of loan covenants). Only once these conditions were met would the Bank task team consider issuing a No Objection for disbursement of the second tranche (the list of additional information requested by the Bank is disclosed in Annex 5. Since then, the Bank repeatedly requested this information from PAT, including as formalized in Aide Memoires. However, this information was not submitted prior to project closure. Page 23 The World Bank Housing Finance Development Program (P131973) 48. Focus group sessions conducted with 6 MFBs and 18 MFB beneficiaries during the ICR mission highlighted positive views regarding the TA service as well as the first tranche disbursed. It is important to note, however, findings from focus group sessions are not sufficient to determine the success of the pilot in absence of a thorough independent evaluation, especially considering the issues recorded in the initial pilot evaluation. Most pilot beneficiaries confirmed that the pilot enabled them to get necessary funds to construct or renovate their business towards increasing revenues and productivity, and that the pilot had enabled them to get loans with lower interest rate, longer tenor, bigger loans and reduced loan processing period than previously available. Most MFB beneficiaries also reported to notice positive changes in MFB’s business strategy and operations since the pilot, such as increased efficiency in interaction with clients, expanded eligibility of customer base, increased proactivity and advocacy, more flexibility in their product offerings. Despite substantial interest from MFBs and beneficiaries, the next tranches of disbursements never occurred given that the issues identified in the initial evaluation report had not been addressed. ii) The Deepening of Secondary Mortgage Market in Nigeria 49. For the purposes of this ICR, the secondary mortgage market will mainly be assessed by: i) Bonds issued by NMRC, ii) Mortgage loans refinanced by NMRC, iii) Number and percentage of PMLs doing refinancing with NMRC. 50. NMRC Bonds issuance: At project design, it was envisioned that NMRC would issue N50 billion in bonds. However, at project closing, NMRC had issued two series of bonds total offing just N19 billion. The first series was issued in July 2015 for N8 billion with yield of 14.9 percent. The second series was issued in June 2018 for N11 billion with yield of 13.8 percent. Both bonds were guaranteed by FGN and both issuances were fully deployed towards refinancing legacy mortgage loan portfolios of the eligible member PMBs. 51. The target of N50 billion bonds issuance was not reached, largely due to the unfavorable economic situation since 2014 and the government’s monetary policy response to the economic situation. As shown in Figure 6 below, Nigeria tightened its monetary policy by increasing the monetary policy rate (MPR) from 12 percent in 2013 to 14 percent in 2016 37 (and has remained at that rate since then). The unfavorable macroeconomic conditions, including a high MPR, and resulting high interest rate environment has made it financially unviable for NMRC to issue bonds. Further, the combination of high T-bill rates and low savings & deposit rates between 2013 and 2018 38 (average of 10.7 percent and 3.5 percent respectively) encouraged investors to put their money in T-bills instead of banks, resulting in liquidity challenges and high costs of funding for banks. Interbank call rate surged from 11 percent in 2013 to 23 percent in 2017 39. Overall, it was not an enticing environment for bond issuance and hence the underachievement in this target within the project period. Despite the unfavorable macroeconomic conditions, NMRC was able to issue two series of bonds, linking the capital markets with housing finance products and established the viability of a secondary mortgage company/provider to support the primary mortgage market in the long-run. Page 24 The World Bank Housing Finance Development Program (P131973) Figure 6 Nigeria Money Market Indicators, 2006-2018 Data source: CBN 52. Mortgage loans refinanced 40 by NMRC: At project closure, a total of 1,036 mortgages with a volume of N17.97 billion (about US$59 MUSD 41) had been refinanced, accounting for 95 percent of the bonds issued by NMRC. The refinancing volume accounted for 15 percent of total mortgage loans created by the market throughout the project period, slightly below the 20 percent target set at project design. The average amount of refinancing was about N17 million, which catered to the population at the top of the income pyramid. When NMRC started refinancing in 2015, the volume was low, partially because there were not enough legacy loans meeting the eligibility criteria for refinancing as well as the macroeconomic down turn contributing to decreased mortgage issuance in the market. The number of qualified mortgages increased with the introduction of the Uniform Underwriting Standards (UUS). Since the country emerged from recession in 2018, the total number of mortgage refinancing surged from 456 in 2017 to 1,036 in 2018. Despite the improvement, currently only 10 PMLs are doing refinancing with NMRC. NMRC needs to expand their membership portfolio and engage refinancing activities with more PMLs, especially with commercial banks to reach critical mass. The number and volume of refinanced mortgages from 2015-2018 is shown in Figure 4 below. 37 CBN 38 ibid 39 ibid 40 With NMRC refinancing, eligible mortgage loans assigned by MLBs will remain on their balance sheet as NMRC require them to continue servicing the mortgage loans as well as conducting proper and periodic due diligence to ensure their continued eligibility. 41 Using exchange rate of $1 USD = N305.8 Page 25 The World Bank Housing Finance Development Program (P131973) Figure 4. Number and Volume of Refinanced Mortgages, 2015-2018 Number and Volume of Refinanced Mortgages 2015-2018 refinanced mortgages (vol.) 1200 18.0 20 18 1000 16 refinanced mortgages (no.) 14 800 12 600 8.2 8.6 10 8 400 6 1.9 4 200 2 120 434 456 1,036 0 0 FY 2015 FY 2016 FY 2017 FY 2018 Accumulated No. of refinanced mortgages Accumulated Volume of refinanced mortgages (billion Naira) 53. Number and percentage of PML’s participation in refinancing: Throughout the project period, NMRC had provided refinancing for four commercial banks (20 percent of total commercial banks), and six Primary Mortgage Banks (17 percent of total PMBs). The percentage of PMB participation is considered low mainly due to several reasons: first, NMRC only received its operating license in February 2015, two years after project approval, and the bonds issuance and refinancing program have only been marketed for several years. Second, NMRC operates under a club structure. Only PMLs that contribute a minimum of N50 million to become a club member can qualify for NMRC’s refinancing scheme. This had limited some PML’s ability and willingness to join NMRC’s refinancing scheme. Thirdly, not all PMBs were interested in getting refinancing from NMRC. Some were able to get liquidity support from other sources and therefore were not interested in the refinancing. Among the PMBs that were willing to do refinancing, some did not meet the eligibility criteria, as they were financially unviable, either in terms of profitability of capital adequacy. Throughout the project period, NMRC contributed a significant amount of advocacy resources around their products towards helping PMBs understand the importance of meeting high quality portfolio standards. NMRC also worked with some PMBs to improve their financial soundness to qualify for the refinancing. With PMBs that were uncomfortable large-scale refinancing, NMRC offered the option of pilot transactions intended to grow the refinancing portfolio gradually. NMRC’s refinancing scheme did succeed in recruiting the participation of numerous PMLs to provide mortgage finance, which enhanced competition in the mortgage market and led to an improvement in the quality of services (i.e. more aggressive market campaign on the new products, better client service, more flexible loan terms, etc.). Iii) Increase Access to Housing Finance in Nigeria 54. For the purposes of this ICR, the increase access to housing finance in Nigeria will be assessed by: i) Mortgage Page 26 The World Bank Housing Finance Development Program (P131973) debt to GDP, ii) Home loan Penetration, and iii) Efforts to create enabling environment for an efficient mortgage market. 55. Mortgage debt as percent of GDP: As of 2016, the mortgage debt accounted for 0.35 percent of GDP and had decreased from 0.58 percent in 2011. 42 Mortgage as percent of GDP was very low compared to other SSA countries such as South Africa (22.54 percent in 2014) and Kenya (3.53 percent in 2013), not to mention countries outside of SSA such as Malaysia (33 percent) and India 9 percent. 56. Home Loan Penetration: The percentage of adults with an outstanding housing loan in Nigeria stood at 2.6 percent in 2017, decreased from 6.3 percent in 2014 43. Home loan penetration for its SSA peers in 2017 were: 8.4percent for Kenya, 7.5 percent for Ghana, 5.9 percent for Cameroon and 5.7 percent for South Africa. This suggests slow growth and lack of depth in the Nigerian mortgage market. 57. Efforts to create enabling environment for an efficient mortgage market. Despite decreased mortgage debt to GDP and home loan penetration, efforts were made throughout the project period to create an enabling environment for an efficient mortgage market. These include the development of Uniform Underwriting Standards (UUS), Model Mortgage Foreclosure Law, as well as the Housing Market System (HMS). 58. The development of the Uniform Underwriting Standards (UUS) for both formal and informal sectors was one of the major achievements under the project. The UUS was developed in collaboration with the CBN, NMRC, Mortgage Bankers Association of Nigeria and other stakeholders. The UUS aims to promote efficiency and mitigate the legal and operational risks inherent in mortgage lending by ensuring quality collateral, adequate property title, proper registration, enforcement of legal mortgages, and maintenance of efficient collection process. Such standards would balance the requirements of responsible finance with lenders’ needs to enforce loan contracts, as well as enhance consumer confidence in the mortgage products. It would also ensure homogenous data to be collected across new originations which will help in developing credit risk management metrics as the sector evolves. The UUS was developed in 2014. Up to date, all NMRC’s member banks had now conformed to the UUS. Table 8 below shows the status of the mortgage loan NPLs for PMLs as of today. Out of the 24 PMLs that provided the data, 16 of them had NPLs below 10 percent and most of them are NMRC’s member banks who had adopted the UUS. Only one NMRC’s member bank had mortgage NPL above 30 percent. By bringing more PMLs to adopt to the UUS, it will help improving their mortgage loan quality and help these institutions achieve financial sustainability. Table 8: Current Mortgage Loans NPLs for Primary Mortgage Lenders (PMLs) NPL Level No. of PMLs NPL < 10% 16 NPL 10%-30% 3 NPL 30%-50% 4 42 NMRC, 2016 data 43 Global Findex, 2017 Page 27 The World Bank Housing Finance Development Program (P131973) NPL > 50% 1 No info on NPL 30 Total PMLs 54 59. Further, NMRC also developed the Model Mortgage Foreclosure Law. The Law intended to help strengthen the regulatory framework for mortgage finance by fast-tracking the process for creating legal mortgages (including reducing the cost and processes for obtaining the relevant consents and registration), timely resolution of disputes arising out of mortgage transactions, consumer protection and creating an efficient foreclosure process. Through this initiative, states are encouraged to: lower the cost of land and mortgage transfers from about 30 percent to 5 percent; to set a standard for completion of foreclosures within 180 days; and for courts to set up specialized mortgage courts to handle these transactions. Since 2015, NMRC has executed Memorandum of Understanding (MoU) with the governments of Kogi, Cross River, Delta, Kaduna, Lagos and Ogun States on the adoption and passage of the draft law to institute the enabling environment for mortgage creation and investment in housing. Negotiations on the passage of the law are at advanced stages with the governments of Akwa Ibom, Edo, Enugu, Kwara, Osun and Rivers states. In July 2017 Kaduna state became the first in the federation to pass the law. The project also collaborated with the CBN and hosted a workshop for Judicial Officers from across the country on foreclosure and related matters, which also provided opportunities for housing sector operators to interact with judicial officers on the need for a clear and streamlined foreclosure process that will attract investments into the housing sector. 44 60. A Housing Market System (HMS), which included the Housing Information Portal (HMIP) and the Mortgage Market System (MMS), was also developed by NMRC (although outside the scope of the project). The HMIP aims to capture demand-side baseline data on housing market in Nigeria (i.e. off-taker registration, real estate brokerage home- matching, pre-sales, etc.). The MMS aims to capture supply-side data on the housing market. It provides a standard operating platform for creating and managing the company’s mortgage portfolios from the stage of origination of the mortgages by the PMBs to the point of refinancing by NMRC, and until the loans are fully repaid. The HMS infrastructure aims to integrate operations of all stakeholders along the housing value chain and enhance the efficiency of operations of the housing (mortgage) sector. It would also provide standardization, simplification of processes with improved transparency and accountability in the mortgage and overall housing sector. The HMS was launched in July 2018 and available information was limited at project close. 61. Structural change in institutional supervision and corporate governance: At the time of project design, there were five existing state owned DFIs in Nigeria that were sector specific, not subject to prudential or internationally recognized governance standards. These DFIs lacked political independence and boards, and their CEOs were appointed by the various line ministries. The activities of these DFIs often deviated from targeted market failures, instead being used to bail out or rehabilitate ailing corporations. The previous Financial Sector Assessment Program (FSAP) pointed out these issues and recommended to set up proper corporate governance structure and institutional supervision for these types of institutions. Taking these recommendations into consideration, the NMRC was set up as a public private partnership 44 NMRC Annual Report 2017 Page 28 The World Bank Housing Finance Development Program (P131973) (PPP) structure, with the Ministry of Finance Incorporated (MoFI) and the Nigerian Sovereign Investment Authority (NSIA) holding 15.68 percent and 20.91 percent respectively. Five (5) commercial banks and twenty (20) primary mortgage banks represented 11.11 percent and 52.30 percent respectively of the paid-up shares. NMRC was licensed and supervised jointly by the Central Bank of Nigeria (CBN) and the Securities Exchange Commission (SEC). This was the first time a co- owned institution was operating with a public-private governance structure. The mix of investor groups was chosen to ensure that each group would bring a unique value addition to the institution. 62. The design of NMRC aimed to ensure that sound corporate governance practices were embedded in all areas of its operations. In this regard, NMRC emphasized strict adherence to the CBN Corporate Governance Code as revised in 2014, the SEC code of Corporate Governance 2011, the provisions of the Banks and Other Financial Institutions Act (BOFIA) 1991 as amended and other Institutional Best Practices and Standards on Corporate Governance. NMRC set up a Board of Directors to provide oversight of the internal control and risk management policies and practices of the company. The Board was supported in its oversight function by the various Board Committees. The Committee’s roles and responsibilities were set out in their respective charters that were reviewed and updated periodically to ensure relevance with contemporary industry operations. The charters also define the roles, responsibilities, scope of authority, composition and procedures for reporting to the Board. All these mentioned above were major structural changes and had contributed to NMRC’s resilience under the unfavorable macro environment and its ability to deliver sustainable value in the long run. Justification of Overall Efficacy Rating 63. The overall efficacy of the project is assessed as Negligible. 64. The project was instrumental in creating a majority private sector-owned liquidity facility that is self-sustaining in the long-run and with sound governance structure. However, the project only achieved three out of ten results indicators, and did not reach critical mass of deepening the primary and secondary mortgage market during the project implementation period, mainly due to inefficient project implementation compounded with unfavorable macroeconomic situation and a short timeframe to achieve an ambitious mandate. In terms of affordability, most mortgages associated with the project catered to the population at top of the income ladder, although there were downward movements in interest rates and upward movements in loan tenor among the PMLs who actively refinanced with NMRC 45. Improvements in the macro situation (e.g. elimination of market distortions such as mortgage interest rate cap) and increased number of PMLs joining NMRC’s refinancing scheme could bring the affordability of mortgages down the income ladder. The development for complimentary mortgage products was not successful due to procurement delays and other implementation challenges. In terms of the increase in access to housing finance, both mortgage to GDP and home loan penetration ratio had shrunk. Therefore, despite progress in building key mortgage market infrastructure and creating enabling environment for housing finance, the overall project efficacy is rated Negligible. 45 Data provided by NMRC Page 29 The World Bank Housing Finance Development Program (P131973) C. EFFICIENCY Assessment of Efficiency and Rating 65. The efficiency of the project is assessed as Modest. 66. Economic Analysis: An economic analysis (EA) was done at the appraisal stage. The EA model made key assumptions based on the macroeconomic situation at that time (summarized in Table 9). Based on these assumptions, the projected number of mortgage loans generated by end of 2018 would be 187,557. The additional mortgages generated by the project would be 81,000. The mortgage to GDP ratio would be 0.58 percent. 67. However, the macro situation changed over the course of the project. Inflation was fluctuating between 8 percent and 19 percent 68. rather than the originally projected 10 percent; the exchange rate surged from US$1/N158 to US$1/N306 due to the oil price shock. The CBN increased the MPR from 11 percent to 14 percent. The spill-over effect also led to slow growth in the mortgage market, which was 5 percent versus the originally projected 16 percent. The IDA resource shrunk due to Naira devaluation and the actual disbursement to the NMRC component was only SDR118,536,258 instead of the original SDR166,300,000. The team re-ran the economic analysis model at project closure (as shown in Table 10 below). Based on the actual inflation rate of 12 percent, MPR of 14 percent, exchange rate US1/N306, and mortgage market growth of 5 percent; the number of mortgages generated by 2018 would be 10,627, which is 94 percent below the original projection. The additional mortgages generated by the project would be 676 and the mortgage to GDP ratio would be 0.14 percent. Given the above, the actual number of mortgages generated by the project was 14,978 and the actual mortgage to GDP ratio was 0.35 percent at project closure, which exceeded the projection calculated based on the actual macro data. Table 9: Economic Analysis: Key assumptions and projections at design stage Assumed Inflation 10% Assumed MPR 11% Assumed Exchange Rate USD/Naira 158 Assumed Growth in Mortgage Market 16% Assumed IDA Funds Allocation to NMRC (SDR) 166,300,000 Assumed Mortgage Debt Outstanding in year 0 (million Naira) 300,000 Projected Number of Mortgage Loans based on the assumptions by end of 2018 187,557 Projected Mortgage Debt/GDP based on the assumptions by end of 2018 0.58% Page 30 The World Bank Housing Finance Development Program (P131973) Table 10: Revised Economic Analysis with real macro data at project completion Actual Inflation (average) 12% Actual MPR 14% Actual Exchange Rate USD/Naira 306 Actual Growth in Mortgage Market 5% Actual IDA Funds Allocation to NMRC (SDR) 118,536,258 Actual Mortgage Debt Outstanding in Year 0 (million Naira) 50,000 Projected Number of Mortgage Loans based on the actuals by end of 2018 10,627 Projected Mortgage Debt/GDP based on the actuals by end of 2018 0.14% 69. Efficiency in Project Design: The selection of the Project Administration Team (PAT) located at the Central Bank of Nigeria led to inefficient project implementation. The Project was administered by the Central Bank of Nigeria (CBN) through a PAT within the Other Financial Institutions Supervision Department (OFISD), which oversees the supervision of non-bank financial institutions including the Nigeria Mortgage Refinance Company (NMRC) and microfinance banks. The CBN’s role as the project implementing entity created a potential for conflict of interest considering its regulatory duties. In addition, almost all PAT staff had full time jobs as bank supervisors and they often had to meet competing deadlines for the project and for their supervisory tasks. The PAT also lacked full-time staff to handle project management, M&E, and fiduciary aspects of the project. The PAT initially insisted on handling most of the project duties without reliance on external consultants. As a result, the project implementation experienced substantial delays. Details are further elaborated in Section III (Key Factors That Affected Implementation and Outcome) below. 70. Efficiency in Project Implementation: The project was approved in September 2013 and was effective in March 2014. However, during the first year of project implementation, performance was slow since key recruitments within PAT did not take place. NMRC had an interim license in the beginning of the project and was granted full license in February 2015. The only achievements of the project during the first year were the setting up of NMRC and the development of the Uniform Underwriting Standards (UUS), together with some progress on improving the legal framework for mortgage loans. By July 2015, NMRC was able to issue and meet the eligible pool of the mortgages set out in the specified disbursement linked indicators (DLI), which triggered disbursement of the second tranche of US$100 M. It was not until June 2018 that NMRC was able to do another bonds issuance due to the unstable macroeconomic environment caused by oil price shocks The sharp decrease in oil prices also caused mismatch between the demand and supply for US dollars, which led to Naira devaluation and project restructuring to adjust exchange rates for the project to meet their DLI targets. Despite the above-mentioned challenges, the NMRC component was the largest and the most successful component. It had achieved 3 out of 6 DLIs and disbursed 56 and 95 percent of the overall project allocation before and after the partial cancellation (which equals to SDR 199,500,000 and SDR 126,297,393 respectively). 71. By December 2018, only 59.73 percent of the project funds were disbursed and therefore a partial cancellation Page 31 The World Bank Housing Finance Development Program (P131973) of SDR 73,202,607 was done right before project closure to enable project funds to be recycled into the Nigeria IDA envelope. After the partial cancellation, disbursement ratio surged to 99 percent. Table 11 below shows disbursement by project category both before and after partial cancellation. As shown in the Table, the “Line of credit to NMRC” was the most efficiently operated category, as 71 percent of the allocated amount before the partial cancellation was disbursed. The undisbursed amount was mainly due to delays in NMRC’s bonds issuance, which were mainly due to unfavorable macro situations and hence outside the control of the Project. The “Mortgage Guarantee product” did not result in any tangible outputs, since NMGC was not operationalized and 0 percent allocated to the mortgage guarantee product (SDR 16.6m before partial cancellation) was disbursed at project close. On the “HMF loan” category, only SDR 5.4 million, or 53 percent of the originally budgeted amount (SDR 10m before partial cancellation), was disbursed for PAT’s administration, and only a portion of that amount was utilized. The pilot tranches were disbursed to the 7 selected MFBs. However, due to the substantial issues identified in the MFB assessment report for the pilot, lack of supporting documentation and lack of compliance with project’s rules, no following tranches were disbursed. On the “Goods, Non- CS, CS, OC and Training” category, only SDR 1.12 million, or 17 percent of the originally allocated amount (SDR 6.6 million before partial cancellation) was disbursed. The list of TA cost breakdown is attached in Annex 7. Table 11: Disbursement by Project Category 46 % Disbursed (on % Disbursed (on Amount (SDR, Amount (SDR, the amount the amount after Category before partial after partial Disbursed (SDR) before partial partial cancellation) cancellation) cancellation) cancellation) Line of Credit Eligible 166,300,000.00 119,837,267.00 118,536,258.05 71% 99% Investment Pt 1 NMRC Mortgage Guarantee 16,600,000.00 0 0 0% 0% Product HMF Loan 10,000,000.00 5,339,258.00 5,339,256.35 53% 100% Goods, Non- CS, CS, OC, 6,600,000.00 1,120,868.00 1,120,867.94 17% 100% Training Total 199,500,000.00 126,297,393.00 124,996,382.34 63% 99% D. JUSTIFICATION OF OVERALL OUTCOME RATING Rating: Moderately Unsatisfactory 72. The overall outcome of the project is rated Moderately Unsatisfactory. Despite that the project PDO was 46Bank’s Operations Portal, data as of April 2, 2019 Page 32 The World Bank Housing Finance Development Program (P131973) highly relevant to the CPF and country strategy, it had achieved negligible efficacy in achieving the intended outcome, as well as a modest efficiency with which the inputs and resources translated into results. E. OTHER OUTCOMES AND IMPACTS (IF ANY) Gender 73. The project design did not provide for any outcome or impact specifically aimed at addressing gender considerations. Institutional Strengthening 74. The project contributed to the set-up of a liquidity facility that is self-sustained via bond issuance and had already shown positive impacts in the market. IDA funds had strengthened NMRC’s balance sheet, facilitated access to capital markets, ensured sustainability of the new model adopted by the WB, and covered administrative expenses for NMRC to recruit highly qualified staffs to run its operations. 75. Awareness campaigns, trainings and workshops were organized throughout the project implementation period. The project strengthened the capacity of primary mortgage lenders (PMLs) by introducing uniform underwriting standards (UUS) for both formal and informal sectors. The UUS provided industry standards for granting mortgage loans to borrowers, promoted efficiency and mitigated the mortgage financing risks for these financial institutions. Capacity building events were held to train PMLs on housing mortgage, underwriting, collateral evaluation, mortgage risk management, housing finance policy, delinquent loans management, mortgage consumer protection and financial literacy, etc. 76. The project had strengthened capacity of microfinance banks through provision of technical assistance on housing microfinance product development, consumer protection, marketing, technical, risk management and loan monitoring skills necessary to administer the HMF products. 77. In support of strengthening the legal and regulatory environment underpinning the housing finance sector, trainings and workshops were conducted to strengthen the capacity of banking supervisors and relevant ministries who supervise the activities of the project. Trainings for judges were also conducted to improve contract enforcement and foreclosure. Mobilizing Private Sector Financing 78. The creation of NMRC mobilized private funding from Primary Mortgage Lenders (PMLs). NMRC had so far refinanced in excess of N17 billion of mortgage loans making possible for mortgage lending institution to continuously access the capital market for long-term funding. NMRC had successfully issued an N8 billion (US$22.2 million) Series 1 Bond and an N11billion (US$30.56 million) Series 2 Bond in July 2015 and May 2018 respectively under its Bond Issuance Page 33 The World Bank Housing Finance Development Program (P131973) Program with both having maturities of 15 years - making NMRC the first mortgage refinance institution in Africa to issue bonds for mortgage refinancing. NMRC addressed the asset-liability mismatch issue that PMLs are experiencing, hence encouraging them to provide longer term funding at a lower interest rate. NMRC also acted as an intermediary between PMLs and the bond market, with the objective of providing long-term funds at better terms and conditions than PMLs might be able to obtain if acting alone. 79. Equity contributions from PMLs accounted for 53.41 percent of the paid-up shares of NMRC 47 . NMRC had N11.59 billion Tier 1 capital as of end of FY17, which was well above the minimum capital requirement of N5 billon as stipulated by the CBN. Five commercial Banks and twenty PMBs represent 11.11 percent and 52.30 percent respectively of the paid-up shares of NMRC. The rest of the shares are held by Ministry of Finance Incorporated (MoFI, 15.68 percent) and the Nigerian Sovereign Investment Authority (NISA, 20.91 percent) respectively. The paid-up capital from the 25 private sector investors demonstrates the mobilization of private funding in the setup of a public-private institution in Nigeria. 80. The project differed from other WB housing finance projects where WB loans were used directly for on-lending, and project support concludes once the credit line is disbursed. In this project, IDA funds have strengthened the balance sheet of NMRC, providing confidence in the credit standing of NMRC as a bond-issuing entity. In addition, IDA funds will be invested in government securities to generate sufficient return to cover administrative expenses and generate sufficient income to grow the capital base in line with growth in the balance sheet. This model departed from some of the previous WB housing finance projects, where World Bank loans were used directly to provide funding for on-lending. NMRC had to issue bonds from the outset before it could begin refinancing operations, through which the project had achieved the maximize finance for development (MFD) outcomes. Poverty Reduction and Shared Prosperity 81. The project encouraged competition in the mortgage market, intending to promote better quality of services and lower costs. It encouraged PMLs and MFBs to offer more flexible loan terms to the applicants, enabled mortgage products to reach lower-income groups, and expanded the eligibility for mortgage product applications. There could also be indirect job creation through the increase in housing resulting from the project, which unfortunately was not captured during the project implementation period. Other Unintended Outcomes and Impacts 82. There are no other unintended outcomes and impacts from the project. 47 NMRC Annual Report 2017 Page 34 The World Bank Housing Finance Development Program (P131973) III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME A. KEY FACTORS DURING PREPARATION 83. The Scope of the Project: The Bank was originally requested to advise on a solution to help unlock the affordable housing market in Nigeria. To achieve this required tackling both the demand and supply side constraints. The demand side constraints included illiquid mortgage market, high mortgage rates, short mortgage tenor and lack of housing finance products, etc. The supply side constraints include high cost of developing residential properties; low developer capacity; as well as issues associated with legal and regulatory frameworks i.e. land administration, mortgage foreclosure law, etc. The engagements on the demand and supply side are complementary to each other. For example, development of the housing mortgage market would not get far without adequate housing stock. In an ideal situation, supply side constraints would need to be addressed first. 84. The design of the Project was streamlined and focused on the demand side constraints only. The project design was also influenced by two key assumptions: i) by creating NMRC and providing liquidity to PMBs, PMBs would give out more housing loans, particularly to low and middle-income groups. Once there is more demand for the affordable housing in the market, the construction industry would respond by supplying more affordable housing stocks; and ii) at the time of the project design, there were complementary projects funded by the Bank and development partners and implemented by the Government, aiming to address cost of residential construction, developer capacities, and legal and regulatory frameworks. It was therefore assumed that the supply-side constraints would be handled by these complementary projects. 85. However, these assumptions did not come to fruition: the construction industry did not respond to the increased liquidity and demand in the mortgage market. As of today, there was still very little stock of affordable housing in the market. There were still plenty of supply-side issues to be addressed, despite the activities of complementary projects which had addressed some areas. Although NMRC had worked on some supply side issues such as the mortgage foreclosure law and developing housing market information portal, many more important actions on the supply side were needed but beyond the scope of this project. As a result, the lack of affordable housing supplies, combined with inadequate legal and regulatory frameworks and the oil crisis between 2014 and 2016, led to low-level increases in the number and volume of mortgages throughout the project period. 86. The selection of the Project Administration Team (PAT): the decision to house the PAT in the CBN created a potential for conflict of interest, both real and perceived, given that CBN was at the same time the regulator for NMRC and MFBs. This led to many delays in project implementation, especially in the technical assistance components. Despite that the PAT had been a recognized champion for housing finance in Nigeria and had shown a lot of dedication to the project, they were carrying out their duties to deliver the work on top of their usual functions within CBN. The project coordinator, which is a full-time job on its own, was also a deputy director within the OFISD department of CBN. Almost all PAT staff were full-time supervisors for other financial institutions at the same time and traveled often to conduct on-site supervisions. In addition, the staff career appraisal structure primarily evaluated staffs’ performance Page 35 The World Bank Housing Finance Development Program (P131973) on their formal supervisory role, without additional incentives to reward those taking on the additional role as PAT staffs. All these factors had an impact in term of PAT staff’s availability to work on the project. As the number of procurement activities increased during the project period, it became evident that PAT was not able to keep pace with these activities, which led to delays in various TA components. In addition to WB procedures connected with the project, PAT was also required to comply with CBN’s own internal procedures and guidelines. In the first years of the project, PAT was unwilling to follow the Bank rules, and it required a lot of effort for the Bank team to facilitate PAT’s compliance with Bank rules, which further complicated and delayed the project implementation process. At the time of the project design, there was low capacity in other institutions to implement the project. The fundamental issue was rather the overall low capacity from the counterparts in project implementation. B. KEY FACTORS DURING IMPLEMENTATION a) Factors subject to World Bank Control: 87. Strong supervision and continuous dialogue with the counterpart. The project benefited from the solid and consistent communication with the counterpart. The Bank’s first two TTLs were based in Abuja. The TTL at closing was part of the core team throughout the project implementation period and maintained continuous and consistent communication with the counterpart during the project period, as well as regular travel to Nigeria. All TTLs were very persistent in requiring the PAT to follow all Bank rules and provide all supporting documents to project milestones whenever they had to approve TORs and give no-objections. The fiduciary team also reached out to the client and provided necessary support in terms of helping the counterpart to transition to the new procurement system. The continuous strong engagement of the task team also ensured that project restructuring, and the partial cancellation of IDA funds, was completed on time in response to the changing environment. The team could have been more proactive on the M&E aspect. When PAT did not provide any updated results for a period, the Bank team should have considered alternative solutions to conduct M&E, including onboarding a Bank M&E specialist to help on this element. 88. Strong governance framework. The team initially expended a lot of effort to design a good governance implementation framework that meets high levels of transparency. A detailed governance supervision matrix was prepared to set out key requirements to be assessed during each WB supervision missions over the course of project implementation. Governance risk was mitigated through these approaches, and in general, the management of NMRC had good reputation. The downside was that the main governance issues arose not from NMRC, but rather the lack of stronger involvement of FMoF, which was a factor beyond WB control and will be elaborated in that section. 89. The external (oil price) shock to Nigeria and the ability of the project to adjust to these circumstances. In response to the oil price shock, the project adjusted the DLIs from USD to Naira through a Level II Restructuring. However, even though it appears the impact of the shock extended beyond the exchange rate, there was no adjustment to the results framework, the issuance of bonds, or other key performance benchmarks. The project should have been restructured earlier to properly adjust to the changing circumstances in Nigeria, putting more focus on the housing market and the lower income segment given the macro circumstances. However, it was not clear if the project could Page 36 The World Bank Housing Finance Development Program (P131973) have achieved this regardless. Even Level II Restructuring to change DLI from USD to Naira, which is comparatively simple project adjustment, was took a long period of time before completion due to lack of counterpart response. Considering this, it would have been difficult to pursue other interventions at the time. 90. Weak presence on urban, land and housing expertise in Nigeria. Even though this project focused mainly on the demand side, supply side remained critical to address the housing sector issue in Nigeria. Despite that there were other WB and development partner projects dealing with supply side issues, these projects still lacked tangible results. One of the difficulties in Nigeria was a relative lack of strong expert presence on urban, land and housing related issues in general. It would have been helpful if the team was able to engage expertise in these areas to advise on the design of the project. b) Factor(s) subject to government and/or PAT control: 91. Lack of high-level implementation. Strong government ownership is critical not only for project development, but also for creating a conducive environment for the project by addressing key demand and supply-side constraints, as well as helping address issues arising from the project during the implementation period. The initial project design envisaged FMoF as a key counterpart, with the PAT (CBN) playing the implementation/administration role. However, shortly after the project became effective, the election started, and the government administration changed. It took at least a year before the new Minister of Finance was appointed. Since then, the project lacked a champion from higher- level leadership, and PAT was the only counterpart to the Bank. The PAT (CBN) effectively took full control of the project in terms of policy direction and administration (in addition to its role as a regulator for NMRC and PMBs). Further, the governance structure put in place at project design with a Project Steering Committee (PSC) to be chaired by the FMoF and consisting of senior representatives (director level) from key ministries, states, regulators and private sectors, was never fully implemented. The supervision team found itself without necessary powers or support from FMoF to effectively remedy the issues encountered during project implementation. 92. Lack of capacity within the PAT. PAT lacked capacity in handling project implementation and administration. As mentioned before, key PAT staff were also full-time supervisors in the OFISD department and often needed to play dual roles to fulfill responsibilities at both ends. Further, due to the absence of full-time staff to handle project management and M&E activities, data for the project M&E framework was not being collected on a regular and systemic basis. Despite the Bank’s attempted efforts to encourage the PAT to hire additional staffs to handle all these tasks (including a full- time project coordinator) this never happened. The absence of a procurement consultant until mid-2016 had been a continued source of concern for the Bank. The procurement consultant that was eventually recruited lacked adequate information and access, including access to the STEP system, which was essential in fulfilling the expectations associated with this role. As a result, there were delays in procurement, M&E and other activities, which negatively affected implementation and results achievement. 93. Mortgage Interest Rate Cap: The CBN Guide to Charges by Banks and Other Financial Institutions in Nigeria (effective May 1st, 2017) introduced an interest cap for mortgage financing which makes mortgage lending uneconomical for mortgage lenders. The cap is “negotiable subject to a maximum of Monetary Policy Rate +5 percent”, Page 37 The World Bank Housing Finance Development Program (P131973) which amounts, at present, to 19 percent. During the project implementation, the team continuously engaged discussions with CBN (including OFISD and Financial Policy and Regulation Department) regarding the removal of the cap given its harmful impact on market development. CBN indicated that the cap would be removed, however, the mortgage cap was not yet formally removed at project closure 48. While interest rate caps are intended to reduce the cost of lending, increase access to finance, and reduce over-indebtedness, experiences from other countries show that the cap could lead to higher fees, the rationing of credit resources, and greater indebtedness through origination of larger loans. The cap remained in place until project closure and was removed only in September 2019. 94. Expanding NMRC’s refinancing footprint among the PMLs. Based on discussion with industry stakeholders and NMRC, the market share of NMRC member bank was estimated to be about 40 percent of the qualified mortgage portfolios in all PMBs. However, when compared to the entire mortgage market including portfolios of the commercial banks, the market share held by NMRC member banks was somewhat negligible due to the size of PMBs’ balance sheet relative to commercial banks’ balance sheet. Therefore, it is crucial for NMRC to recruit more PMLs, especially commercial banks, to reach the critical mass on mortgage generation and refinancing volume in the entire market. However, commercial banks typically do not face liquidity problems given the lack of retail operations in Nigeria and the high interest rate environment that incentivize them to simply invest in T-bills and enjoy high returns. Given that their housing finance lending portfolio was relatively low, commercial banks did not immediately recognize the value of NMRC’s refinancing scheme. The project undertook a lot of education and awareness programs to the banks intended to explain the importance of the refinancing scheme, especially given the direction of Nigeria’s financial sector in the long-term when these banks may need long-term liquidity. These awareness programs appeared to achieve at least some level of success: As of project closure, five commercial banks became NMRC member banks and four of them did refinancing with NMRC. 95. The other key constraints faced by NMRC was that the company operated under a club structure. Member banks were required to contribute a minimum of N50 million. Further, there was also a “20-time rule”, which established that the maximum refinancing volume a member could obtain from NMRC is equal to no more than 20 times their equity contribution. This arrangement discouraged some PMLs to join, as they either did not have sufficient funds to contribute the required equity (for some PMBs), or they were unwilling to do so (in the case of commercial banks). The “20 times” rule also limited the refinancing volume of existing member banks, as some member banks were willing to increasing their refinancing but were not willing to increase their capital contribution and hence were unable to do more refinancing with NMRC. It would have been better if the club structure was designed differently, i.e. to allocate fees based on use of the facility. c) Factors beyond World Bank or government/PAT control: 96. The project was negatively impacted by the economic downturn caused by the fall in oil prices since 2014. Oil prices had fallen by over 50 percent, from US$115 per barrel in June 2014 to under US$50 per barrel. This led to acute 48 The interest rate cap was formally removed in September 2019. Page 38 The World Bank Housing Finance Development Program (P131973) shortage of foreign exchange liquidity impacting the Nigerian banks, as their borrowers had difficulties in accessing foreign exchange to service their foreign-currency denominated loans. The CBN introduced multiple exchange rates starting in early 2015 to accumulate foreign reserve and control domestic liquidity, which exacerbated the problem. The currency control not only restricted free trade of Naira but also created confusions in the market, which further lowered investor confidence. There were strong recommendations from international partners for CBN to convert back to a unified interest rate, to let the currency settle at the point where there was a clearing up of the foreign currency backlog. Since April 2017, foreign exchange reserves began to rise again gradually due to oil price increase, and CBN allowed greater convergence of the official exchange rate with the black-market rate. However, there are still additional actions required (i.e. removing the remaining restrictions and multiple currency practices) before investor confidence can be regained. 97. The impact of the economic downturn affected affordable housing through various channels: the lower overall demand in the economy had curtailed investment into new housing in general. Reduced access to foreign currency and possible future devaluation had led to increases in the cost of building and construction, given that a large proportion of building materials were imported. Higher inflation reduced the time value of money, which discouraged lenders from lending for longer durations. The deterioration in the credit environment also led to rising NPLs in the banking sector, which had consequent effects on the capacity of banks to grow assets and extend mortgage lending. All these factors had negatively impacted the Nigerian mortgage market: the actual size of the market had shrunk by 27 percent. 98. Lack of IFC Investment. IFC initially joined many of the supervision missions given that they were considering investing in the project, including direct equity investment to NMRC, investment in a housing pilot in Ogun state, as well as joint venture partnerships with local and international developers. However, IFC ended up unable to conclude equity investment with NMRC almost 2 year after project effectiveness. The inability to conclude an investment was unfortunate and was a missed opportunity to help strengthen NMRC’s governance by bringing in best practice and experience from elsewhere through IFC. Page 39 The World Bank Housing Finance Development Program (P131973) IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME A. QUALITY OF MONITORING AND EVALUATION (M&E) M&E Design 99. The PDO indicators were: i) number of new mortgage loans provided; ii) number of new mortgages below N 5 million; iii) proportion of mortgage debt outstanding refinanced by NMRC. The PDO indicators were simple, attributable and measurable. The target of the results framework was set, however, based on the entire mortgage market development and did not measure NMRC’s direct contribution. These targets appear to have been rather ambitious given that NMRC only had 25 PMLs and did refinancing with 10 PMLs (6 PMBs and 4 commercial banks), as at project closure. Further, the results framework focused on the direct generation of mortgages and refinancing. It did not capture the project’s efforts in creating the enabling environment for affordable housing finance (i.e. the development of UUS, model mortgage foreclosure law, HMIP, and other reform initiatives). 100. The intermediate indicators selected were appropriate for expected outputs and intermediate outcomes of activities under each project component. However, one shortcoming was that the indicators did not cover outcomes of activities under component 4. While component 4 was mainly TA and capacity building, many of the activities carried out under this component contributed to the creating of the enabling environment for housing finance agenda (i.e. UUS, model mortgage foreclosure law, housing market information portal, etc.). Since no indicators were linked to this component, these outcomes of efforts could not be properly reflected in the results framework. Page 40 The World Bank Housing Finance Development Program (P131973) M&E Implementation 101. The M&E implementation had been a major issue throughout the project implementation period. First, the results framework was never updated during the implementation missions, despite the Bank repeatedly requesting for regular updates of the results framework. During the project mid-term review (MTR), the Bank team insisted that the MTR could only be concluded once the Bank receive the updated M&E data. Only then did the Bank receive the M&E data. However, the Bank assessed the data received from PAT to be unreliable, and it was required to follow up with NMRC to request best estimates in lieu of reliable data. At project closure, the M&E results framework was again not updated, with the most recent data available from the MTR. After numerous communications by the Bank, the PAT shared the updated M&E data only in April 2019, four months after project closure. Further, while the M&E framework was implemented by the PAT, no M&E specialist was hired for the project throughout the implementation period. As a result, data for the project M&E framework was not collected on a regular and systematic basis. M&E Utilization 102. There was no clear evidence on whether the monitoring data collected under the project was utilized, besides for reporting purposes in the Bank ISRs. There was no information on whether the data was relevant to any decisions or resource allocations during project implementation. Also, no specific M&E arrangements were envisaged for long-term monitoring of the project outcomes. Justification of Overall Rating of Quality of M&E 103. The ICR assigns an overall rating of Negligible to the quality of M&E, given that there were no M&E personnel at the PAD, and M&E was not conducted on regular basis. M&E was also not utilized to inform or support the project implementation. B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE Safeguards 104. Most of the transactions covered by the Project are for services or investments that will have little environmental or social risks associated with them. However, it was recognized that there may be some business activities in which the environmental and social risks are moderate and require commensurate assessment and management, such as land acquisition, labor and working standards, inappropriate disposal of wastes, or unhealthy or hazardous working conditions. The project was categorized as FI-2 and had adopted OP/BP 4.03 Performance Standards for Private Sector Activities. With regards to the mortgage refinancing activities under component 1, as well as the housing microfinance activities under component 3, one of the prerequisites of the WBG support was to ensure that all project activities are conducted in compliance with the Environmental and Social Operations Manual (ESOM), and that all PMLs Page 41 The World Bank Housing Finance Development Program (P131973) and MFBs integrate into their lending operations the minimum requirements for screening and managing environmental and social risks in line with the ESOM. 105. These safeguards protocols had been satisfactorily adhered to for component 1, as NMRC initially prepared its Environmental and Social System (ESMS) as per required by OP/BP 4.03 in 2015 and had updated the NMRC E&S Policy and procedures that forms the core part of the ESMS in 2017. The updated NMRC E&S Policy and procedures was approved by the NMRC Board in March 2017 and was cleared by the WB Regional Safeguards Advisor on February 15, 2017. This was followed up with an E&S workshop supported by the WB for NMRC and all PFIs. The focus of the workshop was to strengthen NMRC’s and PFIs’ capacity to manage E&S risks in mortgage finance operations according to the provisions outlined in the NMRC’s Environmental and Social Management System (ESMS). 106. NMRC had also strengthened its capacity in conducting proper E&S screening and due diligence against key risks, including appointing staff to be formally responsible for E&S risk management and conducting on-going due diligence and supervision by NMRC. 107. Under the housing microfinance component, there was a total disbursement of US$1.82 million; one of the reasons for lack of further disbursement under the component was insufficient compliance with project’s E&S framework. The Bank had repeatedly requested PAT to ensure that all MFBs meet the minimum standards of environmental and social risk management before the first disbursement, in line with the requirements set out in the Project’s ESOM during the project phase. However, according to the latest E&S report submitted by the E&S consultant to the PAT, not all MFBs had an approved E&S policy in place. Some MFBs had developed the policy but had yet to seek internal management approval, while others only had a draft policy in place. One of the MFBs did not have specific covenants that addresses E&S commitment in their loan agreement. In addition, it was also not clear from the reporting if the disbursed HMF sub-loan portfolio had been screened. As at project closing, the WB was still waiting for clarification and updates from PAT on the above-mentioned areas. Financial Management 108. The financial management of the project is Moderately Unsatisfactory. 109. There have been substantial FM risks identified by the Bank and the external auditor, and it took substantial effort to facilitate PAT’s responses. The financial supervision report conducted in June 2019 flagged that: i) An outstanding balance undisbursed TA amounting to US$294,522.91 needed to be refunded; ii) The project is yet to refund unutilized balance of US$5,678,168.36 from US$7,500,000 released for line of credit to microfinance banks (MFBs); iii) There is an invoice for US$224,981.25 for ineligible expenditure under the Housing Microfinance Component which requires a refund; and iv) there is a sum of US$7,794 unretired funds from the current FM mission. Therefore, the total refund to be made to the project was US$6,205,466.52. As the amount was not refunded to World Bank by September 2019, the loan was officially declared lapsed. Moreover, further information indicated that the substantial portion of the above-mentioned sum of US$7,794 was used for training cost following project closure, which was not allowed under project rules. Currently, the Bank is reviewing additional training expenditure which may not have been utilized in line Page 42 The World Bank Housing Finance Development Program (P131973) with the project rules. Overall, there was apparent lack of basic documentation for incurred expenditure such as keeping custody of invoices, receipts and attendance sheets for records of participants at events, payment vouchers to track trail of expenditure. The accounting system remain manually operated throughout the project duration. 110. In addition, the external audit review for the year ended December 2018 urged the auditors and the project to reconcile the differences in positions regarding the fact that N62,905,145,790.76 including exchange losses of N22,405,203,819.75 in NHFP books excludes interest accrued, thereby under-stating the indebtedness of Federal government of Nigeria to the World Bank. On the contrary, the project stated that interest due to the bank was provided for in CBN account numbers 2050594 and 2050595 with N1,059,963,972.60 and N713,951,712.60 balances respectively, including commitment and service charge as at December 31, 2018. 111. Reporting: There was no outstanding Interim Financial report49. IFR was prepared up to Q4 2017, though preparation of the IFR was done manually. The IFR produced included sources and uses of funds, Special Account Activity statement, and Reconciliation statement, as well as uses of funds by project activity. The internal audit report prepared by the project was only half-page long and falls short of the required standard of the Bank’s internal audit report. The Bank’s acceptable template was not being used by the project, while the content of the IAR is inadequate. The PAT had not been able to prepare internal audit report using World Bank required template and format. 112. Budgeting: The project did not have a budget/disbursement plan as a guide for the implementation of its activities which needs to be addressed urgently. There was a work-plan and procurement plan which the project used consistently for the implementation of the project activities. The approved annual plans formed the basis for all disbursement. The project’s annual plan was not reflected on the uses of funds by project category. Since the project prepared IFR manually, variances were also not reflected on the uses of funds by category. The Bank advised the PAT to ensure that both planned disbursements, as well as possible variances, were inputted on the uses of funds by project category. 113. Accounting: There were satisfactory accounting documents, guidelines, policies and procedures in place. The project Financial Manual and the Government Financial Regulations as well as CBN guideline formed the basis for the maintenance of financial records. The project maintained and operated a manual accounting system using CBN procedure. The accounting records maintained were in order and adequately reconciled with the Bank statements. 114. Internal Controls: despite adequate policy and procedures, there were weaknesses in implementation. There were adequate internal control policy and procedures documented in the project’s operational manual and the Government Financial Instructions including the CBN guidelines and policies. The application of these policies was adequate, with proper approval, authorization and segregation of duties. However, there had been some weakness in internal control system over staff costs, including lunch allowances, per diems, subsistence allowances/DSA, accommodation (including missing hotel bills/receipts) and inadequate supporting documents as per FY2017 financial management supervision. These had been resolved since then, but it reflects weaknesses in internal control. 49 Based on the latest FM supervision report (FY2018). Page 43 The World Bank Housing Finance Development Program (P131973) Procurement 115. The procurement rating at project close was Moderately Unsatisfactory. The project had suffered repeated delays in procurement processes during project implementation phase, which impacted on the delivery of all components, and especially for the Mortgage Asset Registry System, Mortgage Finance Specialist, Safeguards Specialist, Guarantee Product, as well as on the broader capacity building for both mortgage market and housing microfinance. All these were critical components in supporting the project’s achievement of its PDO. 116. One of the key impediments for procurement was the lack of technical capacity within the PAT to handle procurement-related tasks, exacerbated by initial resistance to use external capacity and expertise in this area. Until mid-2016, no procurement specialist was recruited by the PAT and the project coordinator handled most of the procurement tasks by himself or his team. The PAT also had limited experience with the World Bank procurement requirements. As a result, some procurement documents were rejected due to unacceptable quality and were subsequently revised to the Bank’s satisfaction. Given the lack of capacity within the PAT, the Bank had attempted multiple times to invite them for procurement-related meetings to help them resolve the issues they were experiencing, but the PAT never attended these meetings. The PAT was also not very active in attending procurement clinics and other capacity building activities held by the Bank, despite Bank’s active follow up and invitation. These were important platforms to help clients clarify some of the procurement related issues. The Bank also recommended the PAT should onboard a full-time procurement specialist who had experiences with WB procurement procedures, but it was not until mid-2016 before the procurement specialist was onboarded. 117. After the project MTR in mid-2016, the PAT had finally onboarded a procurement consultant who had experience with WB procurement procedures. However, the consultant experienced restrictions to information and access to the PAT, mainly due to CBN’s institutional policy and internal security related issues. For example, the consultant could not access CBN network and was not provided a work station within the PAT, and as a result had to work outside of the CBN most of the time. Finally, the consultant did not even have access to the STEP system as a procurement specialist which was a fundamental requirement. Therefore, despite that the consultant had worked diligently in handling procurement related tasks and in producing procurement plans, once the deliverables were submitted to the PAT, it still took very long for the PAT to submit the plans to the Bank. In addition, it seems that there were also a lot of CBN internal procedures that the PAT needed to follow on top of the Bank procedure, which further delayed procurement processes. The Bank procurement team reflected that sometimes even after the Bank gave the non-objection, it still took very long for the PAT to finalize the procurement process, and it seemed that the PAT needed to go through the highest level to seek approval within CBN even for smaller procurement packages. These also resulted in PAT having to reconcile between the Bank and the CBN procedure, which further complicated the situation. 118. When the Bank introduced the STEP platform in July 2016, the Bank offered one-on-one trainings to guide the PAT through the process; procurement performance improved as a result. As of project closure in December 2018, the project completed the Mortgage guarantee contracted valued at US$427,000 successfully. The PAT managed to upload the procurement plan to STEP under Bank’s insistence (although implementation, including uploading of related Page 44 The World Bank Housing Finance Development Program (P131973) documents, were not completed in STEP). Procurement Post Review (PPR) conducted in March 2019 revealed that PAT sufficiently followed the procurement procedures as outlined in the Bank’s procurement guidelines for an activity involving the procurement of 7 units of laptop. Some issues pointed out from the PPR were addressed by PAT and well documented by STEPs. Based on the above, the procurement rating is Moderately Unsatisfactory. C. BANK PERFORMANCE Quality at Entry 119. The ICR rates the Project’s quality at entry as Moderately Satisfactory, for the following reasons. 120. The project fitted well with Nigeria's development strategy in meeting the housing demand of Nigerian citizens as well as supporting non-oil sectors such as housing that had strong job creation and multiplier potential. 121. The project design was based on sound analytical work and engagement with the FGN since 2005. A FIRST funded program was conducted ahead of the project to support CBN’s Financial Sector Strategy (FSS) 2020, which included the development of a housing finance strategy. The project, therefore, had well captured the challenges in the housing finance market from both demand and supply side. 122. The project focused largely on the demand side due to the circumstances described in earlier section. It relied on assumptions that the success of the demand side would translate to increased housing stocks on the supply side. However, the economic contraction during the project implementation period had a market effect on the housing flow of new mortgage loans. Alongside this, a fall in domestic revenues for FGN led to increased borrowing from capital markets, which placed upward pressure on domestic interest rates and reduced affordability and attractiveness of the mortgage instrument. At the same time, domestic housing production had underperformed due to high interest rates and low capabilities of developers, which limited the available mortgages that could be refinanced by NMRC. 123. The PDO was overambitious and broader than NMRC’s refinancing targets and focused on overall deepening of primary and secondary mortgage markets. The market impact of NMRC depends on the number of commercial banks and PMBs participating in their refinancing scheme. In reality, only 4 commercial banks and 6 PMBs had done refinancing with NMRC, limiting the scheme’s impact on the entire mortgage market. The PDO did not focus on refinancing volumes which was the project’s core activity. In addition, the PDO could have captured the establishment and operationalization of NMRC, as well as efforts in creating the enabling environment for affordable housing finance (i.e. the development of UUS, model mortgage foreclosure law, HMIP, and other reform initiatives), which were project’s major achievements. Quality of Supervision 124. The ICR rates the quality of supervision as Moderately Satisfactory. Page 45 The World Bank Housing Finance Development Program (P131973) 125. Regular project supervision was undertaken throughout the duration of the project. A total of 11 ISRs over 5- year project life (including an MTR in December 2016), documented the progress made and the key issues to be addressed at regular intervals. Two out of the three task team leaders were based in Abuja for most of the implementation period. The third task team leader was a core member of the project from the outset based in Washington but had regular communication with the client and a co-task team leader based in Abuja, along with regular travel to Abuja. 126. The project risk profile was updated regularly to reflect challenges that the project was facing (i.e. downturn in the macroeconomic environment). While most of the project ratings were reflected properly in the ISR throughout the implementation period, the key project ratings were over-rated in most of the ISR. For example, by Mid-Term Review (MTR), the project had achieved less than one-fifth of its three PDO indicators. Despite this, progress towards achievement of PDO still remained Satisfactory since the first ISR, before being downgraded to Moderately Satisfactory in July 2017 (the 7th ISR) and again downgraded to Moderately Unsatisfactory in Feb 2018 (the 8th ISR). 127. The lack of success in this area can primarily be attributed to project management issues affecting the components managed by the PAT in the CBN, including capacity challenges related to applying World Bank fiduciary rules which caused substantial delays. Despite the challenges, all TTLs had demonstrated great professionalism, persistence and patience by complying with the WB procedures, communicating them with the clients and ensure prudent practices on the fiduciary aspects of the project. 128. The project had issues of inadequate M&E framework reporting from the beginning. However, no action was taken during the implementation to address this other than mentioning this in each ISR. M&E could have benefitted from closer support through an M&E specialist as a member of the World Bank team to provide support to the PAT. 129. Feedback received from NMRC and PAT shows that all Bank TTLs were proactive and responsive throughout the project implementation period. However, according to the PAT, one of the main reasons that delayed implementation was the WB Bank TTL’s stringency in providing no objections. 130. The project conducted its MTR in 2016, followed by a Level II Restructuring to change its DLI currencies and a second Level II Restructuring in December 2019 for partial cancellation of IDA funds. The macro situation had worsened at MTR, which impacted the project in a negative way. As a result, less than one fifth of the PDO indicators were achieved. Additional achievements were made by the project that were not recorded as PDO or intermediate indicators i.e. adoption of uniform underwriting standards, mortgage foreclosure law, etc. While the project was doing its Level II Restructuring to adjust DLI currencies, it seems to have missed the opportunity to adjust the PDO/intermediary indicators to incorporate some of its achievements. If these indicators were incorporated at the time, it could have impacted the project efficacy in a positive manner. The team recognizes such missed opportunity. However, it also stressed that the MTR could not be completed for almost a year because the PAT did not provide results framework data. In addition, at the time of MTR, NMRC was not able to issue a bond and the country management unit (CMU) had decided that the project should only be restructured when NMRC issues a bond. All these aspects had prevented the team to further restructure the project. Page 46 The World Bank Housing Finance Development Program (P131973) 131. In 2018, given the prior implementation delays, the project was yet to establish and operationalize the Mortgage Guarantee company, and the PAT had informally explored the possibility to extend the project for one year. However, the extension request was not approved by Bank’s CMU because the project was a problem project plagued with substantial implementation delays, while at the same time, there was a lending preparation on the new Affordable Housing Finance project that could take forward some of the proposed reform agenda. 132. In addition, the project cancelled an undisbursed balance of SDR 73,202,610.03 towards the end of the project. This was a proactive cancellation to ensure that IDA funds could be recirculated into the Nigerian IDA envelope. Despite the project closure, the Bank is providing continuous support on the housing finance agenda through TA activities funded by external resources. Justification of Overall Rating of Bank Performance 133. Overall, the Bank performance is rated Moderately Satisfactory, due to the moderate shortcomings described above. D. RISK TO DEVELOPMENT OUTCOME 134. Macroeconomic risk: As mentioned previously, there was a major unexpected development during the time of the project implementation, which was the oil crisis that took place between 2014 and 2016. Given that the macroeconomic environment is critical for the stability of the financial sector and the development of the mortgage market; and since Nigeria had an undiversified oil-based economy, the oil price drop led to exchange rate fluctuations and devaluation of the Naira, with implications on the mortgage market as well as the project. Banks experienced deteriorated asset quality, which had weakened their appetite and ability to expand their mortgage finance product line. The high interest rate also affected NMRC’s ability to issue bonds and refinance mortgages. Overall, the project’s outcome had been largely impacted by the unfavorable macroeconomic environment, which was not predicted at the project design stage. The PDO indicator targets were set based on the macroeconomic outlooks before the crisis, which later turned out to be overly optimistic. 135. Political Risk: At design stage, the project had a high-level government champion who helped to accelerate the different reforms and to ensure the successful approval of the project. However, shortly after the project became effective, there was a shift in the government and the project lost the champion, together with a lot of momentum. Since that time, the project lacked high-level government support and there was no steering committee to drive the strategic direction of the project and to solve major issues. The only counterpart left to the project was the PAT embedded in the CBN, who had limited resource and capacity to implement the project. Government should play a pivotal role in the housing finance agenda, and high-level leadership commitments are needed to help push forward the relevant legal and regulatory frameworks, ensure close collaboration between different ministries, monitor project implementation and assess results periodically, etc. It would, therefore, be important that political risk mitigation measures are placed in the future to avoid sudden loss of high-level support and momentum in the cases of changes of governments. Page 47 The World Bank Housing Finance Development Program (P131973) 136. Project Design Risk: Another risk to development outcome was the demand-focused design that the project had adopted. The Borrower had pushed for the focus on demand-side interventions during the design stage. There were also complementary projects conducted by other WB teams and development partners that focused on the supply side at the time. However, these supply-side interventions had still not been addressed as of project closure. Land administration issues and low developer capability still hinder the country’s ability to deliver housing units that meet the needs of middle to lower income groups. Lack of progress on supply-side issues had constrained the ability of the construction industry to supply more affordable housing units, despite increased demand in the mortgage market generated by project activities. 137. Based on these considerations, the risk to development outcomes of the project is assessed as High. V. LESSONS AND RECOMMENDATIONS The following key lessons were learned from the project: 138. The project allocated US$250 million of WB funding into NMRC’s tier 2 capital base. This was one of the major innovations in the project design. The sound capital base enabled NMRC to strengthen its balance sheet, helping to generate sufficient return to cover administration expenses, and sufficient income to grow the capital base in line with growth in the balance sheet. It also helped in mobilizing private sector funds to NMRC’s capital base. Further, it provided confidence in the credit standing of NMRC as bond-issuing entity. Bond issuance had been the major failures of the Tanzania and Egypt Housing finance projects, whereas the Nigeria project was successful in having an early bond issuance even under economic downtown. On the other hand, more analytical work needs to be conducted to understand the “right” level of capital needed to create such a liquidity facility in Nigeria, and to explore whether there could be alternative design options to divide WB resources between capital (that can be leveraged) and credit lines (that target populations in the lower income brackets and provide some immediate kick-starter resources which may also be concessional, as designed in the most recent housing finance projects in Rwanda, Kenya and West Africa). Annex 7 compares the allocation of IDA resources for liquidity facilities on various housing finance projects, both before and after the Nigeria project, as well as some of the outcomes of these designs. 139. A right implementation partner is key to project success. A PAT that is well-resourced with sufficient implementation capacity is key to a smooth project implementation, while avoiding institutional conflict of interest. At the time of project design, there was no capacity within FMoF to administer the project. Therefore, as per discussion with FMoF, a Project Administration Team (PAT) was set up within the Other Financial Institutions Supervision Department (OFISD) of CBN, given the department’s long-engaged activities in the housing finance agenda. There was potential for conflict of interest in this type of set up given that the PAT sits within the OFISD department of CBN, and are supervisors for the project beneficiaries, which were the NMRC and MFBs. It became apparent that the PAT lacked resources to handle administration and meet the strict fiduciary requirements, disbursement speed and M&E tasks. Further, CBN had their own internal systems and procedures that differ markedly from the WB procedures Page 48 The World Bank Housing Finance Development Program (P131973) (procurement, accounting, financial management, audits, reporting, etc.). The project had to meet the requirements of both WB and CBN systems and procedures, which required additional time and efforts by the PAT. In the beginning of project implementation, the PAT was reluctant to follow WB rules and procedures, and it took a lot of effort and communication to resolve this issue. All these factors contributed to the implementation delays despite the dedication that PAT staffs had demonstrated during project implementation. The project would have benefitted from a stand-alone PAT with full-time staff to handle all project management aspects including fiduciary, safeguards and M&E aspects of the project. This could result in more efficient project implementation. 140. Strong supervision and governance of the project is Key. Governance issue was a major risk flagged throughout project preparations. The Bank dedicated significant time and expertise in helping the government design a good governance framework with clear targets and requirements to be followed up during each supervision mission. There was also high level of interagency collaboration between supervisory regulators, close partnership among development partners, as well as collaboration with other ongoing projects to strengthen governance. This effort seems to have be worthwhile, as the project did not encounter any major governance issue. NMRC had been operating under good reputation and this approach which should serve as a model for other MRC projects in terms of governance requirements. 141. Limited direct solutions to lower income housing finance. The project had limited focus on lower income housing finance solutions. The project attempted to design a focus on lower income groups through a N5 million quota on refinancing and through the HMF component. Neither of these activities delivered to the extent they should have for different reasons. HMF component also went through big hurdles in getting itself operational because of PAT’s limited technical and operational capacity. The quota on mortgages issued below N5 million was included in the M&E framework but lacked reporting on this due to the weak M&E implementation. It was hence difficult to take supervisory action without any clear numbers. Given these circumstances, the project had limited direct impact on lower income housing finance. Future similar scheme could include simpler solutions to channel the funds to lower income housing populations and might be monitored and supervised effectively through proper M&E framework and strong counterpart collaboration. 142. DLIs should have been set in Naira. The original DLIs targets were set in dollar. This proved to be a design issue which did not anticipate large scale devaluation. When the oil crisis in 2014 caused huge Naira depreciation and the project was unable to achieve its DLI targets because of that, a Level II Restructuring had to be carried out to bring the remaining tranches of DLI targets to Naira, using the historical exchange rate to ensure that the targets could be met. The process of correcting this through a restructuring was also very lengthy due to the lack of responsiveness from counterparts despite that it was a relatively straightforward restructure. Given this experience, it is recommended that DLI targets should be set in Naira from the start, to avoid the potential negative macroeconomic impact on the project’s ability to achieve its targets. 143. All factors are important to consider when designing housing finance projects. In addition to the demand side interventions that were carried out throughout the project, the fundamental prerequisites to the success of housing finance also included supply side interventions such as a proper land administration, property registration and titling Page 49 The World Bank Housing Finance Development Program (P131973) process, effective legal and judicial process for mortgage foreclosure, serviced infrastructure, developer finance and capacity are pre-requisites for a successful mortgage market, etc. Analytical work was done to understand supply side issues, and complementary projects were considered when designing the project. However, the project decided to engage only in demand-side intervention, under the following key assumptions: i) the set-up of refinancing facility would naturally direct PMLs to lend to people at lower income ladder, hence solving the small mortgage critique; ii) the construction industry would respond with supplying more low-cost housing; and iii) key supply-side constraints would be handled by other complementary projects. It had later turned out that it would be risky to narrow down the project scope based on these assumptions, given that mortgages were still offered to the populations in the top income ladders, the supply of lower-cost housing remained low, and not all supply-side issues have been addressed by complementary projects. This confirms that that supply-side interventions were essential to the success of the overall housing finance development in the Nigerian market. 144. Housing finance should be addressed via a longer-term programmatic approach rather than a stand-alone operation. Housing finance is a complex and multi-sectoral agenda that needs to be addressed in a focused, long-term and multi-phased approach. Once a liquidity facility is set up, it takes time for it to do business development, attract financial institutions to adopt its instruments, issue bonds, reduce the asset-liability mismatch issues among the PMLs, improve PMLs’ loan quality and make an impact. It is therefore difficult to make a judgement on the success or failure of such project just after five years, especially given that Nigeria experienced a turbulent economic period throughout the project phase, and no conducive legal and institutional infrastructure was in place at the time of project implementation. The establishment of the liquidity facility might help to make the market more sustainable in the long term but may not have a significant developmental impact as the number of mortgages provided by PMB was still quite small compared to the market size. It was also overly ambitious to think that NMRC could deepen secondary mortgage market when there was an underdeveloped primary mortgage market and a limited number of qualified legacy loans to refinance in the first place. 145. Consistent long-term engagement from the World Bank Group and development partners are needed. The project was designed with a view to ultimately change the housing finance and affordable housing market and to make the market perform in response to the populations at the lower to bottom end of the income pyramid. The instruments designed in the project, including the liquidity facility and the HMF product, were all intended to create a sustainable market solution that would provide pro-poor services without any further need for public funding in the long-run. To push the instruments down the market, it will take longer than the project implementation period, given the macroeconomic situation in Nigeria and the market’s dependency on many external actors who are outside of public sector (i.e. developers, PMBs, consumers, etc.). To change the way the market behaves, strong standing power and consistent engagement from the WBG and development partners is needed. Through this type of continuous engagement, the project and all relevant stakeholders will have an opportunity to take stock on the lessons learned and improve on interventions intended to gradually push the instruments down the market, close the housing gap and better serve the intended final beneficiaries. . Page 50 The World Bank Housing Finance Development Program (P131973) ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS A. RESULTS INDICATORS A.1 PDO Indicators Objective/Outcome: Increase access to housing finance Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of new mortgage Number 4200.00 50000.00 50000.00 14978.00 loans provided 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of new mortgages Number 1050.00 7500.00 7500.00 10348.00 below NGN5 million 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 Comments (achievements against targets): Page 51 The World Bank Housing Finance Development Program (P131973) Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Proportion of mortgage debt Percentage 0.00 20.00 20.00 13.00 outstanding refinanced by NMRC 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 Comments (achievements against targets): A.2 Intermediate Results Indicators Component: Housing Microfinance Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Portfolio at risk for housing Percentage 0.00 2.50 2.50 0.06 microfinance (HMF) pilot 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion 3 - 5 pilot MFBs Number 0.00 5.00 5.00 7.00 Page 52 The World Bank Housing Finance Development Program (P131973) develop/adapt/adopt 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 housing Microfinance loan products, access project credit line and establish first cycle of micro financing lending programs Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of HMF loans Number 0.00 2860.00 2860.00 889.00 disbursed 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Volume (USD) of HFM loans Amount(USD) 0.00 8.00 8.00 1.80 disbursed 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 Comments (achievements against targets): Page 53 The World Bank Housing Finance Development Program (P131973) Component: NMRC Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion NMRC Efficiency Indicator - Percentage 0.00 42.00 42.00 55.00 Cost to Income Ratio 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion NMRC New Bond issuance Number 0.00 50.00 50.00 19.00 volume (years 3 - 5 are without sovereign 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 guarantee) Billion Nigerian Naira Comments (achievements against targets): Component: Guarantee Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of Guarantees Number 0.00 900.00 900.00 0.00 Page 54 The World Bank Housing Finance Development Program (P131973) 31-Dec-2013 31-Dec-2013 31-Dec-2018 16-Apr-2019 Comments (achievements against targets): Page 55 The World Bank Housing Finance Development Program (P131973) A. KEY OUTPUTS BY COMPONENT Objective/Outcome 1 Deepen Primary Mortgage Market in Nigeria 1. Number of new mortgage loans provided Outcome Indicators 2. Number of new mortgages below N 5 million 1. Portfolio at risk for housing microfinance pilot 2. 3-5 pilot MFBs develop/adapt/adopt housing microfinance loan products, access project credit line and establish first cycle of micro Intermediate Results Indicators financing lending programs 3. Number of HMF loans disbursed 4. Volume (USD) of HMF loans disbursed 5. Number of guarantees 1. Component 1 (Establishment and operationalization of NMRC) Key Outputs by Component 2. Component 2 (Guarantee) (linked to the achievement of the Objective/Outcome 1) 3. Component 3 (Housing Microfinance Pilot) Objective/Outcome 2 Deepen Secondary Mortgage Market in Nigeria Outcome Indicators 1. Proportion of mortgage debt outstanding refinanced by NMRC Intermediate Results Indicators 1. NMRC New Bond Issuance Volume Key Outputs by Component 1. Component 1 (Establishment and operationalization of NMRC) (linked to the achievement of the Objective/Outcome 2) Objective/Outcome 3 Increase Access to Finance in Nigeria 1. Number of new mortgage loans provided Outcome Indicators 2. Number of new mortgages below N 5 million Intermediate Results Indicators 1. Portfolio at risk for housing microfinance pilot Page 56 The World Bank Housing Finance Development Program (P131973) 2. 3-5 pilot MFBs develop/adapt/adopt housing microfinance loan products, access project credit line and establish first cycle of micro financing lending programs 3. Number of HMF loans disbursed 4. Volume (USD) of HMF loans disbursed 5. Number of guarantees 1. Component 1 (Establishment and operationalization of NMRC) 2. Component 2 (Guarantee) Key Outputs by Component 3. Component 3 (Housing Microfinance Pilot) (linked to the achievement of the Objective/Outcome 3) 4. Component 4 (Technical Assistance and Capacity Building, establishment of uniform underwriting standards, mortgage foreclosure law, etc.) Page 57 The World Bank Housing Finance Development Program (P131973) ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION A. TASK TEAM MEMBERS Name Role Preparation Supervision/ICR Andrej Popovic Task Team Leader(s) Anas Abba Kyari Procurement Specialist(s) Arigu Yusufu Kudu Financial Management Specialist John Amedu Eimuhi Team Member Joyce Chukwuma-Nwachukwu Team Member Magalie Pradel Team Member Joseph Ese Akpokodje Environmental Specialist Andrey Milyutin Team Member Simon Christopher Walley Team Member Nneamaka Pamela Okechukwu Team Member Elijah Abiodun Siakpere Social Specialist A. STAFF TIME AND COST Staff Time and Cost Stage of Project Cycle No. of staff weeks US$ (including travel and consultant costs) Preparation FY13 37.838 279,492.60 FY14 9.575 52,891.14 FY15 0 494.79 Total 47.41 332,878.53 Supervision/ICR Page 58 The World Bank Housing Finance Development Program (P131973) FY14 28.259 162,582.18 FY15 32.007 213,912.61 FY16 27.515 196,051.07 FY17 34.326 218,624.30 FY18 36.110 245,308.38 FY19 44.444 206,425.85 FY20 2.175 8,744.04 Total 204.84 1,251,648.43 Page 59 The World Bank Housing Finance Development Program (P131973) ANNEX 3. PROJECT COST BY COMPONENT Amount at Approval Actual at Project Percentage of Approval Components (US$M) Closing (US$M) (US$M) 1. Mortgage Refinance 0 250.00 0 Company 2. Mortgage Guarantee 0 25.00 0 3. Housing Microfinance 0 15.00 0 4. Technical Assistance and 0 10.00 0 Capacity Building Total 0.00 300.00 0.00 Page 60 The World Bank Housing Finance Development Program (P131973) ANNEX 4. BORROWER’S COMMENTS TO THE ICR 146. The ICR was disseminated to the Borrower for their review and comments on August 12, 2019. The Borrower was expected to send the consolidated comments, if any, formally via the FMoF by September 20th, 2019. As of October 2nd, 2019, the Borrower did not provide any formal comments or feedback to the ICR. The ICR was informed with inputs and feedback received from PAT. Page 61 The World Bank Housing Finance Development Program (P131973) ANNEX 5. SUPPORTING DOCUMENTS (IF ANY) 147. As mentioned in the main text, the MFB component disbursed one tranche of US$1.82 million after 7 MFBs went through TA. In August 2018, an assessment was conducted on the utilization of the project funds August 2018. The assessment had identified substantial irregularities, including inconsistencies in loan agreements and incomplete E&S policy for some MFBs. The report also highlighted a case of misuse of funds by one MFB, which disbursed part of the funds to related parties (including managing director of the MFB) and for ineligible activities and indicates its lack of compliance with eligibility criteria required to remain a participating financial institution under the project. 148. Based on the assessment, the Bank requested additional information on the utilization of funds and compliance with project criteria before providing no objections to disburse the second tranche. The additional information requested by the Bank for further clarification is summarized as below: • Addressing the identified discrepancies in loan agreements through relevant amendments for all active loan agreements and provide relevant copies of amendments. • Ensuring that loan covenants are implemented (e.g. E&S, consumer protection) through inclusion in sub-loan agreements. • Confirming compliance of each PFI against project eligibility criteria, including full compliance with all prudential requirements and minimum “sound” rating, and include overall current PAR and CAR for information purposes. • Providing brief information on the use of the disbursed funds, in line with the sub-loan eligibility criteria included in the loan agreement, along with updated PAR for the disbursed HMF portfolio for each PFI. • Provide updated E&S Report, addressing comments in the attached and most importantly confirming if the disbursed HMF sub-loan portfolio had been screened (which was not clear in the attached report for all MFBs) • In light of the upcoming project closure, for those MFBs that had achieved 100 percent and fully satisfactory utilization, provide a loan demand schedule not to exceed the allocation for the originally planned 2nd tranche and which could be fully on-lent to eligible sub-borrowers immediately or within 2 weeks from the receipt of funds from CBN and no later than by November 30. In case of lack of compliance, the funds would be withdrawn (this may require update in the loan agreement). • Provide an update on disqualification of the MFB that did not follow the project rules and retrieval of disbursed funds in line with provisions of the loan agreement, and/or subsequently documented evidence of corrective action. 149. The Bank had since then repeatedly requested this information from PAT (including putting the requests in the Aide Memoires but had not received any feedback from the PAT. The Bank had also subsequently requested evidence Page 62 The World Bank Housing Finance Development Program (P131973) of corrective action with numerous deadlines which had not been complied with in form of substance. Given the lack of these information, the Bank could not provide no objection to release the second tranche of microfinance loans. In addition, the funds disbursed to one of the MFBs that did not comply with project rules were ultimately declared as ineligible for financing under the project. Page 63 The World Bank Housing Finance Development Program (P131973) ANNEX 6. DETAILS OF THE FOCUS GROUP STUDY SESSIONS 150. During the ICR mission between January 28 and February 1st, 2019 the Bank conducted six focus group study sessions. It is important to note that the purpose of the focus group discussion was aimed at recording general impressions about the project and interviewed a very small sample of final beneficiaries (10 for PMBs and 18 for MFBs). The PAT helped to arrange the sessions with Microfinance Banks (MFBs) and its beneficiaries. NMRC helped to arrange the sessions with Primary Mortgage Lenders (PMLs), and its beneficiaries. The sessions were conducted in both Abuja (for participants located in the northern region) and Lagos (for those located in the southern region). 7 staffs from 4 different PMBs and 13 staffs from 6 different MFBs participated in the sessions. These staffs were selected by NMRC and PAT based on their roles and responsibilities. PMB staffs mainly consisted of heads of mortgages, credit risk managers, chief risk officers and relationship managers. MFB staffs mainly consisted of housing loan officers, credit risk officers and housing loan coordinators. 151. The beneficiary interviewees were selected by PMBs and MFBs themselves. 10 beneficiaries were selected by the PMBs. They were mainly white collar works with professions such as engineers, investment bankers, HR managers, with annual salary ranging from N12m to N25m, and applied loan amount ranging from N13m to N48. 16 beneficiaries were selected by MFBs and they consisted of a mix of MSMEs who needed the loan to do renovation/incremental housing/building of their businesses, as well as individuals who wanted to construct their own houses. Their annual salary ranges from N2.5m to N24m with applied loan amount ranging from N1m to N6m. It is evident that the PMB and MFB beneficiaries that participated the interviews represented upper middle income to upper class populations. Key takeaways from the focus group sessions are summarized below: PMLs and beneficiaries’ sessions: 152. The participating PMBs had shared a lot of positive impact brought by the project. Doing refinancing with NMRC had enabled them to reduce their interest rates from 20-25 percent to 17-18 percent. It also enabled PMBs to provide loans with longer tenor (20 years vs. 5 – 7 years before). Refinancing enabled PMBs to eliminate asset liability mismatch and improved their liquidity situation. 153. Beneficiaries had perceived both increased eligibility and affordability for mortgages. They perceived that PMBs had improved level of transparency and literacy as they did a lot of advocacy to ensure that the products are known and well understood. Most banks developed new incentives to their loan officers after joining the refinancing scheme. Some PMBs expanded customer base (i.e. informal sector clients increased from 0-10 percent for on PMB). 154. Thanks to the Uniform underwriting standard (UUS), the process of getting mortgages had been simplified and the time to process mortgage had reduced. UUS had also set an industry standard and provided PMBs with the underwriting tool kit, which is very valuable. Initially the UUS had received some push bank from member PMBs due to its high standards, but after realizing that by adopting the UUS, it would help PMBs improve the quality of their mortgage portfolio and reach long-term financial sustainability, PMBs were willing to comply with the high standard set by the UUS. 155. The participating PMBs and MFBs also made a few suggestions to the project. One of them was that the housing finance program is currently limited to first time home buyers and there is also a “one refinance per person” limitation. They were hoping that this limitation could be removed and that PMBs should conduct more risk-based appraisal in determining one’s capacity to take on more refinancing. Page 64 The World Bank Housing Finance Development Program (P131973) 156. On the advocacy, it was suggested that NMRC should do more mass media, especially for the informal sector and diaspora customers. In addition, most of the advocacy events are done in Abuja and Lagos area, which is okay for the pilot but there is a need to spread to the rest of the country to reach a wider range of population, including those living in rural areas. 157. Among the participating PMBs, there was a call for NMRC to open its refinancing to more financial institutions. As of now, NMRC operates in a “club structure”, and only member banks who contributed minimal capital of 50 million Naira could participate in the refinancing. This had limited some PML’s willingness and ability to join the scheme. In addition, there is a “20-time refinancing cap” that that the maximum refinancing volume a member could get from NMRC equals to 20 times of their equity contribution. This also limits the refinancing volume of existing member banks, as some member banks were willing to refinance more but were not willing to increase their capital contribution and therefore were unable to do more refinancing with NMRC. To achieve the project development indicator set at design stage, NMRC needs to expand its refinancing portfolio and engage more PMLs, especially commercial banks, to participate on the refinancing scheme. The removal of 20-time refinancing cap could also potentially help existing member banks to expand their refinancing portfolio. MFB and beneficiaries’ sessions: 158. The HMF pilot was very much welcomed by the participating MFBs. Most of the MFBs already had a housing microfinance product at the time when the product was launched. The beneficiaries interviewed claimed that their interest rate they received through HMF loan was 3.5 percent (as versus to 4.75 percent available to them before the pilot), as well as a higher tenor (from maximum 12 months before to 24 months after the pilot). Most of the participating MFBs had experienced growth in customer and expanded their customer base from entrepreneurs to salaried earners. All MFBs experienced increased portfolio. One MFBs increased from 3 branches to more than 40 branches, and increased customer base from 100 before the pilot to 400 after the pilot. Others reported increased customer base by 20 percent. 159. Most pilot beneficiaries had confirmed that the pilot had enabled them construct/upgrade their business facility, which increased their revenues and productivity by more than 10 percent and 90 percent respectively, and that the pilot had enabled them to get loans with reduced interest rate (from 6 percent to 5 percent), longer tenor (from 6 months to 1 year and above), bigger loans (from maximum N 500k to above N 1M) and reduced loan processing period (from more than 2 weeks to less than a week). Most MFB beneficiaries also noticed positive changes towards their MFB’s business strategy and operations since the adoption of the pilot including increased efficiency, expanded customer base, increased proactivity and advocacy, more flexibility in their product offerings, etc. 160. The MFBs also appreciated the capacity building aspect of the pilot and appraised the technical team for their high-quality work. The capacity building covered the entire process of HMF and helped MFBs with their marketing, technical, risk management and loan monitoring skills necessary to administer the HMF products. The MFBs reflected that the most valuable trainings included i) the loan amount to provide to the client based on the needs assessment; ii) education on securing title document and getting true ownership of collateral; as well as product categorization, etc. The “train the trainer” feature embedded in the capacity building curriculum also enabled these knowledges to sustain Page 65 The World Bank Housing Finance Development Program (P131973) within the participating MFBs in the long run. 161. MFBs had expressed their frustration regarding the failure of the second tranche disbursement. Most MFBs had set up special HMF unit since the launch of this component and had increased their resources and capacity to host the product. They claimed that the discontinuation of the pilot would lead to their operating loss. Further, some MFBs had exhausted their own resources to generate HMF loans, with the hope of receiving the 2nd tranche disbursement afterwards. Now these MFBs don’t have liquidity to meet other loan commitments, are unable to meet customer demands and facing reputational risks. Lastly, MFBs had not been formally communicated by the PAT on the status of the HMF pilot. They were asked to submit documents including the projection of their next tranche disbursements until end of 2019. During the focus group studies, MFBs pleaded for a PAT’s clarification on this program. It is worth to clarify that despite these comments, the pilot had always intended to help the MFBs roll out the scheme and MFBs would use their own funds for further expansion, otherwise the whole scheme wouldn’t have been sustainable. 162. Table 12 below shows the number of participants from each institution, together with their business functions (where applicable). Table 12: No. of participants from each institution for the focus group sessions Name of Institution No. of Participants Business Function PML Session Imperial Homes Mortgage Bank 2 Head of credit risk and admin, relationship manager, Home Base Mortgage Bank 2 Head of mortgages, risk management Trust Bond Mortgage Bank 3 Mortgage Banking, Head of Business Development, Credit Risk Management Total Attendance for PML session 7 PML Beneficiary Session Beneficiaries associated with Trust Bond 3 Beneficiaries associated with Homebase 4 Beneficiaries associated with Imperial Homes 1 Beneficiaries associated with GT mortgage 1 Total Attendance for PML Beneficiaries Session 9 MFB Session Baobab MFB 2 Microfinance Loan officer, Hasal MFB 2 Microfinance Loan Page 66 The World Bank Housing Finance Development Program (P131973) officer, Accion MFB 2 Housing loan coordinator, Housing loan officer AB International MFB 2 Housing loan officer NPF MFB 2 Credit risk officer, credit operations Lapo MFB 3 Microfinance Loan officer, Housing Loan coordinator, Manager Total Attendance for MFB Session 13 MFB Beneficiaries Beneficiaries associated with Baobab MFB 3 Beneficiaries associated with Hasal MFB 3 Beneficiaries associated with Accion MFB 3 Beneficiaries associated with AB International MFB 3 Beneficiaries associated with NPF MFB 3 Beneficiaries associated with Lapo MFB 3 Total Attendance for MFB Beneficiaries Session 18 Page 67 The World Bank Housing Finance Development Program (P131973) ANNEX 7. IDA RESOURCE ALLOCATION ON DIFFERENT HOUSING FINANCE PROJECTS Country Project Board Total Amount to IDA Resource Design on Outcome (at project closure) Name Approval Project Liquidity Liquidity Facilities Year Amount Facilities Jordan Housing 2002 US$20m US$19.6m - US$19.6, for on- - Refinance operations Finance and lending to JMRC totaled US$67m (30 Urban Sector percent of total eligible Reform - JMRC funds these mortgage loans for Project operations through its refinancing) own capital (6.8 percent of total - Outstanding bonds of assets), WB loan (17.8 US$73 million, whose percent) and bond original yields were issues (72.6 percent) between 100 and 200 bps below other - US$0.4m to support corporate bond issues housing sector reform - TA was not carried out. Tanzania Housing 2010 US$100 US$ 73M - US$73M (33 for - US$73M plus the Finance ($40 m (33 for original and 40 for AF) remaining unused Project Original original and for on-lending to amount from other IDA and 40 for AF) TMRC components as of today. $60 M AF for TMRC - US$20 M HMFF (in which - Bond issued 2 private US$10 placement (Nov 2017) million is equivalent to USD 1.74M, conditioned and 2 public bond issuances (May 2018 and for the June 2019) totaled equivalent equivalent to US$ 9.4 M. amount of Grand total of US$ 11.1 bond M (it also results of issuance) unlocking US$ 10M IDA) US$ 20 M for HMFF Egypt Mortgage 2012 US$37.1m US$37.1m - US$37.1 m for on- - LoC fully disbursed Finance lending to EMRC as - EMRC didn’t issue bonds Project LoC - EMRC rely on its own fundraising capability Page 68 The World Bank Housing Finance Development Program (P131973) (i.e. bond issuance or private placements) to continue operations Nigeria Housing 2013 US$300m US$250m - US$250m as tier 2 - NMRC successfully issued Finance capital for NMRC N19 bn bonds despite oil Development crisis Project - Disbursement-linked - Some level of excessive indicators are liquidity for NMRC introduced to ensure - Limited achievements for project only housing finance for lower disbursement once income groups the project meets - Mortgage guarantee and certain objectives housing microfinance product unsuccessful due - US$40m to create to project mortgage guarantee implementation and housing challenges microfinance products - US$10m on TA and capacity building WAMEU Affordable 2017 US$130m US$130m - US$80 IDA Scale-up WAMEU, Rwanda and Kenya Housing facility (SUF) to CRRH are still in too early stage to Finance for on-lending to summarize outcomes. banks (for loans below or equal to US$26k) - US$50m for on- lending to NBFIs (for loans below US$17k) - IDA Grant US$15m on TA & capacity building, design of regional credit guarantee product and support on housing policies Rwanda Housing 2018 US$150m US$147m - US$147 LoC for on- Finance lending to financial Project institutions for targeted beneficiaries. - US$3m TA assistance and capacity building Page 69 The World Bank Housing Finance Development Program (P131973) Kenya Affordable 2019 US$250m US$240m - Us$50 allocated Housing against DLIs as equity Project and tier 2 capital/subordinated debt which will mobilize private sector capital. The subordinated debt will be invested in income generating securities to build KMRC’s capital and income, provide a cushion for lending to SACCO’s MFBs, and raise investors’ confidence - While building KMRC’s capacity to issue bonds, IBRD will provide a credit line of US$190m to KMRC through NT to refinance affordable housing loans - US$10m TA support to Ministry of Lands and Physical planning to improve property registration Page 70 The World Bank Housing Finance Development Program (P131973) ANNEX 8. DESCRIPTIONS AND BACKGROUNDS OF VARIOUS INSTITUTIONS A. Federal Mortgage Bank of Nigeria The Federal Mortgage Bank of Nigeria (FMBN) was created under the Federal Mortgage Bank Act (1992) to provide long-term credit funding to mortgage institutions and encourage and promote mortgage financing throughout the country. To date, however, FMBN has not acted as a mortgage bank that lends to retail lenders. Rather, it acts as the body that manages the National Housing Fund (NHF), a compulsory provident scheme that finances through a tax on salaries imposed on workers and employers. The proceeds can only be lent to retail lenders for their use in financing home purchases. These private mortgage institutions (PMIs) disburse subsidized loans financed by the NHF, usually combined with their business of making market-based loans funded almost entirely by retail deposits. It has resulted in a limited number of loans and a system which Is both inefficient and time consuming. The assets and finances of the NHF are effectively commingled with those of the FMBN, and both operate without sufficient transparency or accountability. Less than 1 percent of contributors have received loans or can expect to receive any. The FMBN is the apex mortgage institution in Nigeria, and has a broad mandate to: • link the capital market with the housing markets; • encourage the emergence and growth of viable primary mortgage loan originators to serve the need for housing finance in Nigeria; • mobilize domestic and foreign funds into the housing sector; • collect and administer the NHF in accordance with the provisions of the NHF Act. B. Nigeria Mortgage Refinance Company (NMRC) At project design, NMRC is to be created in partnership between private financial institutions, development finance institutions and the Ministry of Finance. The NMRC will be registered as a non-bank under publicly listed limited liability company rules and regulation with majority ownership coming from the financial institutions and will act as a bridge between mortgage lenders and the capital markets. It will issue standard corporate bonds in the capital markets and use this funding to make loans to mortgage lending institutions. These loans will be backed by collateral in the form of an assignment of rights over part of the lenders' mortgage portfolio. This collateral will protect NMRC's loans and minimize the amount of credit risk that the agency will take. The minimized risk, in turn, will help to achieve the lowest possible rate in raising its bond financing. A key distinction of this model compared to securitization is that the underlying loans offered as collateral stay on the balance sheet of the lender, and, should they go into default, would need to be replaced by performing loans. Thus, unlike mortgage-backed securities (MBS), this model does not take direct credit risk on mortgage borrowers. Page 71 The World Bank Housing Finance Development Program (P131973) A total of US$250 million equivalent will be disbursed in 6 tranches to NMRC and sit on the balance sheet as Tier 2 capital. The role of the IDA funds will be twofold: (a) Strengthen Balance Sheet. They will provide confidence in the credit standing of NMRC as a bond-issuing entity that critical in ensuring its ability to raise bond financing at just above sovereign debt levels; (b) Ensure Sustainability of Model. The IDA funds will be invested in Government securities to generate sufficient return, to cover administrative expenses, and to generate sufficient income to grow the capital base in line with growth in the balance sheet. This model is a departure from previous models in which World Bank loans were used to directly fund on-lending. One of the consequences of this new model is that NMRC will have to issue bonds from the outset before it can begin refinancing operations. Pricing from NMRC to mortgage lenders will depend on the marginal cost of funds raised by NMRC in the bond market together with an operating margin of 1 percent to cover costs and return on equity. Initially, to give confidence to bond holders in the new entity, IDA funds will represent a large proportion of the balance sheet. As the balance sheet grows and NMRC generates its own capital resources through accruing reserves, the IDA funds will become less significant, although they still will be critical in creating sustainability. The Federal Government of Nigeria will provide limited support to NMRC in its initial phase through a guarantee of its bond issuance. The guarantee will be both time limited and amount limited, but, in the context of the Nigerian market, is seen as a necessary credit enhancement. Institutional investors will not be familiar with NMRC and will require some initial reassurance to support their investments. Alongside its initial support, the Government will seek a number of regulatory concessions for NMRC bonds to ensure that they become attractive and liquid capital market assets. The project will help expand the reach of mortgage finance down the income distribution by setting targets for the refinancing of smaller loans. As part of the project design and to encourage greater access to mortgage finance, there will be a requirement that the portfolios of mortgage loans presented as collateral in a refinance operation must contain a specified minimum proportion of smaller loans. A starting target could be in the range of 30 percent of loans being below N5 million (US$32,000). This requirement would be subject to review and regular update through the Project Operations Manual (POM). Governance Structure of NMRC: The NMRC is organized as follows: Shareholders The NMRC is majority owned by private sector institution and is registered as a publicly listed company owned by Deposit Money Banks, Primary Mortgage Banks, International Financial Institutions and a 20 percent share owned by the Ministry of Finance. The NMRC falls under the supervisory authority of the Securities Exchange Commission (SEC) and the Central Bank of Nigeria. NMRC will provide liquidity only to those mortgage lenders Page 72 The World Bank Housing Finance Development Program (P131973) that also are shareholders in NMRC. Because the NMRC is a private sector company with a public mission, the shareholding structure is intended to reflect the public private partnership. The NMRC should be: (a) majority privately owned; and (b) any public stake limited. The proposed shareholder mix will comprise the Ministry of Finance Incorporated (MOFI) and 3 groups of investors, namely: (a) Banks and PMBs; (b) International Finance Institutions such as the IFC, Shelter Afrique, ADB, (FMO), and KFW; and (c) Private Investors such as Pension Funds, Investment Funds, Private Equity, and Insurance Companies. The FMoF will hold up to 20 percent preference shares, valued at 1.2 billion Naira (US$7.6 million). The remaining shareholders will hold up to 90 percent ordinary shares, valued at 4.6 billion Naira (US$29 million). The shareholdings of MOFI and Sovereign Wealth Fund will be diluted gradually over time as other private shareholders participate in the equity. Government support is envisaged as a temporary measure to facilitate the establishment of NMRC. Board There could be 7-10 members of the Board of Directors who are the Chief Executive Officers (CEOs) of the participating financial institutions. The Board will be responsible for the formulation of policies and strategies of the Company. More detailed and technical issues will be handled by Board Committees, which would receive delegated authority to decide on certain issues and to recommend the more important issues to the Board for approval. The following Board Committees and organization structure are suggested: a) Board Audit Committee - Reviews the audit of the Company's operations as well as the effectiveness of the financial and internal controls. The Internal Auditor and External Auditors will report directly to the Board Audit Committee. b) Board Business Committee - Responsible for overseeing the business operations of the Company, both the refinancing and the funding. c) Board Risk Management Committee (BRC). Overall responsibility to ensure there is proper oversight of risks of the Company. The Board sets the risk appetite and tolerance level that are consistent with the Company's overall business objectives and risk principles. The BRC assists the Board by ensuring that there are effective oversight and development of strategies, policies, and procedures to manage risks. The Chief Executive Officer (CEO) should have an entrepreneurial flair with good knowledge of risk management and the bond market. Knowledge in housing finance will be a critical advantage. The CEO preferably should be an experienced banker/investment banker who is well respected within the banking community and by all the authorities. Charter The NMRC will be a licensed non-deposit-taking institution authorized by the CBN to conduct its business, and by the Securities and Exchange Commission (SEC) to issue bonds. The Charter of NMRC reflects its unique role Page 73 The World Bank Housing Finance Development Program (P131973) in being a single-purpose vehicle. The charter restricts the Company from undertaking any other activities (such as taking deposits, commercial loan refinancing, or lending directly) apart from the ones for which the Company is established. The restrictive charter is to ensure that the Company focuses on supporting the growth of the mortgage market and the bond market by limiting its activities to the refinancing and prefinancing of mortgage loans and the issuance of bonds in support of these activities. IDA's Financing Agreement will include a negative covenant/remedy that the NMRC Charter will not be amended without IDA's prior written approval. Regulation The main regulator for NMRC is the CBN. NMRC will be licensed as a non-depository financial institution with clear limitations on its scope of business limiting it to its core activity of refinancing mortgages. As NMRC seeks to fund itself in the bond market, it also will have to be licensed and regulated by the SEC as a bond issuer. C. Mortgage Guarantee Product To further support reach of mortgages down the income distribution, US$25 million equivalent was designed to develop a guarantee product to bear some of the credit risk for lower income lenders. The guarantee will provide credit default loss protection to mortgage lenders. It will be established in year 2 of the project and will be subject to detailed pricing and market analysis. The project will fund product development including testing and financial modeling and then will launch a competitive process among private or publicly owned non-bank financial institutions to implement the project. Criteria will include: (a) being financially sound and solvent as confirmed by the CBN; (b) providing a matching contribution to the proposed product; and (c) submitting a business plan based on scope and outreach to lower income groups. In addition, a detailed section will be added to the POM that will describe pricing and operations of the guarantee product and how the relationship with the implementing firm is to be managed. The project will finance the TA required to design the guarantee product, the development of the detailed operation manual, the bidding criteria, and the consultations events; and will support the implementation and monitoring. The TA financing is allocated under Component 4. Furthermore, the project will make available a line of credit to cover the issuances of guarantees. The guarantee line of credit allocation will be made available upon the signing of the agreement with the service provider. The selected service provider will issue a quarterly report with the list of issued guarantees and the guarantee terms, including social income data and GPS location of the mortgagor. D. Housing Microfinance The proposed objective is to support the development and piloting of new and/or emerging formal Housing Microfinance (HMF) products and to demonstrate a sustainable business case for this activity. The activities under this component will include (a) tailored TA to select microfinance banks (MFBs) to test the new HMF products, develop new and/or improve emerging HMF products (to be funded from component 4; (b) provision Page 74 The World Bank Housing Finance Development Program (P131973) of risk capital to MFBs, in form of limited soft loans; and (c) a line of credit to MFBs to roll out the tested HMF products. E. Project Implementation Arrangements A Project Steering Committee (PSC), chaired by the Federal Ministry of Finance will supervise project implementation. The involved ministries, regulators, states, and the banking sector will be represented in the PSC through senior level representatives of at least director level. The PSC will be supported by the Secretariat of the Steering Committee (SSC) based in the Ministry of Finance. The SSC will report directly to the PSC and support the PSC in its supervisory role. The day-to-day operations will be managed by a Project Administration Team (PAT) hosted and funded by the CBN, which will report quarterly to the PSC. Project Administration Team (PAT). A PAT will be established in the Other Financial Institutions Department (OFISD) of the CBN and be funded by the CBN. The OFISD also is responsible for the regulation and supervision of PMBs, MFBs, and NMRC and therefore is the optimal location for the PAT. The PAT will be led by a project director and a deputy director. The project director will work closely with the SSC. Other members of the PAT will include a procurement specialist, project accountant, M&E specialist, and sustainability specialist. The PAT will be supported on a needs basis by a procurement specialist of CBN's procurement department and a financial management specialist of CBN's financial management unit. The PAT will be fully embedded in the CBN; therefore, contracts, and reports will be approved by the governance structure of the CBN. The PAT and the secretariat will be the main contacts with the World Bank project team. The PAT in close collaboration with the Secretariat will develop a full operational plan before the project becomes effective. The primary responsibilities of the Project Administration Team are to: * Provide overall project management of the project. * Collect data together with OFISD and CBN's statistical department of NMRC's operations and assess their impact. * Supervise NMRC's compliance with the Environmental and Social Risk Management Operations Manual (ESOM). * Procure consultants for a market analysis and TA for the MFBs. * Consolidate reports to the World Bank, including quarterly financial reports and half-yearly and annual implementation progress reports aligned to agreed project indicators and synchronized with half-yearly Bank review missions. * Contract and coordinate with a consulting firm to conduct the annual technical audits. Page 75