Expanding Access to Housing Finance. Task 1: Business and Sustainability Plan for Affordable Housing Finance Pilot Projects FINAL REPORT

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Expanding Access to Housing Finance Task 1: Business and Sustainability Plan for Affordable Housing Finance Pilot Projects Prepared for: FIRST Initiative Prepared by: Sally Merrill, Urban Institute Duncan Kayiira, UMACIS William Kalema, UMACIS Submitted by: FINAL REPORT Public Disclosure Authorized The Urban Institute UMACIS Consulting 2100 M Street, NW 29, Luthuli Avenue, Bugolobi Washington, DC P.O. Box 9113 (202) Kampala UGANDA JULY 2009

2 Table of Contents 1.0 Summary and Overview Features of the Pilot Projects Loan Product Development for the Pilot Projects Recommendations for Pilot Implementation and Long-term Sustainability Annexes Annex Annex I: Cost Detail for the Illustraitve PIlot Projects 19 Annex II Illustriative Housign Design for the Pilot Projects.25 Annex III: Ilustrative Loan Development Forms for the Pilot Projects Annex IV: Profiles of Potential Developers for the PIlots Annex V: Profiles of Potential Lenders for the Pilot Projects List of Tables and Boxes Box 1.1: Inter-related Goals... 3 Table 2.1: Overall Summary of Pilot Designs... 6 Box 2.1: Demonstrating the Incremental Building Approach... 7 Table 2.2: Potential Pilot Partners: Financial Institutions and Developers... 8 Table 2.3: Design and Materials Specifications... 8 Table 2.4: Cost of a 500m2 Plot in selected urban areas near Kampala City... 9 Table 3.1: Illustrative Features of Potential Housing Pilot Loan Products Table 3.2 Loan Design Features for Microfinance for Housing (MFH) Table 3.3 Loan Design Features for Mini-mortgages Table 4.1 Advantages and Disadvantages of the Pilot Projects i

3 1.0 Summary and Overview 1.1 Overview of the FIRST Project in Uganda The objective of the FIRST Initiative project in Uganda is to expand the access of households to housing finance, especially modest and lower income households, by introducing new and innovative housing loan products, by introducing innovative loan products combined with affordable housing designs. The project has delivered two studies to the Bank of Uganda: i) A study addressing the financial and banking sector context for housing finance, liquidity and liquidity management, and the resultant potential for use of retail funds for mortgage lending; and ii) A feasibility study for housing finance pilots targeted at modest and lower income households, seeking to introduce innovative loan products combined with lower cost house design in a planned urban setting. These companion studies, designed to be read together, provide (i) policy advice on liquidity and the regulatory framework for mortgage banking, and (ii) preliminary recommendations for pilot housing finance projects. A summary of the two reports is provided below. (i) Examining the Use of Retail Funds in Mortgage Lending. Assessing the ability of the banking sector to enable more long and medium term funds to be devoted to housing finance is addressed in the Study to Examine the Use of Retail Funds in Mortgage Lending. 1 The paper provides an overview of the conditions facing mortgage lending, including macroeconomic constraints and the liquidity profile of mortgage lenders. A series of recommendations on regulatory structure for bank and MDI have been provided for BOU. The banks, MDIs, and MFIs most likely to lend to the modest and lower income groups are currently those with least liquidity. Thus, among other findings, the results support the need to provide enhanced liquidity for low and modest income housing, and a housing finance liquidity facility is recommended as a long-term goal. In the short-term, relaxation of the two-year lending cap imposed on MDIs will be necessary for the pilots. (ii) Design of Pilot Projects for Modest and Lower Income Households. 2 Expanding housing and housing finance in Uganda faces numerous barriers from both the demand and supply sides. Drawing on findings from the first paper, as well as examples of low income housing development worldwide, the paper addresses the design of pilot projects to provide housing finance and low cost housing to modest and lower income groups. The paper provides an overview of constraints to low income housing development and housing finance, and suggests pilot projects for assisting two groups that lack access to housing finance in Uganda: (1) the modest income group, or the missing middle, which is under the radar screen of banks providing mortgage loans, and (2) low-modest income households who do not have access to either mortgage loans or microfinance for housing. 1 See Richard Ketley, Jessica Kramer, Michael Hanouch, and Christo Wiese, Expanding Housing Finance in Uganda: Task 2, Study to Examine the Use of Retail Funds for Mortgage Lending, Genesis- Analytics for the FIRST Initiative, June See Sally Merrill, William Kalema, and Duncan Kayiira, Design of Pilot Projects for Modest and Lower Income Households, The Urban Institute and UMACIS for the FIRST Initiative, June

4 This paper Business and Sustainability Plan for Affordable Housing Finance Pilot Projects - provides supporting technical detail for the feasibility study for Design of Pilot Projects for Modest and Lower Income Households, including the need for more liquidity for lenders involved in lending to lower income households. Two types of pilot projects have been developed: one for starter homes for modest income households and one for an incrementally built home for low-modest income households. The recommended loan products include a downmarket or mini-mortgage for the starter home and microfinance for housing for the lower income group. It is recommended that one of the two pilot schemes be developed. Both groups lack access to housing finance in Uganda: The modest income group, or the missing middle, is just under the radar of banks providing mortgage loans, while the low-modest income households, which make up a large proportion of Uganda households, do not yet have adequate access to microfinance for housing. The paper (i) outlines technical details regarding house design and loan product specifications; (ii) recommends the technical assistance and regulatory changes deemed necessary to implement successful pilot; and (iii) provides a brief commentary on actions that would assist such pilots reach sustainability and scale. 1.2 Goals of the Pilot Projects The proposed pilots address constraints to low income housing development from both the demand and supply sides. A major focus is the development of affordable housing loan products, such as microfinance for housing and mini-mortgage. The pilots also focus on the supply side, with the goal of joining new loan products with lower cost housing, so that the overall package is feasible and affordable. The target household groups with monthly income in the range of UGX 0.7 million ($350) and 2.3 million ($1150) - have had limited access to affordable starter homes or incrementally built structures, as the few developers in the country have focused mainly on the higher income market. There is considerable interest in supporting one or more pilot projects in Uganda. A number of lenders have expressed interest in offering new loan products and participating in the pilots. Similarly, several developers have expressed their willingness to provide affordable housing, and both the developers and Ugandan architects have helped the team provide a number of options for house size, building materials and finishes, thereby providing a good range of affordability. Importantly, the recommended pilots will be entirely market-based: they will not depend on finance subsidies, or on funds for land or infrastructure. This will help ensure the sustainability of the model and enable it to be replicated and scaled up without the need for large budget allocations or donor funds. The key criteria for the housing pilot designs include the following: The pilots must serve households with limited or no access to either housing finance loan products or affordable housing options; The house designs must be affordable to the targeted groups; The pilots will be market based, and not dependent on subsidies of any kind financial, land, or infrastructure; and To help support sustainability, the affordability package must represent a compatible match among four key factors: i) the income of target households; ii) housing designs affordable to these households; 2

5 iii) affordable loan products relevant to the targeted households and the housing costs; and iv) be in accordance with the preferences of the target households linked to what the household can afford. The affordability package diagram below represents these inter-related goals. Box 1.1: Inter-related Goals 1. TARGET HOUSEHOLDS: THOSE NOW WITHOUT ADEQUATE ACCESS Effective Demand Housing is affordable to target households Microfinance for Housing Loans & Minimortgages Household preferences addressed in affordable design 2. LENDERS: AFFORDABLE LOAN PRODUCTS Loan appropriate to house cost and construction mode Affordable design given income and loan 3. DEVELOPERS: AFFORDABLE HOUSE DESIGNS: STARTER & INCREMENTAL 1.3 Summary of the Design Process for the Pilot Projects Pilot House Design. The Business and Sustainability Plan seeks to provide the ground work for finalization of both house design and loan products for three representative house designs. As noted, two basic household groups will be targeted: the modest income Ugandan households (monthly income UGX million) and the low-modest income households (monthly income of UGX million). 3 For the modest income pilot, two house designs were developed starter houses of 80m2 and 95m2, while for the low modest income group an incremental self-build house of 60m2 has been detailed. The building materials and cost estimates for all three designs are provided in Annex I. Annex II provides the design layout for the 60m2, 80m2 and 95m2 designs. 3 As discussed in the Design Report, the available income distribution represents nation-wide income, and is not broken down by urban and rural households. It would be desirable to develop an income distribution able to describe urban Ugandan households, in order to have a better idea of where in the distribution the pilot affordability falls. As noted below, it is recommended that improved income data be obtained in order to better support the affordability analysis. 3

6 The design variables include the following, with numerous details provided in the specifications in the Annexes. Lot size House size Roofing material Building materials Floors, windows, and doors Exterior and interior finishes Sanitation Utilities and access infrastructure Starter home or Incremental building approach The house design process has involved extensive discussions with Ugandan developers and architects. Construction would take place on the developers planned estates near Kampala, with house design tailored to the affordability constraints for the targeted households. Please note that the house designs are not necessarily final, but represent pivotal or illustrative designs that can be used to determine an optimal combination of features for size, layout, and building materials. These elements can ultimately be mixed and matched against affordability and household preferences to find the most viable set of characteristics for successful pilots. Pilot Loan Design. The loan design suggestions for microfinance for housing and minimortgage loan products have been assisted via discussions with numerous lenders, including banks, MFIs, and MDIs. 4 An illustrative loan design form is provided in Annex III. However, specification of exact loan terms will require additional information on both prospective buyers, particularly on their income level and sources, as well as the quality of the house designs and locations. This information will be made available by the recommended focus groups and market surveys and through assessments by the lenders of the building sites, materials, layouts, and so forth, in order to better determine collateral value. 1.4 Summary of Recommendations Design of Pilot Projects. It is recommended that one of the above two pilot schemes be developed. As mentioned earlier, both groups the modest income and low-modest income lack access to housing finance in Uganda. The modest income group, or the missing middle, is just under the radar of banks providing mortgage loans, while the lowmodest income households, which make up the vast majority of households, do not yet have access to microfinance for housing. Regulatory Change. It will be crucial that a regulatory change be made regarding the two year cap on MDI lending. While the two year limit on MDI loans may be relevant for SME lending to very low income households, it will severely curtail development of microfinance for housing (MFH) loan products, which, worldwide, tend to have longer terms than SME loans. MFH loans often range between 3 and 7 years; they may be provided as a series of loans, for example, a number of sequential 3 to 5 year loans to finance an 4 Please see the interview list in Annex I of the Design Report. 4

7 incrementally built structure. It is recommended that the cap be lifted or relaxed, at least for MFH. The loan designs will ensure that the loan terms and underwriting follow very prudent standards. Technical Assistance to the Pilots. Numerous types of technical assistance--including additional loan product design, affordable housing design, and market surveys to determine affordability and test acceptance of lower cost housing units--will be important to the success of the pilots. Land Use Planning and Infrastructure Provision. Intensified GOU efforts in land use planning and delivery of infrastructure are highlighted for expanding low income urban housing as a means to both address the current informality in the housing sector as well as curtail future formation of slums and informal settlements. Increased Liquidity in the Longer-Term. In the long run, it is expected that in order to achieve any degree of sustainability and scale, a greater volume of funds will be needed to provide loans to modest and low-modest income households. However, as noted, there appears to be a mismatch between the liquidity profiles of Uganda s banks and the banks willingness to develop down-market lending. Several international banks are not growing their mortgage books. However, these banks are relatively more liquid than the smaller local and regional banks, and it is from this latter group that there is a willingness to undertake housing lending and explore new avenues. Thus, developing a liquidity facility for Uganda, although not immediately on the agenda, may be an important consideration for the future. In addition, MFH products should be included in such an effort (a liquidity facility along similar lines is to be tested by the World Bank in Tanzania in ). 5

8 2.0 Features of the Pilot Projects As noted, three illustrative housing designs are presented in this report. The original designs presented in the Design Report 5 have been modified, and the pilots now focus on smaller dwellings and smaller lot sizes -- in order to increase affordability and to set a pattern for more cost-effective long-term use of land in Uganda. For the modest Income group pilot, two size and design choices are tentatively proposed: 80m2 and 95m2 starter homes; For the low-modest income group, one design has been detailed: a 60m2 incrementally built home. As currently designed, the pilots would be undertaken in planned estates in urban areas near to Kampala. Table 2.1 summarizes the key features of the pilots. The cost of the land, which is owned by developers who have offered to participate in the pilots, is included in the total. Box 2.1 illustrates a possible approach to the incremental building approach suggested for the 60m2 house. Table 2.1: Overall Summary of Pilot Designs Housing Pilot Features Modest Income Group Low-modest Income House Size Options 80m2, 95m2 60m2 House Type Starter or modest home Incremental Construction All-in Cost 80m2: UGX 44.6 million ($22.3k) 95m2: UGX 53.6 million ($26.8k) 60 m2: UGX 20.8 million ($10.4k) Construction Cost 80m2: UGX 34.9 million ($17.5k) 95m2: UGX 42.4 million ($21.2k) 60m2: UGX 15.9 million ($7.95k) Construction Cost per m2 80m2: UGX 436k ($218 per m2) 95m2: UGX 446k ($223 per m2) 60m2: UGX 265k ($ per m2) Lot size 200M2 150M2 Land cost per m2 UGX 10,000 ($ 5) UGX 8,000 ($ 4) Loan Type Down-market/mini mortgage: 7-12 years Microfinance for Housing: repeat loans of 3-5 years Loan Terms 7-12 yrs., mortgage rate + 1-2% 3, 5, or 7 years, maybe same as microfinance rate Household Income UGX million monthly ($ ) UGX million monthly ($ ) Technical Assistance House design, underwriting informal income, market surveys House design, MFH loans, underwriting, market surveys Other Subsidies None None This report also provides considerably more detail on design and building materials, providing illustrative variations on costs, given alternative layouts, finishes, sanitation, and other features. The pilots seek to provide a completed house either a starter home or via incremental or 5 The Design Report discussed a larger number of potential pilot designs to address the needs of two household target groups. However, this report details only 3 designs and has made a number of modifications - decreasing the lot size and house size and using more modest building materials for the 60m2 house all in order to enhance affordability. 6

9 modular design. All are to be built without subsidy. As a result, Uganda s lowest income households are not targeted in this particular effort: households with income significantly lower than the current target groups generally require subsidies to achieve a complete house - at a minimum to obtain the necessary urban land and infrastructure. Box 2.1: Demonstrating the Incremental Building Approach The 60m 2 House 1. Stage One: At this stage, prospective homeowners would buy land at UGX 1.7 million ($ 850), including the cost of infrastructure and titling fees, pay for the house design at UGX 0.2 million ($ 100), and undertake the initial stages of construction; laying the foundation and building the concrete slab, priced at UGX 2.1 million ($ 400) and building part of the wall, priced at UGX 2 ($ 1,000). 2. Stage Two: This stage will involve completing the wall, at UGX 1.3 ($ 650), roofing the house, priced at UGX 2.4 million ($ 1,200), cementing the floor, priced at UGX 0.23 ($ 115), fitting windows and doors, priced at UGX 1.2 ($ 600), plastering inside and outside the house, priced at UGX 1.2 ($ 600), constructing the ceiling, priced at UGX 0.46 ($ ), painting the cost, at UGX 1 million ($ 500), installing water pipes, a shower basin and electricity, priced at 1.2 and fitting the kitchen, priced at UGX 200,000 ($ 100). 3. Stage Three: This stage will involve, among other finishes, constructing a 2,000 litre septic tank, priced at UGX 400,000 ($ 200) and paving the road to the house, priced at UGX 0.9. ($ 450). 2.1 The Pilot Design Process Successful implementation of the pilots will need local champions. Table 2.2 indicates the Ugandan partners who have not only expressed willingness to participate in the pilots, but have been generous in providing time and effort for specification and re-specification of pilot options. In sum: Ugandan developers are engaged: The project has succeeded in finding champions for both types of pilots, including both lenders and developers as noted in table o o Modest income pilot: Akright Project, Low-modest income pilot: Jomayi Property Consultants, Ugandan lenders are engaged both banks and MFIs o o Modest income pilot: DFCU Bank, and Housing Finance Bank (HFB) Low-modest income pilot: Centenary Bank, FINCA Uganda Ltd and PRIDE Uganda Microfinance Ltd. Ugandan architects are engaged: Skil-plan Architects and Eco-Shelter assisted UMACIS in both design and costing. 6 Refer to Annex IV for a brief sketch of the developers who have expressed an interest in participating. 7

10 Table 2.2: Potential Pilot Partners: Financial Institutions and Developers Components Modest Income Pilots Low-Modest Income Pilot Potential Financial Institution Partners Potential Development partners Land and Infrastructure Architecture and Cost Consultants DFCU and HFB, Developer: Akright Architect: Eco-Shelter Urban estate development, existing, surveyed land parcels owned by the developer Skil- Plan Architects; Eco-Shelter Centenary Bank, PRIDE, FINCA Developer: Jomayi Architect: Eco-Shelter, Construction Firms Urban estate development, existing, surveyed land parcels owned by the developer Skil- Plan Architects Illustrative House Designs and Alternative Materials. The houses are designed to accommodate the affordability constraints for the targeted households, but represent to the extent possible the revealed preferences of typical Ugandans in the respective target groups. As has been discussed, no subsidies would be involved in lending or construction. As a result, Uganda s lowest income households are not targeted in this particular effort: households with income significantly lower than our target groups generally require subsidies to achieve a complete house - at a minimum to obtain the necessary urban land and infrastructure. Table 2.3: Design and Materials Specifications Design & Materials Features Low-Modest Income Pilot Modest Income Pilot House Design 60m2 House 80m2 House 95m2 House Prototype Design Incremental Starter home Starter home Lot size 150m2 200m2 200m2 Walls Clay bricks Concrete bricks Concrete bricks Floor Cement screed Ceramic tile Ceramic tile Windows Casement Casement Casement Doors Hardwood Hardwood, steel Hardwood, steel Plastering Internal & external Internal & external Internal & external Paint 3-coat 3-coat High finish Roof Pre-coated iron sheet Kajansi tiles Kajansi tiles Ceiling Timber, mesh Timber, mesh Timber, mesh Water Piped, indoor Piped, indoor Piped, indoor Plumbing Flush toilet complete facilities Flush toilet complete facilities Flush toilet complete facilities, high finish Sanitation Septic tank Lagoon Lagoon Kitchen Basic, indoor Sink, cabinets, tiled Sink, cabinets, tiled, high finish Electricity Installed Installed Installed Access Road Access to paved road Access to paved road Access to paved road 8

11 Table 2.3 summarizes the main design features and construction materials of the three prototype pilots (again, details are found in Annexes I and II). As is evident, there are some major differences in materials and approach, particularly in the following inputs which have a major impact on costs: Roofing material (iron sheet vs. tile) Building materials (clay bricks vs. concrete bricks) Flooring( cement vs. tiled) Kitchen (basic vs. complete and tiled) Incremental building approach vs. starter home Sanitation (septic tank, common lagoon) Lot Size and Land Costs: The proposed lot size in the pilots deserves some discussion. The plot sizes recommended for the pilots 150m2 for the 60m2 house and 200 m2 for the 80m2 and 95mr homes, are considerably smaller than the 500m 2 plots now offered in the Akright and Jomayi developments. Ultimately, in the longer-term, from both affordability and developmental standpoint, it is desirable for most housing to utilize relatively small amounts of land. Clearly, smaller plot sizes will impact the overall affordability of housing. More importantly, however, in the long-run, it would be desirable for Ugandans to focus on smaller plots, most especially in urban areas, to make cost-effective use of land. Together with land use planning, titling, and high-rise and row house construction, more intense and organized use of urban land can help forestall expansion of slums. Thus, if one or more of the current pilot plans are implemented, further discussion as to lot size will need to take place. Table 2.4: Cost of a 500m2 Plot in selected urban areas near Kampala City Location Distance from Kampala City Centre (km) Cost of a 500m 2 (UGX, millions) Cost per m2 (UGX, 000) Luzira Bweyogerere, along Jinja Road Namugongo, along Jinja Road Nansana, along Hoima Road Seeta, along Jinja Road Mukono, along Jinja Road (Jomayi) Kasangati, along Gayaza Road Bulenga, along Mityana Road Bwebajje, along Entebbe Road (near Akright s Kakungulu Estate) Abaita Ababiri, along Entebbe Road During the last decade, developers, including Akright and Jomayi, bought large tracts of land and typically subdivided them into 500m 2 plots. In their opinion, this size is cost effective in the delivery of infrastructure (roads, water and electricity), and allows for orderliness within the housing estates. In addition, this plot size has been readily mortgaged by various banks. A 500m 2 plot in Jomayi s Mukono estate, 15km from Kampala City, where the low-modest income pilot would be undertaken costs UGX 4 million ($ 2,000). At Akright s Entebbe road estate, 18km from Kampala City, where the modest income pilot would be undertaken, the same plot size costs UGX 5 million ($ 2,500). With other developers like Hossana Real Estates Ltd and CANAANSITES Ltd, who sell land in the same area, the cost of a 500m 2 -plot is higher by about 30 percent. The difference in the costs is attributable in part to economies of scale achieved by 9

12 the Jomayi and Akright developments; also. both developers purchased these very large tracts of land some years ago. More generally, the cost of land depends, of course, on the desirability of an area, its level of development, and the distance from a major city. As seen in table 2.4, the cost per m2 is generally less as the distance from Kampala increases. However, the cost of 500m 2 plots along Entebbe Road (Bwebaije and Abaita Ababiri, 18km and 30km from Kampala) is relatively more, given the desirable location on the main Kampala/Entebbe access road. 10

13 3.0 Loan Product Development for the Pilot Projects. As discussed, the households targeted for the pilot currently lack adequate access to both housing finance and affordable dwellings. However, the income levels of the target groups suggest that they will be able to afford modest homes. Two distinct types of loans are being proposed for the pilots: Mini-mortgage. The mini-mortgage is a collateralized mortgage loan product expected to be reasonably similar to the conventional mortgages now being offered by a number of banks in Uganda. Mortgage lending is a mainstay of the proposed pilot partners DFCU and HFB. The risk profile of the customers for mini-mortgages is likely to differ somewhat from that of current customers, however, especially with regard to the level of income, the availability of a credit history, and the likelihood of having to underwrite some amount of informal income in addition to salary income. This risk profile may suggest a somewhat shorter term and slightly higher rate than conventional mortgage loans. The participating lenders have initially suggested a mark-up over the going mortgage rate of 100 to 200 basis points. Microfinance for housing (MFH). Worldwide, MFH loans are most frequently used for home improvement, as they are generally much smaller and shorter-term than mortgages. Since the proposed low-modest income pilot is geared toward completion of a complete home built in three incremental stages, it is proposed that a MFH loan product be developed to support this approach. The MFH loans would be geared to the 3 proposed stages of development, with clear quality control milestones for each stage that the lender would monitor. More work with the participating lenders will be needed to establish a fair but affordable rate. One MFI lender has initially suggested the current microfinance business loan rate as a reference point. This tends to be quite high - e.g. 2% per month - which would certainly reduce affordability in the pilot. This interest rate is not at all unusual for SME loans, particularly those with fairly short terms i.e. 6 to 24 months - and servicing businesses from the informal sector. However, the MFH loans proposed for the pilots are expected to be somewhat longer term and provided to borrowers with higher income than may be the case for current SME loans. It is expected that the target households will have some measure of salary income and could service larger loans than the micro business loans. The MFIs/MDIs must also be helped to become comfortable with the approach: an incremental building methodology resulting in a complete house, which ultimately provides collateral value. In addition, a savings plan approach may be suggested for some would-be borrowers to show ability and willingness to pay and/or to provide a substantial downpayment. Thus, it is expected that a rate somewhere between the micro-business loan and the minimortgage could prevail for the pilot. 11

14 Table 3.1: Illustrative Features of Potential Housing Pilot Loan Products Components Modest Income Group Low-Modest Income Group Pilot Lenders DFCU, HFB Centenary, PRIDE, FINCA Target Groups Monthly income: UGX millions ($) ($ $1250) ($350 - $650) House Type Starter Home Incremental/Modular House Size Options 80m2, 95m2 60m2 Lot Size 200m2 150m2 Type of Loan Product Mortgage Loan Housing Microfinance Loan Use of funds Purchase of a lower priced home from a developer, with 20% to 30% down payment. Incremental construction of a modest home using a series short-term loans of 3 to 5 yrs Loan to Value Options 70% or 80% LTV 70% or 80% LTV Loan Term (Years) Interest rate (%) BP over mortgage rate, or TBD Microfinance lending rate for proven customers, or TBD Discussions with the participating lenders are ongoing. These lenders have either already initiated loan products with some of the relevant characteristics or are in the process of designing and testing them. DFCU staff have indicated that they are already going more downmarket. HFB has a house that grows loan product, although management notes that it has not been as successful as expected. Centenary currently offers short-term loans for home improvement, while both PRIDE and FINCA are now in the process of introducing microfinance for housing loans. Table 3.2 Loan Design Features for Microfinance for Housing (MFH) Loan Parameters Underwriting House Quality LTV: Maximum & minimum loan amount for 3 stage, 60m2 Incremental Building Options for term of each repeat loan: 3 to 5 years (assuming 2-year cap is removed) Interest Rate: Rate relative to new 2 year MFH home improvement loans now being developed? Rate relative to mini-mortgage rate? Will the interest rate vary by? Term? LTV? Credit history, underwriting score? Previous microbusiness loan? Will term vary by income, credit history, loan-to-value, payment to income, previous loan history? Maximum payment to income ratio as % of gross income? as % of net income? Prior successful micro-loan required? Savings required? If yes, what period and amount? If yes, is savings balance kept for life of loan? Process to determine informal income? Documents required? Co-signors required? Other collateral required? Title required? Bank evaluation of house location, land, infrastructure Bank evaluation of incremental design Bank evaluation of quality of materials Repeat Loans: Quality control procedures for achievement of stages in design prior to granting the next loan in the sequence 12

15 Illustrative loan product design parameters are summarized in table 3.2 for a MFH loan and in table 3.3 for a mini-mortgage. The share of informal income in total income is one key parameter, for example, that might distinguish clients of conventional mortgages from those of mini-mortgages. MFH loans, expected to be on a repeat basis for an incremental building approach, are for relatively longer terms than most micro business loans. This distinguishes them from the existing portfolio for lenders such as PRIDE and FINCA. And as discussed above, the MFH loan product proposed for the pilot also differ from the shorter-term home improvement MFH product now being introduced by these lenders. In sum, the participating lenders have indicated that the final loan terms will depend on the results of further knowledge of the prospective borrowers, decisions regarding the underwriting process, and the location and quality of the housing. This will help establish the borrowers presumed risk profile (how does it differ from that of current clients); the costs of underwriting and servicing (are they greater than currently and if so, how?), and the ability to sell any foreclosed housing or land (including partially built structures). Annex III provides a chart that could be used for discussion with the lenders in developing the new loan products for the pilots. Table 3.3 Loan Design Features for Mini-mortgages Loan Parameters Underwriting House Quality LTV: maximum loan amount for 80m2 or 95m2 starter home at costs of UGX at cost of X Maximum payment to income ratio as % of gross income Bank evaluation of house location, land, infrastructure Maximum Term Prior loan required? Bank evaluation of house quality and design Interest Rate: what mark-up over mortgage rate, if any? Other indicators Will the interest rate vary? If yes, according to: Term? Credit history, underwriting score? LTV? Other? Does term vary by income, credit history, loan-to-value, payment to income, etc. Savings required? If yes, what period and amount? If yes, is savings balance kept for life of loan? Process to determine informal income? Documents required? Co-signors required? Title required 13

16 4.0 Recommendations for Pilot Implementation and Long-term Sustainability of Lower Income Housing Development It is recommended that either the low-modest income pilot or the modest income pilot be chosen for the demonstration pilot. The rationale for going forward with one of the pilots includes the following: Willing lender, developer, and architect partners. Both of these pilots appear feasible, and discussions with potential partners have been positive. For the modest income, DFCU and HFB have stated that they are willing to participate in a pilot project. For the low-moderate income group, Centenary Bank. PRIDE and FINCA are willing to participate. Developer partners include Akright and Jomayi, while the architectural assistance for house design and costing was offered by Eco-Shelter and Skil-Plan. Prototypes for affordable housing and appropriate loan products. The pilot projects will provide demonstrations of new approaches to affordable housing, combined with new loan products geared to the target income groups and the housing costs. Ugandan development has heretofore concentrated on the higher end of the market. Land and infrastructure are already integral to the projects. Obtaining serviced land is often one of the most difficult aspects of lower income housing development. Fortunately, the developer partners already own large urban planned estates, which will speed up the pilot process. Subsidies are not required. As noted above, land and infrastructure are integral to the pilots. In addition, no subsidies are required for construction finance or take-out loans. This will be an important step toward proving the sustainability of similar low and modest income lending and building schemes. Section 4.1 provides recommendations regarding the technical assistance that would be necessary to support implementation of the pilots. Section 4.2 provides further recommendations aimed at the long-term sustainability of low and modest income lending and affordable housing development. 4.1 Recommendations for Technical Assistance to Support Pilot Implementation While substantial progress has been made in pilot housing development and loan design, more effort is needed for a number of aspects of the proposed pilots. Thus, although there are no direct subsidies involved, it is expected that technical assistance will be needed to support the further development of loan products, underwriting methodologies for households with informal income, and affordable house designs. Importantly, market surveys will be needed to bring the elements of the pilot together that is, to match household affordability and preferences with size, design, and materials. Pilot Implementation Technical Assistance. The affordability diagram presented in section 1.0 emphasized the inter-related nature of the design features of the pilots. As noted there, 14

17 conformance must be reached among household income, housing cost, loan design, and household preferences in order to have both marketable and affordable homes. Clearly, this design effort cannot be done in the abstract and a number of design variables need to be considered simultaneously. Technical assistance is desirable to help refine the following: MFH and mini-mortgage loans, at rates appropriate to risk profiles, with prudent standards regarding loan-to-value and payment-to-income ratios. Underwriting and servicing methodologies should be suited to modest and lower income borrowers The tenor is either medium-term, suitable for moderate income groups, or shorter, MFH loans, which may be repeated to accommodate incremental house design. As noted, setting the interest rate and other terms for the MFH pilot loan will need further analysis, as this product is entirely new to Uganda and not yet common practice worldwide because of the incremental build/repeat loan requirement. Stromme could also assist with housing microfinance loan products, as these efforts are to be undertaken region-wide. 7 More information is needed on income formal, informal, and variable. The willingness and ability of banks to underwrite informal income will be an important consideration; More knowledge regarding Ugandan household preferences, important to develop tradeoffs among size, alternative building materials, quality of finishing, plot size, and type of infrastructure (especially sanitation); Review and finalize the house designs for expandable starter homes, modular designs, and designs for incremental building, and Help Ugandan lenders analyze the demand for these housing products i.e. that there is adequate effective demand (ability and willingness to pay) and long-term collateral value. Focus Groups and a Market Survey to Assess Income and Preferences. Ideally, an affordability analysis should be done of income (by type and source) for Kampala, or at least for urban areas. However, the available income information in Uganda is not sufficiently detailed to provide the requisite distribution. Focus groups and market surveys would assist in determining the income levels and income sources for the relevant portion of the urban population. In addition, Akright, for example, has focused on various professionals (teachers, police, military) as candidates for modest housing, and the distribution of income for these groups could be analyzed. In addition, a typical description of Ugandan housing preferences has suggested a desire for large lots, numerous bedrooms, and high end materials and finishes. Clearly, only a small fraction of Ugandan households can afford this type of home. The preferences of more modest income households, however, do not seem to be well understood. Again, focus groups and market surveys can assist this process. In sum, accurate information on the inter-active process that links income and affordability with specific house designs and specific preferences -- so that the package works in the sense that respondent links what they can afford to pay with a loan product and a preferred house design is notoriously difficult to collect. Respondents often tend to prefer homes that they cannot afford; they also are reluctant to state their full income. Similarly, they are reluctant to reveal their savings. Monthly loan payments need to be linked to a specific dwelling, and income must often be estimated via questions on expenditures. 7 Stromme Foundation, Proposed Terms Of Reference For A Feasibility Study For A Micro Housing Facility For The Poor In East Africa 15

18 Thus, it is recommended that one or more market surveys be done, and that they be preceded by focus group discussions in order to get the questions and the question sequence in the best order and phrasing. Based on the focus group findings, the survey would include house diagrams and affordability tables to assist the respondent in making realistic choices linking the following: diagrams and descriptions of various house designs, including incremental; house prices linked to these descriptions; monthly payments by loan size and terms, again, linked to specific designs monthly income needed to make the payments; savings needed for downpayments; a process for iteration with the respondent so that their preferences reflect what they can actually pay for: a realistic preference set and estimates effective demand Regulatory Change. It will be crucial that a regulatory change be made regarding the two year cap on MDI lending. While the two year limit on MDI loans may be relevant for SME lending to very low income households, it will severely curtail development of microfinance for housing (MFH) loan products, which, worldwide, tend to have longer terms than SME loans. MFH loans often range between 3 and 7 years; they may be provided as a series of loans, for example, a number of sequential 3 to 5 year loans to finance an incrementally built structure. It is recommended that the cap be lifted or relaxed, at least for MFH, while ensuring that household loan terms and underwriting follow very prudent standards. Savings Products Linked to the Pilots. A recommendation to increase savings throughout the financial sector is noted below. However, in the short-term saving products may assist the pilots, as they have proved elsewhere to be crucial components of underwriting for both microfinance and microfinance for housing lending. As lenders observe steady savings habits, those with informal and/or variable income can be observed with regard to ability to pay. The savings efforts bolster ability to provide lenders with the required down payment, often the major barrier in low income housing lending. Also, for those with little other collateral, some portion of the savings can be held by the bank until the loan is paid. 4.2 Longer-term Sustainability and Scale The pilots should be designed to maximize sustainability and ultimately reach for scale. As has been discussed, the recommended pilots are market-based: they do not depend on finance subsidies, or on funds for land or infrastructure. This will help ensure the sustainability of the model as it enables the pilots to be replicated and scaled up without the need for large budget allocations or donor funds. Economies are achieved via innovative design, alternative building materials, and utilizing modest -sized starter homes and incremental building approaches. Thus, rather than depend on government, NGO, or donor funds, sustainability and scale will be within the province of the market actors the lenders and developers combined with effective demand by Ugandan households. The constraints to sustainability and scale, however, are also market-based: availability of adequate liquidity for both construction and take-out finance; the confluence of innovative house design and building materials with the affordability and the preference profile of Ugandan households; a reasonable rate of return for developers of the modest housing estates; 16

19 the ability and willingness of lenders to underwrite modest and lower income households, including those with informal income; banks concerns about capability of developers and cost overruns which abrogate the affordability calculations; and banks concerns about the collateral value of the homes and locations. Table 4.1 provides an evaluation of the pilot projects in terms of classic criteria: sustainability, scale, efficiency, and relevance. The plus column is fairly robust: these projects do not need subsidy and given the presumed income distribution in urban Uganda, effective demand is expected to be large. The approach is certainly a relevant way to improve the quality of housing. However, the negatives are also robust as far as scale and sustainability. Thus, the recommendations that follow address liquidity in the long-term as well as the land use planning necessary to alleviate a major supply side constraint. Table 4.1 Advantages and Disadvantages of the Pilot Projects Evaluation Criteria Advantages of the Pilot Plans Constraints Against Sustainability and Scale Scale/ Effectiveness Efficiency Utilizes the market experience & Loan product trade-offs among expertise of existing lenders who banks, MFIs, & MDIs; liquidity is a are willing to extend down-market long-run constraint; credit & use new loan products enhancement needed? Effective demand expected to be large; introduce affordable loan products building design Relevance Highly relevant to substantial % of households with modest & low salaries, informal & underreported income Sustainability High interest & building costs & lack of affordable design will slow development Will remain relevant, but preferences need careful linking with affordability and design No major finance or construction Liquidity will be a major issue; subsidies required for low-modest Will need improved land & & modest income; lenders find infrastructure policies. new market niche Savings Products and Bank Liquidity. UN-HABITAT has assisted the FIRST project with research on savings products. 8 DFCU s Smart Plan and HFB s Shelter Account are already in place, offering would- be mortgage borrowers the opportunity to save and provide a demonstration effect for repayment as well as a fund for the down-payment. However, as discussed in the companion paper, Study to Examine the Use of Retail Funds for Mortgage Lending, with deposit rates consistently below the inflation rate, most savers in the banking system face negative real interest rates. In 2008, the real interest rate was approximately -11 percent. As long as negative real deposit rates persist, there is little incentive for the average Ugandan to save in the formal financial sector. Thus, the goal here might be to assist banks in modeling the impact on the balance sheet of combining higher savings deposit rates with increased mortgage lending. Assess development of a Liquidity Facility. The Companion paper also addresses liquidity in the Ugandan market. Notably, the larger international banks, which currently have little 8 See UN-HABITAT, Saving Products for Low-Income Groups 17

20 interest in either mortgage finance or down-market lending, do not appear to be liquidity constrained. However, the smaller banks including those willing to participate in pilot projects, already face liquidity constraints, which would be exacerbated as their lending expands. Also, for several lenders that are potential pilot partners, down-market loan products are already being utilized; thus, lack of liquidity may be their main barrier to expansion. As noted, it is precisely these smaller local and regional banks, MFIs, and MDIs that are currently most interested in making mortgage loans and most willing to investigate down-market entry and new loan products. Liquidity Facility is likely to be developed in Tanzania, on that will assist MFIs as well as banks, and this could provide valuable guidance for any decision on a facility for Uganda. 9 Another approach to enhancing liquidity might be the Microfinance Enhancement Facility, sponsored by IFC and the German development bank KfW. The Microfinance Enhancement Facility is a short- to medium-term facility of up to $500 million with initial contributions of $150 million from IFC and $130 million from KfW. The Facility is expected to provide refinancing to more than 100 microfinance institutions in up to 40 countries. Given increased experience worldwide, Uganda will be able to benefit from exptensive experience in MFH lending. Credit Guarantees and the Microfinance Enhancement Facility. Credit enhancements, such as loan guarantees to banks or guarantees of on-lending from banks to MDIs and/or MFIs are another way to help support down-market lending and boost liquidity for these efforts. In fact, a bank/mfi partner approach, pioneered in Latin America, is now utilized in many emerging markets. Several models are utilized in the partner relationship, but in most cases, the bank provides funds while the MFI undertakes origination and servicing. Credit enhancement provided by the government or donors can help kick-start this type of down-market push. just as for the liquidity facility, a credit enhancement mechanism is not recommended in the shortterm, although in the longer-term is may be useful. Pension Reform. Uganda is now addressing pension reform, including appointing a regulator for the pension industry. NSSF would also be subject to this regulation. More appropriate access to NSSF funds via the capital market would be an important step for banks and in the longer-term, the possible liquidity facility. It is recommended that NSSF work through the financial sector for its housing and construction goals. Such intermediation would be a positive step for the capital market and banking sector, rather than NSSF s current plan to utilize 60 percent of its portfolio directly in development. Land use planning and infrastructure development. While improved loan products and house design are a necessary condition to expansion of housing lending in Uganda, they are far from sufficient for most low income groups. A key supply side barrier is limited availability of land, infrastructure, and planned estates, exacerbated further by some inappropriate building/ planning regulations. Although interest rate subsidies are not included in the proposed pilots, the no-subsidy rule is difficult to sustain for land and infrastructure for lower income groups. Thus, government efforts to improve land titling, land use planning, and infrastructure are vital, along with a reform in the building/ planning code. Since these reforms are likely to take a long time, in the interim it is recommended that some of the land development constraints associated with formal planning code (e.g. minimum plot size, and the building code) simply be waived or relaxed for the pilots, so as to test the model for smaller, but more decent and affordable housing 9 See Sally Merrill, Tanzania: Action Plan for Development the Mortgage Market: Report on the Development of a Liquidity Facility ; Hassler, O., & Walley, S. Mortgage Liquidity Facilities. World Bank. 18