Housing Finance for the Low-income Population in South Africa: A Market Demand Assessment

Transcription

1 Housing Finance for the Low-income Population in South Africa: A Market Demand Assessment Prepared by The Development Innovations Group December 2008

2

3 Table of Contents Executive Summary 5 A. Introduction 7 B. Background 7 B.1. Economic Context 7 B.2. Housing in South Africa 8 B.3. Housing Finance Market for the Low-income Populations 9 C. Overall Results of the Market Assessment 11 C.1. Socio-Economic Profile of the Sample 11 C.2. Household Income 12 C.3. Housing Quality 13 C.4. Enterprise Characteristics 13 C.5. Availability of Informal Financing 15 C.6. Effective Demand for Formal Financial Services 16 C.7. Potential Demand for Individual Loans 17 C.8. Potential Demand for Home Improvement Loans 18 D. Variations by Type of Employment 19 D.1. Socio-Economic 20 D.2. Housing Quality 21 D.3. Use of Financial Services 21 E. Regional Variations 23 E.1. Socio-Economic 23 E.2. Housing Quality 24 E.3. Use of Alternative Financial Services 24 F. Gender Variations 26 G. Market Opportunities 27

4

5 Table of Figures Figure C 1: Type of Employment Figure C 2: Sectors of Employment Figure C 3: Educational Level Figure C 4: Average Household Income and Repayment Capacity Figure C 5: Sources of Informal Loans Figure C 6: Types of Formal Loans Figure D 1: Categories of Employment Figure D 2: Sectors of Employment by Type Figure D 3: Household Income and Repayment Capacity by Employment Type Figure D 4: Use of Financial Services by Employment Type Figure E 1: Household Income Level and Repayment Capacity by Region Figure E 2: Use of Financial Services by Region Figure E 3: Formal Loan Type by Region Figure F 1: Use of Financial Services by Gender Figure G 1: Use of Financial Services

6 Currency Equivalent Exchange rates for the South African Rands (R): 1 U.S. Dollar (US$) = 7.07 R (Average rate for May-July 2007) Acronyms DIG GDP FSC RDP RoSCAs SACCOs Development Innovations Group Gross Domestic Product Finacial Sector Charter Reconstruction and Development Program Rotating Savings and Credit Associations Savings and Credit Cooperatives

7 Executive Summary The following report presents the results of a market demand assessment conducted in South Africa to survey the landscape of financial markets in low-income communities with a particular focus on housing finance. The survey conducted between May and July of 2007 considers the employment, geographic, and gender variables representative of South Africa s low-income population. Based on the survey results, the report provides an analysis of the types and nature of financial services that are currently available, effectively used, and potentially in demand; while at the same time examining current housing conditions and areas where appropriate financing can help support the development of this market. The survey was designed by the Development Innovations Group (DIG) and conducted in association with FinMark Trust and the Progressus Consulting Group (local surveyors). The market demand assessment for housing finance in South Africa helps to shed light on both the effective and potential demand for financial services among the low-income population in general, as well as their particular needs and requirements for housing finance given the current housing situation in the country. The survey examined a range of economic agents, which included microentrepreneurs, lowincome employees, and laborers (i.e. non-contract employees, piece job laborers, and domestic workers). By drawing respondents from a diverse range of urban settings, including townships and poor neighborhoods from Johannesburg s central and metropolitan areas to Cape Town, the survey has based its results on a representative sample that takes into account several social, economic, and ethnic factors. A typical survey respondent is a 40-year-old head of household with a family of three dependents, living in his/her own home. He or she generally supplements his/her income with contributions from other family members thus rendering the households cash on hand, after expenditures, on average in the amount of R1, Additionally, the effective monthly repayment capacity is estimated at R759 2, which could be used to gauge the amount of monthly loan repayments in order to meet current and future housing needs. In general, loan assessments involving monthly loan repayments up to this value reduce the risk of non-repayment among potential clients due to debt overburdening. Among the microentrepreneurs in the sample, the majority are new business owners who have been running microenterprises for an average of five years. Microenterprises are predominantly sole proprietorships, typically with no hired labor. They are ambulant or operate from homes, mainly buying goods and raw materials in cash, seldom using supplier credit, and selling their goods in cash, with exceptional cases of microentrepreneurs extending credit facilities to their clients. The most challenging problems reported by microentrepreneurs are domestic competition and customer problems. For the most part respondents meet their current financing needs through the informal financial sector, primarily participating in Savings and Credit Cooperatives (SACCOs), relying on family and friends, or in some cases employers or even money 1 Equivalent to USD $ household cash on hand after expenditures 2 Equivalent to USD $ effective monthly repayment capacity 5

8 lenders, to borrow for a few months at a time. The vast majority of respondents have never requested a loan from a formal financial institution and very few expressed a desire for formal loans. However, perhaps some of these figures have more to do with the respondents psychological barriers to formal finance reinforced by the general lack of information on potential products and services available rather than real lack of demand. Of those very few who do have access to formal bank loans, about one-third use the loans for home improvement. There is, however, a much more dynamic market penetration by formal institutions when it comes to savings mobilization, with the vast majority of respondents holding a savings account. While the potential demand for financial services clearly exists, it has not been effectively translated into access to formal finance. This is due in part to a general unfamiliarity with and a lack of experience navigating the formal financial sector, as well as fears about the ability to repay and successfully meet financing requirements. The real barrier, however, largely resides in the lack of supply of suitable formal financial services and the inadequate delivery methods adopted by the traditional financial institutions. Despite the general reluctance to take a formal loan, when asked more specifically about meeting their housing needs, there was a much larger response asserting potential demand for these types of loans. The majority of the respondents were interested in undertaking some sort of housing improvement in addition to building or acquiring new homes in some cases. While there was some interest in business and personal loans it involved much smaller amounts. In all cases, respondents were reporting reasonable amounts that could be repaid in monthly installments, in line with their estimated repayment capacity calculated at R600. There are of course variations in the differing uses of and demand for financial services, particularly when examining different types of employment groups and regions. Among the most important variations are the differences in the repayment capacity, which is higher among microentrepreneurs than employees, with laborers having the lowest repayment capacity of all groups. At the same time entrepreneurs have exhibited less use of formal loans. That said, when entrepreneurs do borrow they have the highest loan amounts, making them a very interesting potential target market for housing finance. Employees demonstrated more use of formal loans and savings services in banks than microentrepreneurs and laborers. This greater use of formal financial services on the part of employees is likely due to the salary payment system that relies on direct bank deposits. In addition to assessing the potential demand for existing financial services among low-income earners, this study begins to identify opportunities for new financial products that can be developed to meet the specific needs of this market segment. When examining the housing sector niche in particular, this study clearly shows that there is ample room for expansion and growth. By assessing existing and potential housing loan demand, this study sheds light on the nature of this demand, giving specific indications on the size, terms, conditions and costs of individual loans that respondents would be willing to take to meet their ongoing housing needs. Financial service providers who are willing to better understand the specific characteristics of low-income clients and the nature of their economic activities will be extremely successful in penetrating this new market. In particular, financial service providers should look at the estimated monthly repayment capacity, as a key indicator of the loan amounts and monthly repayment rates that on average could be afforded by the group.

9 A. Introduction The Development Innovations Group (DIG), under a grant from the Bill and Melinda Gates Foundation, is undertaking an action research program in the area of housing finance for the poor in several countries around the world including: India, Indonesia, Mexico, Morocco, Pakistan Angola and South Africa. The program aims to provide donors, policy makers, and development finance practitioners with a better understanding of the housing markets in these countries to help shape future investments and efforts toward the promotion of housing finance in the developing world. As a part of the applied research strategy, DIG is conducting assessments of low-income communities in select countries, with a particular focus on housing finance. These market assessments look at the variety of formal and informal financial services currently available, how they are effectively used, and the potential demand for additional products and services given existing and desired housing conditions. This market assessment report presents the results of the demand survey conducted in South Africa. The survey was designed and conducted between May and July of 2007, in association with FinMark Trust 3 and Progressus Research and Development Group 4, and covered a number of areas of enquiry ranging from an assessment of household income and expenditures, current access to financial services, and potential demand for housing specific financial products. The survey sample was constructed to take into account differences in employment categories, geography, and gender variables representative of South African low-income earners, which are viewed as significant determinants of variations within the population. B. Background B.1. Economic Context South Africa, a middle-income emerging market with a population exceeding 43 million, is endowed with an abundant supply of natural resources; well-developed financial, legal, communications, energy, and transport sectors; a stock exchange that ranks among the ten largest in the world; and a modern infrastructure supporting an efficient distribution of goods to major urban centers throughout the region. The economy grew at a healthy five percent in 2006, and is largely dependent on the service sector, which accounts for 66 percent of GDP, with manufacturing comprising an additional 31 percent and agriculture three percent of GDP 5. However, in spite of the country s impressive achievements, its economic growth has not been strong enough to lower South Africa's high unemployment rate a daunting economic problem stemming from the poverty and lack of economic opportunity for disadvantaged groups institutionalized during the apartheid era. South Africa continues to host one of the most unequal income distribution patterns in the world; almost half of the population (48.5 percent or 22 million people) falls below the national poverty line 6. South Africa s two-tiered economy is split between that of a developed country and an underdeveloped country with the poor having access to only the most basic education, infrastructure and economic opportunities. 3 FinMark Trust is an independent trust whose business is controlled by five trustees from countries in Southern Africa. FinMark Trust aims to promote and support policy and institutional development towards the objective of increasing access to financial services by the un- and under-banked in Africa. 4 Progressus Research and Development Group is a South African marketing and survey company based in Johannesburg, South Africa. 5 CIA The World Fact book; 6 UNDP, South Africa Human Development Report

10 B.2. Housing in South Africa 7 Housing has been a key focus of South Africa s new democratic government since its first election in Since then, the challenges faced by government have been many. With an estimated 86 percent of households earning less than R3,500 per month, housing affordability is seriously constrained and in obvious need of subsidy support. However, existing subsidies from the previous regime were designed to support the racially-defined framework of the government s apartheid policy. They were also expensive and unable to support the breadth of need defined by a democratic administration. The availability of end-user finance was also limited. Retail lenders lacked the capacity to extend downmarket, and there was an explicit reluctance on part of some formal financial institutions to lend in certain areas and to certain groups of people. At the same time, the advent of the country s first democratic government brought with it enormous expectations from a society that was tense with anticipation regarding the promise for development and a better life for all. The 1994 Housing White Paper responded to this situation with three policy interventions: Subsidised housing delivery. A capital housing subsidy providing a basic starter house with services and freehold tenure to qualifying beneficiaries who earn a household income of less than R3,500 per month has resulted in the delivery of just over two million units since its launch in In addition, a rental subsidy has allowed for the delivery of just over 50,000 units within newly formed social housing institutions, since Finally, the subsidised housing programme allowed for the transfer of old township stock (that is, housing stock constructed by the former, apartheid regime for rental on a racially segregated basis) to residents at basically no cost to the resident. This stock has become the basis for the development of a nascent property market in former black residential townships. Enabling environment for housing finance. The policy included a specific strategy to stabilise the housing environment and thereby encourage greater lending down-market by the existing banks that comprised the country s financial lending sector. This strategy led to a range of initiatives. Early on, for example, various institutions were established to encourage banks to extend their loans down-market 8. Mobilising housing finance. Two wholesale financiers, the National Housing Finance Corporation and the Rural Housing Loan Fund, were established to support the emergence of a cadre of non-bank housing micro lenders to offer unsecured, small loans to low income earners for housing purposes. A number of housing micro lenders have been established as a result of this initiative, operating in both rural and urban areas. In 2004, Cabinet approved the Comprehensive Plan for Sustainable Human Settlements, otherwise known as Breaking New Ground. This policy broadened the Department of Housing s role to consider the long-term sustainability of human settlements in addition to the scalable delivery of housing to meet the backlog. One consequence of this policy shift has been an increase in the quantum of the housing subsidy and the raising of minimum norms and standards in Despite the tremendous success of over 2 million houses built, South Africa s housing backlog 7 This section is drawn from Rust (2006) Analysis of South Africa s Housing Sector Performance, with permission of the author. 8 For more detail, see Rust (2006).

11 has remained virtually static in her 2007 budget speech, Minister of Housing Lindiwe Sisulu said the backlog was estimated at 2.2 million households. Of the 8 million households that earn less than R3,500, 36 percent live in informal dwellings. An additional 16 percent live in overcrowded conditions (i.e. more than 2 people per room), and 21 percent live in formal housing that they own but which is structurally inadequate 9. B.3. Housing Finance Market for the Low-income Populations Substantial progress has been made in the housing finance sector since the 1994 Housing White Paper. While lenders were originally reticent to extend their products down-market, this has changed, largely as a result of the Financial Sector Charter 10 (FSC), where between January 2004 and 31 December 2007, lenders made R37.99 billion worth of mortgage, fully guaranteed (pension-backed), unsecured, developer and wholesale loans (Table B1). Table B1: Unaudited Banking Association Figures of FSC Bank Lending for Housing 11 Product Number originated (Jan Dec 07) % Value (R million) % Mortgage % % Fully Guaranteed % Unsecured % % Wholesale % Development loans 274 n/a % Total The value of loans extended to date is R37.9 billion, not far short of the R42 billion target being sought for the end of This activity notwithstanding, the housing market in South Africa still falls short of meeting the housing needs of its low-income population who continue to lag behind in terms of national home ownership. According to the Income and Expenditure survey of 2005/06, 65 percent of South Africa s approximate 12 million households earn less than R3,500 per month and are therefore eligible for the housing subsidy Income and Expenditure Survey, 2005/ This is an industry-led commitment to address broad black economic empowerment issues through governance and management of financial sector institutions, as well as through access to the products and services of the financial sector. It was developed in response to the Black Economic Empowerment Act, but is not legislated in any other way. 11 Rust (2008) ACCESS housing No 10 May/June Provided they also satisfy other qualifying criteria including that they are SA citizens or permanent residents, have financial dependents, and have never owned a house before. 9

12 Figure B1: Monthly Household Income Distribution A factor complicating the housing backlog, however, is that most households earning above the R3,500 income per month threshold cannot afford housing in today's economy. To afford the average "affordable" house 13 that in the third quarter of 2007 cost R265,000, a household would have to earn at least R12,000 per month 14 (Figure B1). Resale market housing is not much more affordable: few if any houses are available for less than R150,000, affordable only to households earning more than R6,500 per month. As a result, newly established households often live in the backyards of their parents, or share housing between multiple families. Given such housing affordability constraints, which exist for at least 77 percent of the population, there is considerable opportunity for the development of additional housing on existing properties. An important area for growth in the housing finance sector, therefore, must be housing microlending. Certainly, from the demand side, housing micro-loans would appear well suited to at least the two million government subsidized houses that have already been constructed, the more than 200,000 site and service properties delivered prior to the 1994 election and the 868,000 or so old township stock houses that have been transferred to their residents. Although banks report that they have extended slightly less than 220,000 such loans, the demand must be much greater; though it has not yet been quantified and is not well understood. In addition, it is not clear whether housing micro-loans are integrated in banks suites of housing loan products. Housing lending continues to be thought of as secured lending by definition, and the mortgage product is understood as fundamental when lending for housing purposes. Unsecured loan products, as reported in the FSC statistics, have been delivered through the personal banking divisions of banks rather than their home loans sections. Non-bank housing lenders struggle to access the capital they need to operate at scale, and investors are wary of this market given the limited information available regarding the characteristics and repayment capacity of potential low-income clients. 13 According to the ABSA Residential Property Perspective, an affordable house is defined as 40-79m2 and costing less than R370, This assumes a twenty-year mortgage at 15.5% prime, with 10 LTV and at 3 installment to income.

13 C. Overall Results of the Market Assessment C.1. Socio-Economic Profile of the Sample The market demand assessment involved a sample of 1,034 observations with roughly equal representation of entrepreneurs (37%), permanent employees (34%), and piece job laborers (26%), along with a small percentage of non-contract employees (3%) (Table A1 and Figure C1). The sample was randomly selected across three metropolitan areas of South Africa, namely Johannesburg, Ekurhuleni and Cape Town. In Johannesburg, the sample was selected from the townships of Eldorado Park (19%) and Orange Farm-Stretford (16%); in Ekurhuleni from Chris Hani (16%) and Katlehong Nhlapo (17%); and in Cape Town from Netreg (16%) and Suburban (16%), (Table A1). The sample is additionally representative of the main sectors of employment for low-income earners, namely services (37%), trade (3), manufacturing (22%) and domestic workers (1), (Figure C2). As a result, the gender distribution in the survey is relatively proportionate by design. The gender factor coupled with the type of employment and regional distribution are very representative of the South African population and business environment. Figure C1: Type of Employment Figure C2: Sectors of Employment The sample is almost equally distributed between men (48%) and women (52%) with the average age of the respondent 41 years. The majority of respondents are married (59%), one-third (28%) are single, and the remaining (13%) are divorced, widowed, or separated. The study shows that the vast majority of respondents (97%) have received some level of education and only a small number (3%) are illiterate with no formal schooling, which is in line with the 86.4% national literacy rate. The majority of respondents have received some secondary education or completed their secondary degree (77%); less than one-fifth (16%) have completed primary education; and a small percentage (4%) received some college education or completed a college degree (Table A2 and Figure C3). 11

14 Figure C3: Educational Level Given the housing shortage and backlog of affordable new housing in the country, a surprising majority of the survey respondents are in fact homeowners (73%) (Table A2) 15. Only a few (6%) of the respondents reported occupying rentals or living in rent-free accommodations (2). Additionally, the majority of the homeowners hold some type of legal documentation such as a registered title deed (73%), an agreement between parties (7%), or a registered sales contract (2%), leaving only one-fifth of home owners (18%) having no form of ownership documentation (Table A2). It is important to highlight that the vast majority of dwellings were built or purchased using personal savings (97% and 62%, respectively). Of those who reported purchasing their homes, 30 percent used formal loans, whereas only two percent of those who constructed their homes used formal loans (Table A2). The higher rate of borrowing for home purchase is likely tied to borrowers ability to use the homes as collateral, providing greater security to the lender. C.2. Household Income In general, households were comprised of a head of household (the main income earner) who supports an average of three dependents. The average monthly income generated by the respondent from his/her primary source of income was R2,461 (R1,975 median) 16, and R126 (0 median) from the secondary source of income. Some respondents reported an average supplementary income from their spouse s activities of R388 (0 median) and/or from other family members of R490 (0 median). The gross household income averages R3,468 (R2,750 median) with total household expenses averaging R2,065 (R1,670 median). These figures yield a net household income after expenses of R1, (R809 median), (Table A3). Using these figures we can calculate an estimated monthly repayment capacity for housing (based on one-fourth of total gross income) equal to R867 (R688 median) 18. To get a more conservative figure, we calculate the effective repayment rate by comparing the estimated monthly repayment capacity with the net household income after expenses and taking the lower amount of the two. This yields an effective monthly repayment capacity for the sample of R759 (R580 median) 19 (Table A3 and Figure C4). The household effective monthly repayment capacity is indicative of the minimum cash on hand available within the household after netting out all expenses. This excess liquidity could be drawn upon to cover the costs of any 15 There may be some bias in the sample particularly as South Africa s total homeownership rate teeters around 50 percent overall, and is assumedly less for lower-income households. 16 $348 mean ($279 median) average monthly income for head of household. 17 Net household income is the difference between total household income and total household expenses. 18 $122 mean ($97 median) effective monthly repayment capacity. 19 $ mean ($82.04 median) effective monthly repayment capacity.

15 emergencies or to service a potential debt. The monthly installment on any future loan should not exceed the value of this available cash. Figure C4: Average Household Income and Repayment Capacity C.3. Housing Quality Respondents reported having an average of three rooms for four inhabitants in their houses (Table A18). Almost half of the respondents (45%) reported living in old township stock houses, some in Reconstruction and Development Program (RDP) houses in townships (18%), in shacks in informal settlements (16%), and in shacks on site and service areas (1), while few reported living in houses on site and service areas (8%), and in rare cases in flats () (Table A18) 20. An overwhelming majority (98%) stated that their homes have a roof. Kitchens are typically inside the home (72%) and bathrooms are often outside (58%). Walls are made of bricks (63%), corrugated iron (27%) and cement blocks (9%). Homes are usually connected to public or private water connections (84%) or a local water line (15%), and have access to electricity (82%), except for some respondents who reported living in houses with no electrical wiring (15%) (Table A18). C.4. Enterprise Characteristics For those respondents who identified themselves as entrepreneurs or self employed (37%), the absolute majority (98%) own and operate their own microenterprises (Table A5). On average, these microenterprises have been in operation for about five years, having started with an initial investment of R2,919 (R500 median) 21 and now owning physical assets with a present market value of R11,960 (R1,500 median) 22, excluding land and buildings (Table A4). The relatively low median of physical asset ownership indicates a large variation among entrepreneurs and in the scale of their operations. Similar to entrepreneurial activities in many countries, start-up funds for businesses in South Africa are largely self-financed by the entrepreneur (88%) from their own savings (Table A6). A very small number of microentrepreneurs reported taking loans from banks (), employers (), stokvels or SACCOs 23 (), friends and family members (2%), or other sources to establish their businesses (Table A6). 20 Households earning below R1,500 per month qualify for RDP housing at no cost and households earning up to R 3,500 qualify for this option at a minimal cost. 21 $413 initial start-up capital for a microenterprise. 22 $1, mean ($ median) present market value of physical assets, excluding land and buildings. 23 Stokvels or SACCOs which stands for Savings and Credit Cooperatives are the same as the more commonly known Rotating Savings and Credit Associations (RoSCAs), however the latter term is not commonly used in South Africa. 13

16 The majority of businesses are solely operated by their owners (77%) with no other employees. Those who have hired labor reported an average annual labor cost (both part and fulltime) of R8,223 (0 median). Additionally, microentrepreneurs pay on average R5,983 (R3,000 median) 24 annually to cover their various operational expenses (Table A4). The bulk of these businesses are home based (76%) while some are ambulant (1), operate from a store (5%), construction site (7%), or a market stall (). Those who operate out of their homes are largely owners of their dwellings (7), while some rent (6%) or use space for free (18%). The majority of microenterprises are proprietorships (94%), with only a few partnerships (6%) and only one-fourth are registered businesses (25%) (Table A5). The majority of the microentrepreneurs surveyed in the study (88%) reported purchasing raw materials or goods for their businesses. The average annual value of these purchases reached R36,476 (R15,600) 25, (Table A7), and are predominately paid for in cash (10) (Table A8). A very small percentage of entrepreneurs reported using supplier credit (2%) in addition to cash, while none reported using advance payments or purchasing on consignment (Table A8). The average annual value of sales among microenterprises is R81,205 (R48,180 median) 26, (Table A9). Similar to their purchasing arrangements, microentrepreneurs predominantly sell their goods on a cash basis (98%) (Table A10). However, one-third of the entrepreneurs (32%) reported extending credit for goods as well, which is a common mechanism used by microentrepreneurs to promote their sales. Consignment () and advance sales (4%) were not commonly reported among the sample (Table A10). The primary business challenges that the microentrepreneurs reported in the survey include: domestic competition (24%), weak demand (16%), and various customer problems (15%). Secondary problems faced by entrepreneurs included again domestic competition (2), weak demand (1), and the lack of proper trading places (1) 27. Other challenges cited include: customer problems (8%) and general services (3%) (Table A11). 24 $ mean ($ median) microentrepreneurs annual operational expenses 25 $5, mean ($2, median) microentrepreneurs average annual value of purchases 26 $11, ($6, median) microentrepreneurs average annual value of sales 27 Those were the entrepreneur responses when asked about the most significant and second significant challenge.

17 C.5. Availability of Informal Financing (1) Effective Demand for Informal Loans Respondents were asked about their general use of alternative financial services to reflect how low-income economic agents are able to meet their demand for financial services outside of the formal financial sector. One quarter of the sample (25%) reported having requested a loan at some point in the past from an informal source (Table A12) including friends (38%), family (32%), moneylenders (15%), and employers (13%) (Figure C5). Of the subgroup who had requested an informal loan, very few () reported having been denied the loan (Table A12). Therefore, while most start-up businesses reported to be selffinanced, it is clear that informal borrowing is a rather widespread financial management practice that is used by a fairly sizable percentage of households within the low-income group. When asked about the use of informal loans over the past 12 months, almost one-fourth of the sample (22%) reported using such loans, again validating a relatively active demand for loans by the low-income group that is currently being supplied by informal providers (Table A12). The predominant sources of these informal loans were concentrated among friends (4) and family (32%), with the remainder coming from moneylenders (16%) and employers (9%). These informal loans were typically disbursed within two days. The average value of these loans is R1,194 (R500 median) 28 with an interest amount of R106 (0 median) and a three-month repayment period (Table A12). The high variation in the value of interest paid on informal loans was due to the fact that only 30 percent of those receiving informal loans actually paid interest, which amounted to an average of R351 (R150 median) 29. Clearly this is significantly dependent on the source of informal loans. The average value of collateral used as a percentage of the loan amount was 4 percent (0 median) due to the fact that only 9 percent of borrowers were required to present collateral. These informal loans were used for a variety of purposes, including largely personal (85%), business (13%), home improvements or repairs (), and home purchases () (Table A12). Figure C5: Sources of Informal Loans 28 $169 average value of informal loans 29 $49.65 average value of interest on informal loans 15

18 (2) Participation in Informal Groups In addition to individual loans from friends and family, other informal savings and credit channels such as Rotating Savings and Credit Associations (RoSCAs) play an important role in household financial management among low-income households. In South Africa, RoSCAs are known as stokvels or SACCOs. The survey indicates that almost a third of the interviewees (3) reported participating in informal savings and loans groups or money transfer schemes. These groups are generally comprised of 15 members (Table A16). On average, the members of the group make one monthly contribution of R156 (R100 median) which amounts to a yearly contribution and payout of R2,922 (R1,578 median) 30. By and large the groups rotate long after the first cycle is completed and can exist up to five to ten years (Table A16). Additionally, a very small percentage of the respondents (2%) reported saving with informal collectors at an average amount of R 2, annually (Table A16). Contributions with pension funds are relatively common as one-third of respondents (29%) reported making regular contributions from their salary. Although remittances are common among the low-income population worldwide, only a small group of the sample (4%) reported receiving remittances at an average of R4,380 (R3,600) annually, despite the fact that 27% reported sending and receiving money transfers domestically to friends or family members via Western Union, Post Bank, and others (Table A16). C.6. Effective Demand for Formal Financial Services Respondents were also asked an array of questions on availability and use of formal financial services in order to better understand the obstacles that low-income groups face in accessing the mainstream financial sector. Not surprisingly, the majority of respondents (79%) reported never having requested a formal loan from a bank or other formal financial institution (Table A13). The main reason for not applying for a formal loan is fear of the inability to repay (42%), followed by a lack of financial documentation (26%), and a perception of high interest rates and fees (14%) (Table A13). This trend of self-selection from the formal credit market by low-income groups in South Africa is not unique and is similar to the behavior exhibited by low-income earners around the globe. Of those who had applied for formal loans in the past, only a small percentage (3%) reported being denied access. The main reason for rejection was the lack of required financial documents (49%), followed by bad credit history (2), lack of collateral (15%), or lack of business registration (6%). Very rarely () respondents reported not collecting an approved loan from the bank due to fear of being unable to repay (Table A13). Of the few respondents (13%) who reported taking a formal loan over the past three years (Table A14) the majority (58%) used these loans for personal purposes, while one-third (33%) used the loans for home improvements and a few (2%) for mortgages. Interestingly, very few formal loans (5%) drawn over the past three years were reported for business purposes (Table A14 and Figure C6). The majority of this sub-sample with formal loans (75%) is currently repaying a loan. The bulk of these loans were provided by unspecified banks (78%), and the remainder were distributed among Absa Bank (13%), Capitec (), and the other financial providers (8%) (Table A14). 30 $ average yearly contribution in stokvels. 31 $ average annual savings with informal collectors.

19 Figure C6: Types of Formal Loans The average size of formal loans used by the respondents over the past three years was approximately R25,434 (R8,000 median) 32 with an annual interest rate of 44% (3 median) (Table A14). The average installment value was R727 (R547 median) 33 associated with an average 40-month term (24 months median). The borrowers typically provided up to 35 percent (0 median) of the loan value as collateral, while only 40 percent of the respondents provided collateral (Table A15). In fact, the majority (6) reported not presenting any kind of collateral (Table A14), while few (16%) reported using real estate as collateral, or bank accounts (8%), and other types of collateral (16%). It is important to note that the limited use of loans from the formal financial sector is observed notwithstanding the fact that the majority of respondents (7) reported holding at least one deposit account with a formal bank (Table A15). Thus, whereas the low-income group largely self-selects itself out of the formal credit market, it still has a high demand for formal deposit facilities and frequents banks for the purpose of depositing and withdrawing savings. This is likely to be a result of the payment system in South Africa, which relies heavily on direct deposits and bank transfers for salary payments and other economic transactions. C.7. Potential Demand for Individual Loans Informal social structures and formal financial institutions provide an array of mechanisms which can be used by low-income economic agents to meet their financing needs. The survey identifies the current use of these alternative financial services as effective demand given that it is based on the real transactions in the market. However, the survey also goes beyond this demonstrated demand in an attempt to identify the terms, conditions and services, which are either currently unavailable or perceived to be so by the target population, which if available would spur latent or potential demand for financial services. Of the respondents who were asked of their desire and interest to receive a loan from a bank, less than one-third (28%) of the sample responded positively (Table A17). Among those who expressed interest in accessing individual loans, the highest demand was for housing loans (68%), followed by business loans (34%) and personal loans (18%) 34. Respon- 32 $ 3,597 average formal loan taken by respondents. 33 $ average installment value for formal loans. 34 Potential demand and loan type were not mutually exclusive and respondents could select more than one type of loan. 17

20 dents interested in taking housing loans reported desired potential loan amounts of R63,495 (R30,000 median) 35 with an 85-month (60-month median) average repayment term (Table A17). Many of those respondents reported that they would use the housing loan to construct a new house (4), to undertake home improvements (38%), or to buy a new house (2). Respondents interested in accessing business loans in the future reported being interested in average loan amounts of R28,932 (R10,000 median) 36, and an average loan term of 38 months (25 months median) (Table A17). Lastly, respondents interested in accessing future loans for personal purposes reported average loan amounts of R21,471 (R10,000 median) 37 with an average loan term of 32 months (24 months median) (Table A17). Respondents interested in taking an individual loan for a variety of puprposes reported a willingness to pay an average annual interest charge of 12% (1 median) with processing fees of 4% (3% median) for an average loan amount of R57,366 (R20,000 median) 38 with a monthly installment of R840 (R600 median). These loans would have to be granted for periods extending from two to five years. Additionally, respondents reported that they could provide various types of collateral including salary guarantees (34%), fixed assets (26%), bank accounts (1), furniture, guarantors, pension funds, machinery and others. Lastly, when asked about their preference for requesting a formal loan from a bank or a micro lender, the majority of the respondents (87%) preferred taking a loan from a bank and only very few (3%) were interested in taking a loan from a microfinance lender, while some (1) did not have any preference (Table A17). Those respondents who were not interested in accessing an individual loan for housing were additionally asked why. The majority of the group reported a number of reasons including: a preference to save and pay in cash (37%) waiting for government subsidy (24%) being content with their current housing structure (23%) and not being interested in taking the risk and burden of a loan (9%) (Table A17). C.8. Potential Demand for Home Improvement Loans Almost half of the sample (4) reported having undertaken home improvements over the past five years, (Table A19), and spending an average of R20,846 (R10,000 median) 39. These improvements included activities such as painting (7), flooring (65%), room additions (53%), and windows (48%). The majority of households reported having financed these improvements with personal savings (83%), while few (8%) used formal loans or relied on family assistance (6%). When asked specifically about their plans for future home improvements, almost half of the respondents (45%) reported an interest in improving their homes over the coming five years (Table A20). Four-fifths (79%) of the respondents indicated that they would use the funds to build an additional room or paint, while many would repair their floors or doors (72% and 7, respectively). On the whole, they estimated an average cost of improvements totaling R44,430 (R30,000 median), and reported having an average value of R2,574 (0 median) in savings that they could use towards home improvements (Table A20) 40. Some respondents (13%) suggested that they could complete the work without hiring labor and thus save on labor costs. Almost one quarter of the respondents who were interested in undertaking home improvements (24%), reported an interest in accessing a home improvement loan. This group indicated that they would be able to pay a monthly installment of about R 1,031 (R600 median). 35 $8,981 potential housing loan demanded. 36 $4,092 potential enterprise loan demanded. 37 $3,037 potential personal/consumption loan demanded. 38 $8,114 potential individual loans demanded. 39 $2, average cost of home improvements undertaken over the past five years. 40 $6, potential home improvement loan demanded.

21 The survey further investigated the interest of homeowners, landowners, and renters, in acquiring or constructing a new house. About one-fourth of homeowners (23%) indicated a willingness to move to an existing or new house. On average, this group estimated that a house costs R77,656 (R15,000 median) 41. Respondents were willing to repay at about the same average monthly installment of R776 (R500 median). Homeowners who were interested in moving to a new house reported an average resale value of their present house of R39,635 (R2,500 median) (Table A22). Of those who currently rent (6% of the total sample) or live free of charge (2 of the total sample), more than half (58%) were interested in acquiring land to build a house. Half of those (57%) were looking to government subsidies as the main source of financing for land purchase, while one-third (37%) cited formal purchase without any subsidy (Table A21). The majority of non-homeowner respondents (93%), who either own land or are willing to acquire land, would like to buy or build a home if given access to financing. They would like to finance the purchase or construction of their new home with a government subsidy (49%), savings (23%), or formal loan (25%). Affordable monthly installments for housing loans of R725 (R500 median) were reported by the respondents, in line with the previous estimates (Table A21). D. Variations by Type of Employment The survey targeted three main employment categories, namely (1) microentrepreneurs or the self-employed (37%), (2) salaried employees (34%), and (3) laborers performing piece jobs as well as non-contract employees (29%) (Figure D1). Section D of this report examines quite extensively the business characteristics of microentrepreneurs in South Africa, as well as employees and laborers, to give a better sense of the financial inter-dynamics between these businesses and the households, as well as their effective demand for business financing. Apart from this, socio-economic characteristics reflected in the sub-samples did not vary significantly. There are, however, a number of similarities and differences that are worth highlighting, particularly with regard to the use of financial services and potential demand for financial products. Figure D1: Categories of Employment 41 $10, estimated cost of new house for homeowners. 19

22 D.1. Socio-Economic Respondents share more or less similar education levels regardless of type of employment, with employees being slightly more educated than other respondents. Home ownership rates vary and are slightly higher for entrepreneurs (78%) than for employees (73%) or laborers (67%). Though not measured in this survey, the lower rate of laborer home ownership could be attributed to the large portion of laborers who work in construction sites or employer homes where they are often provided accommodation, or to the free housing afforded by the RDP to laborers whose monthly income falls below R3,500. Additionally, the majority of microentrepreneurs (76%) work from their homes, which suggests that home ownership may actually facilitate self-employment. When examining the sectors of employment, a few significant differences do arise. While entrepreneurs are concentrated in the trade (69%) and services (24%) sectors, employees and laborers are positioned more in the manufacturing (33% and 28%) and services (5 and 36%) respectively. Domestic workers were treated separately because of the specificity of this sector, but included in the study because of its relative importance in the economy, particularly with regard to the low-income group 42. Figure D2 depicts the variations in types of employment across economic sectors. Figure D2: Sectors of Employment by Type Among the three employment categories, it is interesting to note that there is little difference in total average monthly income between employees (R3,987) 43 and microentrepreneurs (R3,962) who both earn close to two times that of a laborer (R2,229) 44. The higher income for employees, however, is accompanied by higher household expenses (R2,508) yielding a lower effective repayment capacity than microentrepreneurs. Thus, microentrepreneurs have a marked higher effective monthly repayment capacity at R1, than employees at R861 46, with a significant difference from laborers who have a more limited effective repayment capacity of R (Figure D3). 42 The overwhelming majority of domestic workers in South Africa are low-income women with little to no job security. Though different labor unions are working to secure minimum wage requirements as well as formalize the occupation, this still is a very insecure profession. 43 $ employees total average monthly income. 44 $ laborers total average monthly income. 45 $ entrepreneurs effective monthly repayment capacity. 46 $ employees effective monthly repayment capacity. 47 $70.01 laborers effective monthly repayment capacity.

23 Figure D3: Household Income and Repayment Capacity by Employment Type D.2. Housing Quality Across the various types of employment, respondents reported similar housing conditions. Laborers, in particular, live in the most diverse and often run-down type of housing; many reported residing in old township stock houses (36%), RDP houses (2), shacks in informal settlements (19%), and shacks on sites (18%). Old township stock houses were, however, the most common type of dwelling for microentrepreneurs (5) and employees (46%). The vast majority of the respondents reported having used savings as the major source of financing for current home renovations (94% of laborers, 86% of microentrepreneurs and 74% of employees). D.3. Use of Financial Services The use of and demand for financial services from the informal and the formal sectors varies to some degree by type of employment (Figure D4). While the trends generally are not too diverse, there are some variations especially in the use and access to particular financial services. Figure D4: Use of Financial Services by Employment Type 21

24 Results of the survey indicate that in the past year, laborers have had a higher rate of borrowing (3) from informal sources than employees (2) or microentrepreneurs (18%). This higher rate of borrowing is accompanied by loan amounts that are smaller (R762) than those of other employment types (R1,666 for employees and R2,470 for microentrepreneurs). This is mirrored in the laborers lower repayment capacity. At the same time, friends and family are the predominant sources of informal loans for all respondents regardless of the type of employment. Self-selection outside financial markets varied to some degree by type of employment in South Africa. About two-thirds of the employees (64%) reported never requesting a formal loan, whereas most of the laborers (88%) and microentrepreneurs (85%) have mainly abstained. The reasons for not requesting a formal loan vary among the respondents; the reason most noted by employees and laborers was the fear of being unable to repay (5), while microentrepreneurs noted the lack of financial documents (36%). Not surprisingly, two-thirds of the respondents who had access to a formal loan in the past three years were employees (7). This correlates with the high percentage of employees (89%) holding deposit accounts with banks compared to the lower percentage of microentrepreneurs (65%) and laborers (57%). Interestingly, microentrepreneurs borrowed at the highest average loan amounts reaching R47,902 (R15,000 median), followed by employees at R23,742 (R7,000 median), and laborers at R5,844 (R5,000 median), which would seem to confirm their higher effective repayment capacity. Laborers reported paying the highest interest rates at 76 percent, while the interest rates paid by employees and microentrepreneurs were 42 and 27 percent, respectively. This, again, is consistent with global experiences where those with the least resources (i.e. lower total income and repayment capacity) tend to pay the highest premium on access to credit. It should also be noted that the interest rates and collateral provided appear to be inversely related. Potential demand for individual housing loans from the formal sector is relatively modest across all types of employment, with laborers reporting the potential demand with the lowest loan amounts at R36,698 compared with R67,666 and R98,897 for employees and entrepreneurs respectively. Not surprisingly, microentrepreneurs are the most interested in business loans (75%), compared to laborers (25%) and employees (7%). More employees (88%) and laborers (7), however, were interested in housing loans than microentrepreneurs (4).

25 E. Regional Variations Respondents in the survey were selected from three urban areas in South Africa, namely Ekurhuleni (33% or 339 observations), Eldorado Park and Orange Farm (35% or 364 observations) and Cape Town (32% or 331 observations). The geographical segments were further broken down by township: Chris Hani (49%) and Katlehong Nhlapo (5) were selected in the Ekurhuleni sub-sample; Kliptown (54%) and Stretford (46%) were selected from Eldorado Park and Orange Farm; and Netreg (5) and Suburban (49%) in Cape Town. Additional details on the nature of the townships are provided in Appendix E. E.1. Socio-Economic Average household incomes in Eldorado Park and Orange Farm (R3,931) and Cape Town (R3,734) are significantly higher than those in Ekuhurleni (R2,721). Nonetheless, the sectors and type of employment do not vary significantly. The one exception being domestic workers, a small subset whose income levels are well below the average in the overall tables. Examination of household income and expenses sheds further light on household cash flows and the differences in net income across the regions. While Eldorado Park, Orange Farm and Cape Town follow a very similar income and repayment capacity trend, households in Ekurhuleni have much lower net incomes and repayment capacities. Some of the indicators across the three regions are depicted in the following chart (Figure E1). Figure E1: Household Income and Repayment Capacity by Region 23

26 E.2. Housing Quality The survey interviewed a diverse group of respondents living in a variety of different types of housing structures. Two observations stand out. First, half of the group in Cape Town (5) lives in RDP houses compared to very few in Eldorado Park and Orange Farm () and Ekurhuleni (5%). Second, half of the respondents in Ekurhuleni (48%) live in shacks in informal settlements, which is not a characteristic that was observed in the two other areas, and thus translates into the lower value of the homes noted above. Figure E2: Use of Financial Services by Region E.3. Use of Alternative Financial Services The use of financial services varied to some degree across the three regions (Figure E2). Whereas there was a similar trend in the use of formal savings among the three regions, participation in informal groups varied and was markedly higher in Ekurhuleni. This is not surprising given the lower income levels and repayment capacity, which limit both perceived and actual access to formal financial services. Participation rates in informal groups (stokvels) were as high as half of all respondents in Ekurhuleni (48%), while only one-fourth or less in Eldorado Park and Orange Farm and Cape Town (2 and 25%, respectively). Consistent with this trend, the use of informal loans is lower in Cape Town compared to the two other regions. The variation in the use of informal loans across the three regions is mainly manifested in the source of informal borrowing. More Cape Town respondents request loans from employers than their peers in the other two regions (29% in Cape Town compared to 2% Ekurhuleni and 7% Eldorado Park and Orange Farm). In addition, it should also be noted that the average value of informal loans used by the respondents in Cape Town is much higher than the average of the three regions with an average loan value of R2,635 (R1,000 median) compared to the overall average of R1,194 (R500 median).

27 The use of formal loans in Ekurhuleni is also markedly less than in the other regions, where the majority of all respondents (88%) have never applied for a formal loan, citing lack of financial documents (37%) as one of the major obstacles. This is considerably higher than in Cape Town where it was only the case for a few (1) who had never applied. It is also interesting to note that of the respondents who had taken a formal loan at some point in time, those from Eldorado Park, Orange Farm and Cape Town were overwhelmingly more likely to have title to a building to use as collateral than those in Ekurhuleni who were more reliant on bank accounts and salaries. Of the formal loans that were taken by respondents in Ekurhuleni, the vast majority (85%) were for personal use compared to those in Eldorado Park and Orange Farm (55%) and Cape Town (5), (Figure E3). The subgroup using formal loans in Ekurhuleni, however, reported a lower average loan value than the other two regions; R10,875 (R5,000 median) in Ekurhuleni compared to R31,095 (R8,000 median) in Eldorado Park and Orange Farm and R23,078 (R9,000 median) in Cape Town. Figure E3: Formal Loan Type by Region Finally, potential demand for general formal individual loans is highest among respondents from Eldorado Park and Orange Farm (5). This group also exhibited the strongest interest (62%) in housing finance. This is not surprising given the higher income levels and limited availability of informal borrowing such as from employers which is more broadly available in Cape Town. In contrast, potential demand for general formal loans is more limited in Ekurhuleni (2) and Cape Town (13%). However, in both locations housing loans are still in highest demand among all products. Additionally, it is not surprising, given the lower standard of available housing in Ekurhuleni, that the majority of all respondents (88%) were interested in housing loans, in spite of or perhaps because of the fact that they have the lowest reported average income and repayment capacities in the survey. 25

28 The average value of R131,140 for potential housing loans reported by respondents from Cape Town (R100,000 median), is almost three times the amount reported by respondents in the two other areas 48. Additionally, the interest for a 133-month loan term and 19 percent interest rate are almost double those reported by the other regions. Respondents from Cape Town were not interested in requesting housing loans for construction, but indicated their preference for either renovating their existing house, or buying a new one. F. Gender Variations Some variations are observed while examining the data by gender. First of all women respondents were significantly more likely to be single or widowed (53%) than their male counterparts (28%). Additionally, women were more concentrated in domestic work (16%) compared to male respondents (3%) where wages are lower and there is less income security. This is directly reflected in lower primary incomes for women (R2,088) than men (R2,870), which in turn is translated into lower repayment capacity. When looking at the use of alternative financial services (Figure F1), it is surprising that the interest rates females pay for informal loans (7) are much less than what male respondents pay (15). When it comes to formal financing, slightly less women seek formal loans than men; however, the high overall rates of abstention reflect a somewhat gender neutral self-selection process. Finally, more female respondents reported that they depend on subsidies and remittances than males. Figure F1: Use of Financial Services by Gender 48 R55,563 in Eldorado Park-Orange Farm and R46,508 in Ekurhuleni.

29 G. Market Opportunities The market demand assessment for housing finance in South Africa helps to shed light on both the effective and potential demand for financial services among the low-income population in general, as well as to give insight into their particular needs and requirements for housing finance. By looking at a range of economic actors (including microentrepreneurs, low-income employees and laborers) from a diverse set of urban settings, the survey is able to provide results that reflect the characteristics of a representative sample of low income earners in South Africa, taking into account social, economic, and ethnic factors. From the survey we can build a profile of the typical respondent: a 40-year-old head of household with a family of three, living in their own home. He or she earns roughly R2,461 from his or her primary income activity and supplements from other sources, bringing total income amounts to R3, Taking the average household expenses of R2, into consideration, households are left with an average of R1,402 cash on hand, yielding a R effective monthly repayment capacity, which could be used to pay for loans to meet current and future housing needs. In general, a loan assessment involving monthly loan repayments up to this value would not induce risk of non-repayment among potential clients due to debt overburdening. For the most part respondents meet their current financing needs through the informal financial sector having largely self-selected themselves outside the formal lending market. They primarily participate in SACCOs, rely on family and friends, or in some cases employers or even money lenders, to borrow for a few months at a time amounts that hover around R1,200 or less. Most of this financing is used for personal expenditures, or consumption purposes, with very little being used for home improvement or purchases. The vast majority of respondents reported never having requested a loan and only one-third expressed a desire to apply for formal loans. However, it seems that these figures have more to do with psychological barriers to formal finance reinforced by a general lack of information on potential products and services than real demand. 49 $ average household income. 50 $ average household expenses. 51 $ effective monthly repayment capacity. 27

30 Those who do have access to formal bank loans borrow typically around R25,000 or less, repaid over a couple of years. And of those who reported taking a formal loan within the past year, one-third used the money for home improvement but very few for mortgages. There is, however, a much more dynamic market penetration by formal institutions when it comes to savings mobilization, evidenced by the majority of all respondents holding a savings account. The diagram below summarizes both the use of informal loans and participation in informal groups, as well as the use of formal loans and savings accounts offered through the financial markets in South Africa. Additionally, it presents the interest in future individual loans, and specifically home improvement loans. Figure G1: Use of Financial Services While the potential demand for financial services clearly exists, it has not been effectively translated into formal access to finance. This is due in part to a general unfamiliarity with and lack of experience navigating the formal financial sector, as well as fears about the ability to repay and successfully meet the financing requirements. In spite of the general reluctance to take a formal loan, when asked more specifically about meeting their housing needs, there was a much larger response in favor of these types of loans. Many respondents were interested in building or acquiring new homes, which they estimated to cost between R60,000 and R80, When asked specifically about their plans for future home improvements, almost half of the respondents expressed an interest in undertaking home improvements at an everage estimated cost of R44,430. While there was some interest in business and personal loans it involved much smaller amounts. In all cases, respondents were reporting reasonable amounts that could be repaid in monthly installments, in line with their estimated repayment capacity calculated at about R For new housing, this estimate is unrealistic. According to ABSA, the average cost of the most basic, newly built house in 2006 was R216,857; by the third quarter of 2007 this had risen to R265, 917 (this category includes all houses between 40-79m2 and costing less than R ). While houses in the resale market were available at a lower price, this market was very thin and few would have cost less than R100, 000. It is unlikely that households in this income segment would have entered a mortgage transaction with equity therefore a 10 loan to value ratio could be expected (and was permissible given most banks policies at the time). Such loan amounts therefore could only apply to home improvements or incremental construction.