Financing Housing Development in an Underdeveloped Financial Market: Learning from Developers’ Financing Adaptations?

Abstract

Research shows that most households in developing economies rely on informal housing finance and self-help because of the challenges imposed by the underdeveloped capital markets. How housing developers navigate these challenges is less well understood, but this understanding is necessary to develop innovative financing solutions that efficiently meet the escalating housing demand in these economies. We fill this gap by examining the entrepreneurial adaptative financing of developers in Ghana. Our interest is the alternative financing mechanisms developers are innovating to economize the transaction costs (TCs) of accessing capital. These are examined through the lens of TC economics. Instead of debt (only 1.2% of developers used exclusively formal financing sources), we observe private developers’ efforts to coordinate finance hierarchically, in some cases adopting the roles of traditional banks. For most financing mechanisms, the homebuyer either co-creates or temporarily co-owns the property with the developer. The eight identified adaptive financing models economize TCs and traditional financing usage and support housing consumption, offering buyers pathways to homeownership without conventional mortgages. However, they do not support volume development and expose homebuyers to significant ex-post contractual hazards. Our study provides insights into creating responsive policies that increase housing finance accessibility, while protecting homebuyers.

Keywords:

• Capital sources
• financial markets
• Ghana
• housing developers
• housing financing
• transaction costs

Housing finance plays a significant role in housing delivery, and without financing, housing delivery may not occur. The importance of housing finance is evident in both the demand and supply sides of the housing delivery chain. Households and investors rely heavily on financing to decide on housing tenure and investment strategies. Similarly, housing developers who are in the business of supplying houses are also reliant on financing in various forms to develop the mass housing projects required to deal with the enormous housing deficit confronting many regions of the world, particularly Africa. Consequently, housing finance is a critical area occupying the attention of policymakers around the globe.

Housing delivery is a capital-intensive venture, and investments have long time horizons before potential profits are realized. Construction finance from banks and mortgages is the primary source for financing housing in countries with mature financial markets (Reed & Sims, 2014). These financing sources are less accessible in developing economies, and Bah et al. (2018) argue that housing finance on the African continent has not kept pace with the demand for housing. Inaccessibility to formal housing financing sources hampers the ability of private housing developers to supply houses, leading to a high and growing unsatisfied demand for housing among African households, requiring pragmatic financing solutions. Countries in Sub-Saharan Africa (SSA) have persistent precarious housing conditions, overcrowding and deficits, with over 59% of the urban population occupying substandard housing (UN Habitat, 2016). It was estimated that 200 million people would live in African slums by 2020, and this number is expected to grow at 4.5% per annum (Parby et al., 2015).

The underdeveloped Ghanaian financial market typifies that of developing economies, where formal housing finance is scarce and expensive, causing households to rely on informal financing mechanisms (Addaney et al., 2015; Asante et al., 2018). As such, informal housing finance (e.g., incremental self-financing, personal savings, loans from friends and family, and sweat equity) supplies 90% of housing in the country (Habitat for Humanity (HFH), 2011). Housing mortgages are expensive and inaccessible to most households due to unfavorable lending conditions (Adade et al., 2022; Afrane et al., 2016; Aha et al., 2014; Ayitey et al., 2010; Warnock & Warnock, 2008). The average interest rates on mortgages are 12–24% (Centre for Affordable Housing Finance in Africa, 2021). The literature has recognized the unsupportive financial market conditions, the limited access to finance, and the associated challenges (Addaney et al., 2015; Amos et al., 2015; Ayitey et al., 2010; Donkor-Hyiaman and Owusu-Manu, 2015). Berner (2000) concludes that households rationally undertake self-help housing financing outside the established regulatory frameworks in developing economies as formality is expensive.

Housing developers, like households, also face unconducive conditions in the finance market, including prohibitive cost of finance, contributing to the expensive housing they supply. According to developers, the cost of financing constitutes 15–20% of the total housing development cost for those using bank loans (HFH, 2011; Karley, 2008). Ehwi (2020) and Bank of Ghana (2007) observe the widespread use of equity among private developers in Accra. Many developers do not meet the credit requirements of banks, which includes a track record within the industry, membership in the Ghana Real Estate Developers Association (GREDA), 3 years of financial performance, management strength, and experience (Acquah, 2011), because most are recently established companies. According to Owusu-Manu et al. (2015), developers often use bank loans as a last resort.

Market conditions provoke different adaptive responses from households and entrepreneurs, leading them to abide, alter, evade, or exit specific institutional mechanisms (Bylund & McCaffrey, 2017). What remains unclear is how developers respond to the unconducive financial market conditions—because debt is taken as given in mature markets, where most real estate literature originates. Studies on property companies’ capital decisions in Western markets (Brounen & Eichholtz, 2001; Morri & Cristanziani, 2009; Westgaard et al., 2008) do not illuminate the situation in underdeveloped financial markets. Specifically, our interest is in the Ghanaian housing developers’ entrepreneurial adaptations to the underdeveloped financial market in financing supply and facilitating housing consumption. de Haas and Peeters (2006) observe that companies generally use little debt where the finance market is underdeveloped, suggesting that developers have found “nonstandard” financing mechanisms.

Although studies have examined ways to develop formal housing finance, such as leveraging pension funds (Donkor-Hyiaman & Owusu-Manu, 2016), the contention of this paper is that we can learn from developers’ adaptive finance strategies in formulating responsive affordable housing financing policies. Transaction cost economics (TCE) provides a theoretical framework to understand and examine these financing adaptations. Transaction cost (TC) theory argues that the cost of transacting finance between two agents (e.g., due diligence search and negotiations) influences the choice of capital sources entrepreneurs use, with a preference for capital with comparatively lower TCs. Thus, developers will use or create the most cost-effective capital, which should be noticeable from the prevalence of a source of capital.

This paper contributes to the housing financing literature in at least three primary ways. First, we explore housing developers’ capital practices in Ghana in delivering housing and facilitating consumption. Although developers perceive debt as the essential capital source (Owusu-Manu et al., 2015), our findings show that capital practices differ, with many informal elements, although developers are operating in the formal economy. Second, we examine the emergent entrepreneurial innovations of developers in response to the unconducive financial market conditions. Developers have devised solutions to finance housing delivery and facilitate home purchases, in some instances adopting some traditional banking functions. Our findings show that developers are not passive actors in the financial market; they actively create capital sources to thrive where available capital sources are associated with high TCs. Eight unorthodox finance models are observed and detailed, with implications for other developing economies. We provide evidence that housing developers and their business models differ from Western practices reported in extant literature. Finally, this paper concludes with lessons and policy implications.

The remaining sections of this paper progress as follows: section 1 presents the theoretical inspiration for the paper to elucicate the character of the finance market and developers’ solutions. It also discusses relevant literature on Ghanaian finance market conditions and housing development financing. Section 2 details the methodology. Section 3 presents the research findings, whereas section 4 comprises the discussion, with section 5 used for conclusions and practical implications.

1. Literature Review

1.1. Theoretical Background—Transaction Cost Economics

TCE recognizes and emphasizes the cost of economic exchanges: due diligence, coordination costs, contract haggling costs to resolve ex post misalignment of interests, policing, and resolving disputes. Because exchanges of funds are transactions (even where one is accessing one’s own funds in a bank), they have associated TCs. Both borrowers and financiers incur and minimize transaction hazards. For example, borrowers’ TCs include administrative expenses, communication costs, opportunity costs of time, information search costs, broker fees, and commissions. In financial intermediation, interest costs of debt capital are not TCs per se, but they contribute significantly to high interest rates, especially on microloans (Shankar, 2007).

Williamson (1988) insists that it is meaningful to conceive debt and equity as governance mechanisms. The transacting parties align their interests with the hazards, risks and uncertainties associated with each financing mechanism. TCE suggests that people’s choice of investment financing depends on the TCs involved and their ability to economize such hazards in exchanging funds. Debt is akin to market transactions (financial transactions across two independent firms or individuals). Equity is transacted in firms (coordinating capital hierarchically within an independent firm). Even where equity is traded, holders of a company’s stock are part owners who benefit from its profits and are liable to the business’s associated risks up to the nominal value of their investments.

As governance structures, “debt and equity have different levels of ex-ante costs (benefits) and ex-post costs (control rights)” (Kochhar, 1996, p. 717). Debt in the context of TCE is more expensive than equity and more suitable for less asset-specific investments. Equity is more tolerant of investments generating highly specific assets in illiquid markets. Thus, the cost of using debt and equity depends on asset specificity and uncertainty associated with the intended investment. Asset-specific investments have high sunk costs and are more challenging to redeploy. Consequently, the more specific the investments (e.g., locational, firm-specific, technologically specific), such as housing in an underdeveloped market, the less likely these assets can easily be repurposed to their original value in a liquidation (Williamson 1991). Because debtholders are unforgiving, they are likely to provide funding with adverse debt terms in such situations. Corporations prefer equity over debt funding to reduce scheduled debt servicing and liquidation risks (Smith, 2012) which are more consequential in housing. Debt and equity financing have distinct ramifications for entrepreneurs (Schmidt, 2003; Ueda, 2004).

From the TCE perspective, there is a spectrum of governance mechanisms for financing investments, depending on the associated TCs. At one end is financial autarky, where one is completely financially self-sufficient and thus enters no contracts with other individuals or economic organizations and incurs zero TCs. The other end is debt. Between the ends of this spectrum are varieties of financing mechanisms with different TCs. For example, borrowing from family and friends incurs lower due diligence, contracting, compliance, and negotiation costs than borrowing from banks and issuing bonds. Existing filial trust protects the transactions in the former. Venture capital, private equity, angel investing, and mezzanine financing, among others, all have different scales of associated TCs, including accounting fees, cost of due diligence, account establishment fees, negotiation, legal fees, and closing fees, among others.

Imperatively, where a financing mechanism is regulated, and transactions are at arm’s length, the associated TCs are higher because of the need to close the asymmetric information gap between lender and borrower, intermediation costs, and regulatory compliance. Such financing also provides a high level of protection to transacting parties because it reduces potential ex post opportunistic behavior. We see that such capital sources are formal. These financing mechanisms are governed by legal and contractual frameworks, and typically involve third-party intermediaries. Furthermore, financial contracts and disputes are typically enforced through the statutory legal system, which includes courts. Informal finance is distinguished as relationship-based, nonarm’s length. Financial decisions and transactions include private information and have less protection for lenders in case of defaulting (e.g., peer borrowing). The capital is transacted along personal relationships rather than through intermediaries. Trust, reputation, cultural norms, and beliefs are the main ways of enforcing contracts instead of complex legal contracts and intermediaries. As such, they tend to have lower TCs but can have substantial ex post hazards. The fund’s relative costs and benefits and the firm’s circumstances influence its choice of formal or informal financing (Nguyen & Canh, 2021). In other situations, the developer may fail the requirements of raising funds, such as through bonds, limiting their ability to use that source of (formal) finance.

From this understanding, it becomes intuitive why personal savings, peer lending, networks, and borrowing from friends and family predominantly finance housing by households in developing countries. As for households, the extent of TCs influences entrepreneurs’ capital choices—create, use, evade, or exit a capital source (Kerins et al., 2004). The TCE framework is useful because it allows for a wide range of financing models to be studied—as we intend to do in this article. For example, Graham and Harvey (2001) found weak evidence that TCs influenced executives’ capital decisions. TCs are more influential in the capital of large than small firms for market-based finance (Beattie et al., 2006).

The predominant use of specific funding mechanisms shows the TC-minimizing governance mechanisms that actors have identified (Buckley & Chapman, 1997; Wang, 2003; Williamson, 1988), because the choice of financing structure depends on the trade-off between its benefits and costs, including contractual hazards associated with that financing channel. It is unreasonable for entrepreneurs to choose expensive funding mechanisms if cheaper alternatives exist. Thus, entrepreneurs create and facilitate new institutions to emerge or adapt existing ones to foster their profit interests (Klein, 2016).

1.2. The Housing Market in Ghana

Ghana’s housing deficit witnessed a 33% reduction, from 2.8 million (2010) to 1.8 million (2021) (Ministry of Works and Housing, 2022), partly attributable to the real estate boom and a 72.8% increase in residential structures within the period (from 3,392,745 in 2010 to 5,862,890 in 2021) (Ghana Statistical Service [GSS], 2022). The most prevalent types of dwelling units in the country are separate houses (detached, 63.3%) and compound houses (20.9%). Nationally, 64.1% of dwelling units have cement blocks or concrete as the primary construction material for outer walls, with Greater Accra and Ashanti regions accounting for close to half (47.5%) of these dwelling units and an urban representation of 82.5%. About half (48.4%) of households live in owner-occupied dwelling units (Ghana Statistical Service [GSS], 2022).

The above statistics indicate that Ghana’s housing market has undergone significant changes over the past few decades. Across developed and transition economies, such changes are driven by housing policies and institutional and regulatory frameworks formulated explicitly or implicitly for addressing housing delivery—quantity, quality, and cost. Ghana’s National Housing Policy of 2015 is the most recent policy document that guides housing delivery. The Government of Ghana envisions a country where everyone can access safe, secure, decent, and affordable housing, either owned or rented. The current government’s housing policy emphasizes creating an enabling environment for private sector participation in housing delivery. Nevertheless, like previous policies, implementation remains modest, with the few affordable housing projects overpriced and uncompleted.¹

The country’s housing delivery system consists of two approaches—informal and formal (HFH, 2011). The informal approach, an incremental process led by owners (individual householders) rather than institutional players, dominates the system. The process typically starts with acquiring land from customary landholders, followed by the engagement of a mason, other tradespeople, and laborers whom they find through personal recommendation. Construction work starts with or without formal architectural and engineering drawings and building permits and continues so long as there are funds for materials and labor. The owner may occupy the uncompleted house while directly negotiating for infrastructure with service providers. Although efficient at getting many housing projects to start simultaneously, the approach is overall inefficient because completion is usually delayed due to financial constraints.

The formal system, which is dominated by private development companies (i.e., companies that are members of GREDA) and government (and its parastatals, including the State Housing Company, Tema Development Corporation, and Social Security and National Insurance Trust), contributes about 10% of all dwellings and an even smaller percentage of rooms. It mainly supplies the middle-class and high-income households in the formal urban sector and Ghanaians living in the diaspora (Teye et al., 2015). This role is arguably due to the challenges formal developers face, including financing, which is discussed in the next section.

1.3. Financial Market Conditions and Housing Financing in Ghana

Ghana’s financial market is fragmented and developing. Ghana’s domestic credit to the private sector (11.5%) in 2020, as a measure of the level of financial market development, is lower than those of South Africa (107.9%) and lower middle-income countries (46.9%) (World Bank, 2020). The cost of mortgages (18.7–31.7% per annum) is expensive (Bank of Ghana, 2019). The total mortgage debt to gross domestic product (GDP) ratio is 0.78% (Centre for Affordable Housing Finance, 2021). The stock market is similarly underdeveloped, with 38 listed companies, with low trading volumes, and a market capitalization of GH₵56.8 billion as of December 2019 (Securities and Exchange Commission, 2022).

Characteristic of inefficient financial intermediation and developing and noncompetitive financial markets, Ghana’s interest rate spreads (IRS) ranged from 16.3% to 29.5% from 1980 to 2012. These IRS are among the highest in Africa (Mathisen & Buchs, 2005; Obeng & Sakyi, 2017). Interest on loans to the construction sector range from 19.0% to 39.3% per annum (annual percentage rate) (Bank of Ghana, 2019). Asymmetric information worsens the cost of capital, including the TCs that small- and medium-scale enterprises (SMEs) incur to borrow. In some sectors, TCs incurred are equivalent to 10% of the total approved loan amount (Antwi & Ohene-Yankyira, 2017).

Obeng and Sakyi (2017) and Kwakye (2012) posited that competitive government borrowing, high cost of bank funds and high lending risks, low savings, and macroeconomic volatility had sustained the high lending interest rates banks charge. Conversely to the high overhead cost and high credit risks, Ghanaian banks’ pretax returns on assets and equity are the highest in SSA (Nissanke & Aryeetey, 2008; Obeng & Sakyi, 2017). These returns contradict the conventional view that credit risks should depress banks’ profitability. According to Mathisen and Buchs (2005), the entry barriers and industry characteristics have shielded the high profits banks earn and prevented international competition. Banks in fragmented markets rarely face competition and charge monopolistic interest rents (Nissanke & Aryeetey, 2008).

Because of the housing finance market conditions, long-term financing options are scarce, making capital (extremely) expensive and limiting developers’ finance choices (Nissanke & Aryeetey, 2008; Owusu-Manu et al., 2015). Other sources of capital in developed financial markets, including syndicated funds, bonds, and mortgage-backed securities, are either unavailable or underdeveloped. Consequently, developers’ capital sources are incongruent with housing development in Western markets. Ehwi (2020) found that Ghanaian developers frequently used equity rather than debt. Of the sampled developers, 72% use internal equity, and 64% use clients’ deposits. Only 34% of developers use debt from financial institutions (the sum exceeds 100% as developers combined multiple financing sources). Many developers use short-term bank loans as a last option (Owusu-Manu et al., 2015). On the other hand, in the UK, for example, capital from banks, insurance companies, and pension funds constitute a significant share of housing finance (Ratcliffe & Stubbs, 2009). Superannuation funds and insurance companies also engage in direct housing investments, for example acquiring a housing developer (Reed & Sims, 2014).

Banks’ lending to developers is minimal, with less than 1% of total bank loans and industry advances annually going to the housing sector in Ghana (PricewaterhouseCoopers (Ghana), 2019). At the same time, they impose strict credit conditions and collateral requirements and unclear and lengthy loan application processes on developers (Amos et al., 2015). Additionally, banks require housing developers to provide a track record within the industry, membership in GREDA, 3 years of financial performance, management strength, and experience (Acquah, 2011). The short history of many developers and builders and their limited collateral affects their risk profiles (Amos et al., 2015). Loans extended are typically repayable within three years (Amos et al., 2015).

Furthermore, uncertainties with land tenure, weak land titles in Ghana, and weakly enforced statutory property rights have reduced land security and bankability (Gough & Yankson, 2000; Kasanga et al., 1996). Similarly, Leon (2008) indicated that fragmented land ownership inhibits housing development financing. Domeher et al. (2016) observed that banks preferred forms of collateral other than titled landed property or imposed additional requirements when titled landed property was accepted. Registered titles as collateral neither affected loan conditions nor improved banks’ lending decisions favoring SMEs (Domeher et al., 2016). Limited transparent information on the Ghanaian housing market elevates banks’ perceived risks.

Despite the unconducive financial market conditions, which adversely affect developers and homebuyers, housing supply and consumption have been steadily growing. The real estate contribution to GDP increased from GH₵1,173.3 million in 2013 to GH₵1,629.2 million in 2020 (at constant prices). In 2020, the real estate sector grew by 9.6% on a quarter-on-quarter basis with the persisting COVID-19 pandemic (Ghana Statistical Service, 2021). This situation suggests that housing developers have entrepreneurially devised alternative (nonmarketed) financing mechanisms to facilitate both supply and consumption of residential properties, as mortgages and development finance remain expensive and relatively inaccessible. As already stated, the focus of this paper is to identify and examine these innovative financing mechanisms of Ghanaian housing developers. The diversity of developers’ financing solutions (evading financing market conditions) empirically examined in this paper has practical and public policy implications.

1.4. Housing Finance Sources in Africa

Own equity, which is the cash contribution by a developer to a project, is one source. Equity is the last to be repaid in terms of order of payment to various sources of finance after all debts have been repaid (Wilkinson et al. 2008). Equity contribution is determined by the loan-to-value ratio accepted by the bank providing senior debt. For example, the maximum contribution by banks in Australia is about 70% of the project cost, and 30% is financed via equity or combined with mezzanine debt. However, this is not the case in many African countries, as banks provide less financing for real estate projects. As observed in our study, the developer’s equity tends to be significant, with many developments financed entirely with equity.

Second, historically, many African countries, including Ghana, Cameroon, Ethiopia, and Algeria, set up state-owned banks to provide housing finance to households and developers (Bah et al., 2018). Bah et al. (2018) indicated that many of these financial institutions have collapsed, leaving only those receiving government bailouts and still providing funding for housing developments.

Third, there are traditional commercial banks that provide short-term development financing. Owusu-Manu et al. (2015) argue that developers use short-term bank loans, as they are aligned to project duration, to decrease borrowing costs. Developers source senior debt from commercial banks. When necessary, mezzanine debt (which covers the gap between senior debt, available equity, and total development costs) may be arranged from other sources, including syndicates, private equity, and foreign banks (Rowley et al., 2014). Moreover, in some cases, housing developers engage other equity partners who provide land as equity. In this case, the land is valued, and the amount becomes the partner’s equity contribution.

A housing bond is a debt instrument secured against a housing development project (London House Exchange, 2018). For example, Shelter Afrique (based in Nairobi, Kenya) successfully issued five housing bonds between 2005 and 2014. However, the issuance of bonds in Ghana is rare due to a lack of good credit ratings and developers’ financial credibility (Bank of Ghana, 2007). Another funding source for some housing developers is to adopt land banking or land flipping. Land banking is acquiring and holding land before development (Bao et al., 2012). Some developers purchase more extensive tracts of land than required for a project and resell portions after prices have increased. Van Noppen (2012) suggests that if there is significant land price appreciation, the effect is cost-free land for the actual development upon completion, as the price appreciation may far outweigh the cost of the land. Moreover, in some cases, housing developers engage private equity holders who provide funding as equity. These may be individuals with high net wealth.

2. Materials and Methods

We used a mixed-methods research design, with interviews and surveys happening concurrently. Greene (2007, p. 20) refers to mixed methods as “multiple ways of seeing and hearing.” Combining findings from quantitative and qualitative designs allows researchers to deepen understanding of complex phenomenon (Creswell & Clark, 2018). The nature of the research problem motivated the mixed methodology. First, it is essential to understand the nature of development financing mechanisms and their distribution among developers. The survey allowed researchers to identify the diversity of financing mechanisms among the general population of developers. The interviews assisted in identifying and detailing developers’ financial creativity while understanding their financing behavior.

The criteria for participating in the survey were: (a) membership in GREDA, and (b) being registered and operating within either the Greater Accra Metropolitan Area (GAMA) or the Greater Kumasi Metropolitan Area (GKMA). Although membership in GREDA is not compulsory for developers, it is the only association of real estate developers in the country, recognized nationally, and represents developers in the country in all government engagements. The directory thus served as the sample frame for the study. Developers are different from landlords who built for themselves and may lease any unused rooms in their homes. Informal landlords, in contrast to developers, are not legally separate entities from owners. There are no complete available accounts that permit a financial separation of their production activities from the other activities of their owner(s) (cited in Hussmanns, 2004). Although affluent landlords may own multiple houses, they are not registered as real estate developers and operate largely in the unregulated market.

We purposely selected these two metropolitan areas because they constitute the country’s gateway housing markets and have the highest congregation (96% of GREDA members as of 2020) and diversity of developers. To improve the response rate, developers who had no active phone number in the directory, had stopped housing development, or were uninterested in the survey upon initial phone contact were eliminated. One hundred and forty-four residential developers out of 180 remained in the sample frame, 93% and 7% in GAMA and GKMA, respectively. The entire sample frame was sent either a web-based questionnaire using Qualtrics or a hardcopy questionnaire if that was their preference. Ninety-six developers answered the questionnaires, and 82 of the returned questionnaires were completed appropriately and used for further analysis. After examination, a few variables had missing values (<4.9%). Item missingness of less than 5% is unproblematic for statistical analysis (Schafer, 1999). Nonetheless, these data were imputed using joint multiple imputations (multivariate normal imputation) and 20 imputations without postestimation naïve rounding (McNeish, 2017).

Two developers operated in the GKMA; the rest were from GAMA. The survey targeted senior management of the respective companies. This response rate exceeds Owusu-Manu et al. (2015)’s response rate of 53 survey respondents, which included builders/contractors. Ehwi (2020) reported findings based on 51 residential developers, and the Bank of Ghana (2007) studied 22 housing developers.

For the interviews, participants were selected using stratified random sampling. At least one developer was randomly selected and interviewed from each stratum (quasi-state developer, large-scale, medium-scale and small-scale developers). Interviews continued until no new financing mechanisms were identified, indicating a saturation point had been reached. The semistructured interviews were flexible, allowing developers to expand on their experiences with capital and their entrepreneurial capital solutions. All interviews were confidential. In total, 15 interviews were conducted. Creswell and Poth (2017) suggested that saturation can be reached with 10 in-depth interviews. The distribution of interviews per stratum is available on request to reduce the probability of reidentification of developers because there are only two quasi-state developers in the country. The quantitative data are analyzed descriptively, with relative importance indexing and nonparametric techniques. Thematic analysis is used to analyze the qualitative data, following the process outlined by Elo and Kyngäs (2008).

3. Results

3.1. Respondents’ Characteristics

All respondents were in management positions, with most being general/project managers (40.2%) and chief executive officers (CEOs)/directors (23.7%). In terms of educational attainment, 48.8% had postgraduate degrees (see Table 1).

Table 1. Qualifications and positions of respondents (%).

	N	Prebachelor’s	Bachelor’s	Postgraduate
General/project manager Operations manager	33	15.2	42.4	42.4
Chief Executive Officer/director	19	26.3	15.8	57.9
Quantity surveyor	7	0.00	57.1	42.9
Architect	5	0.0	0.0	100.0
Project/site engineer Site manager	7	42.9	28.6	28.6
Sales/financial/marketing manager	6	0.0	33.3	66.7
Procurement manager	1	0.0	0.0	100.00
Others	4	0.0	100.0	0.00
Total	82	15.9	35.4	48.8

Note. Prebachelor’s qualifications = vocational/technical/apprenticeship (2.4%) and postsecondary diploma/higher national diploma (13.4%).

Source: Authors’ field survey (2020).

3.2. Residential Developers’ Characteristics

3.2.1. Years of Operation and Staff Size

Housing developers differ in characteristics other than size, scope, and operational norms (Coiacetto, 2001). On a broad scale, developers in Ghana resemble each other in terms of the organization of housing development, staff size, the scale of production, and annual sales or turnover. Few developers (9.8%) have been operating for 26+ years. Most developers (57.3%) have been operating for 10 years or less; the average was 12 years. The findings in Table 2 are similar to Ehwi’s (2020) findings that 38% and 26% of developers had less than 5 and 5–10 years of experience, respectively. The age of housing developers indicates a nascent residential market and developers’ experience in Ghana’s residential market.

Table 2. Developers’ characteristics.

Years of operation
Up to 5 years	24.4%
6–10 years	32.9%
11–15 years	18.3%
16–20 years	8.5%
21–25 years	6.1%
26+ years	9.8%
N	82
Full-time staff
Up to 10	22%
11–20	18.3%
21–30	9.8%
31–40	14.6%
41–50	20.7%
51–60	3.7%
61+	11%
Mean^b	34
Average housing units built annually
Up to 10 units	42.7%
11–20 units	15.9%
21–30 units	3.7%
31–40 units	6.1%
41–50 units	13.4%
51 units +	18.3%
N	82
Mean^a	31
P50	15
P90	80
SD^a	35

Note. ^aExcludes a developer building 400 units per annum.

^bExcludes a developer with 500 staff.

Source: Authors’ field survey (2020).

Indeed, private formal housing development in Ghana is a recent phenomenon. In the 1970s, there were no houses advertised for sale in the national newspapers (Grant, 2007). Householders, contractors, informal sector builders, and paragovernment agencies supplied housing in urban areas. According to Grant (2007), public and quasi-public housing accounted for 15% of dwelling units in Accra by 1990. Before the structural adjustment programs and market liberalization in the 1980s, people seldom bought or sold housing (Quarcoopome 1992, cited by Grant, 2007). With market liberalization came formal private housing development, foreign direct investment in housing, and gated communities. GREDA was formed in 1988 with 38 members who were mostly housing contractors (Grant, 2007). It is thus unsurprising that of the eight developers in business over 30 years, N = 2 are quasi-state organizations (initially state housing organizations).

Similarly, housing developers have relatively small sizes. Forty percent have up to 20 staff. One developer reportedly had about 500 employees but was a quasi-state organization, initially established by the state in 1952 to plan, develop and manage lands in the Tema Acquisition Area. It is one of the oldest housing developers in the country. Staff size (measured by the nature of staff contracts—permanent staff) is, however, not a good predictor of the developer’s size because a developer can outsource, for example, design and construction. Also, some land developers may not engage in housing development. If we conceive the developer as a nexus of contracts (Demsetz, 1997), a firm’s staff size becomes less relevant. The entrepreneur can use various modes of production governance to supply housing. What is essential is the relationship between staff size and the governance modes of housing development.

3.3. Annual Supply of Housing Units

Most developers (58.6%) are small-scale developers, delivering up to 20 housing units per year (Table 2). One developer was new, with no developments as yet. Three developers reported the highest number of housing units supplied, an average of 150. This study’s findings corroborate those of the Bank of Ghana (2007), which found that the leading housing developer in terms of market share from 2000 to 2006 built 105–180 housing units annually. Other developers surveyed within the same period typically built fewer than 20 units annually. Gated communities within GAMA have an average of 115 completed housing units (out of 305 units planned; Ehwi, 2020). Estimates of the production volume by formal private sector housing developers are inconsistent. The Bank of Ghana estimated that residential developers built over 10,954 homes between 1988 and 2006 (Bank of Ghana, 2007). Notwithstanding the variations in estimates, the formal housing supply is relatively low, comparable to what has been observed in other African countries, such as Kenya, Tanzania, and Uganda (Arvanitis, 2013).

Given the scale of housing supply by housing developers in Ghana, relying on formal developers’ ability to supply the volume and diversity of housing needed in the Ghanaian housing market is currently impractical. Coupled with the type of capital developers used, supplying the volume of housing affordable to low- and middle-income households is contentious. The financing used in housing developments is directly linked to the scale of housing delivery. Without the use of external debt, production cannot be leveraged.

Production scale deficiencies are obvious, but also, developers supply largely to the middle and high-end markets. This study found that 29.3% of developers build housing priced at US$50,000 and below. Most developers (61%) operated in the US$50,001–100,000 market segment, followed by those at US$100,001–200,000 (50%). Twenty-eight percent of developers regularly supply housing priced over US$200,000. The percentages exceed 100% because some developers operate in multiple categories. Real estate developers usually price houses in U.S. dollars in Ghana to hedge against inflation and forex flux. Similarly, sales revenues are low, with only 9.8% of developers exceeding annual sales revenues of US$1 million. Most developers (28%) generate revenues of US$250,000–500,000.

3.4. Sources of Residential Development Finance

First, it should be recognized that all but one of the respondent firms were developer-builders, active in both the land market and construction. The study did not observe the typical land developer in Western markets whose business model is limited to developing land and selling serviced plots to clients (builders, individuals, and householders). Second, one source of observed funding (client deposits) is only used after the developer has already acquired the land and resells it to a client. All the other funding mechanisms (Table 4) can be used for both land acquisition and construction, except loans from mortgage companies, which are typically used for construction rather than land acquisition.

Retained profits, clients’ deposits, and private funds were the principal sources of development finance (by frequency of usage). Cumulatively, 39.3% of residential developers always used these funding sources. It is significant to note that 63.7% of developers reportedly never or rarely used bank loans (Table 3), contrary to what persists in Western property markets. The other sources of funding consisted of spousal support, barter trade, and personal mortgages, which 12.2% of developers reportedly used. These findings align with Ehwi (2020) and Amos et al. (2015). Ehwi (2020), for example, reported that developers’ sources of development financing of gated communities in GAMA were institutional investors (18%), developers’ equity (72%), loans from financial institutions (34%), and installment payments from prospective homebuyers (64%). Our findings provide a historical perspective of developers’ financing behavior and the diversity of creative financing models.

Table 3. Developers’ sources of capital and frequency of use for the past 5 years (%).

	Never	$\to$			Always	P50	P90
	(1)	(2)	(3)	(4)	(5)
Retained profits	4.9	15.9	30.5	30.5	18.3	3	5
Clients’ deposits	15.9	8.5	31.7	32.9	11	3	5
Bank loans	48.8	15.9	19.5	13.4	2.4	2	4
Loans from mortgage companies	57.3	11	23.2	7.3	1.2	1	3
Loans from rural banks	86.6	11	1.2	1.2	0	1	2
Private funds	28	15.9	26.8	19.5	9.8	3	5
Credit from suppliers	40.2	19.5	26.8	13.4	0	2	4
Others	87.8	0	12.1	0	0	0	1

Source: Authors’ field survey (2020).

A relative ranking importance index (RII) indicates the prominence of equity—retained profits and developers’ internal equity—over debt (Table 4). The top-ranked funding sources developers use are retained profits, clients’ deposits, and private funds. Thirty-two percent of developers use loans from mortgage companies (sometimes, most of the time, or always). Bank loans are less frequently used by developers.

Table 4. Relative ranking of sources of development finance.

Source	RII	Rank
Retained profits	0.685	1
Clients’ deposits	0.639	2
Private funds	0.624	3
Credit from suppliers	0.534	4
Bank loans	0.409	5
Loans from mortgage companies	0.368	6
Loans from rural banks	0.234	7
Others	0.224	8

Source: Authors’ field survey (2020).

Tables 3 and 4 clearly show that residential developers are bank debt-averse in favor of equity. Retained profits, client deposits, and private equity are all forms of equity financing. Generally, debt sources are ranked lower than equity sources. The use of retained profits and client deposits takes away the pressure from bankers. Why did housing developers generally prefer such capital sources? Given the practice’s prevalence, explanatory factors must affect large- and small-scale developers alike. Loan application processes are unclear and lengthy, with cumbersome financial arrangements (Amos et al., 2015a). The cost of capital in Ghana ranged from 19.0% to 39.3% per annum (annual percentage rate on loans to enterprises for the construction sector in 2019) (Bank of Ghana, 2019). This high cost of borrowing is a strong disincentive to use debt.

Table 5 shows either no correlation or a weak but significant correlation among the capital sources of developers. The correlations are essential to see whether some funding mechanisms move together. The choice of one source of housing development finance was less related to the choice of another source. Loans from mortgage companies and bank loans have the highest correlation coefficient, suggesting that developers’ use of loans from mortgage companies increases with their use of bank debt. Using client deposits correlates with credit from suppliers. Generally, debt sources are correlated among themselves rather than correlated with equity sources, suggesting that the use of one form of debt tends to increase with the use of other forms of debt.

Table 5. Spearman rank correlation of housing development finance sources.

	Retained profits	Clients’ deposits	Bank loans	Loans from mortgage companies	Loans from rural banks	Private sources	Credit from suppliers	Others
Retained profits	1.000
Clients’ deposits	−0.179	1.000
Bank loans	−0.113	0.061	1.000
Loans from mortgage companies	0.013	0.089	0.458^c	1.000
Loans from rural banks	−0.073	−0.123	0.365^c	0.379^c	1.000
Private sources	0.105	−0.030	−0.173	−0.087	0.134	1.000
Credit from suppliers	0.063	0.306^c	0.272^b	0.311^c	0.193^a	0.147	1.000
Others	0.018	−0.071	0.024	−0.015	0.173	0.006	−0.068	1.000

Note. ^aCorrelation is significant at the 0.1 level.

^bCorrelation is significant at the 0.05 level.

^cCorrelation is significant at the 0.01 level.

Source: Authors’ field survey (2020).

3.4.1. Relationship Between Developers’ Characteristics and Funding Sources

Company characteristics—company size, cost of debt, ownership/control, and operational risk—affect property companies’ capital decisions (Morri & Beretta, 2008; Ooi, 1999; Westgaard et al., 2008). Such findings make it reasonable that developers with common identifiable characteristics should use similar funding mechanisms. Small developers without long histories should use equity more frequently than debt, and large-scale developers more debt because the latter has a more credible borrowing capacity than the former. This section examines this premise: whether a relationship exists between developers’ characteristics and funding mechanisms.

Spearman rank correlation tests show a positive relationship, albeit a weak association, between using retained profits, using loans from mortgage companies, and the number of years the developer has been operating. Older developers tend to use more retained profits and mortgage loans than young developers do, although the association is weak, as evidenced by the low correlation coefficient (rho 0.26 and 0.27, respectively; p < .05).

None of the measures of a developer’s size (measured by its annual sales turnover, full-time staff, and the number of housing units delivered per annum) is related to any capital source. In other words, we find no evidence to suggest that scale of development is linked to the type of funding developers use, which is reasonable given the broad commonality among capital sources of developers. Indeed, we observed that 45% and 40% of developers with an annual sales turnover of up to US$250,000 and over US$1 million, respectively, used mainly private funds. Relatively young developers may not have the credit history, market reputation, balance sheets, and assets to facilitate bank credit access. Small-scale enterprises’ credible credit information is often cumbersome to assemble and verify. Available credit data from credit reference bureaus only goes as far back as 2004 (Kusi et al., 2016). Our findings indicate that the choice of source of housing development finance was exclusive of the developer’s size, but two finance sources weakly correlated with the developers’ age.

The type of homebuyers could also influence the funding sources of developers. As the homebuyers of a developer become more foreign, the developer is less likely to use retained profits and instead use client deposits (Table 6). Nonresident Ghanaian homebuyers are associated with the use of bank loans, whereas residents are associated with financing via retained profits and loans from mortgage companies. Further research should explore why these distinctions, albeit moderate, exist.

Table 6. Spearman rank correlation of housing development finance sources and developers’ features using a 5-point Likert scale.

	Retained profits	Clients’ deposits	Bank loans	Loans from mortgage companies	Loans from rural banks	Private sources	Credit from suppliers	Others
Years of operation	0.264^b	−0.147	0.211^a	0.268^b	−0.023	−0.164	0.053	−0.068
Annual sales revenue	0.059	−0.083	0.162	0.091	−0.028	−0.062	−0.006	−0.168
Staff size	0.103	0.003	0.163	0.109	−0.07	−0.159	0.093	−0.17
Average housing units supplied annually	0.122	0.058	0.115	0.075	0.188^a	0.075	0.073	−0.074
Typical homebuyers
Foreigners	−0.246^b	0.422^c	0.157	0.038	0.063	0.093	0.176	−0.312^b
Nonresident Ghanaians	−0.037	0.174	0.304^c	0.17	0.248^b	0.082	0.269^b	−0.219^b
Residents	0.246^b	0.008	0.122	0.257^b	−0.001	−0.246^b	−0.156	0.096
Institutions	0.072	0.035	0.126	0.194^a	0.151	0.238^b	0.149	−0.286^b

Note. ^aCorrelation is significant at the 0.1 level.

^bCorrelation is significant at the 0.05 level.

^cCorrelation is significant at the 0.01 level.

Source: Authors’ field survey (2020).

3.4.2. Formal and Informal Financing and Developers’ Characteristics

Based on the analytical framework, loans from banks (including rural banks) and mortgage companies were classified as formal financing. Retained profits, client’s deposits, private sources, credit from suppliers, and others were classified as informal. Eleven percent of developers used a single funding source. Of these developers, one developer used a formal source. The remainder used only informal finance. The dominance of informality is further noticeable where developers combine multiple financing sources; on average, developers use three funding sources. Two out of the three sources, on average, were informal. In other words, a typical developer combines one formal and two informal sources of finance. The formal source was either bank loans or loans from mortgage companies. Furthermore, 51.2% of developers used no formal financing, compared to 1.2% of developers who used no informal financing. Forty-eight of the developers used at least both funding types.

We further compared the characteristics of developers who used no formal finance with those who used at least one formal finance type. Those who used only informal finance had smaller staff sizes, were younger, and generated less sales revenue. For example, those using no formal financing built 13 fewer housing units than those employing formal financing. Although the results suggest there may be distinct differences between developers based on financing choices, these differences are nonsignificant (Table 7). Where we see differences are the targeted markets of developers. The median values of the developers’ clients were compared using a two-sample Wilcoxon rank-sum test. Developers using at least one source of formal finance sold to more resident (z = −2.11, p = .035) and nonresident Ghanaians (z = −2.17, p = .030) than those using only informal sources.

Table 7. Mean comparison of developers’ characteristics by financing source using T test for equality of means.

	At least one formal finance		Only informal finance		t test
	M	SD	M	SD
Years in operation	14.9	1.9	10.7	1.69	0.09^a
Annual sales revenue^a	3	1.6	2.9	1.4	−1.28
Staff size	50	12.3	30.4	5.1	−1.49
Housing units supplied annually	42.3	10.6	28.9	5.7	0.26

Note. ^aBased on a Two-sample Wilcoxon rank-sum test; median, SD, and z value are reported.

*p < .1. **p <.05. ***p <.01.

These findings are unexpected given that those using at least one formal financing type simultaneously used informal financing compared with those who used only informal financing. Combining formal and informal financing mechanisms should translate into more housing units and sales revenue. The use of informal financing is not limited to a specific market segment. Of the developers supplying luxury homes (selling at over US$250,000 per unit), 43.5% used only informal financing. Similarly, 62.5% of developers who built and sold low-income homes at a maximum of US$50,000 per unit used only informal financing. A two-sample Wilcoxon rank-sum test showed no statistically significant difference in the use of informal financing in low-income housing (z = 1.31, p = .19).

3.5. Developers’ Adaptive Housing Finance Models

This section describes eight adaptive financing models identified among housing developers. The models aid in both the production and purchase of housing. Arguably, these models are entrepreneurial adaptations to the financial market conditions. Developers typically use multiple models simultaneously. The prevalence of unconventional financing mechanisms indicates the capital sources developers have created or identified to be TC-efficient. The spectrum of funding mechanisms is closer to the autarky end of the spectrum than the debt end, further providing evidence of equity-based adaptive financing mechanisms like those identified in the survey. They are efficient ways developers have identified or created to respond to the TCs of accessing debt and equity. This does not necessarily mean that they are also TCs efficient for homebuyers, because there are ex post contractual hazards, particularly to homebuyers. Williamson’s TC theory also illuminates that TCs and indirect bankruptcy costs play crucial roles in developers’ financial decisions.

3.5.1. Model One: Revolving Equity Financing

This model of housing financing was the most prominent among developers and involved incremental, small-scale developments financed by developers’ internal equity (e.g., owner-manager personal savings) and retained earnings. One project is completed and sold, and another is commenced (the former project does not necessarily need to be completely sold out before the subsequent project commences). These projects are typically gated communities with six to 120 housing units. However, gated communities do not necessarily have to be built (i.e., as master-planned and built communities). Serviced-plot gated communities are common in peri-urban areas of GAMA, wherein the developer acquires a tract of land, provides site and services, and does site planning. Individual plots are sold to mostly middle- and high-income households. Observed prices per plot (70 ft × 100 ft) ranged from GH₵50,000 (US$8,405) to GH₵357,000 (US$60,000). Hybrid gated communities combined both serviced plots and built master-planned communities. Developers used retained profits from one project to finance subsequent projects. The scale of development was thus directly linked to the developer’s equity and the profitability of preceding projects. Large-scale developments are impractical due to limited capital and debt avoidance. The survey observed revolving equity financing among mom-and-pop developers and more established developers. On why developers were unleveraged, as one developer succinctly put it, “The interest rates will kill you” (CEO, Housing Developer Accra 6, 2020).

Another version of revolving equity development financing was observed. Instead of gated communities, developers built a few housing units, sold them, and reinvested profits: “We build a few units one-unit, two-unit, three-unit, and four-unit bedroom detached and semidetached houses at a time. Depending on sales, we then build the ones that sold quickly. Currently, our four-bedroom unit is our fastest-selling product” (Marketing Manager, Housing Accra 5, 2020). This mechanism, while reducing the developer’s capability to profit from economies of scale, effectively reduces debt dependence.

Revolving equity is used for land acquisitions and development. Revolving equity avoids the cost of transacting capital across independent actors, which involves complex formal contracting, due diligence, and meeting collateral requirements. Reinvesting profits is straightforward. If a developer makes profits, the size of reinvestments depends solely on the developer. By so doing, they also avoid the interest on capital that banks charge. The nature of housing investments—illiquid assets in an illiquid market—makes equity attractive because the developer can concentrate on their investment interests and company goals without undue interactions with external financiers.

3.5.2. Model Two: Homebuyers’ Deposits

Clients’ deposits are often used to secure interest in proposed developments and help developers secure bank financing rather than actual development finance. It indicates to banks the project’s potential success, which factors into their credit decision-making. In this financing model, clients’ deposits also constituted the primary source of construction finance. The developers acted synchronously as the builders also. After acquiring the parcel of land from the land developer, construction is staged according to a payment plan agreed with the purchaser. Depending on the scale of development, the client typically had 12 to 24 months to complete payments before receiving keys and accepting the handover of the house. The process is comparable to off-plan sales. Nevertheless, instead of the client paying a deposit and taking a mortgage to pay for the home upon completion in the case of typical off-plan sales, as observed in Western property markets, the client’s incremental payments are used to construct the house.

Two submodels were observed: (a) the land is sold separately and can be paid for outright or in installments. Construction could be procured subsequently either from the developer or elsewhere. In either case, construction proceeds relative to payments from the buyer. Furthermore, construction is only permitted once the payments for the land are completed. (b) The second model is the land and house package: The developer sells a package that includes the land price and the construction cost. The development value is sold to the homebuyer, who pays incrementally for the land and housing construction.

Developers offer various payment plans for prospective homebuyers, and the installment plans vary. Upon signing the contract, the homebuyer pays an initial 50% of the house price, then 30% when the building is at roofing level, and a final 20% when construction is completed. Alternatively, the homebuyer pays 40% of the sale price upon signing the purchase contract, followed by 40% when the house is at the roofing stage and the last 20% when construction is completed. These two were standard options. Other developers offered the possibility to pay an initial 40%, with the remaining 60% payable within 12 months. Homebuyers could make payments in either U.S. dollars or Ghanaian cedis. The most extended payment plan observed was a 5-year interest-free payment plan. For this plan, the client must pay an initial deposit of 30% of the selling price and another 20% within eight months of the first deposit. Upon completion of the home, the buyer could possess the house, and the remaining 50% was payable within 48 months.

Below is an excerpt from a contract regarding deposits between a buyer and developer:

…where you elect to pay for the house by yourself on our standard payment plan, you agree to make an initial down payment of 40% or any such agreed percentage of the house’s value. A further 40% or an agreed amount shall be due at the roofing stage (payable within two weeks of notice that we have reached the roofing stage), and the remaining amount shall be due on or before the date of handing over. (percentages and stages of payments are italicized)

This strategic financing model eliminates or minimizes bank finance by the developer-builder, the builder, and the homebuyer. Although development can progress for several years, it is attractive as it significantly uses the buyers’ deposits for development. Depending on the repayment period, developers impose interest charges on plot sales. Land and house packages often did not. Instead, the value was priced in U.S. dollars, payable in Ghanaian cedi equivalents to account for inflation and currency fluctuations.

This model is risky to the developer, and ex post contractual hazards can be severe. Without financial institutions establishing the creditworthiness of clients, the developer self-appraises the prospective homebuyers and absorbs the risks of payment defaults, delays and the costs of monitoring payments. The contract is void if the buyer fails to honor the initial payment within such a time as specified in the contract (often one month). Although the developer can reallocate the land, delayed payments or defaults post the initial deposit are hazardous. The most significant ex post contract hazards relate to payment defaults and delays, which could severely constrain cash flow and lead to project cost variations and viability, and costly legal debt collection. Clients must adhere to contractual terms for the project to be profitable and to minimize the cost of contract monitoring and enforcement.

Interviewer: What happens when clients do not honor their scheduled repayment terms?

Respondent: Most of the time, once we move to the site, we will just complete the building without mobilizing and moving to the site repeatedly. This happens whether the payments are on schedule or not. Mobilizing and going to the site multiple times is more expensive. (Project Manager, Housing Developer 4, 2020)

In one case, the default risks involved in using homebuyers’ payments as construction finance caused the developer to cease building with clients’ deposits. Instead, they currently sell land and allow clients to procure their builder or self-build: “We have stopped this method because we ran into difficulties with repayments in the past. Now we sell the plots and allow the clients to build themselves” (CEO, Housing Developer, 3 2020).

Again, the benefit of using homebuyers’ deposits as development finance is that it saves the developer the interest and TCs from contracting debt in market transactions. Indeed, the developer, by so doing, also does not share equity contribution (and profits thereof) with external agencies. That notwithstanding, it is potentially hazardous to homebuyers (ex post hazards). Due to the nature of the contracts and the inefficiencies in Ghana’s legal system, the use of homebuyers’ deposits, while helping homebuyers minimize interest on mortgages and the associated TCs of securing one with a bank, is nevertheless risky to homebuyers. Failed projects risk locking up homebuyers’ funds. There are instances where developers have failed to raise enough capital to complete developments that sold off-plan, locking up deposits which were used as part of development financing. Developers could misappropriate funds, and due to the illiquidity and asset specificity of housing developments, homebuyers cannot easily retrieve let alone reallocate their committed funds.

3.5.3. Model Three: Developer Banking

A primary motive for this model is avoiding the cost of borrowing from financial institutions for housing acquisition. It is a demand-side solution by developers for homebuyers, like model two. The average percentage rate on mortgages, reflecting the actual cost of loans among banks, ranged from 18.7% to 31.7% in 2019 (Bank of Ghana, 2019). An additional one third of a house’s sale price will be charged interest at nominal values if acquired with a mortgage. This situation drastically worsens affordability and limits the developers’ market size. Accounting for the repayments over 20 years of the mortgage elevates the mortgage financing cost. For example, a 100% mortgage of GH₵140,000 with a 20-year payment period will accrue interest payments alone of over GH₵537,000 (383% of the nominal value of the mortgage).

In addition to using mortgages, developers typically allowed homebuyers to fully pay for the house outright (or with employee support) or in installments with repayment periods of up to two years, depending on the unit’s selling price. This way, developers adopt some banking functions, allowing homebuyers to acquire homes without mortgages and thus saving the interest on mortgages and other associated compulsory charges and commissions. Homebuyers save money by avoiding additional TCs such as credit report fees, information search costs for finding mortgages, mortgage origination/mortgage account establishment fees, appraisal fees, and application fees. Developers do not charge these fees, resulting in a more cost-effective home-buying process. The duration of the relationship between the developer and the homebuyer is also more elongated than in a typical homebuying situation involving third-party finance.

Taking on banking functions also helps homebuyers avoid the interest cost, making housing comparatively affordable to a more significant population. Payments are flexible, with homebuyers able to make balloon payments without penalties, unlike traditional mortgages. This practice aligns with the inconsistent incomes of people in the informal economy. Although repayment periods are mostly interest-free, defaulting or late payment often attracts charges/interest. One developer reported an interest rate of 14% per annum on any outstanding balance. This rate is still high compared to interest rates on US$ denominated mortgages. Houses are priced in U.S. dollars to eliminate forex change fluctuation to the Ghanaian cedi. The challenge is that the short duration can put enormous financial pressure on homebuyers.

This method of financing housing purchases was typical irrespective of developers’ characteristics. Every developer operated some form of “developer banking.” Developers’ additional risks for acting like banks included payment defaults, creditworthiness screening of homebuyers, deferred revenues, and debt retrievals. These risks comprise both ex ante and ex post hazards and would not be incurred should homebuyers use mortgages.

Developer banking introduces a new way of governing home acquisition transactions, moving from the typical trilateral relationship involving the developer, bank and homebuyer to a bilateral relation that excludes the bank. The developer vertically integrates some banking functions and manages the associated risks. The contract is mutually enforceable and retains an equity interest in the property until the total payment is made. If the buyer defaults, the developer can possess the property, resell, and reimburse the initial buyer’s equity interest. This possibility reduces the risks associated with the contract to the developer, as house prices tend to appreciate over time. It also eliminates the need for court decisions to repossess the property, as everything is handled in-house. Unlike car companies, there is no separate in-house finance department dealing with these financing solutions by developers.

3.5.4. Model Four: Sell Uncompleted Housing Units

Only one developer reported this behavior as part of their business model. Selling expandable housing units is not new in the Ghanaian housing market, especially housing targeted at low- and middle-income households and peri-urban areas. It allows families to add more bedrooms as household income increases. However, this method is like joint financing housing development or coproduction. In Kumasi, the developer reported the building to be about 80% complete before selling to a potential buyer(s). The developer then proceeded to complete the unit(s) using the additional finance received from the buyer. In this case, the developer had guaranteed buyers before further investments into the development and had external capital injected into the project.

A check on one of Ghana’s largest classifieds websites (date of search: February 02, 2023) found 436 uncompleted houses listed for sale in the Greater Accra Region. This finding suggests that selling uncompleted homes is unique to neither Kumasi nor this developer. It may be popular among small developers. One of the most expensive listed uncompleted houses had a listing price of GH₵8,000,000 (an apartment block).

Selling uncompleted homes allows developers to raise capital without taking on debt that would accrue interest to a third party. The developer sells their equity to raise capital to complete the project. This process also avoids dealing with lender requirements and the possibility of loan defaults. However, this approach still incurs TCs, such as marketing and sales expenses, to attract specific buyers interested in an uncompleted property. Indeed, the fact that this approach is unpopular among developers suggests that it may be more of a distress sale tactic than a genuine business strategy. Additionally, selling an uncompleted building can lead to lower profits if sold below market value. Despite its drawbacks, equity financing can reduce the overall costs of obtaining and managing debt financing.

3.5.5. Model Five: Land Sale Revenues for Developing Estates

Selling land to finance residential developments took different forms. This method was reported only in greenfield developments. The typical model proceeds as follows. The developer-builder acquires several acres of land in a peri-urban area or already has several acres of land banked decades ago. The land is zoned for phased developments into gated communities. Some parceled areas are sold as (serviced or unserviced) plots. Developers either retained the exclusive rights to build for the clients or did not. Other zones are developed into master-planned and built gated communities. Housing development and land sales can be either concurrent or sequential. With no landholding taxes currently, developers profit from capital growth by holding land stock and reselling it for a few years.

A developer acquires large tracts of land, of which a share is subdivided, serviced, and sold. As part of the contractual agreements for the land purchase, either the developer must act as the builder, or the buyer can either self-build or procure a builder. In cases where the lessor can procure a builder, the developer nevertheless imposes guidelines for the development, including the period within which the land must be developed, the types of housing developments permitted, and demolition. There were also conditions regarding the resale of the land by the lessor. Revenues from land sales are then used to complement housing developments. It is also a mechanism to attract residents to new sites in the peri-urban area.

Interviewer: How much are you selling those plots?

Interviewee: One is 50,000 Ghana cedis. I think about six plots are remaining. We are selling those to try and finance our new estate at [location]. Sales have been going well so far. (Marketing Manager, Housing Developer 2, 2020; purpose of land sale italicized)

Interviewees also spoke of other motivations for selling serviced plots. Instead of selling complete housing units, land sales offered clients the flexibility to build their dream homes customized to their tastes and preferences. This approach thus minimized customer complaints about the monotonous designs of developer-built estate homes. Finally, it should be mentioned that not all residential developers sell plots of land. About half of all developers (51.6%) sold plots of land.

3.5.6. Model Six: Land Retailing to Development

Land retailing—a combination of land banking, brokerage, and conveyancing—is a means to finance housing developments, although it is less common among developers. Technically, those involved were not developers because they had no developments during the fieldwork, although they had been operating for over 5 years. Land retailing was a means to an end, to becoming developers. It was an entry route through which individuals entered housing development due to minor capital requirements. During the field data collection, one respondent indicated that housing developments were to commence in 2020. Their vision was to evolve into a residential developer and builder.

Differences in this model were observed: (a) the company acquires large tracts of land, prepares a site plan, and sells individual plots. No servicing or improvements of the land are undertaken. The retailer may assist with acquiring the land title/deed at an additional fee. (b) The company makes a deposit for the land, prepares a site plan or land subdivision, and subsequently sells plots. Proceeds from the sales are then used to service the outstanding debt with the allodial or land freeholders. Payments to allodial or land freeholders were made in installments. Land may be registered in the company’s name and subleased to buyers or leased directly to tenants, with the retailer acting as a broker. These developers may also provide construction and architectural services or brokerage, and may manufacture and sell sandcrete blocks. Further interrogation with key informants suggested a prevalence of this method among small evolving developers in both Kumasi and Accra.

Land sale revenues for developing estates and land retailing to finance housing developments are similar through the lens of Williamsonian TC theory. In both cases, the developer sells part of their equity to raise capital rather than borrowing debt from a third-party lender. That notwithstanding, it is potentially more expensive than debt as the developer incurs TCs in selling the plots of land and the completed properties. Such costs include negotiating the sale terms with the buyers (land- and homebuyers) and ensuring the transactions are legally binding and enforceable. Future research should quantify the associated cost of this approach and compare it to using debt, mainly because the developer incurs similar TCs twice.

3.5.7. Model Seven: Production and Sale of Building Materials

Producing bricks, wood products, and sandcrete blocks, among other products, are not core housing development activities. To profit from economies of scale and specialization, developer-builders or builders procure such materials from material suppliers. Thirty percent of developers produced sandcrete blocks, which are the primary walling materials in the country. These developers had production plants that produced sandcrete blocks for in-house use and sale. Manufacturing was either in house or through hybrid organizational structures (i.e., subsidiaries and independent companies with the same shareholding as the developer). One developer supplied diverse precast concrete products to individuals, constructors, and developers. Two developers own polystyrene manufacturing factories. Most importantly, producing building materials is not necessarily a way to finance developments but an extension of the developer’s business.

3.5.8. Model Eight: Familial Capital

Some developers raised construction finance from spouses, family members, and friends. These developers could be termed mom-and-pop developers. As mom-and-pop developers, disincentives included not only the high interest on loans but also meeting the qualifying requirements of banks (collateral, balance sheets, credit history, and long-term credibility):

The documentation alone is hectic. The financial sector supports buying and selling. Something that when they put their money two days, three days [in the short term] they can get their money back. Something that will go through a process for six months to a year, they would not do […] Our local banks do not support us. (CEO, Housing Developer 1, 2020)

A resonating theme among these developers is that “bank loans can easily make you hot.” Serving bank loans with astronomical interest rates in an illiquid market was challenging.

Raising capital from family and friends can be an attractive option due to its relatively low TCs compared to other forms of debt financing. The information search and monitoring costs are typically minimal, given the preexisting relationships. Furthermore, contract enforcement is often based on trust and social connections rather than third-party enforcement. However, there are potential down sides, such as the risk of damaging personal relationships if the business venture fails to generate adequate profits. Additionally, borrowing from family and friends may not be optimal for their capital, as they may have otherwise invested their funds in more liquid and profitable assets.

4. Discussion and Implications of Findings

Our study sought to understand residential developers’ financing of housing in Ghana and their adaptations to the unconducive finance market. There is a mature literature on financing adaptations by households and landlords in developing countries and some of these have led to innovative microfinancing schemes, revolving funds, and cooperatives to support housing. This paper, using the case of Ghana, shows that residential developers in the same markets, like those operating in the informal economy, have adapted to the rogue capital market conditions. The market is not taken as given. Rather, entrepreneurial creativity is at play to enable them to coexist with the hustle capital market.

The cost of capital determines developers’ capital decisions. Such cost includes not only the interest rates but also the cost of exchanging funding. Australian real estate investment trusts (REITs) increased debt financing from 10% in 1995 to 35% in 2007 as falling interest rates made debt comparatively cheaper (Newell, 2008). REITs have repeatedly maintained high leverage ratios (Feng et al., 2007; Ott et al., 2005). Le and Ooi (2012) conclude that the maturity of the debt capital market significantly, positively influenced the capital decisions of property companies. Our findings suggest that TCs of debt are more expensive than those of equity. The prevalence of nonstandard financing mechanisms is indicative of the TC-minimizing structures of developers. Instead of direct computation of TCs, the prevalence of these mechanisms is evidence of what developers have identified or created to effectively minimize the associated TCs and hazards of exchanging funding. These funding mechanisms may not be the most efficient, but their adoption is a comparative decision. Rationally, developers would not take on extra risks and uncertainties in creating funding schemes if bank loans were comparatively cheaper.

The frequently used financing mechanisms in financing housing are retained profits, client deposits, and private funds. Bank loans, loans from mortgage companies, and loans from rural banks are at the bottom of the list. In addition to these funding schemes, developers have created other funding channels to finance housing. These models facilitate both supply and housing consumption. On the housing supply side are revolving equity financing, land retailing, homebuyers’ deposits, land sales to finance developments, building materials sales, and familial capital. Methods facilitating housing consumption are developer banking, homebuyers’ deposits, and the sale of uncompleted buildings, which lowers the entry barrier for buyers.

We see these financing mechanisms as informal because they are nonmarketed and rely on reputation and relationships to monitor and enforce contracts rather than formal contractual obligations enforced through a codified legal system (Ayyagari et al., 2008). Admittedly, equity financing is nonmarketed but not entirely an informal approach. However, equity could come from the private funds of the entrepreneur, which is observed as private equity or underground lending. Allen et al.’s (2005) definition of informal financing includes internal finance, capital raised from family and friends of the founders and managers, and funds raised in the form of private equity and loans, which accounts for 60% of total funds of private sector firms operating in China’s poor legal environments.

The identified financing mechanisms are prevalent among developers, indicative of the TC-efficient mechanisms used to coordinate financing for housing development and consumption in Accra. Their widespread use indicates their cost efficiency relative to formal bank financing and mortgages. They eliminate the expensive interest rates and are flexible, accommodating consumer conditions and market characteristics. From a TCE perspective, housing financing is preferentially coordinated either in-house or via hybrid governance structures instead of markets. Developers are not quantitatively evaluating actual TCs but developing governance mechanisms to secure their interests.

The market in which developers operate is mainly informal, where economic and social interactions are structured by informal constraints such as taboos, traditions, and customs instead of legal contracts. The factors of housing production, labor, land, and building materials used by GREDA involve significant elements of informality. Land transactions are in the customary informal market (Gough & Yankson, 2000; Kasanga & Kotey, 2001; Quaye, 2014); subcontractors, tradespersons, and laborers are drawn from the informal and semi-informal market (Darko & Löwe, 2016). This situation makes it reasonable that their financing innovations will resemble their existing practices of coordinating other factors of production.

This study did not estimate the volume of funds generated from these sources, including bank debt. They may play a complementary role to formal sources. Nevertheless, their prevalence makes this unlikely for most developers. Indeed, of the developers who never use or rarely use bank loans, they build, on average, 53 housing units per year, which is greater than the average of the sample. Furthermore, the literature often sees informal financing as servicing the needs of the lower end of the market. Our findings suggest otherwise. Future studies should examine the relationship between the financing of housing supply and developers’ productivity and growth.

Arguably, these sources of finance for housing projects are less well known in the existing literature. The preference for equity and retained profits over debt contradicts traditional approaches to housing project financing in the so-called Global North. For example, housing developers collect deposits from homebuyers as security for off-plan sales in developed financial markets. The deposits are usually held in a trust account until settlement (Consumer Affairs Victoria, 2021). Buyers subsequently service a mortgage after the home purchase. Rowley et al. (2014) argue that debt financing (senior and mezzanine debt) is the traditional source of finance for housing developments in Australia and is taken as a given because it is inevitable. However, the findings of this study establish that housing developers in Ghana use homebuyers’ deposits as a source of construction financing. Ghana’s market innovations are likely replicated in similar market conditions.

Based on the UN-Habitat definition of informal housing,² the consumption of formally produced housing involves the use of informal financing and thus meets the criteria of informal housing. Indeed, based on the financing mechanisms observed, informal and formal housing mechanisms are not duopolies. They are delivered using formal and informal rules of the game—institutional hybrids. According to the structuralist school of thought on the informal economy, some formal capitalist firms engage informal enterprises to reduce costs in their production and supply chain.

A cross-cutting feature of the financing innovations is the active role of the consumer in the process of housing delivery. Housing is not an already made product bought off the shelf; it is coproduced with the homebuyer. For example, in using the buyer’s deposit as construction finance, the homebuyer becomes an active participant in the housing development process. The person(s) can actively monitor construction progress, make suggestions, and provide more funds as their capacity permits. Such a process is also flexible to meet the peculiar financial capacity of the homebuyer. Developers accept balloon payments, which banks do not tolerate, because homebuyers’ income flows can be inconsistent. This solution by developers aligns with the literature on the coproduction of urban spaces and participatory urbanism (Ardill & Lemes de Oliveira, 2018; Day, 2003; Finn, 2014; Fredericks et al., 2018).

Although the exact developers’ models discussed here may not be replicable and do not support volume housing development, they show that workable financing models can be derived based on coproduction principles. We see opportunities in this light leveraging provident funds and special-purpose vehicles channeling remittances into housing delivery and consumption. Systematic structures could facilitate joint ventures allowing landowners to contribute their land as equity for residential development in return for a share of the developed units or an appropriate return. We further see evidence of the possibility to supply homes with shells only, an adaptation of selling uncompleted houses. There is also an indication that banks may exaggerate the perceived risks of lending to the housing sector. Some of the risks that banks avoid are what developers incur to create viable businesses. Mortgage products should realign with the cultural context of the people who prefer to pay short-duration loans. Developers give homebuyers 2–5 years to complete payments for their houses, minimizing or eliminating the exorbitant interest rates that characterize traditional mortgage repayments.

The lessons that can be learned from developers’ adaptive behavior lie with this evidence of viable coproduction. Instead of attempting to replicate Western finance models, which have so far yielded negligible results (Arku, 2009, 2010; Sarfoh, 2010; Sarfoh et al., 2016), our evidence suggests rethinking these strategies. Policymakers and researchers should pay attention to the market’s character in developing pragmatic policies. An emerging strand of literature argues that informality offers potential solutions to the wicked problems urban areas face, calling for bottom-up approaches (Devlin, 2018; Gurran et al., 2022; Heisel & Woldeyessus, 2016; Watson, 2009a, 2009b). For example, Berner (2000) argues that many developing countries can significantly address poor conditions and deficits by incorporating informal-sector strategies of housing supply into public policies.

Informal lending attracts high interest rates compared to traditional financing, making it more expensive. Nevertheless, the developers’ creativity points to solutions that do not bear (direct) high costs. Developers may hedge against inflation and foreign exchange volatilities by pricing in U.S. dollars, but we observed no evidence of them applying similar interest rates to those of banks. Arguably, this finding suggests that developers may be providing housing that would have otherwise been unaffordable if they took out bank loans to build and the homes were bought with conventional mortgages. Further research could test the extent to which these financing strategies contribute to the affordability of formal-sector housing.

The likelihood of opportunistic behavior persists, particularly in informal contracting. This is heightened in many developing countries where regulatory lapses are typical. The opportunistic behaviors on the part of developers arise from the nature of the financing models discussed. In the revolving equity model, without completing some units, a developer cannot have the capacity to initiate new projects. However, to secure excess business for execution later, a developer can deliberately accept to develop several units beyond capacity at a point in time. The effect is a deliberate delay in executing those extra units to the disadvantage of clients because the developer must complete ongoing projects. Furthermore, opportunistic behaviors arise from the developer-banking model of financing due to the imposition of short-term repayment durations on homebuyers. This is because homebuyers do end up forfeiting projects entirely due to their inability to honor those contractual payments, and contractors often pay back the exact amount paid by defaulting homebuyers without accounting for price appreciation in the value of the specific projects, and the time value of money, for their private benefit rather than on the specific projects. For example, Kimaru (2018) reports that a company disappeared with homebuyers’ deposits from off-plan sales in Kenya. Despite these schemes occasionally turning into scams, they remain a popular source of financing for housing development in Kenya (Kimaru, 2018).

Policy solutions should address issues from the broader perspective of TCs, market transparency, and clear and secure property rights that protect homebuyers. On policy, it is suggested that governments of countries facing similar issues should enact legislation to regulate the sector, particularly the use of homebuyers’ deposits for business operations. For example, legislation could be passed to establish the use of escrow accounts to manage the deposits of homebuyers. The effect is to thwart the opportunistic tendencies of unscrupulous developers to ensure that homebuyers can retrieve their deposits when developers fail to raise enough capital to complete advertised projects.

Furthermore, it will be prudent for policymakers to enact legislation to fine-tune these other pervasive hierarchical sources of capital in the industry as they have become established sources of capital for development projects in many developing countries. An important insight from the study is the co-ownership of property, where developers retain an equity interest in the property until payment is complete: the developer-banking model. Governments could establish co-ownership schemes where they pay either the deposit or the interest on the mortgage for an equity interest in the property, payable if the property is ever sold. This scheme should apply only to owner-occupied properties. By so doing, low-income households are assisted to own homes without necessarily burdening (local) governments, which will benefit from the increased property values. A shared equity scheme also prevents the imposition of short-term payment durations on homebuyers by developers. In cases of default, instead of homebuyers forfeiting projects entirely due to their inability to pay, the government can renegotiate favorable payment terms with homebuyers, leading to a win–win situation for all parties involved. Shared equity could also be interesting to insurance companies, and pension funds.

Many of the capital sources identified in this research allow developers time to sell either uncompleted units or land to finance developments. Consequently, development projects can be delayed unnecessarily, to the detriment of homebuyers who may have committed to projects by paying deposits. To this end, it is suggested that policymakers develop a cooling-off period or a mutually beneficial timeline for project commencement, after which homebuyers should have the option of either continuing or pulling out of the purchase arrangement at no cost to the homebuyer in the case of undue delay caused by a developer. This will extinguish the potential for developers to take undue advantage of lax regulations to delay projects deliberately for selfish gains. The results of this paper, including the discussion and policy implications, will be useful for many countries that are situated similarly to Ghana.

5. Conclusion

The most crucial resource in any housing development project is financing. With appropriate financing, housing development and accessibility should be widespread. Contrary to the (literature about the) Western world, where debt is taken as given, in Africa—and in Ghana, to be specific—despite debt funding sources, it is inaccessible due to the underdeveloped financial market. Due to the onerous conditions attached to debt funding for housing developments in Ghana and associated TCs, developers have evaded, altered, or exited specific institutional mechanisms to sustain their business. The results indicate that Ghanaian housing developers have innovatively developed their sources of capital to execute housing development projects.

In this paper, eight models of capital sources were identified and explored. These models show developers are nonpassive actors in the market. They are entrepreneurially creative in developing alternative governance mechanisms to transact capital. Although some of the identified models resemble models used by residential developers and discussed in the existing literature, they differ in their implementation in Ghana, making them an original contribution to the literature on sources of financing housing development in the Global South. For example, although homebuyers’ deposits are used by housing developers in the West, as a form of commitment on the part of a buyer, in Ghana, these deposits are used as part of the available funds to execute a project. But this practice presents contractual hazards to homebuyers and to developers alike in cases where payment schedules are not adhered to by homebuyers.

The results also have implications for policy and legislation to regulate the use of homebuyers’ deposits and other models. On the policy front, it is suggested that policymakers develop policies that incorporate broader TCs associated with housing development financing and consumption, including some of the opportunistic tendencies that characterize the use of specific financing models identified in this paper. We argue that informality is the norm and that policy should actively accommodate it. As our findings show, developers are engaging in profitable business in the informal economy with informal solutions, whereas policy has generally been hostile to informality. The question of how to create pragmatic financing mechanisms for housing is answered by developers who come up with bottom-up approaches to financing. These provide insights that coproduction and co-ownership are suitable attributes of workable financing mechanisms in a largely informal economy.

The models identified could serve as lessons for developers. We anticipate comparative studies that examine the TCs of these models to guide developers’ financing behavior. Finally, the study is exploratory and provides a basis for future in-depth analysis of the financing mechanisms identified and their feasibility to inform policymaking. Other studies should attempt a larger sample size. How can housing financing policy minimize associated risks and improve the probability of success?

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Notes on contributors

Godwin Kavaarpuo

Godwin Kavaarpuo is a research fellow at the Unison Housing Research Lab, RMIT University. His research interests and publications are in sustainable property development and investments; property technology and blockchain applications in property, and affordable and social housing.

Kwabena Mintah

Kwabena Mintah is a Senior Lecturer in the School of Property, Construction and Project Management of RMIT University. He teaches many subjects including residential property valuation, property and capital markets and research methods for built environment. Mintah's research focuses on home ownership dynamics, housing economics and policy, land tenure and proptech applications and commercial real estate investments.

Kenneth Appiah Donkor-Hyiaman

Kenneth Appiah Donkor-Hyiaman is a University Lecturer in Real Estate Finance and Entrepreneurial Management at the Department of Land Economy, Kwame Nkrumah University of Science and Technology in Ghana. He researches institutions and mortgage finance development, innovative housing finance, sustainability, and entrepreneurial finance.

Notes

1 See a summary of the controversial Saglemi housing project at citinewsroom.com.

2 A house is considered informal if it has one or more of the following characteristics: is located at the urban periphery or within the interstices of the formal city, is self-designed and self-built with local materials, does not fully conform to building and land-use standards, is poorly serviced by network infrastructure and public services, was financed out of family/group savings or loans from informal lenders; and/or is incrementally improved by the occupant over a long period of time (Parby et al., 2015).