Real estate financialisation and the production of ‘investable’ spaces in Johannesburg

Abstract

This article contributes to an understanding of how real estate financialisation unfolds within a city in the global South, Johannesburg. It firstly shows that compared to what has been witnessed in the global North, real estate financialisation in South Africa is characteristically conservative: market actors do not engage in high-risk, low-income neighbourhoods. Indeed, in Johannesburg, financialisation has displayed minimal interest in residential markets. Instead, financialisation in the city has been driven by commercial property investments, typically in the form of urban enclaves such as securitised office parks, malls, mixed-use developments, and increasingly ‘satellite cities’ built from scratch. This article considers how the selective character of financialisation in the country is mediated by power relations in the global political economy. In particular, it highlights how South Africa’s status as an ‘emerging market’ encourages a prudent lending regime for domestic financial institutions and real estate actors.

Introduction

The 2008 global financial crisis (GFC) saw an explosion in research exploring the intersection of finance and urban landscapes in areas such as housing, infrastructure, land use planning, and governance (Aalbers, 2016). This research reveals how burgeoning sections of the city are being transformed from public goods into private financial assets. Toll roads, water networks, schools, hospitals, rental housing, and electricity grids from a wide range of countries are incorporated into investment portfolios and often traded on stock exchanges (Torrance, 2008: 2). The concept of financialisation has been extensively mobilised to make sense of the growing interdependency between the built environment and financial markets. Although this has resulted in a wide range of meanings, financialisation broadly entails:

the increasing dominance of financial actors, markets, practices, measurements, and narratives, at various scales, resulting in a structural transformation of economies, firms (including financial institutions), states and households. (Aalbers, 2016: 2)

The state plays a significant role in these processes. It is essential not only for establishing the regulatory, institutional, and other conditions through which financialisation transpires, but also relies on financial markets to finance development needs, maintain its functionality, and fulfil its social, political and economic duties.

Financialisation is therefore a complex phenomenon that challenges conventional understandings of state-market relationships, revealing the contradictions and tensions within contemporary capitalism (Aalbers, 2017). Furthermore, considering the varied contexts in which financialisation takes place, it does not follow a uniform, one-size-fits-all across different spaces. Instead, it is inherently fragmented, path-dependent, and variegated (Aalbers, 2017: 4). However, as research on financialisation expands, the concept is increasingly at risk of becoming a ‘catch all’ phrase, akin to ‘neoliberalisation’ and ‘globalisation’, providing overarching explanations for varied phenomena. In these accounts, the financial market is often treated as an ‘object of critique and resistance, rather than one to study’ (Berndt and Boeckler, 2012: 203).

A burgeoning field of research broadly associated with science and technology studies (STS), has taken up the challenge to examine markets in-depth, highlighting concerted effort, supporting infrastructures, discourses, and practices that allow financialising projects to be realised (Berndt and Boeckler, 2009). Although this article is inspired by these arguments, a practice-oriented approach risks an inward focus, evading questions regarding the socio-spatial and political effects of markets (Fields, 2017a: 3). To this end, a growing number of scholars have been deploying a heterodox approach to the study of financialisation to bridge the concerns of critical political economy and STS approaches (Butcher, 2016; Fields, 2017a; Ouma, 2016). As Fields (2017a: 3) this approach enables a focus on ‘the political-economic import of market formation as well as the material, provisional, and contingent aspects of financialisation’.

This article adopts a heterodox approach to examine the financialisation of real estate (FoRE) in Johannesburg, the economic hub of South Africa. Financialisation turns immovable assets such as buildings, houses, and infrastructure into tradeable commodities on global capital markets. By focusing on Johannesburg, the article aims to shed light on how the financialisation of real estate reshapes urban spaces in cities of the global South. Although initially overlooked in the immediate aftermath of the 2008 GFC, FoRE in the global South has been receiving increasing attention in recent years. Housing, particularly, has been a major focus in this literature, with financialisation expanding through the promotion of private homeownership and mortgage finance (Migozzi, 2020). While this article draws insights from the literature on housing financialisation, it broadens its scope to include all types of real estate. This wider perspective aims to better understand how the interplay among residential commercial and mixed-use markets transforms the urban landscape.

The interplay between FoRE and urban transformation in South Africa has been largely overlooked. Existing literature on financialisation in South Africa has remained national in scope, focusing on the economy in general, and the mining sector specifically (Karwowski, 2018). This is despite an extensive database documenting the historical evolution of housing policies, urban planning, housing finance and the interplay between real estate capital and spatial segregation (Charlton, 2013; Landman and Badenhorst, 2014; Murray, 2011, 2013; Swilling, 1990). Furthermore, the few studies that do explore the dynamics between the real estate market and the production of urban space in South Africa fail to consider how the local characteristics of FoRE are shaped by transnational processes, actors, and power relations.

This article seeks to address this gap by exploring how the spatial manifestation of FoRE in Johannesburg is influenced by a complex interplay of local factors, historical remnants, and transnational power dynamics. By combining perspectives from critical political economy and poststructuralism, it analyses historical discourses and geographical imaginaries that shape Johannesburg’s real estate markets, and the resultant urban fragmentations and spatial inequalities. The article’s discursive analysis looks at how financial institutions such as banks and real estate companies perceive and assess credit risk in relation to space and individuals. Primary documents such as annual reports, risk measuring methodologies, and real estate portfolios provide insight into these entities’ investment decisions on both local and global scales. To understand how these processes interact with urban space production and inequalities, this article leverages existing literature on housing inequalities, mortgage finance policies, real estate capitalism and urban planning in Johannesburg and South Africa at large (Butcher, 2016, 2020; Harrison and Zack, 2014; Mabin, 2014; Migozzi, 2020; Murray, 2011, 2013).

The article is structured as follows. The first section briefly describes the theoretical framework that guides the analysis. The second section examines the selective and risk-averse character of real estate financialisation in South Africa, focusing in particular on the mortgage market. The third section looks at how FoRE unfolds in Johannesburg and the spatial configurations produced by property markets. The final section considers how the character of FoRE in South Africa is mediated by power relations, discourses, and rationalities on a global scale.

Discursivity and the urban geographies of real estate finance

This article aims to address the call on researchers to “open up the black boxes” of financial markets to understand how they are ‘kept opaque, how they structure their “contexts,” and how those contexts are inscribed within them’ (MacKenzie, 2005: 554). The underlying assumption here is that finance, markets, and the economy are not detached entities, but are interwoven within social, political, and cultural relations. Markets do not simply appear ‘out of thin air but are continually produced and constructed socially with the help of actors who are interlinked in dense and extensive webs of social relations’ (Berndt and Boeckler, 2009: 536). In this sense, financialisation can be considered a practical accomplishment, produced through a network of actors, technologies, processes, and knowledge.

Drawing on the concept of performativity from Callon (2007), this approach to financialisation contends that financial and economic discourses including theories, models, and measurements – do not describe reality, but constitute it. In other words, the financial modelling of markets, such as Johannesburg’s real estate market, is not merely a tool used to analyse and predict the performance of that market; it helps define the market itself. Central to the understanding that financial markets are performed into being is an investigation into the historically grounded and power-laden discourses that can shape financial knowledge and practices such as calculation, risk modelling, and price formation. Discourses are systems of meaning that constitute institutions, practices, and identities in contradictory and disjunctive ways (Fairclough, as cited in Larner, 2006: 205). By attaching meaning to social realities, discourses make intelligible ways of being in and acting towards the world. As they attach meaning to social realities, discourses govern modes of conduct that bring financial markets into existence. This means that risk and valuation metrics, expert advice and media commentaries on property markets do not exist in addition to ‘real’ material financial and economic structures. They are precisely how financial markets materialise (de Goede, 2005: 7). To illustrate, consider the expropriation without compensation (EWC) bill adopted by South Africa’s ruling party, African National Congress (ANC) in 2017, to accelerate land redistribution (Silungwe and Fourie, 2018). Shortly after the announcement of EWC, the rating agency Standard & Poor’s, alongside institutions like the International Monetary Fund (IMF) and International Finance Corporation (IFC) released a report highlighting that the adoption of EWC would create significant uncertainty around property rights and deter fixed investment, lower government revenue, and subdue Gross Domestic Product (GDP) growth (Silungwe and Fourie, 2018). Fearing that the returns on their investments might be compromised, investors have abstained substantially from investing in real estate over the past five years, despite government reassurances that foreign investors have ‘nothing to fear’ (News24, 2019). Although real estate growth has been subdued since 2014 and the global run on emerging markets, this decline accelerated after 2018, in both direct property investments and the listed sector (SimplyWallSt, 2022). Without suggesting that investor concerns are entirely unfounded, this example illustrates how market commentaries do not merely reflect on phenomena, but also materialise within a network of meanings, practices and institutions.

This example brings into focus two relevant points of discussion. First, it is not the case that any market commentary or expert forecast performs that depiction into being. The ability of S&P, the IMF, and the IFC to perform certain representations into being underscores the power relations within which discourses are formed. From a Foucauldian perspective, discourses refer to bodies of knowledge that not only define the subjects authorised to speak and to act but also the knowledgeable practices that these subjects undertake. The ability of S&P and the IMF to speak about the financial and economic realities of states, and for these discourses to be accepted as ‘true’ or ‘scientific’ claims, reveals the inextricable link between power and knowledge. Here, the question of power should not focus on why actors say certain things, but rather how discursive systems enable them to act in a particular way and influence the decision-making of others.

However, discourses and the meanings and the subject-positions they create are not neutral. They are tied to historical modalities of truth production that entail, at some point or another, the exclusion of ‘subjected knowledge’. Here ‘truth’ should not be seen as something to be revealed, but instead, the unconscious structures that determine what gets considered as ‘the truth’ (Foucault, 1978: 132). The ability of a subject’s discourse to be viewed as ‘true’, for instance, S&P’s rendering of South Africa’s land reform bill, reveals the functioning of discourse in relation to power. In financial markets, geography plays a formative role in linking regimes of truth to spaces of power. The discursive authority of financial centres like Wall Street and the City of London has been mediated by a long history of colonialism and imperialism (Schultz, 2021). This history refers not only to the practices of economic, social, and military domination and subjugation, but is intricately linked to how Western discourses of reason, progress, and ‘scientific rationality’ have been normalised as the only way to know, interpret, and be in the world. Today, this history has been projected into ‘spaces of hegemony’, whereby the financial centres of London, New York, and Frankfurt enjoy significant authority in not only determining what counts as ‘information’, but also what this information means and how it should be acted upon by market participants (de Goede, 2005: 7).

Second, based on the discursivity of financial practices, poststructuralism problematises the ideal/material dichotomy and enables us to consider how representations of risk and reward in property markets engage the production of space. Markets are produced in, and productive of, space and place and, importantly, imaginaries of those places (Butcher, 2020: 175). These imaginaries are circulated in media reports, political discourse, and everyday conversations that forge representations about what certain places are like in relation to others. Financial markets act upon these spatial representations through sophisticated modelling technologies that calculate the risks and rewards associated with investments in a certain place. This may lead to a withdrawal or influx of capital into an urban area that significantly influences the property values and the wealth of residents and tenants in a particular neighbourhood. As such, it is not only that private investment shapes cities, but social ideas and geographical imaginations shape private investment (Jacobs, 1961: 313). It is at this point where poststructuralism and critical political economy approaches especially complement each other. On the one hand, poststructuralism allows us to explore the power relations inscribed in the financial discourses that perform real estate markets into being, such as creditworthiness, risk and reward. On the other hand, critical political economy perspectives consider how these discourses, which are inherently political and historically contingent, interact with processes of spatial production in post-apartheid Johannesburg. The following sections consider how discourses of risk and creditworthiness have shaped FoRE in South Africa, followed by an investigation into how FoRE has shaped and transformed the urban landscape of Johannesburg.

The selective character of real estate financialisation in South Africa

In the period leading up to the 2008 GFC South Africa, like many nations, experienced a ‘property boom’. This period saw escalating house prices, a surge in mortgage lending, a flourishing construction industry and significant innovations in mortgage finance (O’Neill, 2012: paragraph 33). In 2001, the first residential mortgage-backed security (RMBS) was established by SA Home Loans (SAHL) with JP Morgan equity, following an amendment in securities legislation by the South African Reserve Bank (SARB) that allowed non-banks to perform banking functions and play multiple roles in a security transaction (van Vuuren, 2004). In 2004, several property loan stock (PLS) companies set up single-borrower commercial mortgage-backed securitisation (CMBS) programmes (Freybote and Karoly, 2008: 197). Although mortgage-backed securitisation remains a small, yet expanding industry, it reached levels of innovation and sophistication in a few years ‘that took the US and European markets nearly two decades to achieve’ (Freybote and Karoly, 2008: 188). The booming property market coincided with pressures from the post-apartheid government to deracialise and democratise finance, particularly the property market, which resulted in the Financial Sector Charter (FSC) of 2004. Although housing finance was only a small part of the FSC’s terrain of ‘transformation targets’, signs of a US-style subprime story were starting to show, with higher interest rates and mortgages bundled into securities (Butcher, 2016: 85).

Yet, contrary to expectations, the South African real estate boom did not result in a subprime mortgage crisis. Instead, the prime mortgage market – higher value properties in higher-income areas – saw the most substantial decrease in value (Butcher, 2016: 103). Contrary to the overinclusion of mortgage loans to low-income households and neighbourhoods in the global North through predatory loans, analysts noted a decreasing share of mortgage finance to low-income households and neighbourhoods (Melzer, 2010: 6).

Mortgage lending and real estate investment in South Africa follow a risk-averse approach, characterised by an extreme selectivity in terms of location, sector developments, and borrower/tenant profiles (Butcher, 2016; Migozzi, 2020). Post-crisis analyses revealed a distinct preference particularly in mortgage lending, for high-income borrowers with permanent employment in higher value properties and more expensive, historically white, neighbourhoods (Butcher, 2020; Haferburg and Huchzermeyer, 2017; Migozzi, 2020). Nationwide, only around 36% of mortgage applications are approved, with a large majority (87.87%) valued over R700 000 (NCR, 2022: 7), i.e., properties located in the ‘conventional’ rather than the ‘affordable’ and/or ‘government’ market. The mortgages underpinning RMBSs are similarly conservatively selected, characterised by low loan-to-value (LTV) and low payment-to-income (PTI) ratios (Freybote and Karoly, 2008; Migozzi, 2020). Although the lack of georeferenced, disaggregated and accessible data on the mortgages underpinning RMBSs make it difficult to trace how they translate into urban space, available evidence reveals that they have a clear metropolitan focus (Migozzi, 2020: 647). For the various RMBSs grouped under the Amber House Fund, backed by SAHL, Gauteng (Johannesburg) represents 46.56% of the value of distributed mortgages, followed by Western Cape (Cape Town) at 16.2% and Kwa-Zulu Natal (Durban) at 12.92% (SASF, Amber House Fund, 2022a). These RMBS deals do not include subprime mortgages but are instead composed of issuances that have an average LTV ratio of around 70% and a PTI ratio of 20% or less (see SASF 2022b: 2020). Securitisation, therefore, operates by selecting the most privileged sections of society who are able to put up a large down payment for a home loan, with larger salaries that allow for low PTI ratios (Migozzi, 2020: 647).

Likewise, commercial property developments, particularly in the office, retail and entertainment sectors typically prioritise wealthy urban nodes (Murray, 2011: 187). The geographical composition of Real Estate Investment Trusts’ (REITs) investment portfolios reveals not only an overwhelming preference toward commercial properties (Nurick, Boyle, Allen, Morris, and Potgieter, 2018) but also established wealthy nodes in urban areas, a ‘risk-averse strategy that favours caution over experimentation’ (Murray, 2011: 187). Additionally, private real estate developers often partner with public authorities on megaprojects aimed at elevating a city’s ‘world-class’ status. These projects include signature buildings, prestigious headquarters, mixed-use developments, and cutting-edge infrastructural services like telecommunication technologies and transport networks. Although some of these megaprojects include mixed-use developments in low-income areas, such as the N2 Gateway Project in Cape Town, the substantial involvement required by the private sector means that private developers are often able to dictate their target market. To mitigate the potential risks associated with undertaking megaproject developments, private real estate actors are typically very strategic and meticulous in choosing where developments should be located and how to yield the best possible returns. As such, their targeted location and population market often involves growing middle- and upper-income groups and areas.

Housing provision and segregation in apartheid South Africa

The conservative approach to real estate investments in South Africa is rooted in a long history where housing provision, city building, and urban planning constituted the main mechanisms through which colonial and apartheid government implemented their racial segregation policies. This entailed the creation of separate locations for different racial groupings and limiting the land and property rights of non-white populations. These areas were characterised by stark contracts in terms of living conditions, green belts, building densities and access to infrastructure and municipal services to the benefit of white neighbourhoods and populations. Additionally, several thriving mutual building societies, benefitting from various regulatory and tax exemptions, were able to offer mortgages at lower-than-market interest rates to the white population especially (Luus, 2003: 152). Benefitting from property rights, privileged access to employment and financial services, and a growing economy, white ‘standards of living’ in South Africa became consolidated with ‘European standards of space’ (Butcher, 2016: 53).

Although some building societies offered their services to the non-white population, up to the late 1970s, Coloured, Black, and Indian households faced severe racial discrimination when trying to access housing finance, while Black people were denied property rights in urban areas until 1986 (Migozzi, 2020). Around this time, the then National Party (NP) government launched a range of policies to facilitate the privatisation of housing, which entailed the selling of 500,000 state-owned housing units – the majority of which were in Black townships – and the demutualisation of building societies (Marais, Sefika, Ntema, Venter, and Cloete, 2014). The demutualisation of building societies meant that they become, like banks, listed, equity-funded institutions that compete for funding on capital markets with a ‘profit motive’ (Luus, 2003: 152). Thus, at a time when Black households were finally allowed to enter the property market, they were unable to benefit from the same lending privileges that white households enjoyed for generations before them. Moreover, several legacies associated with socio-spatial and racially segregated urban planning undermined the integration of historical township areas and households into real estate financial markets.

First, as mortgage lending became financialised, criteria became increasingly caught up in risk-reward calculations, which severely restricted mortgage lending and other forms of real estate investment in township areas. Many banks and building societies refused to offer loans of less than R35 000, which excluded around 90% of those needing homes, as they were deemed less profitable and introduced higher risks, even though the correlation between low-income borrowers and higher default probably has been disputed (Swilling, 1990: 19). By the late 1990s, the ‘township market’ represented only a fraction of banks’ mortgage lending portfolio, a mere 8% (Tomlinson, 1998).

Second, banks started redlining many township areas during the 1990s, as racialised notions of creditworthiness became increasingly entangled with space (Butcher, 2016). During this time, several consumers in township areas started boycotting their bank payments for several reasons, primarily the poor workmanship of bonded houses (Bond, 2000: 132). These boycotts took place at a time of intense anti-apartheid struggles, with efforts to make towns and cities ungovernable being a key strategy in this regard. Banks perceived the non-payment of bonds and instances of intimidation against repossession (Parliamentary Monitoring Group, 2000: Banking Council Submission, paragraph 3) as a part of this political struggle, and used it to justify the introduction of redlining (Haferburg and Huchzermeyer, 2017: 64). Township areas became articulated as areas where ‘common protest’ around the issue of property is high, coupled with illegal occupation, and where evictions are impossible (Parliamentary Monitoring Group, 2000). Instead of viewing these events as short-term ‘non-commercial’ risks, produced by political instability, financial institutions began referring to this as a ‘culture of non-payment’ that is endemic to black townships (Butcher, 2016: 77).

Although these criteria are typically presented as apolitical, value-neutral algorithms that link decisions to outcomes in conditions of uncertainty, the calculation of creditworthiness, risk, and reward is inherently political. Poststructural enquiries into the history of credit emphasise the sociality of creditworthiness, which is an inherently subjective, political judgement (de Goede, 2005). Creditworthiness refers to the probability that a borrower will repay its debt obligations at some time in the future. In other words, it is an assessment of how ‘worthy’ or deserving a borrower is of credit (CFI, 2023). Historically, creditworthiness exhibited the attributes of the colonial gentleman, which was perceived as being honest, speaking the truth, and conducting himself in a rational, disciplined manner (Schultz, 2021). Women, the working class, and non-European populations are subsequently viewed as ‘irrational’, their decision-making based on lust, greed, and emotions (Schultz, 2021). Indeed, during apartheid, white males were the major beneficiaries of mortgage finance and other forms of credit, as these perceptions informed not only financial actors’ decision-making but also banking legislation (Butcher, 2016). Although these overt forms of discrimination became illegal after 1994, they nevertheless continue to shape the categorisation of access to formal credit in contemporary South Africa. Moreover, in their numerical form, an inherently political question relating to social and spatial justice becomes depoliticised and relegated to a realm of technocratic decision-making, where they appear to be discrete from social values. As a result, the historical social categories underpinning credit risk scores in South Africa such as race, gender, and neighbourhood, become normalised and reproduced (Singh, 2015).

South Africa’s shift toward credit metrics and calculative technologies

South Africa’s shift towards credit rating models reveals how historical social categories of race and space influence credit assessments. International credit bureaus like Experian and Transunion have been building databases on South Africans since the early 2000s and now track around 24 million accounts (Migozzi, 2020: 646). Credit bureaus collect information about an individual’s creditworthiness based on their financial histories, such as defaults, repayment history, outstanding debts, and the age of accounts. This data gets compiled into a scorecard that ranges between 0 (poor) and 750+ (excellent), which are then sold to financial institutions that use them in their daily vetting routines (Migozzi, 2020). Although these scoring methodologies have been celebrated for minimising bias (Banking Council in Parliamentary Monitoring Group, 2000: Banking Council’s Submission, paragraph 9), their underlying variables and proxies sediment the historical racial boundaries produced by apartheid. In particular, as Butcher (2016: 106) found in her research on racialised risk framings of township areas in Johannesburg, certain indicators such as neighbourhood risk and property size continue to undermine mortgage access in historically disadvantaged neighbourhoods. While financial institutions, particularly mortgage originators typically avoid lending in these areas, when they do, they adhere to prudential risk management, often resulting in higher interest rates due to ‘risk-based pricing’ methods (Butcher, 2016; Migozzi, 2020).

Nationwide, a very small proportion of mortgages are allocated to households earning less than R15 000 per month (Migozzi, 2020: 645–46). In 2020, only 1% of all residential transactions were bonded in the ‘entry’ submarket (CAHF, 2022a: 10), which represents government-subsidised housing that caters for around 60% of the population earning less than R3,500 per month (Standard Bank, 2017: 1). In the affordable and ‘gap’ market, i.e., households earning more than R3, 5000 per month and therefore do not qualify for a government-subsidised house, only 11% of all transactions were bonded (CAHF, 2022a: 10). This is despite government capital subsidies that aim to facilitate access and deracialise the mortgage market., For instance, the 2006 Finance-linked Individual Subsidy Programme (FLISP), a one-off subsidy that ranges between R27,960 and R121,626 (depending on income, the higher the income, the lower the subsidy) to South African households that fall in the affordable housing market R3,501 and R22,000 (National Housing Finance Corporation, 2022) has seen limited success. Against a target of providing 70,000 FLISP subsidies between 2014 and 2019, only 4,400 were dispersed after three years (Turok and Scheba, 2018). Indeed, banks continue to view the affordable housing market as too risky for the expected rewards they can gain. Furthermore, when banks and real estate developers do engage in this sector, they carefully select the spaces where affordable housing gets territorialised. This pattern extends to the rental market, driven by a few REITs who have pioneered themselves as the solution to the country’s ‘affordable housing crisis’ (Butcher, 2020; Migozzi, 2020). Like banks’ affordable housing stock, REITs’ properties are found in selected areas around urban peripheries that are ‘less associated with the territorial stigmas of apartheid that affects property values and access to finance’ (Migozzi, 2020: 650).

Banks and financial institutions are much more involved in conventional, high-end and luxury submarkets. For instance, 87% of all bonded transactions in 2020 occurred in the conventional, high-end, and luxury markets, with the latter making up 45% of the share (CAHF, 2022a: 10). These markets typically constitute historically ‘white only’ suburbs, separated from high-risk, low-income areas through various buffer zones, such as open spaces, transport routes, or industrial zones. Property values in these markets have risen significantly since the 1990s, with the annual house price inflation rate rising annually by 20% until 2006 (Delmendo, 2021: paragraph 1). Although this number fell considerably after the mortgage boom, averaging between 4 and 6% (Lightstone, 2021: 7), values in these areas remain significantly higher than those in affordable and entry markets.

Furthermore, in a struggling economy with high unemployment and interest rates, attempts to expand bank access and unsecured credit have left many South Africans, particularly Black and Coloured working-class families ‘drowning in debt’ (Le Roux, 2019). Despite banks’ financial inclusion initiatives set out in the Financial Charter such as credit cards, mobile money transactions and retail cards for clothing and food the mortgage finance sector has yet to be deracialised (James, 2018). Tracked by bureaus, these initiatives have created a ‘group of unwilling subjects of financialisation’ (Fields, 2017b), who find themselves excluded from the mortgage market, based on their low credit scores. Indeed, ‘unacceptable credit track record[s]’ are the main reason for declined mortgage applications (Eighty-20, 2017: 21). Likewise, REITs in the rental market meticulously select tenants based on their credit scores, with mediocre scores leading to rejection. Like banks, the selection of tenants follows a direct financial rationale: they want to maintain low vacancy rates and avoid late payments to minimise financial losses and attract investors (Migozzi, 2020: 651). In the following section, I consider how this selective character of real estate financialisation unfolds within Johannesburg.

Social and urban structures of housing financialisation in Johannesburg

Founded just over one hundred years ago with the discovery of gold along the Witwatersrand, Johannesburg has grown to be South Africa’s most populated city and home to the ‘richest square mile in Africa’, called Sandton. It is also a city of extreme contrasts, ‘a telling touchstone for South Africa’s extremes and excesses, the mise-en-scene for displays of ostentatious wealth in the midst of deplorable impoverishment and deprivation’ (Murray, 2011: 1). These contradictory experiences largely take place around a geographic divide that separates the leafy northern suburbs from the industrious mining areas in the south. The northern suburbs exhibit claims of Johannesburg being the world’s largest ‘man-made urban forest’, with literally millions of trees, lakes, and parks (Mabin, 2014). Separated by a mining belt, the southern suburbs historically constituted the residential core of white working-class families (Harrison and Zack, 2014). Here, residents face a much harsher micro-climate and landscape, which has undermined investment in the area for decades. The east-west axis runs immediately to the north of the historic Main Reef Road, which traverses the city along the northern side of the gold reef and mining belt. This separates Soweto (South Western Townships) from Johannesburg’s southern suburbs and the almost exclusively formerly ‘whites only’ suburbs to the north (Haferburg and Huchzermeyer, 2017: 67).

The unpleasant climate and landscape south of the mining belt pushed the city’s upper classes and social elite to the northern suburbs, where large tracts of land were demarcated for residential purposes (Larsen, 2005: 31). Properties tend to be markedly higher than those in the southern suburbs, and they constitute the main focus of mortgage market activity in the city (Butcher, 2016). For instance, in 2021, the number of bonded resales and new transactions in Johannesburg was 23,013. Of these, nearly half (47%) took place in the luxury market, where properties are priced above R1.2 million (8558 bonded transactions) (CAHF, 2022b: 8). These market segments are almost exclusively located in the northern suburbs, as Figure 1 illustrates.

Figure 1. Distribution of residential properties by market segment, City of Johannesburg 2021.

Source: CAHF (2022b).

From the 1970s, the residential character of Rosebank, Randburg, Sandton, and Parktown gradually changed, as retail and office decentralisation from Johannesburg’s CBD was starting to gain momentum (Beavon and Larsen, 2014). Reasons for this include ‘white capital flight’, cheaper land, and better working conditions, as well as the financial deregulation and increasing commodification of real estate (see Goga, 2003; Larsen, 2005; Murray, 2011). Indeed, during the 1980s, real estate globally became increasingly valued for its financial performance Locally, commodification occurred alongside an overaccumulation crisis, with excess capital from financial institutions prompting investment in decentralised nodes (Goga, 2003). Backed by an oligopolistic investment structure, this situation spurred a herd mentality in real estate markets that depreciated CBD properties and drove ‘capital flight’ into suburban nodes. ‘Decentralised office investment … became the “flavour of the month”’ (Da Silva, cited in Goga, 2003: 75).

Real estate financialisation and privatised urbanism in post-apartheid Johannesburg

Hopes that the democratic transition would curb decentralisation and restore the historic CBD as Africa’s premier commercial hub were unfulfilled. Instead, the 1990s saw significant urban decentralisation, as large mining houses, banks, and financial institutions moved to suburban nodes (Larsen, 2005; Murray, 2011). Towards the end of the millennium, most CBD portfolios lost half of their value and became abandoned due to their high operational costs at low occupancy levels (Murray, 2011: 89–90).

By contrast, suburban and retail space vastly surpassed the CBD by 1999. Sandton Central tenants paid up to 91.4% more than the highest rental at any time for the best office space in Johannesburg’s CBD area (Beavon and Larsen, 2014: 384). These developments became increasingly financed through speculative investment channels, funded by a growing amount of institutional investors participating in the real estate market, like banks, pension funds, and specialised investment companies (Murray, 2011). Alongside the inner-city’s tarnished reputation, the continuous search for ‘safe’, high-yielding real estate assets fostered the development of secured office complexes, malls, and gated communities in the city’s northern suburbs (Herbert and Murray, 2015: 472). With high walls and extensive security measures, these spaces intensified urban fragmentation in post-apartheid Johannesburg. Urban fragmentation entails the de-solarisation of the city, implying that the city no longer functions as a system, but becomes increasingly juxtaposed, uncoordinated, with different fragments drifting apart (Bénit-Gbaffou, 2009: 1934).

The financialisation of real estate is closely entangled with patterns of urban fragmentation (Herbert and Murray, 2015). The growing presence of financialised capital in real estate markets associates with macro-economic shifts and urban planning changes spurred by the democratic transition. First, post-apartheid government legislation amplified the deregulation of the financial sector initiated in the 1980s. Consequently, the financial sector flourished, capturing around a quarter of the gross operating surplus in the country by 2017 (StatsSA, 2018: 13). The Johannesburg Stock Exchange (JSE) market capitalisation has more than doubled since 1990, accounting for more than three times GDP today (Karwowski, 2021). The spatial transformation of post-apartheid Johannesburg, particularly the growing importance of decentralised suburbs like Sandton, therefore, epitomises the transition from an economy rooted in mining and industry to a finance-centric one. The expansion and dominance of the financial sector have led to an increase in entities, products, and capital involved in real estate. Numerous pension funds, banks, insurance companies, investment firms, and top-listed JSE companies have diversified their portfolios via speculative real estate investment (Butcher, 2016; Goga, 2003). These investments encompass financing construction projects, acquiring shares in high-end properties, or buying shares in listed real estate companies like REITs.

Second, in its quest to establish itself as a ‘world city’, Johannesburg often relies on partnerships between private real estate developers, financial institutions, and public authorities to construct megaprojects like signature buildings, prestigious headquarters, office complexes, and cutting-edge infrastructural services like telecommunication technologies and transport networks. For the most part, however, megaprojects have not included inner-city and township areas, despite insurance companies and large-scale banks paying lip service to the need to revitalise the central city, often refusing to lend money to projects designed to upgrade downtown infrastructure (Murray, 2011: 104). Instead, Johannesburg’s pursuit of becoming a ‘world city’ revolves around projects in wealthy northern suburbs, like Sandton and Rosebank, and include ‘architectural masterpieces’ like the Michelangelo Towers. Recently, Nedbank partnered with The Legacy Group to Africa’s tallest building, the Leonardo in Sandton. This mixed-use property stands 234 m high and boasts 55 floors. Advertised as ‘a beacon of hope in the heart of Africa’s richest square mile’ (The Leonardo, 2022: bottom of logo), the Leonardo offers high-end business facilities, retail shopping, and luxury condos averaging around R10 million per unit. Akin to Dubai’s Burj Khalifa – the world’s tallest building – and the Warisan Merdeka Tower in Kuala Lumpur, skyscrapers like The Leonardo mirror a society’s ambitions. Their aesthetic and architecture resemble idyllic spaces, allowing residents to dissociate from African city realities and immerse themselves in a simulation of European landmarks (Dirsuweit, 2014). To this end, demarcating the boundaries of ‘investable’ geographies in Johannesburg entail the production of spaces that enable consumers to escape post-apartheid South Africa. They function to ‘override historical memory and to replace it with the common sense of consumption and fantasy’ (Mbembe and Nuttal, 2008: 26).

Finally, the rise of REITs in recent years has intensified real estate financialisation in Johannesburg REITs, defined as companies that own and operate income-producing immovable property traded on the JSE publicly traded on the JSE, were introduced in 2013 to ‘modernise’ and ‘internationalise’ the listed property market (Nurick et al., 2018). Since their inception, REITs have ascended to prominence, both in the city’s real estate sector and global indices. South African REITs have accumulated substantial foreign and local capital (SimplyWallSt, 2022: third graph). While REITs can focus on various real estate types, most South African REITs concentrate on office, retail, and industrial real estate in affluent, low-risk suburban areas. In general, these properties embody the forms of enclaves mentioned earlier and are accompanied by sophisticated security infrastructures. One of the largest REITs, Fortress, owns 35 office properties, of which 28 are in the city’s northern suburbs, such as Bryanston, Fourways, Rivonia, and Sandton (Fortress REIT Limited, 2022: Office Properties).

In recent years, REITs and several other corporate real estate actors have displayed a growing interest in constructing new, privately managed ‘satellite cities’ from scratch (Herbert and Murray, 2015). These private cities signify a new form of fast-tracked urbanism that minimises reliance on public infrastructural services. Entirely managed by private corporate entities, often financed by REITs, they include utilities, sewage, roadways, landscaping, security services, and even educational facilities. Waterfall City, for instance, financed by the REIT Attacq, seeks to ‘become a role model for future international urban developments’ (Attacq, 2022: line 1). Situated halfway between Johannesburg and Pretoria, Waterfall City, boasted by Attacq (2022: line 2) ‘urban enclave’ (Attacq, 2022: line 2), has been built from scratch using ‘smart, cutting-edge technologies’, supported by ‘state-of-the-art infrastructure’ (Attacq, 2022: line 2). Similar to other satellite cities, Waterfall City promises a diverse range of mixed-use facilities within its self-contained location. Embracing the ‘live-work-play’ concept, it includes luxury hotels, gated residential estates, enclosed corporate office parks, a logistics hub, shopping centres, private schools, a hospital, parks, fitness centres, and even a heliport (Waterfall, 2022a: A place where you can belong).

Satellite cities embody a new property investment frontier by developing semi-autonomous cities in spaces away from the main city and rapid urbanisation problems (Van Leynseele and Bontje, 2019). Obtaining residency in satellite cities is expensive. In Waterfall City, a four-bedroom house cost up to R13 million (Waterfall, 2022b), while in Steyn City, near Fourways, Johannesburg, this price increases to R29 million (Pam Golding, 2022). These urban enclaves are not confined to South Africa but are spreading across the global South (Van Leynseele and Bontje, 2019). Contrary to ordinary city development planning, where informal aspects and inequalities infringe on the ambitions of urban planners and officials, satellite cities embody spaces of ‘vacant’ land, where rules still have to be made. This form of privatised urbanism intensifies an earlier impulse toward ‘privatopia’ – entrepreneurially driven ‘landscapes of privilege’ – that separate and divide the urban landscape into fragments (Murray, 2013: 132). Thus, in the greater Johannesburg metropolitan region, real estate financialisation and the production of ‘investable’ spaces, have overshadowed local government authorities’ efforts to rectify apartheid legacies. Instead, efforts to brand Johannesburg as a ‘world city’, which required the deregulation of the financial sector and laissez-faire urban planning, resulted in a trend toward spatial fragmentation, with the city splitting into self-sustaining, secure precincts (Landman and Badenhorst, 2014).

Governing financialisation in emerging markets

This concluding section briefly explores the interaction between real estate financialisation in South Africa and global power dynamics. While real estate financialisation stands as a variegated and fragmented process, research shows that its manifestation in the global South is distinct from what has been witnessed in countries like Spain, Ireland, the UK, and the US (Fernandez and Aalbers, 2020). In their recent article, Fernandez and Aalbers (2020: 681) urged researchers to example the relationship between housing and finance developments in the global South and how they relate to those in the global North. Commonly, responses to this question reference structural disparities between the global North and South, including challenges in securing foreign loans in domestic currencies and exchange rate volatility due to foreign liquidity dependency (Kvangraven, Koddenbrock and Sylla, 2021). While these are important insights, it is also necessary to consider the shared political perceptions of nations in the global South, particularly the idea of ‘emerging markets’ that influence the global political economy and potentially reinforce the power and knowledge dynamics within which they are framed.

The idea of ‘emerging markets’ has been framed within a geopolitical imagination rooted in colonial history (Agnew, 2003: 2). Originating from European and later American encounters with the world, this imagination often views the places and peoples beyond the horizon of the West as sources of chaos and danger (Agnew, 2003: 2). Contrary to the civilised, rational, and enlightened colonial gentleman, non-European people and territories are associated with darkness, savagery, and backwardness (Agnew, 2003). This binary representation of societies and spaces has been projected onto the rest of the world, shaping the geographical representations, assumptions, and understandings that guide the interactions among states, governments, markets, and societies (Agnew, 2003: 5).

Consider, for instance, the complex intersection of power, knowledge, and geography in forming the International Financial Architecture (IFA). The IFA defined as ‘collective governance arrangements for promoting the stability of the international financial system’ (Elson, 2010), mainly focuses on preventing financial crises. A key concern with the IFA has been navigating the inclusion of ‘less developed’ countries from the global South into the international financial system. The IFA, in its various revisions and adaptations often brought on by global financial crises, has consistently highlighted the lack of transparency as the most significant threat to global financial system stability (Langley, 2004). While financial crises often originate from Wall Street activities, ‘lack of transparency’ is typically characterised as an ‘emerging market’ problem. Compared to developed markets such as the US, UK, and Europe, emerging markets are considered ‘unknown territories’: their markets are deemed less efficient and novice, their regulatory bodies lack substance, and they grapple with issues like political corruption, societal unrest, and market volatility. Yet, akin to the allure that captivated the imaginations of colonial explorers, these areas are potentially lucrative: emerging markets are geographies to be conquered and resisted.

To manage this risk-reward dichotomy, emerging markets are urged to implement reforms representing ‘standards of civilisation’. These are the benchmarks against which states, banks and companies should align to be considered legitimate international financial system members. These standards are advocated by a network of regulatory bodies such as the Bank of International Settlements (BIS), IMF, World Bank, and private sector actors like credit rating agencies (CRAs). Together, these institutions not only set the standards and codes of conduct governing market actors’ (countries, banks, listed companies, etc.) engagement in global capital markets but also monitor their adherence to these regulations (Langley, 2004). Here, the geopolitical imagination of emerging markets as sources of uncertainty and risk engenders stricter compliance monitoring against global norms for ‘good’ corporate governance and ‘free markets’. For instance, in the sovereign credit rating space, the portrayal of corrupt politicians, inexperienced traders, and opaque market information impacts the evaluation of emerging market sovereign creditworthiness (Schultz, 2021). This calculation assesses a government’s capacity and willingness to repay its financial debt obligations in full and on time. While ability is assessed using quantitative economic growth, debt burdens, and GDP data, willingness involves assessing a government’s trustworthiness. Assumptions of emerging markets as ideological ‘hotbeds’ (Paudyn, 2014: 18) and therefore potentially uncertain increase their ‘political risk’ score, which significantly lowers the overall credit rating. Indeed, this geopolitical perception of emerging markets biases the rating process. As Canbaloğlu, Alp, and Küçükkocaoğlu (2019) demonstrate, CRAs do not apply the same variables when rating emerging market governments and advanced countries, with ‘home countries’ often receiving better scores than their economic and political fundamentals warrant (Fuchs and Gehring, 2017).

However, in a context where governments, banks, and corporations rely on capital markets for funding, emerging market actors typically adopt necessary standards to enhance their creditworthiness. For South Africa, this meant implementing macroeconomic reforms that facilitated economic financialisation and made national banks, listed companies, and the government subject to the surveillance of rating agencies and supranational regulatory bodies. These reforms include fiscal and monetary policies such as central bank independence and an inflation-targeting regime (Padayachee, 2019). These changes, aimed at demonstrating the government’s commitment to prudent economic management, significantly influenced the selective character of housing and real estate financialisation.

First, the rationality of creditworthiness has subjugated commercial banks to the scrutiny of rating agencies and the prudent lending regulations stipulated in the Basel Accords. Established by the Switzerland-based Basel Committee on Bank Supervision (BCBS), the Basel Accords aim to regulate the capital levels of international banks about the risk profiles of their assets (Bruneau, 2023). The BCBS determines the risk levels for different assets, including sovereign debt. Based on the calculation of risk-weighted assets (RWA), banks must maintain minimum capital reserves as a safety measure against credit default (Kruck, 2011: 53). RWA refers to the estimated value of a bank’s assets, adjusted for credit risk. Higher credit risk assets require higher capital reserves. Here, the question of how to evaluate the risk of sovereign exposures is contested Evaluating sovereign exposure risk remains contentious. The first Basel Accord determined that not all sovereigns were equal: debt of a select group of OECD countries was considered risk-free, whereas non-OECD members were seen as fully risky. In other words, lending to the latter group of sovereigns, their banks, and public sector entities would receive a 100% risk weight (Bruneau, 2023: 158). Although subsequent Basel Accords removed this advantage for OECD countries, sovereign credit risk remains influential in calculating RWA and capital reserve requirements. Thus, in emerging markets like South Africa with ‘below investment grade’ credit ratings, banks must set aside additional capital reserves. For example, under the Basel III Accord, real estate exposures with higher LTV ratios contribute to a higher risk weight, which means higher capital reserve requirements (BCBS, 2017: 21). South African banks have been lauded for their overcompliance with these international best practices and the implementation of Basel reforms, which have secured favourable credit ratings on the one hand, and strengthened their conservative lending criteria in a context where the government’s creditworthiness remains volatile on the other (Butcher, 2016: 102).

Second, cautious real estate investment decisions are underpinned by the discourse valuing risk and influencing capital flows in emerging markets. Because they are perceived as ‘higher credit risk’ investments, emerging market assets and stocks are often undervalued or require a considerable risk premium for potentially high rewards (Lee, 2003: 65). The hypermobility of capital leads to short-term procyclical investment into emerging markets. Risk-loving investors tend to ‘play emerging markets, favouring assets when risk aversion is low and swiftly withdrawing investments when risk aversion escalates (Schultz, 2021). This tends to exacerbate a slowdown in growth in emerging markets, subsequently increasing credit risk and borrowing costs (Lee, 2003). These volatile capital movements fortify the conservative, risk-averse investment strategies of real estate actors like REITs, commercial banks, and institutional investors, as short-term investment horizons deter ventures into presumably higher-risk markets, potentially devaluing their assets and necessitating increased capital reserves. Indeed, overexposure to the South African property market is perceived as a risk by credit rating agencies (Moody’s, 2022), prompting many REITs to divest from South African assets and invest in European assets.

Conclusion

This article explored the spatial aspect of real estate financialisation within a global Southern context using the case of South Africa, highlighting its divergence from trends in the global North where financialisation often targets low-income, minority households and neighbourhoods through predatory lending regimes (see also Migozzi, 2020). By contrast, the financialisation of real estate in South Africa is characteristically conservative, targeting only low-risk areas and the most creditworthy households. This selectivity traces its roots to a long colonial and apartheid history where access to property markets was dictated by racial imaginaries of risk and reward expressed in spatial terms. Although the post-apartheid government has outlawed any form of discrimination based on race, gender, and ethnicity, this article has shown how these rationalities and assumptions still underpin the boundaries of the housing market and the role that credit-scoring technologies play in this regard.

The article also emphasises the discriminatory nature of housing and real estate financialisation in Johannesburg, revealing how the surge of real estate capital post-apartheid has accentuated urban fragmentation, characterised by the creation of segregated spaces like secured office parks, gated estates, malls, and entertainment centres. The emergence of privately managed ‘satellite cities’ has exacerbated this issue, having, as Murray (2013: 122) states, ‘turned their backs on the surrounding cityscape…[and] any sense of responsibility for knitting the urban social fabric together in ways that promote racial harmony and spatial justice’. The spaces resulting from real estate financialisation starkly oppose the city’s aspirations for social and spatial inclusion. Finally, the article underscores how these conservative real estate investment practices are mediated by relations of power and knowledge rooted in the colonial global economy. South Africa’s status as an ‘emerging market’, conceptualised within the modern geopolitical imagination, significantly shapes real estate actors’ lending and investment strategies. The pressures to adhere to societal, financial, and political standards of creditworthiness – set by former imperial powers – reinforce prudent asset management by banks and real estate companies. The analysis extends the geography of financial markets to consider the role of spatial representations, assumptions, and understandings in cross-border capital organisation. Essentially, this article has sought to draw attention to the urgent need to analyse the geographies configured through financial markets. An analysis of the geographies of financial markets enables us to consider how financialisation operates as a form of spatial fixity, where ‘credit, leverage, financial speculation and financial engineering act to displace in time and across space’ (French, Leyshon, and Wainwright, 2011: 812).