Financial markets,developers and the geographies of housing in Brazil: A supply-side account

Introduction

The literature dealing with the financialisation of the urban built environment devotes attention to the pervasiveness of financial markets and how it affects access to homeownership, household indebtedness, housing prices and the geographical patterns of mortgage lending. However, this focus on access to homeownership and its social and spatial implications overlooks how financial markets influence the supply side of the housing sector – that is, the development industry, its business strategies and the nature and location of its products. Following this research agenda, the paper aims to analyse the evolving interrelations between financial markets and the development industry, and how these interrelations contribute to redefining the geographies of housing production in Brazilian cities.

Contrasting with a ‘mechanic, relational concept of the economy’ (Froud et al., 2006: 69), the paper questions a view where developers’ strategies and the resulting geographies of housing would be a straightforward embodiment of the ‘logics’ of shareholder value. Rather, combining two strands of yet unrelated approaches to financial markets (cultural economy and conventionalist economics), the paper develops the concept of resonance to grasp the interactions between a given industry (housing development in this case) and financial markets. Considering with Froud et al.’s cultural economy perspective that the shareholder value theory is a ‘pliable rhetoric’ (2006), the paper pays attention to the narratives on individual firms that are co-authored by the stock market community (investors, analysts, the business press, etc.) and firm managers. While such narratives are performative in the sense that company managers have to make strategic moves, there is nothing predetermined in the content of such moves. Thus, developers’ managers are given back the agency they lost in more mechanistic accounts. Yet to understand how the fragmented narratives and strategic moves relating to each individual firm may contribute to reshaping the geographies of housing, one has to look at their aggregated outputs. This calls for a complementary perspective at a meso-level that stresses how shared social representations (or conventions) support the coordination of financial actors. Conventionalist economics helps to move from the description of the idiosyncratic narratives on individual organisations and to analyse how financial markets adopt dominant valuation and interpretation conventions that affect a group of firms considered to share similar characteristics (such as belonging to the same industry). There is however a need for a unifying concept to reconcile both strands of literature. The paper brings forward the concept of resonance to explain the interactions whereby (1) fragmented narratives on individual firms, and their related strategic moves, eventually vibrate in unison and stabilise into a shared representation on an industry’s fate, and reciprocally, (2) how conventions evolve as financial results diverge from dominant narratives and call for alternative strategic moves by managers. Resonance does not only ground the analysis in actual economic practices, in contrast to macro-political economy accounts, it also contributes to a literature that attempts to go beyond restrictive binary distinctions (between micro- and meso-levels, between stock market communities and company’s management). It constitutes, therefore, a heuristic tool to analyse the interactions between financial markets and developers, the related changes in Brazil’s housing industry, and their potential impacts on the urban built environment and its geography.

Focusing on Brazil provides a unique vantage point for clarifying such relation. If the housing sector was mostly made up of small-scale family-owned firms, a wave of initial public offerings (IPOs) in the mid 2000s brought a group of developers under the pressure of the stock market community. Relying on a qualitative analysis, the paper explores (1) how a convention centred on growth (of outputs and revenue) took shape as these firms went public between 2005 and 2007; (2) how this convention unravelled between 2010 and 2011 as financial results diverged from shared expectations, leading to mounting tensions between the actors involved; and, finally, (3) what consequences the emergence and dissolution of this convention centred on growth had on the geographies of housing provision in Brazil.

Financialisation, housing and developers: The case for a supply-side analysis

‘Centring housing in political economy’, authors such as Aalbers and Christophers (2014) or Schwartz and Seabrooke (2008), have stressed how housing finance systems are key in understanding today’s political economy, if only because homes are a major financial asset in most economies (Schwartz and Seabrooke, 2008: 238). With the rising importance of mortgages, and, in ‘liberal’ residential systems, the securitisation of such mortgages (Rolnik, 2013; Schwartz and Seabrooke, 2008), housing and financialisation are said to be intrinsically linked, households and their homes becoming financialised subjects (Forrest, 2015; Langley, 2007). The genesis, politics, modus operandi and resulting spatial, social and racial consequences of subprime lending offers a vivid illustration of such growing interdependencies between finance and housing (see the contributions to Aalbers, 2012, as well as Ashton, 2009; Carruthers and Stinchcombe, 1999; Poon, 2009).

Yet, financial markets also connect to housing through the development industry, i.e. not only via homeownership by households, but also through the production of housing by development firms. While taking into account this supply-side perspective is important in all countries, it is of particular interest for the Global South where the influence of mortgages and their securitisation remain limited. In so-called ‘emerging’ countries that adopted neoliberal reforms, the role of financial markets in the financing of firms increased, while, at the same time, global capital availability has escalated, especially in the 2000s. The development industry is no exception: cash-consuming developers are eager to resort to private equity funds and stock markets to boost their growth (Rouanet and Halbert, 2016).

For research focusing on the supply-side, there is a tradition concerned with how debt lavishly provided by financial institutions to property developers leads to boom-and-bust cycles (from Haussmann’s Paris through 1980–1990s London’s Docklands to 2000s Halifax, Canada: Fainstein, 1994; Harvey, 2003; Rutland, 2010). However, only a handful of work has touched upon how finance capital transforms housing provision by investing equity into development firms. Ball (1983) observed, for instance, how Initial Public Offerings (IPOs) could alter development companies by providing the means to increase their total volume of production, and to rescale country-wide. However, it was only recently that this initial probing was revived by works exploring how finance capital flows are ‘anchored’ in urban spaces (David and Halbert, 2014a; Theurillat and Crevoisier, 2014), making a case for a systematic analysis of the interrelations between financial markets and the development industry.

In this light, some recent contributions have considered how their ties with financial markets can transform not only developers’ strategies but also the resulting geographies of the built environment. It is demonstrated how in India, developers are among the intermediaries of the ‘transcalar territorial network’ that channels foreign and domestic finance capital into metropolises (Halbert and Rouanet, 2014). These developers leverage such capital flows to increase their market share over a wider number of regional markets. Consequently, they gain political agency in the making of contemporary Indian metropolises (Rouanet and Halbert, 2016). In Mexico, David (2013) analysed how small local developers changed their organisation, business model and reporting practices to meet foreign financial investors’ expectations. Already dominant developers, having access to alternative sources of funding, nonetheless managed to shut out global financial investors from central markets of the Mexico city-region (David and Halbert, 2014b). This hints at recognising the agency of developers that pursue their own agenda and interests (Searle, 2014).

In spite of this growing number of empirical accounts on the interactions between two parallel processes (increasing presence of finance capital in the development sector, on the one hand, and changing business strategies of developers that impact the geographies of the urban built environment, on the other), more work is needed to formulate a conceptual framework to explain such interactions.

Financial markets and corporate management: Towards a theory of resonance

Froud et al. (2006) discusses existing conceptual frameworks that theorise the relations between financial markets and the corporate economy. For them, mainstream financial economics, in the form of agency theory, as well as its political economy opponents (such as the regulation school or the varieties of capitalism approach) share a unidirectional, arrow-to-box understanding (Froud et al., 2006: 130) where financial markets act as the driving force behind the evolutions in a given industry. Consequently, it is shareholders’ capital accumulation strategy that would impose itself onto development firms and be reflected in the evolving geographies of housing provision. However, by claiming the efficiency of financial markets in the allocation of capital, or by criticising the hegemony of shareholders, both accounts dismiss the agency of firms’ management. This results in downplaying the interactions between financial actors and development firms, two sets of organisations whose interests, values and views have empirically been observed to diverge at times (David and Halbert, 2014b; Searle, 2014). To move beyond such ‘mechanistic’ thinking (Froud et al., 2006: 69), two yet unrelated fields in the studies of financial markets, i.e. cultural economy and conventionalist economics, can fruitfully be combined.

Considering that economics – i.e. discourses and representations on the economy – formats economic objects rather than the other way round (Pryke and Du Gay, 2007), Froud et al. (2006) pursue a cultural economy perspective to reconsider the dominant representation in financial markets known as the shareholder value theory (SVT) from the 1980s onwards. For them, SVT is a highly ‘pliable rhetoric which can be borrowed, used and influenced ‘ (Froud et al., 2006: 38). If its bottom line focuses on potential pecuniary returns for shareholders, it is enacted through the daily interactions between the stock market community and company managers. The analysis of these interactions reveals how a company’s strategy is thus the subject of multiple narratives co-authored by fund managers, analysts, the business media but also by firms’ managers, and borrowing from different discursive levels (company-, industry- and grand macro-economic narratives). Following the cultural economy approach, narratives are performative: they constrain corporate managers to demonstrate that they act accordingly. Yet, under a British intellectual tradition revolving around MacKenzie’s works, performativity remains for Froud et al. (2006) essentially open in the forms it effectively takes. Managers devise the ‘strategic moves’ they see fit as long as they remain discursively tied to the shareholder value rhetoric. Yet, claiming that financial returns are redistributed to consumers rather than to shareholders because of the intense price competition in present-days’ product markets, Froud et al. observe that numbers, read financial results, may frequently not corroborate the narratives. These discrepancies lead to a new round of narratives and ‘strategic moves’ that might equally prove financially unfruitful, and so on.

This conceptual framework that holds together narratives, strategic moves and numbers, is particularly fit for analysing individual firms through thick historical case studies (see examples on GE or Ford, Froud et al., 2006). However, since the work is rooted in management studies, and because of the reticence, in contrast with political economy works, to engage in any totalising account, the priority is given to the description of idiosyncratic interactions between the stock market community and the managers of a given firm, i.e. at micro-level. It does not recognise that the myriads of narratives and strategic moves may converge across firms of a given sector, as if vibrating in unison. It consequently weakens any attempt to examine how idiosyncratic interactions relate to the meso-level, and how their convergence can influence, in our case, the evolving geographies of housing.

While not considering micro-level interactions between financial actors and corporate management, the conventionalist approach to finance provides elements to explore such a meso-level. Extending the interest paid by conventionalists to the coordination of economic agents, this approach insists on the role of shared social representations (also known as conventions) in the valuation of firms and the on interpretation of their outlook by financial actors. Orléan’s seminal work (1999) discusses how shared ‘interpretation conventions’ (Orléan, 1999: 145) gain legitimacy among financial actors, possibly leading to boom-and-bust cycles, as more investors abide by the dominant convention. Other authors have complemented this approach by stressing how ‘valuation convention’ are key to their coordination (Lavigne, 2002; Tadjeddine, 2006: 195): financial actors share similar interpretations not only of the future values of companies, but also on the elements that need to be looked at to value companies (such as preference for EBITDA multipliers over Price/Earning Ratios). The attention paid to the mechanisms by which financial actors coordinate their investment decisions through shared social representations goes beyond the fragmented analyses at micro-level. But only as long as one does not fall into the trap that, akin to mechanistic accounts, would consider the predominance of financial markets as inevitable, or at least, unmediated by company managers. Influenced by a proclaimed Marxist approach (Orléan, 2004: 17), the conventionalist approach of finance may be inclined to adopt such an assumption, so that conventions are often depicted as endogenous to the stock market community, paying little if any attention to the agency of management firms. Using a conventionalist perspective, D Lorrain’s analysis of the French electricity sector demonstrates however that in their interactions with financial intermediaries, ‘managers find some room for manoeuvre, allowing them to negotiate with “what the [financial] market says”’ (Lorrain, 2009: 62).

It is pressing to reconcile both perspectives to analyse how firms’ business strategies respond to the pressure of financial markets, and, in our case, how the production of housing by developers may be affected. To do this, we consider how the narratives co-authored by the stock market community and developers’ managers over individual firms resonate into financial valuation and interpretation conventions on the wider housing development industry. By resonance, we point to how the multiple interactions happening between corporate managers and shareholders around each individual organisation mutually influence those occurring around other firms, and temporarily stabilise, as if progressively vibrating in unison, into a dominant convention applied to a subset of firms considered by the actors to be part of an industry. In this light, the circulation of industry-wide and grand economic narratives is key in putting individual company narratives in resonance (arguably irrespective of the fact that organisations do effectively share common elements or not, as evoked by Froud et al., 2006). Resonance thus provides a dynamic understanding of how conventions are made and remade over time as an interactive process between the stock market community and developers’ management. As discrepancies between narratives and numbers are repeatedly observed at micro-level, dissonances in the dominant convention occur, leading to another round of narratives and performative actions which will resonate anew, at least until another wave of disappointment crashes against the (renewed) promises of pecuniary returns for shareholders.

Resonance is thus a concept that brings back the question of power relations into economic activity and of their effects. On the one hand, the process of resonance may encourage convergence in management initiatives. As such, it constitutes an important but still overlooked factor to explain the transformation in the provision and geographies of housing by developers. On the other hand, it may also not go unchallenged: some actors, such as corporate managers in firms whose capital is still partially controlled by founding families and individuals, can raise diverging views, and offer resistance to the adoption of the ‘strategic moves’ accompanying dominant narratives (David and Halbert, 2014b for empirical illustrations; Searle, 2014) as will demonstrate the case-study of housing production in Brazilian cities.

To fruitfully analyse this process of resonance, the research protocol seeks to go past binary distinctions (between micro- and meso-levels, as well as between the financial market community and its object, i.e. firms’ management). Thus the research collected evidence that clarifies (1) the business strategies of listed developers and how they affect the provision of housing in Brazilian cities, and (2) the narratives shared by the stock market community, and occasionally contested by other actors, concerning the adequate evaluation of the housing industry. This material consists of secondary sources including selected business press material (media outlets such as Exame, Valor Econômico, Construção Mercado, Istoé Dinheiro, etc.), periodical reports issued by developers to shareholders containing operational and financial information, as well as transcripts of teleconferences between managers and shareholders. This documentary corpus was triangulated with primary data from 15 interviews with development firms’ managers and business analysts conducted between August 2011 and May 2012 in the cities of São Paulo and Porto Alegre.

Fifty years of housing finance in Brazil

To understand the interactions between financial markets and developers in the 2000s, one first needs to see how growing demand for housing challenged the existing housing finance system in Brazil.

Residential development activity took shape as a distinctive sector in Brazil in the 1960s (Fix, 2011; Ribeiro, 1996). Previously, lack of a national mortgage finance system had restricted homeownership to higher-income households. The Housing Finance System (HFS), tightly regulated by a National Housing Bank, was introduced in 1964. The HFS encompassed a market-oriented development sector, under which savings deposits made by households provided mortgages for middle-income buyers with interest rates defined by the central government, and a social sector which used compulsory savings (deposits made by employers on behalf of employees) to develop lower-income housing projects, most of them managed by municipal government-controlled housing cooperatives (Arretche, 1990; Maricato, 1987; Valença, 1999). Combined with an accelerating pace of economic and demographic growth, as well as rapid urbanisation, the new availability of housing finance supported strong demand for housing in the 1970s (Maricato, 2011; Royer, 2009). This opened up opportunities for development firms to grow (Farah, 1985; Fix, 2011). Yet, the sector remained predominantly composed of small and medium-sized family-controlled firms operating in their hometown markets. This was due in part to the hurdles that managing a nationwide business in a large territory with limited transportation and communication infrastructure presented. Additionally, and of central importance to subsequent events, was that the HFS restricted development finance solely to the covering of construction costs. As the capital markets of Brazil were still weak in the 1960s, finance to scale up activity through land banking was scarce.

Brazil’s economic growth stalled in the 1980s as the Latin American foreign debt crisis unfolded. Mortgage lending plummeted as a result of rising default rates among borrowers, and the National Housing Bank was dissolved in 1986. During most of the 1990s, faltering growth, rising unemployment, monetary austerity and poor coordination among the agencies managing the HFS system further sapped mortgage lending and construction finance (Azevedo, 1996; Royer, 2009; Valença, 1999; Valença and Bonates, 2010). Development firms were forced to adapt by directly financing homebuyers through off-plan sales or through housing co-operative schemes, often on five- or ten-year loan contracts (Castro, 1999). By immobilising capital in such loans, firms lacked the resources to expand their businesses. Developers thus remained small and medium-sized firms, controlled by their founding owners, with a strong specialisation both in terms of geographical reach (targeting a single regional market) and of segment (catering for middle and upper income households who could afford short-term loans).

A new scenario began to emerge at the beginning of the 2000s. Brazil’s economic outlook improved as a result of higher growth rates, falling unemployment and rising incomes. Furthermore, a series of regulatory changes and new tax incentives enhanced support for mortgage lending. In order to introduce mortgage securitisation in Brazil, deeds of trust were permitted on mortgage contracts (Law of Sistema de Financiamento Imobiliário – SFI, 1997). It was argued that these would improve security for lenders who had long complained about the obstructions posed by the judicial system on foreclosures (Royer, 2009). While mortgage securitisation did not subsequently grow as expected, ¹ deeds of trust nevertheless helped to boost conventional mortgage lending through the HFS, especially when interest rates started to fall in 2005: the number of households annually financed increased from 200,000 before 2005 to more than 1 million in 2010.

Development firms responded to this surging demand in different ways. Some opted for organic growth, eschewing radical changes in managerial and organisational structures. Others embraced more radical expansion plans, but lack of finance for land banking remained an obstacle. The solution found was to team up with local and international financial actors. This was facilitated by a sudden abundance of liquidity which resulted from the combination of (1) incentives and regulatory changes from the government to make investing in Brazil’s stock exchange more attractive (Paulani, 2008), and (2) growing numbers of foreign investors seeking to boost their portfolio performance via investments in supposedly higher-yield ‘emerging’ markets. Between 2003 and 2007, developers were thus able to tap into the financial markets to acquire land and expand their businesses.

Financial markets as growth accelerator

The first movers were developers such as Gafisa and Even, who partnered with private equity firms. The former received an initial investment in 1997 from GP Investimentos, a Brazilian private equity fund that would take full control of the firm in 2004. In 2005, the US-based Equity International made an additional BRL135 million investment in exchange for a 32% stake in the same Gafisa (Mandl, 2006). Similarly, in 2006 Even set up a partnership with London-based Spinnaker Capital, a private equity fund that brought BRL 72 million (Aranha, 2011). More common, however, was the case of developers that attracted financial investors by issuing stocks and bonds between 2005 and 2007 with the aid of underwriters such as Credit Suisse, UBS Pactual, Merryl Linch, Banco Votorantim and Deutsche Bank. With the exception of Rossi, which moved from a lower listing segment to the newly created Novo Mercado segment of the Bovespa Stock Exchange ² in 2006, the other 16 residential developers that issued stocks in this period were not publicly listed previously. It should be noted, however, that differences in terms of ownership structure persisted, depending on how influential founding members remained after dilution.

Partnering with financial actors had transformative effects on the housing development industry. First, while small, family-based local firms still populated the sector, access to finance capital supported the concentration of development activity, especially in metropolitan areas. While data on the extent of concentration are scarce, our own empirical estimate is that in 2010, four developers were responsible for one-quarter of all new housing launchings in the metropolitan area of São Paulo. This is consistent with Ball’s observation (1983): when developers issue stocks, the influx of capital allows them to grow their market share because of rising total output.

Alongside concentration through growing output, what made possible this rapid gain in market share was a wave of mergers, acquisitions and partnerships that followed the IPOs (Cichinelli, 2008a). By means of this wave of activity, developers, hoping to capitalise on expertise embedded within the partnering or acquired firms, began to develop for new market ‘segments’ and to enter unfamiliar regional markets. PDG’s chairman explained the thinking thus:

Where were the good professionals [in this market], which were few? The good professionals were the owners of the firms that had managed to survive in the 80s, 90s. These guys are not available [for hiring] in the market. The only way to get them to work for you is to acquire a stake in their firms. This strategy has become widespread in the sector. (Blanco, 2008)

Between 2007 and 2010, acquisitions grew substantially. They involved large developers, as in the case of Gafisa’s acquisition of the low-income segment specialist Tenda, or PDG’s acquisition of Agre (Cançado, 2010). However, joint ventures were more frequent, associating listed developers with smaller local developers and builders, especially when entering a new regional market (Rufino, 2012). Parallel to tighter horizontal interactions between development firms, denser vertical linkages were also established between large developers and a range of real estate organisations, such as business consultancies, architects, brokers and banks.

A second transformative effect concerned the business organisation and governance of developers. A clearer distinction was drawn between owners and management teams, even in cases where members of the owning family continued to perform managerial functions. This also involved attracting managers with professional careers in other sectors to positions that were either previously filled by family members, such as financial management, or that were created after the IPOs, such as investor relations (Lindemann, 2008). The effects of these organisational changes, perceived as beneficial by many, were not limited to listed companies:

I believe it was good for shaking up the market, for improving governance […]. Development activity had been going on since 1964 […] but I believe it was upgraded with […] new funding, with regulatory improvements, and with the IPOs, [all of which] qualified governance in real estate development. This is the role they [listed firms] had. Competition with them was healthy for smaller and medium-sized firms. (Interview no. 3, Analyst, Sinduscon, November 2011)

Alongside the organisational dimension, to attract private equity funds or as a pre-requisite for taking part in Bovespa’s Novo Mercado, developers had to demonstrate a stronger commitment to transparency and improved public relations with shareholders. This involved the release of quarterly reports with detailed operational and financial results, holding teleconferences with investors and analysts, adoption of international standards-based accounting procedures, having balance sheets audited by one of the ‘big four’ multinationals, and arranging key information releases in the business press (Oliveira, 2008a).

Lastly, access to finance capital had a transformative effect via the industrialisation of developers’ business models. Developers embraced project management technologies and introduced information technology to improve information exchange and enhance cost control in the construction process (Shimbo, 2012). Projects and product design became standardised (Fix, 2011; Shimbo, 2012). Economies of scales were sought via larger projects to reduce the marginal costs (of land, project approvals, marketing …) (Sanfelici, 2013; Sigolo, 2014). Lastly, most firms embraced nationwide (as opposed to local and regional) strategies of growth, as well as segment diversification across income strata (Olivion, 2010; Rufino, 2012; Sanfelici, 2013).

Concomitant to the access to financial markets, developers grew significantly, as demonstrated by both output trends (Figure 1) and the evolution of their land banks. For instance, Cyrela’s land bank grew from 3 million m² in 2005 to 13.6 million m² in 2010 and the value of Gafisa’s land bank rose from BRL3 billion in 2006 to BRL18 billion in 2010.

OPEN IN VIEWER

Figure 1.

Housing starts by the seven largest listed developers (2004–2011).

In sum, the housing development industry evolved rapidly in the course of the 2000s, with a group of developers asserting dominance in larger geographical markets. Even though such changes did not affect all markets equally, the restructuring had a distinctive impact on the urban built environment in many Brazilian cities, both directly and indirectly.

Changing the geographies of housing provision

As developers expanded into new regions and pushed for scale economies, their activity left an imprint on major cities. While large-scale projects were not a novelty, they became more widespread as these firms made headway in the market. Apart from conventional apartment buildings, which continue to dominate densely built-up areas where land is scarce, three types of projects prevailed. Products for the so-called lower-income ‘segments’ consisted of highly standardised apartment blocks (usually five-storey buildings with no elevator) and a few amenities, such as sports courts. For middle- and upper-income ‘segments’, two types of projects became more common: large-scale land developments and mixed-use projects. The former includes gated communities where land plots or houses are sold to homeowners as well as high-rise condos with in-house amenities and some common infrastructure (streets, lighting, parks) provided by the developer in co-operation with local governments. ³ Mixed-use projects usually group office, residential and retail functions in a single land plot.

The spread of such projects have transformed the geographies of housing, especially in large cities (see Figure 2). Whereas new developments had generally been concentrated in inner city, higher-income neighbourhoods throughout the 1990s and much of the 2000s, they now spring up more and more outside this core area (see Figure 3). Because they require larger and cheaper land plots to exploit economies of scale, lower-income projects have been generally provided in the outskirts of urban areas, contributing to urban sprawl in cities that lack efficient transit systems (Ferreira, 2012; Klink and Denaldi, 2014; Maricato, 2011; Cichinelli, 2008b). Middle- and upper-income land developments, although generally located outside denser, higher-value neighbourhoods, are either near the city core or within easy access of it through major thoroughfares. Developers built mixed-use projects on higher-value land, catering to young professionals believed to prefer areas of the city where cultural activities and entertainment are within easy reach (see Figure 3). As already suggested (Ferreira, 2012; Maricato, 2011; Rufino, 2012; Volochko, 2012), these projects have generally reinforced fragmented patterns of land use. Many are inward-oriented as a result inter alia of their combining different functions, the availability of amenities that act as a substitute for public space (parks, sports courts, pools, etc.) and the security apparatuses that inhibit or prevent circulation of outsiders.

OPEN IN VIEWER

Figure 2.

Large-scale projects in Porto Alegre: two higher-income land development projects (top) and two lower middle-income condominiums (bottom).

OPEN IN VIEWER

Figure 3.

Types of large-scale projects by selected developers in Porto Alegre (2007–2011) .

The expansion of development firm activities into new regional markets has resulted in the wider spatial circulation of these products into second-tier cities, a process supported by the subsidies included in Brazil’s lower-income housing programme for homeownership. ⁴

There is thus strong evidence of systemic relations between three processes: (1) the growing role played by the stock market in the ownership of housing development firms; (2) changing business strategies of these firms; and (3) redefinition of the geographies of housing in Brazil. It remains to be explained how such interdependences have occurred, by inquiring into the interactions between the stock market communities and development firm managers.

The growth narrative

Following the cultural economy assumption that discourses and representations on the economy format economic objects, we can observe how, in the 2005–2010 period, a narrative in favour of growth was co-authored by four main groups: developers, underwriters, financial analysts and fund managers. This narrative emerged with the first IPOs of 2005 and encouraged the view that development firms should be valued on the basis of their ability to grow quickly. The faster a firm grows, it was assumed, the higher its turnover, and the higher capital gains for shareholders thanks to rising stock prices.

It was developers who originally devised this narrative. Because, as noted in section ‘Fifty years of housing finance in Brazil’, conventional bank lending could only be used to cover construction costs, developers turned to financial markets in order to facilitate land acquisition. In doing so, developers argued that, by solving the land bank bottleneck, financial actors could legitimately expect a strong growth in market capitalisation, reflecting soaring levels of housing production (Amato, 2008).

This was the key insight of firms in tapping into capital markets: the business is too intensive in resource consumption. So firms saw it in this way: ‘my turnover cannot be accelerated if I’ve got no money to invest [in land acquisition, approval, etc]. How can I speed up turnover? By getting access to investors’. Thus the motivation behind IPOs was to obtain land. They [developers] saw an opportunity [in the improving economic environment]. But then they’d say: ‘[…] I have all I need except for money for buying land’. This was their motivation […]. They called investors and said: ‘Look, Brazil is doing well, this is the economic outlook, from now on we have a strong growth potential, but I need you as my partner […]’. Most developers pay only the minimum [required by law] in dividends, because all investors made a bet on growth. (Interview no. 5, Analyst, Banco do Brasil, November 2011).

Underwriters also endorsed this emphasis on future growth since their fee-based revenue model encouraged them to boost firms’ capital values at the date of IPO launch (Lima, 2007). Analysts corroborated development firms’ narrative, first by espousing a broader macro-economic forecast of strong growth for the Brazilian economy as a whole, second by mobilising industry-wide narratives ranging from government incentives for the sector to promises of efficiency gains associated to larger firms. Lastly, as excess liquidity in the mid 2000s drove down interest rates in ‘developed’ economies, fund managers were eager to reap the benefits of Brazil’s growth. The capital that flowed in from the Global North was looking for high-risk high-return investments to complement lower-return assets in their home economies. In other words, investors were looking for a scenario of quick capital gains (through share price increase) rather than long-term growth (and returns based on dividends).

These narratives, which were elaborated investment decision after investment decision, i.e. at the micro-level of each individual developer, circulated within and across the development and stock market communities that were both mainly concentrated within the city of São Paulo. Developers were forming an increasingly tightly bound community, based on similar professional interests and interpersonal acquaintances. Interviews reveal how the narrative based on the land bank bottleneck was circulated among this community as a way of attracting investors:

The idea of [emphasising] land bank appeared in […] 2005 when Cyrela went public and had to show to investors [that] […] [it] could grow in the future. That’s how Cyrela managed [to go public]. Until then no developer in Brazil had had this idea of going to the stock exchange. And this turned into a frenzy. (Interview no. 3, Analyst, Sinduscon, November 2011)

According to a manager at Cyrela, the exchange of information between developers at the IPO moment was key in shaping their strategies:

The strategy [on IPOs] was one of growth. If you look at all [firms’] reports, they sort of say the same thing, because firms copy each other. There’s something called IPO prospectus. I myself received calls from people of [other firms] asking, ‘how should I do this or that’. He […] was preparing a prospectus. So all firms followed [similar strategies]. (Interview no. 2, Project manager, Cyrela, March 2012)

This narrative was taken up by the stock market community, which progressively expanded the idea of a growth scenario to the entire development industry, thus seeing in developers’ ‘investment thesis’ (Interview no. 5, Analyst, Banco do Brasil, November 2011) an opportunity for profit. Through the interactions within and across these two communities, these narratives started to resonate, culminating in the IPOs ‘window’, which was concentrated between 2005 and 2007. All in all, the ‘land bank bottleneck’ provided a convenient story with its focus on fast growth: under conditions of strong housing demand and government incentives; if the bottleneck could be removed, the turnover would quickly soar, giving a pecuniary return to shareholders because of the resulting market capitalisation increase.

This dominant narrative supported – and was enacted through – a valuation convention that spread across the stock market community. The focus being on the capital gains that could be derived from soaring share prices, developers were valued on the basis of elements that could signal future growth in product output. Yet, the stock market community (investors, analysts, business media) was uncertain as to what indicator should be given priority in assessing prices, bringing evolutions in the valuation convention.

Part of the impact of the markets on large cities [came from] this need for generating results. Investors didn’t know where to look. There was a history of analyzing firms […] in manufacturing, in services, but the construction sector was […] new to the stock exchange. What criteria [should investors] use? In this uncertainty, one of the criteria they began using was land bank. If you had a good land bank, it meant you’d launch more, and at this point developers rushed to build up their land banks, causing land prices to rise. (Interview no. 4, Analyst, Fundação Getúlio Vargas, October 2011)

This initial emphasis on land banks was also supported by IPO underwriters. Considering the stock of land as an indicator of future margins, they valued land not at current purchase value but at the Potential Sales Value (PSV – i.e. as the value when entirely developed). Furthermore, they speculatively bet on stability of construction costs and on ever-increasing housing demand (Lima, 2007, 2012). However, recognising that development would take time, and that land banks could not be directly transformed into next year’s turnover, the stock market community shifted from gross land banks to estimates of launches such as guidance value at one- or two-year terms. Likewise, mergers and acquisitions were also a positive indicator later encouraged by stock markets, as with the praises received by Gafisa’s management when they bought Tenda. In other words, all throughout the 2005–2010 period the focus remained on the growth of firms’ total profits, as illustrated by the fund managers’ use of EBITDA multipliers (Gregorio, 2010; Huerta and Motta, 2011; Lima, 2012). When the promises of capital gains were not corroborated by end-of-year financial results, adaptations in the valuation convention (from land bank to launches and M&A) were necessary to preserve the dominant narrative for growth.

This was also accompanied by a shift in power relations. As concern grew with disappointing financial results, fund managers and analysts increasingly pressed managers to take a series of strategic initiatives to ensure stronger growth. Pressure was exerted through specialists’ reports, during quarterly teleconferences, and through business media interviews. It was enacted through more benevolent pricing of shares for developers that most eagerly complied with the proposed managerial evolutions, whereas the shares of more conservative firms were underpriced:

at the time of the IPOs, […] investors wanted everyone to spread, to diversify [investment] across the country, to raise launching projections, and we had a more conservative profile. We even believe our stocks have a […] lower valuation [than our competitors] due to our management’s more conservative strategy. (Interview no. 1, Investor relations manager, Even, November 2011)

If initially housing developers and underwriters attempted to lure investors into an interpretation convention revolving around the land bottleneck problem, the interactions evolved with the stock market community progressively exerting a stronger grip on developers, disciplining them through the use of a valuation convention focusing on profits (Oliveira, 2008b).

Promises and numbers

By 2010–2011, with developer financial performance numbers stubbornly continuing to diverge from the promises underlying narratives for growth, the optimistic assumptions on which share prices had been originally estimated were becoming increasingly discredited. Multiple reasons explain why profitability for shareholders turned out to be weaker than expected. First, construction costs that had been accounted at current value by underwriters rose because of an increase in labour costs, itself an outcome of diminishing unemployment rates resulting from economic growth:

firms […] launched a lot, and because they have a turnover of 3 years, in general they build in the last two […]. The problem is that, […] by the time firms had to build […] in 2009, 2010, 2011, input prices were much higher, with a labor shortage. […] So in the projects we sold, the margin we’d expected to be 36% or 38%, dropped to 28%. […] And there are firms that had worse results […]. Investors got angry with firms. (Interview no. 2, Project manager, Cyrela, March 2012)

Second, confounding business media and analysts expectations, business strategies did not yield improved margins as the increased number of partners pushed coordination costs up; economies of scale were disappointing because of the learning curve, especially in new ‘segments’ and markets; poor execution is said to have resulted in multiple delays, which, in turn, increased financing costs:

[firms’] focus in the first years was on scaling up. Investors began putting a premium on […] rising projections of Potential Sales Value (PSV), which indicate the potential of new projects in terms of revenue generation. ‘Since [the stock] market was buying launches, many partnerships were hastily set up’, says Christian Faricelli, equity manager at Capitânia. According to the expert, [firms’] aim was to diversify regionally, but they lost control [of the operation]. ‘There was a very fast and a bit haphazard growth by most firms’, says Faricelli. (Tauhata, 2012)

Disappointing financial figures led to dissonant views over responsibilities. Developers (particularly those reluctant to adopt aggressive growth strategies and those that stayed away from the stock market) became more vocal at denouncing the gregarious investors and analysts that forced them to pursue misguided objectives while also being volatile in their expectations and eager to follow ‘fads’ (Investor Relations manager, Even, November 2011). In an interview for Exame, Cyrela’s chairman Ellie Horn was asked how he felt about the stock market’s negative reactions to the firm’s recent report in 2009:

[…] The first time they [analysts] talk about your firm you follow them. Today I don’t pay much attention, because otherwise I’d destroy the firm. We need to look at the [stock] market’s reactions with sound judgment so that the company is preserved. (Correa, 2009)

Some analysts, probably reflecting investors’ views, blamed instead developers on their inability to properly execute their projects. Thus Marcelo Motta, an analyst for JP Morgan, declared that:

Many firms ended up focusing only on growth and paid no attention to the matter of profitability. […] Most of the firms […] with falling profits suffered due to a lack of planning. Cost overruns, penalties incurred on projects, allowances [for bad debts] in balance sheets and other factors drove down profits in 2011. (Corsini, 2012)

Other analysts struck a more balanced tone, relating developers’ rash moves to the use of inadequate criteria for assessing firms by fund managers. A Banco do Brasil analyst confided that investors had, until 2011, failed to understand the cycles of building and their impact on the funding needs of firms, while acknowledging that as investors learned how the sector works, the focus on growth alone was gradually changing (Corsini, 2012).

Such dissonances led to general disenchantment and to the unravelling of the prevailing convention around which these actors had initially coalesced. This unravelling made way for new competing narratives that progressively started to vibrate in unison again and to resonate into an interpretation convention where quality and profitability substitutes quantity and turnover. Concomitantly, and mutually reinforcing each other, this led to an evolution in the valuation convention as well, with fund managers and analysts seeking different variables to assess the value of firms:

Given the difficulties experienced by most firms, investors and analysts shifted focus of their demands. It is no longer expected a high volume of launchings, but instead priority is given to projects that are more profitable. Positive cash flow and enhanced profitability now dominate teleconferences. (Corsini, 2013)

Illustrative of this shift in focus, the business press has thrown a spotlight on firms whose strategies have been based on slower-paced expansion through organic growth, be they listed on the stock exchange, such as Eztec (Barra, 2014), or not, such as CFL (Bueno, 2013).

As in the 2005–2010 period, this evolution in stock market assessment had performative effects upon developers. If Brookfield bought back its shares to remove itself from the grip of financial markets (Rostas, 2014), other developers remained in the game by claiming in their reports and in public announcements to be taking strong actions to enhance profitability. These include cutting back on new projects to better control costs and construction schedules (Quintão, 2012); reducing the number of partnerships; retreating from some regional markets (Pereira, 2012); and abandoning ‘segments’ where firms had less expertise, such as lower-income housing (Fernandes, 2012).

[the company has decided to] slow down. We are not going to grow as much as we did [in the past]. Our pace of growth was reduced. So Cyrela informed the [stock] market in March 2011 that growth would not be at 30% a year, but 10–15%. […]. We [also] reviewed [our plans for regional expansion]. We always sought to expand where incomes are, in cities with at least 500,000 inhabitants, etc. In 2008, we had 14 partnerships. Today we have [only] 4. In 2010 and 2011 we limited [these partnerships]. When construction costs went up, we thought: where did it happen? [It’d happened] where we had the least control, […] [that is] with partners outside São Paulo, Rio, and the South. […] So we decided to strengthen our activity where our execution is better. […] This all [aimed at improving] cash flow. (Interview no. 2, Project manager, Cyrela, March 2012)

With such evolutions, a new round of spatial consequences is likely to occur in the production of housing in Brazilian cities. As developers seek to improve profitability, their concentration in selected metropolitan areas in the Southeast and South will rise, leaving the other markets again for smaller, family-owned firms. For the same reason, their projects are likely to be more focused on middle- and higher-income groups. There are also signs that firms will concentrate their efforts in large and very large projects as a way of enhancing margins through scale economies (Gazzoni, 2013) and, maybe, to increase the profitability of their land bank. This approach will prove sustainable as long as the financial outputs corroborate their strategies, or until other narratives resonate into another convention on how best to satisfy financial markets’ pecuniary return expectations through the production of homes for Brazilian households.

Conclusion

Amid the multifarious factors at play in shaping the geographies of housing, the paper set out to take stock of the recent evolutions affecting the development industry, especially in ‘emerging’ countries. With neoliberal reforms, alongside, and more often than not, above the growing importance of mortgage and of mortgage securitisation, the provision of housing and their associated geographies is transformed with the rising importance of financial investors in the development sector. This paper claims that, in order to understand the geographies of housing production in Brazilian cities, it is necessary to analyse the interactions between financial markets and development firms. This is done by an empirical analysis of Brazil’s housing sector between the mid 2000s and 2012.

The paper combines two strands of literature applied to the study of financial markets (cultural economy and conventionalist economics) by developing the concept of resonance. This concept permits to recognise the pressure exerted by financial market players, but also the agency of development firms’ management. It also enables us to straddle the micro- and meso-levels by dynamically linking the interactions happening around each individual organisation with the conventions that are made and remade on the development industry.

This heuristic demonstrates that there is no straightforward, unidirectional relation leading from the expectations of financial markets to the actually existing geographies of housing. Instead it argues for an in-depth analysis of the narratives co-authored by developers and the stock market community (investors, analysts, business media, etc.) as well as the related strategic moves that transform the business practices of developers and, arguably, the geographies of housing provision. Indeed, when idiosyncratic interactions between shareholders and firm managers start vibrating in unison, interpretation and valuation conventions contribute to converging business strategies that have consequences on the production of housing. This is illustrated by the 2005–2010 period, in which a dominant narrative around growth stimulated firms to pursue more aggressive development plans. Throughout this period, developers prioritised large-scale projects, targeted a wider range of income groups, and replicated such developments across a larger number of cities. This resulted in a more fragmented pattern of housing production, as many such projects assumed the form of urban enclaves.

This offers two insights into the relations between financial markets and developers. First, financial market conventions are not predetermined but vary over time in response to observed financial results. Disappointing numbers are accommodated through evolutions in the dominant convention, especially as investors are discovering a new sector, as was the case with the development industry. Yet, these adaptations occur only up to a point: dissonant narratives may break up the harmony, with actors blaming each other. Second, and relatedly, power relations between actors evolve over time. If developers initially transformed their land bank bottleneck into an investment opportunity, the stock market community, and in particular fund managers, gradually tightened their grip on developers, attempting to force them into business strategies and rewarding – or sanctioning them – through share pricing. Interactions are thus often fraught with frictions, which are particularly heated when discrepancies between the expectations that the convention originally embodied and the actual results arise. Additionally, as a new convention progressively substitutes the other, the repercussions on developers’ business strategies are likely to once again transform the geographies of housing.

At a more theoretical level, and without opposing the financing of homeownership (namely through mortgages and their securitisation) and the financing of development activity, we have argued a need to more directly take into account how financialisation affects the provision of housing by developers.

This is important both in countries of the Global South and Global North, since the development industry entertains multiple ties with financial markets. As we have seen in this paper, financial markets directly provide equity and debt to developers at corporate level. But they are also increasingly involved in the direct ownership of properties (see Fields and Uffer, 2016; Guironnet et al., 2016; Halbert et al., 2014) and in the financing of development operations (direct investment at project level). These evolutions press for further analyses on the relation between financialisation and the production of the urban built environment from the perspective of the supply-side.