Geographies of the financial crisis

Authors


Abstract

Real estate is, by definition, local as it is spatially fixed. Mortgage lending, however, has developed from a local to a national market and is increasingly a global market today. An understanding of the financial crisis is ultimately a spatialised understanding of the linkages between local and global. This article looks at the geographies of the mortgage crisis and credit crunch and asks the question: how are different places affected by the crisis? The article looks at different states, different cities, different neighbourhoods and different financial centres. Investors in many places had invested in residential mortgage backed securities and have seen their value drop. Housing bubbles, faltering economies and regulation together have shaped the geography of the financial crisis on the state and city level in the US. Subprime and predatory lending have affected low-income and minority communities more than others and we therefore not only see a concentration of foreclosures in certain cities, but also in certain neighbourhoods. On an international level, the long-term economical and political consequences of this are still mostly unknown, but it is clear that some financial centres in Asia (including the Middle East) will become more important now that globalisation is coming full circle. This article does not present new empirical research, but brings together work from different literatures that all in some way have a specific angle on the financial crisis. The aim of this article is to make the geographical dimensions of the financial crisis understandable to geographers that are not specialists in all – or even any – of these literatures, so that they can comprehend the spatialisation of this crisis.

Introduction

A mortgage is money lent on the security of property owned by the borrower, usually in order to enable the borrower to buy property. When we speak of mortgages today, we generally refer to home mortgages: mortgages issued on residential real estate. Real estate by its very nature is local as it is spatially fixed. Mortgage lending, however, has developed from a local to a national market and is increasingly a global market today. An understanding of the mortgage crisis and the credit crunch is ultimately a spatialised understanding of the linkages between local and global, between the space of places and the flow of spaces. My aim here is not to come to a geographical understanding of the linkages between local real estate, national mortgage lenders and international markets for residential mortgage backed securities. Others have provided the building blocks for such an approach (most notably Gotham 2006, 2009; see also Dymski 1999; Sassen 2001, 2009; Wyly et al. 2004, 2009; Aalbers 2008, 2009; Wainwright 2009). Instead, this article looks at the geographies of the mortgage crisis and credit crunch and asks the question: how are different places affected by the crisis? In doing this I look at different states, different cities, different neighbourhoods and different financial centres. In four different takes, this article describes what the crisis is currently doing to these places: the first explains how mortgages in the US are connected to global financial markets, the second and third look at the geography of the ‘default and foreclosure crisis’ in the US (the second take focusing on states and cities, the third on neighbourhoods) and the fourth take looks at the global reshuffling of money and financial centres. This article does not present new empirical research, but brings together work from different literatures that all in some way provide a specific take on the financial crisis. The aim of this article is to make the geographical dimensions of the financial crisis understandable to geographers that are not specialists in all – or even any – of these literatures, so that the spatialisation of the crisis becomes tangible.

What's Norway got to do with it?

The current financial crisis originates in the housing and mortgage markets, but it affects financial markets around the world. A few decades ago most mortgage lenders were local or regional institutions. Today, most mortgage lenders are national lenders who tap into the global credit market. This is not so much the case because lenders are global financial institutions – most lenders are national in scope – but because they compete for the same credit in a global market. Before the financial crisis of the late 1980s, savings and loans institutions (S&Ls) granted loans based on the savings that got into the bank. Generally speaking, the savings and loans were made in the same geographical market. The fact that the S&Ls only worked in local markets was seen as a problem: what if the savings are available in one area, but loans are needed in another?; and what if a local housing market busts? The solution was to connect local markets and thereby to spread risk. Interest rates on loans would fall because there was now a more efficient market for the demand and supply of money and credit.

The trend from local to national mortgage lenders was one thing, but, it was argued, mortgage markets could be even more efficient if they were connected to other financial markets and not just to savings. In the wider credit market it would be easy for mortgage lenders to get money as mortgages were considered low risk. Mortgages would be an ideal investment for low-risk investors. Cheaper credit, in return, would lower interest rates on mortgage loans. Securitisation was already introduced in the 1960s by Fannie Mae and Freddie Mac, two government-sponsored enterprises that were meant to spur homeownership rates for low- and middle-income households. Securitisation enables mortgage lenders to sell their mortgage portfolio on the secondary mortgage markets to investors. Following the S&L crisis, deregulation favoured securitisation, not only through Fannie Mae and Freddie Mac, but also through so-called ‘private labels’ (Gotham 2006; Immergluck 2009).

Securitisation also meant that mortgage lenders could work according to a new business model whereby mortgages are taken off balance. This frees up more equity for more loans. It also enabled non-banks to enter the mortgage market. Mortgage portfolios could now be sold to investors anywhere in the world and because these investors thought mortgage portfolios were low risk and there was a lot of money waiting to be invested, especially after the dot-com bubble crash (2000–2), there was a great appetite for such residential mortgage-backed securities (RMBS). In other words, the S&L crisis, the following bank merger wave (Dymski 1999), securitisation, the entry of non-bank lenders and the demand for low-risk investments together shaped the globalisation and financialisation of mortgage markets.

The credit crisis started in 2007 when foreclosure and default rates went up and housing prices went down. This implied that investing in mortgages was not as low risk as people thought. The value of RMBS fell even more dramatically. This is not only the case because many people had mortgage loans that were granted without down payments, but also because RMBS were sold with the idea of high returns. These high returns were partly based on high interest rates and not just on the value of the house, and partly on speculation, which increased the value of RMBS beyond what they were actually worth. In sum, not only were risks underestimated, returns were also overestimated (even if there hadn't been rising defaults). In addition, even though housing prices on average fell by 20 per cent, the impact on the RMBS market was much bigger. This is not just a result of inflationary prices, but also of leveraging. Major players in the RMBS market like investment banks basically invested with borrowed money (ratios of 1:20 were not uncommon, 1:14 being the average) and because of this leveraging both profits and losses would be disproportionally big.

It now becomes easier to understand why the impact of partly local and partly national problems in housing and mortgage markets (I will return to this in the next two sections) is global in scope and also affects other credit markets. The crisis does not just hit investment banks on Wall Street, European banks and pension funds that bought RMBS, but also individual investors and small towns around the globe. The example of Narvik in the far north of Norway is widely discussed. The city council of Narvik (population 18 000) and three small, nearby municipalities had invested $78 million of the revenues of a nearby hydroelectric plant in RMBS and lost most of it. The city of Narvik was advised by the Oslo-based firm Terra Securities to buy RMBS that were put together by US bank Citigroup. Terra had not explained these products to the city and most city councillors actually thought their revenues were being invested in low-risk Norwegian funds. The city's investments were meant for the construction of a new school, a nursing home and a child-care facility. Instead, the city has cut the budget and as a result several small rural schools will be closed, budgets for elderly care have been cut, the city is behind with payments to civil servants, and the fire department will cease their 24/7 service and will switch to daytime service (in a city with mostly wooden houses). The Norwegian state has declared it will not help Narvik and other municipalities, as it does not want to set a precedent by which the national state has to bear the losses of local authorities.

Unfortunately Narvik is not the exception as municipalities around the world had invested in RMBS or the financial institutions that went down in this crisis such as the American Lehman Brothers or the Icelandic Landsbanki. Narvik illustrates well how connected the world has become in the twenty-first century. It also illustrates well the world is not flat today: the old geography of local housing markets has not been replaced by a global housing market, but by a chain that starts with local (a mortgage loan on a particular property), turns national (through lenders), then global (in the RMBS market) and then local again (by the effects in places like Narvik). Nearly everyone in North America, Europe and Australia, and many in Asia are in some way involved in this crisis, often as passive investors, e.g. through pension funds or investments by their municipalities. We may not all be capitalists now, but most of us are investors, whether we want – or know – it or not.

From the Rustbelt to the Sunbelt

Subprime mortgage lending has been growing fast, from about $35 billion (5% of total mortgage originations) in 1994 to $600 billion (20%) in 2006 (Avery et al. 2006), 75 per cent of which is securitised. In some states like Nevada, subprime loans accounted for more than 30 per cent of the loans originated in 2006. Subprime lending is often defined as lending to a borrower with poor credit, but this would be a misrepresentation of the essence of subprime lending, which is lending at higher fees and interest rates whether or not the borrower actually has bad credit. Some estimates suggest that more than half of the subprime loans went to prime borrowers (Brooks and Simon 2007). In 2006, 13 per cent of outstanding loans were subprime, but 60 per cent of the loans in foreclosure were subprime, up from 30 per cent in 2003 (Nassar 2007). Recent subprime loans not only go into default and foreclosure more often, but also much faster.

The default and foreclosure crisis that was at the origins of the credit crunch has hit American households across the country, but people in some states and cities are more likely to be in foreclosure. The rise in default rate started some years ago in the Rustbelt where housing prices went down and unemployment up. The combination of lack of employment and falling housing prices is perilous as people who lose their job in a high unemployment area not only have a smaller chance of finding a new job within a few months, but they also run a bigger chance of not being able to pay off their mortgage loan and might then be faced with negative equity. This implies that banks will not only see higher foreclosures as a direct result of default, but also because homeowners with financial problems in declining housing markets are less likely to sell their house which would have enabled them to pay off their loan.

The rise of subprime lending started in the early 1990s with refinance loans, often in the poorer parts of Rustbelt cities (see next section). Increasing default rates led to a first subprime mortgage crisis in 1997–8, 10 years before the second subprime crisis. Party as a result of the crisis, many specialised subprime lenders were acquired by general banks. The growth of subprime lending halted for a few years, but picked up again after 2000 when subprime loans were no longer exclusively targeted at borrowers with low credit scores; ‘exotic’ mortgages were designed to be sold to middle class borrowers, in particular in rapidly growing parts of the country (Ashton 2009). As a result, the fastest increases in defaults and foreclosures in recent years were not in the Rustbelt but in the Sunbelt where housing prices had been going up most rapidly and exotic mortgages with adjustable interest rates and teaser rates were more common.

Compared with fixed-rate mortgages, adjustable rate mortgages (ARMs) are more than twice as likely to default (Schloemer et al. 2006). Since most of these ARMs were originated in 2005 and 2006 with 2 or 3-year low-interest teaser rates (so-called 2/28 and 3/27 loans), their interest rates were going up rapidly in 2007 and 2008, with even higher defaults and foreclosures as a result. It is estimated that about $1 trillion in ARMs were subject to resetting interest rates in 2007 – more than 10 times as much as in 2005. The loan sum of almost a quarter of the ARMs was bigger than the value of the house (Cagan 2007). This is already problematic, but becomes very dangerous when interest rates go up and housing prices down. Cagan shows that with housing price declines of 10 per cent, almost half of the ARMs would have negative equity. In fact, just 2 years after the top in 2006, housing prices have fallen by 20 per cent nationwide, 30 per cent in big cities like Las Vegas, Miami and Phoenix and even more in some smaller cities.

A few years ago, the top 10 of foreclosure cities almost exclusively consisted of Rustbelt cities. In 2007, when the crisis started, the list was a mix of Rustbelt and Sunbelt cities, but since the summer of 2008, the top 10 of foreclosure cities is entirely made up of Sunbelt cities. There are several of such foreclosure lists and they look slightly different, but generally speaking cities in California now make up more than half of them, with cities in the South and Florida completing the list. The differences across the US are huge: the foreclosure rate in Richmond, VA is almost 100 times as low as in Stockton, CA, the foreclosure capital, as about 5 per cent of Stockton's homes are currently in foreclosure.

Housing prices can go down because of a structurally faltering economy, like in the Rustbelt, but also because they have been going up extremely fast, like in many cities in the Sunbelt. Housing prices in the Sunbelt were simply more inflated than elsewhere in the US: the housing bubble was bigger and more likely to burst. In addition, some local and regional economies in the Sunbelt also show signs of a declining economy, perhaps not structurally, as in the Rustbelt, but conjuncturally. Finally, high economic growth also meant a lot of new construction and more homeowners who recently bought a house, thereby increasing the pool of possible victims of falling housing prices.

A weak economy and a housing bubble together explain a great deal of the variation in default and foreclosure rates across the US, but one other factor is also important: regulation. Lately, we have read a great deal about how regulators have failed to see what could go wrong in the mortgage market. This is partly true: many regulatory agencies have not done a lot to prevent the current crisis. Yet, it is also the deregulation and re-regulation of mortgage and credit markets that have enabled both securitisation and subprime lending. Rather than turning a blind eye, the re-regulation of the mortgage market has been a political project and, as such, part of a wider neoliberalisation of states and markets alike. The outcomes are partly an intended consequence of continuous negotiation between lenders, securitisers, politicians and regulators, and partly an unintended consequence of related legislation. One important outcome is that non-bank lenders, who provide the lion's share of mortgage loans, face a dramatically weaker regulatory regime than bank lenders. This is not the place to review how regulation and deregulation have enabled, and indeed stimulated, securitisation and subprime lending – other have done this job in more detail (most notably Gotham 2006; Immergluck 2009; see also Squires 2004; Dymski 2007; Ashton 2008). Here it is important to note that because of failing federal regulation of mortgage markets, several states in the US have geared up their own regulations to impede some of the many excesses of subprime lending. As Wyly et al. (2009) show, states like North Carolina, New Mexico, Massachusetts and West Virginia regulate a wide range of practices, related to foreclosure rules, loan flipping, prepayment penalties amongst others. New Mexico, for example, introduced the Home Loan Protection Act (2003).

Not only has the level of regulation varied from state to state, but also the quality of oversight. In many cases local and state initiatives, however, were blocked by state and federal institutions. North Carolina passed tougher predatory lending laws but faced opposition from two federal agencies (pre-emption). Since 2000 Cuyahoga Country, Ohio, which includes Cleveland, has seen 80 000 foreclosures on less than 400 000 properties (Simon 2008). The city of Cleveland, one of the early Rustbelt victims, replied by introducing local measures to make predatory lending more difficult, but the state of Ohio – heavily lobbied by Ohio banks (Hirsch 2008) – deemed these illegal as authority lay with the governor and not the city (pre-emption).

The geography of predatory lending

The financial crisis is not simply a result of the growing number of defaults and foreclosures – these are symptoms of the crisis. The crisis originates in a subset of subprime lending known as predatory lending. Predatory loans are designed to exploit vulnerable and unsophisticated borrowers. A predatory loan has one or more of the following features (NCRC 2002, 4; cited in Squires 2004):

  • 1higher interest and fees than is required to cover the added risk of lending to borrowers with credit imperfections;
  • 2abusive terms and conditions that trap borrowers and lead to increased indebtedness;
  • 3fails to take into account the borrower's ability to repay the loan;
  • 4violates fair lending laws by targeting women, minorities and communities of colour.

Most subprime and predatory loans are granted to people to refinance their mortgages or as second mortgages. This means that most of these loans do not enable homeownership among low-income and minority groups, as is often argued. Subprime, and in particular predatory, loans are targeted at low-income and minority populations, often living in areas with high unemployment and declining housing values (Hernandez 2009; Newman 2009; Pennington-Cross 2002; Squires 2004; Wyly et al. 2009). Even after controlling for borrower and lender characteristics, African–Americans relatively speaking receive more than twice as many high-priced loans as Whites (Schloemer et al. 2006). Almost half of the loans in minority areas are high-priced compared with 22 per cent in predominantly white areas (Avery et al. 2007); 20 per cent of all loans in minority areas are classified as ‘high risk’ compared with only 4 per cent in white areas (Bromley et al. 2008).

Many of these communities were ‘underserved’ as a result of earlier waves of redlining, the place-based exclusion of financial services. Decades of financial deregulation have not resulted in wider access to mainstream financial services, but in a two-tier banking system with mainstream finance in most places next to a landscape of financial exclusion and predatory lending where banking services and the number of bank accounts have declined and fringe banking (pawn shops, payday lenders etc.) and predatory lending flourishes (Caskey 1994; Dymski 1999; Immergluck 2009; Leyshon and Thrift 1997; Squires 2004). Both quantitative and qualitative research shows that ‘subprime loans are making credit available in communities where credit likely historically has not been – and likely still is not – as readily available’ (Goldstein 2004, 40). The old geography of redlining and financial exclusion has not disappeared, but has been replaced – and to a large extent reproduced – by a new geography of predatory lending and overinclusion. This is why Squires (2004) refers to predatory lending as ‘the new redlining’. In line with Harvey (1985), we can see how, through subprime lending, the urban has become the place of capital extraction (Wyly et al. 2004; Newman 2009).

Predatory lenders make profit ‘by stripping equity and wealth from home owners in underserved communities through highcost refinance loans’ (Taylor et al. 2004, 27). Teaser rates and other tricks are used to sell these loans. Rapidly increasing interest rates and balloon payments are used to increase returns, and in some cases also to increase the likeliness of default. Repeated default then allows lenders to repossess homes and acquire equity. In this way lenders do not enable homeownership but effectively strip home equity from borrowers. This frequently leads to mortgage foreclosures at the individual level and housing abandonment at the neighbourhood level. It is not just defaulting borrowers who are hit; in addition, there are severe spill-over effects on housing prices, crime and neighbourhood decline (Immergluck 2009; Lardner 2008; Newman 2009). Areas that are over 80 per cent minority have more than five times as many foreclosures as areas that are less than 10 per cent minority (Smith 2008). Although some cities in the Sunbelt are now hit harder than those in the Rustbelt, on a neighbourhood level the Rustbelt still tops the foreclosure lists. On the list of most foreclosed zip codes, four are now in Detroit; the Slavic Village in Cleveland has the most foreclosure filings. This neighbourhood was hit hard by redlining and suburbanisation from the 1950s onwards. Due to a combined economic and foreclosure crisis, demand for housing has fallen so dramatically, that one can now buy many homes in the Slavic Village that are priced under $30 000; on eBay you could even buy one for less than $5000. The Slavic Village, now referred to as foreclosure's ground zero, has seen a rapid increase in crime (Christie 2007).

Besides borrowers and neighbourhoods, municipalities are also hit hard because tax income goes down due to foreclosed properties and lower real estate prices, while expenses are increasing as a result of foreclosures and property crime. Local governments around the country have cut expenses on education, infrastructure and social services. The public school system in California alone faces a loss of $4 billion in funding. Foreclosure rates are now so high that we may expect three million foreclosures by the end of 2008. If housing prices keep on falling, job losses increase and inflation keeps on rising, 2009 may be even worse, especially since very little of the $700 billion plan is designated to help defaulting homeowners from being foreclosed on. The American Housing Rescue and Foreclosure Prevention Act (2008) will probably help up to 500 000 homeowners (most of whom are not in default), but since foreclosures for 2007–9 may add up to 8 million, this is far too little.

The next Wall Street

Just as the financial crisis of the late 1980s led to a bank merger and acquisition (M&A) wave (Dymski 1999), the present crisis will also lead to a new M&A wave. So far, much of the M&A activity is government initiated: at the time of writing (September/October 2008) banks and other financial institutions are collapsing, get bailed out, bought up and nationalised on a daily basis. We can see how some big financial conglomerates are growing even bigger, partly as a result of government-engineered M&As. JP Morgan Chase and Bank of America seem to be two of the big banks that become even bigger, but outside the US, we can also see M&A activity. Banco Santander, Spain's biggest bank, for example, has been buying national and foreign banks for a number of years, but is now increasing its activity with the acquisition of Sovereign Bancorp (US), former building society Alliance & Leicester (UK) and parts of savings bank Bradford & Bingley (UK). The credit crunch will likely force more financial institutions into M&As, not just nationally, but also internationally.

Another consequence of this crisis is a shift in who owns financial institutions. For example, the Abu Dhabi Investment Authority, a sovereign wealth fund, is now the largest shareholder of Citigroup; Prince Alwaleed Bin Talal Al Saud of Kingdom Holding of Saudi Arabia the second largest. In the past, crises in ‘the West’ often had severe consequences for ‘the Rest’ as investors would usually flee and re-concentrate on the West. Oil-exporting countries as well as the so-called BRIC countries (Brazil, Russia, India and China) are all hit by this crisis, but this crisis has generally hit the West more than the Rest. Brazil is said to have $200 billion in reserves, China $1800 billion. Even though many of these countries are still export dependent, they are now less export dependent than they were 10 or 20 years ago, as their national markets have grown and trade between non-Western countries is also on the rise.

As a result, we may also see a shift in the dominance of financial centres. The financial crisis does not directly lead to the fall of Wall Street (New York) and The City (London), but it does accelerate the trend towards a shift in financial centres. There are now more secondary financial centres in the world and the centres of increasing importance are to be found outside Northern America and Europe. As the German finance minister, Peer Steinbrück, recently said: ‘This world will become multi-polar’ (cited in Benoit 2008). He expects the emergence of stronger, better capitalised centres in Asia and Europe, but considering how much Europe is hit by this crisis and how fast Asia's economies are growing and changing, a shift towards Asia seems more likely. As Asian economies grow and their current account surpluses grow alongside them, Asian money becomes more and more important globally. While the US over the last 10 years had an average current account deficit of $800 billion, several East Asian countries together had an average $400 billion surplus, more than a third of the world's surplus. China and Japan have the largest surpluses in the world (currently $372 billion and $213 billion). Other countries with big surpluses (in absolute rather than relative terms) are European countries like Germany, Switzerland, Norway and the Netherlands; the Middle Eastern states of Saudi Arabia, Kuwait and the United Arab Emirates; and Eurasian giant Russia. The chief economics editor of the Financial Times, Martin Wolf (2008), suggests that

among the most important tasks ahead is to create a system of global finance that allows a more balanced world economy, with excess savings being turned into either high-return investment or consumption by the world's poor, including in capital-exporting countries such as China. A part of the answer will be the development of local-currency finance in emerging countries, which would make it easier for them to run current account deficits than proved to be the case in the past three decades.

Related to this, Thaksin Shinawatra (2008), the former prime minister of Thailand, recently proposed to construct a basket of Asian currencies which could be used for packing Asian bonds and then be traded on the financial markets of Tokyo, Singapore and Hong Kong. Asian economies may increasingly want to turn their current account surpluses into profitable investments while at the same time withdrawing from the US dollar as the global reserve currency. Many Asian currencies may be considered too volatile and in some cases too small, but a basket of currencies – in some ways like the Euro started out when it was still the European Currency Unit (ECU) – would spread volatility risk, disengage from the risky US dollar, and shift power from those on the receiving end to those on the giving end. In addition, stronger Asian currencies, whether weighed in one basket or not, would allow Asian customers to buy goods from outside their own country more easily, which in return would benefit other economies. The increased role of Asian investment, combined with a rise in the importance of Asian currencies, will help to increase the role of Asian financial centres.

Many in the financial sector now consider Chung Wan/Central (Hong Kong), not Marunouchi (Tokyo), the runner-up behind The City and Wall Street. But the rise of Hong Kong is contested by other fast-growing financial centres such as DIFC (Dubai), Lujiazui/Pudong (Shanghai), Bandra-Kurla (Mumbai), and Shenton Way/Raffles Place (Singapore). Nevertheless financial centres do not simply pop up where the money comes from – the availability of skilled personnel and the regulatory environment are at least as important as the access to financial markets (Mainelli 2006). The stability of a country and its political regime are also crucial. Therefore, Singapore and Hong Kong (with its special status) are better positioned than Mumbai and Shanghai. In addition, quality of life and cultural factors also play a role: London and New York are not just global financial centres, they are also global cultural centres. Most of the financial centres in Asia are merely national cultural centres. Dubai may be investing heavily in high culture; it lacks the cultural networks of cities like London and New York. Shanghai and Tokyo may represent important markets, but these cities may be too Chinese and Japanese respectively to become expatriate magnets. Again, Singapore is positioned better. Much will depend on which cities the Chinese and Arabs will accept as their offshore market and which cities will be accepted by international financial conglomerates and workers as attractive financial centres. In the coming years we will see a continued dominance of London and New York and a number of rising stars in Asia. In time, one or two of these cities may be able to rival the big two.

But the geography of financial centres is not merely a global geography, it is also has a local geography as the case of New York illustrates well. Publications such as the New York Times and New York Magazine had headlines like ‘Wall Street, R.I.P.’ (Creswell and White 2008) and, with a reference to Tom Wolfe, ‘Good-bye, Masters of the Universe’ (Cramer 2008). In a city where 20 per cent of personal income tax and 45 per cent of business income tax come from Wall Street and many others are dependent on Wall Street employees’ spending, the crisis has its own geographies. Almost a quarter of the 188 000 Wall Street jobs may be lost and since every Wall Street job is said to support two others in the city, the loss of jobs may be dramatic (Gapper 2008). Prices for huge mansions in the rich suburb of Greenwich, CT – the hedge fund capital – have fallen dramatically: the New Yorker presented examples of mansions that were priced $25 million and sold for $8 million (Paumgarten 2008). Twenty-five miles south, 15 Central Park West, the most expensive recent condo development in Manhattan, accommodates mostly Wall Street people, including a dozen investment bankers. They are not expected to sell – many of them paid cash thanks to the huge bonuses of the past years (totalling $33 billion in 2007) – but further price increases are considered unlikely and the small but very lucrative market for eight-figure apartments is drying up. Not only the financial services sector, but also anyone from luxury retailers to restaurants and from nannies to hotels is already seeing the impact: one high-end massage therapist for example lost 50 per cent of her Wall Street clientele (Dominus 2008). Cornell Medical College received $250 million from Citigroup in 2007, the New York Public Library $100 million from private equity group Blackstone (Gapper 2008) – in the light of the current crisis such contributions can only go down.

Conclusion

The financial crisis is redrawing the world in lots of ways and at many levels. Financial globalisation's impacts are logically speaking global, but that doesn't mean the impacts are the same around the globe. Investors in many places had invested in RMBS and have seen the value of their securities drop. This has affected both financial institutions and non-financial institutions. There seem to be few financial institutions that remain unaffected by this, but even if they were not involved in RMBS, they now feel the pains of the credit crunch. The impact on non-financial institutions has been more spread and will, for example, affect some towns in Norway. Housing bubbles, faltering economies and regulation together have shaped the geography of the financial crisis on the state and city level in the US. Subprime and predatory lending have affected low-income and minority communities more than others and we therefore not only see a concentration of foreclosures in certain cities but also in certain neighbourhoods, often those places inhabited by low-income and minority groups that have been excluded by earlier rounds of exclusion and exploitation. Locally, the impact is also felt in New York, a city that is heavily dependent on its financial sector. Finally, we see how investors from outside Northern America and Europe come to the rescue by not selling stocks in financial firms when others do – and by even investing additional billions of dollars, thereby slowly but surely expanding their reach. Increasingly, stocks in financial firms are held by investors outside the developed world. After decades – or one could argue, centuries – of uneven globalisation, we can witness the rise of what we could call ‘reversed globalisation’ or ‘globalisation has come full circle’ as investments and returns are now moving in both directions. Structurally increasing oil prices (despite temporarily declining oil prices) and the rise of many emerging economies, two of them backed by billion-people populations, have contributed to a shifting global power balance. The long-term economical and political consequences of this are still unknown, but it is clear that some financial centres in Asia will become more important now that globalisation is coming full circle. Yet, the meanings of globalisation, not unlike the causes and consequences of this crisis, remain geographically uneven. It is important to understand that geography is an essential element in both, and that the fates of places like Stockton and Narvik are not only related to each other, but also to those of Wall Street and Raffles Place. The space of places is intrinsically linked to the flow of spaces.

Acknowledgements

The author would like to thank the referees as well as Ewald Engelen, Rodrigo Fernandez and Martijn Konings for their comments on an earlier draft of this article. The usual disclaimer applies.

Ancillary