Transnational corporations as financial groups

. ABSTRACT Whilst the 2008 crisis provided evidence of the strong (not to say devastating) interrelations between production and finance, giving a boost to the ‘financialisation’ approach, there is a need, forty years after the initial research on transnational corporations (TNCs), to re-explore the nature of these organisations. Issues that require investigation include the reshaping of international trade and production, the close interactions between non-financial TNCs and financial (bank and non-bank) TNCs, the development of global networks, and the strength of the relationships entertained by most of these companies with ‘their’ governments. A basic hypothesis of this paper, which is focused on non-financial TNCs, is that they cannot only be defined by the fact that they are bigger and more internationalised than other firms. The paper argues that, on the contrary, they constitute a category of their own, based upon a centralisation of financial assets and a specific organisational structure (with the core role held by the holding company). TNCs, organised and structured as groups of enterprises, are a locus for a global valorisation of capital, where productive and financial valorisation are closely intertwined. In the context of a global macro-economic regime of accumulation dominated by finance capital, financial logic takes on a preeminent role in the strategies of TNCs. Unfettered deregulation of financial markets and a multiplication of financial innovations (products and institutions) have given a further boost to the transformation of TNCs, which can be defined as financial groups with industrial activities.


Introduction: TNCs in a category of their own
The considerable weight of large non-financial 1 TNCs in the world economy cannot be denied.According to UNCTAD, in 2008 there were 82,000 transnational corporations with a total of 820,000 foreign affiliates.Their aggregate activity was immense.The gross product of foreign affiliates worldwide accounted for 10.5% of global gross 1 I use the term 'non-financial' in the meaning given by the System of National Accounts (SNA, 2008, 4.62): 'Nonfinancial corporations are corporations whose principal activity is the production of market goods or nonfinancial services' .domestic product (GDP) in 2009, compared with only 6.6% of GDP in 1990. In 2009, total sales by foreign affiliates amounted to US$ 29.2 trillion (UNCTAD, 2010).
At the heart of this large web can be found the top 100 TNCs which account for 10% of foreign assets, 16% of foreign sales and 12% of the foreign employment of the total 82,000 TNCs (author's computation from UNCTAD, 2010).The prosperity of these TNCs, despite the current crisis, is equally undeniable.The aggregate profits of the top 100 most profitable companies reached 803,470 billion $US in 2010 (2010 Forbes World's Leading Companies).To put this in perspective, France's GDP in 2009 was only 1,907 €billion.
Despite a steady rise in the number of companies from developing and transitional economies among the largest firms (The Financial Times includes 124 of them in its list of the 500 largest companies in the world, according to UNCTAD, 2010), developed countries dominate the top of the TNC scoreboard, with five countries (the United States, the United Kingdom, Japan, France and Germany) accounting for 730 of the top 1000 firms (UNCTAD 2010).Overall, in 2008, TNCs belonging to developed countries accounted for 92% of the 5000 top TNCs' foreign assets, and 90.9% of their total foreign sales (UNCTAD, 2010).The predominance of large global OECD-based companies in high-tech activities is still more compelling.According to a study by the European Commission, the top 1,400 companies (of which 400 were EU-based and 1,000 non-EU-based), most of them belonging to OECD countries, invested €402 billion in R&D in 2009 (Industrial Research and Innovation, 2010) 2 .This corresponds to approximately 80% of world business expenditure on R&D.Even within this already concentrated group, there is a further concentration, with the top 10 companies accounting for 13.6% of total R&D investment by the 1,400 Scoreboard companies.
Such quantitative indicators do not provide the whole story about the real influence of TNCs, however.TNCs play a decisive role in reshaping international trade and production, and the strength of the relationships entertained by most of them with 'their' governments gives them a critical edge in world competition.Contrary to claims that globalisation means that global companies have become 'stateless' , it is clear that governments and TNCs have, in the past decade, taken advantage of the benefits of mutual cooperation in the context of the exacerbation of economic competition, compounded by the financial and economic crisis that has been spreading all over the world since the end of 2007.
It is the main thrust of this paper that any theoretical examination of TNCs should start from the fact that they cannot be defined, as is usually the case, only by the fact that they are bigger and more internationalised than other firms.They constitute a category of their own, which, to be understood, requires additional tools to those conventionally used to study 'firms' .Building new analytical tools to understand TNCs was one of the main objectives of Hymer, a forerunner in the analysis of the TNC.Dunning and Pitelis remind us that, one of Hymer's lasting contributions as long ago as the early 1970s, was to consider that 'the theory of international operations is part 2 There are statistical divergences between different sources.Jaruzelski and Dehoff (2010) report that the top 1,000 R&D Companies spent 503 billion $ US in 2009 They found a decline in R&D expenditure in 2009, due to the economic crisis (-3,5% in 2009 compared with -1,9% in 2009 for Industrial Research andInnovation).
of the theory of the firm' (Hymer, 1976:21, cited in Dunning & Pitelis, 2008).This insight led to radical conclusions on the nature of multinational enterprises, the scope of their control, their power and value capture.Grounding his analysis in a political economy framework, Hymer predicted that the bleak future he saw being delivered by capitalism was epitomised by the multinational corporation, which constituted 'its swan song ' (1972: 110).This puts Hymer's findings at some distance from the somewhat conventional 'internalisation' theorist he is sometimes likened to (for a clear recognition by a prominent scholar of transaction costs that Hymer cannot be identified with a 'left' wing of a Coase-inspired3 analysis, see Teece, 2006).
Forty years after Hymer's seminal inputs, the analysis of TNCs remains dominated by the approach of conventional economics to 'firms' .We argue here that, given their dramatic transformations in the last decade, there is an urgent need for a reassessment of TNCs.Our hypothesis that they constitute a category of their own is supported by many characteristics, including the following: First, they are organised as groups of enterprises, a reality that has begun to be seriously addressed by statisticians in recent years (see below), although it has been of little concern in the theoretical literature.The latter is dominated by issues such as principal/agency relationships (in which the firm is seen as a nexus of contracts), and cost transactions.Here, the multidivisional firm, as analysed by Williamson (1981) is the most efficient and optimal structure.In the national reporting accounts, enterprise groups are statistically defined as 'associations of enterprises bound together by legal and/or financial links which imply control' (Economic Commission for Europe, 2010:4).TNCs as groups represent a category of firms in their own right, based upon a centralisation of financial assets and a specific organisational structure (with the core role held by a holding company).By definition, such groups constitute a structure in which financial control dominates industrial activities.TNCs have long developed financial activities, but they have been given further opportunities to do so in the two last decades.Their active management of financial assets has considerably increased in recent years, challenging once well-established categories, such as 'Foreign Direct Investment' (FDI).In other words, TNCs are now effectively financial groups with industrial activities.As such, they represent one modality of contemporary 'finance capital' , which cannot be defined, as it was by Hilferding in the early 20 th century, as industrial capital that can be disposed through the banks, making industry dependent on banks.(1910: Chapter 14).In 21 st century capitalism, a major characteristic of finance capital is that it is no longer an alliance of industrial companies and banks under the dominance of the latter; it has been subjected to a blurring of boundaries between financial and non-financial activities within non-financial TNCs.
Second, a major feature, and a critical edge, of TNCs as financial groups lies in their ability to build an integrated global space, with financial and industrial operations being addressed in a combined way.It is a global space because it overcomes national boundaries and governmental regulations.It is an integrated space, because hundred of affiliates (production, R&D, financial, etc.) are co-ordinated under the control of a central office which manages resources and capabilities with the objective of giving coherence and efficiency to the process of valorisation of capital.Again, we need analytical tools to address the dramatic changes brought about by the creation of such integrated global spaces, something which cannot be encapsulated within the internationalisation debate.
Third, from an economics of production vantage point, this integrated global space can be analysed through a global value chain (GVC) approach (Gereffi, Humphrey & Sturgeon, 2005).This approach goes beyond the observation of techno-productive sequences; it takes into account the balance of power among the actors in the GVC, the interconnections through which they structure world industries and markets, their mode of governance and the TNCs' strategies for influencing the (de)regulation agenda.Finally, a global value chain approach draws attention to the category of rent, its sources and how it is shared among the different companies.
This paper is organised as follows: Part two explains why TNCs can be considered as a 'contemporary modality of finance capital' .This concept is explored, drawing on a reading of Marx, which is used for analysing the global strategies of TNCs.Part three addresses a major feature and a critical edge of TNCs: their ability to build an integrated global space, with financial and industrial operations addressed in a combined way.Part four shows that FDIs reflect this combination of financial and industrial operations, while Part five shows that intangible assets, which are largely a creation of the financial community, play an increasingly important role in the financial activities of TNCs.

Finance capital as a concept
A clarification of the concept of finance capital, drawing on our reading of Marx, is needed, not least because it differs significantly from the definition given in Hilferding's (1910) seminal book.In his major works (Capital, Economic Manuscripts, Theories of Surplus Value), Marx puts finance capital in a different context and gives it differentalthough not contradictory -meanings.In some places, Marx refers to finance capital as a specific business -'the trade with money as a commodity' -linked to the specific role played by money in the process of accumulation: 'Capitalist function consists exclusively in performing the financial operations for the entire class of industrial and commercial capitalists' (Volume 3, chapter 29).In turn, as capitalism become a dominant form of socio-economic organisation and those operations linked to money become more complex, this business sub-divides into different businesses: 'Large offices, many bookkeepers and cashiers, a far-going division of labour, disbursement of money, receiving of money, balancing of accounts, keeping of current accounts, storing of money, etc., all these things, separated from the acts that necessitate these technical operations, make of the capital advanced for these functions a financial capital' (Volume 3, chapter 29).In other places, and even in the same chapter, finance capital is presented as synonymous with interest-bearing capital or 'moneyed capital' , i.e. an accumulation of such claims on production; an accumulation of the market price, the illusory capitalvalue (Volume 3, chapter 29).

Passing the buck: corporate restructuring and the casualisation of employment
A Marxist framework offers a unitary approach to capital.Capital, as an abstract category based upon (antagonistic) social relations, finds its concrete form both in productive capital (fixed equipment, etc.) and financial capital, i.e. money capital generating more money thanks to ownership of financial assets and loans, as well as property rights which gives (to those who monopolise these resources) the right to collect rents.
Because finance capital existed well before the development of industrial capitalism, its modern configuration has its roots in the specific role held by money in capitalism.Capitalism is not an economic (and social) organisation based on barter 'plus' money, i.e. with money acting as oil, lubricating and easing the expansion of commodity exchanges (a view typical of the neoclassical approach which narrows down money to its function as a means of payment).On the contrary, capitalism is the only mode of production to be driven by a relentless quest for accumulation, resulting in more money being collected at the end of the cycle than at the beginning, with money acting as the universal form of value.To phrase it in Marx's words, the cycle of capital is M (Money)-P (Production)-M' (M'>M), a formula approvingly referred to by Keynes (1971Keynes ( -1989))4.In this way, the production (and sales) of commodities, based on the ability of capitalists to accumulate surplus value, is always a means (an intermediate), never the end of the process 5 .
This 'logic' of capitalism paves the way for a growing autonomy of the circulation of money capital, property titles and claims.The autonomy of money capital eases the growth of fictitious (interest-bearing) capital, a concept developed by Marx in particular in Capital.The market value of this capital is set on the basis of the capitalisation of the revenues of property titles and claims (for the core role of this category, see Chesnais, 1994).Marx saw the origin of fictitious capital in the development of the credit system and the joint-stock company system (and correlated stock exchange markets), with the active involvement of governments through their public debts.Against the framework of capitalist relations, the vertiginous and autonomous growth of fictitious capital is made possible and gives its capitalist owners the right to extract revenues from value created and existing wealth.Interest-bearing or moneyed capital defines what the essence of money capital in capitalism is: to produce revenues thanks to its ownership (Marx, 1861(Marx, -1863)).
Economising Böhm-Bawerk's (1884-1909) 'detour of production' , money capital is able to self-valorise through advances of money capital and produce revenues, 'as 4 See ' A meaningful observation made by Karl Marx, though the subsequent use to which he put it was highly illogical (was...) that the nature of production in the actual world is not, as economists seem often to suppose, a case of C-M-C, i.e., of simply exchanging a commodity (or effort).That may be the standpoint of the private consumer.But it is not the attitude of business, which is a case of M-C-M' , i.e., parting with money for commodity (or effort) in order to obtain more money' .(Keynes, 1971-89:81) Still, the strong differences existing between the meaning for Keynes of this formula and Marx's analysis should not be underestimated, and there has been considerable literature on these issues (see, among others, Sardoni,1997). 5 'Just because the money-form of value is the independent, tangible form in which value appears, the form of circulation M... M' , the initial and terminal points of which are real money, expresses most graphically the compelling motive of capitalist production -money-making.The process of production appears merely as an unavoidable intermediate link, as a necessary evil for the sake of money-making.All nations with a capitalist mode of production are therefore seized periodically by a feverish attempt to make money without the intervention of the process of production' .a pear-tree gives pears' (Marx, Capital, Volume 3, chapter 24:2).Autonomy does not mean independence; instead, property titles and claims (finance capital) represent for the most part the social power of private property rights (based in law) to extract value from labour and wealth: 'In so far as we have hitherto considered the peculiar form of accumulation of money-capital and of money wealth in general, it has resolved itself into an accumulation of claims of ownership upon labour' (Marx, Volume 3, chapter 30).Capital is based on social relations which give its owner the right to command labour and to capture some part of the value created in the production and commercialisation process.Ultimately, finance capital represents an accumulation of claims on labour.Whilst it is possible for individual units, that is agents and institutions, to be pure rentiers and/or financial revenue collectors, this is quite impossible at the aggregated level -say a country -where value has to be produced by the workforce through its labour.
The very existence of such a duality in capital and its consequences were also later analysed by Veblen, one of the most acute observers of the transformation of capitalism brought about by the incorporation process (Veblen, 1904).In his analysis of 'modern business capital' (Veblen, 1904:Chapter 6) he observes that, while capital can be seen 'as a stock of the material means by which industry is carried on' by the 'received body of (economists') doctrines' , for business, it also represents 'a fund of money values' (Veblen, 1904: 135).Likewise, Veblen stated that investments, in industry or real estate, in interest-bearing securities or in loans represent 'nothing more substantial than a fictitious duplication of material items that cannot be drawn into the industrial process' (Veblen, 1904: 103).

Applying the concept to TNCs
'Finance capital' is a concept which lost momentum during a period when the term 'financialisation' was increasingly used (Krippner 2004;Epstein ed., 2005; for a tentative taxonomy of the financialisation literature and its comparison with a finance capital approach, see also Serfati, 2009).For those familiar with a Marx-inspired analysis, a major reason for rejecting the former in favour of the latter is because of the seminal definition given by Hilferding, who famously wrote of 'this bank-capital -that is, capital in money form -which in this way is converted in reality into industrial-capital, finance capital' (1910).Notwithstanding Hilferding's groundbreaking findings, his definition has been criticised for a number of reasons.First, it is geographically limited: even at the time Hilferding was writing, the role of commercial banks as providers of money capital was not universal.The point was made by Sweezy (1942) who said that in the United States, private bankers and other dealers in domestic and foreign exchange, first entered the field of new securities and in this way gradually evolved the institution of investment banking 6 .Likewise, the British system of finance capital involved a fusion of financial and industrial capital through the practices of the agencies of institutional capital, and was neither bank-dominated nor organised into banking 'empires' of the kind depicted by Hilferding (Scott, 1976).

6
Sweezy changed his mind in the mid 1990s, observing that a financial superstructure (made up of banks and a host of dealers) was now sitting on top of the world economy and most of its national units (Sweezy, 1994).
Passing the buck: corporate restructuring and the casualisation of employment Second, the concept is historically circumscribed: in the last three decades, the irresistible rise of stock markets as a result of market-friendly governmental policies (or, as a mainstream economist would put it, the replacement of a debt-based financial market by a market-based one) has put banks in a secondary position compared to the role of financial markets in the funding of non-financial corporations.
A third consideration is that Hilferding's concept of finance capital is open to criticism because it does not adequately capture the complexity and range of relations between industrial and banking capital that developed in the course of the 20th century, though it is nevertheless important because it focuses attention on the organic and institutional links between these two types of capital (Lapavitsas, 2007:19).
It is also our view that finance capital remains a useful concept for analysing contemporary capitalism.The concept starts from a definition of capital as a certain type of social relations which is incarnated in a distinct way into money and productive capital, and this facilitates an understanding of the structural dynamics of contemporary capitalism.This is particularly useful for an analysis of the nature of industrial groups (i.e. the set of the holding/parent company and its affiliates).Here, drawing on research conducted in France in the late 1970s, Morin, has proposed a description of financial groups as a unitary structure of governance made up of overlapping but hierarchical levels: a financial level which orients and monitors resource allocations and an economic level which encompasses the production level, including such questions as how work is organised (Morin, 2006).In an approach which is convergent with Morin's, we have underlined the dominance of a financial logic in the strategy of industrial groups; this makes it possible to define them as an 'organisational modality of finance capital' , (Serfati, 1996:144).
Finance capital, as a concept, has two facets: one institutional and the other functional.While distinct, these are intertwined (Serfati, 2000).Finance is an institutional sector, made up of firms, the business of which is based on financial activity (the financial industry as distinct from, for instance, the automotive, chemical or energy industry), resulting from the division of labour underlined by Marx.It also defines a function, the ability of money capital to produce revenues 'as a pear-tree produces pears' .
These two dimensions of finance capital are conflated in the definition of the financial sector given in national accounts, which is the following: 'the financial corporation sector consists of all resident corporations and quasi-corporations principally engaged in financial intermediation or in related auxiliary financial activities' (IMF, 2008) 7 .The definition of financial services has been enlarged to give due weight to the increase in financial services other than financial intermediation, specifically financial risk management and liquidity transformation.Now, as is clear from this definition, financial institutions are defined by their main function, which is to provide financial services.Such a definition is, however, untenable from a finance capital-based perspective.Bank and non-bank financial institutions (mutual, pension,

7
A slightly different definition of financial institutions is given by the OECD: 'the set of institutions, instruments, and the regulatory framework that permit transactions to be made by incurring and settling debts; that is, by extending credit' , See http://stats.oecd.org/glossary/detail.asp?ID=6815.investment funds, etc.) on the one hand, and 'financial services' functions on the other cannot be conflated because the latter are not performed only by financial institutions, whether they are banks or not.
In contemporary capitalism, this functional opportunity is also offered to industrial groups through the centralisation and circulation of financial assets and other rentgenerating assets.They rely on numerous affiliates, regardless of whether the latter are registered as 'financial' or 'non-financial' units in their home country.One of our arguments with Hilferding's analysis is that he conflates the institutional and functional aspects, of finance capital (interest and dividend-bearing capital).In our view, he is wrong to assert that 'through this relationship, the banks' capital assumes the form of finance capital, its supreme and most abstract expression' (Hilferding, 1981:12).Although he correctly documents the core role of money capital in capitalism's dynamics (his analysis of money is still close to Ricardo's quantitative theory), he takes it for granted that there is a universal (in the abstract sense) institutional configuration which is historically and geographically bounded.Notwithstanding these debates, 'finance capital' remains a powerful concept both for understand the theoretical foundations of 21st century capitalism and for obtaining empirical evidence of the steady and oppressive power of financial asset ownership through the capture of rent revenues (interest, dividends, royalties).

Institutional separation of productive and finance capital
Joint stock companies (JSCs) provide glaring evidence of the duality of capital (in capitalism, capital-property and capital in production are separated) and the ascendancy of revenue-bearing capital (interests and dividends, capital gains, royalties, etc.. belong to a similar category, because they are rent-like revenues generated by the mere ownership of property titles) 8 .The separation in joint-stock companies between the functions needed to carry out the production process based upon productive capital and the ownership of capital was noted by Adam Smith (1776).Smith was mainly concerned about the possibility that managers might 'rob' shareholders 9 , even though, according to some commentators, he saw some superior organisational efficiency in joint stock companies (Anderson & Tollison, 1982).
Ever since Berle and Means' seminal findings on the emergence of a powerful class of professional managers insulated from the pressure of stockholders (1932), these issues have triggered a considerable debate in the academic and managerial literature.Sociologists, lawyers and political scientists entered the debate very early, while mainstream economists turned their attention to this issue only in the 1960s.For the latter, the connection between financial and productive activities within firms 8 Engels, commenting late in the 19th century on Marx's analysis of joint stock companies, indicated that 'new forms of industrial enterprises have developed, as we know, representing the second and third degree of stock companies' (Marx, Volume 3, chapter 23).For him, it was clearly based on the creation of fictitious capital.Commenting on Marx's development of this category, Engels wrote 'This doubling and trebling of capital has developed considerably further in recent years, for instance, through financial trusts' (Chapter 29). 9 'Being the managers rather of other people's money than of their own, it cannot be well expected that they should watch over it with the same anxious vigilance with which partners in a private copartnery [copartnership] frequently watch over their own (…) Negligence and profusion, therefore, must always prevail, more or less, in the management of the affairs of such a company' (Smith, 1776).
Passing the buck: corporate restructuring and the casualisation of employment is seen as being effected through inter-individual relations, since the firm is seen as a nexus of individual contracts (Jensen & Meckling, 1976).The need to explore the 'double nature' of capital, both as physical and financial assets, and to analyse their functional interrelations, as a unique object of investigation, was an issue of little concern in mainstream economics, after the 'Cambridge Capital controversy' and the convincing inputs, in particular by J. Robinson (1953Robinson ( -1954) )  10 .More recent attempts have been made, in particular by Williamson (1981), to consider that the 'M-form' corporation achieves superior efficiency because 'it takes on many of the properties of (and can usefully be regarded as) a miniature capital market' (Williamson, 1981(Williamson, :1554)).Numerous critiques have been made of the underlying assumption that internal capital markets do not suffer the failures experienced by external capital markets (Bolton & Schlafstein, 1998) as well as use of Chandler's historical research to incorporate it into a transaction-cost framework (Lazonick, 1991: chapter 7).
Marx, following Smith, did not ignore the possibility that some people might build their fortunes on 'swindling' and 'cheating' 11 .Still, he was less concerned with the 'agency' problem created by the management-ownership separation than by the theoretical implications for capitalism of capital being separately (and both) productive and revenue-bearing capital.For him, joint-stock companies buttressed the separation between productive-capital and money-capital 12 .
The reason why the issues related to the distinct functions of capital as property (financial assets) and capital in the production process are given precedence over the 'agency' problem raised by the institutional separation between shareholders and managers is important to explain here.This duality of capital exists even if it is the same people who manage and own the firm -or the corporation 13 .The existence of finance capital as money generating more money, (that is, financial claims producing interest, dividends, capital gains, royalties, etc.) is neither due to the division of labour between shareholders and managers nor to the existence of financial institutions.In other words, functional separation between productive and financial capital does not exist because of this division of labour.It is the other way round: institutional separation between different activities and business has grown in importance because in capitalism money 10 Regarding the financial/monetary realm and its relationship with the ill-named 'real' economy, the presumption of the production and primary appropriation of value as separated issues has a cost in terms of analysis.As late as the end of 2007, a confident OECD report noted that 'the effect of financial turmoil on total activity stemming from the sector itself is likely to be small ' (2007:25).The reason for such a huge misunderstanding of what was already more than a simmering crisis is that, according to the report's authors, the financial sector accounts for less than 10% of the value added in most OECD countries.Money is 'neutral' , has no other role but in nominal price fluctuation, and finance is seen as an institutional sector like any other industrial sector.11 The development of joint stock companies 'reproduces a new financial aristocracy, a new variety of parasites in the shape of promoters, speculators and simply nominal directors; a whole system of swindling and cheating by means of corporation promotion, stock issuance, and stock speculation.It is private production without the control of private property' (Marx, volume 3, chapter 27).12 'Transformation of the actually functioning capitalist into a mere manager, administrator of other people's capital, and of the owner of capital into a mere owner, a mere money-capitalist' (Capital,vol.3,chapter 27).

13
'The employer of capital, even when working with his own capital, splits into two personalities -the owner of capital and the employer of capital; with reference to the categories of profit which it yields, his capital also splits into capital-property, capital outside the production process, and yielding interest of itself, and capital in the production process which yields a profit of enterprise through its function (Capital,vol.3,chapter 2) is the ultimate form of wealth, into which everything has, at the end of the process, to be transformed 14 .That said, there is no doubt that the evolution of the institutional design of firms since the mid 19th century, as explored below, has considerably helped to reinforce (but not create) the autonomy and ascendancy of revenue-bearing capital (interest, dividends, etc.).

Three main stages in the autonomy of finance capital
The separation between productive and finance capital (revenue-bearing capital) within capitalist enterprises has been growing since the mid 19 th century.For the purposes of this article, three major stages are briefly reviewed.The first stage corresponds to the introduction of limited liability in Joint Stock Companies ( JSCs) introduced in Britain in 1855 after a gradual extension since the repeal of the Bubble Companies Act (1825, passed in 1702, which had sought to prohibit unincorporated joint stock companies and the Bubble Act, under which it became a crime to organise such corporations without explicit royal consent).
In the USA, limited liability began early in the 19 th century (in New Hampshire in 1816, Connecticut in 1818 and Maine in 1823), as New York ( 1811) and New Jersey (1816) adopted statutes for the incorporation of manufacturing companies that provided for double liability (Blumberg, 2006).In Japan, JSCs spread very rapidly from the outset of modern economic development, because they were propped up by Government officials, despite businessmens' reluctance to adopt this form of modern enterprise (Daito, 1989).In France, the Code of Commerce of 1807 authorised the creation of sociétés anonymes (corporations), as well as sociétés en commandite par actions (limited partnerships).Still, at the turn of the 1880s, for a number of reasons, French entrepreneurship lagged behind that of other developed capitalist countries.It was only in the 1920s that the 'triumph of limited liability' was achieved, with twice as many companies created as before WWI (Bouvier & François, 1993:774).
The creation of JSCs reflected a structural thirst for capital in some industrial sectors, such as transport infrastructures (canals and railways), public utilities, and, of course, the financial industry (insurance and banking).This process was extended to new industrial sectors at the end of the 19 th century, including steel and oil, as well as more banks.As economists have long acknowledged, some industries require monopolisation, because of the need for economies of scale and positive externalities.Nevertheless, this still leaves open the precise organisational forms that will be adopted, and this structural impulse for centralisation of capital should not be confused with the institutional design adopted by JSCs.Incorporated associations with public or quasi-public status, performed well in many countries, and when and where (mainly in the USA) it was decided that corporations would be given the status of public utilities, this did not entail the introduction of limited liability and creation of holdings (Handlin & Handlin, 1945).

14
The distinct existence of corporations and their individual owners was confirmed by the English House of Lords case of Salomon vs Salomon.The court's decision allowed private actors to organise their business through the corporate legal form, even though the enterprise was composed of the entrepreneur and members of his household alone, and so was not strictly a true joint stock enterprise but a sole trader who would, up to that point, have been personally liable to all creditors (Muchlinski, 2010) Passing the buck: corporate restructuring and the casualisation of employment Corporations with limited liability were not an outcome of capitalism per se but resulted from a complex set of causes, among which political ones were prominent.As showed by Ireland (2010), limited liability, with low denominations of shares, resulted from political pressures aimed at meeting the demands of rentier.Ireland concludes that this made it possible to give shareholders controlling rights in a company while exonerating them from personal responsibility for the debt and liabilities of that company.
The second stage in the development of organisational forms of enterprises that reflected the autonomy of capital as property vis-à-vis productive capital was the creation of holdings.According to a well-known definition given by Bonbright and Means (1969:10) a holding company is 'any company, incorporated or unincorporated, which is in a position to control, or materially to influence, the management of one or more other companies by virtue, in part at least, of its ownership of securities in the other company or companies' .An important landmark was in 1899 when Jersey Standard, while retaining operating functions, became the parent holding company of 33 corporations.Holdings came to become a major, if not the dominant, organisational form in the USA.
The debate about the dangers and benefits of limited liability, already contentious in relation to JSCs, became still more acute with the development of holding companies.That, with the creation of JSCs, risk-taking capitalism could be exonerated from liability and that rentiers should be given increasing momentum was indeed distant from the capitalist ideal-type as wished for by some theorists and policymakers.Still, in the case of the extension of limited liability to holding companies, it was not only rentiers who were exonerated from liability, but corporations themselves, owning a majority of shares in other corporations, thus insulating the parent company's shareholders by means of 'a second layer of protection' (Blumberg, 1986:607).For this reason , the advantages of limited liability can be regarded as irrelevant to the 'incredibly complex' special world of corporate groups [Blumberg, 1986: 624] While 'groups' as organisations, are defined by a hierarchical structure as well as by strategic and financial control by the head (parent) company, their institutional design spans a range of specificities which could be historically-and country-dependent (Jones & Khanna, 2005).Arguing against the mainstream view that such structures are inefficient and rent-seeking (in particular insofar as they are closely connected to governments), Khanna & Yaffeh (2007) find that they are not an exception that would result from resistance to 'good governance' , but are largely present in many countries outside the United States.Business groups centred on a familial form of control are a dominant form in many Asian countries (Goto, 1982).Corporate pyramids are vertically-controlled groups but there are also horizontally-linked groups, where cross shareholdings are important (Khanna & Yaffeh, 2007).Likewise, the holding form became dominant in Europe in the post WWII decades, but numerous and varied types of relations between parent holding companies and their affiliates in industrial production-oriented sectors could be found in European countries (Amatori & Colli, 2007).Some commentators differentiate between 'pure' holdings, in which the investment company may be viewed as a 'cousin' of its affiliates and, corporate pyramids with a corporation acting as a holding company at the apex of the group (Bank & Cheffins, 2011).
Despite the many differences between holding companies and conglomerates which are strongly emphasised, including by Williamson (1981) who found that the latter are more efficient than the former, a similar trend could be found all around the world in the post-WWII decades.This trend was towards the creation or consolidation of industrial groups based on centralisation of parent companies exercising control through financial assets ('capital-property') of numerous affiliates which were in charge of production.Because the creation of holding companies is aimed at overcoming industrial constraints (such as economies of scale and coordination of distinct technical or managerial operations), this trend clearly reflected what was primarily a financial centralisation, i.e. the centralisation of property claims, the objectives of which were to reinforce the market dominance and political power of the group, as well as increasing financial wealth for shareholders.In most cases, whatever their national singularities, these different types of industrial groups were defined by strong inter-corporate relationships relying on more or less dense equity interlocks, but also tight interpersonal and other informal relationships rather than market-based transactions, with the holding company centralising and controlling financial assets and designing the strategy of the group as a whole.
The next stage was the development of transnational corporations.The deepening and widening of financial opportunities opened up by the radical changes that occurred in the new international monetary and financial setting of the 1980s and 1990s, along with shareholder value-based governance, dramatically increased TNCs' scope for financial development (Lazonick & O' Sullivan, 2000).The statistical definition of enterprise groups is as follows: 'associations of enterprises' bound together by legal and/ or financial links which imply control' (Economic Commission for Europe, 2010:4).Valorisation of their assets has become increasingly based on active management and further sophistication has been made possible by the deregulation of financial markets, including product innovations and offshore institutions (for debates on the role of TNCs' financial objectives in the early 1970s, see United Nations, 1973).
The disconnection within TNCs between their financial flows (whether these are for payments, cash netting, trade finance, loans/debts or cash pooling) and their goods flows has been further exacerbated by the creation of hundreds, some claim thousands, of Special Purpose Entities (SPEs) whose main purpose is to maximise the profitability of money capital 15 (Sola, 2006).Other names given to institutions with similar financial objectives to those of SPEs include 'special purpose vehicles, shell companies, special financial institutions, brass plate companies, mailbox companies or international business companies' (Economic Commission for Europe, 2010).
The innumerable SPEs created by TNCs have been developing in parallel with the centralisation of tasks and functions.There is no paradox in having on the one hand a centralisation of control, in particular of financial operations, and on the other hand a decentralisation of operations through the creation of hundred of entities dispersed across the world, mainly in low-tax jurisdictions.That does not mean that all these SPEs are registered and classified as financial corporations: the IMF (Sola, 2006) distinguishes between non-financial (shared service centres for administration, marketing, accounting,

15
A conventional definition of an SPE by the OECD (2008a) is that it is an entity 'in which the parent companies are resident, and engaged primarily in international transactions but have few or no local operations' .sales, management of patents, etc.) and financial SPEs proper.Most SPEs are still financial ones (cash pooling centres, treasury centres, conduits entities, etc. 16 ) and, because they are present in many different countries, TNCs can benefit from the differences between domestic financial markets or between their different segments.To shed some light on the financial dimensions of industrial groups, we collected data on the French case.These data provide clear evidence that the group, as an organisational structure, is designed to drain financial revenues (see Figure 1).One can observe that in France while the share of holdings in the overall creation of value added and workforce payroll (corresponding to what is conventionally described as the 'real' activity of the firm) is quite modest (around 2%), the share of holdings in interests paid and net financial debt is very large (around 45-50%).This picture confirms that holdings can be regarded as a structure (or 'conduit') designed to centralise and drain financial revenues.Groups' holdings are built on a 'reversed pyramid' the basis of which is mainly financial, with a very thin tip of productionrelated activities (Figure 1).

16
A cash pooling system can optimise the use of excess cash and interest yield (by maximising the return by means of proper allocation of short-term investments), reduce interest expense (by minimising the cost of borrowing by borrowing in different money markets) and avoiding costly intra-company transactions.Treasury centres are usually in charge of managing the treasury activities of their group (cash flow and cash position forecasting, banking and cash management, liquidity management, funding management, risk management).Conduit entities are those whose main activity is to raise funds from international markets and lend the proceeds to their group (Sola 2006).

Global valorisation of capital
A major feature of TNCs as financial groups, giving them a critical edge, lies in their ability to build an integrated global space, with financial and industrial operations being addressed in a combined way, a situation remarkably envisioned by Hymer more than forty years ago 17 .It is a global space because it overcomes national boundaries and governmental regulations.It is an integrated space because hundreds of affiliates (production, R&D, financial, etc.) are ultimately under the control of a central office which manages resources and capabilities with the objective of giving coherence and efficiency to the process of valorisation of capital.To borrow from conventional industrial and firm economics, the world is now an 'internalised' area for large TNCs.This process is not restricted to the industrial activities of the corporation.As already noted, it is precisely the blurring of boundaries between industrial and financial activities that reinforces the qualitatively distinctive features of TNCs as 'firms' .
Interactions between productive and financial valorisation of capital have become denser, reflecting the 'TNCs' global valorisation of capital logic' (Serfati, 1996:148).We define the logic of global valorisation of capital, using the term 'global' in two distinct senses: first, in relation to the fact that for large TNCs the whole world has become a playing field for locating and sourcing their activities and inputs (as revealed by the debates about offshoring); and second, in relation to the fact that top managers are offered a wide range of modalities for the valorisation of TNCs' financial resources.The valorisation of capital may thus encompass a broad spectrum and have a global reach, with at one extreme the implementation of industrial investments (new equipment, etc.), and, at the other, 'pure' financial investments made on financial markets.Between these two 'polar' forms of valorisation, other 'mixed' forms occur, resulting from the growing interaction between finance and production, blurring the boundaries between the two spheres of finance and production.For example, the development of Intellectual Property Rights (IPR) does not only reflect successful innovative activities; it also results from the ability of large TNCs to capture a share of the value created by other firms, often in Small and Medium-sized Enterprises (SMEs) and start-ups.This can be achieved through acquisitions or so-called 'market power' 18 , by constraining smaller firms to license their patents 17 'The modern multidivisional or conglomerate enterprise (...) appears capable of integrating world production and exchange to a much larger extent' (Hymer, 1970:444).In another paper, Hymer connects financial internationalisation and TNCs, stating that 'the multinational corporation and the international capital market should be seen as parallel, symbiotic developments ' (1972:99).18 It would be more accurate, as we did elsewhere, to speak of a 'relational power' held by large TNCs, as the latter reflects their ability to build powerful linkages, many of them created outside of the 'market' and relying on political lobbying, social networks built by executive managers and shareholders, influence on firms of lower size, etc.Market power is an outcome of this kind of power (Mampaey & Serfati, 2006).That holding corporations are not only economic institutions but also a political power was forcefully claimed by Bonbright and Means (1969:6).Analysing the devastating role of financial institutions, Dymski stresses the importance of network-based power, and rejects the narrow approach to market power, arguing that 'concentration' can no longer stand in for 'power' (Bonbright & Means, 1969:82).(Jaffe & Lerner, 2004 19 ).It is increasingly acknowledged that IPR has become an object of financialisation: the explosion of patents in the last two decades has little to do with a Schumpeterian 'gale of creative destruction'; rather it is evidence that TNCs' strategy is oriented towards extracting rents, defined as regular flows of IPR-generated revenues (Zeller, 2008).Biopharmaceutical TNCs are particularly strongly involved in this strategy (Montalban & Sakinc, 2009).
Overall, we interpret TNCs' strategies during the era of globalisation as connected with the blurring of the frontiers between value appropriation through a direct value-creating production process on the one hand and through rent capture on the other hand.Rent is a complex category which dates back to the very origins of political economy.In our view, rents exist when people and institutions hold private property rights, allowing them to be in a monopoly situation or/and create a (relative) scarcity, from which they can obtain a flow or revenues from other people and institutions in exchange for the use of resources.Because a rent could exist as a result of 'natural' scarcity (e.g.land, or natural resources), the extension of private property rights (thereby producing a monopoly situation and thus generating rents) is a socio-economically embedded process which is endorsed by the political institutions (generally the state) that are in charge of enforcement and the protection of private property rights.
While the differences between profits and rents were strongly emphasised by Ricardo and his followers, the progressive blurring of their boundaries since the end of the nineteenth century has made it in reality highly challenging to distinguish between what proceeds from 'entrepreneurial' profits and what proceeds from rent appropriation.
In the two last decades, there has been a significant broadening of private property rights in new realms of economic and social life (intellectual activity; life processes; even pollution has become a sphere for the creation of tradable permits).In this context, TNCs have become more oriented toward the generation of revenues based upon their financial and intellectual property rights than on the production process proper.

Bundling financial assets and 'slicing up' the global value chain
Corporations, under pressure from shareholders and 'financial markets' (which is not an invisible hand but made up of quite visible consulting companies, auditors etc.) and as theorised by mainstream economics (in particular in principal-agent theory) are now considered as an agglomeration of distinct segments of capital which have to be monitored by their own return on investment (generally shareholder-value oriented).As corporations tend to be seen by financial markets as bundles of assets, the latter have to be highly liquid if they are to be quickly invested or divested with gains in capital on request on stock exchanges.In the United States, where this radical transformation first emerged, large modern corporations came to be run by managers who endorsed the financial conception of firm control (Fligstein, 1990, Chapter 6).Governmental policies oriented towards the deregulation of financial markets, which were implemented in all industrialised countries during the 1990s, reinforced this managerial ideology.
It is now conventionally admitted that the wave of mergers during the 1980s and 1990s was, under the influence of 'market for corporate control' approaches, largely finance-driven and that these mergers reflected an underlying shift in the dominant conception of the firm toward a financial model (Davies & Stout, 1992: 629).
This short-termist, financial gains-motivated strategy, seems to have been reinforced in the USA by the 'new finance capitalism' in which a small handful of mutual funds became the most significant large-scale corporate owners.This combines the double benefits of concentration and liquidity, resulting in the ownership of vast, but evanescent, networks of portfolio companies (Davis, 2008).Whether that industrial restructuring of their global value chain through external control by stock markets was successful, even from a shareholder perspective, is debatable, since nearly 70% of mergers failed to achieve their expected revenue synergies; not to speak of their negative effects on labour (Milberg & Winkler, 2010;Huws, 2006).The role of finance in cross-border mergers and acquisitions (M&A) was still more evident than in M&A confined within the national boundaries of developed countries.The conundrum was that the cross-border M&A wave occurred and persisted over two decades (the 1980s and 1990s) despite the fact that many of them fell short of delivering the anticipated positive results in terms of either share prices or productivity (UNCTAD, 2000).
This view of corporations as a set of disposable financial assets has coincided with dramatic changes in the production process.In any case, a major goal set by management was to cut down labour costs, refocusing on core competences by dropping non-core activities and maximizing synergies, searching for economies of scale and cost reductions through the closure of plants, etc.Some of the terms used to describe these processes include: 'vertical disintegration' , divestment of segments of the supply chain and 'slicing up the value chain' .For these strategies to be implemented, companies were advised to improve the overall efficiency and effectiveness of their resources, economise on organisational costs (assumed to have become higher than market transaction and agency costs), refocus on 'core competences' and give priority to scale economies.There has been a general trend for management to drop production assessed to be insufficiently value-creating and/or 'non strategic' .In practical terms, 'upgrading' by refocusing on the two ends of the global value chain has been the objective, coupled with shorter time horizons in investment decisions.At the top end of the chain, this usually means keeping strategic activities, such as trans-divisional research, technology and business intelligence, development and design in-house.At the bottom end of the chain, the final integration of the product (often designed and described as a 'system') strategies have been developed to ensure that it generates high margins through branding, marketing, protection of intellectual property, etc.
Restructuring on the production side and the quest for financial revenues are interrelated, resulting from the ability of large TNCs to act across a global integrated space.This point is also underlined by Milberg (2008), who shows that global production strategies have helped to sustain financialisation (see also, Palpacuer, 2008).

Intra-TNC trade and transfer pricing
The restructuring of global value chains and the fragmentation of production processes within these chains are mirrored by the growing international sourcing of intermediates.The share of intermediate manufactured products in non-fuel world trade was around 40% in 2008 (WTO 2010:2) 20 .Trade in intermediate inputs (primary goods, parts and components, and semi-finished goods) takes place mostly among developed countries and represented respectively 56% and 73% of overall trade flows in goods and services over the 1995-2005 period (Miroudot, Lanz & Ragoussis, 2009).
This large development of intra-company trade is largely an outcome of outsourcing and offshoring, two processes which dramatically affect the organisation of global value chains.These trade flows are associated with financial flows, or, to be more precise, it could be said that the distinction between trade and financial flows has become difficult to draw, as the next sections on FDI and Intangible assets demonstrate.It can be concluded that trade in intermediates forms a large part of intra-TNC trade.TNCs have been able to increase their grip thanks to networked forms of organisation and coordination, which have allowed them to capture part of the value added created in institutions participating in these networks (including small and medium-sized firms, public research centres, etc.).The development of an integrated space is evidenced by the large share of intra-TNC trade in total world trade.According to OECD data, in 2006, the export propensity by affiliates under foreign control in the manufacturing sector was almost 100% in Ireland, 60% in Finland, almost 40% in France, but only 10% in the USA (Hatzichronoglou, 2010).Although is difficult to collect robust data on this, according to some estimates, 60% of all world trade consists of internal transfers within multinational companies (Sikka, 2009).
Not surprisingly, intra-TNC trade provides support for profit shifting and transfer pricing (TP).Profit shifting can apply to anything that affects the profits that are subject to corporate income tax (Huizinga, 2009).Estimates of the revenue losses from corporate profit shifting vary substantially.In the USA, where the research is by far the most advanced, estimates range from about $10 billion to about $60 billion for US corporations (Gravelle, 2009).In France, the opportunity given to TNCs by a steady decline in capital controls has been so large that, according to a study released by the Conseil National des Impôts (National Council on Tax), in recent years large companies have paid significantly lower taxes in proportion to their revenues than other companies (Conseil des Prélèvements Obligatoires, 2009).One of the reasons for this might be the high proportion of affiliates of French TNCs that are located in tax havens.In 2009 1,470 such entities were created by the 39 French blue-chip companies, amounting to 14% of all their foreign affiliates (Chavagneux & Rinuy, 2009).
Mainstream economics has generally addressed transfer pricing following the Hirshleifer Rule (1956) which states that where and when an external market price 20 It should be noted that the reliability of the statistical data on the volume of intermediates traded on a world scale could be improved.The 2008 WTO report notes that because each time goods cross the frontier an international transaction is recorded, the compilation of merchandise trade statistics by customs administrations results in the recording of these goods more than once.The impact of this 'double counting' can be significant, especially in the case of international supply chains that include a number of tasks -as is the case for transport equipment and electronics -with the result that unfinished goods may cross frontiers several times during the assembly process.exists, the TNC's efficient transfer price should be the external market price.The main criticism of this approach is that the integrated global space designed by TNCs means that affiliates are under the centralised control of the parent company.In other words, intra-TNC relations (exchanges of resources, financial flows, management of workforce, implementation of property rights, etc.) are by definition distinct from those that exist in a (hypothetical or external) free market, supposedly ruled by 'arms-length' relations between separate and equal agents.A second criticism of the mainstream approach to transfer pricing is the difficulty of measuring precisely an 'arms-length' price, to which transfer pricing could be compared.A growing share of intra-TNC trade is made up of intermediate inputs, largely intangibles, for which a pure competitive market with comparable products does not exist.
Transfer pricing plays a central role in the strategies of TNCs.It offers an elegant and opaque way to meet their financial goals.The latter include managing cash flows, supporting R&D, funding capital expansion, paying interest on debt, meeting tax liabilities in accordance with overall group tax strategies and funding dividend payments to shareholders (PricewaterhouseCoopers, 2009).Intangible assets give a serious opportunity to TNCs to engage in transfer pricing policy, as can be seen in a typical piece of subliminal advice from a major consultant: 'It is not necessary that the asset appears on the balance sheet for it to have significant value for transfer pricing purposes' (PricewaterhouseCoopers, 2009:47).Surveys of top TNCs reveal that their management is fully aware of the crucial importance of transfer pricing.Transfer pricing was the single most important issue for 76% of respondents in a 2009 survey of parent companies in the pharmaceutical sector, an increase of 19% compared with a similar 2005 survey.Pharmaceutical companies were nearly twice as likely as companies in any other industry to have experienced an adjustment of transfer prices, and parent respondents in the pharmaceutical sector said that 56% of transfer pricing investigations since 2003 had resulted in adjustments (Ernst & Young, 2009).
Simple in its principle, transfer pricing represents a real challenge for regulatory authorities (e.g.OECD, 2010).The strict application of the 'arm's length principle' , as mentioned earlier, is often problematic in practice.This explains the magnitude of financial flows generated through transfer pricing practices.On the basis of trade statistics, it is estimated that the scale of manipulated transfer pricing in trade to and from developing countries alone amounted to roughly 500 US$ billion in 2006, equivalent to 6.5% of their foreign trade, and almost 50% of their total capital flight (NOU, 2009).

Foreign direct investment at the crossroads of production and finance
Industrial strategies have driven the dramatic restructuring of their global value chains by TNCs that has taken place in recent years.However this forms only a part of the story: financial objectives are also clearly present in this restructuring.In particular, outsourcing through contract manufacturing or any other form may also be motivated by 'international tax avoidance' (Gravelle, 2009).

Passing the buck: corporate restructuring and the casualisation of employment
The strong rise in the flows of FDI (whether measured in absolute terms or as a proportion of GDP) is generally interpreted as evidence of the globalisation of production processes, reflecting the development of offshoring.FDI is given a different status from Foreign Portfolio Investments (FPI), which are seen as reflecting more short-term, financial objectives, because they are generally carried out, at least in emerging-market equity, by large mutual funds and privately-held hedge funds (Global Development Finance, 2004).That FPI has little to do with productive issues is confirmed by a recent study which found that they are used as a tax evasion device: two thirds of all US FPI is hidden from the authorities, and, conversely, US FPI is more than two and a half times as large as one might expect on the basis of official figures (Dharmapala & Hines, 2009).This finding is confirmed by IMF research that found huge discrepancies between portfolio assets and liabilities in selected offshore centres.Portfolio assets held by foreigners in Luxembourg were estimated to be worth US$1.5 trillion at the end of 2008, while portfolio investment liabilities reported by the government stood at US$2.5 trillion (Lane & Milesi-Ferretti, 2010).A black hole indeed.
Even in the case of FDI, things are not so clear.Indeed, the real meaning of FDI as supposedly productive investment is brought into question if the play of financial motivations in FDI operations is addressed (Forssbaeck & Oxelheim, 2008).Some literature has for years drawn attention to the status of cross-border M&A, which accounts for over 80% of FDI between developed countries and over 40% of FDI from developed to developing countries (UNCTAD, 2010).M&As do not add manufacturing capacities; indeed they often subtract them.They only involve a change in ownership, and as such they should be distinguished from 'green field' investment, involving the creation of new industrial capacities or 'brown field' investment, involving an increase in existing industrial capacities.M&As reflect the need to distinguish carefully between the ownership of capital (including changes in this ownership) and productive activities (the 'double nature of capital'), evidencing the extent to which large corporations can carry out their productive and financial (control of ownership) objectives.
Painstaking research and discussion among statistician accountants over many years has revealed that it is increasingly the case that large segments of FDI reflect activities that are primarily financial on the part of TNCs.Flows of FDI are fed by three components: equity, reinvested earnings and intra-company loans.A 2004 World Bank report underlined the fact that, between 1990 and 2002, inter-company loans and reinvested earnings were often used as a means to adjust FDI.Once considered as quite distinct from FPI, (which were seen as more short-term and financially-oriented) FDIs became strongly volatile in the 1990s, especially in the case of inter-company loans and reinvested earnings, which were nearly as volatile as debt flows (World Bank, 2004).The factors affecting the composition of FDI listed in the World Bank report were mainly finance-related (e.g.tax costs, ownership control and investment regulation) while the macroeconomic environment was an additional factor.
More recently, and in the same vein, a report commissioned by the French government claimed that 'direct investment reflects intra-firm financial activities' (Fontagné & Toubal, 2010:14) because most of the three funding components (equity, reinvested earnings and intercompany loans) are classified as generating direct investment, even when their purpose has nothing to do with the creation or acquisition of physical capacities of production (fixed investment) but is purely financial (for instance transfers of funds for fiscal engineering or the higher rate of return offered in some countries).
Using a reassessment of inter-company loans and a new methodology recommended by the OECD (2008a), data compiled by statistical institutions on FDI have produced astonishing results.The data were adjusted by reclassifying inter-company loans according to the country of residence of the ultimate controlling parent of the group.The conventional directional principle on which FDI statistics are compiled was extended to cover loans between fellow enterprises (defined as entities with no direct links).The OECD's recommended rule states that lending and borrowing between resident entities of a resident group and foreign fellow enterprises must be recorded as outward FDI and, conversely, lending and borrowing between resident entities belonging to a non-resident group and foreign fellow enterprises must be recorded as inward FDI.
In the case of France, thanks to research based on this new methodology -that is a reclassification of inter-company loans according to the country of residence of the ultimate controlling parent of the group -the figures relating to inward and outward FDI were transformed.The importance of countries, such as Luxembourg, that are known to host large numbers of SPEs as sources or destinations of FDI was now greatly reduced.In 2007, according to this new method of calculation, the most important country investing in France was -France!French-owned companies accounted for no less than 26.1% of total inward investments into France in that year (Terrien, 2009;2010).Two observations emerge.First, France ranks as the leading ultimate investing country in France in 2008 because of the investments of non-resident subsidiaries of French groups in their French subsidiaries in the form of equity capital investments or reinvested earnings.Second, French TNCs' intra-company loans accounted for as much as 39.6% of total inward FDI, reflecting the scope of intra-affiliate financial flows.There is little doubt that the case of France is not substantially different from most other developed countries.
As a step forward in an attempt to trace cross-border flows within TNCs, this reform in FDI accounting methodology falls far short of exhausting all the channels used by large global corporations.For example, the OECD's recommendations do not apply to equity capital transactions between fellow companies.This means that if a Luxembourg holding company of a French group injects funds into its direct subsidiary, which is also resident in France, this transaction is recorded as inward FDI, even when the ultimate controlling entity is a French TNC.Further research is needed in order to find better means of tracking these TNC strategies.

A new generation of fictitious capital generated by financial markets
This is not the place to develop the hypothesis, addressed in previous research, that intangible assets are a creation of financial markets and communities.That their value is at least in part fictitious (or that they are by nature fictitious capital) can be attested by the scale of their growth over the two last decades 21 (Serfati, 2008).The rising attraction of the category of 'intangible assets' has resulted from a convergence between, on the one hand, economists keen to put figures on what they call 'knowledge capital' and, on the other, the financial community who, from the end of the 1990s onwards, needed some explanation for the growing gap that could be observed between firms' book value and their stock market value.Equipped with this new category, analysts were able to conclude that intangible assets accounted for as much as over 50% of large companies' stock market value.In other words, intangible assets, for all the vagueness of the concept, were deemed to account for the bulk of the financial value of top world companies (figure 2), with a decline in their paper value as a result of the 2008 financial meltdown.To give a flavour of the extraordinary increase in the importance of intangible assets, we can note that they were estimated to account for only for 17% of the total stock value of companies in 1975 and 32% in 1995 22 .This is not to say that intangibly assets do not reflect the rise of new inputs and drivers in the creation of value.Intangible assets 23 have become a key component 21 After the wave of 'creative accounting' which thrived in the 1990s, the fictitious nature of the 'value' created by intangible assets began to be noticed, even among originally strong supporters of the concept.See this comment posted in a post-financial crisis Business Week's paper: 'With the stroke of a pen, companies can make themselves appear more financially fit than they are' thanks to 'new discretionary accounting rules (which) have made it easier for companies to engage in such behaviour' .(Der Hovanesian, 2009) 22 Source: Ocean Tomo 23 'Intellectual' , 'knowledge' , 'intangible' and 'immatériels' are often used interchangeably in the literature. of developed countries' economies, and are often identified as representing the 'knowledge economy' .Nevertheless, their definition is plagued with imprecision.
Neither economists nor accountants have produced an agreed definition of intangible assets.Indeed, it is possible to identify over 80 different approaches or frameworks of value and performance measures (Value measurement, n.d.).Three core characteristics are generally agreed upon in definitions of intellectual assets: first, they are sources of probable future economic profits; second, they lack physical substance; and third, they can be retained and traded to some extent by a firm.The list generally includes at least R&D, patents, and trademarks (OECD, 2008b:9).They also include: human capital, defined as the knowledge, skills and know-how that employees 'take with them when they leave at night' (OECD, 2008b:10); relational capital, which concerns the resources arising from the external relationships of the firm with customers, suppliers and R&D partners; and structural capital, which refers to the knowledge that stays with the firm 'after the staff leave at night' , e.g.organisational routines, procedures, systems, cultures and databases.The imprecision of this definition is not really improved by other analysts who define organisational capital as whatever makes a group of people and assets more productive together than apart, a definition that assumes that alternatives exist, both for individuals and for society as a whole, between producing individually and producing collectively (Jovanovic & Rousseau, 2001).
In the two last decades, the valuations of financial markets have increasingly taken precedence over accounting ones.In addition to disclosed intangible assets (especially when they have been formally protected through trademarks, patents or copyright), another category has increased in importance.This is 'goodwill' , which is defined as the amount above the fair net book value (adjusted for assumed debt) that is paid for an acquisition, and which accounts for a significant share of intangible assets, but on which accurate information is not available from companies 24 .As observed by experts '[our] research shows there is very little disclosure of the nature of goodwill.Only a few companies have provided brief details of what the goodwill is, but this is quite uninformative as it lacks any real analysis or insight' (Intangible Business, 2008).Goodwill is said to include the workforce in place and group synergies, which represent cost savings.Most companies have chosen not to recognise acquired intangible assets separately, but include them within 'goodwill' .Goodwill is then reported in the company accounts (and charged in the income statement).
Finally, 'Undisclosed Intangible Asset values' , which have no reality, even in company books, make up the bulk of intangible assets.They are defined as 'reflecting the overall premium attached to quoted companies by investors' 25 .In more concrete terms, they reflect the mood of the (financial) markets, and when the 2007 financial meltdown began they were severely discounted.

24
The International Financial Reporting Standards IFRS 3, Business Combinations, defines goodwill as follows:: ' A payment made by the acquirer in anticipation of future economic benefits from assets that are not capable of being individually identified and separately recognised' (IFRS 3: paragraph 52). 25 Could they even be 'disclosed' to external scrutiny?See a comment by the authoritative International Accounting Standards Board: 'Greater transparency in the accounting for intangible assets would potentially undermine the competitive advantage that (…) some entities presently derive from unrecognised and undisclosed (secret) intangible assets that, if presented in the financial report, would cease to provide the level of benefits that would otherwise be expected' (IASB, 2007).
Passing the buck: corporate restructuring and the casualisation of employment

Some appealing features of intangible assets
The large-scale offshoring of intermediate services 26 , of which a major share is made up of intangible assets, further complicates the compilation of statistics on cross-border flows for national accounts.One consulting company has estimated that, just for branding, the increase in value that would be created by moving the most valuable 500 brands in the world to offshore locations would be $700bn, a 30% uplift in value and equal to the $700bn spent by Hank Paulson in his US bank bailout (Brand Finance, 2009).
There are many additional reasons, other those usually described as linked to the transformation of the production process and to the rise in the role of knowledge, why trade in intangibles has become so fashionable.Intangible assets exhibit at least three features which are appealing to TNCs' finance-oriented management and shareholders.
Firstly, as is self-evident, intangible assets have no materiality; they offer opportunities, including profit-shifting (below) and invoices corresponding to transfers of an output the physical reality of which is impossible to trace by external observers 27 .As noted by a top world consulting company, there is scope for large TNCs to find an opportunity for increasing the active management of their transfer pricing policy.This construction is useful in the inter-company pricing context when the parent wishes to conduct R&D in several countries but wishes to retain legal ownership of the intangibles (and therefore the profit created by the R&D) in a single country.Contract R&D places the risk in the country that will ultimately own the technology (PricewaterhouseCoopers, 2009).
Secondly, intangible assets are often shared across TNCs' affiliates, making it difficult to allocate an exact price and/or value to activities.Not only are prices and quantities generally not recorded separately, but most of them can be considered as services produced collectively, and the production of a 'unit' of output is difficult to define.R&D, by its nature, produces unique products, whereas the measurement of quantity and prices generally relies on standardised products with prices that can be repeatedly observed (Economic Commission For Europe, 2009).The case of Sun Microsystems is significant here (Lynch & Clayton, 2003).Attempts by UK statisticians to measure the value of software investment activity to be allocated to that country proved unsuccessful.Firstly, this was because the software developed in Sun UK was used worldwide within the company.A second reason was that much of the internal systems software used in Sun UK was written in North America and Asia.In a generalisation of this case, a manager of the UK Office of National of Statistics stated that 'any attempt to measure software capital formation accurately in a firm like this -except at the level of the whole enterprise group on an international basis -is likely to fail' (Lynch & Clayton, 2003:52).Again, this difficulty in clearly separating the production -and value -of intangible assets expands the opportunities for transfer 26 As trade in intermediate products has increased in recent years, the increase in trade in intermediate services has become higher than that for trade in goods, accounting for over 70% of exports in services (Miroudot, Lanz, Ragoussis, 2009).The share of intermediate to total trade in services is based on twenty OECD countries for which data on trade in intermediate services is available for the entire period 1999-2005. 27 See: 'Member states encounter considerable problems in getting reliable data about these transactions, particularly when intra-group transfers of R&D results are involved' (Economic Commission For Europe 2009).
pricing.Thirdly, their (stock) market value is extremely unstable, paving the way for pure financial speculation.Indeed, intangible assets possess outstanding characteristics of fictitious capital, to the extent that their prices, as evidenced by large fluctuations, are less the reflection of 'real' value existing within the firm than the outcome of a convention which, as explained by Keynes (1971-89) in chapter 12 of his Theory, holds only so long as people go on relying on the maintenance of the convention.
Once again, TNCs are well-placed to draw on the benefits of the integrated space they have constituted, compared to the still territorially-bounded national economies.This is clear in the case of taxes, the raising of which is one of the oldest attributes of national sovereignty.There is general agreement on the fact that intangible assets are a preeminent driver in profit shifting and location of capital in tax havens.Industries with a high share of intangible assets, notably the pharmaceutical and drugs industry and the computer and electronic equipment industry, set the pace in profit shifting (Gravelle, 2009).Dischinger and Riedel (2008) found that a decrease in the tax rate by 1% raised a subsidiary's level of intangible assets by 1.6% among European multinationals between 1995 and 2005.Likewise, Grubert has estimated that about half of income shifting was due to transfer pricing of intangibles and most of the remainder to the shifting of debt (2003).
What is the content of the intangible services traded at the international level which ease this profit shifting?R&D, software, Intellectual property (patents), accounting and management and marketing, are the main services traded between and within TNCs.
Here are two examples.First, there has been a dramatic increase in 'management fees' .This rise is all the more daunting if we take into account that the term 'management fee' is often used rather loosely to describe any inter-company charge for a transaction that is not clearly either a transfer of tangible property or the right to use an intangible property.The term could include charges paid for general administrative or technical services or payments for commercial services that are provided intra-group from one or more providers to one or more recipients (PricewaterhouseCoopers, 2009).Second, there has been a huge growth in the offshoring and outsourcing of R&D.R&D activities are increasingly offshored and outsourced from contract R&D firms.In the USA, the all-industries ratio of contracted-out R&D to company-funded, company-performed R&D increased from 3.7% in 1993 to 7.8% in 2007.For manufacturers, the ratio reached 8.5% in 2007, up from 3.3% in 1993(National Science Foundation, 2010).
Again, because of the immaterial nature of R&D, there is some evidence that geographical separation of the R&D and patent management locations could be used by executive managers as a tool to reduce the cost of tax paid by the group ('tax planning').From this perspective, it could be tempting, for similar motives, to include activities which are far from meeting what is usually defined as R&D in this category.Conversely, intangible assets are favourably treated in the USA because, as costs, they are eligible for a tax credit (Gravelle, 2009).

Conclusion
This article has argued that exploring finance-production interrelations in TNCs requires a conceptual framework different from the one which is used to analyse firms in general.By using the concept of finance capital, it has tried to lay out some

Figure 1 :
Figure 1: Holdings (parent companies) of French non-financial industrial groups: the reversed pyramid