2 Corporate Governance and Institutional ... - CiteSeerX

2

Corporate Governance and Institutional Complementarities: the Co-Evolution of ‘Finance’ and ‘Technology’ by

ANTONIO NICITA*

and

UGO PAGANO°

ABSTRACT

In recent years, an extensive body of studies have dealt with the issue of convergence versus diversity in corporations’ ownership and control in contemporary economic systems. Recent works have emphasized the role of historical conditions and legal origins in shaping path-dependency and diversity in corporate governance patterns, while some others have announced ‘the end of history’ in corporate governance models. In this paper we argue that, despite the a high degree of uniformity achieved by corporate law reforms in many developed countries, the question of diversity in corporate governance is still an issue, as far as the emergence of institutional complementarities among corporate governance domains pushes towards self-reinforcing equilibria shaped by local historical conditions. We apply the notion of institutional complementarity to the study of the well-known trade-off between equity and debt financing. While the ‘Transaction Costs Approach’ (TCE) to corporate governance and finance considers the choices in financial domain as endogenous adaptation to a given technological domain, we outline that an opposite direction of causality may also hold when technological choices are an endogenous adaptation to given financial choices. Moreover, when both the directions of causality hold, some self-enforcing equilibrium across the two domains could prevail. Our simple setting thus provides a framework within which explaining a comprehensive theory which both explains ambivalent tensions towards diversity either towards convergence of existing corporate governance models. Keywords: transaction costs, corporate governance and finance, de-quity, institutional complementarity, organizational equilibria. JEL: D23, G32, G33, L22, L23

* °

DEPT. OF ECONOMICS, UNIVERSITY OF SIENA, [email protected] (corresponding author). DEPT. OF ECONOMICS, UNIVERSITY OF SIENA, AND CENTRAL EUROPEAN UNIVERSITY, BUDAPEST, [email protected]

We thank A. Dalmazzo, R. Tamborini and participants to the thirteenth Annual Meeting of the American Law and Economics Association 2003, for helpful comments. Usual disclaimers apply.

3 1. Introduction In recent years, an extensive body of studies have dealt with the issue of convergence versus diversity in corporations’ ownership and control in contemporary economic systems. Most of these works have compared corporate governance models under the lens of the New Institutional Economics (NIE) theory of the firm, stressing the role of the legal nature of corporations beside the traditional agency costs theories (Allen and Gale, 2000; Becht, Bolton and Röell, 2002). The main question addressed recently by a wide scholarly literature is whether one or another national corporate governance system possesses some relative competitive advantage in the global market so as to overcome any national diversity towards a global model of governance which shapes markets and firms. According to this literature two main systems of corporate governance might be distinguished (Bratton and J.A. McCahery, 1999; Allen and Gale, 2000): a market system characterised by dispersed shareholding and thick, liquid trading markets, and a hierarchical control system characterised by a hard control exerted over the management by a principal or a coalition of principals (banks, families, etc.) and thin trading and non-controlling stakes. While the former system may be found in Usa and UK, the latter has been experienced, with many differences, in Germany, in Italy and in Japan, among other countries. Some of these systems, as the German and the Japanese systems are now subject to economic and institutional crises. This leaves unsolved the problem of convergence versus diversity in corporate models. Recent works have emphasized the role of historical conditions and legal origins in shaping path-dependency and diversity in corporate governance patterns (Bebchuk and Roe, 1999: Schmidt and Spindler, 2002; Beck, Demirgüç-Kunt and Levine, 2003; Djankov, Glaeser, La Porta, Lopez-de-Silanes and Schleifer, 2003), while some others have announced ‘the end of history’ in corporate governance models (Hansmann and Kraakman, 2003). In this paper we argue that, despite the a high degree of uniformity achieved by the recent wave of corporate law reforms in many developed countries, the question of diversity in corporate governance is still an issue, as far as the emergence of institutional complementarities among corporate governance domains pushes towards self-reinforcing equilibria shaped by local historical conditions. In this respect, diversity of governance

4 systems calls for an explanation of path-dependency phenomena in governance as in financial structures which shape, at the same time, firms and markets, sheltering national systems from external competition (Bebchuk and Roe, 1999; Hall and Soskice, 2001; Aoki, 2001; Schmidt and Spindler, 2002). Corporate governance changes are not merely financial either technological matters, since they occur in a given institutional framework, in which economic, legal and organisational issues are bundled in a complex institutional order, shaping all the relevant agents and their actions (La Porta, Lopez-de-Silanes, Schleifer and Vishny, 1998; Becht, Bolton and Röell, 2002) and crafting “institutionalized linkages between the organization domain and the financial transaction domain [..] (Aoki, 2001)”. In this paper we apply the notion of institutional complementarity (Milgrom and Roberts, 1990; Topkis, 1998; Aoki, 2001) to study the relationship between corporate governance and corporate finance in order to enlighten new insights in the well-known trade-off between equity and debt financing. Our focus here is not on agency costs rather on an extension of ‘Transaction Costs Approach’ (TCE) to corporate finance. Williamson (1988) emphasized the relationship between corporate governance and corporate finance, considering the financial choices in corporate governance as endogenous adaptation, in a world of incomplete contracts, to technological choices. According to the TCE’s framework, projects for which physical asset specificity is low are easy to finance with debt and ought to be financed by debt, whereas as asset specificity increases equity should be the preferred financial instrument. This is due to the intuition that, as the degree of assets specificity increases, redeployability in alternative uses (or asset ‘liquidity’) is fairly limited. As a consequence, pre-emptive claims of the bondholders against the investments afford limited protection, in the case of bankruptcy1. In order to finance projects characterised by high levels of specificity the board of directors should thus switch to the selective intervention allowed by equity finance. Since holders of common stocks are the firm’s ultimate residual claimants, entitled to get what is left after everyone else is paid in the event of bankruptcy, they have a very limited interest in the risks associated to the degree of assets specificity to be financed. The main result of the transaction-cost approach is that, as the 1

Another explanation based on debt capacity constraints is developed by Hart and Moore (1994).

5 transactions costs become relevant in the analysis of corporate finance, a new governance structure, called de-quity, might be implemented. This governance structure combines the best properties of debt and equity and allows some form of selective intervention which in turn enables the firm to select the appropriate combination of debt and equity which provides the appropriate degree of assets specificity. This conclusion leads us to pose the following questions: why don’t we observe, in a world of growing interdependence and globalisation, a convergence towards a unique model of ‘dequity’? Why, instead of observing such a convergence process we still face persistent diversity of corporations’ ownership and control systems? We provide two possible explanations: (i) the emergence of path-dependency between ‘governance’ and ‘finance’ in corporate systems; and (ii) the structuring of multiple organisational equilibria2 as institutional complementarities between the degree of assets specificity in the firm (i.e. its technological structure) and its financial structure. While the transaction-cost approach assumes the existence of a infinite elasticity of substitution between debt and equity financing, depending on the nature of the assets to be produced by the investments financed, we outline that financial markets may differ from one country to another and access to financial market might be inhibited thus implying that initial conditions of financial markets may strictly affect firms’ technical choice and even influence the emergence of some self-enforcing equilibrium between technology and finance in firms. In order to illustrate this point we first show that, with reference to the TCE’s approach, an opposite direction of causality may also hold, i.e. that the nature of financial markets may affect assets’ specificity degree within the firm3. We show how institutional complementarities between finance and technology may result either in unique or in multiple self-reinforcing equilibria financial and technological choices and explain the diversity that we find in real-life institutions. The paper proceeds as follows. In the following section, we briefly recall the theory of Williamson (1988) on the relationship between finance and governance in 2

See U. Pagano (1993), Pagano and Rowthorn (1994, 1996). This explanation is based on the degree of competition occurring in financial markets. However, also the degree of competition in the goods markets – which may differ among countries – might well be much more important. See Allen and Gale (2000). 3

6 firms; in section three we extend the TCE approach by considering the nature of interdependency between the choices made in technological domain and of those made in the financial domain within a corporate structure; in section four we propose a simple formalization that tries to capture the fundamental institutional complementarities between finance and technology in firms. In section five we draw the main conclusions. 2. Corporate finance and corporate governance: the TCE’s approach The TCE’s approach to corporate finance is mainly based on the extensions of the transactions cost minimizing paradigma to the analysis of the financial attributes of different transactions. Williamson’s (1988) starting point is to move away from standard studies based on a composite-capital set-up towards a study of the investment attributes of alternative corporate projects so as to regard debt and equity as governance structures rather than as merely financial instruments. The intuition behind this, is to evaluate the emergence of alternative financial instruments in terms of their compared transaction costs, rather than on the basis of their composite-capital characteristics. The main result is that two apparently disparate phenomena such as those of debt-based and equity-based financing – beside internally raised funds - might be analysed under the very same transaction costs economizing scheme. According to Williamson, what distinguishes debt and equity – considered as polar financial devises - is the asset specificity characteristics of investment projects. Let us assume that in a given corporation there is only one form of financing, debt, and that the management has to choose among several alternative projects, within the investment set J= (j1, j2, …., jn), ordered in terms of their degree of asset specificity, so that the costs H to be sustained in order to redeploy the project in alternative uses increases as assets involved in the project become more specific with ∆H(jn)=[H(jn)− H(jn-1)]>0 being an indicator of the degree of specificity or irreversibility of the investment jn with respect to the investment jn-1. Suppose further that the investment set is given by J=(j1, j2) with ∆H=[H(j2)− H(j1)]>0, i.e. with j1 being a general-purpose investment and j2 being a specific investment. Thus suppose that our management has to finance the project j1. In the simplest case debt financing might be defined as a financial

7 claim which imposes on the firm an obligation to pay a specific amount (by stipulated interest payments to be transferred at regular intervals) either to face the risk to be forced into bankruptcy.

In the event of bankruptcy, scheduled payments result in debt-

reorganisation or in liquidation. In this last case, firm’s assets are liquidated and their liquidation degree depending on their re-deployability in alternative uses. As a consequence the nature of physical assets plays a very important role in the event of bankruptcy, given that the value of a pre-emptive claim declines as the degree of assets specificity deepens. Under this framework the cost of debt financing increases, not only as long as ‘bet capacity’ declines, but also as assets specificity increases. If the only possibility for a firm to finance its projects

is given by debt financing, the more

specialised investments will require growing costs and the firm might be forced to some investment rationing, in favour of greater assets’ re-deployability. Assets’ redeployability acts as a powerful safeguards for bondholders, given that a debt contract typically imposes a clause which shifts, under the realization of some contingency, property rights on firm’s assets in the hands of creditors (Aghion and Bolton, 1992; Shleifer and Vishny, 1992; Triantis and Daniels, 1995; Hart and Moore, 1995). In this case, bondholders could liquidate the assets and thus recover the amount financed: the higher is the degree of asset re-deployability the higher will be the safeguard for debt financiers. As a consequence, for a given financial constraint of the firm, the higher is the costs of debt, the greater is the probability of selecting a project characterised by a very low degree of assets’ specificity (a general-purpose project). In any case, regardless of any consideration about the costs of debt financing, this form of financing truncates the set of possible investments to be financed due to the budget constraint if the firm. Now let us assume that another financial instrument, equity, is available and that firm could also recur to this form of financing in order to realise a given project. The main difference with debt financing, in the simplest form, is that equity does not have to be repaid and it is junior to debt, given that holders of common stock are the firm’s ultimate residual claimants. According to Williamson, the Board of Directors, elected by the prorata votes of those who hold tradable shares, “evolves as a way by which to reduce the costs of capital for projects that involved limited redeployability”. The Board of directors indeed has the power to replace management and to monitor operating investment and the

8 way in which the firm is managed4. In a sense, while debt financing is more market-like kind, equity implies some form of (vertical) integration and the shareholders are more interested in investments that are characterised by higher returns in no-bankruptcy event, rather than in higher liquidity in the bankruptcy event. According to Williamson, for any given set J of investment projects, we can define as D(∆H) and E(∆H), respectively, the costs of debt and equity financing, with D(0)E’(∆H). The expressions above mean that, when assets are highly redeployable, equity is more costly than debt, since it involves higher transaction costs (implementing a rule for governance). By contrasts, when assets become more specific, debt financing is more costly, since it induces the firm to face its budget constraint and to truncate the range of potential projects to be financed and realised. Now, let ∆H* be the value for which D(∆H)=E(∆H), thus the optimal financial choice for a given project J is to use debt finance for all project for which ∆H∆H*. This leads us to the formulation of a new financial instrument or governance structure called de-quity. Under ‘dequity’, the best properties of debt and equity are combined. According to the degree of assets specificity the firm can use discretionally alternative financial instruments so as to minimise the transaction costs involved in the realisation of a given project. Several contributions5 have been developed along the lines traced by Williamson and all of them share a very important feature: as long as the relationship between Finance and Technology becomes pervasive, governance problems, i.e. the problems 4

Williamson’s analysis neglects in this formulation the role played by minority shareholders as the possibility of having a complex system of control over the management. 5 Similarly to Williamson, Shleifer and Vishny (1992) developed a model which shows how investments and/or acquisition of specific assets is very difficult to finance by debt financing. Zeckhauser and Pound (1990) estimated the value of assets specificity in some US industries, showing how assets specificity implies a higher difficulty in monitoring management with respect to general purpose investments. P. Worthington (1995) found that the effect of cash flow on investments is larger in industries whose capital expenditure are likely to be “highly sunk” than in low capital industries. Worthington interpreted this finding as evidence that external financing of capital investment is more difficult when the assets being financed have low recovery (resale) values or are sunk (specific). Vilasuso and Minkler (2001) develop a dynamic model that incorporates the insights of both the agency cost and asset specificity literature about corporate finance. In general, they find that neither can be ignored, and that the optimal capital structure minimizes agency cost and asset specificity considerations. A key finding in their work is that the conditions most favorable for reducing transaction costs due to asset specificity are the same as those for reducing the agency costs of debt. Empirically, they find that agency costs and asset specificity are

9 related to the mis-alignment of managers’ incentives, loose their centrality and are residually solved. Since, in the Williamsonian approach, the technological structure of the firm endogenously selects the most efficient financial structure, financiers are able to ‘control’ ex-ante the nature of the assets to be realised by the investments financed. This ‘pre-commitment effect’ between technology and finance reduces the problem of managers’ moral hazard, which represented a central issue in traditional agency costs analyses of the financial structure of the firm. However, the TCE’s approach contracts with the evidence of multiplicity and diversity in modern corporate governance models. Instead of observing a convergence process towards a de-quity structure we still face persistent diversity of corporations’ ownership and control systems. Several authors (Bebchuk and Roe, 1999: Schmidt and Spindler, 2002; Beck, Demirgüç-Kunt and Levine, 2003) have recently suggested that initial conditions on financial markets, technology and legal system may strictly affect corporate structure. However most of these explanations fail in providing a comprehensive theory which both explains ambivalent tensions towards diversity either towards convergence of existing corporate governance models. In the next sections we try to provide a framework which takes into account the emergence of multiple selfenforcing equilibria between ‘technology’ and ‘finance’ by a simple formalization of the notion of institutional complementarity between the two domains.

3. Institutional Complementarities between ‘Technology’ and ‘Finance’ The TCE approach recalled above constitutes an original way to show the underlying relationship between the degree of asset specificity within the firm and the Debt/Equity ratio characterising its financial structure. However the notion of ‘de-quity’ introduced by Williamson and the associated ‘manager discretionality’ craft a general static case in which past investments and financing decisions have negligible effects on future manager’s choices. The analysis provided by Williamson (1988) describes a direction of causality moving from asset specificity to firm’s financial structure, and then to firm’s governance. The notion of , seeking appropriate safeguards for financial investments in the firm, could ‘influence’ the emergence of generic (re-deployable) or specific assets. However, if both the directions of causality between

significant determinants of a firm's capital structure in the transportation equipment and the printing and

10 technology and finance hold, then some self-enforcing equilibrium could prevail institutional complementarity relies on the idea that, in a given institutional framework, economic agents face different domains and do not strategically coordinate their choices across domains games. As a consequence, the institutional choices in one domain act as exogenous parameters in other domains and constitute the ‘institutional environment’ under which choices are being made.

In Williamson’s (1988) approach the only domain over which the firm makes its choices is the financial domain. Technological parameters constrain the financial domain but they are not the result of a deliberated choice in the technological domain. However there it would be easy to show that an opposite direction of causality, between Technology and Finance, could also work: financiers, seeking appropriate safeguards for financial investments in the firm, could ‘influence’ the emergence of generic (re-deployable) or specific assets. Thus not only the technological structure may affect financial choices, but also the financial structure may affect technological choices. When both the directions of causality between technology and finance hold, some self-enforcing equilibrium could prevail. One way of extending the TCE approach is thus to jointly analyse the interaction between financial and technological domains within a firm. Let us assume two independent domains: a) financial domain (F): financial experts and intermediaries choose the financial structure of the firm (including the appropriate ratio of equity and debt) given the type of assets or investments that characterize the technology of the firm. b) technological domain (T): managers on behalf of the present financiers choose the firms’ investments (including the degree of specificity characterizing its technology) given the nature of existing financial structure.

In the financial domain, financial intermediaries make their choices regarding the best financial governance of the firm, according to the existing technological structure. When assets are generic, and when the market value of generic assets is proportionate to the value of debt, financial intermediaries will force a financial policy governed by debt financing: in this case, beside the periodic interest payments received, debt-holders may obtain with certainty, in the event of bankruptcy, a financial return based on the market value of re-deployable assets. In financial markets the firm will be very attractive for individuals lending capital that are truncated in their riskrelated earnings and wish to minimize the risks associated to default. By contrast it will turn out to be less attractive for share-holders that are not truncated in their earnings and cannot avoid to lose a part (bounded by limited liability) in case of default. The opposite situations when assets are specific. In this case financial intermediaries will ‘suggest’ a financial policy governed by equity financing. Choosing a debt financing policy will raise very high transaction costs in terms of the appropriate contractual safeguards to be created for debt-holders that are in any case truncated in the extra-earnings generate by specific assets. By contrast the firm will be attractive

publishing industries.

11 for equity financiers who can get a share of the extra-value generated by specific asset and, in any case, face a default risk associated to their equity capital. In the technological domain, managers make their choices regarding the best technological structure of the firm, according to the existing financial structure. When the firm is driven in the interest of debt financiers, in order to minimize ex-post transaction costs, production managers will select generic investments, to an extent that generates a market value for generic assets which is proportionate to the value of debt. By contrast, when the firm is driven in the interest of equity financiers, production managers will select (specific) investments which maximize the internal value of the firm so as to transfer to shareholders an extra-value to their investments. As we can see, the technological domain shows that an opposite direction of causality may also hold, with respect to that shown by the Transaction costs approach: financiers, seeking appropriate safeguards for financial investments within the firm, would select generic (re-deployable) or specific assets according to their preferences on expected residual return. Assume, for instance, that the ‘type’ of financier selected is willing to finance a low risk investment. By consequence, he will finance an investment characterized by a positive rate of liquidity in the case of bankruptcy, showing a direction of causality which moves from technology to finance. However, as it happens, this means also that the degree of asset specificity will be determined by the ‘type’ of financier selected, suggesting an opposite direction of causality with respect to the one suggested by Williamson, for which ‘finance’ affects ‘technology’. In this context, generic/specific capital ratio and the debt/equity ratio are institutional complements in the sense that a high values as well as low values of both ratios fit each other. This means that we can have some multiple equilibria in corporate governance and finance and that some self-enforcing equilibrium (even Pareto-inferior) could prevail and, by cumulative causation, persist over time. This result sharply contrasts with the flexibility features of ‘dequity’ financing in Williamson (1988). In Williamson’s analysis the emergence of a flexible instrument ‘de-quity’ is given by the fact that one domain is considered as exogenous so that there is no institutional complementarity but only a constrained choice. However, when we consider the institutional complementarity between Technology and Finance, some path dependent equilibrium could prevail. The interdependence among domains may generate multiple institutional arrangements and initial conditions may affect the available choices in one domain so as that some Pareto-inferior institutional arrangement may emerge. In the next section we provide a simple formalization of this interdependence.

4. A simple formalization Assume that a firm’s technological structure is given by h = k/K (with K≠0, k≠0), where K indicates the stock of specific assets, whereas k denotes of general-purpose assets. The technological choice domain of the firm is thus given by values of k/K which fall in a range that goes from a very generic technology (Tg) to a very specific one (Ts ) 6. .

Assume also that the financial structure of the firm belongs to a range of values that fall between a shareholder (S) financial scheme (Fs) and a debt or bondholder (B) financial 6

In this particular respect, concerning the choice of the appropriate level of asset specificity we assume that there are no agency costs between production managers and the individuals financing the firm.

12 scheme (Fb). Let z be the economic return generated by general-purpose assets, while Z is the economic return of specific assets7 (with Z≠0, z≠0). Suppose also that Z is the nobankruptcy extra-return received by agent S for financing the project j(K), while zL , with L=(S,B) is the return-share received by each agent in the event of Bankruptcy with 0≤zS0, zS approaches to zero, and vice-versa. Suppose also that the cost of employing one unit of specific or general-purpose assets is given by C and c, respectively. From the assumptions made above it turns out that agents S (shareholders) and B (bondholders) have alternative preferences over the debt/equity ratio: (i)

agent S receives a “bounded return” in the bankruptcy event (denoted by the probability p=0), zS (due to the fact that shareholders have bounded or eventually zero financial claims on firm’s assets) whereas S receives a “unbounded return”, Z>0 in the no-bankruptcy event;

(ii)

agent B obtains an “unbounded return”8 zB in the bankruptcy event, receiving however a “bounded” return z in the no-bankruptcy event.

Pay-offs of agents S and B

pay-offs

p=1 (No-bankruptcy)

p=0 (Bankruptcy)

S

(z+Z) with Z> zS≥0

zS≥0

B

z

zB>0

Let us assume now that managers choose, in the T domain, the amounts of K and k on the basis of a given financial structure (i. e. acting either in the interest of shareholdes or in the interest of the bondholders) and that financial intermediaries choose in the F domain

7 8

All these variables are expressed in monetary units. We assume however that this value cannot exceed the return associate with general purpose assets.

13 the bond-holder or equity-holder typology of the firm on the basis of the generic/specific capital ratio h9. Define US and UB as, respectively, shareholders’ and bondholders’ pay-offs:

(1) U S =

{ p (zk + ZK)

+ ( 1-p) z S k - [c(k) + C(K)]

{

( 2) U B = pz (k + K) + ( 1-p) z B k - [c(k) + C(K)]

}

}

and assume that zS = 0 and zB = z.

(a) Optimal choices in the financial domain (F) Let us first analyze optimal choices made by financial intermediaries in the financial domain (F). Given the existing technology (K,k), and the range of values (z, Z) in the financial domain, intermediaries will choose equity financing when US ≥ UB , that is when

(3)

p(Z-z)K ≥ (1 − p) zk

or:

(3')

p( Z − z ) k ≥ (1 − p ) z K

Vice-versa, debt financing will prevail when:

UB ≥ US that is

9

In the above framework we have explicitly excluded self-financing by internally raised funds as an alternative way to structure investment decisions. As Allen and Gale (2000) show, self- financing is one of the most diffused way of investment financing among firms through different corporate governance systems. In our framework, self-financing could however be introduced as a particular case of equity financing. In this case, internally raised funds can be treated as a particular form of equity which gives no claims in the case of bankruptcy and which is selected when the degree of asset re-deployability prevents any debt contract. According to this assumption, thus, the result outlined above still apply to selffinancing.

14

(4) (1-p )zk

≥ p(Z-z)K

or

(4')

k p( Z − z ) ≥ K (1 − p ) z

Recall that h= k/K is the ratio between the generic and the specific capital of the firm. When firm’s assets are very specific, the value of k/K will eventually approach to zero and this will increase the probability that the value of p(Z-z)/(1-p)z is higher than k/K. On the contrary, when firm’s assets are very generic, the value of k/K will eventually approach to infinity and this will increase the probability that the value of p(Z-z)/(1-p)z is lower than k/K.

Denote now, as above, by F the financial domain where the choice between the financial schemes Fs and Fb is made by the intermediaries and by T the technology domain where the choice between Ts and Tg is made by production managers. Recall that Tg and Ts are such that the asset specificity ratio, k/K, is greater under Tg than under Ts (recall that Tg is the general purpose technology and Ts is the specific technology). From (4) we get immediately the following proposition:

Proposition 1 In the domain F, the additional benefit of having a financial scheme as Fs instead of Fb is greater when a technology as Ts (instead of a more general purpose one Tg) is chosen in the domain T. That is: U(Fs,Ts ) - U(Fb,Ts ) ≥ U(Fs,Tg ) - U(Fb,Tg ).

Proposition 1 tells us what happens to different financial schemes (FS, FB) for given alternative technologies (Ts,Tg). An analogous result could be shown with reference to the additional benefit of having a financial structure as Fb instead of Fs when a technology as Tg (instead of Ts) is chosen in the domain T.

15 Let now investigate what happens to different technologies for given alternative financial structures.

(b) Optimal choices in the technological domain (T) Given the ‘financial structures’ (FS, FB), the degree of assets specificity will be chosen maximizing, with respect to k and K, the following functions10:. b.1) under shareholders scheme Fs:

Max U S =

{ p (zk

+ ZK) + ( 1-p) z S k - [c(k) + C(K)]

}

that implies: (5)

∂U S = pZ − C ' ( K ) = 0 ∂K

( 6)

∂U S = pz − c' ( K ) = 0 ∂k

b.2) under debt-holder scheme, Fb:

Max U B =

{pz (k

+ K) + ( 1-p) zk - [c(k) + C(K)]}

that implies (7 )

∂U B = pz − C ' ( K ) = 0 ∂K

(8)

∂U B = pz + (1 − p ) z − c' (k ) = 0 ∂k

Now, define by KS and kS the arguments that maximize US and by KB and kB the arguments that maximize UB. Comparing (5) and (7) we have that11:

16 (9) KS ≥ KB and comparing (6) and (8) we have that12: (10) kB ≥ kS It follows, thus, from (9) and (10) that: (11)

k kB ≥ S KB KS

Recalling that Tg and Ts are any two technologies such that the generic to specific capital ratio k/K is greater under Tg than under Ts, the (11) implies the following proposition.

Proposition 2. In the domain T the additional benefit of having a more general purpose technology Tg over a more specific technology Ts increases when a debt-bondholder scheme FB (instead of shareholders rights FS ) is chosen in the domain F. That is: U(Tg, Fb) - U(Ts, Fb) ≥ U(Tg, Fs) - U(Ts, Fs).

Observe that proposition 1 and 2 imply that the choices made in the financial domain and in the technological domain satisfy the standard super-modularity conditions, recalled in section two. This implies that multiple financial equilibria (Fs ,Ts) and (Fb, Tg) are possible, where (Fs ,Ts) is characterized by the complementarity of shareholders financing and specific technology and where (Fb, Tg) is characterized by the complementarity of debt-holders financing and general purpose technology. Let now show the conditions under which the multiplicity of financial equilibria occurs.

10

We continue to assume that zS = 0 and zB = z. Since Z≥z. 12 Since (1-p)z ≥ 0. 11

17 A shareholder financial equilibrium (Fs ,Ts) is defined by the set of values for which shareholders financing FS brings about the highest value of the firm given a technology Ts and, in turn, a technology Ts maximizes firm profits under the shareholders financing FS . This occurs when the values of the arguments (kS,KS), that maximize (1), satisfy also (3'), that is when

k p( Z − z ) ≥ S . (1 − p) z KS

(12)

A debt-holder financial equilibrium (Fb, Tg) is defined by the set of values for which debt-holders financing FB brings about the highest value of the firm given a technology Tg and, in turn, a technology Tg maximizes firm profits under FB . This occurs when the values of the arguments (kB,KB) that maximize (2) also satisfy (4'), that is when:

(13)

kB p( Z − z ) ≥ . KB (1 − p) z

Denote now ERGS =

p( Z − z ) (1 − p ) z

as the ratio between the firm’s expected extra-return coming from specific investment K and the return coming from general purpose investments k, given the bankruptcy probability (1-p). The ratio kS/KS represent the asset specificity ratio relative to values of k and K associated to the more specific technology Ts operated by shareholders, whereas kB/KB are the relative values of k and K associated to the more general purpose technology Tg that is operated by the debt holders. Because of (11), ERGS must either fall within the range of values defined by kS/KS and kB/KB or in the range defined by 0 and kS/KS or in that defined by kB/KB and infinity.

18 Thus we have the following proposition on the existence of multiple financialtechnological equilibria. Proposition 3 (multiple equilibria). Multiple financial-technological equilibria (Fs,Ts) and (Fb,Tg) exist when ERGS falls between the values kS/KS and kB/KB . Proposition 3.1 (unique equilibrium). A unique debt-holder equilibrium (Fb,Tg) exists when ERGS is smaller than kS/KS while a unique shareholder (Fs,Ts) equilibrium exists when ERGS is greater than kB/KB. Proof. Propositions 3 and 3.1 follow from the following. When

(14)

kS p( Z − z ) k B ≤ ≤ (1 − p) z K B KS

both (12) and (13) are satisfied.

When

(15)

kS k p( Z − z ) ≤ B ≤ KS KB (1 − p) z

then (12) is satisfied but (13) is not.

When

(16)

k p( Z − z ) k S ≤ ≤ B (1 − p) z K S K B

19 then (12) is not satisfied whereas (13) it is.

Propositions 3 and 3.1 can be visualized by the following scheme.

0

kS/KS (FB,TG)

∞

kB/KB (FB,TG) or (Fs,Ts)

(Fs,Ts

)

Range of possible values for ERGS =p(Z-z)/(1-p)z

The results coming from the above proposition have an interesting meaning in our setting. The expression in (16) means that when the probability of success for the project (in terms of the bankruptcy event) is low and the ratio between the return of specific and general capital is also low, then only debt-holders financial equilibria are possible. By contrast, when the probability of success is high and the ratio between the return of specific and general capital is also high, then only share-holders financial equilibria are possible. For intermediate values of these parameters, multiple self-enforcing financial equilibria will exist. Initial conditions affect the selection of the institutional complements (F,T). However, initial conditions on financial markets and on the technological structure of the firm can shape future decisions concerning the institutional setting of the two domains, as long as initial financial structure affects technological structure in t=1 and viceversa13.

4. Conclusions

13

In this simple setting we neglect internal funds raised by cash flows self-financing as a third way of financing investments, beside debt and equity. Worthington (1995) compares the trade-off between debt and cash flows financing instead of debt and equity.

20 The emergence of self-enforcing equilibria between (F) and (T) puts together a transaction costs economics explanations with the observations of diversity in corporate governance patterns and given an explanation of the fact that independently of their historical origins (which may be different in different countries) corporate governance models persist in their diversity over time. The diversity of corporate governance models observed in economic systems may thus be explained in terms of the historical conditions governing the path dependency between ‘Technology’ and ‘Finance’. Under this setting, the financial policies reproduce themselves via technology and the technology reproduces itself via the particular financial policy which has originated that technology and so on. Financial equilibria can be interpreted as a way in which technological standards affect financial standards and vice-versa. Network externalities may imply that any pressure to standardize finance will lead to a standardization of technology and vice-versa. According to the way in which we assume that the initial conditions of the system were given, a corporate equilibrium, in terms of the particular combination of (F) and (T) can be interpreted as a "financial equilibrium" or as a "technological equilibrium". If we assume that the initial conditions of the system were given in terms of a "strong" financial markets shock, then an corporate structure can be interpreted as a "financial equilibrium" where the initial condition in financial markets reproduced themselves via technology. By contrast, assume that a technological innovation or a change in the structure of demand has changed the technological characteristics of the resources to be employed. In this case the initial conditions have occurred in terms of a strong technological shock and the structure of the firm, both financial and technological, can be interpreted as a technological equilibrium where the initial technological shock has reproduced itself via an appropriate financial policy. In this respect, independently of their historical origins (which may be different in different countries) the opposite structuring of corporate governance models, such as the American corporate governance system, the German system and the Japanese firm, might be viewed as alternative finance-technology equilibria which endure over time (Hall and Soskice, 2001). These explanations of alternative governance systems, based on the notion of financial

21 equilibrium, might so-far introduce further insights in valuing the question of convergence versus diversity in corporation’s ownership and control systems.

References Aghion, P., Bolton, P., (1992), “An Incomplete Contracts Approach to Financial Contracting”, Review of Economic Studies, 59: 473-94 Allen F., Gale, D. (2000), Comparing Financial Systems, MIT Press Aoki M. (2001) Toward a comparative institutional analysis, MIT Bebchuk L. and M. Roe (1999), "A Theory of Path Dependence in Corporate Governance and Ownership," 52 Stanford Law Review 127-170. Becht, M., Bolton P., and Röell A. (2002) “Corporate Governance and Control”, ECGI Finance Working Paper 02. Beck T., Demirgüç-Kunt A. and Levine R. (2003) “Law and Finance: Why Does Legal Origin Matter?” in Journal of Comparative Economics, vol.31, Issue 4 Bratton W. and McCahery J. (1999) “Comparative Corporate Governance and the Theory of the Firm: The Case against Global Cross Reference”, Columbia Journal of Transnational Law 38, 213-297. Djankov S., Glaeser E., La Porta R., Lopez-de-Silanes F. and Schleifer A. (2003), “The New Comparative Economics” in Journal of Comparative Economics, vol.31, Issue 4 Hall P. and D. Soskice (2001) Varieties of Capitalism: the Institutional Foundations of Comparative Advantage, (edited by) Oxford University Press. Hansmann H. and Kraakman R. (2003) “The End of History for Corporate Law” in Jeffrey Gordon and Mark Roe Eds., Are Corporate Governance Systems Converging? University of Chicago Press Hart O. and J. Moore, (1994) "A Theory of Debt Based on the Inalienability of Human Capital", Quarterly Journal of Economics Hart O. and J. Moore, (1995) "Debt and Seniority: An Analysis of the Role of Hard Claims in Constraining Management", American Economic Review La Porta R., F. Lopez-de-Silanes, A. Shleifer, and R. W. Vishny (1998) "Law and Finance" Journal of Political Economy, Vol. 106 (6) pp. 1113-1155 Milgrom P. and Roberts J. (1990) “Rationalizability, Learning and Equilibrium Games with Strategic Complementarities” Econometrica v. 59 pp. 511-528

22 Pagano U. (1993) “Organisational Equilibria and Institutional Stability” in Bowles S., Gintis H. , Gustafson B. eds. Markets and Democracy. Cambridge University Press, Cambridge Pagano U. (2004) “Legal Positions and Institutional Complementarities” in F. Cafaggi, A. Nicita, U. Pagano (Eds) Legal Orderings and Economic Institutions, Routledge forthcoming Pagano U. and Rowthorn R. (1994) “Ownership, technology and Institutional Stability” Structural Change and Economic Dynamics. Vol.5 No 2 pp. 221-243. Pagano U. and Rowthorn R. (1996) “The Competitive Selection of Democratic Firms” in Pagano U. and Rowthorn R. (eds) Democracy and Efficiency in the Economic Enterprise, Routledge Schmidt R.H. and G. Spindler (2002) “Path Dependence, Corporate Governance and Complementarity” in International Finance 5:3 Shleifer A. and Vishny R. (1992) “Liquidation Values and Debt Capacity: A Market Equilibrium Approach”, Journal of Finance 47(4), pp. 1343-1366 Topkis, D.M. (1998), Supermodularity and Complementarity, Princeton University Press Triantis G. and R. J. Daniels, (1995) “The Role of Debt in Interactive Corporate Governance”, 83 Calif. L. Rev. 1073 Williamson O. (1988) “Corporate Finance and Corporate Governance” (1988), Journal of Finance, 43 Worthington P. (1995) “Investment, Cash Flows and Sunk Costs”, The Journal of Industrial Economics, Vol. XLIII, No.1, p 49-61