«Tobias Schmidt (Deutsche Bundesbank) Wolfgang Sofka (Department of Organisation and Strategy, CentER, CIR, Tilburg University) Discussion Paper ...»
Herrera & Minetti, 2007; Benfratello, Schiantarelli & Sembenelli, 2008). We adopt a novel perspective based on information economics theory by challenging the dominant assumption that all banks are equally subject to suffering from information asymmetries in financing private R&D projects. We acknowledge that banks cannot produce the relevant information on technological innovation themselves, although they are uniquely positioned to aggregate the outcomes of the information production of other firms in the industry in the client portfolio, i.e. there exist information externalities (Stiglitz, 2002) originating from heterogeneous client portfolios of different banks. We contrast this perspective by relying on portfolio theory which posits the opposite relationship: Correlated risks in specialized bank portfolios should make R&D investment in client firms less likely (Markowitz, 1991).
Finally, we allow for a proactive role of firms in signaling the value of their R&D activities to banks through patenting, obtaining government R&D subsidies or venture capital investment.
We test this theoretical framework empirically for more than 7,000 firm observations on R&D investments in Germany between 2002 and 2007. Unique access to the database of Germany’s leading credit rating agency on the population of German firms and their main bank relationship allows us to construct novel variables on the overall portfolio of each of the firms’ main bank. We have the rare opportunity to link this information to firm characteristics, R&D investment, patent statistics and venture capital investments based on a direct, non-heuristic link. In terms of industry, firm and variable coverage, the dataset is, to the best of our knowledge, unique in its breadth and representativeness.
The empirical results corroborate our theoretical model. Firm R&D investment is higher if its main bank is either highly diversified or very active in its particular industry. The information externality effect is restricted to sectors in which ther are major uncertainties about commercial application. Firms can shift the threshold of a bank’s risk considerations to lower levels of specialization if they can signal the value of their R&D activities through successful patent activities. Successful applications for government R&D subsidies and venture capital investment, though, are valuable in themselves for firm R&D investment but fail to alter bank risk assessments significantly. On the basis of these findings, implications are derived for academic research, management and policy making.
The remainder of the article is structured as follows. Section 2, following this introduction, outlines our theoretical framework culminating in the derivation of hypotheses in section 3. In section 4 we present our empirical study including data, variables and methodologies. Section 5 presents the results of these analyses followed by derived conclusions in section 6.
We choose information economics and related signaling theory as our main theoretical building blocks (e.g. Stiglitz, 2002; Ahuja et al., 2005). We shall focus more precisely on signaling theory’s relevance to bank financing of private R&D activities. To achieve this we combine research from finance literature on bank lending decisions (e.g. Rajan & Zingales,
2001) with the literature on knowledge production through R&D. We start by modeling the
R&D investment decision of any given firm:
R&D (Industry characteristics, existing knowledge, funds) A firm’s R&D investment decision can be described as a function of the characteristics of its industry, its existing knowledge, and the available funds. This study will focus on available funds, while industry characteristics and knowledge are largely treated as control variables2.
We conceptualize a firm’s available funds as a general liquidity pool from which a firm can draw financial resources for its R&D investment. We explicitly acknowledge that R&D investment competes with other firm functions (e.g. marketing) for these funds. The pool of available funds determines the cost of capital for a company. Firms will invest in projects (R&D or other) only if the expected returns exceed the cost of capital based on a net present
value rational. The pool of funds has three primary components:
Funds (Internal cash flows, equity finance, bank loans)
Most firms rely on internal cash flows for their R&D projects (Kim, Mauer & Sherman, 1998; Bond, Harhoff & van Reenen, 1999; Haid & Weigand, 2001; Harhoff, 1998). When it comes to external financing of innovations, venture capital financing has received a lot of attention in the literature (e.g. Gompers & Lerner, 2001a; Bottazzi & Da Rin, 2002;
Audretsch & Lehmann, 2004). It is generally acknowledged that access to venture capital is constrained for the majority of firms because of limited availability and the highly selective nature of venture capital investors who target a small number of investments with the potential for high returns (e.g. Eckhardt, Shane & Delmar, 2006). Banks, however, as the primary provider of external financing for the vast majority of firms, appear ill-equipped to finance R&D investments (Bozkaya & Potterie, 2008).
Assuming a perfect market for capital, financing R&D investments should not be different from any other investment decision and firms should opt for all projects with a positive net present value (Modigliani & Miller, 1958). However, the assumption does not hold because of the nature of R&D (for recent reviews, see Hall, 2005 and Hall, 2009). The outcomes of R&D are generally uncertain. This uncertainty has two primary dimensions (Amit et al., 1990).
First, there is a substantial degree of technological uncertainty about the success of an R&D project. Materials and procedures are almost by definition new and largely untested.
Probability distributions for the success of an R&D project are difficult or even impossible to Important industry characteristics encompass the degree of competition in the product market requiring investments in new products and processes (e.g. Schumpeter, 1942; Aghion, Harris, Howitt & Vickers, 2001), technological and legal opportunities for appropriating returns (e.g. Teece, 1986) as well as providing technological opportunities (e.g. McGahan & Silverman, 2006). Existing knowledge stocks (e.g. patents or employee skills) allow the firm to benefit from complementarities in their current R&D investments (e.g.
Cohen & Levinthal, 1989; Dierickx & Cool, 1989).
predict at the early stages (Hall, 2005). R&D investments provide very little collateral. Half of all R&D expenditures finance wages for skilled scientists and engineers (Hall, 2005).
Investments in physical research assets and laboratories are often highly specific to a firm or even a project making it difficult to re-deploy, sell or use for others (Herrera & Minetti, 2007). Secondly, there is a large degree of uncertainty about whether the firm will be economically successful with its technologically new products and processes. A significant proportion of product innovations end up as economic failures because they do not meet customer needs or because competitors are quick in their imitation or substitution activities, which erodes margins from the pioneering advantage (Dos Santos & Peffers, 1995; Gourville, 2006).
Even so, research shows that these underlying uncertainties are not equally exogenous to managers and external capital providers. Endogenous uncertainty can be overcome by firm activities over time while exogenous uncertainties exist independently of any firm actions (Folta & O'Brien, 2004; Cuypers & Martin, 2009). Firms perform R&D to resolve endogenous uncertainties through experimentation, testing and simulation. In that sense, R&D is a sequential process in which firms uncover information and reduce endogenous uncertainty at each stage of the process (Roberts & Weitzman, 1981). There is a long time span between the start of an R&D project and the appearance of revenues from it, i.e. when the uncertainty is ultimately resolved and success or failure is apparent to actors outside of the firm. Empirical estimates predict this time duration to be between four and five years albeit with significant differences across industry (e.g. Pakes & Schankerman, 1984). Hence, firms have significant time advantages in discovering potentials or failures within R&D projects oder external partners, from whom the same uncertainties remain exogenous. This gives rise to an information asymmetry that insiders can exploit (Aboody & Lev, 2000, Ahuja et al., 2005).
Banks are even more disadvantaged in this situation than equity investors because banks’ have limited opportunities for directing/monitoring the use of their funds and they do not benefit from any resulting profits beyond the contractually fixed interest rate (Hennart, 1994).
Moreover, in their lending decisions, all banks are equally subject to legal and regulatory constrains imposed by their institutional environment (e.g. Bank for International Settlement, 2005).
Financing R&D investments is therefore characterized by a combination of information that is either unavailable (exogenous uncertainty) or asymmetrically distributed between the firm and its external capital providers (endogenous uncertainty). Appropriate risk premiums for individual borrowers cannot be assessed (Stiglitz, 2002) and firms find themselves creditconstrained because banks will only set high, average risk premiums (Aghion, Fally & Scarpetta, 2007). As a result, the pool of funds available for R&D investment in a firm is deprived of bank financing.
An emerging stream of literature from finance and strategic management is beginning to emphasize heterogeneities among banks and firm’s signaling. The literature on banks and innovation financing is still scarce. Positive relations have been found for the development of the regional banking system (Benfratello et al., 2008) and relationship length (Herrera & Minetti, 2007). Negative relationships stem from government ownership of banks (Sapienza, 2004; Haselmann, Katharina & Weder di Mauro, 2009). Levitas et al. (2009) contribute to strategic management literature. They discern the value of signaling (through patents and distribution agreements) for overcoming financial constraints in small firms by attracting venture capital investors.
3 Hypotheses development
All of the theory presented so far rests on the important assumption that all banks draw from identical pools of information and should therefore suffer from identical degrees of information deficits. We question this assumption and argue that information is distributed asymmetrically between banks, too. We concede that the specific uncertainties related to R&D investments of a particular firm are equally exogenous to all banks. However, information - on technological feasibility as well as market success - is produced by other firms in the same industry. At least parts of the uncertainties are therefore endogenous to these firms. We argue that banks are heterogeneous in their ability to access this information.
What is more, we will set out the opportunities for the focal firm i to signal the value of its R&D activities and influence the availability of bank financing.
3.1 Information externalities from a firm’s main bank client portfolio
Banks differ in their level of engagement with client firms. Boot et al. (2000) present a dichotomy of bank lending with varying levels in between. Transaction lending is closely related to brokerage activities where it is sufficient for the bank to lend based on a standardized transaction. Relationship lending, though, requires borrower-specific information for activities such as screening and monitoring (for a review see Boot, 2000). The relation specificity can provide banks with access to private data about the financed firm which can lead to a quasi monopolistic banking position and superior benefits from future business with the particular client (Boot & Thakor, 2000). We shall focus on a specific relationship, i.e. a firm’s main bank, defined as the bank that a firm considers its primary source for all banking services. Within all relationship lending, main banks are uniquely positioned for acquisition of information about their client firms (Herrera & Minetti, 2007).