Dynamic R&D choice and the impact of the firm's financial strength

ABSTRACT

This article investigates how a firm's financial strength affects its dynamic decision to invest in R&D. We estimate a dynamic model of R&D choice using data for German firms in high-tech manufacturing industries. The model incorporates a measure of the firm's financial strength, derived from its credit rating, which is shown to lead to substantial differences in estimates of the costs and expected long-run benefits from R&D investment. Financially strong firms have a higher probability of generating innovations from their R&D investment, and the innovations have a larger impact on productivity and profits. Averaging across all firms, the long-run benefit of investing in R&D equals 6.6% of firm value. It ranges from 11.6% for firms in a strong financial position to 2.3% for firms in a weaker financial position.

KEYWORDS:

• R&D choice
• financial strength
• innovation
• productivity
• dynamic structural model

Acknowledgments

We are grateful to Eric Bartelsman, Bronwyn Hall, Jordi Jaumandreu, Hans Lööf, Florin Maican, Jacques Mairesse, Pierre Mohnen, Matilda Orth, Hongsong Zhang and seminar participants at the universities Bielefeld (2015) and Kassel (2015) for helpful comments and discussions. We thank the Centre for European Economic Research (ZEW) for providing data access and research support.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCiD

Bettina Peters http://orcid.org/0000-0002-8611-2016

Mark J. Roberts http://orcid.org/0000-0003-4333-293X

Van Anh Vuong http://orcid.org/0000-0002-9977-8335

Notes

1. Roberts and Vuong (2013) provide a nontechnical overview of the PRVF framework.

2. See Hall, Mairesse and Mohnen (2010), Hall and Mohnen (2013), and Mairesse, Verspagen and Notten (forthcoming) for recent reviews of the literature.

3. A positive relationship between cash flow and R&D investment may simply result because both variables reflect common confounding factors, such as growing market demand, and the correlation is not sufficient to indicate financial constraints (Kaplan and Zingales 1997).

4. Olley and Pakes (1996) specified productivity evolution as an exogenous stochastic process Aw, Roberts and Xu (2011) and Doraszelski and Jaumandreu (2013) endogenize the productivity evolution process by letting it depend on the firm's choice of R&D, and PRVF reformulated it in terms of the firm's innovation outcomes

5. Each firm is characterized by three exogenous variables, its capital stock which enters the profit function, its financial strength which enters the cost function for innovation and the innovation and productivity evolution processes, and its industry which enters all of the structural components. To simplify the notation, we suppress these exogenous characteristics and explain the dynamic decision to invest in R&D focusing on the endogenous variables in the model ω and rd. In the empirical model we treat the firm's capital stock, financial strength, and industry as defining an exogenous firm type and solve the firm's value function for each firm type.

6. Creditreform is the largest German credit rating agency. This information has been used as a measure of financial constraints in previous studies by Czarnitzki (2006) and Czarnitzki and Hottenrott (2011). A measure of credit constraints based on the repayment of trade credits has been used in Aghion et al. (2012).

7. In terms of Standard and Poor's rating system, the high category corresponds to ratings above BBB, the medium category to ratings above BB to BBB, and the low category to ratings BB and below.

8. PRVF compare innovation rates for these high-tech industries and a group of seven low-tech manufacturing industries that have much lower rates of R&D investment. They find that, while product innovations are still generally more common, product and process innovation rates are much more similar in the low-tech industries.

9. The benefits also depend on the industry demand elasticity. The elasticity estimates we construct are: chemicals −3.075, machinery −5.078, electronics −3.713, instruments −4.213, and vehicles −4.891.

10. The decline in investment probability, however, is not observed in all industries. Firms in the medium and low financial categories of the chemicals, electronics, and instruments industries, have a higher average investment probability than firms in the high financial category.

11. There are no standard deviations in these cells because the estimate does not vary within a cell. The estimate does not depend on firm productivity or capital stock. In the model, it only varies across firms with differences in industry and financial strength category.

12. In their review of the literature, Hall, Mairesse and Mohnen (2010) report that production function-based estimates of this elasticity vary from 0.01 to 0.25 and are centered around 0.08. Doraszelski and Jaumandreu (2013, Table 5) report summary statistics of the distribution of firm-level estimates for 10 Spanish manufacturing industries. The average over all firms is 0.015, and the average at the industry level varies from −0.006 to 0.046 across the ten industries, with half of the industries falling between 0.013 and 0.022.