«Abstract This paper studies how the ﬁnancial development in a country inﬂuences ﬁrms’ ﬁnancing and growth decisions. We ﬁrst document ...»
Firm Dynamics and Financial Development∗
Cristina Arellano Yan Bai
University of Minnesota and Arizona State University
Federal Reserve Bank of Minneapolis
University of Michigan
This paper studies how the ﬁnancial development in a country inﬂuences ﬁrms’ ﬁnancing and
growth decisions. We ﬁrst document empirically the debt ﬁnancing and growth patterns of
ﬁrms with a large and comprehensive dataset from 22 European countries. We ﬁnd that in economies with worse ﬁnancial development, small ﬁrms grow disproportionately faster than large ﬁrms yet have lower leverage ratios, than those in more ﬁnancially developed countries.
We build a dynamic model where ﬁrms’ dynamics are linked to ﬁnancial institutions. We ﬁnd that ﬁnancial considerations can rationalize quantitatively the diﬀerence in growth rates between ﬁrms of diﬀerent sizes across countries.
This version is preliminary. Arellano: firstname.lastname@example.org; Bai: email@example.com; Zhang:
firstname.lastname@example.org. We thank V. V. Chari, James Hines, Chris House, Patrick Kehoe, Narayana Kocher- lakota, Ed Prescott, Vincenzo Quadrini, and Richard Rogerson for useful comments and suggestions. We thank Katya Kartashova and Jacek Rothert for excellent research assistance.
1 Introduction Do small and large ﬁrms grow at diﬀerent rates across countries? Are growth rate diﬀerences linked to diﬀerent debt ﬁnancing behavior? This paper documents that small ﬁrms grow disproportionately faster than large ﬁrms in less ﬁnancially developed countries. In these countries, small ﬁrms also ﬁnance their assets with disproportionately less debt than large ﬁrms. We then develop a model to study and quantify these ﬁndings where ﬁrm dynamics are linked to ﬁnancial decisions. We ﬁnd that ﬁnancial development can go a long way in rationalizing the diﬀerence in growth rates across ﬁrms of diﬀerent sizes across countries.
Theoretical models of ﬁrm dynamics have shown that ﬁnancing restrictions can have a real impact on ﬁrm size and growth.1 These theories largely predict that small ﬁrms grow faster than large ﬁrms because they are the ones for which ﬁnancing frictions are the more severe. But if ﬁnancial considerations matter for ﬁrm dynamics, examining ﬁrm growth and debt ﬁnancing across countries that diﬀer in their ﬁnancial institutions can prove useful in evaluating the implications of these theories. Hence, in this paper we use ﬁrm level data on debt ﬁnancing and ﬁrm growth across many countries to evaluate and quantify the impact of ﬁnancial development on ﬁrms’ behavior.
In the empirical part of the paper, we document the growth and ﬁnancing patterns of ﬁrms using a large and comprehensive ﬁrm level database from 22 European countries. We focus on the relative growth and ﬁnancing behavior of ﬁrms of diﬀerent sizes across countries with varying ﬁnancial development. First, we ﬁnd that small ﬁrms grow faster than large ﬁrms.
And that this diﬀerence is stronger in countries that are less ﬁnancially developed as indicated by the ratios of private credit to GDP and the availability of credit information of consumers.
Second, small ﬁrms in more ﬁnancially developed countries use more debt ﬁnancing than large ﬁrms relative to those in less ﬁnancially developed countries. Importantly, these ﬁndings are robust to controlling for country, industry or age speciﬁc characteristics. To the best of our knowledge, this work is the ﬁrst to document these relations for comprehensive ﬁrm level datasets that include a large number of small private ﬁrms across many countries.2 The paper then develops a dynamic model of ﬁrms’ growth and debt ﬁnancing to evaluate and quantify the mechanisms by which ﬁnancial development interacts with ﬁrms decisions.
1 See for example Cooley and Quadrini (2001), Albuquerque and Hopehayn (2004), and De Marzo and Fishman (2006) among others.
2 The negative relation between size and growth has been documented for U.S. ﬁrms in Rossi-Hansberg and Wright (2005) and Klette and Kortum (2004).
2 The framework is a stochastic dynamic model of heterogeneous ﬁrms that builds on Cooley and Quadrini (2001). In our model, debt is constrained by the likelihood of default, which varies across heterogeneous ﬁrms and economies with diﬀerent ﬁnancial development.
Creditors oﬀer ﬁrms defaultable debt contracts but have to pay a ﬁxed cost to obtain credit information when issuing any loan. Diﬀerences in these credit costs across economies proxy for diﬀerences in ﬁnancial institutions across countries.
Given the set of loans available, ﬁrms’ choices over new borrowing depend on their loans due, output, investment needs, and precautionary incentives to self-insure for the future.
Firms who have received a sequence of bad shocks are small and face very restricted loan opportunities. These ﬁrms are very sensitive to good shocks, because they can use the additional output to decrease their debt and increase their capital. Firms’ sensitivity to shocks is more ampliﬁed with more severe ﬁnancial restrictions. And in economies with high credit costs, the set of loans available are limited throughout, but disproportionately for the small ﬁrms. Hence, in less ﬁnancially developed economies, small ﬁrms portray simultaneously disproportionately low debt to asset ratios and high growth rates.
We use the ﬁrm level data of Bulgaria and the United Kingdom to calibrate our model such that we match the ﬁnancial structure observed in the cross section of ﬁrms in each country. We choose these countries as representative countries with weak and strong ﬁnancial development. We show that our model can deliver the relationship between sales growth rate and ﬁrm size observed in the data in both countries. For Bulgaria, we calibrate the leverage ratios of ﬁrms as in the data; for small ﬁrms in the ﬁrst asset quintile to be 0.45 and for the mean size ﬁrm to be 0.53. The model then delivers the observed sales growth diﬀerences of 0.77 for the small ﬁrms and 0.40 for the large ﬁrms, which matches the data counterpart of 0.73 versus 0.39. For the United Kingdom, we do a similar exercise. We calibrate the leverage ratios of ﬁrms as in the data; for small ﬁrms in the ﬁrst asset quintile to be equal to 1.18 while and for the mean size ﬁrm to be equal to 0.84. The model generates a growth rate of small ﬁrms equal to 0.17 while 0.08 for large ﬁrms, which are in line with the data counterparts of 0.23 and 0.05 for small and large ﬁrms respectively. Hence, we conclude that ﬁnancial considerations are important determinants of the diﬀerential growth rates observed across ﬁrms of diﬀerent sizes.
We also perform the comparative static of changing the degree of ﬁnancial development in our model economies. In particular, we decrease the credit costs in Bulgaria to the British level. We ﬁnd that following this experiment, the diﬀerence in growth rates between the smallest and largest ﬁrms decrease from 37 percentage points to 18 percentage points. Lower credit costs also increase the diﬀerence in leverage ratios from -21 percentage points to 9 3 percentage points. Thus, our model is fully consistent with the evidence that in economies with better ﬁnancial development the diﬀerence in growth rates of small versus large ﬁrms is lower yet the diﬀerence in leverage ratios of small versus large ﬁrms is bigger.
Changing credit costs also has a level eﬀect for the overall average debt ﬁnancing and for the output of the small ﬁrms. In this experiment, the average debt to asset ratio increases from 0.60 to 0.73 and the output of the small ﬁrms increase by 19%. However, we ﬁnd that although ﬁnancial development appear as a major determinant for the diﬀerential relation of ﬁrms size with growth and leverage across countries, the overall average growth rate and size of ﬁrms are driven by other factors such as the productivity process for ﬁrms.
The paper is related to the literature that studies the implications of ﬁnancial frictions and agency problem for the dynamics and sizes of ﬁrms. Our model is closest to Cooley and Quadrini (2001) who develop a model where ﬁnancing restrictions arise from limited commitment in debt contracts and show that these frictions can deliver quantitatively large diﬀerences in the growth rate of small versus large ﬁrms. However, we concentrate on how the degree of ﬁnancial development across countries can help explain the diﬀerential ﬁnancing and growth patterns of small ﬁrms. Moreover, contrary to Cooley and Quadrini’s model, our model does not impose an upper bound on the level of liquid assets, which limits the precautionary motive for debt use. In addition, we use ﬁrm level data to discipline the extent to which ﬁnancial considerations impacts growth rates. Our paper is also related to Albuquerque and Hopenhayn (2004) who focus on the eﬀects of enforcement problems and solve for the optimal state contingent contract. Our environment is diﬀerent from theirs in that we consider an incomplete set of assets. Incomplete markets allow for ﬁrms in the model with a history of bad shock to decrease their value while increasing debt holdings through time and allow precautionary savings to play a role.
Clementi and Hopenhayn (2006) and Quadrini (2004) study ﬁnancial constraints that arise due to informational asymmetries between the lender and the entrepreneur and show that moral hazard considerations can rationalize borrowing constraints, which makes investment sensitive to cash ﬂows. They show that information asymmetries can also provide a rationale for the relation between growth and size. De Marzo and Fishman (2007) study the eﬀects on investment within a large set of agency problems between equity and debt holders. They ﬁnd that the positive relation between investment and current and past cash ﬂows is robust to a large class of optimal contracting models that feature informational asymmetries. Moreover, they ﬁnd that leverage ratios are larger for small ﬁrms. Our model generates a similar 4 dependency of investment to cash ﬂow due to imperfections in ﬁnancial markets; however, our model provides a rationale for why small ﬁrms can have higher or lower leverage ratios relative to large ﬁrms depending on the degree of ﬁnancial development.
The paper is also related to a vast literature on the optimal capital structure of ﬁrms in corporate ﬁnance, which is motivated by the classic Modigliani-Miller irrelevance propositions.3 Among various strands of theories on capital structures, our model falls into the dynamic trade-oﬀ theory. Two close works in this strand of literature are Hennessy and Whited (2005) and Miao (2005). Hennessy and Whited (2005) develop a dynamic trade-oﬀ model with the presence of taxes and ﬁnancial distress costs. They model corporate taxes to be progressive which induces larger ﬁrms to use debt more because they face higher tax rates. Interestingly, this theory is at odds with the data in the United Kingdom where corporate taxes are progressive, yet the relation between size and leverage is negative. Miao (2005) develops a model where ﬁrms choose debt and exit decisions to optimally respond to stochastic productivities. However, the debt decision is static as all ﬁrms choose only the same perpetuity level of debt when they enter. In his model, small ﬁrms with low productivity shocks have higher leverage ratios because their equity value is small. In our paper, the ﬁrm’s debt decision is dynamic, ﬁrms can choose to exit, and the interest on debt is time varying, reﬂecting endogenous default probabilities. In addition, our framework rationalizes that small ﬁrms can have smaller debt to asset ratios because debt is more constrained for them given their higher default probabilities.
2 Empirical Facts In this section, we study the empirical relation of ﬁrm size with debt ﬁnancing and growth across countries. Contrary to other studies on growth and ﬁnancing that have analyzed individual countries or only the public ﬁrms, our work documents these relations for comprehensive ﬁrm level datasets that include a large number of small private ﬁrms across many countries. We ﬁnd that these relations vary systematically with the degree of ﬁnancial development across countries as follows. First, in countries with stronger ﬁnancial development, small ﬁrms use more debt ﬁnancing than large ﬁrms relative to those in countries with weaker ﬁnancial development. Second, small ﬁrms tend to grow faster than large ﬁrms in all countries, but by more in countries with weaker ﬁnancial development.
We ﬁrst describe the ﬁrm-level database, Amadeus, which we use for the analysis of ﬁrms in 22 European countries. We then highlight our ﬁndings with two example countries: the 3 See Harris and Raviv (1991) for a comprehensive review.
5 United Kingdom and Bulgaria. The United Kingdom is much more ﬁnancially developed than Bulgaria. Relative to large ﬁrms, small ﬁrms use more debt ﬁnancing in the UK, but less in Bulgaria. Small ﬁrms grow faster than large ﬁrms in both countries, but by more in Bulgaria than in the UK. We then present our main empirical ﬁndings regarding the debt ﬁnancing and growth patterns of ﬁrms in 22 European countries that vary in their ﬁnancial development.