Entry and capital structure mimicking in concentrated markets: The role of incumbents’ financial disclosures

https://doi.org/10.1016/j.jacceco.2020.101379Get rights and content

Abstract

We examine whether the public availability of product market incumbents' financial disclosures leads to greater capital structure mimicking of incumbents by entrants. Exploiting a change in disclosure enforcement for German private firms in the mid-2000s, we find entrant-incumbent mimicking rises substantially in concentrated markets once incumbents' financial statements are publicly available. Additional tests exploring potential mechanisms are more consistent with interfirm learning underlying the effect than alternative channels. Our findings shed light on the effects of competitor financial statement disclosure on private firms’ initial financing decisions and highlight how capital structure dependencies among peer firms arise.

Introduction

Little explains firms' capital structure decisions better than peer firms' capital structure decisions. Empirical work suggests that such “mimicking” arises not only because firms share exposure to institutional environments, industry conditions, and other common factors, but also because financing decisions are direct functions of peers’ decisions (e.g., Leary and Roberts, 2014; Frank and Goyal, 2009). Possible reasons for mimicking range from agency and reputation incentives (e.g., Scharfstein and Stein, 1990) to effects of product market competition (e.g., Bolton and Scharfstein, 1990), but central to nearly all explanations is the observability of financial information. Thus, much as information frictions figure prominently in canonical theories about capital structure decisions (e.g., Myers, 1984), these frictions likely have an important role in explaining how mimicking arises.

We investigate the role of public financial disclosure in facilitating capital structure mimicking. We focus specifically on private firms to exploit variation in disclosure mandates that affect whether even basic financial information (including that about capital structure) is available. Our analyses examine effects of incumbents' public financial disclosures on entrants' initial financing policies in concentrated markets. We study entrants’ initial financing decisions because setting these financing policies is difficult but critical; de novo businesses lack industry experience, yet these initial policies shape firm outcomes such as growth and survival and are hard to adjust after entry (e.g., Hanssens et al., 2016; Brav, 2009; Lemmon et al., 2008). Our emphasis on concentrated markets stems from prior work showing that competitive and contracting costs of deviating from industry financing norms are substantial in concentrated markets but are fairly low in dispersed markets (e.g., Chevalier, 1995; Opler and Titman, 1994). Incentives to mimic are therefore amplified in concentrated markets.

Our empirical analyses exploit the unique disclosure environment for private firms in Germany in the late-2000s. Like most other European countries, Germany has long required that all private limited liability firms publicly disclose certain annual financial statement information. Unlike most other European countries, Germany did not enforce the requirement for decades and the vast majority of firms did not comply. In November 2006, under pressure from the European Commission, the German government implemented a strict enforcement regime. This change dramatically altered the information environment by making nearly all firms’ financial statement information publicly available (Henselmann and Kaya, 2009).1 Thus the setting provides a source of substantial variation in the extent of observable disclosures that is plausibly exogenous to firm characteristics that affect financing decisions.2

Our main empirical prediction is that, in concentrated markets, the sensitivity of entrants' initial leverage ratios to incumbents’ leverage ratios increases more for German entrants after the disclosure shock than for entrants in other major European countries where there was no change in enforcement. We use the Bureau van Dijk database Amadeus to identify de novo product market entrants between 2006 and 2009 based on year of incorporation and retain those in concentrated markets. We match entrants to plausible incumbents based on characteristics such as geographic location and industry to construct a simple test of the sensitivity of entrant to incumbent leverage, which we build out to a difference-in-differences (DD) design using French, Italian, and UK private limited liability firms as the control group. Although the enforcement change became effective for fiscal years ending on December 31, 2006, we use 2008 and 2009 as the post period due to a 12-month lag between fiscal year-end and when disclosures usually become publicly available. Fig. 1 maps the timing of the enforcement shock to our design.

Consistent with our prediction, we find that German entrants' leverage ratios become more sensitive to incumbents’ median leverage ratios in concentrated markets after the disclosure shock relative to control firms. This change in sensitivity (i.e., the increase in mimicking attributable to mandated disclosure) is roughly equal to the magnitude of the unconditional sensitivity of entrant to incumbent leverage in the full sample, suggesting a key role for public disclosure in explaining capital structure peer effects.3 The effect is incremental to those of macroeconomic factors, characteristics of the entrant and incumbent set, and differences across regions, industries, and years in average entrant-incumbent sensitivities.

An identifying assumption of the DD estimator is that entrant-incumbent sensitivities would have similar trends between the treatment and control firms in the absence of the disclosure shock (the parallel trends assumption). While we cannot observe counterfactual trends, graphical evidence shows little difference in pre-treatment trends and a large and persistent jump in entrant-incumbent sensitivities once disclosures become widely available. We also find that the disclosure shock's effect on mimicking is exclusive to markets in which no incumbents voluntarily disclose financial information before the shock as well as those in which incumbents disclose relatively complete and audited information after. These results help address the possibility that a concurrent event, rather than the disclosure change, explains our findings.

Several mechanisms within the disclosure shock could underlie the change in mimicking. Ascribing the results to any single mechanism is challenging in this context, given the scope of the disclosure change was so expansive and its effects so varied. Concurrent papers that examine changes in disclosure mandates for private firms acknowledge the same (e.g., Breuer et al., 2019b; Breuer, 2018). However, features of the setting and additional analyses can speak to which mechanisms are more plausible.

One possibility is the change in mimicking is a result of changes in product market competition, banking competition, or other shared latent factors induced by the shock. Breuer (2018) finds that disclosure mandates increase the dynamism of product markets, leading to greater product market entry and lower concentration. Similarly, Breuer et al. (2018) find that private firm disclosure creates more transactional banking relationships, as the availability of financial information allows firms to contract with a greater number of and more distant banks. We use the Oster (2019) test to address the possibility that the change in mimicking is simply due to these changes in competition or other shared factors. The results provide no evidence to suggest these channels account for the change in mimicking.

A more direct (albeit unobservable) channel that could explain the change is interfirm learning (e.g., Devenow and Welch, 1996). Entrants have less information than incumbents about inputs to financing policies, such as the variability of market demand and operating costs or characteristics of growth opportunities (e.g., Caves, 1998). These information asymmetries make it difficult for entrants to weigh competitive and contracting costs of more debt financing against expected tax savings and other benefits, including reductions in owners' equity risk (e.g., Robb and Robinson, 2014). As a result, incumbents’ financial information is likely useful to entrants to minimize the risk of entering excessively over- or under-levered.4 If so, the change in mimicking due to the availability of public disclosures operates, at least in part, via learning.

We conduct several tests along these lines. First, we illustrate related effects consistent with learning from incumbent disclosures. Given that private firm managers are often uncertain about disclosure requirements (e.g., Gassen and Muhn, 2018; Collis, 2008), we provide evidence that German entrants mimic incumbents' disclosure choices more after the shock relative to entrants in the control countries. These results support the idea that German entrants' managers acquire and use incumbents' newly available filings.5 Analogously, we re-estimate our primary tests based on debt maturity instead of total leverage. If entrants use financial disclosures to mimic total leverage, then they plausibly use financial disclosures to learn about and mimic incumbents’ term structure of debt as well. Our results are consistent with this prediction.

Second, we examine three sources of cross-sectional heterogeneity. If the change in mimicking is because of learning, we expect the effect to be weaker when incumbents exhibit unusual levels of leverage, which can occur when inimitable idiosyncratic factors drive financing choices, and when entrants have high asset tangibility, as collateralization reduces the need to benchmark against peers. We also expect the effect to be weaker for smaller entrants, for which financial sophistication is lower. The results are consistent with these predictions and show effect sizes roughly 100%–300% greater in those portions of the cross-section where the ability or incentives to mimic are not muted, relative to the baseline DD specification. Overall, the findings illustrate the importance of incentives for interfirm learning.

Finally, we conduct a falsification test that examines mimicking of smaller incumbents. De novo entrants tend to start small (to limit sunk costs) and grow to efficient scale, suggesting that larger incumbents are the most relevant benchmarks for entrant capital structure decisions (Caves, 1998; Dunne et al., 1988). Consistent with this reasoning, Leary and Roberts (2014) find that public firms react to changes in the financial structure of larger but not smaller peers. A learning mechanism fits well with this evidence: as smaller incumbents tend to be less relevant for entrants, the availability of smaller incumbents' financials is unlikely to affect entrants' financing decisions. The results provide little evidence that the disclosure shock affects entrants’ mimicking of smaller incumbents. Evidence that firms mimic larger incumbents but not smaller incumbents again supports a learning mechanism and helps to differentiate our results from spillover effects of the peer information environment on lending (e.g., Shroff et al., 2017).6

The results contribute to our understanding of the effects of public financial statement disclosure on entry behavior and private firms' financing decisions. In this way, our study sits at the intersection of several streams of literature. A dominant theme in the capital structure literature is that economic factors such as agency incentives or product market considerations could account for why firms mimic peers' financing decisions, but there is little evidence on how mimicking arises (i.e., what conditions facilitate this mimicking).7 Similarly, empirical evidence suggests a variety of consequences of public disclosure mandates, but few of these studies have direct implications for firm-pair financing sensitivities. For example, Vanhaverbeke et al. (2019) show that corporate credit ratings become more conservative after private firm disclosure increases, presumably due to rating analysts' reputation concerns. However, their results are driven by changes in discretionary assessments of analysts, not by changes in fundamentals or the business environment. Evidence of the effects of financial disclosures on product market entrants is even more limited. While several models consider the interplay of information and entry decisions, these models often take information asymmetries with incumbents as given (e.g., Milgrom and Roberts, 1982) or model an incumbent's discretionary disclosure choice (e.g., Darrough and Stoughton, 1990). Empirical work is largely similar. For instance, Tomy (2019) studies incumbents' use of reporting discretion to discourage entry in a US banking setting but does not examine effects of incumbent disclosure on entrants' post-entry policies.

Our findings complement this work. Our results suggest public disclosure of financial information facilitates capital structure mimicking among private firm entrant-incumbent pairs. In addition, although we cannot pinpoint any single mechanism underlying this effect, the cross-sectional and additional analyses appear to be more consistent with an interfirm learning mechanism than other potential channels documented in the literature. Regardless, evidence that greater incumbent disclosure results in more homogenous financing decisions by entrants should help managers and regulators better understand the implications of mandatory disclosure requirements. Further, in highlighting the role of public information in facilitating financing choices, the evidence contributes to the body of work that examines product market-capital structure interactions (e.g., Chevalier, 1995) and to the literature on determinants of corporate policies of private firms (e.g., Gao et al., 2013; Minnis, 2011; Brav, 2009).

Section snippets

Institutional setting

European countries have long required private limited liability firms to publicly disclose certain financial statement information on an annual basis. These requirements, outlined in EU Directives, are implemented and enforced by each national government. Most countries have maintained high levels of compliance with these requirements. However, the German government failed to effectively enforce public disclosure requirements for decades over concerns that public disclosure could reveal firms’

Data and sample selection

We construct our sample from the Bureau van Dijk database Amadeus, which covers publicly available financial and non-financial data for European private firms (e.g., Bernard, 2016; Burgstahler et al., 2006). The initial sample consists of all firms on Amadeus with fiscal years ending between 2005 and 2010 headquartered in one of the four largest European economies (by GDP): France, Germany, Italy, or the UK.9

Primary model estimation

Our main results are presented in Table 3. Column 1 provides a baseline model where entrant leverage is regressed on incumbent leverage and country-industry, region, and year fixed effects, and column 2 adds control variables. We find a positive and significant relation between entrant and incumbent leverage; in column 2, for instance, we find that a one standard deviation increase in incumbent leverage is associated with an increase in entrant leverage of roughly one percentage point (0.037

Evidence of a potential mechanism—interfirm learning

The remainder of the paper provides evidence on plausible mechanisms underlying the change in mimicking. Because the change in disclosure reshaped the information environment of most private firms in Germany, multiple mechanisms could contribute to (or moderate) the main results. Further, these mechanisms could be partially unobservable or operate via chains of causal links. These possibilities make it challenging to draw definitive conclusions about potential channels, so we caution against

Conclusion

We exploit a shock to the enforcement regime for German private firms' annual financial disclosures in the mid-2000s to examine the effects of product market incumbents' financial disclosures on entrants' initial financing decisions. In this setting, ex ante voluntary disclosure is uncommon, low-cost alternative information sources are scarce, and disclosures made ex post are both easily accessible and reliable (e.g., Hanlon et al., 2014; Burgstahler et al., 2006). These institutional

References (67)

  • A. Demirgüç-Kunt et al.

    Institutions, financial markets, and firm debt maturity

    J. Financ. Econ.

    (1999)
  • A. Devenow et al.

    Rational herding in financial economics

    Eur. Econ. Rev.

    (1996)
  • H. Gao et al.

    Determinants of corporate cash policy: insights from private firms

    J. Financ. Econ.

    (2013)
  • J.R. Graham et al.

    The theory and practice of corporate finance: evidence from the field

    J. Financ. Econ.

    (2001)
  • J. Hanssens et al.

    The evolution of debt policies: new evidence from business startups

    J. Bank. Finance

    (2016)
  • R. Korajczyk et al.

    Capital structure choice: macroeconomic conditions and financial constraints

    J. Financ. Econ.

    (2003)
  • C. Martinelli

    Small firms, borrowing constraints, and reputation

    J. Econ. Behav. Organ.

    (1997)
  • M. Massa et al.

    Mimicking repurchases

    J. Financ. Econ.

    (2007)
  • R. Morck et al.

    Banking system control, capital allocation, and economy performance

    J. Financ. Econ.

    (2011)
  • N. Shroff et al.

    When does the peer information environment matter?

    J. Account. Econ.

    (2017)
  • N. Shroff

    Discussion of Is the risk of product market predation a cost of disclosure?

    J. Account. Econ.

    (2016)
  • R.E. Tomy

    Threat of entry and the use of discretion in banks' financial reporting

    J. Account. Econ.

    (2019)
  • J. Asker et al.

    Competition and the structure of vertical relationships in capital markets

    J. Polit. Econ.

    (2010)
  • P. Bolton et al.

    A theory of predation based on agency problems in financial contracting

    Am. Econ. Rev.

    (1990)
  • J.A. Brander et al.

    Oligopoly and financial structure: the limited liability effect

    Am. Econ. Rev.

    (1986)
  • O. Brav

    Access to capital, capital structure, and the funding of the firm

    J. Finance

    (2009)
  • M. Breuer

    How Does Financial-Reporting Regulation Affect Market-wide Resource Allocation?

    (2018)
  • M. Breuer et al.

    When You Talk, I Remain Silent: Spillover Effects of Peers' Mandatory Disclosures on Firms' Voluntary Disclosures

    (2019)
  • M. Breuer et al.

    Mandated Financial Reporting and Corporate Innovation

    (2019)
  • M. Breuer et al.

    How does financial reporting regulation affect firms' banking?

    Rev. Financ. Stud.

    (2018)
  • D. Burgstahler et al.

    The importance of reporting incentives: earnings management in European private and public firms

    Account. Rev.

    (2006)
  • M.C. Bustamante et al.

    Does Firm Investment Respond to Peers' Investment

    (2018)
  • J. Cai et al.

    Interfirm relationships and business performance

    Q. J. Econ.

    (2018)
  • Cited by (20)

    • The real effects of financial reporting: Evidence and suggestions for future research

      2024, Journal of International Accounting, Auditing and Taxation
    • Mandatory financial information disclosure and credit ratings

      2024, Journal of Accounting and Economics
    • Peer performance and the asymmetric timeliness of earnings recognition

      2023, International Review of Financial Analysis
      Citation Excerpt :

      In line with this argument, previous studies document that managers are more likely to make decisions based on the decisions of their peers. For example, managers take peer financing decisions as an essential reference for their own financing decisions (Bernard, Kaya, & Wertz, 2021; Leary & Roberts, 2014; MacKay & Phillips, 2005). Firms have an incentive to imitate peer innovation (Bessen & Maskin, 2009; Im & Shon, 2019; Zeng, 2001).

    View all citing articles on Scopus

    We thank Bob Holthausen (editor) and Hans Christensen (reviewer), participants at the EAA Conference 2018, workshop participants at Free University Berlin, Karlsruhe Institute of Technology, Ruhr University Bochum, Technical University Munich, University of Passau, University of Pittsburgh, and University of Texas at Austin, as well as Karthik Balakrishnan, Dave Burgstahler, Nicole Cade, Peter Demerjian, Diane Denis, Petroula Glachtsiou, Daniel Hoang, Chris Higson, Phil Quinn, James Ryans (discussant), Andrew Sutherland, Sara Toynbee, Florin Vasvari, Roberto Vincenzi, and Toni Whited for helpful comments on earlier versions of the paper.

    View full text