Corporate Financing and Stock Price Crash Risk

Corporate Financing and Stock Price Crash Risk
A thesis submitted to The University of Manchester for the degree of
Doctor of Philosophy
in the Faculty of Humanities
2018
Yangke Liu
Alliance Manchester Business School

List of Content
ABSTRACT 4
DECLARATION 5
COPYRIGHT STATEMENT 5
ACKNOWLEDGEMENTS 6
INTRODUCTION 8
CORPORATE DEBT MATURITY AND STOCK PRICE CRASH RISK 18
ABSTRACT 18
2.1. Introduction 19
2.2. Related literature and hypotheses 24
2.2.1. Short-maturity debt 24
2.2.2. Stock price crash risk 26
2.2.3. Hypotheses development 29
2.3. Research design 32
2.3.1. Data and sample 32
2.3.2. Measuring short-maturity debt 32
2.3.3. Measuring stock price crash risk 33
2.3.4. Control variables 34
2.4. Empirical results 35
2.4.1. Descriptive statistics 35
2.4.2. Baseline regression results 36
2.4.3. Identification strategies and robustness checks 38
2.4.4. Corporate governance mechanisms and short-maturity debt 46
2.4.5. Information asymmetry and short-maturity Debt 48
2.4.6. Risk-taking and short-maturity debt 50
2.5. Conclusion 51
SUPPLIERS FINANCING AND STOCK PRICE CRASH RISK: MONITORING
VERSUS CONCESSION? 82
ABSTRACT 82
1
3.2. Related literature and hypothesis development 90
3.2.1. Literature on trade credit 90
3.2.2. Literature review on stock price crash risk 91
3.2.3. Hypothesis development 92
3.3. Empirical methodology 94
3.3.1. The sample 94
3.3.2. Measuring trade credit 94
3.4. Empirical results 96
3.4.2. Multivariate regression analysis 97
3.4.3. Addressing endogeneity 99
3.4.4. Robustness checks 101
3.5. Cross-sectional variations and underlying channels 102
3.5.1. Governance mechanisms 102
3.5.2. Bank monitoring mechanisms 104
3.5.3. Market power mechanisms 106
3.5.4. Financial distress mechanisms 108
3.6. Conclusion 109
DOER BANK DEREGULATION AFFECT STOCK PRICE CRASH RISK? 135
ABSTRACT 135
4.2. Literature review and hypotheses development 143
4.2.1. Bank branching deregulation and relevant literature 143
4.2.2. Literature on stock price crash risk 145
4.2.3. Hypothesis 147
4.3. Data and methodology 148
4.3.1. Sample selection 148
4.3.2. Measuring branch deregulation 149
4.4. Results 152
2
4.4.2. Baseline specification and results 152
4.4.3. Endogeneity tests 154
4.4.4. Robustness tests 156
4.4.5. Robustness tests: Alternative measures 157
4.5. Mechanisms 159
4.5.1. External financial dependence 159
4.5.2. Firm risk 161
4.6. Conclusion 162
CONCLUSION AND SUGGESTIONS FOR FUTURE RESEARCH 191
This thesis contains 52,378 words including title page, tables, and footnotes.
3
Abstract
The University of Manchester

Yangke Liu
Doctor of Philosophy (PhD)
Corporate Financing and Stock Price Crash Risk
10 April 2018
In this thesis, I examine the impact of three corporate financing issues on the probability of
firm equity value collapse, namely stock price crash risk. The thesis consists of three essays.
In the first essay, I find that firms with a larger proportion of short-term debt have lower
future stock price crash risk, consistent with short-term debt lenders playing an effective
monitoring role in constraining managers’ bad-news-hoarding behavior. The inverse relation
between short-maturity debt and future crash risk is more pronounced for firms that are
harder to monitor due to weaker corporate governance, higher information asymmetry, and
greater risk-taking. These findings suggest that short-term debt substitutes for other
monitoring mechanisms in curbing managerial opportunism and reducing future crash risk.
Our study implies that short-maturity debt not only preserves creditors’ interests, but also
protects shareholders’ wealth.
The second essay empirically tests two opposing views on the relation between supplier
financing and future stock price crash risk: monitoring versus concession. I present robust
evidence that trade credit is negatively associated with firm-specific stock price crash risk,
consistent with the view that trade credit, as an important source of short-term financing,
effectively monitors buying firms and therefore constrains their bad-news-hoarding behavior.
Further analyses reveal that the role of trade credit in mitigating stock price crash risk is
more pronounced among firms that demand a higher level of monitoring such as those with
weaker governance, less bank monitoring, lower market power, and higher distress risk.
Overall, our results shed light on how trade credit shapes managerial disclosure incentives.
In the third essay, I examine the effect of the deregulation of bank branch restrictions on
nonfinancial firms’ stock price crash risk. Most U.S. states lifted restrictions on intrastate
branching from the 1970s to the 1990s, which improved bank monitoring and lending quality.
I find robust evidence that the bank deregulation leads to lower levels of firms’ stock price
crash risk, consistent with branch reform improving bank monitoring efficiency and enabling
banks to better constrain borrowers’ bad-news-hoarding behavior. This mitigating effect is
more pronounced for firms that are riskier and more reliant on external finance. Our study
suggests that bank branch deregulation is beneficial for protecting nonfinancial firms’
market value.
4
Declaration
No portion of the work referred to in the thesis has been submitted in support of an
application for another degree or qualification of this or any other university or other institute
of learning.
Copyright statement
The author of this thesis (including any appendices and/or schedules to this thesis) owns any
copyright in it (the “Copyright”) and s/he has given The University of Manchester the right
to use such Copyright for any administrative, promotional, educational and /or teaching
purposes.
Copies of this thesis, either in full or in extracts, may be made only in accordance with the
regulations of the John Rylands University Library of Manchester. Details of these
regulations may be obtained from the Librarian. This page must form part of any such copies
made.
The ownership of any patents, designs, trademarks and any and all other intellectual property
rights except for the Copyright (the “Intellectual Property Rights”) and any reproductions of
copyright works, for example graphs and tables (“Reproductions”), which may be described
in this thesis, may not be owned by the author and may be owned by third parties. Such
Intellectual Property Rights and Reproductions cannot and must not be made available for
use without prior written permission of the owner(s) of the relevant Intellectual Property
Rights and/or Reproductions.
Further information on the conditions under which disclosure, publication and exploitation
of this thesis, the Copyright and any Intellectual Property Rights and/or Reproductions
described in it may take place is available from the Head of School of Alliance Manchester
Business School (or the Vice-President) and the Dean of the Faculty of Life Sciences, for
Faculty of Life Sciences’ candidates.
5
Acknowledgements
First and foremost, I would like to give special thanks to my PhD supervisors, Dr.
Viet Dang, Professor Edward Lee, and Dr. Cheng (Colin) Zeng. Very few students could be
so blessed like me to be supervised by three excellent supervisors. In 2014, Viet and Edward
chose to supervise me after we had a long chat, which made my academic journey possible.
I owe a debt of gratitude to Viet for his time, consideration, and careful attention to detail.
He is always so meticulous about my thesis from the macro-level research design to micro-
level writing format. Every time I open the revised paper from Viet, I would be astonished
at the numerous changes and comments. I am very thankful for the excellent example he has
provided as a committed finance researcher.
To Edward, I thank him for his untiring support and guidance throughout my journey.
When I was undertaking my first year study, Edward guided me to be positive, confident,
and motivated. Thinking out a promising research topic is not easy because it requires the
researchers to have solid academic knowledge and, the most important, perseverance. In
order to help me pass the first year annual review, Edward suggested various interesting
topics and eventually established the pivotal theme of my PhD research. I really appreciate
all his contributions of ideas and constant encouragement to make me survive to the end. It
has been my great honor to be Colin’s first PhD student. He joined the supervisory team in
my second year and has taught me, both consciously and un-consciously, how good doctoral
research is done. He is willing to help me with detailed questions, such as data processing,
software programming, paper writing, and presentation. For example, after my first essay
was accepted by European Financial Management Association in 2016, Colin decided to
come to Switzerland together with me to ensure that I can do well in my presentation.
6
For this thesis I would like to thank Professor Marie Dutordoir and Dr. Fangming Xu
for accepting to be the examiners of my PhD viva. Great thanks go to Marie, the PhD
Director of Accounting and Finance Division, for her constructive comments and
suggestions in my PhD reviews as well as inspirational encouragement in my difficult period.
In addition, I must thank many researchers in Alliance Manchester Business School who
taught me or gave me valuable comments on my thesis: Dr. Amedeo De Cesari, Professor
Konstantinos Stathopoulos, Dr. Maria Marchica, Dr. Ning Gao, Professor Norman Strong,
Dr. Roberto Mura, and Dr. Susanne Espenlaub. Also, I am grateful for the helpful comments
from Professor Michael Brennan at UCLA, Professor Karin Thorburn at Norwegian School
of Economics (NHH), and Professor Murillo Campello at Cornell University. Special thanks
go to Dr. John Doukas, the editor of European Financial Management, and two anonymous
reviewers for their efforts in the publication of my essay.
Furthermore, I will forever be thankful to my former postgraduate supervisor,
Professor Hong Hao, at China University of Petroleum, Beijing. She granted me the rich
knowledge of corporate finance and cultivate the interest in academic research. She is one
of the most important reasons why I pursue my PhD study in the UK.
Above all I would like to acknowledge the tremendous sacrifices that my parents
made to ensure that I had an excellent education. They encourage me in all my pursuits and
inspire me to follow my dreams. My PhD journey would not have been possible without
their emotional and financial support. For this and much more, I am forever in their debt. It
is to them that I dedicate this thesis.
7
Chapter 1
Introduction
1.1. Background and motivation
How corporate financing activities affect firm value has been one of the primary
questions in corporate finance research. It has long been established that the reliance on
various financing sources exerts an important influence on firms’ investment decisions and
market value (e.g., Myers and Majluf, 1984; Fama and French, 1998; Myers, 1984; Titman
and Wessels, 1988). Traditionally, the accounting and finance literature attempts to examine
how financial policies impact the change in equity value. However, a burgeoning strand of
research has started to focus on the extreme downside risk of firm market value—stock price
crash risk.
A stock price crash refers to an extreme collapse in equity value that causes a severe
loss of equity investors’ wealth. Different from passive investors, active investors aim to
maximize returns over the short run. An active investor with a large position in a stock that
encounters a crash can suffer a vast loss of wealth. Anecdotal evidence is not rare in history,
such as the cases of Enron, Tyco, and WorldCom.1 However, there is often little guidance
for investors to avoid such stock price crashes. The academic research on firm-specific stock
price crash risk has only emerged since the beginning of this century. In two seminal studies,
Chen, Hong, and Stein (2001) and Jin and Myers (2006) use the conditional skewness of
firm-specific stock return distribution as the measure of “crash prone” and contend that stock
price crashes are typically driven by the release of accumulated bad news. The theoretical
framework of Jin and Myers (2006) argues that the existence of information asymmetries
1
Information sources for those cases include Patsuris (2002), US. Congress (2003), and Desai (2005).
8
allows managers to hoard bad news for an extended period due to career and compensation
concerns. When the accumulated negative information suddenly comes out in the market,
stock prices collapse accordingly, leading to crashes. Following this framework, stock price
crash risk has become a topical issue in accounting and finance research. Recent empirical
studies suggest various determinants of crash risk, such as financial statements opacity (e.g.,
Hutton, Marcus, and Tehranian, 2009), tax avoidance (e.g., Kim, Li, and Zhang, 2011b),
accounting conservatism (e.g., Kim and Zhang, 2016), corporate social responsibility
performance (e.g., Kim, Li, and Li, 2014; Zhang, Xie, and Xu, 2016), employee welfare(e.g.,
Ben-Nasr and Ghouma, 2018), political officials’ incentives (e.g., Piotroski, Wong, and
Zhang, 2015; Lee and Wang, 2017), regulations and laws (e.g., DeFond, Hung, Li, and Li,
2014; Ali, Li, and Zhang, 2015).
Despite the proliferation of stock price crash risk research, studies investigating from
the perspective of a firm’s financing policy remain sparse. This thesis aims to contribute to
the literature by thoroughly exploring whether and how stock price crash risk is affected by
a set of corporate financing issues, including short-term debt financing, supplier financing,
and bank deregulation.
My thesis is structured around three topics that address the influence of all the three
corporate financing issues on stock price crash risk. Specifically, the first and second essays
provide robust evidence that short-term debt financing and trade credit, which is an
alternative external financing resource, have statistically and economically significant
impacts on firm-specific future stock price crash risk. The third essay examines the impact
of deregulation laws in the U.S. banking industry on nonfinancial firms’ crash risk. This
essay is also linked with corporate financing as the deregulation of bank branching
restrictions has greatly affected borrowers’ financing activities and their relationship with
9
banks (e.g., Jayaratne and Strahan, 1996; Black and Strahan, 2002). I will briefly introduce
each of the three essays in the following section.
1.2. Thesis overview and contributions
In my first essay, I find that firms with a larger proportion of short-term debt have
lower future stock price crash risk. This finding is consistent with the monitoring role of
short-term debt serving as an effective tool to curb managerial bad-news-hoarding behavior,
which in turn reduces stock price crash risk. Further analyses indicate that the negative
relationship between short-term debt and stock price crash risk is more pronounced for firms
with weak corporate governance, high information asymmetry, and high firm riskiness,
suggesting that short-term debt may act as a substitute for corporate governance mechanisms
in reducing managerial bad news hoarding.
This essay contributes to the literature on both debt maturity and stock price crash
risk. First, this is the first study to investigate the equity market consequences of corporate
debt maturity. Before this study, there is little research examining the impact of short-term
debt on corporate disclosure behavior and, ultimately, shareholder wealth. Second, this essay
adds to a growing stream of literature on stock price crash risk by showing that creditors
from debt capital market can exert a significant impact on stock price crash risk. Overall,
this study extends prior research by showing that shareholders can benefit from the
monitoring function of external creditors and in particular short-term debt lenders..
In my second essay, I attempt to examine two opposing views on the relation between
trade credit and future stock price crash risk: monitoring versus concession. On one hand,
trade creditors have an incentive to monitor buyers’ financial status and credit-worthiness
for the purpose of minimizing the default risk (Mian and Smith, 1992). Hence, trade credit
financing should curb managerial bad-news-hoarding behavior, leading to lower stock price
10
crash risk. On the other hand, suppliers may grant concessions to customers in order to
maintain a long-term supplier-buyer relationship, which might create opportunities for
managers to hide bad news and thereby increase crash risk. My finding shows that trade
credit is negatively associated with firm-specific stock price crash risk, consistent with the
monitoring view that suppliers may use trade credit to monitor buyers and constrain their
bad-news-hoarding behavior. Further analyses suggest that the role of trade credit in
mitigating crash risk is more pronounced among firms with weaker governance, less bank
monitoring, lower market power, and higher distress risk than firms with the opposite
characteristics.
This essay contributes to the literature in several ways. First, it adds to recent research
on the equity market consequences of trade credit. Different from prior research on the
relationship between trade credit and firm value (e.g., Martínez-Sola, García-Teruel, and
Martínez-Solano, 2013; Aktas, Croci, and Petmezas, 2015), this study is the first to examine
the impact of supplier financing on extreme negative stock returns. Second, this study joins
the debate concerning the monitoring versus concession effects of suppliers in supplier-
buyer relationships. My findings show that suppliers can effectively monitor customers and
restrict their bad-news-hoarding activities, which is consistent with the notion that trade
creditors have information advantage and enhance buyers’ information transparency (e.g.,
Petersen and Rajan, 1997; Cuñat, 2007). Finally, this study adds to the literature on stock
price crash risk by proposing a new determinant of stock price crash risk, namely trade credit
financing.
In my third essay, I focus on an exogenous shock in the banking industry—bank
deregulation and find that intrastate branching deregulation causes a significant reduction in
stock price crash risk for nonfinancial firms. This finding is consistent with the argument
that bank branch deregulation improves bank monitoring efficiency and enables banks to
11
better constrain borrowers’ bad-news-hoarding behavior. My additional findings show that
the mitigating effect of bank deregulation on crash risk is more pronounced for firms with a
high degree of external financial dependence and risk-taking behavior, suggesting that banks
tend to more effectively monitor firms that are more reliant on external finance and are
riskier. Overall, these results indicate that the dependence of firms on bank lending and firm
risk are the underlying mechanisms driving the relation between intrastate bank deregulation
and firms’ crash risk.
The contributions of this study are as follows. First, it adds to the literature examining
the effects of bank deregulation on corporate policies of non-financial firms (e.g., Black and
Strahan, 2002; Ceterolli and Strahan, 2006; Hombert and Matray, 2016). Second, this study
also adds to the rich literature on stock price crash risk by showing that intrastate branching
deregulation laws help mitigate firm-level stock price crash risk and protect shareholders’
wealth. Last, this study contributes to the literature linking bank monitoring with borrowers’
agency problem. The findings support the notion that improved bank monitoring can
alleviate borrowers’ opportunistic misbehavior (Rajan and Winton, 1995; Datta, Iskandar-
Datta and Raman, 2005; Roberts and Sufi, 2009).
The methodology used in the three essays is econometric analysis. The research
objects are a large sample of U.S. publicly listed nonfinancial firms. Consistent with prior
literature on crash risk (e.g., Chen, Hong, and Stein, 2001; Kim, Li, and Zhang, 2011a, b), I
measure crash risk as negative conditional skewness of firm-specific weekly returns
(NCSKEW) and down-to-up volatility of firm-specific weekly returns (DUVOL). Each essay
has shown the detailed calculation procedure for those stock price crash risk measures.
1.3. Thesis structure
12
The thesis structure follows the format accepted by the Manchester Accounting and
Finance Group, Alliance Manchester Business School. It allows chapters to be incorporated
into a format suitable for submission and publication in peer-reviewed academic journals.
Therefore, this thesis is structured around three essays containing original research in
Chapters 2, 3, and 4. The chapters are self-contained, i.e., each chapter has a separate
literature review, answers unique and original questions, and employs a distinct analysis with
different datasets. The equations, footnotes, and tables are independent and are numbered
from the beginning of each chapter. Page numbers, titles, and subtitles have a sequential
order throughout the thesis.
The thesis continues as follows. Chapter 2 examines whether corporate debt maturity
influences stock price crash risk. Chapter 3 investigates the relationship between firms’
usage of trade credit and stock price crash risk. Chapter 4 studies the impact of branch
deregulation laws in the banking sector on nonfinancial firms’ stock price crash risk. Chapter
5 concludes.2
2
In Chapters 2, 3, and 4, I use the third person (we, our) rather than the first person (I, my) as these chapters
have already been in the form of published or working papers co-authored with my supervisors.
13
References
Aktas, N., Bodt, E., Lobez, F., and Statnik, J. C., 2012. The information content of trade
credit. Journal of Banking & Finance, 36(5), 1402–1413.
Ali, A., Li, N., and Zhang, W., 2015. Managers’ career concerns and asymmetric disclosure
of bad versus good news. Working paper.
Ben-Nasr, H., and Ghouma, H., 2018. Employee welfare and stock price crash risk. Journal
of Corporate Finance, 48, 700–725.
Black, S.E., and Strahan, P.E., 2002. Entrepreneurship and bank credit availability. Journal
of Finance, 57(6), 2807–2833.
Cetorelli, N., and Strahan, P.E., 2006. Finance as a barrier to entry: Bank competition and
industry structure in local US markets. Journal of Finance, 61(1), 437–461.
Chen, J., Hong, H., and Stein, J., 2001. Forecasting crashes: trading volume, past returns,
and conditional skewness in stock prices. Journal of Financial Economics, 61(3), 345–
381.
Cuñat, V., 2007. Trade credit: suppliers as debt collectors and insurance providers, Review
of Financial Studies, 20(2), 491–527.
Datta, S., Iskandar-Datta, M., and Raman, K., 2005. Managerial stock ownership and the
maturity structure of corporate debt. Journal of Finance, 60(5), 2333–2350.
DeFond, M.L., Hung, M., Li, S., and Li, Y., 2014. Does mandatory IFRS adoption affect
crash risk?. The Accounting Review, 90(1), 265–299.
Desai, M.A., 2005. The degradation of reported corporate profits. Journal of Economic
Perspectives, 19(4), 171–192.
14
Fama, E.F., and French, K.R., 1998. Taxes, financing decisions, and firm value. Journal of
Finance, 53(3), 819–843.
Fazzari, S.M., Hubbard, R.G., Petersen, B.C., Blinder, A.S., and Poterba, J.M., 1988.
Financing constraints and corporate investment. Brookings papers on economic activity,
1988(1), 141–206.
Hombert, J., and Matray, A., 2016. The real effects of lending relationships on innovative
firms and inventor mobility. Review of Financial Studies, 30(7), 2413–2445.
Hutton, A.P., Marcus, A. J., and Tehranian, H., 2009. Opaque financial reports, R2, and crash
risk. Journal of Financial Economics, 94(1), 67–86.
Jayaratne, J., and Strahan, P.E., 1996. The finance-growth nexus: Evidence from bank
branch deregulation. The Quarterly Journal of Economics, 111(3), 639–670.
Kim, Y., Li, H., and Li, S., 2014. Corporate social responsibility and stock price crash risk.
Journal of Banking & Finance, 43, 1–13.
Kim, J. B., Li, Y., and Zhang, L., 2011a. CFO vs. CEO: equity incentives and crashes.
Journal of Financial Economics, 101(3), 713–730.
Kim, J. B., Li, Y., and Zhang, L., 2011b. Corporate tax avoidance and stock price crash risk:
Firm level analysis. Journal of Financial Economics, 100(3), 639–662.
Kim, J.B., and Zhang, L., 2016. Accounting conservatism and stock price crash risk: Firm-
level evidence. Contemporary Accounting Research, 33(1), 412–441.
Lee, W., and Wang, L., 2017. Do political connections affect stock price crash risk? Firm-
level evidence from China. Review of Quantitative Finance and Accounting, 48(3), 643–
676.
15
Martínez‐Sola, C., García‐Teruel, P. J., and Martínez‐Solano, P., 2013. Trade credit policy
and firm value. Accounting & Finance, 53(3), 791–808.
Mian, S., and Smith, C. W., 1992. Accounts receivable management policy: theory and
evidence. Journal of Finance, 47(1), 169–200.
Myers, S.C., 1984. The capital structure puzzle. Journal of Finance, 39(3), 574–592.
Myers, S.C., and Majluf, N.S., 1984. Corporate financing and investment decisions when
firms have information that investors do not have. Journal of Financial Economics, 13(2),
187–221.
Patsuris, P., 2002. The corporate scandal sheet. Forbes.com, 1–6.
Petersen, M., and Rajan, R., 1997. Trade credit: Theories and evidence. Review of Financial
Studies, 10(3), 661–691.
Piotroski, J.D., Wong, T.J., and Zhang, T., 2015. Political incentives to suppress negative
information: evidence from Chinese listed firms. Journal of Accounting Research, 53(2),
405–459.
Rajan, R., and Winton, A., 1995. Covenants and collateral as incentives to monitor. Journal
of Finance, 50(4), 1113–1146.
Roberts, M.R., and Sufi, A., 2009. Renegotiation of financial contracts: Evidence from
private credit agreements. Journal of Financial Economics, 93(2), 159–184.
Titman, S., and Wessels, R., 1988. The determinants of capital structure choice. Journal of
Finance, 43(1), 1–19.
16
U.S. Congress, 2003. Joint Committee on Taxation. Report of Investigation of Enron
Corporation and Related Entities Regarding Federal Tax and Compensation Issues, and
Policy Recommendations, Volumes I–III. JCS–3–03. Washington, DC.
Zhang, M., Xie, L., and Xu, H., 2016. Corporate philanthropy and stock price crash risk:
Evidence from China. Journal of Business Ethics, 139(3), 595–617.
17
Chapter 2
Corporate Debt Maturity and Stock Price Crash Risk
ABSTRACT
We3 find that firms with a larger proportion of short-term debt have lower future stock price
crash risk, consistent with short-term debt lenders playing an effective monitoring role in
constraining managers’ bad-news-hoarding behavior. The inverse relation between short-
maturity debt and future crash risk is more pronounced for firms that are harder to monitor
due to weaker corporate governance, higher information asymmetry, and greater risk-taking.
These findings suggest that short-term debt substitutes for other monitoring mechanisms in
curbing managerial opportunism and reducing future crash risk. Our study implies that short-
maturity debt not only preserves creditors’ interests, but also protects shareholders’ wealth.
3
I use “we” hereafter because the three essays were co-authored with my supervisors Viet Dang, Edward Lee,
and Cheng Zeng,
18
2.1. Introduction
Debt is one of the primary means of capital acquisition for firms in the US and around the
world (e.g., Graham et al., 2015; Öztekin, 2015). In the context of debt contracting, the
structure of debt maturity significantly influences the decision making of both firms and
investors. The existing academic literature on debt maturity comprises two pathways. One
stream of literature has extensively documented the determinants of firms’ debt maturity
choices (e.g., Barclay and Smith, 1995; Guedes and Opler, 1996; Stohs and Mauer, 1996;
Ozkan, 2000; Datta et al., 2005; Antoniou et al., 2006; Brockman et al., 2010; Custódio et
al., 2013). The other strand of literature investigates the interaction between debt maturity
and other corporate policies, including financial leverage (Barclay et al., 2003; Johnson,
2003), debt covenants (Billett et al., 2007), cash holdings (Harford et al., 2014), and real
investment (Aivazian et al., 2005; Duchin et al., 2010; Almeida et al., 2011). Despite the
growing awareness of the role of debt maturity in shaping corporate finance and investment
policies, relatively limited research is available on whether and how the monitoring of short-
term debt lenders affects shareholder wealth through its impact on stock prices. Our study
fills this gap in the literature by examining the effect of short-term debt on future stock price
crash risk.
Stock price crash refers to an extreme collapse in equity value that causes a severe
decline in shareholders’ wealth. This downside risk is of serious concern to investors and
firms alike because it affects their risk management and investment decision making. Prior
literature suggests that the primary cause of stock price crash is managers’ tendency to hoard
and withhold unfavorable information from outsiders in the presence of potential agency
problems (e.g., Jin and Myers, 2006; Kothari et al., 2009; Hutton et al., 2009; Callen and
Fang, 2015a). Incentivized by empire building, as well as career and compensation concerns,
managers may attempt to conceal bad news over an extended time, and upon subsequent
19
revelation of such accumulated information the market value of their firms corrects sharply
downward, leading to stock price crashes.
We hypothesize that short-term debt can reduce a firm’s stock price crash risk for the
following reasons. Since the repayment of debt financing is fixed, lenders face an
asymmetric payoff, that is, they are exposed to downside credit risk with a capped upside
payoff. Under such circumstances, the timely disclosure of bad news is of particular
importance to debtholders. Compared to long-term debt, debt with short maturities involves
more frequent renewal or refinancing (Myers, 1977; Diamond, 1991a), thus serving as an
effective tool for lenders to monitor managerial behavior and enhance information
transparency (Ranjan and Winton, 1995; Stulz, 2001; Datta et al., 2005; Graham et al., 2008).
This is because incomplete debt contracts only allocate lenders’ control rights ex ante, hence
giving lenders strong incentives to use the credible threat of not renewing debt contracts to
deter managers’ opportunistic behavior ex post (Giannetti, 2003). Lenders of short-term debt,
in particular, can protect their rights by requiring managers to provide timely and reliable
information about firms’ financial condition and future investments when negotiating the
renewal of debt contracts. This distinct feature of short-term debt enhances managerial
information revelation, curbs the likelihood of bad news hoarding, and hence reduces future
stock price crash risk. While it is possible that long-term debt holders can also play a
monitoring role, especially through the use of debt covenants, the monitoring function of
long-term debt tends to be less effective than that of short-term debt, because long-term debt
holders can act only when a covenant violation occurs (Rajan and Winton, 1995). As a result
of this limitation, the ability of long-term debtholders to curb managerial hoarding of adverse
information may be relatively weaker than that of short-term debtholders. Overall, our
arguments predict that short-maturity debt is negatively related to future stock price crash
risk.
20
To test this prediction, we regress future stock price crash risk on short-term debt,
while controlling for several important firm-specific determinants of crash risk. Consistent
with prior studies (e.g., Chen et al., 2001; Hutton et al., 2009; Kim et al., 2011a, 2011b; Kim
and Zhang, 2016), we use two main measures of stock price crash risk, namely (i) the
negative conditional skewness of firm-specific weekly returns and (ii) the “down-to-up
volatility” of firm-specific weekly returns. Following the debt maturity literature, we
measure short-maturity debt as the fraction of debt due within three years, which is a well-
established cutoff point for computing the short-term debt ratio (e.g., Barclay and Smith,
1995; Johnson, 2003; Brockman et al., 2010; Harford et al., 2014).
We document empirical evidence in support of our main hypothesis. Using a sample
of 7,712 unique firms and 53,052 firm-year observations from 1989 through 2014, we find
that firms using more short-term debt exhibit lower future stock price crash risk. This finding
is in line with managers being less likely to conceal and hoard bad news in the presence of
external monitoring by short-term debt lenders. Our results are robust to a battery of tests
addressing endogeneity concerns and those using alternative measures of key variables,
including crash risk and short-maturity debt. Importantly, using a sample of new debt issues,
we find that the maturities of those debt issues are positively related to future stock price
crash risk, which further strengthens our main inference of a causal relationship between
short-term debt and future crash risk.
We next investigate whether short-term debt effectively substitutes for other
monitoring mechanisms in curbing managerial bad-news-hoarding behavior. These
additional empirical analyses are motivated by the extant studies on the agency perspective
of debt maturity (e.g., Rajan and Winton, 1995; Datta et al., 2005). If short-term debt indeed
reduces stock price crash risk due to creditors’ monitoring, then we would expect such an
effect to make a bigger difference among firms that are more susceptible to agency problems
21
and information asymmetry. Consistent with this conjecture, we show that the mitigating
effect of short-term debt on crash risk is more pronounced when firms have weaker
governance, such as less (long-term) institutional ownership and lower shareholder rights.
Meanwhile, we find that the negative relation between short-term debt and future crash risk
is stronger among firms with a higher degree of information asymmetry, measured by analyst
forecast errors, a dispersion in analyst earnings forecasts, and research and development
(R&D) intensity. Finally, we show that the inverse relation between short-term debt and
future crash risk is more salient for firms that engage in greater risk-taking, including those
with higher leverage or without bond rating. Taken together, these findings shed light on
how short-term debt lenders can substitute other corporate monitoring mechanisms in
mitigating managerial discretion and future stock price crash risk.
Our paper contributes to at least two strands of literature. First, to the best of our
knowledge, this is the first study to investigate the equity market consequences of corporate
debt maturity, with a focus on the impact of short-term debt on high moments of stock return
distribution (i.e., extreme negative returns). Prior research suggests that short-maturity debt
plays a significant role in reducing agency costs (Myers, 1977; Childs et al., 2005; Datta et
al., 2005), risk-taking incentives (Barnea et al., 1980; Leland and Loft, 1996; Brockman et
al., 2010), and audit risk (Gul and Goodwin, 2010) through the frequent and stringent
monitoring of external creditors (Rajan and Winton, 1995; Stulz, 2001). However, there has
been little, if any, research testing the impact of short-term debt on corporate disclosure
behavior and, ultimately, shareholder wealth. Our empirical evidence therefore extends this
literature by showing that short-term debt can reduce stock price crash risk through curbing
managers’ bad-news-hoarding activities.
Second, our study enriches a growing stream of research on stock price crash risk.
As a special feature of stock return distribution, the issue of stock price crash risk is attracting
22
increased attention among academics and practitioners. Recent studies show that various
internal and external factors influence firms’ stock price crash risk, consistent with the bad-
news-hoarding argument (see Habib et al., 2016 for a literature review). Among the internal
mechanisms affecting managerial incentives to withhold adverse information are executive
compensation (Kim et al., 2011a), tax avoidance techniques (Kim et al., 2011b), accounting
conservatism (Kim and Zhang, 2016), and chief executive officer (CEO) overconfidence
(Kim et al., 2016). Examples of external monitoring mechanisms include institutional
ownership (An and Zhang, 2013; Callen and Fang, 2013), accounting standards (DeFond et
al., 2014), short-selling (Callen and Fang, 2015b), religion (Callen and Fang, 2015a),
auditing service (Habib and Hasan, 2015; Callen and Fang, 2016), and stock liquidity (Chang
et al., 2016). Our study adds to this literature by providing novel evidence of how (i) a
corporate financial policy such as debt maturity structure and (ii) creditors from debt capital
market can exert a significant impact on stock price crash risk, above and beyond the effects
of several factors affecting crash risk reviewed above.
Our paper is closely related to Boubaker et al. (2014) and Andreou et al. (2016), who
document evidence of the influences of corporate governance attributes on stock price crash
risk. However, our study differs from them in an important way. While we examine the
monitoring role of short-term debt, an external corporate governance mechanism, in
mitigating crash risk, Boubaker et al. (2014) and Andreou et al. (2016) mainly focus on the
role of internal governance mechanisms and particularly those relating to shareholders and
managers (e.g., controlling shareholders, ownership structure, board structure and processes,
and managerial incentives). Our study is the first to provide systematic evidence that short-
maturity debt plays an effective monitoring role in constraining managers’ bad news
hoarding and thus reducing stock price crash risk. Hence, we extend prior studies by showing
that shareholders can benefit from the monitoring function of external creditors and in
23
particular short-term debt lenders. In addition, our study is of practical relevance as it shows
how debt maturity structure may have important implications for stock selection by equity
investors.
The rest of the paper is organized as follows. Section 2.2 reviews prior research on
debt maturity and stock price crash risk and develops our hypotheses. Section 2.3 describes
the sample and research design. Section 2.4 presents the empirical results. Section 2.5
concludes.
2.2. Related literature and hypotheses
2.2.1. Short-maturity debt
The finance literature has identified several benefits of short-term debt. From lenders’
perspective, a distinct advantage of short-maturity debt is that it gives them control rights ex
post, with which they can effectively monitor borrowers. Due to incomplete debt contracting,
lenders generally do not have control rights over every future contingency in the initial
contract terms. Debt with short maturities, however, provides them with better protection
and greater bargaining power because they can threaten borrowers with rejection of
refinancing when the short-term debt comes up for renewal (Giannetti, 2003). Put differently,
the frequent renegotiations and renewals of short-maturity debt help fill the void of
contractual incompleteness by allocating lenders’ control rights ex post (Roberts and Sufi,
2009; Roberts, 2015). This advantage of short-term debt prompted Myers (1977, p. 159) to
suggest that “permanent debt capital is best obtained by a policy of rolling over short
maturity debt claims.”4
4
From a borrowing firm’s perspective, although short-term debt exposes the firm to refinancing risk (Diamond,
1991a), it can help alleviate incentive problems arising from the conflicts of interest between shareholders and
creditors. For instance, existing studies suggest that short-term debt helps reduce underinvestment (Myers,
24
A key benefit of short-term debt is that it exercises a monitoring function over
borrowers, thus reducing information problems and increasing corporate transparency. Prior
research shows that short-maturity debt subjects managers to more frequent and stringent
creditor monitoring (Stulz, 2001; Datta et al., 2005), thereby forcing more timely
information disclosure (Rajan and Winton, 1995). The reason is that short-term debt lenders
must periodically evaluate the borrowing firm’s creditworthiness, especially during the debt
renegotiation and renewal processes. This feature of short-term debt gives its lenders an
important advantage over long-term debt lenders, who may also play a monitoring role but
mainly do so by relying on ex ante covenant terms to gather limited and verifiable
information. Rajan and Winton (1995) argue that short-maturity debt provides lenders with
the flexibility and unlimited ability to act, even when the debt covenants have not been
violated. Specifically, the frequent scrutiny of short-term debt lenders leads to greater
disclosure of company information, including relevant information that may be unverifiable
and imperfectly correlated with covenant terms.5
Empirical evidence provides support for the notion that short-term debt plays an
important role in scrutinizing firms and alleviating information asymmetry. Graham et al.
(2008) show that banks shorten the maturities of loans provided to firms that have previously
engaged in financial misreporting, consistent with lenders using short-term debt to enhance
1977; Childs et al., 2005), asset substitution (Leland and Toft, 1996), and excessive risk-taking (Barnea et al.,
1980; Childs et al., 2005; and Brockman et al., 2010).
5
However, some research suggests that under certain circumstances long-term debt is able to discipline
management effectively. For instance, Rajan and Winton (1995) argue that long-term debt may be preferable
to short-term debt if the former’s covenants depend on publicly available or less costly information. On the
other hand, some studies contend that the liquidity and refinancing risk of short-term debt may incentivize
managers to conceal negative information (Roberts and Sufi, 2009; Fields et al., 2016). Since long-term debt
may not impose such a refinancing pressure, it may be more effective in curbing firms’ misreporting behavior.
Nevertheless, these arguments are likely to bias against us finding evidence of an inverse relationship between
short-maturity debt and stock price crash risk.
25
managerial scrutiny and information gathering in an environment of increased risk and
information asymmetry. In a similar vein, Gul and Goodwin (2010) demonstrate the
importance of short-maturity debt in improving corporate transparency, especially among
risky firms. They find that short-term debt is negatively related to audit risk and that this
relation is more pronounced for firms with low quality ratings. The latter finding suggests
that the monitoring of short-maturity debt has a greater impact on corporate transparency for
firms that are considered more uncertain and risker by credit rating agencies. While these
studies show how creditors’ monitoring reduces firms’ information asymmetry, they have
not investigated whether the use of short-term debt influences stock price crash risk and
shareholder wealth, as we do in this paper.
2.2.2. Stock price crash risk
A large and growing body of literature has examined what determines stock price
crash risk, reflecting the increasing importance of this issue to academics and practitioners
(e.g., Habib et al., 2016). Chen et al. (2001) find that the trading volumes and returns over
the past several months can forecast future crashes. Importantly, Jin and Myers (2006)
theoretically show that inside managers who are in charge of revealing firm-specific
information have incentives to absorb certain downside risk by withholding bad news.
However, once the hoarded bad news reaches a critical threshold, managers may no longer
be able to conceal bad news and the revelation of such information to the public becomes
inevitable. This revelation in turn leads to extreme downward stock price corrections or
crashes that are manifested as a long left tail in the distribution of returns.
Existing studies identify various determinants of future stock price crash risk and
provide empirical evidence in support of the bad-news-hoarding argument. Jin and Myers
(2006), Hutton et al. (2009), and Callen and Fang (2015a) show that information opacity and
low transparency, particularly due to the use of discretionary accruals or accounting
26
irregularities, lead to higher future crash risk. In a similar vein, Kim and Zhang (2016) find
that a high degree of conditional conservatism neutralizes managers’ tendency to delay bad
news and accelerates good news recognition, thus lowering future crash risk. While those
studies provide direct evidence in favor of the bad-news-hoarding argument, other works
have documented additional evidence of several internal factors affecting crash risk via the
bad-news-hoarding channel. For instance, Kim et al. (2011a) argue that equity incentives
induce managers to purposely hide negative information and manipulate market expectations,
leading to an increase in future crash risk. Similarly, Xu et al. (2014) find that executives
who enjoy excess perks are more likely to conceal bad news, thus resulting in higher future
crash risk. Kim et al. (2011b) show that corporate tax avoidance increases crash risk because
the tax avoidance techniques used by managers reduce corporate transparency. On the other
hand, Kim et al. (2014) argue that managers committed to corporate social responsibility
tend to maintain greater transparency and have less incentive to withhold bad news, which
in turn leads to a lower probability of price crashes. However, Kim et al. (2016) show that
overconfident CEOs tend to overestimate and misperceive negative net present value (NPV)
projects, thus leading to an accumulation of bad performance and unfavorable information
and, subsequently, higher stock price crash risk. Recent studies find that several governance
mechanisms play an important role in determining crash risk. Boubaker et al. (2014) argue
that firms with substantial excess control rights due to the presence of large controlling
shareholders tend to disclose less firm-specific information and therefore experience more
stock price crashes. In another study examining several attributes of corporate governance,
Andreou et al. (2016) find that crash risk is positively related to CEO stock option incentives
but is negatively associated with board size and inside directors’ ownership. Our paper is
related to those two studies, although we focus on an external governance mechanism across
27
capital markets and associated with a different class of investors, namely external creditors
and particularly short-term debt lenders.
In parallel, several studies highlight the influence of external monitoring mechanisms
on future stock price crash risk. For instance, DeFond et al. (2014) find that the adoption of
better accounting standards reduces future crash risk by improving the disclosure of firm-
specific information and comparability. An and Zhang (2013) and Callen and Fang (2013)
both show that the presence of institutional investors can improve governance monitoring
mechanisms, thereby constraining managerial discretion and reducing future crash risk.
Specifically, institutional ownership of dedicated investors can limit managers’ ability to
conceal unfavorable news, while transient institutional holdings incentivize managers to
hide bad news to prevent transient investors from large short-term selling. Similarly, Chang
et al. (2016) show that stock liquidity gives rise to crash risk because high liquidity attracts
more transient investors, who focus on firms’ short-term earnings and hence induce
managers to withhold adverse information. On the other hand, Callen and Fang (2015a)
further find that strong religion acting as a social norm can inhibit managers’ bad-news-
hoarding activities and render lower stock price crash risk. Callen and Fang (2015b) contend
that sophisticated short sellers can identify managers’ bad news hoarding and seek profit
from those firms; therefore, high levels of short-selling inflate future price crash risk.
Recently, Callen and Fang (2016) provide evidence that auditor tenure reduces the likelihood
of stock price crashes, consistent with the auditor-client relationship enabling auditors to
develop client-specific knowledge and enhancing their ability to deter and detect managerial
withholding of adverse information. Although the above studies demonstrate several
external monitoring mechanisms affecting the likelihood of future stock price crashes, they
have not examined the role that external financing decisions, especially those regarding the
28
choice of debt maturity, plays in reducing stock price crash risk. It is this gap that our study
seeks to fill.
2.2.3. Hypothesis development
A substantial body of research suggests that debt plays an important role in corporate
governance, in large part through its reliance on contractual provisions, which require the
borrowers to meet minimum financial criteria, increase information disclosure, and operate
within bounds specified by creditors (e.g., Williamson, 1988). Debt contracting terms,
especially restrictive covenants in a loan contract, are largely written on information from
the borrowers’ financial statements. Thus, debt holders, including banks, have an incentive
to monitor the financial statements provided by the borrowers in order to detect any
accounting irregularities in a timely manner and protect their financial stake in the borrowers
(e.g., Fama, 1985; Diamond, 1991b; Rajan and Winton, 1995). However, there is also a
significant amount of loan research arguing that monitoring by debt holders, particularly
banks, may not necessarily improve the borrowers’ information transparency because the
banks can achieve information monopoly due to their private access to inside information of
the borrowers (Sheard, 1989; Boot, 2000; Vashishtha, 2014). This argument is corroborated
by Dass and Massa (2011), who suggest that increased monitoring by banks creates a higher
level of information asymmetry between borrowers and other market participants, leading
to lower stock liquidity. One possible reason for the inconclusive findings regarding the
relation between debt financing and information transparency is the failure of prior research
to take debt maturity structure into account, given the monitoring effect can differ
substantially among different types of debt.
We therefore focus on the relation between debt maturity and stock price crash risk
and formulate our hypotheses by intersecting the intuitions associated with the two strands
of literature reviewed above (i.e., those on debt maturity and stock price crash risk). Short-
29
term debt subjects managers to frequent monitoring, thus effectively reducing managerial
discretion and enhancing information disclosure (Rajan and Winton, 1995). While it is
possible that long-term debt holders can play a monitoring role (see footnote 5 above), they
are generally less effective monitors than short-term debt holders. Short-term debt lenders
are able to scrutinize borrowers and gather information about their financial conditions and
future prospects, especially when the short-term debt comes up for renewal and borrowers’
creditworthiness is re-evaluated (Graham et al., 2008).6 Overall, the stringent monitoring of
short-term debt forces firms to release relevant and reliable information that is likely to be
above and beyond the disclosure required by covenant terms. Importantly, the frequent
scrutiny by short-term debt lenders and debt markets restrains managers from arbitrarily
concealing bad news. Put simply, short-maturity debt gives managers fewer opportunities to
withhold adverse information, which in turn facilitates the reflection of such information in
stock prices on a more timely and regular basis. To the extent that bad news hoarding leads
to higher stock price crash risk (e.g., Jin and Myers, 2006; Hutton et al., 2009; Callen and
Fang, 2015a; and Kim and Zhang, 2016), the decrease in bad-news-hoarding opportunities
induced by short-term debt should reduce extreme downward price corrections upon sudden
revelation of the previously accumulated and concealed negative information. Hence, we
formulate our first hypothesis as follows:
H1: Firms with a higher proportion of short-term debt are associated with lower
future stock price crash risk.
If the negative relation between short-term debt and crash risk is indeed attributed to
the monitoring role of short-term debt in reducing bad news hoarding by opportunistic and
6
In addition to short-term lenders, other debt market participants such as investors, underwriters, rating
agencies, and analysts may also exercise their monitoring function (Datta et al., 2005), thus helping to promptly
reflect corporate information in stock prices.
30
self-serving managers, then we expect this relation to be more pronounced among firms
likely to be associated with greater agency or information problems, such as those with
weaker governance, higher information asymmetry, and excessive risk-taking (Kim et al.,
2011a; Callen and Fang, 2015a; Andreou et al., 2016). The rationale for this prediction is
that the monitoring function of short-term debt is most effective at enhancing information
disclosure among firms in which managers are poorly disciplined, harder to monitor, and
allowed greater discretion. In firms with weaker governance monitoring mechanisms,
managers would be less accountable for not releasing information on a timely basis (Bhojraj
and Sengupta, 2003) or for not providing high-quality information (Bae et al., 2006).
Meanwhile, in firms with lower corporate transparency, investors are less able to monitor
managerial performance (Bushman and Smith, 2001) and are more likely to misprice
securities (Lee et al., 2014). Furthermore, managers of riskier firms are both more difficult
to monitor (Demsetz and Lehn, 1985) and more likely to conceal unfavorable news to avoid
being perceived by investors as taking on excessive risk (Callen and Fang, 2015a). Overall,
to the extent that short-term debt lenders effectively serve as a substitute for alternative
monitoring mechanisms that are lacking in those firms, short-term debt should exert a more
pronounced impact to curb managers’ bad-news-hoarding activities and contribute to the
reduction of future crash risk. These arguments lead to three further hypotheses:
H2: The relation between short-term debt and future stock price crash risk is
stronger for firms with weaker governance.
stronger for firms with a higher degree of information asymmetry.
stronger for firms with greater risk-taking.
31
2.3. Research design
2.3.1. Data and sample
We measure U.S. firms’ crash risk using stock return data from the Center for
Research in Security Prices (CRSP) from 1989 through 2014. We collect firm accounting
and financial data from Compustat annual files. Our sample period starts in 1989 because
our explanatory variables (lagged by one year) are computed from 1988, the first year for
which we can estimate discretionary accruals using the cash flow statement method.
Following previous research (e.g., Hutton et al., 2009; Kim et al., 2016; Chang et al., 2017),
we exclude firms (i) with year-end share prices below $1, (ii) with fewer than 26 weeks of
stock return data in each fiscal year, (iii) with negative total assets and book values of equity,
(iv) operating in financial (SIC codes 6000–6999) or public utility (SIC codes 4900–4999)
industries, and (v) with insufficient data to calculate the variables used in our regressions.
Our final sample consists of 7,712 firms and 53,052 firm-year observations.
2.3.2. Measuring short-maturity debt
Following prior studies on debt maturity (e.g., Johnson, 2003; Datta et al., 2005;
Brockman et al., 2010; Harford et al., 2014), our main proxy for short-maturity debt is the
proportion of total debt maturing in three years or less, ST3. In our robustness tests, we
consider alternative measures of short-term debt, namely, the fraction of debt maturing
within one (ST1), two (ST2), and five years (ST5), as well as the ratio of short-term debt to
total assets (ST_TA). We also use a new measure of very short-term debt, that is, short-term
debt due within one year net of the current proportion of long-term debt that is maturing
(STNP1). In addition, in further analysis using a sample of new debt issues, we measure debt
maturity as the maturities of new bond or loan issues, in logarithmic form.
32
2.3.3. Measuring stock price crash risk
Stock price crash risk reflects the tendency of extreme negative returns on individual
firms. We follow Jin and Myers (2006) and compute alternative measures of crash risk using
firm-specific weekly returns. Based on Hutton et al. (2009), we first estimate the following
expanded market model to compute those weekly returns:
𝑟𝑗,𝜏 = 𝛼𝑗 + 𝛽1,𝑗 𝑟𝑚,𝜏−1 + 𝛽2,𝑗 𝑟𝑖,𝜏−1 + 𝛽3,𝑗 𝑟𝑚,𝜏 + 𝛽4,𝑗 𝑟𝑖,𝜏 + 𝛽5,𝑗 𝑟𝑚,𝜏+1 + 𝛽6,𝑗 𝑟𝑖,𝜏+1 + 𝜀𝑗,𝜏 (1)
where rj,τ is the return on stock j in week τ, rm,τ is the return on CRSP value-weighted market
index, and ri,τ is the Fama and French value-weighted industry index in week τ. Following
Dimson (1979), we include the lead and lag terms to correct for nonsynchronous trading.
Following Wang et al. (1997) and Bartholdy and Peare (2005), we estimate weekly returns
from Wednesday to Wednesday to avoid the contaminating effects from weekends and
Mondays. The firm-specific return for stock j in week τ (Wj,τ) is measured by the natural
logarithm of one plus the residual return from Eq. (1).7
As in Chen et al. (2001) and Kim et al. (2011a, 2011b), our first crash risk measure
is the negative conditional skewness of firm-specific weekly returns (NCSKEW). We
calculate NCSKEW for firm j over fiscal year t by taking the negative of the third moment
of firm-specific weekly returns for each year and dividing it by the standard deviation of
firm-specific weekly returns raised to the third power. A stock with high NCSKEW
represents a highly left-skewed return distribution and a high probability of a price crash.
The formula for the negative conditional skewness for firm j in year t is as follows:
3 2 3/2
𝑁𝐶𝑆𝐾𝐸𝑊𝑗,𝑡 = −[𝑛(𝑛 − 1)3/2 ∑ 𝑊𝑗,𝜏 ]/[(𝑛 − 1)(𝑛 − 2)(∑ 𝑊𝑗,𝜏 ) ] (2)
7
Jin and Myers (2006) and Kim et al. (2011a, 2011b) estimate firm-specific weekly returns using an alternative
market model that does not include the industry index. In untabulated robustness checks, we employ this model
to recalculate our crash risk measures and obtain qualitatively similar results.
33
where Wj,τ is the firm-specific weekly return as defined above and n is the number of weekly
returns in fiscal year t.
Our second measure of firm-specific crash risk is “down-to-up volatility,” which is
calculated as follows:
2 2
𝐷𝑈𝑉𝑂𝐿𝑗,𝑡 = 𝑙𝑜𝑔{(𝑛𝑢 − 1) ∑𝐷𝑜𝑤𝑛 𝑊𝑗,𝜏 /(𝑛𝑑 − 1) ∑𝑈𝑝 𝑊𝑗,𝜏 } (3)
where nu and nd are the number of up and down days over the fiscal year t, respectively. For
each firm j over year t, we separate firm-specific weekly returns into down (up) weeks when
the weekly returns are below (above) the annual mean. We separately calculate the standard
deviation of firm-specific weekly returns for each of the two groups. Then, DUVOL is the
natural logarithm of the ratio of the standard deviation in the down weeks to the standard
deviation in the up weeks. Chen et al. (2001) suggest that a high DUVOL indicates a more
left-skewed distribution. We note that DUVOL is less likely to be affected by the number of
extreme returns as it does not involve third moments.
2.3.4. Control variables
Following prior studies of stock price crash risk (e.g., Chen et al., 2001; Jin and
Myers, 2006), we employ the following set of control variables: stock turnover (DTURN),
stock return volatility (SIGMA), firm size (SIZE), market-to-book ratio (MB), leverage (LEV),
return on assets (ROA), lagged negative conditional skewness (NCSKEW), and earnings
quality (ACCM). The control variables are all lagged one period and measured as follows:
DTURNt–1 is the difference between the average monthly share turnover over fiscal year t1
and t2. SIGMAt–1 is the standard deviation of firm-specific weekly returns in fiscal year t1.
RETt–1 is the average firm-specific weekly returns in fiscal year t1. SIZEt–1 is the log of the
market value of equity in year t1. MBt–1 is the market value of equity divided by the book
value of equity in year t1. LEVt–1 is the book value of total liabilities scaled by total assets
34
in fiscal year t1. ROAt–1 is income before extraordinary items divided by total assets at the
end of fiscal year t1. NCSKEWt–1 is the negative conditional skewness for firm-specific
weekly returns in fiscal year t1. ACCMt–1 is defined as the absolute value of discretionary
accruals, where discretionary accruals are the residuals estimated from the modified Jones
model (Hutton et al., 2009). Finally, we control for Fama and French 48-industry and year
effects.8 To mitigate the influence of outliers, we winsorize all the continuous variables at
the 1% and 99% levels. We provide detailed variable definitions in the Appendix.
2.4. Empirical results
2.4.1. Descriptive statistics
Table 1 presents the descriptive statistics for all the variables used in our regressions.
The mean values of two stock price crash risk measures, NCSKEW and DUVOL, are –0.085
and –0.055, respectively, which are quite similar to those reported in Kim et al. (2011b).
Short-term debt, ST3, has a mean value of 0.541, which is in line with the reported means in
Johnson (2003) and Custódio et al. (2013).9 The summary statistics of the other variables
are largely consistent with those reported in prior research, and thus are not discussed herein
to preserve space.
Moreover, in untabulated correlation analysis, the two crash risk measures, NCSKEW
and DUVOL, are significantly and negatively correlated with short-maturity debt, ST3; their
correlation coefficients are –0.059 and –0.064, respectively. This finding lends initial
support to our prediction that short-term debt induces a lower probability of future stock
8
Our results are robust to controlling for industry fixed effects defined by 2-digit or 4-digit SIC codes.
9
Custódio et al.’s (2013) measure of debt maturity is the fraction of debt maturing after 3 years, that is, 1 
ST3.
35
price crashes. Consistent with prior research, we find the two crash risk measures to be
positively correlated with each other, with a very high correlation coefficient of 0.961.
[Insert Table 1 about here]
2.4.2. Baseline regression results
We examine the impact of short-term debt on future stock price crash risk by
estimating the following regression model:
𝐶𝑟𝑎𝑠ℎ 𝑅𝑖𝑠𝑘𝑗,𝑡 = 𝛽0 + 𝛽1 𝑆𝑇3𝑗,𝑡−1 + 𝛽2 𝐷𝑇𝑈𝑅𝑁𝑗,𝑡−1 + 𝛽3 𝑆𝐼𝐺𝑀𝐴𝑗,𝑡−1 + 𝛽4 𝑅𝐸𝑇𝑗,𝑡−1 +
𝛽5 𝑆𝐼𝑍𝐸𝑗,𝑡−1 + 𝛽6 𝑀𝐵𝑗,𝑡−1 + 𝛽7 𝐿𝐸𝑉𝑗,𝑡−1 + 𝛽8 𝑅𝑂𝐴𝑗,𝑡−1 + 𝛽9 𝑁𝐶𝑆𝐾𝐸𝑊𝑗,𝑡−1 +
𝛽10 𝐴𝐶𝐶𝑀𝑗,𝑡−1 + 𝜀𝑗,𝑡 (4)
Table 2 presents the regression results for this model. In Columns (1) and (4), we
regress two crash risk measures, NCSKEW and DUVOL, on short-maturity debt, ST3, and
the control variables. In Columns (2) and (5), we include year fixed effects to control for a
secular increase in short-term debt (Custódio et al., 2013). We further control for both year
and industry fixed effects in Columns (3) and (6). The results across the table show that
short-term debt is significantly and negatively associated with one-year ahead stock price
crash risk. For example, in our preferred baseline models in Columns (3) and (6), the
coefficients on ST3 are –0.048 (t-stat = –4.30) and –0.023 (t-stat = –4.57), respectively. This
finding suggests that firms with more short-term debt experience lower future stock price
crash risk, consistent with the notion that the monitoring of short-maturity debt restricts
managers from hiding bad news, thus leading to a lower likelihood of firms’ future stock
price crashes.
We further evaluate the economic significance of the effect of short-maturity debt on
future crash risk. Following Hutton et al. (2009), we compare the values of cash risk
corresponding to the 25th and 75th percentile values of short-term debt (0.217 and 0.938,
36
respectively), while keeping all other control variables at their sample means. In Columns
(3) and (6), we find that the decrease in NCSKEW (DUVOL) is 0.035 (0.017) or 40.72%
(30.15%) relative to the sample mean, when there is an increase from the 25th to 75th
percentiles of the distribution of short-maturity debt. Moreover, the economic impact of
short-term debt is twice as large as the impact of earnings quality (ACCM) on crash risk
(17.12% and 7.75%, respectively, for NCSKEW and DUVOL). These results suggest that the
effect of short-maturity debt on crash risk not only is statistically significant but also has
large economic significance. We thus conclude that our baseline regression results provide
strong support for Hypothesis H1.
The results regarding the control variables are generally consistent with prior studies.
The coefficients on stock turnover (DTURN) and stock return volatility (SIGMA) are
significant and positive, which is consistent with Chen et al.’s (2001) finding that stocks
with higher turnover and higher return volatility are likely to experience more price crashes
in the future. The coefficients on past returns (RET) and market-to-book ratio (MB) are also
significantly positive, in line with Harvey and Siddique (2000) and Chen et al. (2001). To
the extent that high stock returns and high market-to-book signal the buildup of a stock price
bubble, these variables are likely to be associated with higher future crash risk. The
coefficient on leverage (LEV) is negative, which seems to reflect firms’ endogenous capital
structure choice as less crash-prone firms may have stronger incentives to accumulate debt
(Hutton et al., 2009; Kim et al., 2011b). Moreover, the coefficients on firm size (SIZE),
lagged crash risk (NCSKEW), and earnings quality (ACCM) are positive, which is consistent
with the evidence documented in prior studies (e.g., Chen et al., 2001; Hutton et al., 2009).
Finally, we find a positive association between profitability (ROA) and crash risk,
corroborating the findings of Kim et al. (2014) and Callen and Fang (2015a).
37
2.4.3. Identification strategies and robustness checks
2.4.3.1. Dealing with endogeneity
One major concern about the baseline results reported in Table 2 is that the debt
maturity structure may be endogenous, in which case the estimated negative effect of short-
term debt on crash risk would be biased and inconsistent, and our inference thus far would
be invalid. One main source of this endogeneity is the potential presence of omitted variables
as short-term debt may be correlated with unobserved firm-specific characteristics that affect
future crash risk. Another common source of endogeneity includes reverse causality and
simultaneity since crash risk may explain variation in short-term debt or that the two
variables may be jointly determined.10 To address these potential endogeneity problems, we
employ several estimation approaches.
First, we run firm fixed-effects (FE) and first-differences (FD) regressions to control
for time-invariant unobserved firm characteristics and alleviate the potential omitted-
variable bias due to heterogeneity. The FD regression further addresses the concern that our
main measure of short-maturity debt may reflect past debt maturity decisions as it includes
the proportion of long-term debt that is maturing. By estimating this model, we can better
capture how a change in debt maturity structure affects a change in the likelihood of future
stock price crashes. In Panel A of Table 3, the results from the FE and FD regressions show
that the relation between short-term debt and future stock price crash risk remains negative
and significant for both crash risk measures. This suggests that our main findings continue
to hold after controlling for unobserved heterogeneity.
10
We note this endogeneity concern is less likely to affect our results because we examine the impact of current
short-term debt on future stock price crash risk. Given that short-term debt evolves over time (Custódio et al.,
2013), it is unlikely that future crash risk affects current short-term debt.
38
Our second strategy to address endogeneity involves using the dynamic system
generalized method of moments (SYSGMM) approach (Blundell and Bond, 1998), which
takes into account the dynamics of stock price crash risk, while accounting for other sources
of endogeneity in the model (e.g., Kim et al., 2014). We employ the SYSGMM estimator
because our estimated model of stock price crash risk is a dynamic panel data model that
includes lagged crash risk as a regressor (NCSKEWt1). Using the traditional ordinary least
squares (OLS) method for estimating the model might lead to biased and inconsistent
estimates of the coefficients because the dynamic term, lagged crash risk, may be correlated
with unobservable firm-specific factors and this potential correlation would not be
eliminated in the FE and FD regressions (Baltagi, 2013). In applying the SYSGMM, we
estimate Eq. (4) in both levels and first-differences using appropriate instruments for the two
endogenous variables, crash risk (NCSKEWt1) and short-term debt (ST3t1). In the levels
equations, our instruments for NCSKEWt1 and ST3t1 include their lagged values in first
differences. In the first-differenced equations, our instruments for NCSKEWt1 and ST3t1
are the lagged values of NCSKEWt1 and ST3t1, both in levels.11
Panel B of Table 3 reports the results from our SYSGMM regressions. We find that
the coefficient on short-term debt is significantly negative in both models, consistent with
the baseline results. In terms of diagnostic tests, the second-order autocorrelation (AR2) and
J-tests provide no evidence of second-order autocorrelation and over-identification. This
suggests that our instruments are valid and that the specifications we use are appropriate.
11
Specifically, in the levels equations, we use NCSKEWt2, NCSKEWt3,…, NCSKEW1 as instruments for
NCSKEWt1. In the first-differenced equations, we use NCSKEWt2, NCSKEW t3,…, NCSKEW1 as instruments
for NCSKEWt1. We construct the instrument matrix for ST3t1 in a similar way.
39
We perform two additional tests to further address the endogeneity concern. First,
we employ the instrumental variable (IV) approach, in which we use the term structure of
interest rates (TERMSTR) as an instrument for short-maturity debt; TERMSTR is measured
as the difference between the yield on 10-year government bonds and the yield on 6-month
Treasury bills. We argue that this instrument plausibly satisfies both the relevance and
exclusion conditions of a valid IV. First, prior empirical studies provide evidence of a
significant and positive relation between term structure and short-maturity debt (e.g.,
Barclay and Smith, 1995; Johnson, 2003; Brockman et al., 2010). For example, these studies
suggest that firms prefer short-maturity debt to long-maturity debt because the former source
of debt financing is typically less costly, unless the yield curve is inverted. Second, regarding
the exclusion condition, term structure and stock price crash risk are not likely to be
correlated, unless via the debt maturity channel. This is because that the changing pattern of
the yield curve is unlikely to affect managers’ bad-news-hoarding activities directly. Our IV
regression results in Panel C of Table 3 show that short-term debt remains significantly and
negatively related to both measures of future stock price crash risk.
Second, we further mitigate the omitted-variable bias, a major source of the
endogeneity concern, by controlling for the following set of variables that may affect future
crash risk but are potentially related to debt maturity choice: (i) firm quality, measured as
abnormal earnings (Custódio et al., 2013); (ii) credit quality, measured as bond rating
(Diamond, 1991a); and (iii) the degree of financial constraint, proxied by the dividend
payout ratio (Faulkender and Wang, 2006). We control for firm quality and credit quality
because high-quality firms are less likely to experience future price crashes, while, under
asymmetric information, they are more likely to issue short-term debt as a signal of good
future prospects (Flannery, 1986; Ozkan, 2000). We include the degree of financial
constraint since constrained firms tend to rely more on short-maturity debt as they are likely
40
to be screened out of the long end of the maturity spectrum (Diamond, 1991a). Firms with a
low dividend payout ratio or with no rating are more constrained than those with the opposite
characteristics.12 Meanwhile, these firms may use discretionary accruals opportunistically to
attract external financing (Dechow et al., 1996; Dechow et al., 2011); their aggressive
earnings management could, in turn, result in higher synchronicity risk and future stock price
crash risk (Hutton et al., 2009). In Panel D of Table 3, the results show that the coefficient
on short-term debt is significantly negative, whether we include the additional control
variables in our models separately or together, or measure crash risk using NCSKEW or
DUVOL.
In summary, the results from the above tests show that, controlling for heterogeneity
and endogeneity, short-term debt exerts a negative impact on future stock price crash risk.
This finding is consistent with our baseline results and provides further support for
Hypothesis H1.
2.4.3.2. Other robustness checks
We perform a series of robustness checks using alternative measures of our key
variables and controlling for a set of fundamental risk variables that may also affect future
stock price crash risk. Following prior research on debt maturity (e.g., Barclay and Smith,
1995), we additionally measure short-term debt as the proportion of total debt maturing
within one (ST1), two (ST2), or five years (ST5). We further use an alternative measure of
12
In untabulated results, we control three additional measures of financial constraint, namely, the Kaplan and
Zingales (1997) index, the Whited and Wu (2006) index, and the Hadlock and Pierce (2010) index. The
untabulated results are consistent with our main findings.
41
short-term debt maturing within three years (ST3_TA) by scaling it using total assets, rather
than total debt as in our main analysis. Following Huang et al. (2016), we consider a new
measure of very short-term debt (STNP1), which we define as the ratio of debt in current
liabilities net of long-term debt due in one year, scaled by total debt. By using STNP1, we
can rule out the effect of the long-term debt that is maturing. Panel A of Table 4 reports the
results for five alternative measures of short-term debt. We find that the coefficients on those
measures, namely, ST1, ST2, ST5, ST3_TA, and STNP1, are all significant and negative for
both crash risk measures.13
Following Jin and Myers (2006) and Callen and Fang (2015a), we further measure
stock price crash risk as the number of crashes minus the number of jumps over the fiscal
year (COUNT). Specifically, we first define crash weeks in a given fiscal year as those during
which a firm experiences firm-specific weekly returns 3.09 standard deviations below the
mean firm-specific weekly returns over the whole fiscal year, with 3.09 chosen to generate
a frequency of 0.1% in the normal distribution. Likewise, when the firm-specific weekly
return is 3.09 standard deviations above its mean in a fiscal year, we define those weeks as
jump weeks. As in Hutton et al. (2009), Kim et al. (2011b), and Chang et al. (2017), we also
measure future stock price crash risk as the likelihood that a firm experiences more than one
price crash week in a fiscal year (CRASH). The results in Panel B of Table 4 show that all
six alternative measures of short-term debt are significantly and negatively related to
COUNT. Furthermore, short-term debt has a negative impact on CRASH, although this
impact is only significant for four measures of short-term debt. In further (untabulated)
13
In unreported results, we further measure short-term debt using the ratio of short-maturity loans (i.e., loans
that have a time-to-maturity of less than three years) to total loans. Our main inferences remain qualitatively
the same.
42
robustness checks, we find that our main results continue to hold when we use two- or three-
year-ahead crash risk measures.
In the main analysis, we measure leverage as the ratio of total debt to total assets,
which is one of the most conventional measures used in prior capital structure studies (e.g.,
Graham et al., 2015; Öztekin, 2015), including those that examine both debt maturity and
leverage in a single model specification as we do in this paper (e.g., Billet et al., 2007).
However, since total debt includes short-maturity debt, one possible concern is that the
coefficient on short-term debt may not clearly identify the effect of short-maturity debt on
crash risk. To mitigate this concern, we use an alternative measure of leverage, which is
calculated as long-term debt divided by total assets. As can be seen in Panel C of Table 4,
the coefficients on short-term debt remain positive and statistically significant for our main
crash risk measures.
Finally, to ensure that our findings are not driven by fundamental risk factors, we
replicate our main analysis by further controlling for a set of variables that capture risk,
including earnings volatility (the standard deviation of the ratio of earnings, excluding
extraordinary items and discontinued operations, to lagged total equity during the past five
years), cash flow volatility (the standard deviation of the ratio of cash flow to total assets in
the past five years), sales volatility (the standard deviation of the ratio of sales to total assets
in the past five years),14 and beta (the covariance between the individual firm return and the
market return divided by the variance of the market return over a fiscal year). In Panel D of
Table 4, the coefficients on short-term debt are significantly negative after including these
additional controls. Overall, we conclude that our main findings are generally robust to the
14
Our findings are unaffected if the volatility measures (i.e., earnings, cash flow, and sales volatility) are
calculated using data in the past three years.
43
use of alternative measures of future stock price crash risk, short-maturity debt, and leverage,
as well as the inclusion of various risk factors.
2.4.3.3. New debt issues
Although our approach of calculating short-maturity debt based on the balance sheet
data is widely used in the literature, as mentioned, one concern about this approach is that
the short-term debt ratio may be affected by the proportion of long-term debt that is coming
due. This fraction of maturing long-maturity debt is unlikely to have the desired monitoring
effect on managers’ bad-news-holding behavior as our hypothesis predicts. We note that in
two robustness checks above, we have, to an extent, addressed this concern by (i) running a
change (FD) regression and (ii) focusing on short-term debt due within one year, net of the
current proportion of maturing long-term debt, STNP1. In this section, we follow prior
research (e.g., Guedes and Opler, 1996; Brockman et al., 2010; Custódio et al., 2013) and
further use an incremental approach in which we focus on newly issued debt. This
incremental approach better captures the relations between debt maturity structure and firm-
specific variables at all points of the maturity spectrum (Guedes and Opler, 1996).
Importantly, it also allows us to better study the causal effect of the maturities of new debt
issues on one-year-ahead crash risk, while avoiding potential endogeneity problems due to
reverse causality and simultaneity.
Following Custódio et al. (2013), we obtain data on both bond issues and private
bank loan issues. Data on new bonds are from the Mergent Fixed Income Securities Database
(FISD) and data on new loans come from the Loan Pricing Corporation’s Dealscan database,
which contains issuance-level information on syndicated bank loans. We first construct a
transaction-level (unconsolidated) sample of new debt issues, including both bonds and
44
loans. Merging data of new debt issues with Compustat leaves us with an unconsolidated
sample of 4,233 unique firms and 32,785 debt issues. We measure the debt maturity of a
debt issue (DEBT_MAT) as the natural logarithm of the maturity of the issue.
As robustness checks, we examine different samples of debt issues. First, we
construct an unconsolidated sample using data on newly issued private bank loans only. Our
sample of loan issues consists of 24,845 transactions. Second, following Brockman et al.
(2010), we further construct a consolidated (firm-level) sample of both bond and loan issues
to deal with the possibility that firms have multiple debt issues within a fiscal year. We
measure the debt maturity of those multiple issues as the natural logarithm of the issue-size-
weighted maturity (WAVG_MAT).15 Our (firm-level) consolidated sample consists of 16,685
firm-year observations.
Table 5 presents the regression results for both the unconsolidated and consolidated
samples. In the first two columns of Panel A, the coefficient on debt maturity (DEBT_MAT)
is significantly positive for both NCSKEW and DUVOL. This finding continues to hold after
we control for the size of new debt issues (DEBT_SIZE). Taken together, our results indicate
that firms that issue debt with longer (shorter) maturity are more (less) likely to experience
future stock price crashes, which is also in line with our prediction. Panel B shows the effect
of the maturities of new loan issues on future stock price crash risk. The coefficient on loan
maturity (LOAN_MAT) is positive and significant, whether we control for the amount of the
loan issue (LOAN_SIZE) or not. Its magnitude seems higher than the magnitude of the
coefficient on debt maturity in Panel A, which is consistent with our expectation. Panel C
presents the results for the consolidated sample of new debt issues. In the last two columns
of the panel, we further control for the total amount of firms’ multiple debt issues within a
15
The results (untabulated) are robust if we use the natural logarithm of the equal-weighted maturity.
45
fiscal year (SUM_SIZE). The results show that the coefficients on WAVG_MAT are positive
and significant, for both crash risk measures, NCSKEW and DUVOL. Overall, the evidence
from different samples of new debt issues is consistent with our baseline regression results
obtained using the balance sheet data and provides additional support for Hypothesis H1.
2.4.4. Corporate governance mechanisms and short-maturity debt
We next examine whether the negative relationship between short-term debt and
future stock price crash risk is attenuated by the strength of firms’ governance monitoring
mechanisms, as predicted by Hypothesis H2. We employ three proxies for the effectiveness
of corporate governance, namely, the proportion of total institutional ownership (INST), the
proportion of long-term-oriented institutional ownership (LTINST), and the governance
index (GINDEX). The former two variables measure the monitoring power of institutional
investors, while the latter proxies for monitoring by equity markets. Using data from
Thomson 13F, we calculate the proportion of total institutional ownership (INST) as the
percentage of shares held by institutional owners. We further compute the proportion of
long-term (non-transient) institutional ownership (LTINST) as the percentage of common
shares owned by dedicated and quasi-indexer institutions.16 We focus on dedicated investors
because prior research finds that they play an important monitoring role in curbing
managerial myopic behavior (e.g., Bushee, 1998, 2001). We further include quasi-indexers
because recent evidence suggests that these investors are not passive owners as they hold
sway over managers and typically demand greater firm transparency and public information
production (e.g., Boone and White, 2015; Appel et al., 2016). We exclude transient investors
16
We follow Bushee (1998, 2001) and categorize institutional investors into quasi-indexers, transient, or
dedicated based on their portfolio turnover, diversification, and expected investment horizon.
46
from our calculation of LTINST because these investors have shorter term horizons and
hence fewer incentives to understand and monitor firms (e.g., Andreou et al., 2016; Chang
et al., 2017). Regarding the governance index (GINDEX), we collect the data on this variable
from the RiskMetrics database. GINDEX measures the strength of antitakeover provisions
(Gompers et al., 2003); firms with fewer antitakeover provisions have better shareholder
rights protection and lower managerial discretion.
We partition the full sample into subsamples based on the annual median values of
the governance measures and report the results in Table 6. In Panels A and B, we divide the
sample into firms with strong (weak) external institutional monitoring, defined as those with
above-median (below-median) INST and LTINST, respectively. In both panels, we find that
the effect of short-term debt on crash risk is always negative, but only significant, at least at
the 5% level, when (long-term-oriented) institutional shareholdings are lower. This finding
holds for both measures of crash risk, NCSKEW and DUVOL. In Panel C, we define weak
(strong) governance firms as those with above- (below-) median GINDEX. The results show
that the coefficient on short-term debt is negative for both groups of firms, but significant at
the 1% level for those with weaker governance (Columns (1) and (3)) while insignificant for
those with stronger governance (Columns (2) and (4)).
To compare the differences between the coefficients on short-maturity debt (ST3)
across the subsamples, we also follow the approach in Kim et al. (2011b) and Callen and
Fang (2015a) and estimate an alternative specification in which we include an interaction
term between ST3 and each governance measure defined above. Our specification is similar
to Eq. (4) for the whole sample, with two additional regressors being the interaction term
and the governance measure of interest. The results (untabulated) suggest that the coefficient
on the interaction term ST3 × INST is significantly negative for both measures of crash risk,
NCSKEW and DUVOL. This finding is consistent with the above evidence that institutional
47
ownership moderates the negative impact of short-maturity debt on future crash risk. We
obtain qualitatively similar results regarding the interaction term between ST3 and either
LTINST or GINDEX, although ST3 × GINDEX becomes less significant.
Overall, our results support Hypothesis H2 that the negative effect of short-term debt
on crash risk is more pronounced among firms with weak corporate governance mechanisms,
consistent with short-term debt serving as an effective monitoring tool to reduce managers’
bad-news-hoarding activities when the monitoring by institutional investors and
shareholders is insufficient. Our evidence is also in line with prior studies on stock price
crash risk. For example, Kim et al. (2011b) find that effective external monitoring can
moderate the positive impact of tax avoidance on crash risk, while Callen and Fang (2015a)
show that the negative relation between religion and crash risk becomes weaker among firms
with strong governance mechanisms.
2.4.5. Information asymmetry and short-maturity debt
We next study whether the relation between short-term debt and future stock price
crash risk is more pronounced among firms with a weak information environment, as
predicted by Hypothesis H3. Following prior research, we measure the degree of information
asymmetry using analyst forecast error (FERR) (Callen and Fang, 2015b), analyst forecast
dispersion (DISPER), and R&D intensity (RD) (Custódio et al., 2013). We obtain analyst
earnings forecast data from the I/B/E/S database. We define analyst forecast error (FERR)
as the absolute value of the difference between actual earnings per share and the consensus
analyst forecast, divided by the consensus analyst forecast. The dispersion of analyst
forecasts (DISPER) is the standard deviation of analyst forecasts divided by the consensus
analyst forecasts. R&D intensity (RD) is the ratio of research and development expenditure
48
to total assets. We expect higher analyst forecast error, higher forecast dispersion, and
greater R&D intensity to be associated with higher levels of information asymmetry (e.g.,
Healy and Palepu, 2001; Zhang, 2006; and Custódio et al., 2013).
In Table 7, we split the sample into firms with high and low levels of asymmetric
information based on those three measures. The results in Panel A show that the coefficient
on short-term debt is significantly negative at the 1% level for the subsample of firms with
high analyst forecast error. However, for firms with low analyst forecast error, the coefficient
is insignificant or only marginally significant at the 10% level. Similarly, in Panel B, the
coefficient on short-term debt is significantly negative for firms with greater dispersions in
analysts’ forecasts, but either insignificantly or marginally negative for those with lower
dispersions. Panel C also reveals that the coefficient on short-term debt is statistically
significant only for R&D-intensive firms.
As a robustness check, we interact short-term debt with each measure of information
asymmetry defined above and add both this interaction term and the information asymmetry
measure in question to the baseline model (Eq. (4)). As mentioned above, this interaction
term approach enables us to test for differences in the estimated coefficients on short-
maturity debt (ST3) across the subsamples in Table 7. The results (untabulated) show that
the coefficient on the interaction term ST3 × FERR is negative and significant, consistent
with short-maturity debt exerting a stronger mitigating effect on crash risk when firms have
weaker analyst monitoring and a higher degree of information asymmetry. We obtain
qualitatively similar results when we repeat this analysis and include the interaction term
between ST3 and either DISPER or RD; the interaction term ST3 × DISPER is significantly
negative, while the interaction term ST3 × RD is negative but insignificant.
49
Overall, our findings are broadly consistent with Hypothesis H3 that the role of short-
term debt in mitigating managers’ information withholding and lowering future stock price
crash risk is stronger when there is less corporate transparency.
2.4.6. Risk-taking and short-maturity debt
Finally, we investigate whether the relationship between a firm’s short-maturity debt
and future stock price crash risk is conditional on its riskiness, as predicted by Hypothesis
H4. We follow Johnson (2003) and use two proxies for the riskiness of the firm, including
leverage (LEV) and bond rating status (RATEDUM). We classify a firm as being risker if its
leverage is above the sample median or if it does not have a bond rating.
In Panels A and B of Table 8, we partition the sample into firms with high and low
risk-taking based on their leverage ratios and rating status, respectively. In Panel A, the
coefficient on short-maturity debt is significantly negative at the 1% level for the riskier
firms but insignificant for the less risky counterparts. Furthermore, the coefficients on short-
term debt are much larger in magnitude for the riskier subsample. The results in Panel B
show that the coefficient on short-term debt is negative and significant at least at the 5%
level among the unrated group, while it is only marginally significant among the rated group.
To further test for differences in the coefficients on short-maturity debt (ST3) across
the subsamples in Table 8, we adopt the interaction term approach as mentioned above, that
is, we add an interaction term between ST3 with each measure of riskiness to our baseline
model (Eq. (4)). The results (untabulated) reveal that the coefficient on the interaction term
ST3 × LEV is negative and significant, again supporting our prediction that short-maturity
debt has a more pronounced effect on future crash risk among highly leveraged firms. The
results are qualitatively similar when we add the interaction term between ST3 and
50
RATEDUM to our baseline model (i.e., we find ST3 × RATEDUM to be significantly
negative).
Overall, the results in Table 8 are consistent with our Hypothesis H4 that the
influence of short-maturity debt on future crash risk is stronger among firms with a higher
level of risk-taking. Our evidence is also in line with the recent finding by Callen and Fang
(2015a) that the impact of religiosity on future crash risk is more concentrated in riskier
firms.
2.5. Conclusion
This study examines the impact of debt maturity on stock price crash risk. We
provide original evidence of a negative relationship between the use of short-term debt and
the likelihood of subsequent stock price crashes. The results are robust to a battery of
robustness checks, including tests addressing endogeneity concerns, using alternative
measures of debt maturity and crash risk, and focusing on various samples of new debt issues.
Our findings are consistent with the monitoring role of short-term debt serving as an
effective tool to curb managerial bad-news-hoarding behavior, which in turn reduces stock
price crash risk.
We also investigate whether the influence of short-term debt on crash risk is
conditional on corporate governance mechanisms, information asymmetry, and firm
riskiness. Our findings indicate that the mitigating effect of short-term debt on future crash
risk is more pronounced when firms have lower shareholder rights and less (non-transient)
institutional ownership. Furthermore, the negative relationship between short-maturity debt
and future crash risk is stronger among firms with weaker analyst monitoring and higher
51
asymmetric information and among riskier firms with high leverage and without credit rating.
Taken together, these results highlight the importance of short-term debt for firms with
weaker governance, a weaker information environment, or greater riskiness. Put differently,
short-term debt may act as a substitute for corporate governance mechanisms in reducing
managerial bad news hoarding and future stock price crash risk.
Overall, our study complements a growing body of research on stock price crash risk
and debt maturity structure. In the crash risk literature, we provide the first novel evidence
that a corporate financial policy such as debt maturity has a significant influence on high
moments of stock return distribution, above and beyond the effects of other determinants of
crash risk identified by prior studies. In the debt maturity literature, we show that firms can
benefit from short-term debt due to its mitigating effect on future stock price crash risk,
supporting the notion that “short-term debt maturity can be an extremely powerful tool to
monitor management” (Stulz, 2001, p. 172). Together, these add to a greater understanding
of how debt financing can contribute to improving corporate governance and reducing
agency costs. Our study suggests that short-maturity debt enables creditors to constrain
managers’ misconduct, which in turn is mutually beneficial to shareholders’ wealth through
the reduction of stock price crash risk. Finally, our findings provide relevant implications
for equity investors who wish to predict and avoid future stock price crash risk based on
information about debt maturity policy.
52
References
Aivazian, V., Ge, Y. and Qiu, J., ‘Debt maturity structure and firm investment’, Financial
Management, Vol. 34(4), 2005, pp. 107–119.
Almeida, H., Campello, M., Laranjeira, B. and Weisbenner, S., ‘Corporate debt maturity and
the real effects of the 2007 credit crisis’, Critical Financial Review, Vol. 1(1), 2011, pp.
3–58.
An, H. and Zhang, T., ‘Stock price synchronicity, crash risk, and institutional investors’,
Journal of Corporate Finance, Vol. 21, 2013, pp. 1–15.
Andreou, P. C., Antoniou, C., Horton, J. and Louca, C., ‘Corporate governance and firm-
specific stock price crashes’, European Financial Management, 2016, forthcoming.
Antoniou, A., Guney, Y. and Paudyal, K., ‘The determinants of corporate debt maturity
structure: Evidence from France, Germany and the UK’, European Financial
Management, Vol. 12(2), 2006, pp. 161–194.
Appel, I., Gormley, T. A. and Keim, D. B., ‘Passive investors, not passive owners’, Journal
of Financial Economics, Vol. 121(1), 2016, pp. 111–141.
Bae, K., Lim, C. and Wei, J., ‘Corporate governance and conditional skewness in the world’s
stock markets’, Journal of Business, Vol. 79(6), 2006, pp. 2999–3028.
Baltagi, B. D., ‘Econometric analysis of panel data’, 5th Ed. Wiley, 2013.
Barclay, M. and Smith, C. W., ‘The maturity structure of corporate debt’, Journal of Finance,
Vol. 50(2), 1995, pp. 609–631.
Barclay, M., Marx, L. and Smith, C., ‘The joint determination of leverage and maturity’,
Journal of Corporate Finance, Vol. 9(2), 2003, pp. 149–167.
53
Barnea, A., Haugen, R. A. and Senbet, L. W., ‘A rationale for debt maturity structure and
call provisions in the agency theoretic framework’, Journal of Finance, Vol. 35(5), 1980,
pp. 1223–1234.
Bartholdy, J. and Peare, P., ‘Estimation of expected return: CAPM vs. Fama and French’,
International Review of Financial Analysis, Vol. 14(4), 2005, pp. 407–427.
Bhojraj, S. and Sengupta, P., ‘Effect of corporate governance on bond ratings and yields:
The role of institutional investors and outside directors’, Journal of Business, Vol. 76(3),
2003, pp. 455–475.
Billett, M., King, T. and Mauer, D., ‘Growth opportunities and the choice of leverage, debt
maturity, and covenants’, Journal of Finance, Vol. 62(2), 2007, pp. 697–730.
Blundell, R. and Bond, S., ‘Initial conditions and moment restrictions in dynamic panel data
models’, Journal of Econometrics, Vol. 87(1), 1998, pp. 115–143.
Boone, A. L. and White, J. T., ‘The effect of institutional ownership on firm transparency
and information production’, Journal of Financial Economics, Vol. 117(3), 2015, pp.
508–533.
Boot, A. W., ‘Relationship banking: What do we know?’ Journal of Financial
Intermediation, Vol. 9(1), 2000, pp. 7–25.
Boubaker, S., Mansali, H., and Rjiba, H., ‘Large controlling shareholders and stock price
synchronicity’, Journal of Banking & Finance, Vol. 40, 2014, pp. 80–96.
Brockman, P., Martin, X. and Unlu, E., ‘Executive compensation and the maturity structure
of corporate debt’, Journal of Finance, Vol. 65(3), 2010, pp. 1123–1161.
Bushee, B. J., ‘Do institutional investors prefer near-term earnings over long-run value?’
Contemporary Accounting Research, Vol. 18(2), 2001, pp. 207–246.
54
Bushee, B. J., ‘The influence of institutional investors on myopic R&D investment behavior’,
Accounting Review, Vol. 73(3), 1998, pp. 305–333.
Bushman, R. and Smith, A., ‘Financial accounting information and corporate governance’,
Journal of Accounting Economics, Vol. 32(1), 2001, pp. 237–333.
Callen, J. L. and Fang, X., ‘Crash risk and the auditor-client relationship’, Contemporary
Accounting Research, 2016, forthcoming.
Callen, J. L. and Fang, X., ‘Institutional investor stability and crash risk: Monitoring versus
short-termism?’ Journal of Banking & Finance, Vol. 37(8), 2013, pp. 3047–3063.
Callen, J. L. and Fang, X., ‘Religion and stock price crash risk’, Journal of Financial and
Quantitative Analysis, Vol. 50(1–2), 2015a, pp. 169–195.
Callen, J. L. and Fang, X., ‘Short interest and stock price crash risk’, Journal of Banking &
Finance, Vol. 60, 2015b, pp. 181–194.
Chang, X., Chen, Y. and Zolotoy, L., ‘Stock liquidity and stock price crash risk’, Journal of
Financial and Quantitative Analysis, Vol. 52(4), 2017, pp. 1605–1637.
Chen, J., Hong, H. and Stein, J., ‘Forecasting crashes: trading volume, past returns, and
conditional skewness in stock prices’, Journal of Financial Economics, Vol. 61(3), 2001,
pp. 345–381.
Childs, P. D., Mauer, D. C. and Ott, S. H., ‘Interactions of corporate financing and
investment decisions: The effects of agency conflicts’, Journal of Financial Economics,
Vol. 76(3), 2005, pp. 667690.
Custodio, C., Ferreira, M. A. and Laureano, L., ‘Why are US firms using more short-term
debt?’ Journal of Financial Economics, Vol. 108(1), 2013, pp. 182–212.
55
Dass, N. and Massa, M., ‘The impact of a strong bank-firm relationship on the borrowing
firm’, Review of Financial Studies, Vol. 24(4), 2011, pp. 12041260.
Datta, S., Iskandar-Datta, M. and Raman, K., ‘Managerial stock ownership and the maturity
structure of corporate debt’, Journal of Finance, Vol. 60(5), 2005, pp. 2333–2350.
Dechow, P. M., Ge, W. L., Larson, C. R. and Sloan, R. G., ‘Predicting material accounting
misstatements’, Contemporary Accounting Research, Vol. 28(1), 2011, pp. 17–82.
Dechow, P. M., Sloan, R. G. and Sweeney, A. P., ‘Causes and consequences of earnings
manipulation: An analysis of firms subject to enforcement actions by the
SEC’, Contemporary Accounting Research, Vol. 13(1), 1996, pp. 1–36.
DeFond, M. L., Hung, M., Li, S. and Li, Y., ‘Does mandatory IFRS adoption affect crash
risk?’ Accounting Review, Vol. 90(1), 2014, pp. 265–299.
Demsetz, H. and Lehn, K., ‘The structure of corporate ownership: Cause and consequences’,
Journal of Political Economy, Vol. 93(6), 1985, pp. 1155–1177.
Diamond, D. W., ‘Debt maturity structure and liquidity risk’, Quarterly Journal of
Economics, Vol. 106(3), 1991a, pp. 709–737.
Diamond, D. W., ‘Monitoring and reputation: The choice between bank loans and directly
placed debt’, Journal of Political Economy, Vol. 99(4), 1991b, pp. 689–721.
Dimson, E., ‘Risk measurement when shares are subject to infrequent trading’, Journal of
Financial Economics, Vol. 7(2), 1979, pp. 197–226.
Duchin, R., Ozbas, O. and Sensoy, B. A., ‘Costly external finance, corporate investment,
and the subprime mortgage credit crisis’, Journal of Financial Economics, Vol. 97(3),
2010, pp. 418435.
56
Fama, E., ‘What’s different about banks?’ Journal of Monetary Economics, Vol. 15(1), 1985,
pp. 29–39.
Faulkender, M. and Wang, R., ‘Corporate financial policy and the value of cash’, Journal of
Finance, Vol. 61(4), 2006, pp. 1957–1990.
Fields, L. P., Gupta, M., Wilkins, M. S. and Zhang, S., ‘Refinancing pressure and earnings
management: Evidence from changes in short-term debt and discretionary accruals’,
Working paper (Texas A&M University, 2016).
Flannery, M. J., ‘Asymmetric information and risky debt maturity choice’, Journal of
Finance, Vol. 41(1), 1986, pp. 19–37.
Giannetti, M., ‘Do better institutions mitigate agency problems? Evidence from corporate
finance choices’, Journal of Financial and Quantitative Analysis, Vol. 38(1), 2003, pp.
185–212.
Gompers, P., Ishii, J. and Metrick, A., ‘Corporate governance and equity prices’, Quarterly
Journal of Economics, Vol. 118(1), 2003, pp. 107–155.
Graham, J., Leary, M. and Roberts, M. R., ‘A century of capital structure: The leveraging of
corporate America’, Journal of Financial Economics, Vol. 118(3), 2015, pp. 658–683.
Graham, J., Li, S. and Qiu, J., ‘Corporate misreporting and bank loan contracting’, Journal
of Financial Economics, Vol. 89(1), 2008, pp. 44–61.
Guedes, J. and Opler, T., ‘The determinants of the maturity of corporate debt issues’, Journal
of Finance, Vol. 51(5), 1996, pp. 1809–1833.
Gul, F. A. and Goodwin, J., ‘Short-term debt maturity structures, credit ratings, and the
pricing of audit services’, Accounting Review, Vol. 85(3), 2010, pp. 877–909.
57
Habib, A. and Hasan, M. M., ‘Auditor-provided tax services and stock price crash risk’,
Accounting and Business Research, Vol. 46(1), 2015, pp. 51–82.
Habib, A., Jiang, H. and Hasan, M. M., ‘Stock price crash risk: Review of the empirical
literature’, Working paper (Massey University, 2016).
Hadlock, C. and Pierce, J., ‘New evidence on measuring financial constraints: Moving
beyond the KZ index’, Review of Financial Studies, Vol. 23(5), 2010, pp. 1909–1940.
Harford, J., Klasa, S. and Maxwell, W., ‘Refinancing risk and cash holdings’, Journal of
Finance, Vol. 69(3), 2014, pp. 975–1012.
Harvey, C. R. and Siddique, A., ‘Conditional skewness in asset pricing tests’, Journal of
Finance, Vol. 55(3), 2000, pp. 1263–1295.
Healy, P. M. and Palepu, K. G., ‘Information asymmetry, corporate disclosure, and the
capital markets: A review of the empirical disclosure literature’, Journal of Accounting
and Economics, Vol. 31(1), 2001, pp. 405–440.
Huang, R., Tan, K. J. K. and Faff, R. W., ‘CEO overconfidence and corporate debt maturity’,
Journal of Corporate Finance, Vol. 36, 2016, pp. 93–110.
Hutton, A. P., Marcus, A. J. and Tehranian, H., ‘Opaque financial reports, R2, and crash risk’,
Journal of Financial Economics, Vol. 94(1), 2009, pp. 67–86.
Jin, L. and Myers, C. S., ‘R2, around the world: New theory and new tests’, Journal of
Financial Economics, Vol. 79(2), 2006, pp. 257–292.
Johnson, S. A., ‘Debt maturity and the effects of growth opportunities and liquidity risk on
leverage’, Review of Financial Studies, Vol. 16(2), 2003, pp. 209–236.
58
Kaplan, S. and Zingales, L., ‘Do investment-cash flow sensitivities provide useful measures
of financing constraints?’ Quarterly Journal of Economics, Vol. 112(1), 1997, pp. 169–
215.
Kim, Y., Li, H. and Li, S., ‘Corporate social responsibility and stock price crash risk’,
Journal of Banking & Finance, Vol. 43(6), 2014, pp. 1–13.
Kim, J.B., Li, L., Lu, L.Y. and Yu, Y., ‘Financial statement comparability and expected
crash risk’, Journal of Accounting and Economics, Vol. 61(2–3), 2016, pp. 294–312.
Kim, J. B., Li, Y. and Zhang, L., ‘CFO versus CEO: Equity incentives and crashes’, Journal
of Financial Economics, Vol. 101(3), 2011a, pp. 713–730.
Kim, J. B., Li, Y. and Zhang, L., ‘Corporate tax avoidance and stock price crash risk: Firm-
level analysis’, Journal of Financial Economics, Vol. 100(3), 2011b, pp. 639–662.
Kim, J. B., Wang, Z. and Zhang, L., ‘CEO overconfidence and stock price crash
risk’, Contemporary Accounting Research, 2016, forthcoming.
Kim, J. B. and Zhang, L., ‘Accounting conservatism and stock price crash risk: Firm-level
evidence’, Contemporary Accounting Research, Vol. 33(1), 2016, pp. 412441.
Kothari, S. P., Shu, S. and Wysocki, P. D., ‘Do managers withhold bad news?’ Journal of
Accounting Research, Vol. 47(1), 2009, pp. 241–276.
Lee, E., Strong, N. and Zhu, Z., ‘Did regulation fair disclosure, SOX, and other analyst
regulations reduce security mispricing?’ Journal of Accounting Research, Vol. 52(3),
2014, pp. 733–774.
Leland, H. E. and Toft, K. B., ‘Optimal capital structure, endogenous bankruptcy, and the
term structure of credit spreads’, Journal of Finance, Vol. 51(3), 1996, pp. 987–1019.
59
Myers, S. C., ‘Determinants of corporate borrowing’, Journal of Financial Economics, Vol.
5(2), 1977, pp. 147–175.
Ozkan, A., ‘An empirical analysis of corporate debt maturity structure’, European Financial
Management, Vol. 6(2), 2000, pp. 197212.
Öztekin, O., ‘Capital structure decisions around the world: Which factors are reliably
important?’ Journal of Financial and Quantitative Analysis, Vol. 50(3), 2015, pp.
301323.
Rajan, R. and Winton, A., ‘Covenants and collateral as incentives to monitor’, Journal of
Finance, Vol. 50(4), 1995, pp. 1113–1146.
Roberts, M. R. and Sufi, A., ‘Renegotiation of financial contracts: Evidence from private
credit agreements’, Journal of Financial Economics, Vol. 93(2), 2009, pp. 159–184.
Roberts, M. R., ‘The role of dynamic renegotiation and asymmetric information in financial
contracting’, Journal of Financial Economics, Vol. 116(1), 2015, pp. 61–81.
Sheard, P., ‘The main bank system and corporate monitoring and control in Japan’, Journal
of Economic Behavior & Organization, Vol. 11(3), 1989, pp. 399–422.
Stohs, M. H. and Mauer, D. C., ‘The determinants of corporate debt maturity
structure’, Journal of Business, Vol. 69(3), 1996, pp. 279–312.
Stulz, R., ‘Does financial structure matter for economic growth? A corporate finance
perspective’, Financial structure and economic growth: A cross-country comparison of
banks, markets, and development, 2001, pp. 143–188.
Vashishtha, R., ‘The role of bank monitoring in borrowers’ discretionary disclosure:
Evidence from covenant violations’, Journal of Accounting and Economics, Vol. 57(2),
2014, pp.176–195.
60
Wang, K., Li, Y. and Erickson, J., ‘A new look at the Monday effect’, Journal of Finance,
Vol. 52(5), 1997, pp. 2171–2186.
Whited, T. and Wu, G., ‘Financial constraints risk’, Review of Financial Studies, Vol. 19(2),
2006, pp. 531–559.
Williamson, O. E., ‘Corporate finance and corporate governance’, The Journal of
Finance, Vol. 43(3), 1988, pp. 567591.
Xu, N., Li, X., Yuan, Q., and Chan, K. C. ‘Excess perks and stock price crash risk: Evidence
from China’, Journal of Corporate Finance, Vol. 25, 2014, 419434.
Zhang, F., ‘Information uncertainty and analyst forecast behavior’, Contemporary
Accounting Research, Vol. 23(2), 2006, pp. 565590.
61
Table 1. Descriptive statistics.
The table reports the descriptive statistics for the variables used in our study. Variable definitions are provided
in the Appendix. All variables are winsorized at the 1% and 99% levels.
Variable N Mean Std. dev. 5th 25th Median 75th 95th
NCSKEWt 53,052 –0.085 0.748 –1.283 –0.502 –0.104 0.302 1.199
DUVOLt 53,052 –0.055 0.352 –0.627 –0.290 –0.065 0.170 0.554
CRASHt 53,052 0.182 0.386 0.000 0.000 0.000 0.000 1.000
COUNTt 53,052 –0.037 0.632 –1.000 0.000 0.000 0.000 1.000
ST3t–1 53,052 0.541 0.356 0.008 0.217 0.515 0.938 1.000
DTURNt–1 53,052 0.028 0.792 –1.100 –0.207 0.000 0.223 1.279
SIGMAt–1 53,052 0.056 0.029 0.021 0.034 0.049 0.071 0.114
RETt–1 53,052 –0.198 0.223 –0.647 –0.247 –0.120 –0.058 –0.022
SIZEt–1 53,052 5.752 2.202 2.330 4.083 5.665 7.278 9.622
MBt–1 53,052 2.675 2.624 0.592 1.184 1.892 3.103 7.499
LEVt–1 53,052 0.239 0.171 0.006 0.098 0.222 0.351 0.558
ROAt–1 53,052 0.014 0.129 –0.225 0.001 0.039 0.075 0.143
NCSKEWt–1 53,052 –0.085 0.776 –1.258 –0.500 –0.108 0.294 1.164
ACCMt–1 53,052 0.069 0.082 0.004 0.019 0.042 0.085 0.230
INSTt–1 43,063 0.455 0.301 0.017 0.181 0.445 0.710 0.951
LTINSTt–1 43,063 0.339 0.238 0.010 0.129 0.319 0.521 0.748
GINDEXt–1 12,230 9.456 2.983 5.000 7.000 9.000 12.000 14.000
FERRt–1 34,298 0.531 1.551 0.006 0.038 0.108 0.318 2.207
DISPERt–1 33,631 0.122 0.794 –0.626 0.031 0.083 0.202 0.886
RDt–1 31,513 0.059 0.086 0.000 0.004 0.026 0.077 0.227
RATEDUMt–1 53,052 0.302 0.459 0.000 0.000 0.000 1.000 1.000
62
Table 2. The impact of short-term debt on future stock price crash risk.
This table presents the regression results for Model (4), in which we regress stock price crash risk on short-
term debt and the control variables. Variable definitions are provided in the Appendix. All variables except year
and industry dummies are winsorized at the 1% and 99% levels. T-statistics are reported in parentheses and are
based on standard errors that are corrected for heteroskedasticity and clustered at the firm level. ***, **, and *
indicate significance of the coefficients at the 1%, 5%, and 10% levels, respectively (two-sided).
Predicted (1) (2) (3) (4) (5) (6)

Sign NCSKEWt NCSKEWt NCSKEWt DUVOLt DUVOLt DUVOLt
ST3t–1 – –0.039*** –0.041*** –0.048*** –0.020*** –0.021*** –0.023***
(–3.53) (–3.74) (–4.30) (–3.89) (–4.17) (–4.57)
DTURNt–1 + 0.036*** 0.036*** 0.037*** 0.017*** 0.017*** 0.018***
(8.58) (8.61) (8.76) (8.94) (9.06) (9.12)
SIGMAt–1 + 3.428*** 3.394*** 3.144*** 1.258*** 1.192*** 1.106***
(7.53) (7.13) (6.47) (5.84) (5.31) (4.83)
RETt–1 + 0.489*** 0.506*** 0.485*** 0.193*** 0.199*** 0.192***
(8.67) (8.83) (8.38) (7.23) (7.37) (7.04)
SIZEt–1 + 0.049*** 0.046*** 0.047*** 0.023*** 0.021*** 0.021***
(23.33) (20.12) (19.93) (23.30) (19.62) (19.14)
MBt–1 + 0.016*** 0.016*** 0.015*** 0.007*** 0.007*** 0.008***
(11.38) (11.27) (10.93) (11.22) (11.24) (11.23)
LEVt–1 – –0.133*** –0.133*** –0.112*** –0.059*** –0.060*** –0.054***
(–6.26) (–6.29) (–5.04) (–5.93) (–6.06) (–5.16)
ROAt–1 +/– 0.388*** 0.380*** 0.373*** 0.194*** 0.189*** 0.183***
(12.89) (12.60) (12.16) (14.04) (13.65) (12.96)
NCSKEWt–1 + 0.021*** 0.021*** 0.019*** 0.010*** 0.010*** 0.009***
(4.42) (4.41) (3.99) (4.61) (4.66) (4.33)
ACCMt–1 + 0.198*** 0.196*** 0.179*** 0.090*** 0.088*** 0.081***
(4.26) (4.24) (3.85) (4.23) (4.15) (3.79)
Intercept –0.468*** –0.539*** –0.598*** –0.220*** –0.256*** –0.283***
(–18.19) (–17.39) (–7.88) (–18.28) (–17.83) (–7.83)
Year FE No Yes Yes No Yes Yes
Industry FE No No Yes No No Yes
N 53,052 53,052 53,052 53,052 53,052 53,052
2
Adjusted R 0.039 0.045 0.047 0.041 0.049 0.050
63
Table 3. Dealing with endogeneity.
This table presents two tests to address endogeneity concerns in the baseline regression of future stock price
crash risk on short-term debt maturity. Panel A presents the results from the firm fixed-effects (FE) regressions
(Columns (1) and (2)) and the first-differences (FD) regressions (Columns (3) and (4)). In the FE regressions,
we report the within R-squared. Panel B reports the system generalized method of moments (SYSGMM)
regression results. In the levels equations, our instruments for NCSKEWt1 and ST3t1 include their lagged
values in first differences. In the first-differenced equations, our instruments for NCSKEWt1 and ST3t1 are
the lagged values of NCSKEWt1 and ST3t1, both in levels. AR1 and AR2 p-values are the p-values of the tests
for first- and second-order autocorrelation in the residuals, under the null of no autocorrelation, respectively.
J-test is the Sargan/Hansen test for overidentification of the instruments, under the null of non-
overidentification; the number of overidentified instruments is provided in brackets. Panel C presents the first-
and second-stage results from the instrumental variable (IV) and two-stage least squares (2SLS) regressions
with the term structure of interest rates (TERMSTR) used as the instrument variable for short-term debt (ST3).
TERMSTR is the difference between the yield on 10-year government bonds and the yield on 6-month Treasury
bills. Panel D presents the results of regression analysis with three additional control variables, namely,
abnormal earnings (ABNEARN), bond rating dummy (RATEDUM), and payout dummy variable (PAYOUT).
VIY refers to a set of controls including all explanatory variables, industry effects, and year effects. Variable
definitions are provided in the Appendix. All variables except year and industry dummies are winsorized at the
1% and 99% levels. T-statistics are reported in parentheses and are based on standard errors that are corrected
for heteroskedasticity and clustered at the firm level. ***, **, and * indicate significance of the coefficients at
the 1%, 5%, and 10% levels, respectively (two-sided).
Panel A: Firm fixed-effects and first-differences regressions

(1) (2) (3) (4)
FE FE FD FD
NCSKEWt DUVOLt NCSKEWt DUVOLt
ST3t–1 –0.035** –0.018*** –0.034* –0.019**
(–2.37) (–2.69) (–1.86) (–2.27)
Controls VIY VIY VIY VIY
N 53,052 53,052 42,414 42,414
Adjusted R2 0.043 0.046 0.273 0.256
Panel B: Dynamic SYSGMM approach
(1) (2)
NCSKEWt DUVOLt
ST3t–1 –0.045* –0.023**
(–1.77) (–1.97)
Controls VIY VIY
N 53,052 53,052
AR1 p-value 0.000 0.000
AR2 p-value 0.122 0.161
J-test p-value 0.185 0.244
J-test 59.87(51) 57.60(51)
64
Panel C: Instrumental variable/two-stage least squares regressions
(1) (2)
First stage Second stage First stage Second stage
Dependent variable ST3t–1 NCSKEWt ST3t–1 DUVOLt
ST3t–1 –1.842** –1.174***
(–2.48) (–2.92)
DTURNt–1 –0.016*** 0.007 –0.016*** –0.002
(–9.82) (0.51) (–9.82) (–0.21)
SIGMAt–1 1.848*** 6.541*** 1.848*** 3.320***
(7.17) (4.31) (7.17) (4.00)
RETt–1 0.059*** 0.580*** 0.059*** 0.257***
(2.02) (6.46) (2.02) (5.17)
SIZEt–1 –0.040*** –0.022 –0.040*** –0.023
(–24.14) (–0.73) (–24.14) (–1.42)
MBt–1 0.008*** 0.029*** 0.008*** 0.016***
(8.67) (4.92) (8.67) (5.02)
LEVt–1 –0.690*** –1.353*** –0.690*** –0.849***
(–40.77) (–2.63) (–40.77) (–3.04)
ROAt–1 –0.092*** 0.212*** –0.092*** 0.079*
(–5.87) (2.58) (–5.87) (1.78)
NCSKEWt–1 0.002 0.023*** 0.002 0.012***
(1.22) (3.76) (1.22) (3.67)
ACCMt–1 0.161*** 0.466*** 0.161*** 0.265***
(7.91) (3.54) (7.91) (3.73)
Intercept 0.832*** 0.991 0.832*** 0.725**
(23.96) (1.56) (23.96) (2.11)
Instrumental variable
TERMSTRt–1 0.005*** 0.005***
(3.90) (3.90)
Diagnostic test
Hausman test (p-value) 0.002*** 0.000***
F-statistic 15.18 15.18
Industry FE Yes Yes Yes Yes
N 53,052 53,052 53,052 53,052
2
Adjusted R 0.28 0.28
65
Panel D: Regression analysis with additional controls
(1) (2) (3) (4) (5) (6) (7) (8)
NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt
ST3t–1 –0.047*** –0.023*** –0.052*** –0.025*** –0.044*** –0.022*** –0.049*** –0.024***
(–4.27) (–4.55) (–4.63) (–4.87) (–3.87) (–4.17) (–4.20) (–4.48)
ABNEARNt–1 –0.341*** –0.172*** –0.344*** –0.176***
(–15.49) (–17.14) (–15.09) (–16.89)
RATEDUMt–1 –0.024** –0.011** –0.025** –0.011**
(–2.42) (–2.27) (–2.39) (–2.21)
PAYOUTt–1 –0.531*** –0.245*** –0.457*** –0.207***
(–6.07) (–5.93) (–5.24) (–5.01)
Controls VIY VIY VIY VIY VIY VIY VIY VIY
N 53,028 53,028 53,052 53,052 49,617 49,617 49,593 49,593
Adjusted R2 0.052 0.056 0.047 0.050 0.047 0.050 0.052 0.057
66
Table 4. Other robustness checks.
This table presents the results of other robustness checks. In Panel A, we use alternative measures of short-term debt, including the ratio of debt in current liabilities to total
debt (ST1), the ratio of debt in current liabilities plus debt maturing in two years to total debt (ST2), the ratio of debt in current liabilities plus debt maturing in five years to total
debt (ST5), the ratio of debt in current liabilities plus debt maturing in three years to total assets (ST3_TA), and the ratio of debt in current liabilities minus long-term debt due
in one year to total debt (STNP1). In Panel B, crash risk is proxied by CRASH and COUNT. CRASH is an indicator variable that takes one if a firm experiences more than one
price crash week in a fiscal year. COUNT is the number of crash weeks minus the number of jump weeks over the fiscal year. In the logistic regressions of CRASH, we present
the marginal effects in square brackets; we also report the pseudo R-squared for those regressions. In Panel C, we use an alternative measure of leverage, i.e., long-term debt
divided by total assets. In Panel D, we further control for variables that capture fundamental risk, including earnings volatility, cash flow volatility, sales volatility, and beta
risk. VIY refers to a set of controls including all explanatory variables, industry effects, and year effects. Variable definitions are provided in the Appendix. All variables except
year and industry dummies are winsorized at the 1% and 99% levels. T-statistics (Z-statistics) are reported in parentheses and are based on standard errors that are corrected for
heteroskedasticity and clustered at the firm level. ***, **, and * indicate significance of the coefficients at the 1%, 5%, and 10% levels, respectively (two-sided).
Panel A: Alternative measures of short-term debt

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt
ST1t–1 –0.037*** –0.015***
(–3.06) (–2.67)
ST2t–1 –0.053*** –0.023***
(–4.64) (–4.49)
ST5t–1 –0.023* –0.011*
(–1.81) (–1.86)
ST3_TA t–1 –0.224*** –0.103***
(–5.98) (–5.97)
STNP1t–1 –0.045*** –0.019***
(–3.49) (–3.23)
Controls VIY VIY VIY VIY VIY VIY VIY VIY VIY VIY
N 53,052 53,052 53,052 53,052 51,646 51,646 53,052 53,052 52,505 52,505
Adjusted R2 0.046 0.050 0.047 0.050 0.046 0.050 0.047 0.050 0.047 0.050
67
Panel B: Alternative measures of crash risk
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
CRASHt COUNTt CRASHt COUNTt CRASHt COUNTt CRASHt COUNTt CRASHt COUNTt CRASHt COUNTt
ST1t–1 –0.033 –0.025**
[–0.005]
(–0.79) (–2.46)
ST2t–1 –0.074* –0.038***
[–0.011]
(–1.84) (–4.00)
ST3t–1 –0.071* –0.033***
[–0.010]
(–1.79) (–3.56)
ST5t–1 –0.008 –0.016
[–0.001]
(–0.17) (–1.50)
ST3_TAt–1 –0.315** –0.172***
[–0.046]
(–2.23) (–5.41)
STNP1t–1 –0.086* –0.025**
[–0.125]
(–1.87) (–2.37)
Controls VIY VIY VIY VIY VIY VIY VIY VIY VIY VIY VIY VIY
N 53,052 53,052 53,052 53,052 53,052 53,052 51,646 51,646 53,052 53,052 52,505 52,505
2
Adj./Pseudo R 0.020 0.027 0.020 0.028 0.020 0.028 0.021 0.027 0.020 0.028 0.020 0.028
68
Panel C: Alternative measure of leverage
(1) (2) (3) (4)
NCSKEWt DUVOLt CRASHt COUNTt
ST3t–1 -0.047*** -0.024*** -0.066 -0.032***
(-3.93) (-4.37) (-1.54) (-3.08)
DTURNt–1 0.037*** 0.018*** 0.093*** 0.024***
(8.78) (9.15) (6.23) (6.57)
SIGMAt–1 3.134*** 1.106*** 2.546 2.149***
(6.44) (4.83) (1.44) (5.31)
RETt–1 0.485*** 0.193*** 0.625*** 0.322***
(8.37) (7.05) (2.87) (6.56)
SIZEt–1 0.047*** 0.021*** 0.031*** 0.033***
(19.96) (19.17) (3.66) (17.11)
MBt–1 0.015*** 0.007*** 0.021*** 0.010***
(10.79) (11.12) (4.59) (8.33)
LT_LEVt–1 -0.078*** -0.042*** -0.053 -0.045**
(-2.97) (-3.43) (-0.56) (-2.01)
ROAt–1 0.381*** 0.186*** 0.718*** 0.253***
(12.41) (13.19) (6.58) (10.01)
NCSKEWt–1 0.019*** 0.009*** 0.062*** 0.010**
(4.00) (4.34) (3.94) (2.46)
ACCMt–1 0.180*** 0.080*** 0.820*** 0.137***
(3.85) (3.78) (5.45) (3.51)
Intercept -0.615*** -0.290*** -2.011*** -0.378***
(-8.13) (-8.03) (-8.33) (-6.42)
Year FE Yes Yes Yes Yes
Industry FE Yes Yes Yes Yes
N 53,052 53,052 53,052 53,052
2
Adjusted R 0.046 0.050 0.020 0.027
69
Panel D: Additional controls for fundamental risk
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt
ST3t–1 -0.049*** -0.024*** -0.050*** -0.024*** -0.048*** -0.023*** -0.043*** -0.021*** -0.045*** -0.022***
(-4.35) (-4.62) (-4.45) (-4.68) (-4.29) (-4.56) (-3.85) (-4.17) (-4.01) (-4.30)
EARNVOLt–1 -0.034*** -0.018*** -0.040*** -0.020***
(-2.98) (-3.45) (-3.24) (-3.45)
CFVOLt–1 -0.127* -0.077** -0.110 -0.064*
(-1.71) (-2.22) (-1.39) (-1.71)
SALEVOLt–1 0.041* 0.015 0.058** 0.023**
(1.74) (1.37) (2.34) (2.00)
BETAt–1 0.049*** 0.019*** 0.050*** 0.020***
(7.92) (6.81) (7.88) (6.76)
N 52,384 52,384 51,676 51,676 53,052 53,052 53,052 53,052 51,037 51,037
Adjusted R2 0.046 0.050 0.046 0.050 0.046 0.050 0.048 0.051 0.047 0.051
70
Table 5. Evidence from new debt issues.
This table presents results regarding the effect of debt maturity on stock price crash risk using data on new debt
issues. Panel A reports regression results of new debt (loan or bond) issues based on an unconsolidated sample
at the transaction level. DEBT_MAT is the natural logarithm of the maturity of a new loan or bond issue.
DEBT_SIZE is the natural logarithm of the amount of a new loan or bond issue. Panel B reports regression
results of new loan issues based on an unconsolidated (transaction-level) sample. LOAN_MAT is the natural
logarithm of the maturity of a new loan issue. LOAN_SIZE is the natural logarithm of the amount of a new loan
issue. Panel C reports regression results of new loan issues based on a consolidated sample. WAVG_MAT is
the natural logarithm of the issue-size-weighted debt maturity. SUM_SIZE is the natural logarithm of the total
amount of new loans or bond issues. VIY refers to a set of controls including all explanatory variables, industry
effects, and year effects. Variable definitions are provided in the Appendix. All variables except year and
industry dummies are winsorized at the 1% and 99% levels. T-statistics are reported in parentheses and are
Panel A: Regressions based on unconsolidated sample of new debt issues

(1) (2) (3) (4)
DEBT_MATt–1 0.016** 0.007** 0.019*** 0.008**
(2.48) (2.39) (2.88) (2.51)
DEBT_SIZEt–1 0.002 0.000
(0.70) (0.31)
N 32,785 32,785 32,785 32,785
2
Adjusted R 0.033 0.038 0.033 0.038
Panel B: Regressions based on unconsolidated sample of new loan issues
(1) (2) (3) (4)
LOAN_MATt–1 0.023*** 0.009** 0.022*** 0.009**
(2.92) (2.42) (2.72) (2.24)
LOAN_SIZEt–1 0.011* 0.005*
(1.92) (1.72)
N 24,845 24,845 24,845 24,845
2
Adjusted R 0.034 0.038 0.034 0.038
Panel C: Consolidated sample: Issue-size-weighted maturity of new debt issues
(1) (2) (3) (4)
WAVG_MATt–1 0.017** 0.008** 0.021** 0.009**
(2.09) (2.08) (2.45) (2.23)
SUM_SIZEt–1 0.003 0.001
(1.36) (0.78)
N 16,685 16,685 16,685 16,685
2
Adjusted R 0.030 0.034 0.030 0.034
71
Table 6. The impact of short-term debt on future stock price crash
risk: governance monitoring mechanisms.
This table presents the results regarding the impact of short-term debt on future stock price crash risk
conditional on the effectiveness of corporate governance mechanisms. In Panel A, we partition the sample
based on the (annual) median value of the fraction of institutional ownership (INST). In Panel B, we split the
sample using the (annual) median value of the fraction of long-term institutional ownership (LTINST). In Panel
C, we split our sample using the (annual) median value of the lagged shareholder rights index (GINDEX). VIY
refers to a set of controls including all explanatory variables, industry effects, and year effects. Variable
definitions are provided in the Appendix. All variables except year and industry dummies are winsorized at the
1% and 99% levels. T-statistics are reported in parentheses and are based on standard errors that are corrected
for heteroskedasticity and clustered at the firm level. ***, **, and * indicate significance of the coefficients at
the 1%, 5%, and 10% levels, respectively (two-sided).
Panel A: Institutional ownership

(1) (2) (3) (4)
Dependent variable NCSKEWt NCSKEWt DUVOLt DUVOLt
Partition Weak governance Strong governance Weak governance Strong governance
(INST ≤ median) (INST > median) (INST ≤ median) (INST > median)
ST3t–1 –0.033* –0.024 –0.019** –0.012
(–1.91) (–1.42) (–2.42) (–1.56)
N 21,538 21,525 21,538 21,525
Adjusted R2 0.044 0.024 0.048 0.026
Panel B: Long-term (non-transient) institutional ownership
(1) (2) (3) (4)
(LTINST ≤ median) (LTINST > median) (LTINST ≤ median) (LTINST > median)
ST3t–1 –0.041** –0.027 –0.022*** –0.014*
(–2.33) (–1.59) (–2.76) (–1.75)
N 21,538 21,525 21,538 21,525
Adjusted R2 0.044 0.025 0.049 0.028
Panel C: Shareholder rights
(1) (2) (3) (4)
(GINDEX ≥ median) (GINDEX < median) (GINDEX ≥ median) (GINDEX < median)
ST3t–1 –0.085*** –0.041 –0.047*** –0.024
(–2.63) (–1.30) (–3.20) (–1.63)
N 7,045 5,185 7,045 5,185
2
Adjusted R 0.019 0.029 0.022 0.030
72
Table 7. The impact of short-term debt on future stock price crash risk: the role
of information asymmetry.
conditional on the degrees of information asymmetry (IA). In Panel A, we partition our sample based on the
(annual) median value of analyst forecast error (FERR). In Panel B, we split our sample based on the (annual)
median value of the dispersion of analysts’ forecasts (DISPER). In Panel C, we partition our sample based on
the (annual) median value of the R&D ratio. VIY refers to a set of controls including all explanatory variables,
industry effects, and year effects. Variable definitions are provided in the Appendix. All variables except year
and industry dummies are winsorized at the 1% and 99% levels. T-statistics are reported in parentheses and are
Panel A: Analyst forecast error

(1) (2) (3) (4)
Dependent variable
NCSKEWt NCSKEWt DUVOLt DUVOLt
Partition High IA Low IA High IA Low IA
(FERR ≥ median) (FERR < median) (FERR ≥ median) (FERR < median)
ST3t–1 –0.077*** –0.021 –0.037*** –0.016*
(–4.26) (–1.11) (–4.30) (–1.77)
N 17,154 17,144 17,154 17,144
2
Adjusted R 0.037 0.023 0.042 0.024
Panel B: Dispersion of analyst forecasts
(1) (2) (3) (4)
Dependent variable
(DISPER ≥ median) (DISPER < median) (DISPER ≥ median) (DISPER < median)
ST3t–1 –0.084*** –0.023 –0.040*** –0.015*
(–4.60) (–1.19) (–4.71) (–1.72)
N 16,822 16,809 16,822 16,809
Adjusted R2 0.030 0.031 0.035 0.034
Panel C: R&D intensity
(1) (2) (3) (4)
Dependent variable
(RD ≥ median) (RD < median) (RD ≥ median) (RD < median)
ST3t–1 –0.048** –0.033 –0.025*** –0.012
(–2.34) (–1.61) (–2.67) (–1.33)
N 15,763 15,750 15,763 15,750
Adjusted R2 0.056 0.044 0.061 0.046
73
Table 8. The impact of short-term debt on future stock price crash risk: firm
risk mechanism.
conditional on the degree of firm risk. In Panel A, we partition our sample based on the (annual) median value
of leverage (LEV). In Panel B, we analyze unrated and rated firms separately. VIY refers to a set of controls
including all explanatory variables, industry effects, and year effects. Variable definitions are provided in the
Appendix. All variables except year and industry dummies are winsorized at the 1% and 99% levels. T-statistics
are reported in parentheses and are based on standard errors that are corrected for heteroskedasticity and
clustered at the firm level. ***, **, and * indicate significance of the coefficients at the 1%, 5%, and 10%
levels, respectively (two-sided).
Panel A: Leverage
(1) (2) (3) (4)
Dependent variable
Partition High risk Low risk High risk Low risk
(LEV ≥ median) (LEV < median) (LEV ≥ median) (LEV < median)
ST3t–1 –0.093*** –0.010 –0.040*** –0.011
(–5.76) (–0.65) (–5.33) (–1.50)
N 26,534 26,518 26,534 26,518
2
Adjusted R 0.045 0.050 0.049 0.053
Panel B: Bond rating
(1) (2) (3) (4)
Dependent variable
Partition High risk Low risk High risk Low risk
(Unrated) (Rated) (Unrated) (Rated)
ST3t–1 –0.029** –0.040* –0.016*** –0.020*
(–2.24) (–1.66) (–2.63) (–1.82)
N 37,013 16,039 37,013 16,039
2
Adjusted R 0.053 0.021 0.055 0.026
74
Appendix 1: Variable definitions
Crash risk variables

NCSKEW is the negative skewness of firm-specific weekly returns over the fiscal year.
DUVOL is the log of the ratio of the standard deviations of down-week to up-week firm-
specific weekly returns.
For both crash risk variables, the firm-specific weekly return (W) is equal to ln(1+residual),
where the residual is obtained from the following expanded market model:
𝑟𝑗,𝜏 = 𝛼𝑗 + 𝛽1,𝑗 𝑟𝑚,𝜏−1 + 𝛽2,𝑗 𝑟𝑖,𝜏−1 + 𝛽3,𝑗 𝑟𝑚,𝜏 + 𝛽4,𝑗 𝑟𝑖,𝜏 + 𝛽5,𝑗 𝑟𝑚,𝜏+1 + 𝛽6,𝑗 𝑟𝑖,𝜏+1 + 𝜀𝑗,𝜏 ,
index, and ri,τ is the Fama and French value-weighted industry index in week τ.
Debt maturity variables

ST3 is the ratio of debt in current liabilities (dlc) plus debt maturing in two or three years
(dd2+dd3) to total debt (the sum of debt in current liabilities plus long-term debt, i.e.,
dlc+dltt).
ST1 is the ratio of debt in current liabilities (dlc) to total debt (dlc+dltt).
ST2 is the ratio of debt in current liabilities (dlc) plus debt maturing in two years (dd2) to
total debt (dlc+dltt).
ST5 is the ratio of debt in current liabilities (dlc) plus debt maturing in two to five years
(dd2+dd3+dd4+dd5) to total debt (dlc+dltt).
ST3_TA is the ratio of debt in current liabilities (dlc) plus debt maturing in two or three years
(dd2+dd3) to total assets (at).
75
STNP1 is the ratio of debt in current liabilities (dlc) minus long-term debt due in one year
(dd1) to total debt (dlc+dltt).
DEBT_MAT is the natural logarithm of new private loan or public bond maturity, in days.
Data source: Mergent Fixed Income Securities Database (FISD) and Dealscan.
LOAN_MAT is the natural logarithm of new private loan maturity, in days. Data source:
Dealscan.
WAVG_MAT is the natural logarithm of the issue-size-weighted debt maturity.
AVG_MAT is the natural logarithm of the equal-weighted debt maturity.
Control variables
DTURN is the average monthly share turnover over the current fiscal year minus the average
monthly share turnover over the previous fiscal year, where monthly share turnover is
calculated as the monthly trading volume divided by the total number of shares
outstanding during the month.
SIGMA is the standard deviation of firm-specific weekly returns over the fiscal year.
RET is the mean of firm-specific weekly returns over the fiscal year, times 100.
MB is the market value of equity (csho*prcc_f) divided by the book value of equity (market-
to-book).
SIZE is the natural logarithm of market value of equity.
LEV is total debt (dltt+dlc) divided by total assets (at).
ROA is income before extraordinary items (ib) divided by total assets (at).
ACCM is the absolute value of discretionary accruals estimated from the modified Jones
model.
76
TERMSTR is the difference between the yield on 10-year government bonds and the yield
on 6-month Treasury bills. Data source: Federal Reserve Bank of St. Louis.
ABNEARN is income before extraordinary items, minus common or ordinary stock (capital)
equivalents (ibadj) in year t+1 and t, divided by the market value of equity in year t.
PAYOUT is the sum of common and preferred dividend and purchase (dvp+dvc+prstkc)
divided by total assets (at).
DEBT_SIZE is the natural logarithm of the total amount of new private loans or the par value
of new public bonds. Data sources: FISD and Dealscan.
LOAN_SIZE is the natural logarithm of the total amount of new private loans. Data sources:
Dealscan.
SUM_SIZE is the natural logarithm of the total amount of new loans or bond issues in year
t.
EARNVOL is the standard deviation of the ratio of earnings, excluding extraordinary items
and discontinued operations, to lagged total equity during the past five years.
CFVOL is the standard deviation of the ratio of cash flow to total assets during the past five
years.
SALEVOL is the standard deviation of the ratio of sales to total assets during the past five
years.
BETA is the beta risk from the CAPM, namely the covariance between the individual firm
return and the market return divided by the variance of the market over a fiscal year.
Conditioning variables
INST is the percentage of shares held by institutional owners, obtained from the Thomson
13F database.
77
LTINST is the percentage of shares held by dedicated and quasi-indexer institutional
investors.
GINDEX is the number of anti-takeover provisions based on Gompers et al. (2003). Anti-
takeover provisions are obtained from the RiskMetrics’ governance database.
FERR is the absolute value of the difference between actual earnings per share and consensus
analyst forecast divided by consensus analyst forecast in year t.
DISPER is the standard deviation of analyst forecasts divided by consensus analyst forecast
in year t.
RD is the ratio of research and development expenditures (xrd) to total assets (at).
RATEDUM is an indicator variable that takes one if a firm has an S&P rating on long-term
debt (splticrm) and zero otherwise.
78
Appendix 2: Untabulated results.
Table A1. Untabulated robustness tests

This table presents the untabulated results of robustness tests in the paper. Panel A presents the regression results of analysis using two- or three-year-ahead crash risk measures.
In Panel B, we measure short-term debt using the ratio of short-maturity loans (i.e., loans that have a time-to-maturity of less than three years) to total loans. VIY refers to a
set of controls including all explanatory variables, industry effects, and year effects. Variable definitions are provided in the Appendix. All variables except year and industry
dummies are winsorized at the 1% and 99% levels. T-statistics are reported in parentheses and are based on standard errors that are corrected for heteroskedasticity and clustered
at the firm level. ***, **, and * indicate significance of the coefficients at the 1%, 5%, and 10% levels, respectively (two-sided).
Panel A: Analysis using two- or three-year-ahead crash risk

(1) (2) (3) (4) (5) (6) (7) (8)
NCSKEWt NCSKEWt+1 DUVOLt DUVOLt+1 COUNTt COUNTt+1 CRASHt CRASHt+1
ST3t–1 –0.042*** –0.023* –0.021*** –0.009 –0.024** –0.022* –0.016** –0.017**
(–3.40) (–1.71) (–3.75) (–1.45) (–2.25) (–1.91) (–2.35) (–2.37)
Controls VIY VIY VIY VIY VIY VIY VIY VIY
N 42,237 35,384 42,237 35,384 42,237 35,384 42,237 35,384
Adjusted R2 0.035 0.030 0.038 0.033 0.021 0.017 0.016 0.015
Panel B: Alternative measure of short-term debt: short-maturity loans
(1) (2)
NCSKEWt DUVOLt
STMLt–1 –0.024** –0.014**
(–1.99) (–2.40)
Controls VIY VIY
N 31,522 31,522
Adjusted R2 0.035 0.039
79
Table A2. Untabulated interaction tests
This table presents the untabulated results of interaction tests in the paper. Panel A presents the regression
results of interaction tests based on Table 6 corporate governance mechanisms. INSDUM is an indicator
variable that takes the value of one if institutional ownership (INST) is lower than the sample median, and zero
otherwise. LTINSDUM is an indicator variable that takes the value of one if long-term institutional ownership
(LTINST) is lower than the sample median, and zero otherwise. GIDUM is an indicator variable that takes the
value of one if governance index (GINDEX) is greater than the sample median, and zero otherwise. Panel B
presents the regression results of interaction tests based on Table 7 information asymmetry mechanisms.
FERRDUM is an indicator variable that takes the value of one if analyst forecast error (FERR) is greater than
the sample median, and zero otherwise. DISPDUM is an indicator variable that takes the value of one if analyst
dispersion (DISPER) is greater than the sample median, and zero otherwise. RDDUM is an indicator variable
that takes the value of one if R&D ratio (RD) is greater than the sample median, and zero otherwise. Panel C
presents the regression results of interaction tests based on Table 8 firm risk mechanisms. LEVDUM is an
indicator variable that takes the value of one if leverage ratio (LEV) is greater than the sample median, and zero
otherwise. VIY refers to a set of controls including all explanatory variables, industry effects, and year effects.
Variable definitions are provided in the Appendix. All variables except year and industry dummies are
winsorized at the 1% and 99% levels. T-statistics are reported in parentheses and are based on standard errors
that are corrected for heteroskedasticity and clustered at the firm level. ***, **, and * indicate significance of
the coefficients at the 1%, 5%, and 10% levels, respectively (two-sided).
Panel A: Interaction tests based on Table 6 corporate governance mechanisms.

(1) (2) (3) (4) (5) (6)
NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt
ST3t–1 0.006 0.002 –0.006 –0.004 –0.030 –0.017
(0.38) (0.23) (–0.36) (–0.53) (–0.98) (–1.15)
INSDUMt–1 –0.077*** –0.033***
(–5.60) (–5.07)
ST3t–1 * INSDUMt–1 –0.070*** –0.035***
(–3.26) (–3.48)
LTINSDUMt–1 –0.061*** –0.027***
(–4.33) (–3.98)
ST3t–1 * LTINSDUMt–1 –0.058*** –0.029***
(–2.65) (–2.82)
GIDUMt–1 0.046** 0.023**
(2.03) (2.15)
ST3t–1 * GIDUMt-1 –0.059 –0.034*
(–1.42) (–1.78)
Controls VIY VIY VIY VIY VIY VIY
N 43,063 43,063 43,063 43,063 12,230 12,230
Adjusted R2 0.048 0.052 0.047 0.050 0.023 0.025
80
Panel B: Interaction tests based on Table 7 information asymmetry mechanisms.
(1) (2) (3) (4) (5) (6)
NCSKEWt DUVOLt NCSKEWt DUVOLt NCSKEWt DUVOLt
ST3t–1 –0.014 –0.012 –0.024 –0.015* –0.041** –0.017*
(–0.76) (–1.43) (–1.35) (–1.80) (–2.14) (–1.86)
FERRDUMt–1 –0.004 –0.002
(–0.28) (–0.36)
ST3t–1 * FERRDUMt–1 –0.076*** –0.030***
(–3.29) (–2.83)
DISPDUMt–1 0.017 0.010
(1.19) (1.53)
ST3t–1 * DISPDUMt–1 –0.060*** –0.026**
(–2.61) (–2.44)
RDDUMt–1 0.003 0.002
(0.13) (0.23)
ST3t–1 * RDDUMt–1 0.007 –0.002
(0.29) (–0.17)
N 34,298 34,298 33,631 33,631 31,513 31,513
Adjusted R2 0.032 0.036 0.030 0.034 0.049 0.053
Panel C: Interaction tests based on Table 8 firm risk mechanisms.
(1) (2) (3) (4)
ST3t–1 –0.013 –0.010 –0.038*** –0.019***
(–0.88) (–1.53) (–2.92) (–3.21)
LEVDUMt–1 0.062*** 0.027***
(4.18) (3.80)
ST3t–1 * LEVDUMt–1 –0.075*** –0.028***
(–3.83) (–3.09)
RATEDUMt–1 0.006 0.003
(0.48) (0.51)
ST3t–1 *RATEDUMt–1 –0.055** –0.023*
(–2.09) (–1.86)
N 53,052 53,052 53,052 53,052
Adjusted R2 0.047 0.050 0.047 0.050
81
Chapter 3
Supplier Financing and Stock Price Crash Risk:
Monitoring versus Concession?
ABSTRACT
This paper empirically tests two opposing views on the relation between supplier financing
and future stock price crash risk: monitoring versus concession. We present robust evidence
that trade credit is negatively associated with firm-specific stock price crash risk, consistent
with the view that trade credit, as an important source of short-term financing, effectively
monitors buying firms and therefore constrains their bad-news-hoarding behavior. Further
analyses reveal that the role of trade credit in mitigating stock price crash risk is more
pronounced among firms that demand a higher level of monitoring such as those with weaker
governance, less bank monitoring, lower market power, and higher distress risk. Overall, our
results shed light on how trade credit shapes managerial disclosure incentives.
82
3.1. Introduction
This study examines the influence of trade credit financing on firm-specific stock
price crash risk. Suppliers usually offer trade credit as a strategy to increase sales and reduce
stock on hand. A buyer is obliged to pay its supplier in the future and records this liability
as trade payable. Trade credit has become an increasingly important financing channel in
real business transactions. At the end of 2011, the U.S. corporate sector reported
considerable aggregate accounts payable of $5.3 trillion (Desai, Foley, and Hines Jr., 2016).
Trade payables represented the second largest liability on the aggregate balance sheet of
nonfinancial businesses in the U.S. (U.S. Flow of Funds, 2016). Using a large sample of
firms across 34 countries, Levine, Lin, and Xie (2016) show that, for an average firm, trade
credit accounts for 25% of total debt liabilities.
In addition to the benefits of boosting sales for suppliers, trade credit also brings
about several benefits to buyers, such as reducing buyers’ holding cost of finance (Ferris,
1981; Emery, 1987) and allowing them to ascertain the quality of products before making
payment (Smith, 1987; Lee and Stowe, 1993; Long, Maltiz, and Ravid, 1993). A growing
body of literature has also documented evidence that trade credit can serve as a substitute
for bank lending, especially during periods of financial crisis (e.g., Love, Preve, and Sarria-
Allende, 2007; Fabbri and Menichini, 2010; Garcia-Appendini and Montoriol-Garriga,
2013; Murfin and Njoroge, 2015). However, few studies have paid attention to whether and
how trade credit financing affects firms’ disclosure choices and subsequent shareholders’
wealth. This study attempts to fill the literature void.
Our research is also motivated by the growing academic efforts to forecast extreme
outcomes in the equity market, namely, stock price crash risk. Due to various incentives,
such as compensation contracts, career concerns or empire building, managers tend to

83
conceal unfavorable information (e.g., Jin and Myers, 2006; Hutton, Marcus, and Tehranian,
2009). When the bad news is accumulated for an extended period, the firm’s share price will
be overvalued. However, once the negative information is suddenly released to the market,
share price will experience an extreme downward correction, leading to a stock price crash.
Overall, the primary cause of stock price crash risk is the lack of corporate transparency and
monitoring that enables managers to withhold bad news.
Due to the important implications of stock price crash risk for portfolio investment
and risk management, a large body of literature has examined a wide range of factors that
may affect managerial bad-news-withholding behavior and future firm-specific crash risk,
such as executive equity incentives, tax avoidance techniques, accounting conservatism,
religion, stock liquidity, CEO age, and governance mechanisms (Kim, Li, and Zhang2011a,
2011b; Kim and Zhang, 2015; Callen and Fang, 2015a; Chang, Chen, and Zolotoy, 2016;
Andreou, Antoniou, Horton, and Louca, 2016; Andreou, Lucas, and Petrou, 2016). In a
recent study, Dang, Lee, Liu, and Zeng (2017) suggest that short-maturity debt, as a formal
financing source, is effective in reducing borrowers’ stock price crash risk. However, thus
far little is known as to whether and how informal financing channels such as trade credit
may affect managerial-bad-news-hoarding activities and future stock price crash risk. As
such, we seek to fill this research gap.
The effect of firms’ use of trade credit on managers’ information revelation and stock
price crash risk is an open question. We examine two competing views: monitoring versus
concession. On the one hand, while suppliers can benefit from providing trade credit, they
have to bear default risk on credit payment from customers. Jacobson and Schedvin (2015)
suggest that trade creditors face increased insolvency and subsequent bankruptcy risks when
their customers default. In order to minimize the default risk, trade creditors tend to collect
information about buyers’ financial status and credit-worthiness (Mian and Smith, 1992;
84
Jain, 2001) or take various measures such as setting certain contract terms (Klapper, Laeven,
and Rajan, 2012). Mian and Smith (1992) argue that trade creditors have a monitoring
advantage over banks because they can regularly make sales contracts and thereby access
buyers’ information as a by-product of selling. Chod, Trichakis, and Tsoukalas (2016)
suggest that suppliers may have access to private information about their customers, which
can facilitate monitoring and reduce information asymmetries between suppliers and buyers.
Such an informational advantage is largely gained from the fact that suppliers and buyers
are often engaged in the same or closely related industry. Consequently, trade credit
suppliers can identify the private financial information about a customer by comparing to
other clients (Ng, Smith, and Smith, 1999). Trade creditors are also able to enforce debt
repayment by threatening to cut off future intermediate goods supply to their customers
(Cuñat, 2007). Given the aforementioned informational advantage and monitoring function
of suppliers, trade credit financing is expected to restrict managerial opportunistic and bad-
news-hoarding behavior, thus reducing stock price crash risk.
On the other hand, trade credit might create opportunities for managers to hide bad
news and mislead investors. Wilner (2000) suggests that suppliers grant more concessions
to customers in debt renegotiations than banks in order to maintain an enduring product
market relationship. Many studies indeed find that suppliers can enjoy the benefits of a long-
term supplier-buyer relationship, including reduced selling costs and higher profitability
(Kalwani and Narayandas, 1995; Petersen and Rajan, 1997; Raman and Shahrur, 2008).
Moreover, this supplier-client relationship is of critical value to suppliers because of the
inter-firm linkages in valuation (Hertzel, Li, Officer, and Rodgers, 2008). The value of
suppliers will be negatively affected by the distress or bankruptcy of a customer. In a similar
vein, Jorion and Zhang (2009) find that borrowers’ financial distress leads trade creditors to
experience negative abnormal equity returns and increases in their credit spreads. Such
85
wealth and distress effects are described as “credit contagion”. As a result, suppliers are
incentivized to extend trade credit to irreplaceable buyers who are facing financial distress
(Wilner, 2000; Cuñat, 2007). Indeed, García-Appendini and Montoriol-Garriga (2015) show
that the majority of trade credit suppliers choose to maintain the relationship with their
customers by providing financial support even though those customers may eventually
default on their debt obligations. Under such circumstances, the monitoring role of suppliers
can be weakened. Managers in buyer firms are likely to take advantage of suppliers’
willingness to make concessions and withhold negative information, thereby leading to an
increase in future stock price crash risk.
Our study seeks to investigate the above competing views as to whether trade credit
mitigates or exacerbates stock price crash risk. We follow previous research (e.g., Petersen
and Rajan, 1997; Fisman and Love, 2003; Cuñat, 2007) and measure trade credit as accounts
payable divided by total assets. Following Chen, Hong, and Stein (2001), Hutton, Marcus,
and Tehranian (2009) and Kim, Li, and Zhang (2011a, 2011b), we use two measures of stock
price crash risk, namely, the negative conditional skewness of firm-specific weekly returns
and the asymmetric volatility of negative and positive stock returns. Using a large sample of
U.S. public firms from 1990 to 2013, we provide evidence that trade credit is negatively
associated with future stock price crash risk, consistent with suppliers exerting effective
monitoring on buyers and preventing them from hoarding bad news. The effect is both
statistically significant and economically sizable. The decrease in future stock price crash
risk accounts for 10.40% to 17.85% of the sample mean when trade credit increases from
the 25th to the 75th percentiles of the distribution. Our results are robust to dealing with
potential endogeneity concerns and using alternative proxies for trade credit and crash risk.
We further examine cross-sectional variations in the negative association between
trade credit and stock price crash risk. Our main findings suggest that trade credit reduces
86
stock price crash risk via enhanced monitoring by suppliers. According to the monitoring
mechanism, we thus expect the above relation between trade credit and crash risk to be more
pronounced for firms that demand a higher level of monitoring. To test this conjecture and
provide more evidence on the underlying channels, we examine the effect of trade creditor
monitoring on stock price crash risk conditional on other governance mechanisms, bank
monitoring, market power, and financial distress. First, if the negative impact of trade credit
on stock price crash risk is indeed driven by suppliers’ monitoring effect, we expect suppliers
to play a more salient role when firms are exposed to insufficient governance mechanisms.
Consistent with this prediction, we find that the negative relation between trade credit and
crash risk is more pronounced for firms with lower analyst coverage, less total institutional
ownership, or more transient institutional ownership.
In addition, many studies argue that trade credit can act as a substitute for bank credit
(e.g., Biais and Gollier, 1997; Petersen and Rajan, 1997; Burkart and Ellingsen, 2004;
Giannetti, Burkart, and Ellingsen, 2011), as trade creditors have a greater informational and
liquidation advantage over banks (Fabbri and Menichini, 2010). Motivated by this stream of
research, we examine whether suppliers exert a stronger monitoring impact on customers
when bank monitoring is weak. Our results show that the adverse effect of trade credit on
crash risk is more pronounced for firms facing weaker bank monitoring, i.e., those firms
having a smaller volume of bank loans, especially loans of shorter maturities.
The supplier-buyer relationship can be influenced by the market power of buyers.
Powerful customers tend to demand more favorable prices, irregular purchases, and more
trade credit extension from their suppliers (Klapper, Laeven, and Rajan, 2012; Kelly, Lustig,
and Van Nieuwerburgh, 2013; Murfin and Njoroge, 2015). In contrast, trade creditors do not
need to cater to the demand of buyers with relatively weaker market power and therefore are
more able to restrict customers’ misbehavior. Consistent with this prediction, our evidence
87
shows that the negative impact of trade credit on crash risk is more pronounced among buyer
firms with weaker market power, that is, those firms with a lower sales-to-industry sales
ratio, a higher amount of standardized inputs, and a higher level of product market
competition.
Last but not least, financially distressed firms require more monitoring because they
are more likely to manipulate their disclosure due to contracting motives (Summers and
Sweeney, 1998; Healy and Palepu, 2001; Graham, Harvey, and Rajgopal, 2005). Thus, the
negative association between trade credit and stock price crash risk is expected to be stronger
among financially distressed firms than for financially healthy ones. We use Altman’s (1968)
z-score and assets-to-debt ratio to proxy for financial distress and obtain results consistent
with the prediction above. Specifically, the mitigating effect of trade credit on crash risk is
more pronounced for distressed firms (i.e., those with a lower z-score and assets-to-debt
ratio).
Our study contributes to the literature in several ways. First, it contributes to research
on the equity market consequences of trade credit. While a large body of literature
investigates various motives of using trade credit, only a very few studies have directly
examined the relationship between trade credit financing and firm value. Goto and Xiao
(2015) investigate how firms’ use of trade credit affects their stock returns while
Albuquerque, Ramadorai, and Watugala (2015) study the role of trade credit relationships
in generating cross-border stock return predictability. Aktas, Croci, and Petmezas (2015)
examine the value effect of working capital management and show the existence of an
optimal level of working capital investment. To our best knowledge, our study is the first to
provide evidence on the impact of trade credit financing on extreme negative stock returns.
88
Second, this paper extends the current understanding of supplier-buyer relationships.
Our findings contribute to the debate concerning the monitoring versus concession effects
of suppliers. Specifically, we provide new evidence that suppliers can effectively monitor
customers and restrict their bad-news-hoarding activities. This finding is also consistent with
trade credit enhancing buyers’ information transparency, which is documented in the prior
literature (e.g., Emery, 1984; Petersen and Rajan, 1997; Cuñat, 2007; Aktas, Bodt, Lobez,
and Statnik, 2012).
Finally, our study contributes to the growing literature on stock price crash risk by
documenting the role of a new factor, trade credit, in mitigating crash risk. Stock price crash
risk is of great importance to both academics and practitioners. As mentioned earlier, while
prior studies have shown various determinants of future stock price crash risk (e.g., Hutton,
Marcus, and Tehranian, 2009; Kim, Li, and Zhang, 2011a, 2011b; Callen and Fang, 2013,
2015a, 2015b; Chang, Chen, and Zolotoy, 2016; Andreou, Lucas, and Petrou, 2016), we
provide new evidence that an informal financing channel such as trade credit can also reduce
stock price crash risk, consistent with trade creditors effectively monitoring buyers and
protecting buyer firms’ shareholder wealth.
The rest of the paper is organized as follows. Section 3.2 reviews the literature on
trade credit and stock price crash risk, and develops our hypotheses. We discuss the sample
and research design in Section 3.3 and present the main empirical results in Section 3.4.
Section 3.5 provides additional analyses. Section 3.6 concludes.
89
3.2. Related literature and hypothesis development
3.2.1 Literature on trade credit
The literature on trade credit has sought to understand why firms borrow from suppliers
instead of borrowing from conventional financial institutions. Three motives have been
suggested to justify the existence of trade credit: transaction, price discrimination, and
financing motives (e.g., Brick and Fung, 1984; Fabbri and Menichini, 2010). The transaction
motive deems trade credit as a way to reduce transaction costs for both sellers and buyers,
i.e., the costs associated with holding cash and goods (Ferris, 1981; Emery, 1987; Mian and
Smith, 1992). Sellers can also make use of trade credit provision to apply price
discrimination (Brennan, Maksimovic, and Zechner, 1988), and to offer implicit quality
guarantees (Smith, 1987; Lee and Stowe, 1993; Long, Malitz, and Ravid, 1993). Overall, the
nonfinancial motives explain how trade credit plays a role in business trade but do not predict
the advantage of supplier lending over financial intermediaries.
The financial motive argues, on the other hand, that firms with substandard credit
worthiness tend to finance from their suppliers in the form of trade credit (Schwartz, 1974;
Petersen and Rajan, 1997). The literature suggests that trade creditors have both a liquidation
and informational advantage over financial intermediates. To the extent that suppliers can
better repossess and resell goods to other buyers than financial institutions, suppliers that
encounter customers’ default are able to extract a greater liquidation value from the inputs
collateralized (Mian and Smith, 1992; Petersen and Rajan, 1997; Frank and Maksimovic,
2004; Cuñat, 2007; Fabbri and Menichini, 2010). In addition, suppliers have a comparative
advantage over banks in getting information about buyers. Specifically, suppliers are able to
obtain their customers’ information as a by-product of selling activities, whereas banks or
other lenders can only obtain such information at a cost (e.g., Schwartz and Whitcomb, 1979;
Ferris, 1981; Emery, 1984; Mian and Smith, 1994; Biais and Gollier, 1997; Goto, Xiao, and
90
Xu, 2015). This argument suggests that trade credit can alleviate buyers’ asymmetric
information problem to a larger extent than banks. In this study, the rationale behind the
relationship between trade credit and stock price crash risk is based on the informational
advantage of trade credit.
3.2.2 Literature review on stock price crash risk
This study is also related to the literature on the determinants of stock price crash
risk. Chen, Hong, and Stein (2001) find that average monthly turnover and past returns can
forecast future stock price crashes. Jin and Myers (2006) introduce a theoretical model in
which stock price crashes occur when managers’ accumulated bad news is revealed to the
public at once. They further examine international stock markets and find that opaque stocks
are more likely to crash. Hutton, Marcus, and Tehranian (2009) complement Jin and Myers’
(2006) findings by showing that financial reporting opacity leads to more managerial bad
news hoarding activities and therefore a higher likelihood of stock price crashes.
Following those papers, a considerable body of literature has identified various
determinants of future stock price crash risk, including factors related to both internal and
external monitoring mechanisms. Research on internal mechanisms suggests that future
stock price crash risk is positively associated with managerial equity incentives (Kim, Li,
and Zhang, 2011a; Xu, Li, Yuang, and Chan, 2014), the utilization of corporate tax
avoidance tools (Kim, Li, and Zhang, 2011b), unreligious managers (Callen and Fang,
2015a), overconfident or younger CEOs (Kim, Wang, and Zhang, 2016; Andreou, Louca,
and Petrou, 2016), the degree of earnings smoothing (Chen, Kim, and Yao 2017; Khurana,
Pereira, and Zhang, 2017), and the ownership-control wedge (Hong, Kim, and Welker,
2017). In parallel, the literature on external mechanisms documents a number of
determinants that mitigate firm-level stock price crash risk, such as the presence of
(dedicated) institutional investors (An and Zhang, 2013; Callen and Fang, 2013), the
91
mandatory adoption of International Financial Reporting Standards (IFRS) (DeFond, Hung,
Li, and Li , 2014), and auditor tenure (Callen and Fang, 2017).
Although the existing literature has examined a number of factors that are associated
with stock price crash risk, few studies have investigated this issue from the perspective of
corporate financing policy. Dang, Lee, Liu, and Zeng (2017) find that firms with more short-
term debt are subject to more efficient monitoring from debt holders, thereby exhibiting less
bad-news-hoarding behavior and lower stock price crash risk. Boehme, Fotak, and May
(2017) argue that managers issuing seasoned equity are prone to withhold bad news for an
extended period, leading to higher future stock price crash risk. We contribute to this
literature by examining an alternative financing source from vendors, i.e., trade credit.
3.2.3. Hypothesis development
Based on those two strands of literature, we formulate our competing hypotheses as
follows. On the one hand, trade credit offered by suppliers enables them to monitor their
buyers in several ways. First, suppliers can gain information about buyers’ financial status
during normal business interactions (Biais and Gollier, 1997; Petersen and Rajan, 1997).
Both suppliers and buyers are usually engaged in similar businesses so that suppliers could
have an easy access to buyers’ information and a good understanding of such information.
Second, suppliers may impose repayment pressures on clients by using the threat of cutting
off future inputs supply (Petersen and Rajan, 1997; Cuñat, 2007). This can incentivize buyers
to improve their corporate governance and information transparency to maintain a long-term
relationship with their suppliers. Third, suppliers can access a buyer’s information through
the network of other similar clients (Ng, Smith, and Smith, 1999). Suppliers often conduct
business with a network of buyers and therefore can identify the financial status and
performance of a specific buyer by comparing its products demand with that of peers.
Specifically, although a buyer may not actively disclose unfavorable news about operating
92
performance to the public, its suppliers could still be aware of such negative information
through the abnormal decline in the frequency and size of orders. Given the above
arguments, we expect that vendors are able to monitor buyers and restrict their managers
from withholding negative information. Thus, we formally state our first hypothesis as
follows:
H1a: Firms’ use of trade credit is negatively related to stock price crash risk.
In contrast, the concession view argues that suppliers tend to be lenient in order to
sustain a long-term relationship with buyer firms. This is because suppliers can benefit from
lower selling costs and better operating performance from a stable relationship with their
customers (Kalwani and Narayandas, 1995; Petersen and Rajan, 1997; Raman and Shahrur,
2008). Meanwhile, such a relationship implies that the wealth of suppliers and customers is
mutually linked. Once buyers encounter an unfavorable situation, such as financial distress
or bankruptcy, suppliers will also suffer a decreased firm valuation (Jorion and Zhang, 2009;
Hertzel and Officer, 2012), especially when buyers are highly concentrated (Campello and
Gao, 2017). These contagion effects have been documented in Cohen and Frazzini (2008)
and Kolay, Lemmon, and Tashjian (2016), who show that a firm suffers valuation downward
when its major customers reveal negative news. Thus, even when buyer firms are
withholding negative information about their financial status, suppliers may not necessarily
exert monitoring and enforce trade credit payment due to potential contagion effects. Instead,
suppliers may grant concessions to buyers by way of extending more trade credit. Such
concessions enable the managers of buyer firms to continue accumulating bad news, leading
to increased stock price crash risk. We thus formulate the following competing hypothesis:
H1b: Firms’ use of trade credit is positively related to stock price crash risk.
93
3.3. Empirical methodology
3.3.1. The sample
We construct a panel data set of U.S. public firms from 1990 to 2014.17 Our crash risk
measures are calculated based on weekly return data from the Centre for Research in Security
Prices (CRSP), while trade credit and other control variables are computed using firms’
annual financial statement data from Compustat. Following prior studies (e.g., Hutton,
Marcus, and Tehranian, 2009; Kim, Li, Lu, and Yu, 2016; Chang, Chen, and Zolotoy, 2017),
we exclude firms with year-end share prices below $1 and those with fewer than 26 weeks
of stock return data in fiscal years, firm-year observations with negative total assets and book
values of equity, financial firms (SIC code 60006999) and utilities (SIC code 49004999),
and firm-year observations with insufficient financial data to calculate relevant variables.
This procedure yields a final sample of 63,722 firm-year observations, with 7,540 unique
firms.
3.3.2. Measuring trade credit
Following Petersen and Rajan (1997), Fisman and Love (2003), Cuñat (2007), and
Giannetti, Burkart, and Ellingsen (2011), we measure trade credit as the ratio of accounts
payable to total assets. As a robustness check, we also consider alternative measures of trade
credit; see Section 4.4.
To compute alternative proxies for firm-specific stock price crash risk, we proceed
in two steps. First, we follow Hutton, Marcus, and Tehranian (2009) and calculate firm-
specific weekly returns by estimating the following equation:
17
The sample begins with 1990, which is the first year that we can use the Cash Flow Statement method to
calculate three-year moving average discretionary accruals.
94
where rj,τ is the weekly return on stock j in week τ, rm,τ is the return on CRSP value-weighted
market index, and ri,τ is the Fama and French value-weighted industry index in week τ.
Following Dimson (1979), we control for the impact of nonsynchronous trading by
incorporating the lead and lag terms. We estimate weekly returns from Wednesday to
Wednesday in order to avoid the contaminating effects from weekends and Mondays (Wang,
Li, and Erickson, 1997; Bartholdy and Peare, 2005). The firm-specific weekly return (Wj,τ)
is calculated as the log value of one plus the residual return from Eq. (1).
Second, we calculate two crash risk variables based on the estimated firm-specific
weekly returns (Wj,τ), following Chen, Hong, and Stein (2001) and Kim, Li, and Zhang
(2011a, b). The primary measure is the negative conditional skewness (NCSKEW), computed
as negative of the third moment of each stock’s firm-specific weekly returns divided by the
standard deviation raised to the third power. For firm j in fiscal year t,
3 2 3/2
where n is the number of observations on weekly returns in fiscal year t. A high NCSKEW
indicates a more left-skewed return distribution and a stock being more likely to crash.
Our second measure of firm-specific crash risk is “down-to-up volatility” (DUVOL),
which is calculated as follows:
2 2
where nu and nd are the number of up and down weeks over the fiscal year t, respectively.
For each stock j over fiscal year t, we separate all firm-specific weekly returns into down
(up) weeks when the weekly returns are below (above) the average value during the year.
We then calculate the standard deviation of firm-specific weekly returns for each of the
95
two groups separately. DUVOL is the log ratio of the standard deviation in the down weeks
to the standard deviation in the up weeks. Chen, Hong, and Stein (2001) suggest that this
alternative measure of crash risk may be less likely to be excessively affected by a small
number of extreme returns as it does not involve the third moments. Similar to NCSKEW, a
higher DUVOL represents a stock being more prone to crash.
Following prior literature (e.g., Chen, Hong, and Stein 2001; Jin and Myers, 2006),
we consider several control variables: DTURNt, the difference between the average monthly
share turnover over fiscal year t and t–1; SIGMAt, the standard deviation of firm-specific
weekly returns over fiscal year t; RETt, the average firm-specific weekly returns over fiscal
year t; SIZEt, defined as the log of total assets at the end of fiscal year t; MBt, the market
value of equity divided by the book value of equity at the end of fiscal year t; LEVt, the book
value of total debt scaled by total assets at the end of fiscal year t; ROAt, income before
extraordinary items divided by total assets at the end of fiscal year t; and NCSKEWt, the
negative conditional skewness for firm-specific weekly returns in fiscal year t. Following
Hutton, Marcus, and Tehranian (2009), we also control for accounting opacity, ACCMt,
defined as three years moving average absolute value of discretionary accruals, where
discretionary accruals are the residuals estimated from the modified Jones model (Dechow,
Ge, Larson, and Sloan, 1995). The Appendix provides the definitions of all variables used
in this study.
Table 1 reports the descriptive statistics for all variables used in our regressions.
The means of NCSKEWt+1 and DUVOLt+1 are 0.091 and 0.062, respectively, which are
96
quite similar to those reported in Kim, Li, and Zhang (2011b) (0.079). The average trade
credit ratio is 0.082, which is consistent with the reported value (0.089) in Fisman and Love
(2003). The descriptive statistics for the control variables further shows that our sample is
fairly in line with studies that utilize data from the same sources (e.g., Kim, Li, and Zhang,
2011b; Callen and Fang, 2015a).
3.4.2. Multivariate regression analysis
To test our prediction about the relationship between trade credit and future stock
price crash risk, we estimate the following regression model:
𝐶𝑟𝑎𝑠ℎ 𝑅𝑖𝑠𝑘𝑗,𝑡+1 = 𝛽0 + 𝛽1 𝑇𝐶𝑗,𝑡 + 𝛽2 𝐷𝑇𝑈𝑅𝑁𝑗,𝑡 + 𝛽3 𝑆𝐼𝐺𝑀𝐴𝑗,𝑡 + 𝛽4 𝑅𝐸𝑇𝑗,𝑡 + 𝛽5 𝑆𝐼𝑍𝐸𝑗,𝑡 +
𝛽6 𝑀𝐵𝑗,𝑡 + 𝛽7 𝐿𝐸𝑉𝑗,𝑡 + 𝛽8 𝑅𝑂𝐴𝑗,𝑡 + 𝛽9 𝑁𝐶𝑆𝐾𝐸𝑊𝑗,𝑡 + 𝛽10 𝐴𝐶𝐶𝑀𝑗,𝑡 + 𝜀𝑗,𝑡 ,
(4)
where the dependent variable Crash Riskt+1 is measured by NCSKEW and DUVOL in year
t+1 and all regressors are measured in year t. The independent variable of interest is trade
credit (TCt). Our model also includes a set of crash risk determinants mentioned above, as
well as year and industry fixed effects, which control for time trends and heterogeneity
across industries.
Table 2 presents the regression results. We regress NCSKEWt+1 on trade credit, TCt,
and the control variables in Column (1). We then include year fixed effects in Column (2)
and both year and industry fixed effects in Column (3). The results show that the coefficients
on TCt are significantly negative for NCSKEWt+1 (in Column (3), for example, coefficient =
0.226 and t-stat = 4.24). The results are qualitatively similar in Columns (4)-(6) when we
measure crash risk using DUVOLt+1. The negative coefficients on trade credit indicate that
97
firms’ purchase on account reduces their future stock price crash risk, consistent with the
monitoring view of trade credit that suppliers play an effective monitoring role in
constraining managers from hiding bad news.
Following Hutton, Marcus, and Tehranian (2009) and Callen and Fang (2015a), we
evaluate the economic significance of the results by taking the values of trade credit at the
25th and 75th percentiles (0.035 and 0.106, respectively), and then comparing crash risk at
those two percentile values while keeping all other variables at their mean values. Using the
results in Columns (3) and (6), we find that, on average, the decrease in NCSKEWt+1
(DUVOLt+1) is 17.85% (10.40%) of the sample mean corresponding to an increase in trade
credit from the 25th to the 75th percentiles of its distribution. These results show that the
mitigating impact of trade credit on crash risk is also economically significant, which
strongly supports the monitoring hypothesis H1a.
The results regarding the control variables are overall consistent with previous
research. In line with Chen, Hong, and Stein (2001), Kim, Li, and Zhang (2011b), and Callen
and Fang (2015a), the coefficients on stock turnover (DTURN), stock return volatility
(SIGMA), past returns (RET), and market-to-book ratio (MB) are positive and statistically
significant. Since stocks with higher turnover, returns and return volatility, and growth are
more likely to be overvalued, they are more prone to have future price collapses. Consistent
with Hutton, Marcus, and Tehranian (2009), we document negative coefficients on leverage
(LEV), which is likely driven by firms’ endogenous capital structure choice, as less crash-
prone firms can accumulate more debt. Finally, the observed positive coefficients on
negative skewness in year t (NCSKEW) and discretionary accruals (ACCM) are also
consistent with the existing empirical evidence in the literature (e.g., Hutton, Marcus, and
Tehranian, 2009; Kim, Li, and Zhang, 2011b; Callen and Fang, 2015a).
98
3.4.3. Addressing endogeneity
The baseline analysis identifies a significantly negative association between trade
credit and crash risk. However, endogeneity concerns such as reverse causality and omitted
variable issues may exist, leading to biased and inconsistent estimates and invalid inference.
In our regressions, we already measure crash risk in year t+1 and trade credit variables in
year t, which could more or less alleviate the potential concern about reverse causality. To
mitigate the potential omitted variable bias, we provide two additional tests.
First, we perform a two-stage least squares (2SLS) estimation, in which we use the
U.S. monetary stance (MS) and inventory ratio (INVENTORY) as instrumental variables
(IVs) for trade credit. Monetary stance is measured as U.S. Federal Fund rates (Gertler and
Gilchrist, 1994; Oliner and Rudebusch, 1996). Inventory ratio is defined as the total
inventories divided by total assets. We argue that both instruments satisfy the relevance and
exclusion conditions of valid IVs. When monetary policy is tight, the supply of bank credit
shrinks and firms tend to finance their projects more with trade credit (Mateut, Bougheas,
and Mizen, 2006). As in-kind financing, trade credit is typically granted in the form of input
supply and thus is positively related to the inventory ratio. Overall, the two IVs clearly meet
the relevance condition. Furthermore, they are plausibly exogenous as neither of them is
likely to directly influence managerial bad-news-hoarding behavior and subsequent stock
price crash risk. Panel A of Table 3 presents the 2SLS regression results. In the first two
models we separately use two instruments to predict trade credit in the first stage. In the last
model we use both instruments together. The results from the first-stage regressions show
that both instruments have significantly positive associations with trade credit, which is
consistent with our prediction. The results from the second-stage regressions in all models
confirm the baseline finding that trade credit is significantly negatively related to stock price
99
crash risk. 18 We also note that all diagnostic tests are satisfactory, again confirming the
validity of the IVs used.
Second, to further mitigate concerns about omitted correlated variables, we include
some additional variables that could influence both the use of trade credit and crash risk.
Dang, Lee, Liu, and Zeng (2017) suggest that due to the frequent renegotiations with lenders,
short-term debt subjects managers to stringent monitoring, thereby reducing managerial
discretion and lowering future crash risk. On the other hand, financially constrained firms
tend to reduce external financing costs by using discretionary accruals to cater to the pressure
of investors (Dechow, Ge, Larson, and Sloan, 2011), leading to more opaque financial
reports and higher stock price crash risk (Hutton, Marcus, and Tehranian, 2009; He and Ren,
2017). While short-term debt and financial constraints directly affect crash risk, they may
also be correlated with trade credit. Prior research shows that trade credit can act as a
substitute for other forms of short-term financing (Petersen and Rajan 1997; Nilsen 2002).
On the other hand, financially constrained firms may increase the use of trade credit,
particularly during a financial crisis when institutional credit is scarce (e.g., Garcia-
Appendini and Montoriol-Garriga, 2013). In sum, both trade credit and crash risk could be
potentially related to short-term debt financing or financial constraints. Hence, to alleviate
the concern that our results are driven by those alternative mechanisms, we include in the
regression model short-term debt (ST3) and conventional measures of financial constraints,
namely the dividend payout ratio (PAYOUT), the Hadlock-Pierce (2010) index (HPINDEX),
the Kaplan-Zingales (1997) index (KZINDEX), and the Whited-Wu (2006) index
(WWINDEX). The results presented in Panel B of Table 3 show that the negative relation
between trade credit and crash risk remains qualitatively unchanged after controlling for
18
In Panel A of Table 3, the dependent variable is NCSKEWt+1. The untabulated results using DUVOLt+1 are
qualitatively similar.
100
those additional variables, suggesting that our main finding is not driven by the effects of
short-term debt and financial constraints.
3.4.4. Robustness checks
To test the robustness of the main results, we repeat the baseline analysis using
alternative measures of stock price crash risk and trade credit. To rule out the effect of the
expected level of purchase on account, we first use abnormal trade credit (ABTC), defined
as the residuals from regressing trade credit on a set of determining factors. Following prior
research (e.g., Petersen and Rajan 1997; Love, Preve, and Sarria-Allende 2007; Garcia-
Appendini and Montoriol-Garriga, 2013; Abdulla, Dang, and Khurshed, 2017), the
determinants of trade credit include ln(1+age), ln(1+age)2, cash flow, cash holdings, current
assets, negative growth, positive growth, one-year maturity debt, and firm size. Second, we
define abnormal trade credit (ABTCIND) as the difference between a firm’s trade credit and
the industry-level median of trade credit for a given year. Finally, according to the matching
principle of trade credit (Petersen and Rajan, 1997), firms’ purchase on account may depend
on their trade credit provision. To test the offsetting effect of trade payables on crash risk,
we use net account payables (NETAP), calculated as the difference between accounts
payable and receivable, divided by total assets. Table 4 presents the regression results using
the above alternative measures of trade credit.19 We find that our evidence of a negative
19
Following Callen and Fang (2015a), we employ another proxy for crash risk, COUNT, measured as the
number of crash weeks minus the number of jump weeks over the fiscal year. Crash weeks are defined as those
weeks during which a firm experiences firm-specific weekly returns 3.09 standard deviations below the mean
firm-specific weekly returns over the whole fiscal year, with 3.09 being chosen to generate a frequency of 0.1%
in the normal distribution. We define jump weeks in the opposite way. The unreported results show that the
negative impact of trade credit on stock price crash risk persists.
101
relation between trade credit and stock price crash risk continues to hold. Overall, we provide
robust evidence in support of our hypothesis H1a.
3.5. Cross-sectional variations and underlying channels
Our empirical results have thus far provided support for the monitoring view of trade
credit rather than the concession view. In this section, we provide additional evidence on the
former view, especially the underlying channels driving the relation between trade credit and
stock price crash risk. We first examine whether the mitigating effect of trade credit on crash
risk is affected by governance and bank monitoring mechanisms. Further, we investigate
whether the monitoring effect of suppliers varies among buyers with different degrees of
market power and financial distress.
3.5.1. Governance mechanisms
Apart from trade credit suppliers, other corporate governance mechanisms, such as
institutional investors, can also impose monitoring pressures on managers in buyer firms
(e.g., Kim, Li, and Zhang, 2011b; Callen and Fang, 2015a). However, when buyers are less
effectively monitored through corporate governance, their managers may have more
opportunities to conceal private information and need to be monitored by other mechanisms.
Thus, we argue that firms will likely face stronger monitoring by their suppliers when other
governance and monitoring mechanisms are weak. Specifically, we expect the negative
relation between trade credit and future stock price crash risk to be more pronounced for
firms with less effective governance mechanisms.
To examine this prediction, we employ three proxies for governance monitoring
mechanisms: analyst coverage (COVER), total institutional ownership (INST), and transient
102
institutional ownership (LTINST). We obtain analyst forecast data from the I/B/E/S database
and define analyst coverage as the number of analysts following a company in a fiscal year.
Financial analysts can act as intermediaries to collect and disseminate information about
firms. It is widely documented that firms with lower analyst coverage have more information
asymmetry (Custódio, Ferreira, and Laureano, 2013). For instance, Yu (2008) argues that
firms followed by few analysts are weakly monitored and more prone to manipulate
earnings. We measure total institutional ownership (INST) as the percentage of shares held
by institutional owners, with the data obtained from the Thomson 13F database. Institutional
investors are more sophisticated and act as external monitors (Kim, Li, and Zhang, 2011b).
Further, following Bushee (1998, 2001), we classify institutional investors into quasi-
indexers, transient, or dedicated ones. We use transient institutional ownership (TRAINST)
as a proxy for weak governance because these investors typically pursue short-term profits
and thus are less inclined to monitor portfolio firms (Andreou, Antoniou, Horton, and Louca,
2016; Chang, Chen, and Zolotoy, 2016). Transient investors are different from dedicated
investors, who are more willing to monitor firms and thereby curb managerial myopic
behavior (e.g., Bushee, 1998, 2001).
In Table 5, we partition the sample into firms with weak and strong governance based
on the three measures described above. In Panel A, a firm is defined as having strong (weak)
governance if its analyst coverage is above (below) the annual median. The results show
that, the coefficients on trade credit are significantly negative at the 1% level for firms with
lower analyst coverage. However, for firms with higher analyst coverage, the coefficients
on trade credit are insignificant. Similarly, in Panel B, a firm is defined as having strong
(weak) governance if its total institutional ownership is above (below) the annual median.
We find that the negative coefficients on trade credit are significant for poorly governed
firms only, but insignificant for well governed firms. Panel C estimates the baseline
103
regression separately for the well governed subsample (i.e., firms with below-median
transient institutional ownership (TRAINST)) and the poorly-governed subsample (i.e., firms
with above-median TRAINST). The results also reveal that the coefficients on trade credit
are significantly negative only for firms with weak governance.
Overall, the above empirical findings are in line with our conjecture that the negative
effect of trade credit on crash risk is stronger when firms have weaker governance
mechanisms. This evidence suggests that firms’ trade credit can serve as a substitute for
those governance mechanisms in constraining managerial bad-news-hoarding behavior.
3.5.2. Bank monitoring mechanisms
In this subsection, we examine the impact of trade credit on crash risk for firms with
different levels of bank monitoring. Bank financing is a common way for firms to obtain
debt capital. To alleviate asymmetric information between banks and borrowers, and control
borrowers’ credit risk and earnings management behavior (Ahn and Choi, 2008), banks may
monitor their borrowers using both qualitative and quantitative measures, such as evaluating
the quality of borrowers’ management, using debt covenants, or requesting information
about the borrowers’ financial condition in a timely manner (Ranjan and Winton, 1995;
Treacy and Carey, 1998; Datta, Iskandar-Datta, and Raman, 2005). On the other hand, many
studies show that trade credit is a substitute form of credit and, importantly, has an
informational advantage over bank credit (e.g., Fabbri and Menichini, 2010; Garcia-
Appendini and Montoriol-Garriga, 2013; Murfin and Njoroge, 2015). Thus, we expect trade
creditors to monitor their buyers in a more effective way when banks could not offer
effective monitoring.
104
Following prior research, we measure the strength of bank monitoring using three
proxies, including the total volume of bank loans (TTLOAN), the magnitude of short-
maturity bank loans (STLOAN), and the length of bank loans (LENGTH). The data of
syndicated bank loans is from the DealScan database. We measure total bank loans
(TTLOAN) as the natural logarithm of one plus the amount of total existing bank loans in
year t. Khalil and Parigi (1998) contend that loan size is an important determinant of lenders’
commitment to monitor borrowers. Sufi (2007) suggests that the lead bank retains a higher
proportion of syndicated loans and forms a more concentrated syndicate when there is a
greater need to monitor borrowers. Similarly, we define short-maturity bank loans
(STLOAN) as the natural logarithm of one plus the amount of existing bank loans due within
three years. Short-maturity loans involve frequent renegotiations with banks and thus should
help to enhance managerial scrutiny (Ranjan and Winton, 1995; Graham, Li, and Qiu, 2008).
We measure the length of bank loans (LENGTH) as the number of years passed since loan
activation. Overall, a higher volume of bank loans and shorter maturity of loans are
associated with more powerful bank monitoring.
We split the sample into firms with strong and weak bank monitoring based on the
annual median values of those three measures, and present the results of our subsample tests
in Table 6. Panel A shows that, for both NCSKEWt+1 and DUVOLt+1, the coefficients on
trade credit are significantly negative for firms with smaller size of bank loans. However,
the coefficients are insignificant for firms with larger size of bank loans. Likewise, in Panel
B, the coefficients on trade credit are significantly negative for firms with less short-maturity
loans, but insignificantly or marginally negative for those with more short-maturity loans.
Panel C also shows that the coefficients on trade credit are statistically significant only for
firms that have longer length of bank loans.
105
To sum up, the empirical evidence in Table 6 indicates that monitoring by trade credit
suppliers can function as a substitute mechanism for bank monitoring in curbing managerial
bad-news-hoarding behavior and reducing stock price crash risk, especially when bank
monitoring is weak.
3.5.3. Market power mechanisms
We next test whether the negative relation between trade credit and future crash risk
could be impacted by market power mechanisms. Our analysis is motivated by Klapper,
Laeven, and Rajan (2012), who argue that large, creditworthy buyers tend to exercise market
power and extract favorable contract terms from small suppliers in order to obtain better
price discrimination. Bhattacharya and Nain (2011) and Fee and Thomas (2004) show that
customers becoming larger through mergers may exert price pressures on their suppliers. On
the contrary, buyers’ weak market power encourages suppliers to set stricter trade credit
terms to minimize customer default risk (Mian and Smith, 1992; Klapper, Laeven, and
Rajan, 2012). Due to the unbalanced market power among suppliers and buyers, the
“weaker” party should be more likely to play a “submissive” role in the supplier-buyer
relationship. According to the monitoring view, we expect to observe a stronger mitigating
effect of trade credit on crash risk for buyer firms with weak market power. However, if the
concession view is to hold, we predict that suppliers will be more likely to grant more
concessions to their powerful customers, implying weaker supplier monitoring and a
significantly positive relation between trade credit and crash risk for buyer firms with strong
market power.
We use three proxies for buyer firms’ market power: sales to industry sales ratio
(SALETIND), standardized inputs index (STINPUT), and fitted Herfindahl-Hirschman Index
106
(FITHHI). 20 Giannetti, Burkart, and Ellingsen (2011) find that buyer firms relying on
standardized inputs have stronger market power, while those using non-standardized inputs
have relatively weaker market power because standardized products are generally not unique
or customized so they may be easy to replace. Consistent with Giannetti, Burkart, and
Ellingsen (2011), we use the standardized inputs index (STINPUT), which is based on the
industry classification in Rauch (1999) and defined as the share of inputs that comes from
sectors producing standardized goods.21 Fabbri and Klapper (2016) suggest that in a more
competitive product market, buyer firms have weaker bargaining power than suppliers. We
employ the fitted Herfindahl-Hirschman index (FITHHI) based on Hoberg and Phillips
(2010) to proxy for market competition; the data is obtained from Hoberg-Phillips data
library. A lower value of FITHHI indicates a more competitive product market and weaker
buyer market power.
Table 7 presents the results separately for firms with strong and weak market power.
In Panel A we split the full sample based on the median SALETIND for each year. The results
show that, for both crash risk measures, the coefficients on trade credit are significantly
negative at the 1% significance level for firms with weaker market power (i.e., below-
median SALETIND), but insignificant for those with stronger market power (i.e., above-
median SALETIND). Panel B repeats the main regression for subsamples with above- and
below-median STINPUT. We find that the coefficients on trade credit are significantly
negative only for firms with weaker market power (those with an above-median standardized
inputs index). In a similar fashion, Panel C partitions the sample based on above- and below-
20
Following Klapper, Laeven, and Rajan (2012), we also proxy for market power using firm size and bond
rating status. We measure firm size as the natural logarithms of total assets and bond rating status as a dummy
variable that is equal to one if a firm has an S&P rating on long-term debt and zero otherwise. The unreported
results are strongly consistent with our findings in Table 7.
21
The index data of inputs is shown in the Appendix of Giannetti, Burkart, and Ellingsen (2011).
107
median FITHHI for each year. The estimation for both subsamples shows significantly
negative coefficients on trade credit only for less powerful firms, that is, firms with above-
median FITHHI and facing more intensive competition.
Overall, the results in Table 7 show that the negative impact of trade credit on crash
risk is more pronounced in firms with weak market power. These findings provide further
support for the monitoring view that suppliers can mitigate buyers’ bad-news-hoarding
behavior and reduce their stock price crash risk most effectively when those buyers have less
bargaining power.
3.5.4. Financial distress mechanisms
Finally, to provide additional evidence on the monitoring view versus the concession
view, we examine the role of financial distress. Financially distressed firms require more
monitoring because they are more likely to manipulate their disclosure (Summers and
Sweeney, 1998; Healy and Palepu, 2001; Graham, Harvey, and Rajgopal, 2005). Thus,
according to the monitoring view, we expect the impact of trade credit on crash risk to be
more pronounced for distressed firms. On the other hand, the concession view argues that
suppliers are willing to grant concessions to their financially distressed customers (Wilner,
2000; Cuñat, 2007). Hence, if this view is to hold, we predict that the impact of trade credit
on crash risk is significantly positive for distressed firms because suppliers would be lenient
and not constrain their distressed buyers from withholding bad news. To test those
competing predictions, we separately analyze financially distressed and non-distressed
firms.
We first measure financial distress following Altman’s (1968) methodology. In
particular, Altman’s z-score (ZSCORE) is computed as 1.2X1+1.4X2+3.3X3+0.6X4+1.0X5,
108
where X1 is working capital divided by book assets; X2 is retained earnings divided by book
assets; X3 is earnings before interest and taxes divided by book assets; X4 is the market value
of equity divided by total liabilities; and X5 is sales divided by book assets. A lower z-score
indicates a higher degree of financial distress. Following Davydenko (2012, 2013), our
second measure for financial distress is the asset-to-debt ratio (ADR), computed as the
market value of assets relative to the book value of debt. A lower asset-to-debt ratio also
indicates a higher degree of distress risk.
In Panel A of Table 8, we estimate our crash risk model separately for the subsamples
of firms with above- and below-median ZSCORE. The results show that the coefficients on
trade credit are significant and negative for firms with below-median ZSCORE (i.e.,
distressed firms) at the 1% significance level, but only marginally significant for firms with
above-median ZSCORE (i.e., non-distressed firms). Panel B splits the sample based on the
asset-to-debt ratio (ADR). The results indicate that the negative effect of trade credit on crash
risk is also significantly negative for the financially distressed group (i.e., firms with above-
media ADR) but insignificant for the non-distressed group (i.e., firms with below-median
ADR).
Overall, the results in Table 8 suggest that the mitigating impact of trade credit on
stock price crash risk is more pronounced among distressed firms, consistent with the
monitoring hypothesis. Again, we find little evidence in support of the concession view.
3.6. Conclusion
This study investigates the impact of trade credit on future stock price crash risk.
Based on prior studies, we develop two contrasting predictions regarding the role of trade
109
creditors: monitoring versus concession. If suppliers utilize their information and liquidation
advantage (Biais and Gollier, 1997; Burkart and Ellingsen, 2004; Fabbri and Menichini,
2010), they would effectively monitor customers and restrict managerial bad-news-hoarding
behavior, implying a negative effect of trade credit on crash risk. However, if suppliers are
willing to grant concessions to buyers, especially when buyers are experiencing financial
distress (Wilner, 2000; Cuñat, 2007), suppliers would be lenient to buyers’ misconduct,
facilitating buyers’ bad-news-hoarding behavior and stock price crash risk.
Using a large sample of U.S. public firms from 1990 to 2013, we show that firms’
use of trade credit is negatively related with future stock price crash risk. This negative
relation remains robust after addressing potential endogeneity concerns and using alternative
measures of crash risk and trade credit. Our findings are in line with the monitoring view
that predicts a negative relation between firms’ trade credit and crash risk.
We further analyze whether the mitigating effect of trade credit on crash risk is
affected by governance, bank monitoring, market power, and financial distress mechanisms.
Our empirical results show that the negative relation between trade credit and crash risk is
more pronounced for firms with weaker governance or bank monitoring, suggesting that
supplier monitoring can serve as a substitute for those mechanisms in monitoring customers’
misconducts. Further analysis shows that this negative relation is stronger for firms with
weaker market power or those with higher distress risk, which further strengthens the
monitoring view of trade credit.
This study contributes to the growing literature on stock price crash risk. We show a
novel factor that can determine future stock price crash risk – trade credit. Our study also
contributes to research on trade credit. In particular, it adds to the ongoing debate on the
benefits versus consequences of supplier-buyer relationships. Our evidence suggests that
110
suppliers are inclined to monitor their buyers and restrict them from withholding negative
information, rather than granting them concessions to avoid negative contagion effects. The
practical implication of this study is that trade credit suppliers can exert a positive impact on
customers’ value by reducing the latter’s likelihood of extreme negative stock returns.
111
References
Aktas, N., Bodt, E., Lobez, F., and Statnik, J. C., 2012. The information content of trade
credit. Journal of Banking & Finance, 36(5), 1402–1413.
Albuquerque, R., Ramadorai, T., and Watugala, S. W. Trade credit and cross-country
predictable firm returns. Journal of Financial Economics, 115(3), 592613.
Altman, E.I., 1968. Financial ratios, discriminant analysis and the prediction of corporate
bankruptcy. Journal of Finance, 23(4), 589–609.
An, H., and Zhang, T., 2013. Stock price synchronicity, crash risk, and institutional investors.
Journal of Corporate Finance, 21, 1–15.
Andreou, P.C., Antoniou, C., Horton, J., and Louca, C., 2016. Corporate Governance and
Firm-specific Stock Price Crashes. European Financial Management, 22(5), 916–956.
Andreou, P. C., Louca, C., and Petrou, A. P., 2016. CEO Age and Stock Price Crash Risk.
Review of Finance, 21(3), 1287–1325.
Barrot, J.N., 2015. Trade credit and industry dynamics: Evidence from trucking firms.
Journal of Finance, 71(5), 1975–2016.
Bartholdy, J., and Peare, P., 2005. Estimation of expected return: CAPM vs. Fama and
French. International Review of Financial Analysis, 14(4), 407–427.
Bhattacharyya, S., and Nain, A., 2011. Horizontal acquisitions and buying power: A product
market analysis. Journal of Financial Economics, 99(1), 97–115.
Biais, B., and Gollier, C., 1997, Trade credit and credit rationing, Review of Financial
Studies, 10(4), 903–937.
112
Boehme, R.D., Fotak, V., and May, A.D., 2017. Crash risk and seasoned equity offerings.
Working paper.
Boone, A.L., and White, J.T., 2015. The effect of institutional ownership on firm
transparency and information production. Journal of Financial Economics, 117(3), 508-
533.
Brennan, M., Maksimovic V., and Zechner J., 1988, Vendor financing, Journal of Finance,
43(5), 1127–1141.
Brick, I.E. and Fung, W.K., 1984. Taxes and the theory of trade debt. Journal of Finance,
39(4), 1169–1176.
Burkart, M., and Ellingsen T., 2004, In-kind finance: a theory of trade credit, American
Economic Review, 94(3), 569–590.
Bushee, B.J., 2001. Do institutional investors prefer near-term earnings over long-run value?
Contemporary Accounting Research, 18(2), 207–246.
Bushee, B.J., 1998. The influence of institutional investors on myopic R&D investment
behavior. The Accounting Review, 73(3), 305–333.
Bushman, R. M., and Smith, A.J., 2001. Financial accounting information and corporate
governance. Journal of Accounting and Economics, 32(1), 237–333.
Callen, J.L., and Fang, X., 2013. Institutional investor stability and crash risk: Monitoring
versus short-termism? Journal of Banking & Finance, 37(8), 3047–3063.
Callen, J.L., and Fang, X., 2015a. Religion and stock price crash risk. Journal of Financial
and Quantitative Analysis, 50 (1–2), 169–195.
113
Callen, J.L., and Fang, X., 2015b. Short interest and stock price crash risk. Journal of
Banking & Finance, 60, 181–194.
Campello, M., and Gao, J., 2017. Customer concentration and loan contract terms. Journal
of Financial Economics, 123(1), 108–136.
Chang, X., Chen, Y., and Zolotoy, L., 2017. Stock liquidity and stock price crash risk.
Journal of Financial and Quantitative Analysis, 52(4), 1605–1637.
381.
Chen, C., Kim, J.B., and Yao, L., 2017. Earnings smoothing: Does it exacerbate or constrain
stock price crash risk? Journal of Corporate Finance, 42, 36-54.
Chod, J., Trichakis, N., and Tsoukalas, G., 2016. A signaling theory of in-kind finance.
Working paper.
Cohen, L. and Frazzini, A., 2008. Economic links and predictable returns. Journal of
Finance, 63(4), 1977–2011.
Cuñat, V., 2007, Trade credit: suppliers as debt collectors and insurance providers, Review
Custódio, C., Ferreira, M.A., and Laureano, L., 2013. Why are US firms using more short-
term debt? Journal of Financial Economics, 108(1), 182–212.
Dang, V.A., Lee, E., Liu, Y., and Zeng, C., 2017. Corporate debt maturity and stock price
crash risk. European Financial Management, 1–34.
114
Davydenko, S., 2012. When do firms default? A study of the default boundary. Working
Paper. University of Toronto.
Davydenko, S., 2013. Insolvency, illiquidity, and the risk of default. Working Paper.
University of Toronto.
Dechow, P.M., Ge, W., Larson, C.R., and Sloan, R.G., 2011. Predicting material accounting
misstatements. Contemporary Accounting Research, 28(1), 17–82.
crash risk? The Accounting Review, 90(1), 265–299.
Desai, M.A., Foley, C.F., and Hines Jr, J.R., 2016. Trade credit and taxes. Review of
Economics and Statistics, 98(1), 132–139.
Dhaliwal, D., Huang, S., Khurana, I.K., and Pereira, R., 2014. Product market competition
and conditional conservatism. Review of Accounting Studies, 19(4), 1309–1345.
Dimson, E., 1979. Risk measurement when shares are subject to infrequent trading. Journal
Emery, G., 1984. A pure financial explanation for trade credit. Journal of Financial and
Quantitative Analysis, 19(3), 271–285.
Fabbri, D., and A. Menichini. 2010. Trade credit, collateral liquidation, and borrowing
constraints. Journal of Financial Economics, 96(3), 413–32.
Fee, C.E., and Thomas, S., 2004. Sources of gains in horizontal mergers: evidence from
customer, supplier, and rival firms. Journal of Financial Economics, 74(3), 423–460.
115
Ferris, J., 1981. A transactions theory of trade credit use. Quarterly Journal of Economics,
96(2), 243–270.
Fisman, R., and Love, I. 2003. Trade credit, financial intermediary development, and
industry growth. Journal of Finance, 58(1), 353–374.
Frank, M., and Maksimovic, V., 1998. Trade credit, collateral, and adverse selection.
Unpublished manuscript, University of Maryland.
Gertler, M., and Gilchrist, S., 1994. Monetary policy, business cycles, and the behavior of
small manufacturing firms. Quarterly Journal of Economics, 109(2), 309–340.
Giannetti, M., Burkart, M., and Ellingsen, T., 2011. What you sell is what you lend?
Explaining trade credit contracts, Review of Financial Studies, 24(4), 1261–1298.
Goto, S., and Xiao, G., 2015. As told by the supplier: Trade credit and the cross section of
stock returns. Journal of Banking & Finance, 60, 296–309.
Graham, J.R., Li, S., and Qiu, J., 2008. Corporate misreporting and bank loan contracting.
Graham, J.R., Harvey, C.R., and Rajgopal, S., 2005. The economic implications of corporate
financial reporting. Journal of Accounting and Economics, 40(1), 3–73.
Hadlock, C.J., and Pierce, J.R., 2010. New evidence on measuring financial constraints:
Moving beyond the KZ index. Review of Financial Studies, 23(5), 1909–1940.
He, G., 2015. The effect of CEO inside debt holdings on financial reporting quality. Review
of Accounting Studies, 20(1), 501–536.
He, G. and Ren, H., 2017. Financial constraints and future stock price crash risk. Woking
paper.
116
Hertzel, M.G., Li, Z., Officer, M.S., and Rodgers, K.J., 2008. Inter-firm linkages and the
wealth effects of financial distress along the supply chain. Journal of Financial
Economics, 87(2), 374–387.
Hertzel, M.G., and Officer, M.S., 2012. Industry contagion in loan spreads. Journal of
Financial Economics, 103(3), 493–506.
Hoberg, G., and Phillips, G. 2010. Product market synergies and competition in mergers and
acquisitions: A text-based analysis. Review of Financial Studies, 23(10), 3773–3811.
Hoberg, G., Phillips, G., and Prabhala, N. 2014. Product market threats, payouts, and
financial flexibility. Journal of Finance, 69(1), 293–324.
Hoberg, G., and Phillips, G. 2016. Text-based network industries and endogenous product
differentiation. Journal of Political Economy, 124(5), 1423–1465.
Hong, H.A., Kim, J.B., and Welker, M., 2017. Divergence of cash flow and voting rights,
opacity, and stock price crash risk: international evidence. Journal of Accounting
Research, 55(5), 1167–1212.
risk. Journal of Financial Economics, 94, 67–86.
Jacobson, T., and Schedvin, E., 2015. Trade credit and the propagation of corporate failure:
an empirical analysis. Econometrica, 83(4), 1315–1371.
Jain, N., 2001. Monitoring costs and trade credit. The Quarterly Review of Economics and
Finance, 41(1), 89–110.
Jin, L., and Myers, C.S., 2006. R2, around the world: New theory and new tests. Journal of
117
Jorion, P., and Zhang, G., 2009. Credit contagion from counterparty risk. Journal of Finance,
64(5), 2053–2087.
Kalwani, M.U., and Narayandas, N., 1995. Long-term manufacturer-supplier relationships:
do they pay off for supplier firms? Journal of Marketing, 59(1), 1–16.
Kaplan, S.N., and Zingales, L., 1997. Do investment-cash flow sensitivities provide useful
measures of financing constraints?. Quarterly Journal of Economics, 112(1), 169–215.
Kelly, B., Lustig, H., and Van Nieuwerburgh, S., 2013. Firm volatility in granular networks
(No. w19466). National Bureau of Economic Research.
Khalil, F., and Parigi, B., 1998. Loan size as a commitment device. International Economic
Review, 39, 135–150.
Khurana, I.K., Pereira, R., and Zhang, E.X., 2017. Is real earnings smoothing harmful?
Evidence from firm-specific stock price crash risk. Contemporary Accounting Research,
Forthcoming.
Kim, J.B., Li, L., Lu, L.Y., and Yu, Y., 2016. Financial statement comparability and
expected crash risk. Journal of Accounting and Economics, 61(2–3), 294–312.Kim, J.B.,
Li, Y., and Zhang, L., 2011a. CFO vs. CEO: equity incentives and crashes. Journal of
Financial Economics, 101 (3), 713–730.
Kim, J.B., Li, Y., and Zhang, L., 2011b. Corporate tax avoidance and stock price crash risk:
Firm level analysis. Journal of Financial Economics, 100 (3), 639–662.
Kim, J.B., Wang, Z., and Zhang, L., 2016. CEO overconfidence and stock price crash risk.
Kim, J.B., and Zhang, L., 2015. Accounting conservatism and stock price crash risk: Firm‐
118
Klapper, L., Laeven, L., and Rajan, R., 2012. Trade credit contracts. Review of Financial
Studies, 25(3), 838–867.
Kolay, M., Lemmon, M., and Tashjian, E., 2016. Spreading the misery? Sources of
bankruptcy spillover in the supply chain. Journal of Financial and Quantitative Analysis,
51(6), 1955–1990.
Lamont, O., Polk, C., and Saaá-Requejo, J., 2001. Financial constraints and stock returns.
Review of Financial Studies, 14(2), 529–554.
Lee, Y.W., and Stowe, J.D., 1993. Product risk, asymmetric information, and trade credit.
Levine, R., Lin, C., and Xie, W., 2016. Corporate resilience to banking crises: the roles of
trust and trade credit (No. w22153). National Bureau of Economic Research.
Long, M., Malitz, I., and Ravid, S., 1993. Trade credit, quality guarantees and product
marketability. Financial Management, 22, 117–127.
Love, I., Preve, L., and Sarria-Allende, V., 2007. Trade credit and bank credit: Evidence
from recent financial crises. Journal of Financial Economics, 83(2), 453–469.
Mateut, S., Bougheas, S., and Mizen, P., 2006. Trade credit, bank lending and monetary
policy transmission. European Economic Review, 50(3), 603–629.
Martínez‐Sola, C., García‐Teruel, P. J., and Martínez‐Solano, P., 2013. Trade credit policy
and firm value. Accounting & Finance, 53(3), 791–808.
Mian, S., and Smith, C. W., 1992. Accounts receivable management policy: theory and
evidence. Journal of Finance, 47(1), 169–200.
119
Murfin, J., and Njoroge, K., 2015. The implicit costs of trade credit borrowing by large firms.
Review of Financial Studies, 28 (1), 112–145.
Ng, C., Smith, J., and Smith, R., 1999. Evidence on the determinants of credit terms used in
interfirm trade. Journal of Finance, 54(3), 1109–1129.
Nilsen, J., 2002. Trade credit and the bank lending channel. Journal of Money, Credit, and
Banking, 34, 226–253.
Oliner, S.D., and Rudebusch, G.D., 1996. Is there a broad credit channel for monetary
policy?. Economic Review-Federal Reserve Bank of San Francisco, (1), 3.
Studies, 10(3), 661–691.
Rajan, R., and Winton, A., 1995. Covenants and collateral as incentives to monitor. Journal
of Finance, 50(4), 1113–1146.
Raman, K., and Shahrur, H., 2008. Relationship-specific investments and earnings
management: Evidence on corporate suppliers and customers. The Accounting Review,
83(4), 1041–1081.
Rauch, J. E., 1999. Networks versus Markets in International Trade. Journal of International
Economics, 48(1), 7–35.
Schwartz, R., 1974. An economic model of trade credit. Journal of Financial and
Quantitative Analysis, 9(4), 643–657.
Schwartz, R., and Whitcomb, D., 1979. The trade credit decision. Handbook of Financial
Economics, 257–73.
120
Smith, J., 1987. Trade credit and informational asymmetry. Journal of Finance, 42(4), 863–
872.
Sufi, A., 2007. Information asymmetry and financing arrangements: Evidence from
syndicated loans. Journal of Finance, 62 (2), 629–668.
Summers, S.L. and Sweeney, J.T., 1998. Fraudulently misstated financial statements and
insider trading: An empirical analysis. The Accounting Review, 131–146.
Treacy, W.F., and Carey, M.S., 1998. Credit risk rating at large US banks. Federal Reserve
Bulletin, 84, 897.
Xu, N., Li, X., Yuang, Q., and Chan., K. C., 2014. Excess perks and stock price crash risk:
Evidence from China. Journal of Corporate Finance, 25, 419–434.
Wang, K., Li, Y., and Erickson, J., 1997. A new look at the Monday effect. Journal of
Finance, 52(5), 2171–2186.
Whited, T.M., and Wu, G., 2006. Financial constraints risk. Review of Financial Studies,
19(2), 531–559.
Wilner, B.S., 2000. The exploitation of relationships in financial distress: The case of trade
credit. Journal of Finance, 55(1), 153–178.
Wintoki, M. B., Linck, J. S., and Netter, J. M., 2012. Endogeneity and the dynamics of
internal corporate governance. Journal of Financial Economics, 105(3), 581–606.
Yu, F., 2008. Analyst coverage and earnings management. Journal of Financial Economics,
88(2), 245–271.
121
Table 1. Descriptive statistics
The table reports the descriptive statistics for crash risk, trade credit, and other variables. The sample consists
of 63,722 firm-year observations for 7,540 public U.S. firms over the period from 1990 to 2014. Variable
definitions are listed in Appendix. All variables are winsorized at 1% and 99%.
Variable N Mean Std 5% 25% Median 75% 95%

NCSKEWt+1 63,722 –0.091 0.771 –1.312 –0.523 –0.117 0.297 1.259
DUVOLt+1 63,722 –0.062 0.356 –0.638 –0.300 –0.072 0.164 0.556
COUNTt+1 63,722 –0.041 0.657 –1.000 0.000 0.000 0.000 1.000
TCt 63,722 0.082 0.070 0.010 0.035 0.064 0.106 0.224
DTURNt 63,722 0.027 0.858 –1.228 –0.222 0.002 0.241 1.381
SIGMAt 63,722 0.057 0.029 0.023 0.036 0.051 0.072 0.115
RETt 63,722 –0.203 0.219 –0.653 –0.256 –0.127 –0.063 –0.025
SIZEt 63,722 6.016 2.075 2.821 4.471 5.877 7.430 9.731
MBt 63,722 2.814 2.753 0.615 1.227 1.978 3.289 7.949
LEVt 63,722 0.197 0.177 0.000 0.024 0.173 0.316 0.527
ROAt 63,722 0.013 0.141 –0.257 –0.001 0.041 0.080 0.157
NCSKEWt 63,722 –0.088 0.750 –1.274 –0.512 –0.115 0.290 1.226
ACCMt 63,722 0.072 0.065 0.014 0.031 0.052 0.089 0.201
MSt 63,722 0.135 0.138 0.000 0.016 0.101 0.206 0.417
INVENTORYt 63,201 3.349 2.299 0.100 1.130 3.880 5.300 6.240
COVERt 63,722 6.936 9.092 0.000 0.000 3.000 10.000 27.000
INSOWNt 50,708 0.472 0.300 0.019 0.202 0.473 0.727 0.956
TRAOWNt 50,708 0.010 0.025 0.000 0.000 0.000 0.009 0.047
LENGTHt 35,661 1.885 2.323 0.000 0.000 1.000 3.000 6.000
TTLOANt 63,722 10.860 9.686 0.000 0.000 16.730 19.740 21.860
STLOANt 63,722 9.618 9.620 0.000 0.000 14.460 19.210 21.470
SALETINDt 63,621 14.900 54.330 0.080 0.515 1.615 6.055 58.980
STINPUTt 63,722 0.504 0.123 0.320 0.360 0.550 0.620 0.650
FITHHIt 39,400 0.057 0.023 0.035 0.043 0.051 0.065 0.102
ZSCOREt 52,320 66.280 303.500 1.415 2.945 4.898 11.210 212.900
ADRt 53,544 121.200 560.000 2.018 3.773 7.050 18.840 397.700
122
Table 2. Impact of trade credit on stock price crash risk
This table presents the regression results of the effect of trade credit on firm-level stock price crash risk. All
variables except dummies are winsorized at the 1% and 99% levels. The standard errors in the regressions are
clustered at the firm level. The numbers reported in parentheses are t-statistics. ***, **, and * denote statistical
significance at the 1%, 5% and 10% levels, respectively. See Appendix for variable definitions.
Predicted (1) (2) (3) (4) (5) (6)

Sign NCSKEWt+1 NCSKEWt+1 NCSKEWt+1 DUVOLt+1 DUVOLt+1 DUVOLt+1
TCt +/– –0.171*** –0.125** –0.226*** –0.068*** –0.039* –0.090***
(–3.51) (–2.54) (–4.24) (–3.15) (–1.81) (–3.75)
DTURNt + 0.011*** 0.011*** 0.011*** 0.004*** 0.005*** 0.005***
(2.90) (2.89) (2.95) (2.63) (2.70) (2.71)
SIGMAt + 3.733*** 4.084*** 3.684*** 1.556*** 1.647*** 1.474***
(8.60) (9.05) (8.02) (7.75) (7.95) (6.99)
RETt + 0.525*** 0.572*** 0.538*** 0.229*** 0.249*** 0.234***
(9.81) (10.56) (9.83) (9.36) (10.09) (9.38)
SIZEt + 0.053*** 0.048*** 0.049*** 0.025*** 0.022*** 0.022***
(26.68) (22.79) (21.97) (27.43) (22.67) (21.49)
MBt + 0.026*** 0.026*** 0.025*** 0.012*** 0.012*** 0.012***
(20.27) (19.92) (18.77) (20.91) (20.60) (19.72)
LEVt – –0.242*** –0.215*** –0.185*** –0.114*** –0.099*** –0.089***
(–12.68) (–11.14) (–9.03) (–12.86) (–11.00) (–9.40)
ROAt +/– 0.381*** 0.391*** 0.397*** 0.198*** 0.202*** 0.204***
(14.47) (14.77) (14.67) (17.06) (17.40) (17.01)
NCSKEWt + 0.038*** 0.037*** 0.034*** 0.018*** 0.017*** 0.016***
(8.52) (8.26) (7.61) (8.72) (8.40) (7.83)
ACCMt + 0.059 0.043 –0.005 0.008 –0.002 –0.019
(1.09) (0.80) (–0.09) (0.34) (–0.08) (–0.74)
Intercept –0.534*** –0.553*** –0.619*** –0.265*** –0.268*** –0.281***
(–22.71) (–20.19) (–10.05) (–24.47) (–20.96) (–9.49)
Year FE No Yes Yes No Yes Yes
Industry FE No No Yes No No Yes
N 63,722 63,722 63,722 63,722 63,722 63,722
Adjusted R2 0.040 0.046 0.049 0.046 0.054 0.056
123
Table 3. Regression analysis to address endogeneity issues
This table presents the regression analysis to address potential endogeneity issues. Panel A presents the results of two-stage least squares regressions. The instruments for trade
credit (TC) are monetary stance (MS) and the inventory-to-assets ratio (INVENTORY). Monetary stance is measured as U.S. federal fund rates. The inventory-to-assets ratio
is measured as the ratio of inventory to total assets. In the first-stage regression, we regress TC on the control and instrumental variables. In the second-stage regression, we
regress crash risk (NCSKEWt+1) on the predicted TC together with other determinants of crash risk. Panel B presents the results of the effect of trade credit on firm-level stock
price crash risk after including additional controls. VIY indicates a set of controls including all explanatory variables, industry effects, and year effects; see Table 2 for details.
All variables except dummies are winsorized at the 1% and 99% levels. The standard errors in the regressions are clustered at the firm level. The numbers reported in parentheses
are t-statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels, respectively. See Appendix for other variable definitions.
Panel A. Instrumental variable/two-stage least squares regressions

(1) (2) (3)
First stage Second stage First stage Second stage First stage Second stage
Dependent variable TCt NCSKEWt+1 TCt NCSKEWt+1 TCt NCSKEWt+1
TCt –1.862*** –0.354** –0.562***
(–2.96) (–1.97) (–3.29)
Instrumental variable
MSt 0.002*** 0.002***
(19.69) (8.54)
INVENTORYt 0.172*** 0.175***
(53.92) (18.82)
Diagnostic test
Hausman test (p–value) 0.000*** 0.000*** 0.032**
F–statistic 387.58 2907.37 1730.06
N 63,722 63,722 63,201 63,201 63,201 63,201
Adjusted R2 0.23 0.29 0.29
124
Panel B. Regression analysis with additional controls
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1
TCt –0.162** –0.067** –0.217*** –0.085*** –0.216*** –0.085*** –0.255*** –0.106*** –0.225*** –0.089***
(–2.54) (–2.34) (–3.73) (–3.28) (–4.06) (–3.58) (–4.62) (–4.33) (–4.21) (–3.71)
ST3t –0.042*** –0.017***
(–3.46) (–2.98)
PAYOUTt –0.009 –0.004
(–1.45) (–1.63)
HPINDEXt –0.109*** –0.046***
(–10.65) (–9.58)
KZINDEXt –0.000 –0.000
(–1.13) (–0.68)
WWINDEXt –0.001*** –0.000***
(–9.82) (–9.41)
N 45,049 45,049 53,254 53,254 63,722 63,722 60,578 60,578 63,164 63,164
Adjusted R2 0.049 0.055 0.048 0.055 0.051 0.058 0.049 0.057 0.050 0.058
125
Table 4. Alternative measures of trade credit
This table presents the regressions of crash risk on alternative measures of trade credit. ABTC is abnormal
trade payables, calculated as the residuals from the regression of trade credit on a set of determinants, namely
ln(1+age), ln(1+age)2, cash flow, cash holdings, current assets, negative growth, positive growth, one-year
maturity debt, and firm size. ABTCIND is the difference between trade credit (TC) and the median industry
trade credit in year t. NETAP is the net account payables, defined as the difference between account payables
and receivables, divided by total assets. VIY indicates a set of controls including all explanatory variables,
industry effects, and year effects; see Table 2 for details. All variables except dummies are winsorized at the
1% and 99% levels. The standard errors in the regressions are clustered at the firm level. The numbers reported
in parentheses are t-statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels,
respectively. See Appendix for other variable definitions.
(1) (2) (3) (4) (5) (6)

NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1
ABTCt –0.208*** –0.092***
(–3.05) (–3.03)
ABTCINDt –0.230*** –0.093***
(–4.30) (–3.90)
NETAPt –0.088** –0.040**
(–2.55) (–2.55)
N 50,247 50,247 63,722 63,722 63,722 63,722
Adjusted R2 0.049 0.056 0.049 0.056 0.049 0.056
126
Table 5. Differential impacts of trade credit on crash risk: External governance
mechanisms
This table presents the results regarding the impact of trade credit on future stock price crash risk conditional
on governance mechanisms. In Panel A, the sample is partitioned based on the median value of the number of
analyst coverage (COVER) for each year. In Panel B, the sample is divided based on the median value of total
institutional ownership (INST) for each year. In Panel C, the sample is divided based on the median value of
transient institutional ownership (TRAINST) for each year. VIY indicates a set of controls including all
explanatory variables, industry effects, and year effects; see Table 2 for details. All variables except dummies
are winsorized at the 1% and 99% levels. The standard errors in the regressions are clustered at the firm level.
The numbers reported in parentheses are t-statistics. ***, **, and * denote statistical significance at the 1%,
5% and 10% levels, respectively. See Appendix for other variable definitions.
Panel A. Analyst coverage

(1) (2) (3) (4)
NCSKEWt+1 NCSKEWt+1 DUVOLt+1 DUVOLt+1
Weak governance Strong governance Weak governance Strong governance
(COVER ≤ Median) (COVER > Median) (COVER ≤ Median) (COVER > Median)
TCt –0.193*** –0.082 –0.068** –0.035
(–2.73) (–1.06) (–2.15) (–1.00)
N 32,291 28,725 32,291 28,725
Adjusted R2 0.058 0.028 0.064 0.036
Panel B. Institutional ownership
(1) (2) (3) (4)
(INST ≤ Median) (INST > Median) (INST ≤ Median) (INST > Median)
TCt –0.208*** 0.020 –0.066* –0.001
(–2.61) (0.22) (–1.84) (–0.02)
N 25,300 25,284 25,300 25,284
Adjusted R2 0.056 0.028 0.065 0.033
Panel C. Transient institutional ownership
(1) (2) (3) (4)
(TRAINST ≥ Median) (TRAINST< Median) (TRAINST ≥ Median) (TRAINST< Median)
TCt –0.208*** –0.181 –0.090*** –0.052
(–3.07) (–1.52) (–2.95) (–0.96)
N 25,300 25,284 25,300 25,284
Adjusted R2 0.058 0.029 0.067 0.034
127
Table 6. Differential impacts of trade credit on crash risk: Bank monitoring
mechanisms
on bank monitoring mechanisms. In Panel A, the sample is divided based on the median magnitude of bank
loans (TTLOAN) for each year. In Panel B, the sample is divided based on the median magnitude of short-
maturity bank loans (STLOAN) for each year. In Panel C, the sample is partitioned based on the median length
of bank loans (LENGTH) for each year. Syndicated bank loans data is from the DealScan database. VIY
indicates a set of controls including all explanatory variables, industry effects, and year effects; see Table 2 for
details. All variables except dummies are winsorized at the 1% and 99% levels. The standard errors in the
regressions are clustered at the firm level. The numbers reported in parentheses are t-statistics. ***, **, and *
denote statistical significance at the 1%, 5% and 10% levels, respectively. See Appendix for other variable
definitions.
Panel A. Magnitude of bank loans

(1) (2) (3) (4)
Less More Less More
TCt –0.342*** –0.093 –0.133*** –0.042
(–4.77) (–1.17) (–4.18) (–1.17)
N 35,046 28,676 35,046 28,676
Adjusted R2 0.058 0.035 0.066 0.041
Panel B. Magnitude of short-maturity bank loans
(1) (2) (3) (4)
Less More Less More
TCt –0.321*** –0.118 –0.127*** –0.047
(–4.61) (–1.47) (–4.08) (–1.31)
N 36,724 26,998 36,724 26,998
Adjusted R2 0.056 0.036 0.065 0.042
Panel C. Length of bank loans (leftover years)
(1) (2) (3) (4)
Long length Short length Long length Short length
TCt –0.237** –0.080 –0.120** –0.014
(–2.17) (–0.89) (–2.44) (–0.33)
N 13833 21825 13833 21825
2
Adjusted R 0.041 0.036 0.047 0.042
128
Table 7. Differential impacts of trade credit on crash risk: Market power
mechanisms
on market power mechanisms. In Panel A, the sample is partitioned based on the median sales-to-industry-
sales ratio (SALETIND) for each year. In Panel B, the sample is divided based on the median standardized
inputs index (STINPUT) for each year. In Panel C, the sample is divided based on the median Herfindahl index
(FITHHI) for each year. VIY indicates a set of controls including all explanatory variables, industry effects,
and year effects; see Table 2 for details. All variables except dummies are winsorized at the 1% and 99% levels.
The standard errors in the regressions are clustered at the firm level. The numbers reported in parentheses are
t-statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels, respectively. See
Appendix for other variable definitions.
Panel A. Sale/industry sale

(1) (2) (3) (4)
Weak market power Strong market power Weak market power Strong market power
(SALETIND ≤ Median) (SALETIND > Median) (SALETIND ≤ Median) (SALETIND > Median)
TCt –0.265*** –0.107 –0.101*** –0.054
(–3.25) (–1.48) (–2.78) (–1.64)
N 32,087 31,538 32,087 31,538
Adjusted R2 0.060 0.026 0.067 0.031
Panel B. Standardized inputs
(1) (2) (3) (4)
(STINPUT ≥ Median) (STINPUT < Median) (STINPUT ≥ Median) (STINPUT < Median)
TCt –0.273*** –0.128 –0.113*** –0.045
(–3.93) (–1.28) (–3.65) (–0.99)
N 27,829 26,442 27,829 26,442
Adjusted R2 0.050 0.046 0.055 0.054
Panel C. Herfindahl Index
(1) (2) (3) (4)
(FITHHI ≤ Median) (FITHHI > Median) (FITHHI ≤ Median) (FITHHI > Median)
TCt –0.303*** –0.124 –0.144*** –0.033
(–3.26) (–1.18) (–3.43) (–0.68)
N 20,027 19,373 20,027 19,373
Adjusted R2 0.071 0.053 0.079 0.058
129
Table 8. Differential impacts of trade credit on crash risk: Financial distress
mechanisms
on financial distress mechanisms. In Panel A, the sample is partitioned based on the median value of the
Altman’s z-score (ZSCORE) for each year. In Panel B, the sample is divided based on the median value of the
asset-to-debt ratio (ADR) for each year. VIY indicates a set of controls including all explanatory variables,
industry effects, and year effects; see Table 2 for details. All variables except dummies are winsorized at the
1% and 99% levels. The standard errors in the regressions are clustered at the firm level. The numbers reported
in parentheses are t-statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels,
respectively. See Appendix for other variable definitions.
Panel A. ZSCORE
(1) (2) (3) (4)
Distressed Non-distressed Distressed Non-distressed
(ZSCORE ≤ Median) (ZSCORE > Median) (ZSCORE ≤ Median) (ZSCORE > Median)
TCt –0.281*** –0.158* –0.115*** –0.062*
(–3.14) (–1.93) (–2.84) (–1.69)
N 26,403 25,915 26,403 25,915
Adjusted R2 0.050 0.045 0.055 0.054
Panel B. Asset-to-debt ratio
(1) (2) (3) (4)
Distressed Non-distressed Distressed Non-distressed
(ADR≤ Median) (ADR > Median) (ADR≤ Median) (ADR > Median)
TCt –0.280*** –0.143* –0.111*** –0.063*
(–3.12) (–1.87) (–2.76) (–1.83)
N 27,017 26,526 27,017 26,526
Adjusted R2 0.049 0.045 0.053 0.054
130
Appendix. Variable definitions
The firm-specific weekly return (W) is equal to ln (1 + residual), where the residual is
estimated from the following expanded market model regression:
Trade credit variables
TC is the ratio of accounts payable (ap) to total assets (at).
ABTC is abnormal trade payables, calculated as the residuals from regression of TC on a set
of determinants of trade credit, namely ln(1+age), ln(1+age)2, cash flow, cash holdings,
current assets, negative growth, positive growth, one-year maturity debt, and firm size.
Cash flow is operating income before depreciation (oibdp), less interest expense (xint),
less net income taxes (txt), less dividends (dvc), scaled by total assets (at). Cash holdings
is the ratio of cash and cash equivalents (che) to total assets (at). Current assets is current
assets (act) minus cash (che), divided by total assets (at). Negative (positive) growth is
sales growth times the negative (positive) growth dummy, which takes one if sales
growth is negative (positive), and zero otherwise. Short maturity debt is short-term
borrowings (np) plus the current portion of long-term debt (dd1), scaled by total assets
(at).
ABTCIND is the difference between trade credit (TC) and the median industry trade credit.
131
NETAP is the net account payables ratio, defined as the difference between accounts payable
(ap) and receivable (rect), divided by total assets (at).
Control and instrumental variables
MB is the market value of equity (csho×prcc_f) divided by the book value of equity (market-
to-book).
SIZE is the natural logarithm of total assets (at) at the end of the fiscal year.
ACCM is the three years moving average absolute value of discretionary accruals, where
discretionary accruals are estimated from the modified Jones model.
MS is monetary policy, measured as Federal Fund rates. Data source: Federal Reserve Bank
of St. Louis.
INVENTORY is total inventories (invt) divided by total assets (at).
132
ST3 is the ratio of debt in current liabilities (dlc) plus debt maturing in two or three years
(dd2+dd3) to total debt (the sum of debt in current liabilities plus long-term debt, i.e.,
dlc+dltt).
PAYOUT is the sum of common and preferred dividend and purchase (dvp+dvc+prstkc)
divided by total assets (at).
HPINDEX is the Hadlock-Pierce financial constraint index, developed by Hadlock and
Pierce (2010). HPINDEX is computed as: (−0.737×SIZE) + (0.043×SIZE2) –
(0.040×AGE), where SIZE is log of inflation-adjusted book assets and AGE is firm age.
KZINDEX is the Kaplan-Zingales financial constraint index. As presented by Lamont Polk,
and Saaá-Requejo (2001), KZINDEX is calculated as: −1.002×(Cashflow/K) +
0.283×(Q) + 3.139×(Debt/Capital) − 39.368×(Dividends/K) − 1.315×(Cash/K), where
K is beginning-of-year property, plant, and equipment (ppent); Cash flow is operating
income plus depreciation (ib+dp); Tobin’s Q is book assets minus book common equity
minus deferred taxes plus market equity (at−ceq−txdb+prcc_f×csho), divided by book
assets (at); Debt is short-term plus long-term debt (dltt+dlc); Dividends are total annual
dividend payments (dvc+dvp); Cash is cash plus marketable securities (che); and Capital
is debt plus total stockholders’ equity (dltt+dlc+seq).
WWINDEX is the White-Wu financial constraint index, developed by White and Wu (2006).
WWINDEX is computed as: −0.091×CF − 0.062×DIVPOS + 0.021×TLTD −
0.044×SIZE + 0.102×ISG − 0.035×SG, where CF is the ratio of cash flow to total assets;
DIVPOS is a dummy variable that takes one if the firm pays cash dividends, and zero
otherwise; TLTD is the ratio of the long-term debt to total assets; SIZE is the natural log
of total assets; ISG is the 3-digit industry sales growth; and SG is firm sales growth.
133
Conditioning variables
COVER is the number of analyst covering from I/B/E/S.
INST is the percentage of shares held by institutional owners, obtained from the Thomson
13F database.
TRAINST is the percentage of shares held by transient institutional investors.
TTLOAN is the log of one plus the amount of exiting total bank loans.
STLOAN is the log of one plus the amount of exiting bank loans maturing in three years.
LENGTH is the length of a bank loan, measured as number of years passed since loan
activation (i.e., the current year–the year of loan activation). Data source: DealScan.
SALETIND is the ratio of firm sales (sale) to median industry sales in year t.
STINPUT is standardized inputs index based on the industry classification in Rauch (1999),
defined as the share of inputs that comes from sectors producing standardized goods.
FITHHI is the fitted Herfindahl-Hirschman index, computed as sales-based Herfindahl ratios
for both public and private firms from each three-digit SIC code, developed by Hoberg
and Phillips (2010). A lower fitted Herfindahl ratio indicates a more competitive product
marketplace. Source: Hoberg-Phillips Data Library (http://hobergphillips.usc.edu/).
ZSCORE is Altman’s z-score, computed as: 1.2X1+1.4X2+3.3X3+0.6X4+1.0X5, where X1 is
working capital (act–lct)/total assets (at); X2 is retained earnings (re)/total assets (at); X3
is earnings before interest and taxes (oiadp)/total assets (at); X4 is market value equity
(prcc_f×csho)/book value of total liabilities (lt); and X5 is sales (sale)/total assets (at).
ADR is total assets (at) plus market value of equity (prcc_f×csho) minus book equity (ceq),
divided by total debt (dltt+dlc).
134
Chapter 4
Does bank deregulation affect stock price crash risk?
ABSTRACT
We examine the effect of the deregulation of bank branch restrictions on nonfinancial firms’
stock price crash risk. Most U.S. states lifted restrictions on intrastate branching from the
1970s to the 1990s, which improved bank monitoring and lending quality. We find robust
evidence that the bank deregulation leads to lower levels of firms’ stock price crash risk,
consistent with branch reform improving bank monitoring efficiency and enabling banks to
better constrain borrowers’ bad-news-hoarding behavior. This mitigating effect is more
pronounced for firms that are riskier and more reliant on external finance. Our study suggests
that bank branch deregulation is beneficial for protecting nonfinancial firms’ market value.
135
4.1. Introduction
During the last quarter of the twentieth century, most U.S. states removed branching
restrictions in the banking sector by allowing banks to open branches within and across state
borders. A large body of research has concluded that bank deregulation has significantly
changed regional banking market structures and lead to economic growth (e.g., Jayaratne
and Strahan, 1996, 1998; Berger, Demsetz, and Strahan, 1999; Kroszner and Strahan, 1999;
Black and Strahan 2002). For instance, the banking system becomes more integrated after
bank deregulation, which stabilizes economic growth (Morgan, Rime, and Strahan, 2004).
Moreover, bank branch reform mitigates income inequality by boosting incomes in the lower
part of the income distribution (Beck, Levine, and Levkov, 2010). Meanwhile, a growing
strand of literature focuses on several micro-level channels through which bank deregulation
affects economic activities and corporate behavior. For example, existing research has
investigated the influence of bank deregulation on corporate financing and investment
(Zarutskie, 2006; Rice and Strahan, 2010), entrepreneurship (Black and Strahan, 2002;
Ceterolli and Strahan, 2006), and corporate innovation (Chava, Oettl, Subramanian, and
Subramanian, 2013; Cornaggia, Mao, Tian, and Wolfe, 2015; Hombert and Matray, 2016).
However, it remains unknown whether and how such deregulation affects firm-specific
downside risk in the equity market. This study therefore attempts to fill this literature void
by investigating the impact of bank branch deregulation on nonfinancial firms’ stock price
crash risk.
Previous literature suggests that bad news hoarding by managers engenders sudden
and extreme declines in a firm’s stock price, hence increasing its stock price crash risk (Jin
and Myers, 2006; Bleck and Liu, 2007; Hutton, Marcus, and Tehranian, 2009; Kothari, Shu,
and Wysocki 2009; Kim, Li, and Zhang, 2011a, b). Managers who have privileged access to
136
the firm’s private information may have incentives to withhold unfavorable information
within the firm or opportunistically manage the timing of such information disclosure (Jin
and Myers 2006; Kim, Li, and Zhang, 2011a, b; Hong, Kim, and Welker, 2017). Although
managers can accumulate bad news for an extended period, they will likely reach a tipping
point, beyond which the cost of hoarding bad news exceeds the benefit of doing so. It is at
this point that the concealed negative information will be made public, leading to a sudden
collapse in stock price, namely a stock price crash (Kim, Li, and Zhang, 2011ab).
We argue that bank branch deregulation lowers stock price crash risk due to enhanced
bank monitoring efficiency. Bank lending is typically hindered by adverse selection and/or
moral hazard problems because borrowers are often reluctant in providing complete and
credible information to their lenders (e.g., Diamond, 1991; Ranjan and Winton, 1995; Bae
and Goyal, 2009; Qian and Strahan, 2008; Graham, Li, and Qiu, 2008). In order to overcome
debt contracting frictions and mitigate information asymmetries, lenders such as banks
usually engage in monitoring by collecting and analyzing information about borrowers’ risk
and profitability (Diamond, 1984; Strahan, 1999, 2017). We argue that the relaxation of
intrastate branching restrictions allows banks to monitor their borrowers more efficiently.
After the deregulation many small banks were acquired and incorporated as branches into
large bank holding companies (BHCs). Indeed, bank branch reform allowed new entry and
bank consolidation, providing an important selection mechanism to remove less efficient
banks (Jayaratne and Strahan, 1996, 1998; Strahan, 2003). As a result, local banking markets
became more concentrated while banks became larger after intrastate deregulation (Chava,
Oettl, Subramanian, and Subramanian, 2013). The increased bank size may help reduce the
costs of monitoring borrowers (Diamond, 1984; Allen, 1990; Strahan, 2017). For example,
Berger, Minnis, and Sutherland (2017) and Chen and Vashishtha (2017) show that large
banks are more cost-efficient in processing “hard” information about their clients. Given the
137
above, after intrastate branching reform, banks should be able to better monitor their
borrowers and prevent them from hiding negative information about their financial
performance. We thus predict that the passage of intrastate branching deregulation mitigates
nonfinancial firms’ future stock price crash risk.
We begin our analysis by estimating a difference-in-differences model to assess the
causal impact of bank branch deregulation on firm-specific stock price crash risk. The results
show a significant and negative association between intrastate deregulation and future stock
price crash risk, which is consistent with our prediction. The economic impact of bank
deregulation is also sizable. After intrastate branching restrictions were lifted, stock price
crash risk, proxied by conditional negative skewness (NCSKEW) and down-to-up volatility
(DUVOL) of firm-specific weekly returns, decreased by 9.25% and 8.91%, respectively. In
additional analysis, we follow Ceterolli and Strahan (2006) and include firm, state-by-year,
and industry-by-year fixed effects to control for unobserved heterogeneity at the firm, state,
industry, and year levels. The mitigating impact of intrastate branching deregulation on firms’
stock price crash risk remains unchanged under those specifications.
Although the staggered nature of bank branch deregulation acts as a plausible
exogenous shock to bank lending activities and thus helps to alleviate endogeneity concerns,
there might be state-level factors that varied across states and impacted the timings of the
deregulation events. If so, our results would be spurious and affected by reverse causality.
To rule out this potential concern and ensure that the parallel trends assumption is valid, we
follow Bertrand and Mullainathan (2003) and conduct a pre-reform trend analysis. We
examine the dynamics of stock price crash risk surrounding the deregulatory events and find
that there are no effects prior to bank deregulation.
Another potential endogeneity concern with our analysis is the presence of omitted
variables or unobserved shocks that might have coincided with bank branch deregulation
138
and, at the same time, could determine the changes in firm-level stock price crash risk. This
omitted-variable problem may invalidate our interpretation of the causal effect of bank
deregulation on stock price crash risk. To address such a concern, we follow Cornaggia, Mao,
Tian, and Wolfe (2015) and conduct placebo tests by randomly assigning states into each of
these deregulation years (without replacement) while maintaining the empirical distribution
of those years. If unobservable shocks related to firm-specific stock price crash risk occurred
simultaneously with the deregulation, then, despite the incorrect assignments of deregulatory
years to states, we would still observe a significant and negative relation between bank
deregulation and crash risk. However, the results of the falsification test indicate that these
counterfactual bank deregulatory events have no effects on stock price crash risk, suggesting
that the omitted-variable bias is unlikely to be a concern in our analysis.
To further verify the robustness of our findings, we control for another other form of
bank deregulation, namely interstate branching deregulation. The passage of interstate
deregulation laws allowed bank holding companies to freely enter other states and to operate
branches across state lines. The results show a significant and negative relation between
intrastate branching deregulation and stock price crash risk but an insignificant relation
between interstate branching deregulation and crash risk, consistent with the former type of
reform playing a more significant role in improving bank intermediation efficiency than the
latter (Calem, 1994; Jayaratne and Strahan, 1996). Next, we examine the robustness of our
results to different sample periods. Consistent with Cetorelli and Strahan (2006), the main
sample used in our main analysis ends in 1994 when bank deregulation was fully completed
with the passage of the Interstate Banking and Branching Efficiency Act (IBBEA). To ensure
all the states have sufficient observations in the post-reform periods, we extend our sample
to three, five, and 10 years after the passage of the IBBEA, and the results remain
qualitatively unchanged. Additionally, to eliminate potential survivorship bias, we require
139
that our sample firms have at least one year of data in both the pre- and post-deregulation
periods. Our results are insensitive to this restricted sample.
As a further robustness check, we employ alternative measures of stock price crash
risk and bank deregulation. Following prior research, such as Hutton, Marcus, and Tehranian
(2009), Kim, Li, and Zhang (2011b), and Chang, Chen, and Zolotoy (2016), we measure
stock price crash risk as the likelihood that a firm experiences more than one price crash
week in a fiscal year (CRASH), and the number of crashes minus the number of jumps over
the fiscal year (COUNT). In addition, following Black and Strahan (2002) and Hombert and
Matray (2016), we use a deregulation index (DERINDEX), rather than an indicator, to proxy
for intrastate branching deregulation. The results are qualitatively similar.
We then explore two underlying mechanisms through which intrastate branching
deregulation could affect stock price crash risk: external financial dependence and firm
riskiness. First, if the observed decrease in stock price crash risk following the passage of
bank deregulation is driven by improved bank monitoring, then the negative impact of the
bank branch reform on stock price crash risk is expected to be more noticeable among firms
with greater reliance on external finance. To the extent that those firms depend more on
external financing resources, they have stronger and more frequent interactions with banks
and therefore benefit more from enhanced bank monitoring. Using the external finance
dependence ratio, net change in capital, and asset tangibility as alternative proxies for
external financial dependence (Rajan and Zingales, 1998; Amore, Schneider, and Zaldokas,
2013; Almeida and Campello, 2007), we find that the negative impact of bank branch
deregulation on crash risk is more pronounced for firms that rely more on external finance.
Second, to the extent that banks can reduce information asymmetry by discerning their
borrowers’ riskiness and profitability, riskier firms are more likely to be monitored by their
lenders than less risky firms. Banks that become more cost-efficient after intrastate
140
branching deregulation should better curb borrowers’ risk-taking behavior. Thus, we expect
the impact of bank deregulation on stock price crash risk to be more pronounced among
riskier firms. Following Johnson (2003) and Callen and Fang (2015a, 2015b), we measure
firm risk as earnings volatility and financial leverage. The results indicate that the mitigating
effect of intrastate deregulation is more pronounced for firms with high earnings volatility
or financial leverage, consistent with our prediction.
This paper contributes to the literature in three ways. First, it adds to the literature on
the economic consequences of bank deregulation, in particular the effects of branch reform
on corporate policies of non-financial firms (e.g., Black and Strahan, 2002; Ceterolli and
Strahan, 2006; Zarutskie, 2006; Rice and Strahan, 2010; Chava, Oettl, Subramanian, and
Subramanian, 2013; Cornaggia, Mao, Tian, and Wolfe, 2015; Bai, Carvalho, and Phillips,
2015; Hombert and Matray, 2016). To our best knowledge, this paper is the first that
investigates the influence of bank branch reform on firm-specific stock price crash risk.
Jayaratne and Strahan (1996, 1997) suggest that bank efficiency improves substantially after
states permit statewide branching, evidenced by sharp declines in loan losses and increases
in loan quality. Our study complements their studies by highlighting that intrastate branching
deregulation can also protect shareholders’ wealth through enhanced bank monitoring.
Second, our study adds to the growing literature on stock price crash risk. Recent
research has documented a number of firm-specific determinants of crash risk, such as
financial reporting quality (Hutton, Marcus, and Tehranian, 2009; Kim, Li, Lu, and Yu, 2016;
Ertugrul, Lei, Qiu, and Wan, 2017; Kim and Zhang, 2016), equity-based executive
compensation (Kim, Li, and Zhang, 2011a; Xu, Li, Yuang, and Chan, 2014), tax avoidance
(Kim, Li, and Zhang, 2011b), and dividend policy (Kim, Luo, Xie, 2016). This literature has
also shown that crash risk is related to managerial characteristics, such as religiosity (Callen
and Fang, 2015a), managerial overconfidence (Kim, Wang, and Zhang, 2016), CEO age
141
(Andreou, Louca, and Petrou, 2016), among others. However, one major challenge of this
stream of research is that the determinants of stock price crash risk may be endogenously
linked with unobserved firm and managerial characteristics, making inference difficult.
Using the staggered passage of bank branch deregulation by different states at different
points in time allows us to establish the causal effect of a new factor, intrastate branching
reform, on nonfinancial firms’ crash risk.
Our study is related to a few recent studies of crash risk that have employed quasi-
natural experiment settings for identification purposes. For instance, Ali, Li, and Zhang
(2015) find that firms’ stock price crash risk is greater in states that have adopted the
Inevitable Disclosure Doctrine (IDD) than in non-adopting states. Balachandran, Duong,
Luong, and Nguyen (2018) document that the staggered passage of mergers and acquisitions
(M&A) laws in 32 countries increases the threat of takeover that disciplines managerial
misbehavior and leads to reduced stock price crash risk. Our study adds to this strand of
research by utilizing a novel quasi-natural experiment, namely, bank branch reform in the
banking industry.
Lastly, our paper is related to the literature linking bank monitoring with borrowers’
agency problem. Prior studies suggest various methods that banks might exploit to monitor
borrowers, such as using short-maturity debt, debt covenants, and debt contract
renegotiations (Rajan and Winton, 1995; Datta, Iskandar-Datta, and Raman, 2005; Roberts
and Sufi, 2009; Roberts, 2015). Our study contributes to this stream of research by showing
that intrastate branching deregulation, which was primarily designed to promote bank
efficiency and economic growth, has produced unintended, yet positive, consequences for
nonfinancial firms’ stock price dynamics. Our results are therefore of importance to a wide
range of stakeholders and provide relevant policy implications.
142
The rest of the paper proceeds as follows. Section 4.2 briefly reviews related
literature and develops the hypothesis. Section 4.3 discusses data and research design.
Section 4.4 presents the empirical results of the main analysis. We discuss the underlying
mechanisms of our baseline results in Section 4.5, and Section 4.6 concludes.
4.2. Literature review and hypotheses development
4.2.1. Bank branch deregulation and relevant literature
Traditionally, U.S. banks were subject to extensive regulations on geographical
expansion. The 1927 McFadden Act clarified the authority of the states over the regulation
of national banks’ branching activities within their borders. In most states bank holding
companies separately owned capitalized and licensed banks within state borders, while in
some states banks were typically allowed to run unit offices. Up to the 1970s, only 12 states
allowed unrestricted statewide branching. The other 38 states progressively relaxed their
branching restrictions between the 1970s and the passage of the IBBEA in 1994. Two classes
of branching restrictions were lifted in the 1970s through 1990s. First, states permitted
multibank holding companies (MBHCs) to convert subsidiary banks (existing or acquired)
into branches. MBHCs could then expand geographically by acquiring banks and converting
them into branches. Second, states permitted de novo branching, whereby banks could open
new branches anywhere within state borders. Table 1 depicts the years each state relaxed the
restrictions on bank branching.
Our study is related to the literature that examines the economic consequences of
deregulating bank branch restrictions. In an early study, Jayaratne and Strahan (1996)
suggest that intrastate branching deregulation significantly increases the rates of real per
capita growth in income and output. Following this study, a few papers have documented
143
additional evidence that intrastate deregulation is beneficial to the economy. For instance,
Black and Strahan (2002) show that following the deregulation the rate of new corporations
increases. Ceterolli and Strahan (2006) find that concentrated banking after branching
reform restrains potential entrants from gaining access to credit. Kerr and Nanda (2009)
document that branch banking deregulation brings about exceptional growth in both
entrepreneurship and business closures. Beck, Levine, and Levkov (2010) contend that bank
branch reform leads to the reduction in total income inequality by boosting the relative
demand for low-skilled workers. However, a recent study of Hombert and Matray (2016)
suggests that intrastate deregulation may result in unintended negative consequences.
Specifically, they find that the number of innovators decreases after bank deregulation
because the increase in competition for lending reduces financial constraints for firms in
more tangible sectors, but tightens financial constraints for small innovative firms. In a
similar vein, Chava, Oettl, Subramanian, and Subramanian (2013) show that intrastate
deregulation leads to less supply of credit and less innovation for young and private firms.
Overall, existing research on the role of bank deregulation in corporate policies
largely exploits bank branch reform as a regulatory shock to bank competition and credit
supply. However, although Jayaratne and Strahan (1996, 1997) argue that intrastate
branching deregulation exogenously changed bank monitoring mechanisms, few studies
have to date provided empirical evidence to that effect. Our study fills this gap by examining
whether bank deregulation reduces nonfinancial firms’ stock price crash risk via its
monitoring function.
We note that there was another form of bank branch deregulation in the U.S., i.e.,
interstate branching deregulation that allowed banks to expand across state borders.
Specifically, under this reform, states gradually lifted branching restrictions for bank holding
companies to expand beyond state boundaries. Both intrastate and interstate branching
144
deregulation was completed following the passage of the IBBEA of 1994. The literature
suggests that interstate deregulation affects state business cycles (Morgan, Rime, and
Strahan, 2004), bank competition and credit supply (Zarutskie, 2006; Rice and Strahan,
2010), as well as corporate innovation (Amore, Schneider, and Zaldokas, 2013; Cornaggia,
Mao, Tian, and Wolfe, 2015). However, in this paper we focus on intrastate deregulation,
rather than interstate deregulation. This is primarily because prior studies have suggested
that the latter type of branching reform has a limited impact on the structure of the banking
markets (e.g., Amel and Liang, 1992; Calem, 1994; Strahan, 2003), which, as argued above,
are important to bank monitoring and stock price crash risk. For instance, Jayaratne and
Strahan (1996) show that the deregulation of restrictions on geographic expansion beyond
state boundaries has little effect on the costs of intermediation. Moreover, we seek to isolate
the effect of an exogenous shock to bank monitoring without any systematic change in banks’
ability to diversify geographically. Given these arguments, our analysis focuses only on
intrastate branching deregulation. Nevertheless, in a robustness check we also control for the
effect of interstate branching deregulation.
4.2.2. Literature on stock price crash risk
Our study is also related to the literature investigating the determinants of firm-
specific stock price crash risk. In an early study, Chen, Hong, and Stein (2001) find that the
recent average monthly turnover and past returns can forecast future stock price crashes. Jin
and Myers (2006) then introduce an analytical model, in which stock price crashes occur
when managers’ accumulated bad news is revealed to the public at once. A key takeaway of
their study is that opaque stocks are more likely to crash than transparent ones.
Extant empirical literature has identified various internal and external determinants
of future stock price crash risk. In terms of the internal factors, Hutton, Marcus, and
Tehranian (2009) show that financial statement opacity leads to less information revelation
145
and more managerial bad-news-hoarding activities, hence higher crash risk. Zhu (2016) finds
a strong positive association between total accruals and stock price crashes, which is
consistent with the bad news withholding explanation. Kim, Li, and Zhang (2011a, 2011b)
document that equity incentives and corporate tax avoidance instruments can incentivize
managers to purposely withhold negative information, leading to a higher likelihood of
future stock price crashes. Using Chinese data, Xu, Li, Yuan, and Chan (2014) find that
excess perks in state-owned enterprises (SOEs) encourage managers to withhold bad news
for personal interest, resulting in high crash risk. Kim, Li, and Li (2014) show that firms
with superior corporate social responsibility (CSR) performance exhibit lower levels of crash
risk. Likewise, Zhang, Xie, and Xu (2016) examine the effect of firms’ philanthropic action
on their stock price crash risk and document a negative relation between corporate
philanthropy and crash risk. Kim and Zhang (2015) find that firms with more conservative
accounting numbers are less likely to hide bad news, which results in lower crash risk. Kim, Li,
Lu, and Yu (2016) reveal that financial statement comparability attenuates crash risk, because
greater financial statement comparability gives investors better access to negative information
about a firm through analyzing its comparable peer firms, thus restricting managers from
withholding bad news.
Some studies examine the role of external corporate governance mechanisms in
mitigating crash risk. For example, An and Zhang (2013) show that stock holdings by
dedicated institutional investors reduce stock price crash risk whereas those by transient
institutional investors increase crash risk. Callen and Fang (2013) also suggest that
institutional investor stability is negatively associated with crash risk. In addition, auditors
can act as an external corporate governance mechanism that mitigates stock price crash risk.
Robin and Zhang (2015) find that industry-specialist auditors play an important role in
reducing managerial incentives to manipulate earnings, leading to lower stock price crash
146
risk. Similarly, Callen and Fang (2016) document that auditors develop client-specific
knowledge over the auditor-client relationship, which enables them to deter clients’ bad
news hoarding behavior and their stock price crash risk. In a similar vein, Kubick and
Lockhart (2016) document that firms located closer to the Securities and Exchange Commission
(SEC) have lower stock price crash risk because SEC oversight induces disclosure practices that
reduce the likelihood of financial misreporting.
Our study adds to the literature reviewed above by examining whether firms’ stock
price crash risk is influenced by bank deregulation, which is a plausibly exogenous
regulatory shock in the banking industry. Importantly, our results reveal that the structural
change in the banking industry, as a consequence of bank branch reform, not only affects
debtholders’ interest, but also protects shareholders’ wealth.
4.2.3. Hypothesis
Jayaratne and Strahan (1997) suggest that branching restrictions retarded the “natural”
evolution of the banking industry by preventing better-run banks from establishing branches.
Once those branching restrictions were removed, banks were able to acquire their peers and
convert them into branches, or were permitted entry via de novo branching within state
borders (McLaughlin, 1995; Jayaratne and Strahan, 1996; Rice and Strahan, 2010). Indeed,
Calem (1994) and Strahan (2002) show that after branching reform large banks acquired
many small banks. These entry and consolidation activities play an important role in
removing less efficient banks and improving bank performance (Jayaratne and Strahan, 1996;
Kroszner and Strahan, 1997, 1999). As a result, banks can better distinguish promising
projects from bad ones, and effectively monitor borrowers after lending relationships are
established (Jayaratne and Strahan, 1997).
To the extent that banks greatly intensified their merger and acquisition (M&A)
activities following the deregulation reforms, the banking sector became more concentrated
147
with lenders having greater bargaining power over borrowers (Jayaratne and Strahan, 1997;
Chava, Oettl, Subramanian, and Subramanian, 2013). Prior studies suggest that banking
concentration can fundamentally change the nature of banks’ monitoring by altering their
organizational structures (e.g., Diamond, 1984; Stein, 2002; Berger, Miller, Petersen, Rajan,
and Stein, 2005). Specifically, M&A activities in the banking industry significantly increase
banking concentration and result in larger, more complex, and more hierarchical banking
organizations, which are able to process borrowers’ “hard” information22 less costly (Berger,
Minnis, and Sutherland, 2017; Chen and Vashishtha, 2017). Theoretical studies show that
complex and hierarchical organizations rely more on “hard” information because such
organizations face higher costs in motivating agents to produce and truthfully present “soft”
information (Stein, 2002; Chen and Vashishtha, 2017). Lifting branching restrictions leads
to higher efficiency in bank monitoring, especially in processing borrowers’ “hard”
information. Thus, we expect that, after intrastate deregulation, banks can better monitor
their borrowers and deter them from withholding negative information, hence reducing their
stock price crash risk. We thus formulate our central hypothesis as follows:
Hypothesis: The passage of intrastate branching deregulation reduces nonfinancial
firms’ future stock price crash risk.
4.3. Data and methodology
4.3.1. Sample selection
We construct a panel dataset of U.S. public firms from 1962 to 1994. Our sample
starts from 1962 because this is the first year that corporate financial data became available
22
Firms’ “hard” information is quantitative and independent of the collection process, while “soft” information
is often communicated in text and difficult to verify (Petersen, 2004). Collecting “soft” information is more
costly and less efficient due to the requirement for lenders to build a close relationship with borrowers (Strahan,
2017).
148
on the COMPUSTAT file. Consistent with Cetorelli and Strahan (2006), our sample ends in
1994 when the deregulation of branching restrictions was completed with the passage of the
IBBEA. Cetorelli and Strahan (2006) argue that it becomes increasingly less plausible to
view markets in banking as local after 1994, because of the completion of deregulation as
well as the fact that new technologies have allowed banks to lend to borrowers not physically
close to their banks.
Our crash risk measures are calculated based on weekly return data from the Centre
for Research in Security Prices (CRSP) database. The control variables are computed using
firms’ annual financial statement data from COMPUSTAT. Following prior studies (e.g.,
Hutton, Marcus, and Tehranian, 2009; Kim, Li, and Zhang, 2011a, 2011b; Kim, Li, Lu, and
Yu, 2016; Chang, Chen, and Zolotoy, 2017), we exclude firms with year-end share prices
below $1, firms with fewer than 26 weeks of stock return data in fiscal years, firm-year
observations with negative total assets and book values of equity, and firm-year observations
with insufficient financial data to calculate relevant variables. Our final sample comprises
61,784 firm-year observations (7,190 unique firms).
4.3.2. Measuring bank branch deregulation
Consistent with Jayaratne and Strahan (1996), we choose the date of branch
deregulation as the date on which a state permitted branching via M&A through the holding
company structure or de novo branching. The bank branch deregulation variable (BRANCH)
is a dummy variable that equals one if a state has implemented intrastate branching
deregulation and zero otherwise. As mentioned, Table 1 shows the years of bank branch
deregulation on a state-by-state basis.23
23
Following Jayaratne and Strahan (1996) and Beck, Levine, and Levkov (2010), we confirm the robustness
of the empirical results by dropping Delaware and South Dakota as their banking systems were heavily affected
by laws that provided a tax incentive for credit card banks to operate. For example, during the mid-1980s the
149
We follow Hutton, Marcus, and Tehranian (2009) and calculate firm-specific weekly
returns by estimating the following equation:
where rj,τ is the weekly return on stock j in week τ, rm,τ is the return on CRSP value-weighted
market index, and ri,τ is the Fama and French value-weighted industry index in week τ. The
lead and lag terms of the market and industry returns are included to account for
nonsynchronous trading (Dimson, 1979). We use weekly returns to avoid the concern caused
by thinly traded stocksand estimate weekly returns from Wednesday to Wednesday to avoid
any contaminating effects from weekends and Mondays (Wang, Li, and Erickson, 1997;
Bartholdy and Peare, 2005). The firm-specific weekly return (Wj,τ) is calculated as the log
value of one plus the residual return from Eq. (1).
We then follow Chen, Hong, and Stein (2001) and Kim, Li, and Zhang (2011a, b)
and calculate our primary measure of stock price crash risk, negative conditional skewness
(NCSKEW), as negative of the third moment of each stock’s firm-specific weekly returns
divided by the standard deviation raised to the third power. For firm j in fiscal year t, this
measure is defined as
3 2 3/2
where n is the number of observations of weekly returns in fiscal year t. Firms with high
NCSKEW are more likely to experience a stock price crash.
Our second measure of firm-specific crash risk is “down-to-up volatility” (DUVOL),
which is calculated as follows:
2 2
banking industry in those states expanded quickly and contributed significantly more to economic growth than
the banking system in other states.
150
where nu and nd are the number of up and down weeks over the fiscal year t, respectively.
For each stock j over fiscal year t, we partition all firm-specific weekly returns into down
(up) weeks when the weekly returns are below (above) the annual mean. We then calculate
the standard deviation of firm-specific weekly returns for each group separately. DUVOL is
the log ratio of the standard deviation in the down weeks to the standard deviation in the up
weeks. A higher value of DUVOL corresponds to a stock being more “crash prone.” This
alternative measure of crash risk may be less likely to be excessively affected by a small
number of extreme returns as it does not involve the third moments (Chen, Hong, and Stein,
2001).
Following prior literature (e.g., Chen, Hong, and Stein, 2001; Jin and Myers, 2006),
we include several control variables: DTURNt, the difference between the average monthly
share turnover over fiscal year t and t–1; SIGMAt, the standard deviation of firm-specific
weekly returns over fiscal year t; RETt, the average firm-specific weekly returns over fiscal
year t; SIZEt, the log of market value of equity at the end of fiscal year t; MBt, the market
value of equity divided by the book value of equity at the end of fiscal year t; LEVt, the book
value of total debt scaled by total assets at the end of fiscal year t; ROAt, income before
extraordinary items divided by total assets at the end of fiscal year t; NCSKEWt, the negative
conditional skewness for firm-specific weekly returns in fiscal year t.24 Appendix 1 provides
the definitions of all variables used in this study.
24
We do not control for discretionary accruals in the baseline regressions because this variable requires
information from the cash flow statements, which were only available after 1988. Including this control
variable in the main analysis will substantially reduce our sample size (Hutton, Marcus, and Tehranian, 2009).
Nevertheless, in a robustness check (untabulated) we control for discretionary accruals and obtain qualitatively
similar results.
151
Table 2 presents the summary statistics of all variables used in our regressions. For
the two price crash risk measures, NCSKEWt+1 and DUVOLt+1, their mean values are –0.270
and –0.153, respectively. These figures are quite close to those reported in previous research
(e.g., Chen, Hong, and Stein, 2001; Callen and Fang, 2015a).25 The mean of the bank branch
deregulation indicator is 0.54, similar to that reported by Cetorelli and Strahan (2006). The
summary statistics of the control variables are largely in line with those reported in prior
studies, and thus are not discussed herein to preserve space.
4.4.2. Baseline specification and results
Our main econometric model focuses on the relationship between bank branch
deregulation and firm-specific stock price crash risk. The empirical specification we estimate
is as follows:
𝐶𝑟𝑎𝑠ℎ 𝑅𝑖𝑠𝑘𝑗,𝑡+1 = 𝛽0 + 𝛽1 𝐵𝑅𝐴𝑁𝐶𝐻𝑗,𝑡 + 𝛽2 𝐷𝑇𝑈𝑅𝑁𝑗,𝑡 + 𝛽3 𝑆𝐼𝐺𝑀𝐴𝑗,𝑡 + 𝛽4 𝑅𝐸𝑇𝑗,𝑡 +
𝛽5 𝑆𝐼𝑍𝐸𝑗,𝑡 + 𝛽6 𝑀𝐵𝑗,𝑡 + 𝛽7 𝐿𝐸𝑉𝑗,𝑡 + 𝛽8 𝑅𝑂𝐴𝑗,𝑡 + 𝛽9 𝑁𝐶𝑆𝐾𝐸𝑊𝑗,𝑡 +
𝑌𝑒𝑎𝑟𝑡 + 𝑆𝑡𝑎𝑡𝑒𝑖 + 𝜀𝑗,𝑡 , (4)
where the dependent variable Crash Riskt+1 is measured by NCSKEW or DUVOL in year t+1
and all independent variables are measured in year t. The independent variable of interest is
the bank branch deregulation indicator (BRANCHt). We control for year and state fixed
effects and cluster standard errors by state in our baseline regressions. Including state fixed
effects helps address the concern that (unobservable) time-invariant omitted variables that
25
In Chen, Hong, and Stein (2001), the mean values for NCSKEW and DUVOL are 0.262 and 0.190,
respectively. In Callen and Fang (2015a), these values are −0.226 and −0.211.
152
generate variation in a state’s stance toward openness to bank branching might be
simultaneously correlated with the stock price crash risk of firms in the state.
Since our paper exploits the staggered introduction of bank branch deregulation
across states, the specification we use is a generalized difference-in-differences model. The
effect of bank branch deregulation is estimated as the difference between the change in stock
price crash risk before and after deregulation, with the change in crash risk being the
difference between the treatment firms and the control group. Under this specification,
control firms are those that do not experience a change in their deregulation status. For
example, a firm in the pre-deregulation periods serves as its own control group for the same
firm in the post-reform periods.
Table 3 presents the baseline results from the OLS regression analysis. In the first
two columns, we regress crash risk, NCSKEWt+1 or DUVOLt+1, on the bank branch
deregulation dummy variable, BRANCHt, without any control variables but with year and
state fixed effects. The results show that the coefficients on BRANCH are significantly
negative (t-stat = –2.25 in Col. (1) and –2.59 in Col. (2)). We further control for a set of
crash risk determinants in the remaining columns. The results show that the coefficients on
BRANCH are still significantly negative for both crash risk measures (t-stat = –2.95 in Col.
(3) and –3.33 in Col. (4)). This suggests that intrastate branching deregulation reduces stock
price crash risk for nonfinancial firms, consistent with our hypothesis that the improved
quality of bank monitoring over their borrowers after bank deregulation allows banks to
better restrict borrower from hiding bad news.
We further evaluate the economic significance of the effect of bank branch
deregulation on firms’ future crash risk. Based on the results from the comprehensive
specification in Columns (3) and (4) of Table 3, when BRANCH changes from 0 to 1, there
is a decrease of 9.25% (8.91%) in NCSKEW (DUVOL) from their sample mean, respectively.
153
This finding suggests that the impact of intrastate deregulation on firms’ stock price crash
risk is not only statistically significant but also economically important.
Turning to the control variables, we find that the coefficients on firm size (SIZE),
market-to-book ratio (MB), profitability (ROA), and lagged crash risk (NCSKEW) are
significant and positive, consistent with the evidence documented in prior studies (e.g., Chen,
Hong, and Stein, 2001; Hutton, Marcus, and Tehranian, 2009; Callen and Fang, 2015a).
To assess the robustness of the results in Table 3, we include additional fixed effects
and present the results in Table 4. We first include firm fixed effects and repeat the main
regression analysis. The coefficients on bank deregulation BRANCH in Columns (1) to (4)
remain negative and significant at least at the 10% level, regardless of the presence of the
control variables. Moreover, following Cetorelli and Strahan (2006), we include state-by-
year and industry-by-year fixed effects to absorb any potential confounding factors (i.e.,
time-varying state and industry heterogeneity) that could invalidate our identification. The
results presented in the last two columns of Table 4 continue to show a statistically
significant and negative association between BRANCHt and future stock price crash risk
(NCSKEWt+1 or DUVOLt+1). Overall, we find robust evidence of the role of intrastate branch
reform in curbing managers’ bad news hoarding behavior and firms’ crash risk, consistent
with our main hypothesis.
4.4.3. Endogeneity tests
Our identification is based on the idea that the staggered deregulation of bank
branching laws can represent an exogenous shock to bank monitoring effectiveness, thus
affecting firms’ stock price crash risk. However, one concern with our setting is that,
although we have controlled for state fixed effects in the main specification, there may
154
remain omitted state-level factors that could potentially trigger the deregulation in different
states. Under this scenario, there might be a reverse causality problem if the states differ in
their firms’ stock price crash risk and such variation further affects the timing of bank branch
deregulation in each state. Following Bertrand and Mullainathan (2003) and Cornaggia, Mao,
Tian, and Wolfe (2015), we address the possible reverse causality concern by investigating
the dynamic trends of stock price crash risk surrounding the deregulatory events. If reserve
causality indeed exists, we should observe significant changes in stock price crash risk prior
to the deregulatory events.
To test the pre-existing trends in stock price crash risk, we require the sample firms
to experience branch bank reform. Following Cornaggia, Mao, Tian, and Wolfe (2015), we
construct four dummy variables indicating four periods around the deregulation: Before2+,
Before1, After1, and After2+. Before2+ takes one for observations up to, and including, two
years prior to deregulation; Before1 takes one for one year prior to deregulation; After1 takes
one for one year post-deregulation; and After2+ takes one for two years or more post-
deregulation. Specifically, we estimate the following model:

2+ 1 1 2+
𝐶𝑟𝑎𝑠ℎ 𝑅𝑖𝑠𝑘𝑗,𝑡+1 = 𝛽0 + 𝛽1 𝐵𝑒𝑓𝑜𝑟𝑒𝑖,𝑡 + 𝛽2 𝐵𝑒𝑓𝑜𝑟𝑒𝑖,𝑡 + 𝛽3 𝐴𝑓𝑡𝑒𝑟𝑖,𝑡 + 𝛽4 𝐴𝑓𝑡𝑒𝑟𝑖,𝑡 +
𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠𝑗,𝑡 + 𝑌𝑒𝑎𝑟𝑡 + 𝑆𝑡𝑎𝑡𝑒𝑖 + 𝜀𝑗,𝑡 . (5)
Table 5 presents the estimation results. In Columns (1) and (2), we regress both
measures of crash risk on the four period indicators without the controls but with year and
state fixed effects. The coefficients on Before2+ and Before1 are not significant, suggesting
that stock price crash risk experiences no significant change prior to brank branch
deregulation. The coefficients on After1 and After2+ are significantly negative, which is
consistent with the baseline findings. In Columns (3) and (4), we include all control variables
and obtain similar results. Overall, the results reported in Table 5 suggest that before
intrastate branching deregulation, there is no pre-existing trend in firms’ stock price crash
155
risk. These results help validate the important assumption about parallel trends and mitigate
the concern about reverse causality.
Another source of endogeneity that would adversely affect our identification strategy
is the existence of potential omitted unobservable shocks that occur at approximately the
same time as state-level bank branch deregulation events. To address this concern, we follow
Cornaggia, Mao, Tian, and Wolfe (2015) and conduct placebo tests by requiring the
deregulation events to happen in years other than the actual deregulatory years. Specifically,
we randomly assign each state into a different deregulation year following the empirical
distribution of years (see Table 1) while allowing the distribution of deregulatory years to
be consistent with our baseline specification but disrupting the proper assignment of
deregulation years to states. If unobservable shocks related to firms’ crash risk exist and
coincide with the deregulation events, they should potentially drive the baseline findings.
Otherwise, the estimation results should be weakened by the random assignments of
deregulatory years to states. Table 6 reports the results of our placebo tests. The coefficients
on BRANCH are statistically insignificant across all columns, suggesting that there are no
unobservable shocks coinciding with bank branch deregulation.
Overall, our tests for reverse causality and omitted variables support the notion that
bank branch deregulation has a causal, negative effect on firm-specific stock price crash risk.
4.4.4. Other robustness tests
As a further robustness check, we include an alternative form of bank deregulation
in the main model. Some studies have considered the implications of both intrastate and
interstate bank branching deregulation; the latter allows banks to establish branches across
states (Rice and Strahan, 2010; Amore, Schneider, and Zaldokas, 2013; Chava, Oettl,
156
Subramanian, and Subramanian, 2013). We thus include an interstate branching deregulation
indicator (INTER), which equals one after a state implemented interstate deregulation and
otherwise zero. As can be seen from Panel A of Table 7, our main results regarding the
impact of intrastate deregulation continue to hold after controlling for interstate deregulation.
On the other hand, the impact of interstate deregulation is insignificant, consistent with our
early argument that this form of branching reform is less relevant for our analysis of bank
monitoring and firms’ stock price crash risk.
Next, we attempt to examine whether our baseline findings remain robust after the
U.S. Congress passed another form of deregulation, namely the IBBEA. To this end, we
expand our sample to 5 years following the passage of the IBBEA and present the regression
results in Panel B of Table 7. 26 The estimation results show significant and negative
coefficients on BRANCH, which lends further support for our hypothesis. Lastly, we repeat
the estimation of Eq. (4) only for firms that exist both before and after the deregulatory event
years. Such tests isolate firms that did not experience the shock of bank deregulation and
allow us to better capture the difference between control group (pre-deregulation) and
treatment group (post-deregulation). As shown in Panel C of Table 7, the results are
consistent with our baseline findings that intrastate branching deregulation reduces firm-
level stock price crash risk.
4.4.5. Robustness tests: alternative measures
Following Hutton, Marcus, and Tehranian (2009), Kim, Li, and Zhang (2011b), and
Chang, Chen, and Zolotoy (2017), we further measure future stock price crash risk as the
likelihood that a firm experiences more than one price crash week in a fiscal year (CRASH).
26
We also expand the sample to three or 10 years after the passage of the IBBEA. The results reported in
Appendix 2 are still consistent with the main findings.
157
Specifically, we define crash weeks in a given fiscal year as those during which a firm
experiences firm-specific weekly returns 3.09 standard deviations below the mean weekly
returns over the whole fiscal year, with 3.09 chosen to generate a frequency of 0.1% in the
normal distribution. As in Jin and Myers (2006) and Callen and Fang (2015a), we also
measure stock price crash risk as the number of crashes minus the number of jumps over the
fiscal year (COUNT). Likewise, we define those weeks as jump weeks when the firm-
specific weekly return is 3.09 standard deviations above its mean in a fiscal year. We present
the OLS estimation results in Panel A of Table 8.27 In line with our main findings, BRANCH
are significantly and negatively related to both CRASH and COUNT.
We have thus far defined bank branch deregulation as a dummy variable. In this test,
we follow Black and Strahan (2001) and Hombert and Matray (2016) and employ a
continuous measure, an intrastate bank deregulation index (DERINDEX). As mentioned,
starting in 1970, all states progressively lifted restrictions on branching within their borders.
They generally relaxed restrictions on within-state bank expansion in three steps: first,
permitting the formation of multibank holding companies; then, permitting branching by
means of mergers and acquisitions (M&As) only; and finally, permitting unrestricted (de
novo) branching, thereby allowing banks to enter markets by opening new branches. We
define the deregulation index to be zero if a state did not permit branching via any of the
three approaches; otherwise, the index equals the sum of the number of ways that banks may
expand within a state. Hence, the value of the deregulation index (DERINDEX) ranges from
zero (full regulation) to three (full deregulation). The results in Panel B of Table 8 show that
27
We use the linear probability model because prior research shows that non-linear logit or probit models with
fixed effects can produce biased coefficient estimates due to the incidental parameters problem (e.g., Neyman
and Scott, 1948; Lancaster, 2000; Greene, 2004).
158
the coefficients on deregulation index (DERINDEX) are significant and negative for both
crash risk measures.
Overall, we conclude that our main findings are robust to a variety of sensitivity tests,
such as controlling for interstate deregulation, using expanded test windows, and employing
alternative measures of future stock price crash risk and intrastate branching deregulation.
4.5. Mechanisms
Thus far we have shown a robust negative effect of bank branch deregulation on
stock price crash risk. In this section, we explore two underlying mechanisms through which
the branch banking reform affects crash risk: external financial dependence and firm risk.
4.5.1. External financial dependence
As argued above, lifting intrastate branching restrictions significantly changed the
structure of banking markets (Jayaratne and Strahan, 1998). The banking industry became
more concentrated and small banks lost their market share after branching reform. Post-
deregulation local banks became larger and more complex, thus being more efficient at
processing “hard” information (Berger, Miller, Petersen, Rajan, and Stein, 2005; Chen and
Vashishtha, 2017). Meanwhile, firms that rely heavily on external financing are likely to
borrow more from larger banks after branch deregulation and undertake more efficient
monitoring. Thus, if bank deregulation reduces stock price crash risk through improved bank
monitoring, we expect that firms with greater reliance on external finance experience more
effective bank monitoring, resulting in a decline in stock price crash risk.
Empirically, we employ three proxies to measure the degree of external financial
dependence: the external finance dependence ratio (EXDEP), net change in capital (NCC),
and asset tangibility (TANG). Following Rajan and Zingales (1998), we define a firm’s
159
external finance dependence as the amount of desired investment that cannot be financed
through internal sources. Accordingly, the external finance dependence ratio (EXDEP) is
calculated as investment plus R&D expenses and acquisitions minus operating income
before depreciation, divided by investment. Following Amore, Schneider, and Zaldokas
(2013) and Frank and Goyal (2003), we compute net change in capital (NCC) as net change
in equity and debt normalized by total assets. Both measures reflect firms’ demand for
finance and sensitivity to credit supply shock. The third measure, asset tangibility (TANG),
is calculated as the proportion of tangible assets to total assets. Asset tangibility also reflects
the sensitivity of firms to financing frictions as firms can access bank credit using tangible
assets as collateral (Almeida and Campello, 2007).
In Panel A of Table 9, we partition the full sample based on these variables and
define more financially dependent firms as those with above-the-median external finance
dependence in year t. The results show that, for both crash risk measures, the coefficients
on the bank branch deregulation indicator are only significantly negative for firms with high
external finance dependence. However, for firms with low external finance dependence, the
coefficients on BRANCH become insignificant. Panel B presents the results based on net
change in capital. Similar to the results in Panel A, we find that the coefficients on BRANCH
are only statistically significant for firms with greater reliance on external finance but
insignificant for firms with less reliance on external finance. Panel C reports the results based
on asset tangibility. As in the first two panels, the coefficients on BRANCH are negative and
significant only for firms with greater reliance on external financing sources. Overall, our
empirical findings are in line with the conjecture that the negative effect of bank deregulation
on crash risk is stronger when firms have higher levels of external financial dependence. In
other words, external financial dependence appears to be an important underlying
160
mechanism through which branch deregulation affects nonfinancial firms’ stock price crash
risk.
4.5.2. Firm risk
Next, we examine whether firms’ riskiness affects the relation between intrastate
branching deregulation and future stock price crash risk. Demsetz and Lehn (1985) suggest
that managers of riskier firms are more difficult to monitor. Kim, Li, and Zhang (2011b) and
Callen and Fang (2015a) contend that managers are more likely to conceal information about
excessive risk-taking behavior in order to support share price. To the extent that banks can
better monitor borrowers following the deregulation, which reduces the potential moral
hazard concern, borrowers with higher levels of riskiness would benefit more from the
enhanced monitoring and exhibit lower levels of price crash risk.
To empirically examine this channel, we follow Johnson (2003) and Callen and Fang
(2015a, 2015b) and use two proxies for firm riskiness: earnings volatility (VOL) and
financial leverage (LEV). We classify a firm as being riskier if its earnings volatility or
leverage is above the sample median. Panel A of Table 10 shows that the coefficients on
BRANCH are significantly negative only for firms with high earnings volatility but
insignificant or marginally significant for those with low earnings volatility. Likewise, the
results in Panel B show that the coefficients on BRANCH are significantly negative among
highly levered firms but insignificant for those with low leverage.
Overall, the evidence reported in this section shows that, after intrastate branching
deregulation, banks can better monitor firms with more risk-taking behavior. As a
consequence, these firms have less bad news hoarding and lower future stock price crash
risk. Put differently, firm risk appears to be another mechanism through which branch
deregulation impacts stock price crash risk.
161
4.6. Conclusion
This study investigates the impact of bank deregulation on nonfinancial firms’ stock
price crash risk. We find that the staggered passage of state-level intrastate branching
deregulation laws leads to lower levels of firm-specific stock price crash risk. This finding
remains robust after addressing potential endogeneity concerns about reverse causality and
omitted variable bias. Our empirical results are also insensitive to the inclusion of interstate
deregulation, as well as the use of alternative event windows and various measures of key
variables. Overall, those results support the argument that after branching reform banks are
able to monitor their borrowers more effectively and prevent them from withholding bad
news.
We further analyze the potential underlying mechanisms through which bank branch
deregulation may affect firms’ stock price crash risk. Specifically, our results show that the
negative relation between branching reform and crash risk is more pronounced for firms that
are more reliant on external finance and are risker.
Overall, our study contributes to research on bank deregulation and stock price crash
risk. The financial economics literature provides robust evidence that the deregulation in the
banking industry is beneficial to economic growth. A growing number of studies extend this
view by providing firm-level evidence (e.g., Black and Strahan, 2002; Ceterolli and Strahan,
2006; Amore, Schneider, and Zaldokas, 2013; Chava, Oettl, Subramanian, and Subramanian,
2013; Hombert and Matray, 2016). Our paper complements the literature by documenting
new evidence that intrastate bank deregulation reduces nonfinancial firms’ stock price crash
risk. Our finding extends the view of Jayaratne and Strahan (1996) that the key to the
beneficial growth effects of bank branch reform is the improvement in lending quality. We
162
show that deregulatory policies in the banking industry are beneficial to borrowers through
not only potentially lower loan rates and greater access to credit supply, but also the
protection of shareholders’ wealth.
163
References
Acharya, V.V., Imbs, J., and Sturgess, J., 2010. Finance and Efficiency: Do Bank Branching
Regulations Matter?. Review of Finance, 15(1), 135–172.
Ali, A., Li, N., and Zhang, W., 2015. Managers’ career concerns and asymmetric disclosure
of bad versus good news. Working paper.
Allen, F., 1990. The market for information and the origin of financial intermediation.
Journal of Financial Intermediation, 1(1), 3–30.
Almeida, H., and Campello, M., 2007. Financial constraints, asset tangibility, and corporate
investment. Review of Financial Studies, 20(5), 1429–1460.
Amel, D.F., and Liang, J.N., 1992. The relationship between entry into banking markets and
changes in legal restrictions on entry. The Antitrust Bulletin, 37(3), 631–649.
Amore, M.D., Schneider, C., and Žaldokas, A., 2013. Credit supply and corporate innovation.
An, H., and Zhang, T., 2013. Stock price synchronicity, crash risk, and institutional investors.
Journal of Corporate Finance, 21, 1–15.
Andreou, P.C., Louca, C., and Petrou, A.P., 2016. CEO age and stock price crash risk. Review
of Finance, 21(3), 1287–1325.
Bae, K.H., and Goyal, V., 2009. Creditor rights, enforcement, and bank loans. Journal of
Finance, 64(2), 823–860.
Bai, J.J., Carvalho, D., and Phillips, G.M., 2017. The impact of bank credit on labor
reallocation and aggregate industry productivity (No. w24081). National Bureau of
Economic Research.
164
Balachandran, B., Duong, H.N., Luong, H., and Nguyen, L., 2018. M&A laws and stock
price crash risk. Working paper.
Bartholdy, J., and Peare, P., 2005. Estimation of expected return: CAPM vs. Fama and
French. International Review of Financial Analysis, 14(4), 407–427.
Beck, T., Levine, R., and Levkov, A., 2010. Big bad banks? The winners and losers from
bank deregulation in the United States. Journal of Finance, 65(5), 1637–1667.
Berger, A.N., Demsetz, R.S., and Strahan, P.E., 1999. The consolidation of the financial
services industry: Causes, consequences, and implications for the future. Journal of
Banking & Finance, 23(2–4), 135–194.
Berger, A.N., Miller, N.H., Petersen, M.A., Rajan, R.G., and Stein, J.C., 2005. Does function
follow organizational form? Evidence from the lending practices of large and small
banks. Journal of Financial Economics, 76(2), 237–269.
Berger, P.G., Minnis, M., and Sutherland, A., 2017. Commercial lending concentration and
bank expertise: Evidence from borrower financial statements. Journal of Accounting and
Economics, 64(2–3), 253–277.
Bertrand, M., and Mullainathan, S., 2003. Enjoying the quiet life? Corporate governance and
managerial preferences. Journal of Political Economy, 111(5), 1043–1075.
Black, S.E., and Strahan, P.E., 2002. Entrepreneurship and bank credit availability. Journal
of Finance, 57(6), 2807–2833.
Bleck, A., and Liu, X., 2007. Market transparency and the accounting regime. Journal of
Accounting Research, 45(2), 229–256.
Brealey, R., Leland, H.E., and Pyle, D.H., 1977. Informational asymmetries, financial
structure, and financial intermediation. Journal of Finance, 32(2), 371–387.
165
Calem, P., 1994. The impact of geographic deregulation on small banks. Business Review,
17–31.
Callen, J.L., and Fang, X., 2013. Institutional investor stability and crash risk: Monitoring
versus short-termism?. Journal of Banking & Finance, 37(8), 3047–3063.
Callen, J.L., and Fang, X., 2015a. Religion and stock price crash risk. Journal of Financial
and Quantitative Analysis, 50(1–2), 169–195.
Callen, J.L., and Fang, X., 2015b. Short interest and stock price crash risk. Journal of
Banking & Finance, 60, 181–194.
Callen, J.L., and Fang, X., 2016. Crash risk and the auditor-client relationship.
Contemporary Accounting Research, 34: 1715–1750
Cetorelli, N., and Strahan, P.E., 2006. Finance as a barrier to entry: Bank competition and
industry structure in local US markets. Journal of Finance, 61(1), 437–461.
Chang, X., Chen, Y., and Zolotoy, L., 2017. Stock liquidity and stock price crash risk.
Chava, S., Oettl, A., Subramanian, A., and Subramanian, K.V., 2013. Banking deregulation
and innovation. Journal of Financial Economics, 109(3), 759–774.
381.
Chen, Q., and Vashishtha, R., 2017. The effects of bank mergers on corporate information
disclosure. Journal of Accounting and Economics, 64(1), 56–77.
166
Cornaggia, J., Mao, Y., Tian, X., and Wolfe, B., 2015. Does banking competition affect
innovation?. Journal of Financial Economics, 115(1), 189–209.
crash risk?. The Accounting Review, 90(1), 265–299.
Demsetz, H., and Lehn, K., 1985. The structure of corporate ownership: Causes and
consequences. Journal of Political Economy, 93(6), 1155–1177.
Diamond, D.W., 1984. Financial intermediation and delegated monitoring. The Review of
Economic Studies, 51(3), 393–414.
Diamond, D.W., 1991. Debt maturity structure and liquidity risk. The Quarterly Journal of
Economics, 106(3), 709–737.
Dimson, E., 1979. Risk measurement when shares are subject to infrequent trading. Journal
Ertugrul, M., Lei, J., Qiu, J., and Wan, C., 2017. Annual report readability, tone ambiguity,
and the cost of borrowing. Journal of Financial and Quantitative Analysis, 52(2), 811–
836.
Frank, M.Z., and Goyal, V.K., 2003. Testing the pecking order theory of capital structure.
Gormley, T.A., and Matsa, D.A., 2017. Playing it safe? Managerial preferences, risk, and
agency conflicts. Journal of Financial Economics, 122(3), 431–455.
167
Graham, J., Li, S., and Qiu, J., 2008. Corporate misreporting and bank loan contracting.
Greene, W., 2004. The behaviour of the maximum likelihood estimator of limited dependent
variable models in the presence of fixed effects. The Econometrics Journal, 7(1), 98–
119.
Hombert, J., and Matray, A., 2016. The real effects of lending relationships on innovative
firms and inventor mobility. Review of Financial Studies, 30(7), 2413–2445.
Hong, H.A., Kim, J.B., and Welker, M., 2017. Divergence of cash flow and voting rights,
opacity, and stock price crash risk: international evidence. Journal of Accounting
Research, 55(5), 1167–1212.
Jayaratne, J., and Strahan, P.E., 1996. The finance-growth nexus: Evidence from bank branch
deregulation. The Quarterly Journal of Economics, 111(3), 639–670.
Jayaratne, J., and Strahan, P.E., 1997. The benefits of branching deregulation. Economic
Policy Review, 13–29.
Jayaratne, J., and Strahan, P.E., 1998. Entry restrictions, industry evolution, and dynamic
efficiency: Evidence from commercial banking. The Journal of Law and
Economics, 41(1), 239–274.
Johnson, S.A., 2003. Debt maturity and the effects of growth opportunities and liquidity risk
on leverage. Review of Financial Studies, 16(1), 209–236.
168
Kerr, W.R., and Nanda, R., 2009. Democratizing entry: Banking deregulations, financing
constraints, and entrepreneurship. Journal of Financial Economics, 94(1), 124–149.
Kim, Y., Li, H., and Li, S., 2014. Corporate social responsibility and stock price crash risk.
Journal of Banking & Finance, 43, 1–13.
Kim, J.B., Li, L., Lu, L.Y., and Yu, Y., 2016. Financial statement comparability and
expected crash risk. Journal of Accounting and Economics, 61(2–3), 294–312.
Kim, J.B., Luo, L., and Xie, H., 2016. Dividend payments and stock price crash risk.
Working paper.
Kim, J.B., and Zhang, L., 2016. Accounting conservatism and stock price crash risk: Firm‐
Kothari, S.P., Shu, S., and Wysocki, P.D., 2009. Do managers withhold bad news?. Journal
of Accounting Research, 47(1), 241–276.
Kroszner, R., and Strahan, P.E., 1997. The political economy of deregulation: evidence from
the relaxation of bank branching restrictions in the United States (No. 9720). Federal
Reserve Bank of New York.
169
Kroszner, R.S., and Strahan, P.E., 1999. What drives deregulation? Economics and politics
of the relaxation of bank branching restrictions. The Quarterly Journal of Economics,
114(4), 1437–1467.
Kubick, T.R., and Lockhart, G.B., 2016. Proximity to the SEC and stock price crash risk.
Financial Management, 45(2), 341–367.
Lancaster, T., 2000. The incidental parameter problem since 1948. Journal of Econometrics,
95(2), 391–413.
Levine, R., Loayza, N., and Beck, T., 2000. Financial intermediation and growth: Causality
and causes. Journal of Monetary Economics, 46(1), 31–77.
Levine, R., and Zervos, S., 1998. Stock markets, banks, and economic growth. American
Economic Review, 537–558.
McLaughlin, S., 1995. The impact of interstate banking and branching reform: evidence
from the states. Current Issues in Economics and Finance, 1(May).
Morgan, D.P., Rime, B., and Strahan, P.E., 2004. Bank integration and state business cycles.
The Quarterly Journal of Economics, 119(4), 1555–1584.
Neyman, J., and Scott, E.L., 1948. Consistent estimates based on partially consistent
observations. Econometrica: Journal of the Econometric Society, 1–32.
Petersen, M. A., 2004. Information: hard and soft. Working paper.
Qian, J., and Strahan, P.E., 2007. How laws and institutions shape financial contracts: The
case of bank loans. Journal of Finance, 62(6), 2803–2834.
Rajan, R., and Winton, A., 1995. Covenants and collateral as incentives to monitor, Journal
of Finance, 50(4), 1113–1146.
170
Rajan, R.G., and Zingales, L., 1998. Financial Dependence and Growth. American Economic
Review, 88(3), 559–586.
Rice, T., and Strahan, P.E., 2010. Does Credit Competition Affect Small Firm Finance?.
Roberts, M.R., and Sufi, A., 2009. Renegotiation of financial contracts: Evidence from
private credit agreements, Journal of Financial Economics, 93(2), 159–184.
Roberts, M.R., 2015. The role of dynamic renegotiation and asymmetric information in
financial contracting, Journal of Financial Economics, 116(1), 61–81.
Robin, A.J., and Zhang, H., 2014. Do industry-specialist auditors influence stock price crash
risk?. AUDITING: A Journal of Practice & Theory, 34(3), 47–79.
Stein, J.C., 2002. Information production and capital allocation: Decentralized versus
hierarchical firms. Journal of Finance, 57(5), 1891–1921.
Strahan, P., 1999. Borrower risk and the price and nonprice terms of bank loans. Working
paper, Banking Studies Function, Federal Reserve Bank of New York.
Strahan, P.E., 2003. The real effects of US banking deregulation. Federal Reserve Bank of
Saint Louis Review, 85(4), 111–128.
Strahan, P.E., 2017. Are concentrated banks better informed than diversified ones?. Journal
of Accounting and Economics, 64(2–3), 278–283.
Xu, N., Li, X., Yuan, Q., and Chan, K.C., 2014. Excess perks and stock price crash risk:
Evidence from China. Journal of Corporate Finance, 25, 419–434.
Wang, K., Li, Y. and Erickson, J., 1997. A new look at the Monday effect. Journal of
Finance, 52(5), 2171–2186.
171
Zarutskie, R., 2006. Evidence on the effects of bank competition on firm borrowing and
investment. Journal of Financial Economics, 81(3), 503–537.
Zhang, M., Xie, L., and Xu, H., 2016. Corporate philanthropy and stock price crash risk:
Evidence from China. Journal of Business Ethics, 139(3), 595–617.
Zhu, W., 2016. Accruals and price crashes. Review of Accounting Studies, 21(2), 349–399.
172
Table 1. Year of state-level branch deregulation and sample distribution
This table reports the distribution of firm-year observations in each state. The sample consists of U.S. public
firms from 1962 to 1994.
State Branch deregulation Freq. Percent

Alabama 1981 345 0.56
Alaska 1960 16 0.03
Arizona 1960 569 0.92
Arkansas 1994 289 0.47
California 1960 7,262 11.75
Colorado 1991 896 1.45
Connecticut 1980 2,461 3.98
Delaware 1960 331 0.54
District of Columbia 1960 194 0.31
Florida 1988 2,265 3.67
Georgia 1983 1,447 2.34
Hawaii 1986 99 0.16
Idaho 1960 272 0.44
Illinois 1988 3,714 6.01
Indiana 1989 869 1.41
Iowa 1999 409 0.66
Kansas 1987 351 0.57
Kentucky 1990 340 0.55
Louisiana 1988 415 0.67
Maine 1975 128 0.21
Maryland 1960 902 1.46
Massachusetts 1984 2,915 4.72
Michigan 1987 1,933 3.13
Minnesota 1993 1,594 2.58
Mississippi 1986 147 0.24
Missouri 1990 1,212 1.96
Montana 1990 46 0.07
Nebraska 1985 135 0.22
Nevada 1960 362 0.59
New Hampshire 1987 343 0.56
New Jersey 1977 3,518 5.7
New Mexico 1991 108 0.17
New York 1976 7,463 12.08
North Carolina 1960 1,121 1.81
North Dakota 1987 55 0.09
Ohio 1979 3,210 5.2
Oklahoma 1988 516 0.84
Oregon 1985 342 0.55
Pennsylvania 1982 3,579 5.79
Rhode Island 1960 267 0.43
South Carolina 1960 418 0.68
South Dakota 1960 31 0.05
Tennessee 1985 601 0.97
Texas 1988 4,845 7.84
Utah 1981 389 0.63
Vermont 1970 53 0.09
Virginia 1978 1,316 2.13
Washington 1985 670 1.08
173
West Virginia 1987 55 0.09
Wisconsin 1990 953 1.54
Wyoming 1988 13 0.02
Total 61,784 100
174
Table 2. Descriptive statistics
This table reports the descriptive statistics for variables used in the empirical analyses. The sample consists of
61,784 firm-years observations for 7,190 public U.S. firms over the period from 1962 to 1994. Variable
definitions are listed in Appendix 1. All variables are winsorized at the 1% and 99% levels.
Variable N Mean Std 25% Median 75%

NCSKEWt+1 61,784 –0.270 0.641 –0.637 –0.254 0.102
DUVOLt+1 61,784 –0.153 0.321 –0.369 –0.155 0.058
BRANCHt 61,784 0.540 0.498 0.000 1.000 1.000
DTURNt 61,784 0.001 0.315 –0.082 0.002 0.084
SIGMAt 61,784 0.075 0.041 0.044 0.066 0.096
RETt 61,784 –0.366 0.429 –0.456 –0.215 –0.096
SIZEt 61,784 4.536 1.840 3.142 4.417 5.821
MBt 61,784 1.907 1.916 0.840 1.322 2.206
LEVt 61,784 0.254 0.184 0.107 0.237 0.371
ROAt 61,784 0.036 0.085 0.014 0.045 0.076
NCSKEWt 61,784 –0.271 0.634 –0.635 –0.254 0.101
INTERt 61,784 0.387 0.487 0.000 0.000 1.000
EXDEPt 56,090 –3.361 524.400 –2.146 –0.695 0.212
NCCt 61,784 0.019 0.079 –0.011 0.000 0.030
TANGt 58,697 0.609 0.383 0.342 0.546 0.839
VOLt 47,740 0.138 1.824 0.024 0.050 0.102
175
Table 3. Impact of bank deregulation on stock price crash risk
This table presents the regression results of the effect of bank branch deregulation on firm-level stock price
crash risk. The standard errors in the regressions are clustered at the state level. All variables except dummies
*** ** *
are winsorized at the 1% and 99% levels. The numbers reported in parentheses are t-statistics. , , and
denote statistical significance at the 1%, 5% and 10% levels, respectively. See Appendix 1 for variable
definitions.
(1) (2) (3) (4)

VARIABLES NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1
BRANCHt –0.019** –0.010** –0.025*** –0.014***
(–2.25) (–2.59) (–2.95) (–3.33)
DTURNt 0.001 –0.002
(0.08) (–0.53)
SIGMAt 0.526* –0.095
(1.72) (–0.61)
RETt 0.047** 0.002
(2.39) (0.18)
SIZEt 0.071*** 0.034***
(24.95) (25.12)
MBt 0.014*** 0.007***
(9.56) (9.16)
LEVt 0.007 0.001
(0.39) (0.07)
ROAt 0.439*** 0.227***
(12.01) (12.47)
NCSKEWt 0.052*** 0.028***
(13.53) (13.99)
Constant –0.216 –0.164* –0.593*** –0.336***
(–1.55) (–1.94) (–4.15) (–3.93)
State FE Yes Yes Yes Yes
Observations 61,784 61,784 61,784 61,784
R-squared 0.023 0.029 0.082 0.089
176
Table 4. Impact of bank deregulation on stock price crash risk: Firm-fixed effects
This table presents the regression results of the effect of bank branch deregulation on firm-level stock price
crash risk. The standard errors in the regressions are clustered at the state level. All variables except dummy
variables are winsorized at the 1% and 99% levels. The numbers reported in parentheses are t-statistics. ***,
**, and * denote statistical significance at the 1%, 5% and 10% levels, respectively. See Appendix 1 for variable
definitions.
(1) (2) (3) (4) (5) (6)

VARIABLES NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1 NCSKEWt+1 DUVOLt+1
BRANCHt –0.026* –0.014** –0.020* –0.011* –0.469*** –0.387***
(–1.93) (–2.17) (–1.68) (–1.83) (–4.33) (–7.39)
DTURNt 0.007 0.001 –0.003 –0.004
(0.71) (0.13) (–0.40) (–1.09)
SIGMAt –1.125*** –0.785*** –1.395*** –0.903***
(–2.94) (–4.11) (–4.25) (–5.64)
RETt –0.073** –0.048*** –0.088*** –0.054***
(–2.40) (–3.15) (–3.51) (–4.42)
SIZEt 0.184*** 0.094*** 0.191*** 0.098***
(26.99) (28.36) (20.78) (22.03)
MBt 0.014*** 0.008*** 0.014*** 0.008***
(5.20) (5.71) (5.19) (5.15)
LEVt 0.061** 0.025 0.078** 0.032**
(1.97) (1.61) (2.50) (2.18)
ROAt 0.151*** 0.067** 0.151*** 0.071***
(2.61) (2.42) (2.92) (2.85)
NCSKEWt –0.066*** –0.029*** –0.074*** –0.033***
(–13.42) (–12.14) (–11.48) (–11.37)
Constant –0.242 –0.169* –0.845*** –0.466*** –0.986*** –0.428***
(–1.48) (–1.74) (–5.73) (–5.29) (–17.45) (–16.38)
Firm FE Yes Yes Yes Yes
State×Year Yes Yes
Industry×Year Yes Yes
Observations 61,784 61,784 61,784 61,784 60,853 60,853
R-squared 0.203 0.205 0.235 0.238 0.278 0.282
177
Table 5. Endogeneity tests
This table presents the estimation results of the time trend analysis using a dynamic specification. We require
firms to experience bank branch deregulation during their available sample lifetime and extend the sample to
a five-year window post-branching deregulatory events. To economize on space, all the control variables (see
Table 3) are suppressed. The standard errors in the regressions are clustered at the state level. All variables
except dummies are winsorized at the 1% and 99% levels. The numbers reported in parentheses are t-statistics.
***, **, and * denote statistical significance at the 1%, 5% and 10% levels, respectively. See Appendix 1 for
variable definitions.
(1) (2) (3) (4)

Before2+ 0.025 0.014 –0.015 –0.006
(1.24) (1.47) (–0.80) (–0.68)
Before1 0.001 0.002 –0.007 –0.002
(0.045) (0.23) (–0.28) (–0.14)
After1 –0.030* –0.015* –0.038** –0.018**
(–1.79) (–1.76) (–2.40) (–2.34)
After2+ –0.012 –0.003 –0.046** –0.021**
(–0.58) (–0.31) (–2.46) (–2.19)
Constant –0.341 –0.243 –0.701** –0.402**
(–1.02) (–1.24) (–2.16) (–2.12)
Control No No Yes Yes
Observations 36,502 36,502 36,502 36,502
R-squared 0.026 0.031 0.089 0.096
178
Table 6. Falsification test: Randomly assign states to deregulation years
This table presents the falsification test results of Eq. (4) with randomized state deregulations. We assume the
deregulatory events do not occur in the actual deregulation years shown in Table 1. To economize on space,
all the control variables (see Table 3) are suppressed. The standard errors in the regressions are clustered at the
state level. All variables except dummies are winsorized at the 1% and 99% levels. The numbers reported in
parentheses are t-statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels,
respectively. See Appendix 1 for variable definitions.
(1) (2) (3) (4)

BRANCHt 0.006 0.004 0.011 0.006
(0.46) (0.60) (0.92) (1.00)
DTURNt 0.001 –0.002
(0.09) (–0.53)
SIGMAt 0.546* –0.085
(1.76) (–0.54)
RETt 0.048** 0.003
(2.43) (0.26)
SIZEt 0.071*** 0.034***
(24.88) (24.99)
MBt 0.014*** 0.007***
(9.58) (9.18)
LEVt 0.006 –0.000
(0.30) (–0.03)
ROAt 0.436*** 0.225***
(12.12) (12.65)
NCSKEWt 0.053*** 0.028***
(13.61) (14.07)
Constant –0.221 –0.167* –0.600*** –0.339***
(–1.59) (–1.98) (–4.23) (–4.00)
Observations 61,784 61,784 61,784 61,784
R-squared 0.023 0.029 0.082 0.089
179
Table 7. Robustness tests
This table presents the regression results of robustness tests. In Panel A, we include interstate deregulation
variable (INTER) as an additional control variable. INTER is a dummy variable that equals one for the years
after a state implemented interstate banking deregulation. In Panel B, we present the regression results based
on the sample of five years post-deregulation. In Panel C, we analyze firms that experienced intrastate bank
deregulation events during their lifetime. To economize on space, all the control variables (see Table 3) are
suppressed. The standard errors in the regressions are clustered at the state level. All variables except dummies
are winsorized at the 1% and 99% levels. The numbers reported in parentheses are t-statistics. ***, **, and *
denote statistical significance at the 1%, 5% and 10% levels, respectively. See Appendix 1 for variable
definitions.
Panel A: Including interstate deregulation variable

(1) (2) (3) (4)
BRANCHt –0.020** –0.011** –0.026*** –0.014***
(–2.37) (–2.65) (–2.88) (–3.16)
INTERt 0.012 0.003 0.004 –0.001
(0.90) (0.47) (0.29) (–0.08)
Constant –0.215 –0.164* –0.593*** –0.336***
(–1.54) (–1.94) (–4.14) (–3.92)
Observations 61,784 61,784 61,784 61,784
R-squared 0.023 0.029 0.082 0.089
180
Panel B. Different windows
(1) (2) (3) (4)
5 years post 5 years post 5 years post 5 years post
BRANCHt –0.016* –0.009** –0.015* –0.009**
(–1.85) (–2.07) (–1.97) (–2.29)
Constant –0.181 –0.140 –0.605*** –0.330***
(–1.30) (–1.66) (–4.14) (–3.80)
Observations 83,698 83,698 83,697 83,697
R-squared 0.021 0.026 0.084 0.090
Panel C: Analysis of sample firms that experience intrastate bank deregulation
(1) (2) (3) (4)
BRANCHt –0.028** –0.013** –0.042*** –0.028**
(–2.28) (–2.17) (–2.90) (–2.28)
Constant –0.317 –0.232 –0.701** –0.406**
(–0.95) (–1.19) (–2.13) (–2.11)
Observations 30,745 30,745 30,745 30,745
R-squared 0.028 0.033 0.093 0.100
181
Table 8. Robustness tests: Alternative measures
This table presents the regression results of robustness tests using alterative measures of crash risk and bank
branch deregulation. In Panel A, crash risk is proxied by CRASH and COUNT. CRASH is an indicator that
takes one if a firm experiences more than one price crash week in a fiscal year. COUNT is the number of crash
weeks minus the number of jump weeks over the fiscal year. In Panel B, we construct a bank branch
deregulation index (DERINDEX), which equals zero if a state does not permit branching via M&As, de novo
branching, or the formation of multibank holding companies; otherwise, the index equals the sum of the number
of ways that banks may expand within a state. To economize on space, all the control variables (see Table 3)
are suppressed. The standard errors in the regressions are clustered at the state level. All variables except
dummies are winsorized at the 1% and 99% levels. The numbers reported in parentheses are t-statistics. ***,
**, and * denote statistical significance at the 1%, 5% and 10% levels, respectively. See Appendix 1 for variable
definitions.
Panel A: Alternative measures of crash risk

(1) (2) (3) (4)
VARIABLES CRASHt+1 COUNTt+1 CRASHt+1 COUNTt+1
BRANCHt –0.010** –0.014* –0.011** –0.018**
(–2.51) (–1.93) (–2.62) (–2.30)
Constant 0.144 0.044 0.110 –0.229
(1.37) (0.33) (1.00) (–1.63)
Observations 61,784 61,784 61,784 61,784
R-squared 0.011 0.010 0.016 0.039
Panel B: Alternative measures of branch deregulation
(1) (2) (3) (4)
VARIABLES NCSKEWt+1 DUVOL t+1 NCSKEWt+1 DUVOL t+1
DERINDEXt –0.008** –0.004** –0.011*** –0.005***
(–2.02) (–2.06) (–2.97) (–3.01)
Constant –0.222 –0.168* –0.601*** –0.340***
(–1.59) (–1.98) (–4.20) (–3.99)
Observations 61,784 61,784 61,784 61,784
R-squared 0.023 0.029 0.082 0.089
182
Table 9. External financial dependence
This table presents the results regarding the impact of bank branch deregulation on future stock price crash risk conditional on firms’ external financial dependence. In Panel
A, the sample is partitioned based on the median value of the external financial dependence variable (EXDEP) for each year. In Panel B, the sample is partitioned based on the
median value of net change in capital (NCC) for each year. In Panel C, the sample is partitioned based on the median value of asset tangibility (TANG) for each year. To
economize on space, all the control variables (see Table 3) are suppressed. The standard errors in the regressions are clustered at the state level. All variables except dummies
are winsorized at the 1% and 99% levels. The numbers reported in parentheses are t-statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels,
respectively. See Appendix 1 for other variable definitions.
Panel A. External financial dependence

(1) (2) (3) (4) (5) (6) (7) (8)
NCSKEWt+1 NCSKEWt+1 NCSKEWt+1 NCSKEWt+1 DUVOLt+1 DUVOLt+1 DUVOLt+1 DUVOLt+1
High reliance Low reliance High reliance Low reliance High reliance Low reliance High reliance Low reliance
(EXDEP ≥ Med.) (EXDEP < Med.) (EXDEP ≥ Med.) (EXDEP < Med.) (EXDEP ≥ Med.) (EXDEP < Med.) (EXDEP ≥ Med.) (EXDEP < Med.)
BRANCHt –0.030** –0.006 –0.038*** –0.008 –0.015** –0.005 –0.020*** –0.005
(–2.43) (–0.37) (–3.44) (–0.53) (–2.40) (–0.58) (–3.58) (–0.75)
Control No No Yes Yes No No Yes Yes
Year FE Yes Yes Yes Yes Yes Yes Yes Yes
State FE Yes Yes Yes Yes Yes Yes Yes Yes
Observations 28,054 28,036 28,054 28,036 28,054 28,036 28,054 28,036
Adjusted R2 0.024 0.030 0.083 0.091 0.029 0.036 0.090 0.097
183
Panel B. Net change in capital
(1) (2) (3) (4) (5) (6) (7) (8)
(NCC ≥ Med.) (NCC< Med.) (NCC ≥ Med.) (NCC< Med.) (NCC ≥ Med.) (NCC< Med.) (NCC ≥ Med.) (NCC< Med.)
BRANCHt –0.027** –0.005 –0.027** –0.019 –0.016*** –0.002 –0.017*** –0.008
(–2.55) (–0.42) (–2.57) (–1.54) (–2.92) (–0.27) (–2.98) (–1.48)
Observations 35,325 26,459 35,325 26,459 35,325 26,459 35,325 26,459
Adjusted R2 0.024 0.027 0.088 0.079 0.030 0.032 0.096 0.084
Panel C. Asset tangibility
(1) (2) (3) (4) (5) (6) (7) (8)
(TANG ≤ Med.) (TANG >Med.) (TANG ≤ Med.) (TANG >Med.) (TANG ≤ Med.) (TANG >Med.) (TANG ≤ Med.) (TANG >Med.)
BRANCHt –0.034** –0.007 –0.038*** –0.012 –0.018*** –0.004 –0.021*** –0.006
(–2.59) (–0.48) (–3.40) (–0.96) (–2.97) (–0.51) (–4.01) (–0.98)
Observations 31,652 30,132 31,652 30,132 31,652 30,132 31,652 30,132
Adjusted R2 0.025 0.027 0.094 0.077 0.031 0.032 0.102 0.082
184
Table 10. Firm risk
This table presents the results regarding the impact of bank branch deregulation on future stock price crash risk conditional on firm riskiness. In Panel A, the sample is partitioned
based on the median value of earnings volatility (VOL) for each year. In Panel B, the sample is partitioned based on the median value of leverage (LEV) for each year. To
economize on space, all the control variables (see Table 3) are suppressed. The standard errors in the regressions are clustered at the state level. All variables except dummies
are winsorized at the 1% and 99% levels. The numbers reported in parentheses are t-statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels,
respectively. See Appendix 1 for other variable definitions.
Panel A. Earnings volatility

(1) (2) (3) (4) (5) (6) (7) (8)
High risk Low risk High risk Low risk High risk Low risk High risk Low risk
(VOL ≥ Med.) (VOL < Med.) (VOL ≥ Med.) (VOL < Med.) (VOL ≥ Med.) (VOL < Med.) (VOL ≥ Med.) (VOL < Med.)
BRANCHt –0.033** –0.012 –0.036** –0.017 –0.017** –0.008 –0.019** –0.011*
(–2.15) (–0.80) (–2.25) (–1.40) (–2.12) (–1.19) (–2.28) (–1.90)
Observations 23,875 23,865 23,875 23,865 23,875 23,865 23,875 23,865
Adjusted R2 0.027 0.026 0.082 0.084 0.032 0.030 0.087 0.088
185
Panel B. Leverage
(1) (2) (3) (4) (5) (6) (7) (8)
High risk Low risk High risk Low risk High risk Low risk High risk Low risk
(LEV ≥ Med.) (LEV < Med.) (LEV ≥ Med.) (LEV < Med.) (LEV ≥ Med.) (LEV < Med.) (LEV ≥ Med.) (LEV < Med.)
BRANCHt –0.031*** –0.007 –0.040*** –0.011 –0.013*** –0.008 –0.017*** –0.010
(–2.96) (–0.52) (–4.04) (–0.87) (–2.69) (–1.18) (–3.79) (–1.65)
Observations 30,898 30,886 30,898 30,886 30,898 30,886 30,898 30,886
Adjusted R2 0.026 0.025 0.074 0.096 0.031 0.031 0.079 0.104
186
Appendix 1. Variable definition
For both crash risk variables, the firm-specific weekly return (W) is equal to ln (1 + residual),
where the residual is from the following expanded market model regression:
Bank branch deregulation variables
BRANCH is a dummy variable that equals one after a state implemented intrastate branching
deregulation and zero otherwise. The years each state relaxed the restrictions on
intrastate branching are shown in Table 1.
DERINDEX is a bank branch deregulation index, which equals zero if a state does not permit
branching via M&As, de novo branching, or the formation of multibank holding
companies; otherwise, the index equals the sum of the number of ways that banks may
expand within a state.
Control variables
187
MB is the market value of equity (csho×prcc_f) divided by the book value of equity (market-
to-book).
SIZE is the natural logarithm of total assets (at) at the end of the fiscal year.
Other variables
Before2+ is an indicator variable that takes one for observations with two years or more prior
to deregulation and zero otherwise.
Before1 is an indicator variable that takes one for observations with one year prior to
deregulation and zero otherwise.
After1 is an indicator variable that takes one for observations with one year post-deregulation
and zero otherwise.
After2+ is an indicator variable that takes one for observations with two years or more post-
deregulation and zero otherwise.
EXDEP is external financial dependence ratio defined as investment (capital expenditure
(capx) + R&D expenses (xrd) + acquisitions using cash (aqc)) minus operating income
before depreciation (oibdp), divided by investment.
NCC is net change in capital defined as net change in equity and debt normalized by total
assets.
TANG is asset tangibility defined as the proportion of tangible assets (ppegt) to total assets.
188
VOL is earnings volatility defined as the standard deviation of the ratio of earnings,
excluding extraordinary items and discontinued operations, to lagged total equity during
the past five years.
189
Appendix 2. Robustness tests: Different windows
This table presents the regression results of robustness tests using extended sample with different windows. In
Panel A, we present the regression results based on the sample of three years post-deregulation. In Panel B, we
present the regression results based on the sample of ten years post-deregulation. To economize on space, all
the control variables (see Table 3) are suppressed. The standard errors in the regressions are clustered at the
state level. All variables except dummies are winsorized at the 1% and 99% levels. The numbers reported in
parentheses are t-statistics. ***, **, and * denote statistical significance at the 1%, 5% and 10% levels,
respectively. See Appendix 1 for variable definitions.
Panel A. Three-year windows

(1) (2) (3) (4)
BRANCHt –0.017* –0.009** –0.019** –0.011**
(–1.80) (–2.06) (–2.29) (–2.62)
Constant –0.182 –0.142 –0.610*** –0.336***
(–1.29) (–1.67) (–4.24) (–3.94)
Observations 73,078 73,078 73,077 73,077
R-squared 0.021 0.026 0.081 0.087
Panel B. Ten-year windows
(1) (2) (3) (4)
BRANCHt –0.019* –0.010** –0.016** –0.009**
(–2.01) (–2.12) (–2.08) (–2.26)
Constant –0.194 –0.146* –0.649*** –0.351***
(–1.39) (–1.73) (–4.46) (–4.07)
Observations 95,959 95,959 95,958 95,958
R-squared 0.024 0.030 0.087 0.094
190
Chapter 5
Conclusion and suggestions for future research
This thesis is an effort to advance our understanding of the impact of three corporate
financing issues on shareholders’ wealth collapse probability, namely stock price crash risk.
First, I examine whether short-term debt affects stock price crash risk. The empirical
evidence shows that financing from short-term debt holders mitigates stock price crash risk,
which is consistent with the notion that short-term creditor monitoring acts as an effective
corporate governance mechanism in curbing borrowers’ opportunistic behavior. Second, I
find that high levels of supplier financing lead to lower stock price crash risk, consistent with
the monitoring view that trade credit suppliers can effectively monitor buyers and constrain
their bad-news-hoarding behavior. Third, I investigate the influence of staggered U.S. bank
branching deregulation laws on stock price crash risk. The results show that the stock price
crash risk of nonfinancial firms decreases significantly following intrastate branching
deregulation. This chapter contributes to the broader literature on how laws affect finance.
All the three issues are closely related to corporate financing policy whereas at the
same time inherently intertwined. Firms usually receive external financial indebtedness in
the form of loans and bonds while trade credit provides firms with an informal source of
short-term financing. However, to the extent that debt maturity structure and trade credit are
firm-level financial policies, examining how they determine firm-level stock price crash risk
has to address challenging endogeneity concerns. The third topic employs staggered
deregulation laws in the banking sector to allow for a causal interpretation of the link
between corporate financing and stock price crash risk. Also, it complements the first topic
on debt maturity structure by providing additional evidence on creditor monitoring
191
mechanisms. I provide below a summary of the findings of this thesis, together with a
discussion of my contributions and suggestions for future research.
5.1. Summary of results
In Chapter 2, I examine how corporate debt maturity policy affects stock price crash
risk. Using a sample of 53,052 firm-year observations from 1989 through 2014, I find that
firms using more short-term debt have lower future stock price crash risk, measured as the
conditional skewness of stock return distribution. This finding is consistent with managers
being less likely to conceal and hoard bad news in the presence of external monitoring by
short-term debt lenders. The empirical results are robust to several tests addressing
endogeneity issues and those using alternative measures of stock price crash risk and short-
term debt. Since the primary measure of short-term debt is based on financial statements, I
follow Guedes and Opler (1996) and Brockman, Martin, and Unlu (2010) and employ an
incremental approach in which I use a sample of new debt issues. I find that the maturities
of newly issued debt are positively associated with future stock price crash risk, which
further strengthens the main inference of the relationship between the maturities of new debt
issues and crash risk.
Furthermore, I examine whether the mitigating effect of short-term debt on crash risk
is conditional on corporate governance, information asymmetry, and firm riskiness. The
results support the hypotheses that the negative association between short-term debt and
stock price crash risk is more pronounced among firms with lower shareholder rights and
less (non-transient) institutional ownership, those with less efficient analyst monitoring and
higher asymmetric information, as well as highly levered or unrated firms. Overall, the
findings are in line with Stulz’s (2001) argument that short-term debt can act as an effective
192
substitute for governance mechanisms in disciplining managerial opportunistic behavior and
future stock price crash risk.
In Chapter 3, I examine whether stock price crash risk is affected by an alternative
source of external financing, namely trade credit. Unlike the focus on short-term debt
financing in Chapter 2, I study a different setting in which firms finance inputs from their
suppliers in the form of trade credit. I develop two competing hypotheses based on the
monitoring and concession roles of trade creditors. On one hand, the monitoring view posits
that trade credit suppliers, with their information and liquidation advantage, are able to
effectively monitor buyers (Biais and Gollier, 1997; Fabbri and Menichini, 2010), thereby
refraining the latter from hoarding bad news. On the other hand, the concession view argues
that suppliers tend to be lenient and willing to grant concessions to buyers, especially when
the latter are financially distressed (Wilner, 2000; Cuñat, 2007); in this case, buyers are more
prone to withholding bad news and experiencing a stock price crash in the future. Using a
sample of U.S. public firms from 1990 to 2013, I find that firms’ usage of trade credit
financing is negatively associated with future stock price crash risk. This negative relation
is robust to a set of robustness checks. The findings are consistent with the monitoring view
that suppliers are able to monitor their customers more effectively using trade credit, which
reduces customers’ bad-news-hoarding behavior and future crash risk.
Furthermore, I conduct several additional analyses conditioning on governance, bank
monitoring, market power, and financial distress. The results show that the negative
association between trade credit and crash risk is more pronounced for firms with weaker
governance or bank monitoring, and those with weaker market power or higher distress risk.
These findings further support the monitoring view by suggesting that supplier monitoring
serves as a substitute for formal governance mechanisms in disciplining customers’
misconduct.
193
A potential limitation of this chapter is that, while my trade credit measure is
conventional in the literature (e.g., Petersen and Rajan, 1997; Fisman and Love, 2003, Cuñat,
2007), it is based on financial statement information and thus may not be able to capture the
incremental effect of trade credit on crash risk. The reason is that trade credit is frequently
used in business activities and so it is difficult for me to collect data or conduct tests based
on new trade credit contracts. Also, due to data and time constraints, my research has not
examined the extent to which the results are affected by supplier-buyer relationships. Recent
studies have used supplier-buyer pairwise data from 10-K disclosures to analyze supplier-
buyer relationships (e.g., Murfin and Njoroge, 2015). Identifying firms’ major suppliers is a
part of my future research agenda, which will enable me to better capture supplier monitoring
mechanisms and examine how they affect stock price crash risk.
In Chapter 4, I test whether the deregulation of bank branch restrictions affects stock
price crash risk for nonfinancial firms. Between the 1970s and 1990s most states in the U.S.
removed branching restrictions in the banking sector by allowing banks to open branches or
undertake M&A activities within or out of states. Using a sample of 61,784 firm-year
observations from 1962 to 1994, I find that the passage of intrastate branching deregulation
laws leads to lower stock price crash risk for nonfinancial firms. I address the potential
endogeneity concerns by estimating a dynamic specification and performing falsification
tests. I also control for related deregulatory events and estimate long-window samples to
ensure the robustness of the results. My findings are consistent with the argument that, after
the passage of intrastate deregulation, banks can monitor their borrowers more effectively
and help constrain their bad-news-hoarding behavior. Furthermore, I show that the negative
relation between bank deregulation and crash risk is more pronounced for firms with high
external financial dependence and risk-taking behavior, suggesting that external financial
dependence and firm riskiness are the two mechanisms explaining the results.
194
5.2. Contributions and suggestions for future research
Taken together, the three chapters of this thesis conclude that corporate financing
policies significantly affect the likelihood of extreme collapses in firm value. Understanding
what drives or constrains the perceived crash risk has the potential to make a significant
contribution toward an understanding and knowledge of extreme equity value decreases.
Managers would learn some new determinants of stock price crash risk from my findings
and act to avoid extreme unfavorable outcomes, such as compensation reduction, executive
dismissal, and infamous lawsuits (see the cases of Enron or Tyco). For equity investors, the
first and second essay would enable them to make better portfolio investment decisions by
predicting future crash risk based on the information about debt maturity and supplier
financing policies. The third essay would be important for policy makers in banking industry
to take nonfinancial firms’ stock price crash risk into considerations. Thus, I argue that my
results have important practical implications for managers, shareholders, debtholders, and
suppliers by providing insights on how to moderate firms’ crash risk using corporate
financing policies.
My thesis also contributes to academic research on corporate financing and stock
price crash risk. First, it adds to the literature on stock price crash risk by suggesting three
new determinants of crash risk. The existing literature has examined a number of
determinants, mainly including corporate governance factors that are related to financial
reporting, corporate disclosure (e.g., Hutton, Marcus, and Tehranian, 2009; Kim, Li, and
Zhang, 2011a, b), and managerial characteristics (e.g., Callen and Fang, 2015; Andreou,
Louca, and Petrou, 2016; Kim, Wang, and Zhang, 2016). Very few studies have paid
attention to the role of corporate financing in influencing stock price crash risk. In this thesis,
I propose three distinct factors that are associated with crash risk, namely debt maturity,
trade credit, and bank deregulation laws. Second, my thesis provides new and exciting
195
research insights on the links between the interests of various finance suppliers and those of
shareholders. Chapters 2 and 3 imply that short-maturity debt and trade credit allow creditors
and suppliers to restrict managers’ misconduct, which in turn is mutually beneficial to
shareholders’ wealth through the reduction of stock price crash risk. Chapter 4 shows that
state-level deregulation laws in the banking sector reduce nonfinancial firms’ crash risk, thus
implying that shareholders’ value also benefits from the improved bank monitoring
efficiency driven by bank reform.
Future research can investigate new facets of stock price crash risk. First, it remains
fruitful to continue uncovering new determinants of stock price crash risk. However, a key
challenge with this strand of research is to address endogeneity concerns or enhance the
identification of causal effects. Atanasov and Black (2016) argue that shock-based designs
(or natural experiments) can often create a stronger basis for reasonable causal inference
than the best available non-shock designs. Thus, it is promising for future research to
investigate the association between stock price crash risk and exogenous events due to
regulatory changes. Second, while a bulk of research on the causes of stock price crash risk
has emerged over the past decade, studies on the consequences of stock price crashes are
surprisingly scarce. Crash incidence has disastrous consequences for shareholders’ wealth
but may also implicitly alter certain corporate policies or exert different pressures on
stakeholders. This will be another interesting avenue for future research.
196
References
Andreou, P.C., Louca, C., and Petrou, A.P., 2016. CEO age and stock price crash risk.
Review of Finance, 21(3), 1287–1325.
Atanasov, V., and Black, B., 2016. Shock-based causal inference in corporate finance and
accounting research. Critical Finance Review, 5(2), 207–304.
Biais, B., and Gollier, C., 1997. Trade credit and credit rationing, Review of Financial
Studies, 10(4), 903–937.
Brockman, P., Martin, X., and Unlu, E., 2010. Executive compensation and the maturity
structure of corporate debt. Journal of Finance, 65(3), 1123–1161.
Burkart, M., and Ellingsen T., 2004. In-kind finance: a theory of trade credit, American
Economic Review, 94(3), 569–590.
Callen, J.L., and Fang, X., 2015. Religion and stock price crash risk. Journal of Financial
and Quantitative Analysis, 50(1–2), 169–195.
Cuñat, V., 2007. Trade credit: suppliers as debt collectors and insurance providers, Review
Fabbri, D., and A. Menichini. 2010. Trade credit, collateral liquidation, and borrowing
constraints. Journal of Financial Economics, 96(3), 413–32.
Fisman, R., and Love, I. 2003. Trade credit, financial intermediary development, and
industry growth. Journal of Finance, 58(1), 353–374.
Guedes, J., and Opler, T., 1996. The determinants of the maturity of corporate debt issues.
197
Murfin, J., and Njoroge, K., 2015. The implicit costs of trade credit borrowing by large firms.
Review of Financial Studies, 28 (1), 112–145.
Studies, 10(3), 661–691.
Wilner, B.S., 2000. The exploitation of relationships in financial distress: The case of trade
credit. Journal of Finance, 55(1), 153–178.
Stulz, R., 2001. Does financial structure matter for economic growth? A corporate finance
perspective. Financial structure and economic growth: A Cross-country comparison of
banks, markets, and development, 143–188.
198

Corporate Financing and Stock Price Crash Risk

Uploaded by

Copyright:

Available Formats

Corporate Financing and Stock Price Crash Risk

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Corporate Financing and Stock Price Crash Risk

Uploaded by

Copyright:

Available Formats

Corporate Financing and Stock Price Crash Risk

A thesis submitted to The University of Manchester for the degree of

in the Faculty of Humanities

Alliance Manchester Business School

CORPORATE DEBT MATURITY AND STOCK PRICE CRASH RISK 18

The University of Manchester

provided as a committed finance researcher.

was accepted by European Financial Management Association in 2016, Colin decided to

come to Switzerland together with me to ensure that I can do well in my presentation.

suggestions in my PhD reviews as well as inspirational encouragement in my difficult period.

taught me or gave me valuable comments on my thesis: Dr. Amedeo De Cesari, Professor

reviewers for their efforts in the publication of my essay.

Furthermore, I will forever be thankful to my former postgraduate supervisor,

is to them that I dedicate this thesis.

1.1. Background and motivation

2014; Ali, Li, and Zhang, 2015).

and bank deregulation.

alternative external financing resource, have statistically and economically significant

each of the three essays in the following section.

1.2. Thesis overview and contributions

short-term debt serving as an effective tool to curb managerial bad-news-hoarding behavior,

in reducing managerial bad news hoarding.

monitoring function of external creditors and in particular short-term debt lenders..

maintain a long-term supplier-buyer relationship, which might create opportunities for

In my third essay, I focus on an exogenous shock in the banking industry—bank

and firms’ crash risk.

Datta and Raman, 2005; Roberts and Sufi, 2009).

measure crash risk as negative conditional skewness of firm-specific weekly returns

1.3. Thesis structure

Finance Group, Alliance Manchester Business School. It allows chapters to be incorporated

order throughout the thesis.

credit. Journal of Banking & Finance, 36(5), 1402–1413.

of bad versus good news. Working paper.

of Corporate Finance, 48, 700–725.

of Finance, 57(6), 2807–2833.

industry structure in local US markets. Journal of Finance, 61(1), 437–461.

of Financial Studies, 20(2), 491–527.

maturity structure of corporate debt. Journal of Finance, 60(5), 2333–2350.

crash risk?. The Accounting Review, 90(1), 265–299.

Perspectives, 19(4), 171–192.

Finance, 53(3), 819–843.

Financing constraints and corporate investment. Brookings papers on economic activity,

firms and inventor mobility. Review of Financial Studies, 30(7), 2413–2445.

risk. Journal of Financial Economics, 94(1), 67–86.

branch deregulation. The Quarterly Journal of Economics, 111(3), 639–670.

Journal of Banking & Finance, 43, 1–13.

Journal of Financial Economics, 101(3), 713–730.

Firm level analysis. Journal of Financial Economics, 100(3), 639–662.

level evidence. Contemporary Accounting Research, 33(1), 412–441.

and firm value. Accounting & Finance, 53(3), 791–808.

evidence. Journal of Finance, 47(1), 169–200.

Patsuris, P., 2002. The corporate scandal sheet. Forbes.com, 1–6.

Studies, 10(3), 661–691.

of Finance, 50(4), 1113–1146.

private credit agreements. Journal of Financial Economics, 93(2), 159–184.

Finance, 43(1), 1–19.

Policy Recommendations, Volumes I–III. JCS–3–03. Washington, DC.

Evidence from China. Journal of Business Ethics, 139(3), 595–617.

Corporate Debt Maturity and Stock Price Crash Risk

constraining managers’ bad-news-hoarding behavior. The inverse relation between short-