Corporate Governance and Firm‐specific Stock Price Crashes

• DOI: http://doi.org/10.1111/eufm.12084
• Date: 01 November 2016
• Author: Joanne Horton,Panayiotis C. Andreou,Christodoulos Louca,Constantinos Antoniou
• Published date: 01 November 2016

Corporate Governance and

Firm-specific Stock Price Crashes

Panayiotis C. Andreou

CyprusUniversity of Technology,Departmentof Commerce, Financeand Shipping, 115 SpyrouAraouzou

Street,3036 Lemesos, Cyprus, andDurham University Business School, Mill HillLane, Durham DH1 3LB, UK

E-mail: benz@pandreou.com

Constantinos Antoniou

Warwick Business School, Warwick University, Coventry CV4 7AL, UK

E-mail: constantinos.antoniou@wbs.ac.uk

Joanne Horton

Exeter University Business School, University of Exeter, Rennes Drive, Exeter EX4 4PU, UK

E-mail: j.horton@exeter.ac.uk

Christodoulos Louca

CyprusUniversity of Technology,Departmentof Commerce, Financeand Shipping, 115 SpyrouAraouzou

Street,3036 Lemesos,Cyprus, andDurham University Business School, Mill Hill Lane, Durham DH1 3LB, UK

E-mail: christodoulos.louca@cut.ac.cy

Abstract

We investigate whether ownership structure, accounting opacity, board structure

& processes and managerial incentives attributes relate to futurestock price crash

risk. Principal component analysis on the 21 attributes that comprise these four

corporate governancedimensions reveals that they can explain between 13.1%and

23.0% of a one standard deviation in crashrisk. Transient institutional ownership,

CEO stock option incentives and the proportion of directors that hold equity

increasecrash risk, whilst insiders’ownership,accounting conservatism, board size

and the presenceof a corporate governance policy mitigatecrash risk. Overall these

relationships are more pronounced in environments that accentuate agency risk.

Keywords: crash risk, corporate governance, agency risk, information environment

JEL classification: G38, G34, M48

We express our heartfelt thanks to the Editor, John Doukas, and three anonymous referees

for their constructive suggestions and helpful comments. We also thank Richard Harris,

Rajesh Tharyan, Grzegorz Trojanowksi, Jay Ritter, David Yermack and seminar

participants at Exeter Business School, Cyprus University of Technology and the European

Financial Management Meeting 2012 for useful comments and suggestions. All remaining

errors are our own.

European Financial Management, Vol. 22, No. 5, 2016, 916–956

doi: 10.1111/eufm.12084

© 2016 John Wiley & Sons, Ltd.

1. Introduction

The great breadth and depth of corporate governance literature, as shaped by a growing

empirical and theoretical research in the past few decades, is indicative of the significant

role that corporate governance systems play in protecting shareholders’welfare. This

study investigates which corporate governance attributes, amongst the many that firms

employ, are the most prominent ones in explaining the link between governance and

future stock price crashes.

Stock price crashes are likely to occur among firms with high agency risk (Callen and

Fang, 2013; Kim and Zhang, 2015). Managers, in such firms, due to career concerns and

incentives arising from compensation contracts, may exploit information asymmetries to

conceal negative information and engage in short-sighted price maximisation to better

serve their own interests (Stein, 1989); either through earnings management (Kothari

et al., 2009) or suboptimal investment decisions that cater to prevailing market sentiment

(Bebchuk and Stole, 1993). However, withholding, delaying or accumulating the

disclosure of bad news is unsustainable for long periods and will eventually lead to

significant stock price crashes when the true fundamentals unexpectedly reveal an

enormous amount of negative information in the market (Bleck and Liu, 2007; Callen

and Fang, 2013; Kim and Zhang, 2015).

A large body of literature finds that corporate governance systems can alleviate agency

risk, curbing such opportunistic managerial behaviour that could be harmful to

shareholders (Karamanou and Vafeas, 2005; Klein, 2002; Masulis et al., 2007; Xie et al.,

2003). Prior research on the effect of corporate governance on various organisational

outcomes typically focuses on individual governance mechanisms or constructs one-

dimensional governance metrics by summing up individual variables (Gompers et al.,

2003). Similarly, the crash risk literature has only focused upon a single governance

mechanism (Callen and Fang, 2013; Hutton et al., 2009; Kim et al., 2011; Kim and

Zhang, 2015).

1

In contrast, we undertake a comprehensive investigation using a broad set

of governance attributes which provides a more appealing and credible approach

(Ashbaugh-Skaife et al., 2006; Bebchuk et al., 2009; Bhagat et al., 2008) and enables us

to measure the overall quality of a firm’s governance system and ascertain which are the

most relevant in limiting crash risk.

Drawing motivation from all these studies we examine which corporate governance

attributes mitigate the 1-year-ahead firm-specific stock price crashes.

2

We jointly

consider four central dimensions of governance mechanisms: (i) ownership structure,

(ii) accounting opacity, (iii) board structure & processes, and (iv) managerial

incentives, which comprise a broad set of 21 attributes. Each of these corporate

governance dimensions are designed to increase or enhance the monitoring of

management’s actions to promote effective decision-making, limit their opportunistic

behaviour and reduce the information asymmetry between the firm and its external

stakeholders (Ashbaugh-Skaife et al., 2006). Although we motivate the use of each of the

corporate governance attributes employed, we do not expect that all 21 would equally

1

Kim and Zhang (2014) also examine the relationship between financial reporting opacity

and ex-ante (or perceived) crash risk, as reflected in the steepness of implied volatility smirks.

2

We use the terms firm-specific stock price crash and crash risk interchangeably to refer to

1-year-ahead crash risk that results from sharp stock price declines.

© 2016 John Wiley & Sons, Ltd.

Corporate Governance and Firm-specific Stock Price Crashes 917

affect a firm’s propensity to stock price crashes since some attributes might have little or

no relevance to the phenomenon under scrutiny. Hence, in this study, we rely on various

regression model specifications from principal component analysis (PCA) to a

horse-race approach to identify which of these dimensions and attributes are responsible

for explaining future firm-specific stock price crashes and under what circumstances.

Our regression approach i s carefully implemented to ta ckle econometric problem s

previously identified in the corpo rate governance literature and w hich often cloud the

interpretation of resul ts, for example simultaneo us causality and dynamic en dogeneity

issues (Bhagat and Bolton, 200 8; Callen and Fang, 2013; Creme rs and Ferrell, 2014;

Kim and Zhang, 2015) and the limiting i nterpretation of a single or comp osite

corporate governance index (Ashbaugh-Skaife et al., 2006 ; Cremers and Ferrell,

2014).

Overall, our results are cons istent with the notion that a firm’s gover nance system

can be set up to mitigate the occu rrence of future stock price cr ashes. Specifically, the

principal component anal ysis reveals that, dependi ng on the regression estim ation

setting, the statistica l significant factors that eme rge from the governance attr ibutes

can explain overall between 1 3.1% and 23.0% of a one standard devi ation of future

crash risk. Thus, the effect of co rporate governance on crash r isk is not just

statistically significant bu t also important in economic ter ms. Further detailed analysi s

reveals that the governance attributes pertaining to both ownership structure and

accounting opacity provide a first-order effect in mitigating the occ urrence of future

crash risk, and can explai n on their own approximatel y 9.0% and 3.4% of a one

standard deviation of futur e crash risk, respectively . Whilst the remaining dimensi ons,

namely board structure & processes and manage rial incentives add very li ttle, if any,

on their own, to the explana tory power of the model that l inks the governance factor s

to crash risk.

Analysis of the 21 governance attributes employed, reveals seven which appear to be

the most prominent ones to explain the occurrence of future stock price crashes.

Specifically we find that crash risk increases with transient institutional ownership, CEO

stock option incentives and the proportion of outside directors that hold equity in the

company. Furthermore, despite the widespread claim that directors’share-ownership

aligns managerial and shareholders incentives, our results with respect to outside

directors suggest that this incentive mechanism may not reduce crash risk, but rather

increase it instead, consistent with Song and Windram (2004). We also find that crash

risk decreases with insiders’ownership, the level of conditional accounting conservatism

in the financial reports, board size and the presence of corporate governance policy in a

firm’s mandate. In this respect, firms with larger boards and a clear defined corporate

governance policy are also more likely to reduce the agency conflicts that induce crash

risk.

Interestingly, unlike prior research, we do not find that the opacity of the financial

reports increases stock crashes (Hutton et al., 2009). Further analysis reveals that this

lack of a positive relationship is driven by the financial crisis, an economic downturn

period in which managers had limited opportunities and fewer incentives to stockpile

negative information through earning management practices (Ahmad-Zaluki et al.,

2011; Bertomeu and Magee, 2011; Chia et al., 2007; Filip and Raffournier, 2014; Jenkins

et al., 2009) a precursor situation to stock price crashes.

Finally, we examine whether industry or firm characteristics, which are known to

affect the effectiveness of corporate governance systems (Giroud and Mueller, 2010),

© 2016 John Wiley & Sons, Ltd.

918 P. C. Andreou, C. Antoniou, J. Horton and C. Louca