The impact of bank merger growth on CEO compensation

• Author: Zhian Chen,Lu Xing,Wing‐Yee Hung,Donghui Li
• Published date: 01 October 2017
• Date: 01 October 2017
• DOI: http://doi.org/10.1111/jbfa.12263

DOI: 10.1111/jbfa.12263

The impact of bank merger growth on CEO

compensation

Zhian Chen1Wing-Yee Hung2Donghui Li3Lu Xing4,5

1Schoolof Banking and Finance, University of

NewSouth Wales, Australia

2RBCCapital Markets, Sydney, Australia

3JinanUniversity, Management School,

Guangzhou,China

4BusinessSchool, University of Edinburgh, UK

5AdamSmith Business School, University of

Glasgow,UK

Correspondence

DonghuiLi, Management School, Jinan University,

601Huangpu Avenue West, Tianhe District,

Guangzhou510632, China.

Email:lidonghui@jnu.edu.cn

Fundinginformation

Liacknowledges financial support from the

NationalNatural Science Foundation of China

(GrantNo. 71362013) and Jinan University.

Abstract

We examine the impact of bank mergers on chief executive offi-

cer (CEO) compensation during the period 1992–2014, a period

characterised by significant banking consolidation. We show that

CEO compensation is positively related to both merger growth and

non-merger internal growth, with the former relationship being

higher in magnitude. While CEO pay–risk sensitivity is not signif-

icantly related to merger growth, CEO pay–performance sensitiv-

ity is negatively and significantly related to merger growth. Collec-

tively,our results suggest that, through bank mergers, CEOs can earn

higher compensation and decouple personal wealth from bank per-

formance. Furthermore, we document a more severe agency prob-

lem in CEO compensation as a consequence of bank mergers relative

to mergers in industrial firms. Finally, we find that the post-financial

crisis regulatory reform of executivecompensation in banks has lim-

ited effectiveness in curbing the merger–pay links.

KEYWORDS

CEO compensation, incentives, bank mergers, financial crisis

1INTRODUCTION

Mergers and acquisitions (M&As) are major,externally observable and discretionary long-term investments that pro-

vide managers with opportunities for their incentives (e.g., hubris, personal risk reduction and perquisites) to diverge

from the interests of shareholders (Bliss & Rosen, 2001). These potential agency problems can also be inferred from

the links between asset growth through M&As and chief executiveofficer (CEO) compensation. Jensen (1986) argues

that managers can grow their firms beyond the optimal size, which brings about increased managerial power and bet-

ter remuneration. Seo, Gamache, Devers, and Carpenter (2015) find that relatively underpaid CEOs tend to engage

in M&As to increase their pay to the level of their peers. However, theoretically and empirically, it remains unclear

whether CEOs receive beneficial compensation through M&A activity. While Bliss and Rosen (2001), Grinstein and

Hribar (2004), & Harford and Li (2007) find a positive effect of M&As on subsequent executivepay, Avery, Chevalier,

and Schaefer (1998) find no such effect. Toclarify the unsettled merger–pay relationship, our paper investigates what

private benefits in compensation accrue to CEOs by undertaking M&As. CEO compensation is estimated as a function

of asset growth, market value and accounting profitability,among many other factors. To study the merger effect, we

distinguish asset growth through M&As from non-merger internal growth.

1398 c

2017 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jbfa JBus Fin Acc. 2017;44:1398–1442.

CHEN ET AL.1399

Toexamine the links between CEO compensation and M&A decisions, this paper focuses on the banking industry,

which provides a natural laboratory on this issue. Minnick, Unal, and Yang (2011) argue that the role of regulatory

supervision in banks can be a substitute for, or a complement of,internal corporate governance (e.g., CEO compensa-

tion), and that some specific governance issues (e.g.,the relationship between CEO compensation and M&As) that are

valid in industrial firms may not be valid in banks (Adams & Mehran, 2003; Barth, Caprio, & Levine, 2004). The risk-

shifting problem is particularly severe in banks because of their high leverage (John, Mehran, & Qian, 2010), which

complicates the agency problems, making the effective design of executive compensation contracts, as a remedy to

mitigate the problems, especially important. Furthermore,according to Minnick et al. (2011), the M&As made by banks

are normally within a single financial industry and are thus not drivenby industry rebalancing. The single industry study

addresses the challenge of identifying industry sectors in fixed-effect regressions, as the industry classifications may

not be detailed enough.

The banking industry in the United States is an appropriate setting for analysing the impact of M&As on CEO com-

pensation arrangements for at least two reasons. First, the federal deregulations of the banking industry opened up

new opportunities for banks to grow via M&As, leading to increased decision-making power of managers through con-

trollingenlarged banks (Jensen, 1986; Hope & Thomas, 2008). The Riegle-Neal Interstate Banking and Branching Efficiency

Act of 1994 removedmost restrictions on interstate bank M&As and allowed banks to open branches in multiple states.

The Gramm-Leach-Bliley Financial Services Modernization Act of 1999 eliminated the barriers that forbade commercial

banks to merge with insurance underwriters, securities brokerages and investment banks. Followingthe creation of

expanded investment sets for banks, a large amount of consolidation occurred. Second, the majority of debt holders

of banks (i.e., dispersed depositors), who are insured by the Federal Deposit Insurance Corporation (FDIC), lack the

incentive, as well as the ability, to monitor the actions of bank managers. It is thus likelythat bank CEOs employ the

expanded investmentoptions via M&As to seek favourable compensation contracts.

The US banking industry exhibits different characteristicsfrom non-financial industries and, hence, may exhibit dif-

ferent patterns when we examinethe links between CEO compensation and M&As. First, the banking industry is heav-

ily regulated, compared to non-financial industries. Becoming large is especially important for banks. M&As, as a means

to achieve rapidsize expansion, benefit bank shareholders, as those banks deemed ‘too big to fail’ enjoy advantageous

regulatory treatments, such as governmentbailout, to prevent extreme downside risk. It remains an empirical question

whether the heavy regulation in banking can re-shape the merger–pay links documented in non-financial industries.

Second, bank CEO pay is based on less equity than CEO pay in non-financial industries (Adams & Mehran, 2003); and

bank CEO pay–risk sensitivity is significantly lower than that of industrial firms, despite efforts bybanks since the mid-

1990s to increase it (Belkhir & Chazi, 2010; DeYoung,Peng, & Yan, 2013). The disparity in their pay structure calls for

careful explanations of the differences in M&A engagement between bank CEOs and non-financial firm CEOs. Third,

corporate governance practices of banks are not identical to those of industrial firms (Adams & Mehran, 2003). It is

meaningful to study whether CEO compensation, as a major internal governance mechanism, functions differently in

linking to M&A decisions across the two groups of firms. Fourth, compared to shareholders of industrial firms, bank

shareholders are likely to take excessiverisks at the expense of debt holders (John et al., 2010). One such activity is

M&As, which are economically significant and highly risky in nature. Banks are debt holders of industrial firms and

have the incentive to monitor the latter using their expertiseand information advantage. In contrast, bank depositors

do not have the incentive to monitor bank managers, as their deposits are largely insured. Given the risk nature of

M&As and the different monitoring roles of debt holders, it is worthwhile to study whether M&As are associated with

CEO compensation differently across banks and industrial firms.

We decompose asset growth in banks into growth driven by bank mergers and internal growth unrelated to merg-

ers. We examine the changes in CEO compensation in response to the two different means of bank growth. Several

key findings emerge. First, we document a positive relationship between CEO total compensation and asset growth

through bank mergers. Among the components of the total compensation, merger growth significantly increases the

level of equity compensation but does not significantly affect the level of cash compensation. Second, CEO compen-

sation is positively related to non-merger internal growth. However, it adds much less to CEO compensation than

does the same dollar amount of merger growth. Acquiring US$ 1 million of new assets through mergers increases

1400 CHEN ET AL.

CEO total compensation by US$ 187.5; by comparison, US$ 1 million of non-merger internal growth increases the

total compensation by only US$ 13. Finally,while CEO pay–risk sensitivity (vega) is not significantly related to merger

growth, we find that CEO pay–performance sensitivity (delta) is negativelyand significantly related to merger growth,

which suggests that, through M&As, CEOs wield their increased power to decouple their equity wealth from bank

performance.

We address endogeneity problems in three ways. First, we include firm-fixed effects to control for time-invariant

firm heterogeneity that may drive the merger–payrelation. Second, as we find the current level of CEO compensation

is associated with its past levels, we use a dynamic panel system generalized method of moments (GMM) estimator

(Arellano & Bover, 1995) to mitigate reverse causality and time-invariant omitted variable bias. Third, we adopt a

difference-in-differences (DiD) estimator combined with propensity score matching (PSM) to further alleviate the

endogeneity concerns. The DiD estimator compares the variations in CEO compensation around M&As between

acquiringbanks and matched non-acquiring control banks, which removes the effect of constant firm-level heterogene-

ity and unobserved common time trends. Matches between acquiring and non-acquiring control banks are identified

by using PSM, which reduces the selection bias due to observed factors that drive bank mergers. The empirical results

reaffirm our main findings.

Wesplit the sample banks into high versus low groups based on the median dollar value of bank merger activity over

the sample period. We find that CEOs of high-merger banks are paid less per dollar value of assets acquired through

M&Asthan CEOs of low-merger banks. The result suggests that high-merger banks discourage CEOs’ excessive merger

activity through compensation arrangements.In addition, we find that merger growth significantly increases CEO vega

of low-merger banks but does not significantly affect the vega of high-merger banks. This indicates that low-merger

banks respond to their below-median M&A intensity by conferring more incentives on CEOs to employ risky invest-

ment opportunities.

We extend our analysis to the compensation arrangements of ‘top five’ executives in banks. We find that asset

growth through bank mergers increases their total compensation. More specifically, it increases the level of equity

component but does not affect the level of cash component, which is consistent with our findings for bank CEOs.

Nonetheless, we do not find that top five executives’delta or vega shifts around bank mergers.

We compare the impact of M&As on CEO compensation in industrial firms with that in banks. Industrial firms oper-

ate in different business and regulatory environments. Their merger impact on compensation structure as well as

the incentives derived may be different. We show that mergers in industrial firms substantially increase CEO cash

compensation, a result absent in bank mergers. One possible explanation is that industrial firms are regulated to a

lesser extent, making the large cash payments more feasible. Although CEOs in industrial firms gain higher total and

equity compensation after M&As, their compensation increases due to mergers are smaller than those in banks. In

industrial firms, both CEO delta and vega increase following M&As, indicating that the post-merger compensation

arrangements better align CEO incentives with shareholder interests and motivate CEOs to take higher risks. The

results taken together highlight a less severe agency problem in CEO compensation as a consequence of industrial

firm mergers compared to bank mergers. In line with this argument, our DiD analysis shows that, relative to matched

mergers in industrial firms, bank mergers significantly increase CEO total and equity compensation but decrease CEO

delta.

During the global financial crisis (GFC), banks experienceddramatic financial losses. Despite a positive merger–pay

relationship existent in most of our sample period, bank mergers significantly reduced CEO compensation during the

breakoutof the GFC. The dramatic losses suffered by banks, and the resulting illiquidity that plagued the real economy,

demanded that policymakers takea closer look at the roots of the crisis. One common view is that the crisis is at least

partly the result of exorbitant risk taking bybanks (Bhagat & Bolton, 2014), which warrants regulative changes in the

banking sector to incentivize bank managers properly through compensation design. Notwithstanding the Dodd-Frank

Wall StreetReform and Consumer Protection Act of 2010 that created new rules to restrict managerial compensation in

financial institutions, we document a positiveeffect of merger growth on CEO compensation and vega in the post-crisis

period. This finding suggests that investing in M&As is one way for bank CEOs to counter the regulatory intention of

curbing excessiveexecutive pay and risk taking.