The wealth effects of mergers and acquisitions by dividend payers

Journal Pre-proof The Wealth Effects of Mergers and Acquisitions by Dividend Payers Mina Glambosky, Surendranath Rakesh Jory, Thanh Ngoc Ngo

PII:

S1062-9769(20)30013-2

DOI:

https://doi.org/10.1016/j.qref.2020.01.013

Reference:

QUAECO 1339

To appear in:

Quarterly Review of Economics and Finance

Received Date:

5 October 2018

Revised Date:

15 November 2019

Accepted Date:

29 January 2020

Please cite this article as: Glambosky M, Jory SR, Ngo TN, The Wealth Effects of Mergers and Acquisitions by Dividend Payers, Quarterly Review of Economics and Finance (2020), doi: https://doi.org/10.1016/j.qref.2020.01.013

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier.

The Wealth Effects of Mergers and Acquisitions by Dividend Payers Mina Glambosky1, Surendranath Rakesh Jory2, Thanh Ngoc Ngo3 1 Corresponding

author: Mina Glambosky, Brooklyn College CUNY, Brooklyn, NY, United States.

1 Corresponding

ro

of

Email [email protected]; Telephone: +1 718 951 3550 2 Surendranath Rakesh Jory, University of Southampton, SO17 1BJ, United Kingdom. Email [email protected]; Telephone: +44 23 8059 5923 3 Thanh Ngo Ngoc, East Carolina University, Greenville, NC 27858-4353, United States. Email [email protected]; Telephone: + 1 252 328 4038

author: Mina Glambosky, Brooklyn College CUNY, Brooklyn, NY, United States.

-p

Email



Jo

Abstract

lP



The market exhibits a more favorable market reaction to the announcement of mergers and acquisitions (M&As) by dividend-paying acquirers when compared to non-paying acquirers. Dividend-paying acquirers are associated with greater improvements in return on assets post-M&A. Dividend-paying acquirers hold higher levels of cash and are more likely to engage in cash financed deals.

ur na



re

Highlights

We document a more favorable market reaction to the announcement of mergers and acquisitions (M&As) by dividend-paying acquirers when compared to non-paying acquirers. Dividendpaying acquirers are associated with greater improvements in return on assets post-M&A. Furthermore, dividend-paying acquirers hold higher levels of cash and are more likely to engage in cash financed deals. We infer that an existing dividend payment policy acts as a disciplinary

mechanism forcing managers to engage in value-adding M&As, restricting self-motivated empire-building acquisitions. In order to preserve their ability to maintain dividends, dividend paying acquirers seek targets that can contribute to free cash flow.

JEL Classification: G34; G35

of

Keywords: Dividends; Dividend Policy; Mergers and Acquisitions; Takeovers; Event Study

ro

The Wealth Effects of Mergers and Acquisitions by Dividend-Payers

-p

1. Introduction

re

Prior studies document negative to zero wealth effects for acquiring firms upon the merger and acquisition (M&A) announcement. Empire-building by egocentric managers is often

lP

cited as an explanation for the poor results. Consequently, the search for disciplining mechanisms to limit managerial discretion may be value adding. Previous research has identified

ur na

debt as a disciplining mechanism, the presence of interest and debt payments restrict possible misuse of free cash flow by management (see Myers, 1977; Hennessy, 2004; Diamond and He, 2014). Our paper contributes to the literature on M&A disciplining mechanisms by examining how an existing acquirer dividend payment policy can align shareholders’ and managers’

Jo

interests. Analysis of dividend policy continues to be a frequent topic of research as a definitive explanation of the impact to the firm from electing to pay a dividend has not been identified. We add to this body of literature by focusing on the mitigating impact of the dividend policy on M&A transactions. The primary inquiry of this study is whether or not investors distinguish between M&A transactions by dividend paying and non-paying bidders. Does the dividend

policy decision effect the market’s perception of the M&A transaction and the return to the dividend paying bidder? We hypothesize that M&A announcements by dividend-paying acquirers should elicit more favorable reactions. An established dividend payment policy acts as a disciplinary mechanism that forces managers to engage in value-adding M&As, limiting self-motivated empire-building. Dividend paying acquirers seek targets that can contribute to free cash flow in

of

order to preserve their ability to pay dividends. We use three theories to explain our supposition

ro

(i) the life cycle theory, (ii) signaling theory and (iii) agency theory as follows.

The lifecycle theory describes that mature firms with fewer investment opportunities tend

-p

to retain higher levels of free cash flow. Thus, they are more likely to pay out dividends, rather

re

than investments (De Angelo et al., 2006; Gizelle et al., 2013; Subba et al., 2015). However, over time, pressure from investors spurs a continuous search for investment opportunities,

lP

including M&As. Managers seek to avoid value-destroying M&As that can negatively affect their ability to maintain the existing dividend payout. Management of dividend paying acquirers

ur na

pursue value-enhancing M&A opportunities that can help to maintain the existing dividend payout and potentially boost firm growth. As firms progress through the lifecycle and elect to pay a dividend the dividend paying firms use of cash is constrained by the payout policy. We seek to determine if this constraint effects management’s choice of target and thus the return to

Jo

the dividend paying bidders. The risk of overinvesting or empire-building is minimized in the presence of dividends. The signaling theory of dividends states that managers, unlike shareholders, are privy to

superior private information. Corporate transactions, such as dividend payments, share repurchases, and takeovers, represent managers’ attempts to communicate their firms’ outlook to

shareholders. Empirical tests of the signaling theory offer mixed results (Allen and Michaely, 2003; Li and Zhao, 2008; Richard, 2010; How et al., 2011; Reza et al., 2014; Baker and Sujata, 2015; Subba 2015; Dewasiri et al., 2019). While the literature fails to document significant improvements in the operating performance of dividend-increasing firms (see DeAngelo et al., 1996, Benartzi et al., 1997, and Grullon et al., 2005), survey-based results appear to be more supportive of the signaling theory (Baker et al., 2011). Extending the signaling theory,

of

acquisitions by dividend-paying firms should signal managerial confidence in the expected

ro

synergy of the combined firms. We examine if the dividend payment decision signals

information that increases investor confidence in the future outlook of the firm. Identifying

-p

whether investors differentiate dividend paying bidder acquisition returns may determine if

re

investors perceive that management of dividend paying bidders are better able to identify value adding targets. We seek to determine if the positive signal sent by the dividend payment

lP

manifests in greater returns from the acquisition process. The decreased information asymmetry resulting from the dividend payout policy is motivation for improved confidence in the ability of

ur na

management.

The agency theory (Jensen, 1986) identifies that managers of firms with excess free cash flows tend to spend the cash on “empire-building” acquisitions, maximizing management welfare to the detriment of shareholders. Accordingly, consistent with the bird-in-the-hand

Jo

theory of Lintner (1956) and Gordon (1959), investors prefer that these firms distribute the excess cash by paying dividends. Previous studies have examined dividend payment policy and the agency cost theory and found support for dividends as a disciplining mechanism (Claudiu and Marilen, 2014; Baker and Sujata, 2015; Subba and Dollery, 2015). The lifecycle theory adds to the agency theory with the inclusion of the firm’s investment opportunity set. Based on these

theories we expect dividend-paying firms to make judicious use of cash and for investors to positively view the M&As they engage in. An existing dividend payout policy prevents managers from diverting funds toward self-gratification, i.e. private benefits of control. In line with the agency theory, an existing dividend payout reduces the inefficiency of marginal investments by the bidder firm (La Porta, Lopez-de-Silanes, Shleifer and Vishny, 2000) and boosts the value of new acquisitions. Similarly, La Porta et al. (2000) find that greater dividend

of

payouts afford better protection of the rights of minority shareholders. Entrenched managers who

ro

take actions that are in their own self-interest, rather than the best interests of shareholders,

would prefer to maintain higher levels of cash. Additionally, investors are concerned about

-p

“empire building” and hubris on the part of management. This study is motivated by the need to

re

examine whether the existence of a dividend payout policy is reflective of constrained management who are more deliberate in their use of cash and selection of M&A targets.

lP

We construe from the theories above that the M&A choices of dividend-payers are less likely to be motivated by “empire-building”. The adverse market reaction to the M&A

ur na

announcement, previously documented in the literature, should be reduced for dividend paying acquirers. Our sample includes US domestic M&A deals valued over $1 million for publiclylisted companies over the period 1985-2014. We compare investor reaction to M&A announcements for dividend-paying vs. non-paying acquirers. We find that investors react more

Jo

favorably to M&A announcements by dividend-paying acquirers. The favorable market reaction for dividend-payers is logical given that dividend-payers are associated with higher changes in operating performance post M&A, measured by the return on assets. We also find that dividend-paying acquirers hold higher levels of cash and are more likely to engage in cash financed deals. Consequently, the favorable market reaction may be

driven by the use of a cash payment rather than the disciplining role of dividends. Turki and Dereeper (2012) documented negative returns for dividend paying acquirers who pay with stock. We control for the cash effect and still find consistently superior performance for dividend paying acquirers. The positive market reaction to the announcement of acquisitions by dividend paying acquirers is incremental to the cash-financed effect. Firms can utilize various sources of cash to finance M&As in addition to their cash reserves. Cash generated internally from

of

operations and other sources, i.e. asset sales, additional borrowing and issuance of new equities,

ro

are not all structured to sustain long running dividend payments. We contend that the

incremental effect of dividend payments is as follows, judicious acquisitions add significantly to

-p

the acquirer's free cash flow and support dividend payments. This aligns with our hypothesis that

re

dividend payers engage in M&As that are less likely to benefit managers over shareholders. The remainder of the paper is structured as follows. Section 2 summarizes the relevant

lP

literature and develops our hypotheses. Section 3 explains the data and sample descriptive statistics. Section 4 presents our empirical findings and discussion, and section 5 concludes the

ur na

paper.

2. Literature Review and Hypotheses Development A body of literature exists that has examined M&As in conjunction with dividends, and

Jo

our hypotheses build upon the relevant literature. The impact of dividend policy on the method of the M&A transaction payment was examined by Turki and Dereeper (2012). They find that dividend-paying acquirers are more likely to finance acquisitions of publicly-traded targets with stock payments. The authors also test the market reaction to acquisitions of publicly-traded firms based on the method of payment only. They include sub-samples of dividend-payers and non-

payers and find a negative market reaction in both groups for M&As financed with stock payments. Turki and Dereeper (2012) find that the choice of method of payment by dividendpayers fails to affect the market valuation of the M&A compared to non-payers. Our study directly tests the effect of bidders’ dividend policy on the value of the M&A transaction. Czerwonka (2011) finds that acquirers do not manage their dividend policies in anticipation of a merger in a sample of 21 companies on the Warsaw Stock Exchange (WSE).

of

The author hypothesizes that a cash payment necessitates the accumulation of funds prior to the

ro

transaction, which may cause the bidder to cease dividend payments. However, the empirical findings fail to support the proposition. Examining a sample of 38 companies listed on the WSE,

-p

Czerwonka (2012) finds that acquirers tend to pay lower dividends compared to non-acquirers

re

post-M&A.

Jeon et al. (2010) find that similarity in the dividend policies of the target and the

lP

acquirer increases the likelihood of a stock payment. Dissimilarity in their dividend policies adversely impacts the M&A announcement returns, although the likelihood of a cash payment

ur na

increases. Dereeper and Turki (2013) find that the likelihood of deal completion increases in the presence of similarity in dividend policies between the two parties. The authors argue that a similarity in dividend policies for stock payment acquisitions would face less resistance from target shareholders, as their dividend income needs are still served. Dereeper and Turki (2013)

Jo

focus on the effect of the similarity and dissimilarity in dividend policies between the two parties in an M&A. We extend on the existing literature by examining the M&A valuation effects of the bidders’ dividend policy. Dereeper and Turki (2016) examine the relationship of pre-merger dividend policy of the target and acquirer on post-merger dividend policy. Using a sample of stock-based acquisitions, they

find the post-merger dividend status and payout ratio of the combined entity is directly related to the target firm’s policies prior to the M&A. The authors conclude that bidders alter their dividend policy to satisfy target shareholders. Hypotheses M&A Announcements – The opportunity costs for dividend paying acquirers can be substantial as cash dividends reduce funds available for M&As. Managers have greater motivation to pursue

of

value-adding M&As so as not to jeopardize their capability to maintain the existing dividend

ro

payout. Value-adding M&A opportunities can help meet the dual objective of increasing firm growth and contributing to the maintenance of the existing dividend payout. The cash obligation

-p

of dividend payments increases the likelihood of dividend payers to seek external funds

re

(Easterbrook, 1984; La Porta et al., 2000); this exposes the firm to additional scrutiny from external investors. To the extent that the high opportunity costs of dividends act as a disciplinary

lP

mechanism on management, dividend-payers face higher constraints on their ability to use free cash flow in value-destroying M&As. Our first hypothesis examines market reaction to M&A

ur na

announcements, specifically if dividend-payers are more favorably received by investors compared to non-paying firms. H1:

The market reaction to M&A announcements by dividend-paying acquirers is positive.

Jo

Consistency of Dividend Policy – Investors in dividend paying firms have an expectation that the acquirer will maintain their dividend policy. If investors can perceive an increased likelihood that an acquirer will maintain their dividend policy they will place a greater value on the M&A transaction announcement. We therefore expect the market to react more positively toward firms that are expected to maintain their dividend payout preferences post-M&A. Specifically, ex ante

investors can only deduce bidders’ dividend intentions by looking at their recent changes in dividend policy. Previous literature suggests that positive market reactions are associated with the announcement of dividend initiations and/or dividend increases (Bhattacharya, 1979; John and Williams, 1985; Miller and Rock, 1985, and Ambarish, John and Williams, 1987, among many others). Conversely, firms that announce dividend decreases and/or cancellations suffer from

of

adverse market reaction (Aharony and Swary, 1980; Asquith and Mullins, 1983). We examine

ro

whether the market reaction to M&A announcements is impacted by changes in bidders’

dividend policy immediately prior to the M&A announcement. We hypothesize that firms with a

-p

track record of maintaining a stable dividend policy will experience higher announcement period

H2:

re

CARs.

Bidders with a consistent dividend policy experience higher M&A announcement CARs,

ur na

lP

relative to bidders that have recently changed their dividend policy.

Post-M&A Operating Performance – Firms subject to the disciplinary mechanism of dividends face greater pressure to pursue value-adding M&As. Subsequently, the takeover activities of dividend-payers should lead to improved operating performance post-M&A. In contrast, non-

Jo

paying bidders are not subject to the restraints on agency cost-based incentives for free cash flow disbursement and retention. The disbursement of free cash flows, in the form of a dividend, mitigates investor concerns about overinvestment. H3:

The post-M&A operating performance of dividend-paying acquirers exceeds that of nonpaying acquirers.

Method of Payment – Firms can disburse free cash flows by paying dividends and/or making new investments. A dividend payout policy reduces the free cash flow available for investments, and vice versa. As such dividend paying firms are more likely to finance a bid with stock rather than cash. Consequently, we test the propensity of dividend-paying acquirers to engage in stockfinanced deals.

of

Dividend-payers are more likely to finance an M&A deal with stock rather than cash.

ro

H4:

3. Sample and Data Description

-p

We collect M&A announcement dates, deal, bidder and target characteristics from the

re

Securities Data Company (SDC) US M&A Database. Our sample period includes announcements from 1985 through 2014. We limit the sample to completed domestic deals

lP

valued over $1 million between publicly-listed bidders and targets, where the acquirer becomes the majority shareholder post-M&A. We exclude firms with SICs 4900-4999 and 6000-6999, as

ur na

they are highly regulated. Accounting data is obtained from COMPUSTAT and share price data is obtained from the Center for Research in Security Prices (CRSP) database. The final sample consists of 1,844 M&A transactions.

The sample description is provided in Table 1. There are 411 dividend-paying targets and

Jo

894 dividend-paying acquirers, and the remaining firms do not pay dividends. We find that 36% of the deals are financed by cash only and 32% are financed by stock only, 39% are related deals. Our sample includes few hostile deals, 1.68% of the overall sample. High-tech firms account for a significant percentage of the targets, 58.57%. [Insert Table 1 about here]

We present additional deal, bidder and target characteristics for the whole sample in Panel A of Table 2. The full sample exhibits a mean deal value of $1,351 million, cash financed deals represent 47% of the sample. The mean bidder’s market capitalization is $18,004 million and the mean target’s market capitalization is $904 million. The bidder’s mean ROA, cash ratio and market-to-book ratio are 5.1%, 1.5% and 3.136, respectively. The respective figures for target firms are -3%, 6.6% and 4.852%.

of

We split the sample between dividend-paying and non-paying acquirers in Panel B of

ro

Table 2. The mean (median) values of the deals completed by dividend-payers and non-payers are $2,145 million ($425 million) and $706 million ($186 million), respectively. The mean

-p

(median) values of the percentage of cash financed deals by dividend-payers and non-payers are

re

59% (37%) and 85% (0%), respectively. The evidence suggests that dividend-paying acquirers are not cash-constrained relative to non-paying acquirers, which casts doubts on previous

lP

findings that dividend-paying acquirers are more likely to pay in stock. Our growth proxy, the market-to-book ratio, shows a mean of 1.937% vs. 4.131% for dividend-paying acquirers and

ur na

non-paying acquirers. This aligns with the lifecycle hypothesis that dividend-payers tend to be low-growth mature firms.

[Insert Table 2 about here]

Jo

4. Empirical Findings

4.1. Acquirer’s dividend payout and M&A announcement returns In Table 3, we present the results of univariate tests of the stock market reaction to M&A

announcements of dividend-paying and non-paying acquirers. We present both market model cumulative abnormal returns (CARs) and the Fama-French three-factor model CARs. The CARs

represent the summation of daily abnormal returns, error terms from either the market or FamaFrench model, calculated from stock price data retrieved from the CRSP database. We designate the date of the M&A announcement as day 0, and use the CRSP equally-weighted index daily returns to proxy for the market portfolio in both models. Consistent with hypothesis H1, all reported CARs are higher for dividend-paying acquirers, the difference between the subsamples

[Insert Table 3 about here]

ro

In Table 4, we present the results of the following regression model:

of

range from 1.350% to 1.535%, significant at the 1% level.

CARi =  + 1ACQDIVi + 2TGTDIVi + 3ACQDIVi*TGTDIVi + 4ACQRUNUPi +

-p

4DEALVALUEi + 6CASHi+ 7HOSTILEi + 8COMPETINGi + 9RELATEDi +

re

10TGTHITECHi + 11ACQROAi+ 12ACQCASHi+ 13ACQMKBKi + i where CARi represents the cumulative abnormal return of the acquiring firm i. Our variable of

lP

interest is the bidder’s dividend policy, ACQDIV, a dummy variable set to 1 for acquirers that paid dividends in the year preceding the M&A announcement. In separate regressions, we

ur na

replace ACQDIV with ACQDIVYIELD the ratio of the bidder’s dividend per share to market price per share at the end of the fiscal year preceding the M&A announcement. To account for the nonlinear relationship between the bidder’s dividend payout and the M&A performance, we include the squared term of the bidder’s dividend yield, ACQDIVYIELD2.

Jo

Jeon et al. (2010) find that dissimilarities in dividend policies of the bidder and target

adversely affect the M&A CARs. In addition, Dereeper and Turki (2013) find that similarities in dividend policies of the bidder and target increase the likelihood of deal completion. Dereeper and Turki (2016) also find that the entity’s combined dividend policy post-M&A is related to the target’s dividend policy in the period pre-M&A. These findings suggest that the target’s dividend

policy affects the wealth implications of M&As. It is important to examine whether the M&A announcement returns for dividend vs. non-paying bidders stems from the market’s anticipation of unchanged or higher dividend payments post acquisition, results may denote an endogeneity problem. As previous studies have documented that the post-M&A policy is related to the target’s dividend policy support of the endogeneity problem would be reflected in significantly higher returns for acquisitions of dividend paying targets. Consequently, we examine the effect

of

of the target’s dividend policy using TGTDIV and TGTDIVYIELD. TGTDIV is a dummy

ro

variable representing target firms that paid dividends in the year preceding the M&A

announcement. TGTDIVYIELD is the ratio of the target’s dividend per share to market price per

-p

share at the end of the fiscal year preceding the M&A announcement.

re

The inclusion of the remaining control variables is guided by prior studies. ACQRUNUP represents the acquirer stock price run-up in the (-30, -10) window prior to the M&A

lP

announcement, day 0. DEALVALUE represents the natural logarithm of deal value; CASH is a dummy variable representing deals that are paid in cash only; HOSTILE is a dummy variable

ur na

representing hostile deals; COMPETING is a dummy variable representing contested deals; RELATED is a dummy variable representing deals where the bidder and target share the same four-digit SIC code; TGTHITECH is a dummy variable representing high-tech target firms based on the SDC database classification; and ACQROA, ACQCASH and ACQMKBK represent the

Jo

industry-adjusted return on assets, cash to asset ratio and market-to-book value of equity ratio of the acquirer, measured at yearend preceding the M&A announcement. The regressions control for year-fixed effects and use heteroscedasticity-robust standard errors to compute the significance levels of coefficients.

We present our findings in Table 4. In Models 1 and 3, the dummy variable ACQDIV is positive and statistically significant at the 1% level, CARs of dividend-paying acquirers are statistically greater than CARs of non-paying acquirers. When we substitute ACQDIVYIELD for ACQDIV in Model 4, the coefficient is positive and significant at the 5% level. Confirming a more favorable market reaction to M&A announcements by acquirers with a higher dividend yield. Our results in Model 2 and 4, for the TGTDIV and TGTDIVYIELD variables show mixed

of

results with no statistical significance. Although previous literature indicates that the target

ro

dividend policy impacts the combined entity’s dividend policy choice (Dereeper and Turki, 2016), it does not significantly impact the M&A announcement return.

-p

[Insert Table 4 about here]

re

The coefficient of the ACQDIVYIELD2 variable in Model 4 is negative and significant, which suggests a nonlinear relationship between dividend yields and CARs. However, the

lP

variable loses significance in tests of endogeneity in Table 5. The curvilinear relationship can be explained by the lifecycle theory. The lower the amount by which a firm’s internal funds exceed

ur na

its M&A opportunities, the lower its dividend yield. Investors would associate a low dividend yield with a greater availability for profitable investment opportunities, accordingly they value M&A announcements by these firms. This is exhibited by the positive relationship between lower dividend yields and high stock returns. Conversely, higher dividend yields significantly

Jo

constrain managements’ ability to participate in the M&A market. The negative investment constraints of high dividend payouts inhibit managers’ selection of value-enhancing risky projects that could potentially undermine the firm’s ability to maintain existing dividend levels. The coefficients of the control variables are consistent with prior studies: acquirer CARs are positively related to (i) the run-up in the acquirer’s stock price in the days leading up to the

M&A; (ii) cash-only bids (consistent with Eckbo, Giammarino, and Heinkel, 1990, and Eckbo and Thorburn, 2000) and (iii) acquirer market-to-book ratio. Conversely, the announcement period acquirer CARs are lower for larger deals, consistent with Moeller, Schlingemann and Stulz (2005) who document larger losses to acquirers for larger transactions. The acquirer CARs are lower for acquirers with more cash at hand, reflecting negative market perception of potential agency problems with free cash flow (Jensen, 1986). For example, managers of firms with larger

of

free cash flows may be more likely to use excess cash to empire build in order to maximize their

ro

compensation. We find lower CARs for acquisitions of high-tech targets, similar to the findings of Benou and Madura (2005) who document negative valuation effects for acquirers of public

re

-p

high-tech targets.

4.2. Bidder’s dividend payout and M&A announcement returns using propensity-matching

lP

analyses

The results we document in Tables 3 and 4 may be driven by fundamental differences

ur na

between dividend-paying and non-paying acquirers. Dividend-paying acquirers may display characteristics that improve their ability to identify value-enhancing targets. To control for such endogeneity, we perform propensity-matching analysis to validate the results from Tables 3 and 4. Li (2013) argues in favor of this method to control for endogeneity and to produce unbiased

Jo

causal effect in studies of management practices such as dividend policy which use observational, rather than experimental data. Using the universe of non-financial and non-utility firms from COMPUSTAT for 1985-2014, we first estimate the probability of a firm to pay dividends following DeAngelo et al. (2008). The probability of a firm to pay dividends is a function of (1) retained earnings to assets ratio (2) return on assets (3) sales growth and (4) firm

size. We extract the predicted probability from the Probit regression. Each dividend-paying acquirer is then matched with one non-paying acquirer based upon its closest predicted probability to pay dividends in the year preceding the acquisition. In Panel A of Table 5, we compare the CARs of the dividend-paying acquirers with the propensity-matched non-paying acquirers. Consistent with the results in Table 3, dividendpaying acquirers experience significantly higher CARs. In Panel B of Table 5, we present the

of

OLS regression results for the dividend-paying acquirer’s CARs in the (-2, +1) window and the

ro

propensity-matched non-paying acquirers. Consistent with the results in Table 4 and in support of Panel A of Table 5, dividend-paying acquirers experience significantly higher CARs.

re

-p

[Insert Table 5 about here]

4.3. Acquirer’s dividend changes and M&A announcement returns

lP

Due to the signaling implications of dividend changes (see Bhattacharya, 1979; Aharony and Swary, 1980; Asquith and Mullins, 1983; John and Williams, 1985; Miller and Rock, 1985;

ur na

and Ambarish, John and Williams, 1987, among others), investors do not view firms that decrease dividends in the same manner as they view those that increase dividends. Additionally, the results for dividend paying acquirers may be related to the market’s perception of dividend payments post M&A. Whether investors perceive that the acquisition will negatively or

Jo

positively impact the dividend payment may influence the announcement results. At the time of the M&A announcement, investors cannot observe the bidder’s dividend policy post-M&A. Nonetheless, investors can seek clues from the existing dividend policy of the bidder. Examining bidders who change their dividend policy pre-acquisition may help shed light on differences in M&A announcement returns for dividend paying bidders.

In Table 6, we examine the wealth effects of dividend changes by dividend-paying acquirers in the year preceding the M&A announcement. We create three dummy variables, ACQ_INIT, ACQ_INCR and ACQ_UNCHG, representing acquiring firms that initiated dividends, increased dividends and kept their dividend payout unchanged in the year preceding the M&A announcement. We run the following regression model, the intercept captures the effect of dividend-decreasing acquirers and the remaining variables are as defined earlier:

of

CARi =  + 1ACQ_INITi + 2ACQ_INCRi + 3ACQ_UNCHGi+ 4ACQRUNUPi +

ro

4DEALVALUEi + 6CASHi+ 7HOSTILEi + 8COMPETINGi + 9RELATEDi +

-p

10TGTHITECHi + 11ACQROAi+ 12ACQCASHi+ 13ACQMKBKi + i

We also test for the effect of bidder dividend increases greater than 5%, ACQ_INCR5,

re

consistent with Amihud and Li (2006) and Andres and Hofbaur (2017) and changes in the acquirer’s dividend yield, ACDDIVYIELDCHG. Our findings are presented in Table 6. In

lP

Model 1, the coefficient of the variable ACQ_INCR is negative, suggesting that acquirers that recently increased dividends are unlikely to perform value-enhancing M&As. However, the

ur na

coefficient is only marginally significant at the 10% level. A decision to change the dividend policy immediately prior to an M&A announcement may be difficult for shareholders to interpret, the decision to increase the dividend may be a consequence of the impending M&A. [Insert Table 6 about here]

Jo

Hypothesis H2 posits that dividend clientele would reward acquisitions that leave the

bidder’s dividend payments unchanged. In Model 1, consistent with hypothesis H2, the coefficient of the variable ACQ_UNCHG is positive and significant at the 5% level, investors reward acquisitions by bidders that maintain their existing dividend policy. The variable ACQ_UNCHG is still positive but marginally significant in Model 2. Likewise, bidders that

initiate dividends in the preceding year are weakly associated with positive M&A announcement period CARs. However, the remaining variables that measure dividend changes at bidder firms are not statistically significant. Taken together, the evidence suggests that it is the presence of dividends rather than the level of dividends that affect investors’ valuation of M&As most consistently.1 Moreover, M&As that are likely to leave the bidder’s dividend policy unchanged are better received by shareholders.

of

4.4. Bidders’ post-M&A operating performance

ro

Insofar as dividend payouts serve as a disciplinary mechanism to prevent self-interested managers from engaging in value destroying acquisitions, we should observe better post-M&A

-p

operating performance for dividend-paying acquirers. To test this proposition, we compare the

re

changes in industry-adjusted return on assets (ROA) prior and subsequent to the M&A announcement for dividend-paying and non-paying acquirers. The results are reported in Panel A

lP

of Table 7. While both groups of acquirers exhibit values below the industry-adjusted ROAs post M&A, dividend-paying acquirers display significantly better operating performance relative to

ur na

non-paying acquirers. [Insert Table 7 about here]

In Panel B of Table 7, we perform a cross-sectional analysis of the one- and three-year changes in the bidders’ ROA from prior to subsequent to the M&A announcement. The

Jo

coefficient of the dummy variable ACQDIV is consistently positive and statistically significant, suggesting that the ROA performance of dividend-paying acquirers is superior. Accordingly, the initial market sentiment at the time of the M&A announcement is reasonable. The dividendpaying acquirers’ better operating performance supports the favorable market sentiment for

1

This argument is similar to Jory, Ngo and Wang (2016) who find that, in M&As, it is the presence of credit ratings rather than changes in the ratings or the level of the ratings that affects investors’ valuations.

dividend paying acquirers’ M&A announcements, consistent with hypothesis H3. The evidence suggests that the superior operating performance by dividend-paying acquirers is consistent with dividends acting as a disciplinary mechanism that motivates managers to pursue value-enhancing M&As.

4.5. Methods of payment

of

The agency theory describes that firms with excess cash may be prone to empire-building

ro

and the pursuit of value-destroying acquisitions. Firms that distribute free cash to shareholders in the form of dividends, contributing to reducing the agency problem, limit managerial

-p

profligacies. Once managers commit to paying a dividend they seek to maintain the dividend

re

payout and avoid a negative market reaction to a dividend reduction. Consequently, managers of dividend-paying firms are motivated to preserve cash and avoid non-value-enhancing M&As.

lP

The impact of free cash on the dividend-paying acquirers’ decision to engage in M&As is uncertain. Dividend-paying acquirers may use their excess cash to pay dividends and more likely

ur na

finance M&A transactions with stock. To the degree that dividend-paying acquirers are cashconstrained they would be disadvantaged in acquisitions where target firms prefer cash. Our data suggests that dividend-paying acquirers are no less active than non-paying acquirers in the market for corporate control. Based on our data selection criteria we retrieved 894 M&As by

Jo

dividend-paying acquirers and 950 M&As by non-paying acquirers over the 1985 to 2014 period, from the SDC database. Conversely, the lifecycle theory suggests that dividend-paying acquirers are mature companies with larger cash holdings and could capably use cash to acquire targets. The empirical evidence better fits the expectations of the lifecycle theory. First, in Panel A of Table 8 we show that both the mean and median percentages of cash payment by dividend-

paying acquirers are higher than the corresponding figures for non-paying acquirers. The mean and median values of the percentages of cash paid by dividend-paying and non-paying acquirers are 59% (37%) and 85% (0%), respectively. The differences are statistically significant at the 1% level. Second, we run a Probit regression for the cash method of payment and present our findings in Models 1 and 2 in Panel B of Table 8. The dependent variable is set to a value of one

of

for cash-only transactions, zero otherwise. The coefficients of both ACQDIV (Model 1) and

ro

ACQDIVYIELD (Model 2) are positive and significant at the 1% level, suggesting dividendpaying firms are more likely to use cash as the method of payment. However, the association

-p

between the dividend variables and cash payment is not linear. High levels of dividend payouts

re

reduce the likelihood that firms will pay cash. The coefficient of ACQDIVYIELD2 is negative and statistically significant in Model 2. The cost pressure on the firm’s cash, stemming from

lP

dividend payout, effects the method of payment. [INSERT TABLE 8 ABOUT HERE]

ur na

Additionally, we perform ordinary least squares (OLS) regressions in Models 3 and 4. The dependent variable in the OLS models is the percentage of the payment paid in cash. Similar to the previous results, the coefficients for both ACQDIV (from Model 3) and ACQDIVYIELD (Model 4) are positive and highly significant. This supports the positive relationship between the

Jo

bidders’ dividend payment policy and their use of cash in M&A transactions. In contrast, at higher levels of dividend payment, captured by the ACQDIVYIELD2 variable, the dividendpaying acquirers are associated with a lower percentage of cash as the method of payment. Both the univariate tests and the Probit regressions fail to support hypothesis H4. There is little

evidence to suggest that dividends constrain the bidders’ ability to use cash in the M&A market, excluding those acquirers with relatively high dividend yields.

5. Conclusion In this paper, we examine whether the bidders’ dividend policy serves to convey information to investors about the quality of their bids. Our findings are consistent with the

of

agency theory, dividend payments serve to reassure investors that managers will be constrained

ro

from using excess cash in empire-building M&As. Consistent with the theory and the benefits of using dividends as a disciplinary mechanism, we document a more favorable M&A

-p

announcement period market reaction to bidders that pay dividends relative to bidders that do not

re

pay dividends. This market sentiment appears to be astute as the post-acquisition operating performance of dividend payers is superior. There is no evidence to suggest that dividend payers

lP

are cash-constrained, excluding firms that distribute a high proportion of the value they generate in dividends. Both the probability to pay cash and the percentage of cash payment are higher for

ur na

dividend paying acquirers. This finding is consistent with the lifecycle theory, which suggests that dividend payers tend to be mature firms with significant amounts of accumulated cash who may self-finance M&A transactions.

In accordance with the efficient approach to marginal investments by dividend-paying

Jo

acquirers, empirical evidence suggests that these firms have profitable operations. For instance, Fama and French (2001) find that dividend-payers are more profitable than non-payers. The authors document a ratio of annual earnings before interest-to-assets of 7.82% for payers versus 5.37% for non-payers. The gap is wider when using earnings available for distribution to common stockholders, 12.75% for dividend payers versus 6.15% for non-payers. DeAngelo et

al. (2004) find that the largest dividend-payers over the period 1978-2000 have the largest increases in real earnings. Our empirical study finds that dividend-paying acquirers can be more efficient than non-paying acquirers in the market for corporate control.

Declaration of interests

Jo

ur na

lP

re

-p

ro

of

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References Aharony, J., & Swary, I. (1980). Quarterly dividend and earnings announcements and stockholders' returns: An empirical analysis. The Journal of Finance, 35(1), 1-12. Allen, F., & Michaely, R. (2003). Payout policy. Handbook of the Economics of Finance, 1, 337429. Ambarish, R., John, K., & Williams, J. (1987). Efficient signaling with dividends and investments. The Journal of Finance, 42, 321–343.

of

Amihud, Y., &Li, K., 2006. The declining information content of dividend announcements and the effects of institutional holdings. Journal of Financial and Quantitative Analysis, 41, 637-660.

ro

Andres, C., & Hofbaur, U. (2017). Do what you did four quarters ago: Trends and implications of quarterly dividends. Journal of Corporate Finance, 43, 139-158.

-p

Arnott, Robert D., & Clifford S. Asness. 2003. Surprise! Higher Dividends = Higher Earnings Growth. Financial Analysts’ Journal, 59(1) (January/February), 70–87.

re

Asquith, P., & Mullins Jr, D. W. (1983). The impact of initiating dividend payments on shareholders' wealth. Journal of Business, 56, 77-96.

lP

Baker, H.K., Singleton, J.C. & Veit, E.T. (2011). Survey Research in Corporate Finance – Bridging the Gap between Theory and Practice Oxford University Press, New York. Baker. H.K., & Sujata K., (2015). "Dividend policy in India: new survey evidence", Managerial Finance, 41 (2), 182 – 204.

ur na

Baker, H. K., & Weigand, R. (2015). Corporate dividend policy revisited. Managerial Finance, 41(2), 126-144. Benartzi, S., Michaely, R., & Thaler, R. (1997). Do changes in dividends signal the future or the past? The Journal of Finance, 52(3), 1007-1034.

Jo

Benou, G., & Madura, J. (2005). High tech acquisitions, firm specific characteristics and the role of investment bank advisors. Journal of High Technology Management Research, 16, 101–120. Bhattacharya, S. (1979). Imperfect Information, Dividend Policy, and 'the Bird in the Hand' Fallacy. Bell Journal of Economics 10 (Spring), 259-270. Claudiu, B., & Marilen, P. (2014). “Dividend Payout-Policy Drivers: Evidence from Emerging Countries”, Emerging Markets Finance and Trade, 50(4), 95-112. Czerwonka, L. (2011). Dividends and M&A transaction financing. Paper presented at the 8th International scientific conference: Financial management of firms and financial institutions. Available at:

https://www.ekf.vsb.cz/export/sites/ekf/frpfi/cs/prispevky/prispevky_plne_verze/Czerwonka_Les zek.uprav.pdf (Accessed on July 22, 2017) Czerwonka, L. (2012). Takeovers as a way of investing versus dividend payments on the Warsaw stock exchange. Journal of International Studies, 5(1), 70-76. Diamond, D. W., & He, Z. (2014). A theory of debt maturity: the long and short of debt overhang. The Journal of Finance, 69(2), 719-762.

of

DeAngelo, H., DeAngelo, L., & Skinner, D. J. (1996). Reversal of fortune dividend signaling and the disappearance of sustained earnings growth. Journal of Financial Economics, 40(3), 341-371.

ro

DeAngelo, H., DeAngelo, L., & Skinner, D. J. (2004). Are dividends disappearing? Dividend concentration and the consolidation of earnings. Journal of Financial Economics, 72(3), 425456.

-p

De Angelo, H., DeAngelo, L., & Stulz, R. (2006). “Dividend Policy and the Earned/Contributed Capital Mix: A Test of the Lifecycle Theory”, Journal of Financial Economics, 8,(1), 227-254.

re

DeAngelo, H., DeAngelo, L. & Skinner, D.J. (2008), Corporate payout policy. Foundations and Trends in Finance, 3(2/3), 95-287.

lP

DeAngelo, H., DeAngelo, L., & Stulz, R. (2006). Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory. Journal of Financial Economics, 81, 227–254.

ur na

Dereeper, S., & Turki, A. (2013). Dividend-Paying Merging Firms and M&A Deal Completion (January 13, 2013). Available at SSRN: https://ssrn.com/abstract=2200159 or http://dx.doi.org/10.2139/ssrn.2200159 (Accessed on July 22, 2017) Dereeper, S., & Turki, A. (2016). Dividend policy following mergers and acquisitions: US evidence. Managerial Finance 42 (11), pp. 1073-1090.

Jo

Dewasiri, N.J., Koralalage, W.B.Y., Azeez, A.A., P.G.S.A., Jayarathne, P.G.S.A., Duminda Kuruppuarachchi, D., & Weerasinghe, V.A. (2019) "Determinants of Dividend Policy: Evidence from an Emerging and Developing Market", Managerial Finance, 45(3), pp.413-429 Easterbrook, F. H. (1984). Two agency-cost explanations of dividends. The American Economic Review, 74(4), 650-659. Eckbo, B. E., & Thorburn, K. S. (2000). Gains to bidder firms revisited: domestic and foreign acquisitions in Canada. Journal of Financial and Quantitative Analysis, 35(01), 1-25. Eckbo, B. E., Giammarino, R. M., & Heinkel, R. L. (1990). Asymmetric information and the medium of exchange in takeovers: Theory and tests. Review of Financial Studies, 3(4), 651-675.

Fama, E. F., & French, K. R. (2001). Disappearing dividends: changing firm characteristics or lower propensity to pay? Journal of Financial Economics, 60(1), 3-43. Gizelle, F. P., Marcus, T. A., & Shelton, H.W. (2013). "Determinants of dividend policies for ADR firms", Managerial Finance, 39(12), 1155 – 1168. Gordon, M.J. (1959). Dividends, earnings and stock prices. Review of Economics and Statistics, 41(2), 99-105. Grullon, G., Michaely, R., & Swaminathan, B. (2002). Are dividend changes a sign of firm maturity? Journal of Business, 75, 387-424.

of

Grullon, G., Michaely, R., Benartzi, S., & Thaler, R. H. (2005). Dividend changes do not signal changes in future profitability. The Journal of Business, 78(5), 1659-1682.

-p

ro

Hennessy, C. A. (2004). Tobin's Q, debt overhang, and investment. The Journal of Finance, 59(4), 1717-1742.

re

How, J.C,, Ngo K., & Verhoeven, P. (2011). Dividend Initiations and Long-run IPO Performance. Australian Journal of Management, 36(2), 267-286.

lP

Jensen, Michael C. (1986). Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers. American Economic Review, 76(2) (May), 323-329. Jeon, J. Q., Ligon, J. A., & Soranakom, C. (2010). Dividend policy and the method of payment in mergers and acquisitions. Unpublished working paper. Dongguk University, Seoul.

ur na

John, K., & Williams, J. (1985). Dividends, dilution, and taxes: A signalling equilibrium. The Journal of Finance, 40(4), 1053-1070. Jory, S.R., Ngo, T. N., & Wang, D. (2016). Credit ratings and the premiums paid in mergers and acquisitions. Journal of Empirical Finance, 39, 93-104.

Jo

La Porta,R., Lopez-de-Silanes, F., Shleifer,A., & Vishny, R.W. (2000). Agency problems and dividend policies around the world. The Journal of Finance, 55, 1-33. Li, K., & Zhao, X. (2008). Asymmetric Information and Dividend Policy. Financial Management, 37(4), 673-694. Li, M. (2013). Using the propensity score method to estimate causal effects: A review and practical guide. Organizational Research Methods, 16(2), 188-226. Lintner, J. (1956). Distribution of income of corporations among dividends, retained earnings and taxes. American Economic Review, 46(2), 97-113.

Miller, M. H., & Modigliani, F. (1961). Dividend policy, growth, and the valuation of shares. The Journal of Business, 34(4), 411-433. Miller, M. H., & Rock, K. (1985). Dividend policy under asymmetric information. The Journal of Finance, 40, 1031-1051. Moeller, S. B., Schlingemann, F. P., & Stulz, R. M. (2005). Wealth destruction on a massive scale? A study of acquiring‐ firm returns in the recent merger wave. The Journal of Finance, 60(2), 757-782.

of

Myers, S. C. (1977). Determinants of corporate borrowing. Journal of Financial Economics, 5(2), 147-175.

ro

Myers, S.C., & 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.

-p

Reza, H.C., Min, M., & Jenny Z., (2014). "Information content of dividends: a case of an emerging financial market", Studies in Economics and Finance, 31(3), 272 – 290.

re

Subba R. Y., (2015)."Corporate governance ratings and the dividend payout decisions of Australian corporate firms", International Journal of Managerial Finance, 11(2), 162 – 178.

lP

Subba, R.Y., & Dollery, Y.B. (2015). "Corporate governance and financial policies", Managerial Finance, 41(3), 267 - 285.

Jo

ur na

Turki, A., & Dereeper, S. (2012). The Information Content of Acquirer Dividend Policy in Acquisitions of Publicly Traded Firms (November 18, 2012). 25th Australasian Finance and Banking Conference 2012. Available at SSRN: https://ssrn.com/abstract=2112605 or http://dx.doi.org/10.2139/ssrn.2112605 (Accessed on July 22, 2017)

Table 1. Sample distribution This table describes the sample distribution by year and by deal characteristics. We report the distribution for the whole sample and for the subsamples of deals in which the target and the acquirer, alternatively, pay dividends at the end of the fiscal year preceding the deal announcement date.

of

Dividend-paying acquirers N Percent 30 3.36 36 4.03 30 3.36 15 1.68 10 1.12 19 2.13 19 2.13 39 4.36 46 5.15 60 6.71 73 8.17 80 8.95 50 5.59 37 4.14 24 2.68 20 2.24 24 2.68 35 3.91 42 4.7 28 3.13 20 2.24 30 3.36 37 4.14 16 1.79 25 2.8 23 2.57 26 2.91 894 100

ro

-p

re

lP

Whole sample N Percent 37 2.01 46 2.49 43 2.33 26 1.41 23 1.25 35 1.9 54 2.93 85 4.61 101 5.48 129 7 159 8.62 156 8.46 129 7 104 5.64 64 3.47 73 3.96 59 3.2 72 3.9 71 3.85 65 3.52 47 2.55 51 2.77 56 3.04 25 1.36 48 2.6 42 2.28 44 2.39 1,844 100

ur na

Jo

Year 1985 1986 1987 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Total

Dividend-paying targets N Percent 20 4.87 27 6.57 16 3.89 5 1.22 4 0.97 12 2.92 8 1.95 24 5.84 19 4.62 32 7.79 31 7.54 31 7.54 25 6.08 12 2.92 13 3.16 3 0.73 9 2.19 17 4.14 18 4.38 13 3.16 9 2.19 11 2.68 10 2.43 7 1.7 11 2.68 12 2.92 12 2.92 411 100

Cash-only deals

668

36.23

157

38.2

409

45.75

Stock-only deals Related deals

591 724

32.05 39.26

91 169

22.14 41.12

186 296

20.81 33.11

Hostile

31

1.68

17

4.14

23

2.57

High-tech target

1,080

58.57

112

27.25

443

49.55

of

ro

-p

re

lP

ur na

Jo

Table 2. Sample descriptive statistics We report the summary statistics of the deal, the target and the acquirer characteristics for the whole sample in Panel A. In Panel B, we separate the sample into two groups based upon whether the acquirer pays dividend at the end of the fiscal year preceding the deal announcement date.

Panel A – Whole sample Variables

Mean

Median

25th percentile 78.542 0.000 0.000 547.324 528.935 0.000 -0.061

75th percentile 890.353 100.000 1.298 10,169.390 8,420.352 0.106 0.072

Jo

ur na

lP

re

-p

ro

of

Deal value ($ million) 1,351.871 263.727 % cash payment 47.276 41.354 % dividend yield 0.956 0.000 Acquirer market cap. 18,004.020 2,277.323 Acquirer asset 11,257.040 1,959.900 Acquirer adjusted ROA 0.051 0.036 Acquirer adjusted cash ratio 0.015 -0.003 Acquirer adjusted market-to-book ratio 3.136 0.724 -0.059 2.165 Target market cap. 904.146 167.390 51.712 622.039 Target asset 797.477 139.587 52.383 461.587 Target adjusted ROA -0.030 0.007 -0.050 0.072 Target adjusted cash ratio 0.066 0.006 -0.038 0.159 Target adjusted market-to-book ratio 4.852 0.119 -0.548 1.335 Panel B – Dividend-paying acquirer subsample versus non-dividend-paying acquirer subsample Dividend-paying Non-dividend-paying acquirers (N = 894) acquirers (N = 950) Variables Mean Median Mean Median Deal value ($ million) 2,145.938 425.048 706.154 186.860 % cash payment 59.023 85.337 37.724 0.000 % dividend yield 0.019 0.015 0.002 0.000 Acquirer market cap. 6,050.95 30,414.090 8 7,917.017 1,129.096 Acquirer asset 5,755.63 21,154.520 3 3,218.389 830.700 Acquirer adjusted ROA 0.074 0.038 0.032 0.034 Acquirer adjusted cash ratio -0.022 -0.010 0.044 0.008 Acquirer adjusted market-to-book ratio 1.937 0.782 4.131 0.639 Target market cap. 1,457.036 271.311 458.232 115.000 Target asset 1,302.153 239.744 387.770 96.593 Target adjusted ROA 0.000 0.012 -0.054 0.000 Target adjusted cash ratio 0.045 -0.001 0.084 0.030 Target adjusted market-to-book ratio 9.615 0.210 0.746 0.053

Table 3. Acquirer short-run stock price reactions depending on the firm’s dividend-paying status

-1.863% -1.808% -1.748% -1.715%

-0.328% -0.339% -0.359% -0.366%

Jo

ur na

lP

Market model (-2,+1) Market model (0,+1) Fama-French (-2,+1) Fama-French (0,+1)

re

Variable

Dividend-paying acquirers (N = 894) Mean

-p

Non-dividend-paying acquirers (N = 950) Mean

ro

of

This table reports the cumulative abnormal returns (CARs) of the acquirer stocks around the announcement dates of the mergers. CAR is measured as the difference between the cumulative realized return and cumulative expected return over (-2, +1) and (0, +1) days around the announcement date. The expected return is estimated from the market model and the Fama-French three-factor model, alternatively, using daily stock returns of the firm and daily return on the CRSP equally-weighted index as the benchmark over the (-252, -30) days window. We compare the acquirer CARs between two groups: non-dividend-paying acquirers and dividend-paying acquirers. *, ** and *** indicate the significance levels of 10%, 5% and 1%, respectively.

Difference Mean 1.535% 1.469% 1.388% 1.350%

t-stat 4.2*** 4.48*** 3.86*** 4.11***

Wilcoxonstat 4.37*** 4.35*** 4.22*** 3.98***

Table 4. Cross-sectional analyses of the acquirer short-run stock price reactions on acquirer dividend payout in the year preceding the merger

0.001 (0.351)

re

ACQDIV* TGTDIV

ACQDIVYIELD2 TGTDIVYIELD

ur na

TGTDIVYIELD2 ACQRUNUP

DEALVALUE

Jo

CASH

lP

ACQDIVYIELD

HOSTILE

COMPETING RELATED TGTHITECH ACQROA

Model 3 0.008 (0.827) 0.010 (2.984***) 0.004 (0.573) -0.007 (-1.045)

0.115 (9.205***) -0.006 (-5.930***) 0.022 (6.937***) 0.011 (1.059) 0.001 (0.192) 0.006 (1.711*) -0.015 (-4.601***) 0.003

Model 4 0.007 (0.744)

ro

Model 2 0.010 (1.155)

-p

Independent variables Model 1 Constant 0.008 (0.854) ACQDIV 0.009 (2.720***) TGTDIV

of

In this table, we perform the OLS regressions of the acquirer’s CARs in the (-2, +1) days around the announcement date on the acquirer dividend payout in the year preceding the merger. The dependent variable, CAR, is measured as the difference between the cumulative realized return and cumulative expected return over (-2, +1) days around the announcement date. The expected return is estimated from the Fama-French three-factor model, using daily stock returns of the firm and daily return on the CRSP equally-weighted index as the benchmark over the (-252, -30) days window. ACQDIV is the dummy variable for acquirers that paid dividends in the year preceding the M&A announcement. TGTDIV is the dummy variable for targets that paid dividends in the year preceding the M&A. ACQDIVYIELD is the acquirer dividend yield in the preceding year. TGTDIVYIELD is the target dividend yield in the preceding year. ACQRUNUP is the acquirer stock price run-up in the (-30, -10) days prior to the deal announcement date. DEALVALUE is the natural logarithm of the deal value. CASH is the dummy variable for deals paid only with cash. HOSTILE is the dummy variable for hostile deals. COMPETING is the dummy variable for deals with more than one bidder. RELATED is the dummy variable for deals by targets and acquirers in the same four-digit SIC codes. TGTHITECH is the dummy variable for high-tech targets as per SDC classification. ACQROA, ACQCASH and ACQMKBK are the acquirer’s industry-adjusted return on asset, cash-to-asset ratio and market-to-book ratio in the year preceding the merger announcement. Heteroskedasticity-robust t-statistics are reported in parentheses. ***, **, and* indicate the significance levels at 1%, 5% and 10%, respectively.

0.115 (9.164***) -0.005 (-5.256***) 0.024 (7.561***) 0.012 (1.154) 0.001 (0.135) 0.004 (1.375) -0.016 (-4.742***) 0.003

0.115 (9.195***) -0.006 (-5.660***) 0.022 (6.876***) 0.011 (1.044) 0.001 (0.189) 0.006 (1.738*) -0.016 (-4.470***) 0.002

0.189 (2.516**) -0.176 (-2.003**) -0.003 (-0.109) -0.003 (-0.120) 0.115 (9.154***) -0.005 (-5.609***) 0.023 (7.539***) 0.013 (1.174) 0.001 (0.198) 0.005 (1.471) -0.016 (-4.741***) 0.001

ACQCASH ACQMKBK

(0.303) -0.026 (-2.295**) 0.000 (3.259***)

(0.182) -0.022 (-1.926*) 0.000 (3.354***)

(0.113) -0.025 (-2.180**) 0.000 (3.259***)

8.97*** 0.194 Yes 1,807

9.01*** 0.191 Yes 1,807

8.59*** 0.194 Yes 1,807

8.84*** 0.192 Yes 1,807

Jo

ur na

lP

re

-p

ro

of

F-statistics Adj. R-squared Year fixed effects Observations

(0.230) -0.022 (-1.951*) 0.000 (3.382***)

Table 5. Propensity-matching analyses for acquirer CARs

re

-p

ro

of

In Panel A, we compare the cumulative abnormal returns (CARs) of the dividend-paying acquirers versus the propensity-matching non-dividend-paying acquirers. Using the whole universe of non-financial and non-utility firms in Compustat from 1985-2014, we first estimate the probability of a firm to pay dividend. We extract the predicted probability from the probit regression. Each dividend-paying acquirer is then matched with one non-dividend-paying acquirer based upon the closest predicted probability to pay dividend in the year preceding the M&A announcement. CAR is measured as the difference between the cumulative realized return and cumulative expected return over (-2, +1) and (0,+1) days around the announcement date. The expected return is estimated from the market model and the Fama-French three-factor model, alternatively, using daily stock returns of the firm and daily return on the CRSP equally-weighted index as the benchmark over the (-252, -30) days window. In Panel B, we perform the OLS regressions of the acquirer’s CARs in the (-2, +1) days around the announcement date for the pooled sample of dividend-paying acquirers and sample of the propensity-matching non-dividend-paying acquirers. The dependent variable, CAR, is measured as the difference between the cumulative realized return and cumulative expected return over (-2, +1) days around the announcement date. The expected return is estimated from the Fama-French three-factor model, using daily stock returns of the firm and daily return on the CRSP equally-weighted index as the benchmark over the (-252, -30) days window. ACQDIV is the dummy variable for acquirers that paid dividends in the three years preceding the M&A announcement. TGTDIV is the dummy variable for targets that paid dividends in the three years preceding the M&A announcement. ACQDIVYIELD is the average acquirer dividend yield in the three years preceding the M&A announcement. TGTDIVYIELD is the average target dividend yield in the three years preceding the M&A announcement. ACQRUNUP is the acquirer stock price run-up in the (-30, -10) days prior to the deal announcement date. DEALVALUE is the natural logarithm of the deal value. CASH is the dummy variable for deals paid only with cash. HOSTILE is the dummy variable for hostile deals. COMPETING is the dummy variable for deals with more than one bidder. RELATED is the dummy variable for deals by targets and acquirers in the same four-digit SIC codes. TGTHITECH is the dummy variable for high-tech targets as per SDC classification. ACQROA, ACQCASH and ACQMKBK are the acquirer’s industry-adjusted return on asset, cash-to-asset ratio and market-to-book ratio in the year preceding the merger announcement. Heteroskedasticityrobust t-statistics are reported in parentheses. ***, **, and * indicate the significance levels at 1%, 5% and 10%, respectively.

Jo

ur na

lP

Panel A – Acquirer CARs by whether the acquirer pays dividends (N=756) Dividend-paying acquirers Non-dividend-paying acquirers (N=756) (N=756) Difference Variable Mean Median Mean Median Mean Median t-stat Wilcoxon-stat Market model (-2,+1) -0.204% -0.176% -1.549% -1.417% 1.345% 1.241% 4.31*** 4.41*** Market model (0,+1) -0.213% -0.124% -1.585% -2.064% 1.372% 1.941% 5.07*** 5.81*** Fama-French (-2,+1) -0.236% -0.144% -1.545% -1.280% 1.309% 1.135% 4.25*** 4.530*** Fama-French (0,+1) -0.244% -0.158% -1.666% -1.607% 1.422% 1.449% 5.11*** 5.69*** Panel B – Cross-sectional analyses of the acquirer short-run stock price reactions using pooled dividend-paying acquirer observations and propensity matching non-dividend-paying acquirer observations. Model 1 Model 2 Model 3 Model 4 Independent variables Coef. T-stat Coef. T-stat Coef. T-stat Coef. T-stat Constant -0.015 -1.927* -0.007 -0.926 -0.017 -2.190** -0.014 -1.768* ACQDIV 0.017 7.314*** 0.018 6.562*** TGTDIV 0.010 2.862*** 0.007 1.575 ACQDIV* TGTDIV -0.007 -1.625 ACQDIVYIELD 0.543 4.174*** ACQDIVYIELD2 -1.854 -1.381 TGTDIVYIELD 0.020 0.742 TGTDIVYIELD2 -0.016 -0.875 ACQRUNUP 0.161 10.047*** 0.159 9.915*** 0.161 10.013*** 0.162 10.108*** DEALVALUE -0.005 -5.380*** -0.005 -6.110*** -0.004 -4.999*** -0.005 -5.561*** CASH 0.007 2.259** 0.007 2.081** 0.007 2.089** 0.007 2.267**

-2.897*** -3.877*** 2.852*** -4.422*** -0.276 -4.676*** 0.114

21.90*** 0.365 Yes 1,530

lP ur na Jo

-0.020 -0.020 0.011 -0.015 -0.018 -0.062 -0.000

-2.637*** -3.927*** 4.158*** -5.020*** -0.908 -4.764*** -0.028

of

25.36*** 0.383 Yes 1,530

-0.023 -0.020 0.008 -0.013 -0.005 -0.062 0.000

ro

F-statistics Adj. R-squared Year fixed effects Observations

-2.429** -3.994*** 4.305*** -5.151*** -0.914 -4.846*** -0.015

-p

-0.017 -0.020 0.012 -0.015 -0.018 -0.063 -0.000

re

HOSTILE COMPETING RELATED TGTHITECH ACQROA ACQCASH ACQMKBK

23.97*** 0.383 Yes 1,530

-0.015 -0.020 0.009 -0.014 -0.010 -0.058 0.000 23.07*** 0.373 Yes 1,530

-2.213** -3.907*** 3.294*** -4.714*** -0.510 -4.374*** 0.124

Table 6 Cross-sectional analyses of the acquirer short-run stock price reactions on acquirer dividend change in the year preceding the merger

ACQ_INCR

ACQ_UNCHG

ACQRUNUP

DEALVALUE CASH

0.012 (2.251**)

ur na

ACQDIVYIELDCHG

lP

ACQ_INCR5

Jo

HOSTILE

COMPETING RELATED

TGTHITECH ACQROA ACQCASH

Model 2 0.008 (0.781) 0.027 (1.852*)

-p

ACQ_INIT

Model 1 0.008 (0.843) 0.023 (1.567) -0.007 (-1.658*)

0.148 (8.129***) -0.005 (-4.261***) 0.010 (3.035***) 0.007 (0.669) -0.009 (-1.296) -0.000 (-0.044) -0.008 (-2.080**) -0.028 (-1.602) -0.023

Model 3 0.015 (1.627)

re

Independent variables Constant

ro

of

In this table, we perform the OLS regressions of the acquirer’s CARs in the (-2, +1) days around the announcement date on the acquirer dividend change in the year preceding the merger. The dependent variable, CAR, is measured as the difference between the cumulative realized return and cumulative expected return over (-2, +1) days around the announcement date. The expected return is estimated from the Fama-French three-factor model, using daily stock returns of the firm and daily return on the CRSP equally-weighted index as the benchmark over the (-252, -30) days window. ACQ_INIT is the dummy variable for acquirers that initiated dividends in the year preceding the announcement. ACQ_INCR is the dummy variable for acquirers that increased dividends in the year preceding the announcement. ACQ_UNCHG is the dummy variable for acquirers that paid dividends and did not change their dividends in the year preceding the announcement. ACQ_INCR5 is the dummy variable for acquirers that increased dividends by at least 5% in the year preceding the announcement. ACQDIVYIELDCHG is the change in the acquirer dividend yield in the year preceding the announcement. ACQRUNUP is the acquirer stock price run-up in the (-30, -10) days prior to the deal announcement date. DEALVALUE is the natural logarithm of the deal value. CASH is the dummy variable for deals paid only with cash. HOSTILE is the dummy variable for hostile deals. COMPETING is the dummy variable for deals with more than one bidder. RELATED is the dummy variable for deals by targets and acquirers in the same four-digit SIC codes. TGTHITECH is the dummy variable for high-tech targets as per SDC classification. ACQROA, ACQCASH and ACQMKBK are the acquirer’s industry-adjusted return on asset, cash-to-asset ratio and market-to-book ratio in the year preceding the merger announcement. Heteroskedasticity-robust t-statistics are reported in parentheses. ***, **, and * indicate the significance levels at 1%, 5% and 10% levels, respectively

-0.002 (-0.620) 0.008 (1.718*)

0.148 (8.116***) -0.005 (-4.268***) 0.010 (2.974***) 0.007 (0.659) -0.009 (-1.313) -0.000 (-0.002) -0.007 (-2.032**) -0.030 (-1.725*) -0.022

0.010 (1.266) 0.148 (8.170***) -0.005 (-4.235***) 0.011 (3.100***) 0.007 (0.655) -0.010 (-1.425) 0.000 (0.102) -0.006 (-1.705*) -0.036 (-2.105**) -0.021

(-1.723*) 0.000 (0.074)

(-1.659*) 0.000 (0.174)

F-statistics Adj. R-squared Year fixed effects Observations

4.32*** 0.219 Yes 820

4.27*** 0.217 Yes 820

4.46*** 0.212 Yes 820

Jo

ur na

lP

re

-p

ro

of

ACQMKBK

(-1.819*) -0.000 (-0.084)

Table 7. Post-Merger ROA Change and Acquirer Dividend Policy Payout In this table, we examine the one-year and three-year industry-adjusted ROA changes of bidder firms from before to after the M&A announcements. In Panel A, we compare the acquirer ROA changes by whether the acquirer pays dividend in the year preceding the announcements. In Panel B, we perform the OLS regressions of the acquirer’s one-year and three-year industry-adjusted ROA changes. ACQDIV is the dummy variable for dividend-paying acquirers. ACQDIVYIELD is the acquirer dividend yield in the year preceding the M&A announcement. ACQRUNUP is the acquirer stock price run-up in the (-30, -10) days prior to the deal announcement date. DEALVALUE is the natural logarithm of the deal value. RELSIZE is the ratio of the deal value to the acquiring firm market capitalization at the end of the preceding year. RELATED is the dummy variable for deals by targets and acquirers in the same four-digit SIC codes. TGTHITECH is the dummy variable for high-tech targets as per SDC classification. ACQCAPX1, ACQDEBT1, ACQCAPX3 and ACQDEBT3 are the acquirer’s industry-adjusted capital expenditure-to-asset ratio in year 1, debt-to-asset ratio in year 1, capital expenditure-to-asset ratio in year 3, and debt-to-asset ratio in year 3, respectively. Heteroskedasticity-robust t-statistics are reported in parentheses. ***, **, and* indicate the significance levels at 1%, 5% and 10% levels, respectively.

ro of

ROA 3-Year Change Mean Median -1.760% -0.901% -2.741% -1.657% 0.981% 0.756% 2.46** 2.18**

ROA 3-Year Change Model 3 Model 4 -0.032 -0.023 (-1.657*) (-1.210) 0.014 (2.532**) -0.027 (-0.125) 0.054 (0.229) -0.002 -0.001 (-1.262) (-0.650) -0.000 -0.001 (-0.173) (-0.305) -0.004 -0.006 (-0.656) (-1.070) 0.005 0.003 (1.006) (0.475)

Jo

ur na

lP

re

-p

Panel A – Comparisons of ROA Changes ROA 1-Year Change Mean Median Dividend-paying acquirer -1.507% -0.818% Non-dividend-paying acquirers -2.099% -1.746% Difference 0.592% 0.928% t-stat / Wilcoxon-stat 2.51** 2.86*** Panel B - Regressions of Industry-Adjusted ROA change ROA 1-Year Change Independent variables Model 1 Model 2 Constant -0.038 -0.035 (-2.092**) (-1.811*) ACQDIV 0.026 (5.115***) ACQDIVYIELD 0.442 (2.194**) 2 ACQDIVYIELD -0.616 (-2.769***) DEALVALUE -0.004 -0.002 (-2.283**) (-1.408) RELSIZE -0.005 -0.005 (-1.436) (-1.454) RELATED 0.000 -0.003 (0.060) (-0.496) TGTHITECH -0.005 -0.008 (-1.117) (-1.681*) ACQCAPX1 -0.109 -0.113 (-2.444**) (-2.487**) ACQDEBT1 -0.069 -0.065 (-3.955***) (-3.749***) ACQCAPX3 ACQDEBT3

F-statistics Adj. R-squared Year fixed effects

3.42*** 0.0544 Yes

2.98*** 0.0447 Yes

0.085 (2.127**) -0.103 (-6.226***)

0.084 (2.075**) -0.102 (-6.259***)

5.07*** 0.104 Yes

5.13*** 0.0993 Yes

1,648

1,648

1,390

1,390

Jo

ur na

lP

re

-p

ro of

Observations

38

Table 8. Merger Payment Method and Acquirer Dividend Policy Payout In this table, we run a probit regression of the methods of payment using the dummy variable representing cash-only transactions as the dependent variable. We run a second model using ordinary least squares (OLS) regression and using the percentage of cash that makes up the payment method as the dependent variable. ACQDIV is the dummy variable for acquirers that paid dividends in the year preceding the M&A announcement. ACQDIVYIELD is the acquirer dividend yield in the preceding year. DEALVALUE is the natural logarithm of the deal value. HOSTILE is the dummy variable for hostile deals. COMPETING is the dummy variable for deals with more than one bidder. RELATED is the dummy variable for deals by targets and acquirers in the same four-digit SIC codes. ACQROA, ACQDEBT and ACQMKBK are the acquirer’s industry-adjusted return on asset, debt-to-asset ratio and market-to-book ratio in the year preceding the merger announcement, respectively. TGTROA and TGTMKBK are the target’s industry-adjusted return on asset and market-tobook ratio in the year preceding the merger announcement. Heteroskedasticity-robust t-statistics are reported in parentheses. ***, **, and* indicate the significance levels at 1%, 5% and 10%, respectively.

Jo

ur na

lP

re

-p

ro of

Panel A – Comparisons of per cent cash payment by subsamples Subsamples Mean Median Dividend-paying acquirer 59.023% 85.337% Non-dividend-paying acquirers 37.724% 0.000% Difference 21.299% 85.337% t-stat / Wilcoxon-stat 10.2156*** 9.651*** Panel B – Probit regression of cash as the only payment method and OLS regression of the percentage of cash payment Dependent variable = CashDependent variable = % cash Only payment Independent variables Model 1 Model 2 Model 3 Model 4 Constant 0.690 0.749 0.703 0.739 (2.720***) (2.907***) (8.362***) (8.646***) ACQDIV 0.616 0.193 (8.330***) (9.011***) ACQDIVYIELD 0.032 0.003 (5.584***) (2.474**) ACQDIVYIELD2 -0.377 -0.003 (-3.651***) (-2.727***) DEALVALUE -0.235 -0.224 -0.050 -0.040 (-10.092***) (-9.802***) (-7.660***) (-5.814***) HOSTILE 0.354 0.423 0.154 0.190 (1.392) (1.634) (1.882*) (2.320**) COMPETING 0.370 0.366 0.185 0.188 (2.598***) (2.555**) (4.522***) (4.644***) RELATED -0.230 -0.255 -0.049 -0.070 (-3.206***) (-3.568***) (-2.355**) (-3.382***) ACQCASH 0.030 0.112 -0.007 -0.074 (0.117) (0.479) (-0.101) (-0.994) ACQROA 1.141 1.427 0.301 0.298 (3.555***) (5.811***) (4.051***) (3.652***) ACQDEBT 0.275 0.340 0.067 0.063 (1.370) (1.747*) (1.147) (1.056) ACQMKBK -0.004 -0.003 -0.000 -0.000 (-0.839) (-0.556) (-1.084) (-1.405) TGTROA 0.535 0.558 0.136 0.140 (3.100***) (3.299***) (2.989***) (3.189***) TGTMKBK -0.001 -0.001 0.000 0.000 39

(-0.208)

(7.995***)

(9.777***)

326.80*** 0.177 Yes 1,680

325.80*** 0.174 Yes 1,680

24.19*** 0.223 Yes 1,680

18.87*** 0.191 Yes 1,680

Jo

ur na

lP

re

-p

ro of

Chi-squared / Fstatistics Pseudo R-squared Year fixed effects Observations

(-0.214)

40