DIVIDEND POLICY AND STOCK ACQUISITION ANNOUNCEMENT RETURNS: A TEST OF ASYMMETRIC INFORMATION THEORY

• Date: 01 March 2019
• Author: Aymen Turki
• DOI: http://doi.org/10.1111/jfir.12164
• Published date: 01 March 2019

DIVIDEND POLICY AND STOCK ACQUISITION ANNOUNCEMENT

RETURNS: A TEST OF ASYMMETRIC INFORMATION THEORY

Aymen Turki

European Business School of Paris, INSEEC U.

Abstract

This study examines 711 U.S. stock-based acquisitions announced between 1985 and

2013 to analyze the relation between the acquirer’s dividend policy and its stock returns

when the acquisition is announced. Asymmetric information theory suggests that the

lower the level of uncertainty about the acquirer’s value, the smaller the acquirer’s price

drop when a stock-based acquisition is announced. In support of this theoretical

prediction, the current study identifies less negative acquirer stock returns around the

announcement of stock acquisitions initiated by dividend-paying firms compared with

those initiated by non-dividend-paying firms.

JEL Classification: G34, G35

I. Introduction

Even as studies of mergers and acquisitions (M&As) have proliferated, following the

development of corporate control activity, their returns remain poorly understood,

including the potential for negative market reactions to stock-based deals. In Myers and

Majluf’s (1984) early models of asymmetric information created to explain negative

returns to stock-based M&As, they assume information asymmetry on both sides, such

that each merging firm retains private information about its own value. This result, as

further developed by Hansen (1987) and Fishman (1989), is common when predicting

the same negative stock price impact of stock payments regardless of the characteristics

of the stock issued. Thus, testing hypotheses pertaining to the specific features of the

stock that firms issue to pay for the transaction might be additionally insightful.

Dividend-paying firms create less information asymmetry (Lang and Litzenberger 1989;

Howe and Lin 1992; Khang and King 2006; Li and Zhao 2008), such that the acquirer’s

dividend policy might help reduce negative market reactions to stock-based M&As.

No extant research addresses this issue directly, though recent studies consider it

empirically in seasoned equity offering (SEO) and initial public offering (IPO) settings.

Booth and Chang (2011) analyze the relation between dividend payment status and SEO

announcement returns and find that since the mid-1980s, the market has reacted less

The author acknowledges the support of the European Centre for Corporate Control Studies, SKEMA

(ECCCS) Research Center and the Institute of Research, European Business School (IREBS) Research Center. He

thanks Sebastien Dereeper, Khamis Al-Yahyai (Australasian Finance and Banking Conference, 2012), Uri Benzion

and Viktoria Dalko (Eastern Economic Association Conference, 2013), and Hubert De-La-Bruslerie (French

Finance Association Conference, 2013) for their helpful comments and suggestions.

The Journal of Financial Research Vol. XLII, No. 1 Pages 115–145 Spring 2019

DOI: 10.1111/jfir.12164

115

© 2019 The Southern Finance Association and the Southwestern Finance Association

negatively to dividend-payer SEO announcements than to non-dividend-payer SEO

announcements. By examining whether the dividend status of newly listed firms can

explain IPO performance, How, Kian, and Peter (2011) offer support for asymmetric

information theory, indicating that dividend payers are more profitable after an IPO. This

effect should also be apparent in stock-based acquisitions: dividend-paying acquirers

should have less negative announcement-induced returns than non-dividend-paying

acquirers, because dividend payers create less information asymmetry about the real

value of their traded shares. The argument underlying this central hypothesis is that

dividends resolve information asymmetry about stock valuation, such that investors who

are uncertain about the true value of the stock used to pay for the transaction react less

negatively to the announcement of a stock offer.

To test this hypothesis, the curren t study relies on a sample of 711 U.S. M&As

announced between 1985 and 2013 tha t features only listed acquirer s, to ensure the

availability of dividend data, and it applies standard data screens to identify significant

deals (Masulis, Wang, and Xie 20 07). Dividend status (i.e. , whether the acquirer pays

dividends) before an acqu isition is the focal variab le. The number of analys ts

following the acquirer’s stock in the month before the deal announcement, forecast

error, and forecast dispersion rep resent acquirer-side infor mation asymmetry. The

computation of the acquirer’s cum ulative abnormal returns (CA Rs) around the

announcement date relies on Brown and Warner’s (1985) standard event-study

methodology, with several em pirical tests. First, univariate tests determine whet her the

acquirer announcement retur n is, on average, less negative fo r stock-based deals

initiated by dividend-payi ng acquirers than for thos e initiated by non-dividen d-paying

acquirers. Second, a cross-sec tional analysis of acquirer ab normal return reveals

whether the dividend status of the acqu irer reduces negative market rea ctions to the

deal announcement. These tes ts account for target status (i.e ., whether the target is

privately held or publicly t raded), given the effect of t his target-specific inform ation

on the level of abnormal returns arou nd the announcement. The data ind icate that

dividend-paying stock ac quirers exhibit less negat ive stock returns around t he time of

the announcement than non-di vidend-paying stock acqu irers, consistent with the idea

that the former establish less unc ertainty about the value of the stock t hey have used to

finance the deal.

This result holds after robustness checks. First, the check for the effect of

dividend payment activity over the sample period involves a cross-sectional analysis of

announcement returns for two subperiods: 1985–2002 and 2003–2013. Second,

alternative measures of dividends include whether investors consider changes in and

the amount of the dividend before reacting on the stock acquisition announcement day.

Third, Heckman’s sample selection model addresses concerns about endogeneity, such

that certain types of firms might choose to pay dividends, and the sample of dividend-

paying acquirers might not be random. Fourth, an assessment of whether dividends solve

information asymmetry for cash acquisitions, by comparing the effect of dividends on

announcement returns for stock versus cash acquisitions, shows that dividends do not

solve stock valuation uncertainty for cash payments. This finding is unsurprising;

dividends solve the information asymmetry that emerges from the inaccurate valuation

of the acquirer’s stock used to pay for the acquisition (Shleifer and Vishny 2003), which

116 The Journal of Financial Research