The Benefits of Overvaluation: Evidence from Mergers and Acquisitions

• Date: 01 March 2020
• Author: Evangelos Vagenas‐Nanos
• Published date: 01 March 2020
• DOI: http://doi.org/10.1111/fima.12247

The Benefits of Overvaluation: Evidence

from Mergers and Acquisitions

Evangelos Vagenas-Nanos∗

Theoretical and empirical evidence debates whether acquirerscan exploit their overvalued equity

and create value by purchasing less overvalued or undervalued target firms. Shleifer and Vishny

(2003) and Savor and Lu (2009) argue in favor of this, while Fu, Lin, and Officer (2013) and

Akbulut (2013) provide evidence against. I revisit this issue and develop a quasi-experimental

design. The misvaluation effect for stock acquirers that are more overvalued than their targets

is isolated and measured. My findings offer direct evidence in favor of the Shleifer and Vishny

(2003) market-timing hypothesis.

I. Introduction

One of the theoretical predictions of Shleifer and Vishny’s (2003) model arguesthat overvalued

acquirers that purchase less overvalued or undervalued target firms by offering their overvalued

equity as a means of financing serve the interests of their long-term shareholders. However,

there is an ongoing debate in the empirical literature as to whether long-term shareholders of

overvalued stock acquirers benefit from takeover activity. Savor and Lu (2009) argue in favor

of Shleifer and Vishny’s (2003) prediction. They compare the long run performance of stock

acquisitions with that of withdrawn stock deals. They find that withdrawn deals earn lower long

run abnormal returns than completed deals indicating that shareholders of acquirers are better

off with, rather than without, the acquisition. In contrast to this evidence, Fu, Lin, and Officer

(2013) and Akbulut (2013) argue against the benefits of overvalued equity being exploited in a

mergers and acquisitions (M&A) framework. Fu et al. (2013) demonstrate that overvalued stock

acquirers underperform overvalued nonacquirers in the long run.

Given the debate and contradictory findings in the empirical finance literature, I revisit the

research question as to whether shareholders of acquirers benefit when employing their over-

valued equity in acquiring target firms. I develop a research framework that more accurately

and precisely isolates and measures the impact of overvaluation for stock acquirers. I further

explain and discuss why the research frameworks of Fu et al. (2013) and Akbulut (2013) are

underspecified and uninformative when drawing conclusions in favor or against the market-

timing hypothesis of Shleifer and Vishny (2003). In Fu et al.’s (2013) and Akbulut’s (2013)

work, it is unclear whether the underperformance of overvalued stock acquirers is driven by

the acquisitions effect, by the methods of payment, or by the overvaluation effect. Thus, their

approach is not suitable when offering evidence in favor or against Shleifer and Vishny’s (2003)

misvaluation hypothesis. Unlike Fu et al. (2013) and Akbulut (2013), the design of my approach

I thank the participants of the EFMA Conference 2012, Barcelona, the participants of the World Finance Conference,

2012, Rio de Janeiro, Monika Tarsalewska, and Milena Petrova for providing useful comments. I also thank Patrick

Verwijmeren, Dimitrios Petmezas, Leonidas Barbopoulos, Jo Danbolt, and Antonios Siganos for their constructive

comments. All remaining errors, omissions, and inaccuraciesare the responsibility of the author.

∗Evangelos Vagenas-Nanos is an Associate Professor in the Adam Smith Business School, at the University of Glasgow

in Glasgow,UK.

Financial Management •Spring 2020 •pages 91 – 133

92 Financial Management rSpring 2020

helps to account for all of the forces that the acquirers’ share price is subject to and success-

fully isolates the effect emanating from offering overvalued equity. The long run performance of

overvalued stock acquirers is subject to four different forces: the misevaluation effect and three

additional ones. The first is a natural long run stock price correction (De Bondt and Thaler, 1985;

Daniel, Hirshleifer, and Subrahmanyam, 1998). Overvalued firms have the greatest incentive to

proceed to acquisitions using stock as a means of financing the deal. A long-term downward

effect is expected due to overvaluation and a stock price reversal. The second effect is a negative

signaling effect. Myers and Majluf (1984) and Travlos (1987) argue that equity issuance signals

negative newsto the market as investors realize that such an announcement implies that the firm is

likely to be overvalued resulting in a negative market reaction.1The third force is the acquisition

effect itself. Empirical evidence argues that acquisitions destroyvalue (Loughran and Vijh, 1997;

Rau and Vermaelen, 1998) in the long run. The comparison of acquirers with nonacquirers (i.e.,

as in Fu et al., 2013 and Akbulut, 2013) does not consider this effect. Finally, the final force is the

effect of the exploitation of overvalued equity. If Shleifer and Vishny’s (2003) story holds, this is

expected to have a positive effect. Conclusively, overvalued stock acquirers are subject to three

potentially negative forces (i.e., signaling, long run price correction, and the acquisition effect)

and one possibly positive force (i.e., theexploitation of overvalued equity). The goal of this paper

is to eliminate the first three forces and isolate and measure the effect of exploiting overvalued

equity in an M&A framework.

I study US listed acquirers that announce completed acquisitions of listed target firms from

1985 to 2016. My sample includes deals that are financed with either 100% equity or 100%

cash. I adopt Rhodes-Kropf, Robinson, and Viswanathan’s (2005) decomposing methodology in

order to identify overvalued and undervalued acquiring and target firms. Takeovers for which the

acquirer is more overvalued than the target firm are classified as high relative misvaluation (high

RM) deals, and all other deals are classified as low RM deals. I develop a quasi-experimental

design and a difference-in-differences (dif-in-difs) approach in an attempt to isolate, capture,

and measure the effect emanating from the exploitation of overvaluation in stock acquisitions.

At the first stage, I compare the long run performance of high RM stock deals with that of

low RM stock deals. This first difference between the two groups captures misvaluation related

effects by employing stock as a method of payment and nonvaluation related effects. If Shleifer

and Vishny’s (2003) hypothesis holds, there should be a positive effect for high RM deals and

a neutral or negative effect for low RM deals. The nonvaluation related effects are primarily

associated with a natural long run price correction. High RM deals are subject to a downward

price correction, while low RM deals are subject to an upward price correction. The second

difference is estimated between the high RM and low RM cash deals. This difference captures

only nonvaluation related effects. Cash acquirers, irrespective of their RM, are not associated

with any exploitation of misvaluation benefits. Thus, the difference in the performance of cash

high RM versus low RM deals onlycaptures the long run price correction for the two groups. The

difference of the two differences [(Stock High RM - Stock Low RM) - (Cash High RM - Cash

Low RM)] cancels out the nonvaluation related effectsand only generates the valuation effect as

1One may argue that signaling is a short run effect. However, announcements of takeover deals convey important news

to the market and long run price reactions are observed. If there are no long run effects, there should be no difference

between cash and stock payment deals in the long run, which is not the case (Rau and Ver maelen, 1998). Markets are

not always efficient. Even in more obvious cases, such as earnings announcements, a well document longer run drift is

observed. If markets were strong form efficient, the signal and the news conveyed upon the announcement of earnings

would be incorporated in the share price immediately. Empirical evidence indicates that a long run drift appears after

the announcement day (Bernard and Thomas, 1990). Similarly,in the case of takeovers, the announcement of deals has

longer term signaling implications.

Vagenas-Nanos rThe Benefits of Overvaluation 93

exploited by stock acquirers. My dif-in-difs estimator indicates a positive effect of around 15%

to 28% for a period of two to five years postacquisition announcement for firms that exploited

overvaluation by undertaking stock acquisitions.

The signaling effect does not bias the dif-in-difs estimator. If stock deals convey a negative

signal (Travlos,1987; Loughran and Vijh, 1997), both high and low RM stock deals will be subject

to downward pressure. Thus, the first difference between high and low RM stock acquisitions

will remain unaffected. The same holds for the second difference between cash takeovers. If

cash acquisitions signal neutral or positive news, both cash subsamples will be affected, but the

difference between them will remain unaffected. I take the assumption that the signaling effect

will be equal for both high and low RM subsamples.2Overvalued cash acquirers may benefit

by acquiring undervalued target firms. However, that happens due to the exploitation of target

undervaluation, not the exploitation of their misevaluation, and that does not affect my research

framework.

On average, takeoverdeals have been shown to destroy market valuein the long r un. The acqui-

sition effect is also unlikely to bias the dif-in-difs estimator. I compare acquirers with acquirers.

I do not compare acquirers with nonacquirers (Akbulut, 2013; Fu et al., 2013) or acquirers with

failed acquirers (Savor and Lu, 2009). The long-term creation or destruction of synergies, on

average, should be captured in all four subgroups. Even if I assume that overvaluation could

incorporate elements of superior managerial skills (Tobin’s Q theory), the difference between

the high and low RM will be affected in favor of high RM deals. However, that would hold

for the differences in both stock and cash subsamples, and the dif-in-difs estimator will remain

unaffected. Malmendier and Tate (2008) argue that overconf ident managers who destroy value

for their shareholders are more likely to use cash as a method of payment. If we assume that

cash acquirers are more likely to choose acquisitions of lower quality, both high and low RM

cash acquirers will be subject to this effect. However, the second difference between high and

low RM cash deals will remain unaffected and, as such, the dif-in-difs estimator is unlikely to

be biased. Fishman (1989) argues that the medium of exchange may have valuation implications

for the target firm. To account for the above acquisition related effects, I run robustness test for

premiums and operating performance. My analysis indicates that synergy gains or overpayment

is unlikely to affect the dif-in-difs estimator.

A major question in the dif-in-difs research framework is whether the two groups, stock and

cash acquirers, are comparable. For instance, stock may be more overvalued than cash acquirers

(Rhodes-Kropf et al., 2005). The descriptive statistics on misvaluationin my sample are consistent

with this finding. The long run price correction effect (the nonvaluation related effect in the dif-

in-difs approach) will be more pronounced for stock acquirers. To alleviate this issue, I adopt

the minimum Mahalanobis distance matching technique (De Maesschalck, Jouan-Rimbaud, and

Massart, 2000). I match high RM stock acquirers with cash acquirers employing the RM measure,

as well as on the acquirer’s misvaluation measure. I repeat the same process for low RM stock

acquirers. In this way, high (low) RM cash acquirers are equally overvalued (undervalued) with

high RM stock acquirers. The reversal effectshould now be equal for both stock and cash acquirers

leading to an even more unbiased dif-in-difs estimator. I further extend the Mahalanobis distance

matching technique by taking into consideration a number of additional observable variables

that could affect the choice of the method of payment. My results remain robust in favor of the

hypothesis that overvaluation has positive effects for the shareholders of stock acquirers.

2Even if I relax this assumption and accept that the negative signaling effect will be stronger for overvalued stock

acquisitions, the first difference between high and low RM stock acquirers will be lower than estimated leading to an

underestimation of the dif-in-difs estimator. That wouldstill work in favorof my f indings and conclusions.