Buy Smart, Time Smart: Are Takeovers Driven by Growth

Buy Smart, Time Smart: Are TakeoversDriven by Growth Opportunities or

Mispricing?

Sjoerd Van Bekkum, Han Smit∗, and Enrico Pennings

Excessively high pricing by bidders and targets can be explained by new growth opportunitiescreated by the merger or by irrational overpricing in financial markets. We integrate both explana-tions through a new decomposition of firm value and investigate whether it is “true” growth valueor mispricing that drives takeover waves. We find that “bidders buy smart.” Bidders primarilyhave high market values because of growth opportunities and overpricing, and select targets thatare less overpriced with similar fundamental growth value. Bidders also seem to “time smart.”Takeover activity increases when bidders are more overpriced, in order to cushion against pricecorrections.

To justify their intended business strategies in takeover announcements to shareholders, man-agement teams frequently argue that new synergies and growth opportunities arise from thecombined entity. Yet, previous research has found conflicting evidence as to what drives mergeractivity. In view of the vast literature on mergers and acquisitions (M&A), explanations for M&Acan be approximately classified into two classes that have both found empirical support: 1) arational explanation and 2) a behavioral explanation.

Consistent with neoclassical rational theory, prior studies indicate that high market valuesof bidders and a takeover premium for targets reflect rational expectations of cost efficiencies,synergies, market power, and growth options accordingly appreciated by financial analysts andequity investors. Mergers may reallocate assets more efficiently in response to industry shocks(Mitchell and Mulherrin, 1996) and deregulation, removing “long standing barriers to mergingand consolidation, which might have kept the industry artificially dispersed” (Andrade, Mitchell,and Stafford, 2001). Consolidation is found to have been the driving force behind acquisitionsin the 1990s (Andrade et al., 2001; Holmstrom and Kaplan 2001). In line with the rational view,real options theory implies that acquisitions confer valuable new growth opportunities for thecombined entity (Klasa and Stegemoller, 2007).

In contrast to rational theories, two distinct behavioral views offer alternative explanations forwhat drives merger waves (Baker, Ruback, and Wurgler, 2007). Apart from the biases of decisionmakers, market inefficiencies may arise when financial markets fail to price firms correctly

This paper corresponds to Chapter 4 of Van Bekkum’s Ph.D. dissertation from Tinbergen Institute, Netherlands. Theauthors would like to thank the members of his dissertation committee for their comments. The authors are especiallygrateful to Anjolein Schmeits, an anonymous referee for valuable comments and suggestions, and to Bill Christie (Editor).

∗Sjoerd Van Bekkum is an Assistant Professor at the Erasmus School of Economics at Erasmus University in Rotterdam,Netherlands. Han Smit is a Professor at the Erasmus School of Economics at ErasmusUniversity in Rotter-dam, Netherlands. Enrico Pennings is a Professor at the Erasmus School of Economics at Erasmus University inRotterdam, Netherlands.

Financial Management • Winter 2011 • pages 911 - 940

912 Financial Management � Winter 2011

(Shiller, 2000).1 High market values are strongly related to merger activity (Rhodes-Kropf andViswanathan, 2004; Jensen, 2005) and managers of overvalued firms buy targets that are lessovervalued (Rhodes-Kropf, Robinson, and Visnawanathan, 2005).

While the different explanations for M&A may be, at least in part, due to a wide rangeof methodologies applied (Baker et al., 2007), they may also indicate common ground. Thismakes intuitive sense. The scope for valuation errors or mispricing is greater when “elusive”growth options value is a major source of expected value than when value gains arise fromexploiting established products and established markets. Hence, if takeovers are jointly explainedby both theories, combining theories for mispricing and investment opportunities offers a strongertheoretical base for high market values in takeovers. To facilitate an integrated analysis, we devisea new measure for fundamental growth options value that adjusts market-based growth optionsvalue for mispricing. Fundamental value is based on company fundamentals such as book value,net income, and leverage, while mispricing is unrelated to fundamental value creation.

We find that takeovers do confer valuable growth opportunities, but these are strongly contingentupon market sentiment. The conventional (market-based) growth options measures should be usedwith care, as they can be driven by overvaluation and need to be corrected for mispricing. Byanalyzing firm properties (e.g., involvement in a merger and the role of the target or bidder)and deal properties (e.g., form of payment, acquisition financing, and deal attitude), we find thatdeal properties that favor overvalued bidders are significantly related to bidder overpricing (andto their growth options value), as well as to the growth options value of targets (but not theiroverpricing). These results lead to two key implications.

First, “bidders buy smart.” Although bidders have high fundamental growth options valuerelative to nonbidders and target firms, they are overvalued by the market. This implies that theiroverpriced equity allows them to finance their acquisitions inexpensively. Their targets are pricedlower when compared to firms not involved in a merger but have a similar growth options valuethat persists after controlling for mispricing. This suggests undervaluation, at least with respect tothe bidder’s equity, while fundamental growth option value is retained. Second, bidders time theirtakeovers accordingly. Specifically, when excess pricing for bidders is high, takeover activity ishigh as well. Thus, acquisitions are made by firms that are overvalued relative to their peers. Addi-tionally, growth options value for targets is negatively related to takeover activity. Therefore, whenbidders are overvalued, they acquire targets that are undervalued relative to the bidder. In effect,this will cushion against an eventual correction in share price while retaining growth potential.

These results are validated using a large international data set. By using market-implied growthoptions value and controlling for pricing errors, we avoid using sophisticated option modelsthat introduce substantial model and parameter uncertainty. In a 12-year panel of over 30,000observations, we use t-tests and multiple equation models in a multivariate setting to demonstratethat bidders buy smart. In order to confirm that bidders also time their acquisitions smart, weestimate the relation between growth opportunities, mispricing, and takeover activity in a similarempirical setting.

Our paper proceeds as follows. In Section I, we position our contribution into the literature. InSection II, we derive the hypotheses. Section III describes the data and decomposition employed,followed by a presentation of the results in Section IV. Section V contains our conclusion andpresents implications for further research.

1 Decision makers, the bidder’s chief executive officers (CEOs), and management teams may act irrationally. CEOs maysuffer from overconfidence and overoptimism (Larwood and Whittaker, 1977; Roll, 1986; March and Shapira, 1987) andmanagers may use their discretion to overinvest (Heaton, 2002; Malmendier and Tate, 2005). Taking a perspective similarto ours, Hietala, Kaplan, and Robinson (2003) disentangle synergies and overpayment for overoptimistic bidders.

Van Bekkum, Smit, and Pennings � Are Takeovers Driven by Growth Opportunities or Mispricing? 913

I. Literature Review

A. Real Options Theory in M&A

Recent real options models explain the dynamics of M&A decisions under uncertainty, suchas the timing of takeovers under competition (Morellec and Zhdanov, 2005), the associateddivision of the surplus between acquirer and target, the procyclical nature of waves when theyare motivated by economies of scale (Lambrecht, 2004), and the timing of mergers when they aredriven by diversification (Thijssen, 2008).2 Smith and Triantis (1995) suggested that aside fromflexibility and divesture options, acquisitions contain significant growth options. In addition tothe value of future cash flows and synergy values, combined forces may enhance the value ofthe option to invest in new products and the option to acquire other competitors. Takeovers alsolimit competition by eliminating the target’s option to grow, and the combined entity may moreeffectively allocate resources between the markets of the bidder and the target. Consistent withthis idea, Klasa and Stegemoller (2007) find that serial takeovers are the result of time-varyingchanges in the growth opportunity set of the acquirer.

For the bidder, executing an acquisition is effectively exercising an option to exchange the ac-quisition price (either in cash or the acquirer’s shares) for the value of the target (Margrabe, 1978).Thus, the opportunity to acquire a target’s equity can be seen as an option while, additionally, thetarget itself is a portfolio of real growth options. The target’s strategic growth options generatefollow-on opportunities that are part of the underlying value of the option to exchange. Relativeto the earnings capacity of assets in place, the value of growth options is elusive and difficultto estimate, which may influence the timing and value of the option for the bidder. Therefore,acquisitions become more attractive when the bidder’s equity is overvalued and the acquisitioncan be paid for or financed with equity, or when the fundamental growth option value of the targetis high and the target is less overvalued or even undervalued. In terms of fundamental value, thereal exchange option is more in the money when the bidder is overvalued relative to its target.This, however, is not necessarily observed in financial markets.

B. Empirical Real Options Analysis

Miller and Modigliani (1961) determined that a firm’s equity value (P) can be partitioned intothe value of a firm’s assets in place (VAIP) and the present value of future investments in growthopportunities (VGO):

P = VGO + VAIP, (1)

where VAIP is an appropriately discounted, perpetual stream of earnings from the assets in place.Under the assumption of no further growth, all earnings are equal to those in the current period.

This assumption is nonrestrictive, since all future growth (either positive or negative) should endup in the VGO term. Data on market value and the value of assets in place are readily available, sothe value of growth options VGO can be “backed out” from Equation (1). Real options theory hasunderpinned growth options value with strong theoretical foundations (Myers, 1977; Dixit andPindyck, 1994; Trigeorgis, 1996). In this study, we will extend Equation (1) so that it incorporatesirrational behavior in financial markets.

2 An exchange option can be viewed as a call option on the benefits of the merger. The costs of a merger are representedas the exercise price. While these studies interpret takeovers as a call option, we refer to these options as exchange optionsto stress the difference from strategic options.


A company’s portfolio of real options can be valued “bottom-up” through the estimation ofindividual option parameters. This entails the use of binomial lattice (Trigeorgis, 1996) or sim-ulation models (Longstaff and Schwarz, 2001) making it possible to adjust the option valuationto case-specific characteristics (Copeland and Antikarov, 2001). In reality, a firm’s growth op-portunities consist of a complex combination of growth options resulting from intrafirm optioninteractions. The combined value of an options portfolio is typically not the sum of the options(Trigeorgis, 1996), and real options should occasionally be combined sequentially, sometimes inparallel and sometimes not at all (Childs, Ott, and Triantis, 1998). Two merging firms complicatethe options portfolio even further. To avoid estimating option and interaction parameters, wefollow a less involved “top-down” approach using the financial market information of publiclylisted firms and calculating the combined growth options value embedded in the stock price.

Differences in growth options value were first empirically studied by Kester (1984). Morerecently, growth options value has been related to systematic risk (Chung and Charoenwong,1991); to research and development (R&D), uncertainty, and the skewness of returns (Smit andVan Bekkum, 2011); to the pricing of initial public offerings (Chung, Minsheng, and Yu, 2005);to idiosyncratic risk (Cao, Simin, and Zhao, 2006); to firm, industry, and country effects (Tong,Alessandri, Reuer, and Chintakanda, 2008); to downside risk of multinational corporations (Tongand Reuer, 2008); and to international joint ventures (Tong, Reuer, and Peng, 2008).

This market-based measure for growth options value, however, hinges on the assumption thatmarket prices adapt fully and immediately to price relevant information. This assumption is quitestrong since the elusive value of growth options can easily be over- or underestimated in financialmarkets. Any erroneous market pricing will introduce noise in the growth options value. Below,we will adjust the growth component in Equation (1) to alleviate this issue.

C. Irrational Capital Markets

Due to market inefficiencies (Shleifer, 2000; Barberis and Thaler, 2003), the market priceP in Equation (1) may not reflect fundamental firm value. Corporate managers have superiorinformation about their own firm (Muelbroek, 1992; Seyhun, 1992), and rational managers cangain by timing anomalies that result from irrationality in the markets (Baker and Wurgler, 2002;Jenter, 2004). Managers may also cater to short-term investors by maximizing appeal (Baker et al.,2007). In these situations, managers influence the market value of a firm without creatingfundamental value. In a takeover context, when potential market value deviates from fundamentalvalues, takeover activity may be strongly related to stock market valuation (Rhodes-Kropf andViswanathan, 2004; Dong et al., 2006). In particular, Gao (2010) finds that long-horizon managerstend to exploit overvalued stock prices by making equity transactions. This is in line with previousevidence regarding information asymmetries. Firms issue equity when they are overvalued orexpect poor future performance (Fu, 2010) as acquisitions are, in effect, a form of equity issuance(Ritter, 1991; Loughran and Ritter, 1995; Jenter, 2004).

Empirical studies regarding overvaluation typically separate market value into an intrinsicportion and a mispricing portion. A popular proxy for mispricing is the market-to-book ratio(M /B) (Barberis and Huang, 2001; Daniel, Hirshleifer, and Subrahmanyam, 2001; Bloomfieldand Michaely, 2004; Rhodes-Kropf et al., 2005; Dong et al., 2006). For instance, Rhodes-Kropfet al. (2005) find that bidders have a larger M /B than targets, but both have higher M /Bs thannonmergers. Alternatively, Dong et al. (2006), who report more studies that use M /B in thiscontext, compare a firm’s M /B ratio with its market-to-value ratio, where value is measured as aperpetual stream of (forecasted) earnings. They too find that both bidders and targets have highM /B ratios due to mispricing. Our findings are consistent with both studies.


In addition to mispricing, however, the M /B ratio is also a well-known proxy for growthopportunities (Berk, Green, and Naik, 1999; Anderson and Garcia-Feijoo, 2006). Therefore, interms of M /B, overpriced bidders and targets, as in Rhodes-Kropf et al. (2005), may also bebidders and targets with a large potential for economic growth. Dong et al. (2006) address thisissue by measuring intrinsic value, and find that targets are less overvalued and have a lower M /Bratio. Failure to compare M /B with nonmerging firms, however, does not explain the selectionof targets from a growth options perspective. Therefore, we will resort to a different, cleanermeasure than M /B and compare it among bidders, targets, and nonmerging firms. Our approachindicates that M /B consists of both growth opportunities and mispricing, and that this distinctionis of great importance when analyzing takeovers.

II. Hypotheses

A. Bidders Are Overvalued

Several takeover motives exist that are related to share prices, but many of them are not aimedat creating fundamental value. For instance, Myers (1976) confirms that takeovers may boostshort-term earnings per share but depress long-term growth. More recently, Rhodes-Kropf andVishnawanathan (2004) propose a theory that explains why merger waves occur during valuationwaves. In their model, the overall valuation error of share prices leads to the overestimationof synergies after a takeover. In addition, Shleifer and Vishny (2003) argue that bidders areovervalued relative to the target and buy smart using their overvalued stock. As such, firmsmay inexpensively acquire targets by taking advantage of financing with their overvalued equity.3

What these models have in common is that managers maximize firm value when financial marketsmay not represent company value.4 We make the following empirical observation that serves asa building block for further theoretical reasoning.

Empirical Observation: Bidders are overpriced relative to nonmerging firms

This empirical observation has implications for a company’s growth options value. Whenmarkets fail to price growth options correctly, the growth options value tends to be overpricedand conventional growth options measures may at times be inappropriate. This bias is likely todiffer over time and across industries. The surveyed literature on mispricing indicates that it isalso associated with takeover activity. Mispricing arguments state that market value in excess offundamental value is a pricing error. The implication is that while the growth options value maybe large, it would be substantially smaller if it could be backed out from the fundamental value,instead of the overvalued market price in Equation (1).

B. Bidders Buy Smart

As noted, a company can be viewed as a basket of assets and growth options. Valuable growthoptions from the merging bidder and target can be an important rational reason to bid. One such

3 This model is in line with evidence regarding information asymmetries, where firms issue equity when they areover-valued since acquisitions are, in effect, a form of equity issuance (Jenter, 2004).4 An acquisition may also result from managers who overestimate their ability to create synergies after the merger (Roll,1986), and managers may have objectives that reduce firm value (Morck, Shleifer, and Vishny, 1990). However, thisinterpretation also leads to our first proposition.


example would be when the acquirer can create synergies and provide instant access to manynew resources and capabilities. Organic growth options may be present in either party as theyresult from novel technologies or reshaped market conditions (Bettis and Hitt, 1995). When thevalue of the company’s growth options develops favorably (e.g., positive changes in productdemand and further development of technological trajectories), a firm appropriates future profitsby exercising its options through investment. In addition to these embedded growth optionsthat are complementary to the bidder, a target may also provide an initial foothold in a marketwith high entry barriers as in Miller and Folta (2002), or function as a platform for follow-onacquisitions (Smit, 2001; Smit and Moraitis, 2010a, 2010b). The target then generates additionalgrowth option value to be obtained through follow-on acquisitions.

Real options are related to mispricing through two untested, but relevant implications fromShleifer and Vishny (2003). We will adopt their terminology. First, targets in acquisitions “areundervalued relative to the bidders” (p. 308) or less overvalued; if not, bidders may well select atarget that offers better market value for their overvalued equity. Therefore, targets should be lessoverpriced. Second, targets may also be “underpriced in terms of fundamentals” (p. 308). Clearly,bidders may also aim to buy targets whose growth options can be acquired at a price below theirown equity. As such, cheaper targets should possess similar or higher growth option value.

When related to the work of Shleifer and Vishny (2003), real options theory offers a novelexplanation as to who is likely to acquire, or who becomes a target. Overpriced bidders withexpensive equity prefer to buy targets that provide fundamental growth options value at a lowerprice. These bidders may still possess genuine growth opportunities, and these targets may alsobe overvalued at times. Yet in contrast to bidders, the difference between the fundamental andmarket-based growth options value should be smaller. Such a real options view of the Shleiferand Vishny (2003) model implies that overvalued bidders, potentially having substantial growthoptions value, exercise an option to exchange by cheaply buying targets with similar or bettergrowth options value. Therefore, we derive the following hypotheses:

Hypothesis 1a: Bidders have high market values that are largely due to mispricing andfundamental growth opportunities.

Hypothesis 1b: Targets have high market values that are largely due to fundamental growthpotential, and less to mispricing.

C. Bidders Time Their Bids Smart

Cited as another implication by Shleifer and Vishny (2003), bidders prefer targets that canlimit losses when prices eventually correct. When expecting a downward correction, overvaluedfirms may use their overvalued equity to acquire strong targets and try to cushion against afuture downfall. We expect that takeover activity increases with overvaluation as inappropriatelyhigh market values will provide an incentive to acquire. In effect, mispricing provides a windowof opportunity to inexpensively purchase targets that give more “value for money” in terms ofgrowth opportunities. If these considerations explain the bidding of firms, we should observethat, over time, takeover activity is positively related to excess pricing. Therefore, we derive thefollowing hypothesis:

Hypothesis 2a: For bidders, mispricing is positively related to takeover activity.

D. Bidders Buy Their Targets Smart

When bidders are overvalued, they derive value by acquiring targets whose price is less subjectto overpricing, relative to their own. While fundamental growth options value makes a firm


more attractive to buy, additional value is created if the equity of the target has not been subjectto overpricing (so that the target is inexpensive). Therefore, while it induces firms to becomebidders, overpricing makes a firm less attractive as a target. It is expected that a target’s mispricingis not positively (and potentially negatively) related to takeover activity. Consequently, we derivethe following hypothesis:

Hypothesis 2b: For targets, mispricing is not positively related to takeover activity.

Mispricing can also be related to several characteristics of M&A transactions. If deal character-istics differentially affect bidders’ and targets’ fundamental growth option value and mispricing,this provides further support for Hypotheses 1a and 1b. Furthermore, by estimating the effect oftakeover activity after controlling for deal characteristics, we also perform a robustness checkfor Hypotheses 2a and 2b. To this end, we will examine how deal payment (through stock), dealfinancing (by means of a stock offering), and deal attitude (i.e., hostile or friendly) are relatedto the growth and mispricing components of share price. These characteristics are related tooverpriced bidders as follows.

First, if a bidder is overvalued it can be expected that this bidder will prefer to pay for a target inequity. In terms of true value, this makes the deal cheaper and protects bidding firms from futuredrops in share price. The target’s board of directors may require a cash deal, however, so that allthe risk of adverse price movements of the merger is borne by the bidder. Bidding firms maystill put their overvalued equity to work, then, by financing the deal through a seasoned equityoffering (Burch, Christie, and Nanda, 2004). Since the offering raises cash that can be used tobuy the target, this too protects bidding firms from a future correction in share price (Jegadeesh,2000). If management still opposes the deal, the bidding firm may resort to a hostile takeoverthat is not approved by the target board.

III. Data and Sample Characteristics

To study the hypotheses, we examine how year-by-year changes in growth opportunities andmispricing are related to year-by-year changes in takeover activity and whether overpriced biddersrelate to targets with fundamental growth options. We then examine growth options value andmispricing as time evolves. Finally, we extend the analysis to a multivariate approach.

From the Security Data Company’s (SDC) Platinum M&A database, we selected announcementdates and stock exchange daily official list (SEDOL) codes in the period from 1995 to 2006 of allcompleted and unconditional public takeovers for a stake of at least 95%. For each deal, we collectaccounting data using Worldscope and use Datastream to obtain yearly market equity values offirms involved in takeovers. To these data sets, we add auxiliary data regarding the cost of equity(using weekly pricing data from Datastream) and the synthetic cost of debt (see Damodaran,2002, using yearly data from Thomson One Banker [TOB]) to construct the weighted averagecost of capital required for estimating the assets in place. We also classify a firm’s primary standardindustrial classification (SIC) code into the Fama and French 12-sector definitions. As the miningsector is characterized by many takeovers, we add an extra sector using the Fama–French 30-sectordefinition for mining.

We merge these sources into a sample at firm level that is used to calculate growth options valueand mispricing. This procedure is described in detail in the Appendix. We follow the procedureas provided in Rhodes-Kropf et al. (2005) by matching “fiscal year end data . . . with . . . marketvalues occurring three months afterward. Because firms have different fiscal year end dates,this involves compensating for, so that the year of the data corresponds to the year in which


the accounting information was filed. Then, we associate [TOB and Datastream] observationswith an SDC merger announcement. If the announcement occurs between the fiscal year endand one month after the [Datastream] market value, we associate the merger announcement withthe previous year’s accounting information.” When a firm’s fiscal year end is unreported, weassume the fiscal year to coincide with the calendar year. To increase sample size and facilitate acomparison between takeovers and nontakeovers, “firms that are ultimately involved in mergers[are grouped] in the nonmerger category in the years in which they have no merger activity.”5

We then count the number of announced merger events for every year. Some targets are privatefirms for which we have no financial data. Other targets are located in other countries thanthose covered by Worldscope and lack (part of) the information needed to calculate the value ofassets in place. It also happens that several firms bid for a single target or that bids are rejected.Finally, targets disappear from the sample after a completed acquisition. For these reasons, thefirm-level sample contains roughly three times more bidders than targets. Due to spurious andinvalid SEDOL codes in the SDC database, we are forced to drop more targets than acquirers thatlack data from Worldscope. We keep the deal’s counterpart if that observation can be matched.This results in a sample that consists of 30,000 firm-year observations over a 12-year period. Werefer to Table I for a comparison with two closely related studies.

Finally, we merge the measures for growth options, fundamental value, and market value backto the SDC data. This sample focuses on the M&A transaction level to further test whetherour mispricing measure is valid. It also allows us to analyze which M&A attributes and firmcharacteristics explain the value components of bidders and targets. We match firm-level data(recorded at fiscal year-end) to the last reported fiscal year on deal level. If this item is notavailable, we match firm-level data to the year prior to the date that a merger was officiallyannounced. Because a firm’s balance sheet is no longer drawn up after a takeover, bidders alsocomparably outnumber targets in the deal-level sample.

A. Variable Definitions

1. Growth Options Value

To back out the value of growth options from Equation (1), we follow the approach suggestedby Tong and Reuer (2008), Tong, Alessandri, et al. (2008), and Tong, Reuer, and Peng (2008)who use an economic profit model for valuing the firm. They define the value of assets in placeon a company level, so that the assets in place in equity (VAIP) equal

VAIP = CI + PV (EP) − D, (2)

where CI is capital invested, PV (EP) is the present value of economic profit, and D equals totaldebt. Equation (2) demonstrates that the value of assets in place represents the replacement valueof capital invested and the present value of perpetual earnings, again assuming no growth anda constant discount rate for each year’s estimate of VAIP. This arrangement has the advantagethat invested capital can be directly calculated from the financial reports. As a result, the annuityPV (EP) becomes smaller and VAIP as a whole is less sensitive to changes in the discount rate.Furthermore, economic profit is arguably a more accurate proxy for profitability than accountingearnings, as it adjusts these earnings to cash measures.

5 To test whether the Rhodes-Kropf et al. (2005) sampling approach may introduce bias into the results, we have repeatedtheir estimation procedure below on the full cross-section of firms (without regard to M&A transactions). The regressionresults turned out to be fairly similar to their 2005 article, with comparable R2s and coefficients.


Tab

leI.

Co

mp

aris

on

of

Stu

die

s

Rh

od

es-K

rop

fet

al.(

2005

)T

on

g,A

less

and

ri,e

tal

.(2

008)

Th

isS

tud

y

Pri

mar

yfo

cus

Ove

rval

uati

onof

bidd

ers

and

targ

ets

Gro

wth

valu

eof

mul

tina

tion

alco

mpa

nies

Gro

wth

valu

ean

dov

erva

luat

ion

ofbi

dder

san

dta

rget

sK

eyva

riab

les

Mar

ket-

to-b

ook

valu

eM

arke

tval

ueof

grow

thop

tion

sM

arke

tval

ueof

grow

thop

tion

sfu

ndam

enta

lval

ueof

grow

thop

tion

sIm

plic

itva

lue

(P̂)

Cou

ntry

,ind

ustr

y,ti

me

effe

cts

Indu

stry

,tim

eef

fect

sm

ispr

icin

g(P

−P̂

)M

etho

dof

anal

ysis

OL

SV

aria

nce

com

pone

nts

SU

RD

efin

itio

nof

indu

stry

4-di

gitS

IC4-

digi

tSIC

4-di

gitS

ICD

efin

itio

nof

sect

orFa

ma

and

Fren

chFi

rstd

igit

ofS

ICFa

ma

and

Fren

chN

umbe

rof

sect

ors

127

13O

bser

vati

ons

Yea

rob

serv

atio

nsfo

rU

Sfi

rms

men

tion

edin

mer

ger

anno

unce

men

ts

Yea

rob

serv

atio

nsfo

rth

ela

rges

tpu

blic

lytr

aded

com

pani

esW

orld

wid

eye

arob

serv

atio

nsm

enti

oned

inm

erge

ran

noun

cem

ents

,M&

Atr

ansa

ctio

nsN

umbe

rof

obse

rvat

ions

110,

877

6,65

523

,419

Num

ber

offi

rms

8,35

02,

352

3,97

6N

umbe

rof

year

s19

77-2

000

1997

-199

919

95-2

006

Dat

aso

urce

sC

RS

P,C

ompu

stat

,SD

CM

&A

Dat

abas

eW

orld

scop

e,D

atas

trea

m,

Com

pust

at,S

tern

Ste

war

tW

orld

scop

e,D

atas

trea

m,

Tho

mso

nO

neB

anke

r,S

DC

M&

AD

atab

ase


Defining market value Pit as the fiscal year-end closing price multiplied by common sharesoutstanding, we calculate VGO from Equation (1). Since equity mispricing is an important elementof this paper, we scale to equity value. We delete the highest and lowest 5% of VGO observationsfrom the sample as the normalized growth option in Equation (1) does not always behave well. Thevalue can be larger than one if earnings are negative or smaller than zero when the market valuefalls below a firm’s perpetual stream of earnings. Also, due to the normalizing of Equation (1),when the market value is negligible relative to its earnings (in absolute terms), the likelihoodof (highly influential) extreme values increases. This becomes more prevalent in highly volatileindustries and for firms in some form of distress.

2. Fundamental Value

Particularly for takeovers, assuming Equation (1) to hold would ignore the vast existing mispric-ing literature. Therefore, we will refer to the previous definition of the VGO residual in Equation(1) as the market growth value, or V M

G O , and rewrite Equation (1) by adding (P̂ − P̂):

P = VAIP + (P̂ − VAIP) + (P − P̂) = VAIP + V FGO + XSP, (3)

where V FG O measures fundamental growth value, VAIP can be estimated using Equation (2), XSP

is the excess price and P̂ is a measure for fundamental value. To estimate P̂ , we adopt themethodology recently applied in a takeover context by Rhodes-Kropf et al. (2005), who estimatefundamental value (as opposed to observed market value) from an asset pricing model by Famaand MacBeth (1973). In this model, fundamental value is estimated as the predicted value froma series of simple ordinary least squares (OLS) regressions, estimated by year and industry. Thisprocedure is also employed in the accounting literature on value relevance (Penman, 1998; Francisand Schipper, 1999; Barth, Beaver, and Landsman, 2001). One advantage of this computation isthat it measures mispricing at the firm level, which serves our purpose in estimating Equation (3).The fundamental value of firm i in an industry j at time t is derived from the regression equation

ln(P)ijt = α0jt + α1jt ln (B)ijt + α2jt ln(NI)+ijt + α3jt IN I<0 ln (NI)+ijt + α4jtLEV ijt, (4)

where P is market capitalization, B is the book value, NI+ is the absolute value of net income,LEV is the leverage of a company defined as total debt/(market equity value + total debt), andI (NI<0) is an indicator function for negative income observations. This allows for including firmswith negative income so that α2jt can be interpreted as an earnings multiple while, at the sametime, firm value is adjusted downward through α3jt if an industry has negative income in a givenyear. Time-averaged regression results are reported in Table II. Accounting for varying discountand growth rates, the regressions coefficients are allowed to vary over time and across industries.

In this setup, under the assumption that similar assets sell at similar prices, we definefundamental value P̂ as the fitted value of the regression in Equation (4), and excess priceas the difference between the observed value and the model-based value

X S Pijt = Pijt − P̂ijt

Pijt. (5)

From Equation (4), it follows that the expected value E[ln(Pijt) − ln(P̂ijt)] is zero for each industryyear. As a result, using the property of the mean of a lognormal distribution, the mean ofXSP is dependent on the standard deviation of the error term in Equation (4). Intuitively, the


Tab

leII.

Reg

ress

ion

Res

ult

sfo

rE

stim

atin

gF

un

dam

enta

lVal

ue

Res

ults

from

regr

essi

ngm

arke

tval

ueon

the

log

ofbo

okva

lue

(ln(

B) ij

t),n

etin

com

e(l

n(N

I+ ijt)

),an

dle

vera

ge.T

hefi

tted

valu

esof

thes

ere

gres

sion

sar

ees

tim

ates

offu

ndam

enta

lva

lue

P̂.T

oin

crea

seth

enu

mbe

rof

obse

rvat

ions

,we

incl

ude

the

abso

lute

valu

esof

net

inco

me

(whi

chca

nbe

nega

tive)

and

incl

ude

adu

mm

yva

riab

leeq

ualt

oon

eif

neti

ncom

eis

nega

tive

(I(<

0)ln

(NI+ ij

t)).

The

colu

mns

indi

cate

tim

e-se

ries

aver

age

regr

essi

onco

effi

cien

tsfo

rthe

Fam

a-Fr

ench

12-i

ndus

try

clas

sifi

cati

on,

plus

min

ing.

Bus

ines

sE

quip

men

t(B

usE

q)in

clud

esco

mpu

ters

,so

ftw

are,

and

elec

tron

iceq

uipm

ent.

Che

mic

als

(Che

m)

incl

ude

chem

ical

san

dal

lied

prod

ucts

.D

urab

les

(Dur

bl)

incl

ude

cars

,T

Vs,

furn

itur

e,an

dho

useh

old

appl

ianc

es.

Ene

rgy

(Enr

gy)

incl

udes

oil,

gas,

and

coal

extr

acti

on(p

rodu

cts)

.H

ealt

hcar

e(H

tlth

)inc

lude

sm

edic

aleq

uipm

enta

nddr

ugs.

Man

ufac

turi

ng(M

anuf

)inc

lude

sm

achi

nery

,tru

cks,

airp

lane

s,of

fice

furn

itur

e,pa

per,

and

com

mer

cial

prin

ting

.M

inin

g(M

ines

)in

clud

esm

inin

gan

dm

iner

als.

Fina

nce

(Mon

ey)

incl

udes

bank

s,in

sura

nce

com

pani

es,

and

othe

rfi

nanc

ials

.N

on-d

urab

les

(NoD

ur)

incl

ude

cons

umer

non-

dura

bles

.Oth

erin

clud

esco

nstr

ucti

on,t

rans

port

atio

n,re

crea

tion

,bus

ines

sse

rvic

es,a

nden

tert

ainm

ent.

Sho

psin

clud

esw

hole

sale

,ret

ail,

and

som

ese

rvic

essu

chas

laun

drie

san

dre

pair

shop

s.Te

leco

m(T

elcm

)in

clud

este

leph

one

and

tele

visi

ontr

ansm

issi

on.U

tili

ties

(Uti

ls)

incl

udes

util

itie

s.Fo

rea

chse

ctor

-yea

rsa

mpl

e,th

ecr

oss-

sect

iona

lmod

elfr

omE

quat

ion

(4)

ises

tim

ated

.Rep

orte

dar

eth

eco

effi

cien

tsth

atre

pres

entt

heti

me-

seri

esav

erag

em

ulti

ple

rela

ting

the

acco

unti

ngva

riab

leto

mar

ketv

alue

.Als

ore

port

edar

eth

eti

me-

seri

esav

erag

esa

mpl

esi

zean

dR

2.F

ama-

Mac

beth

(197

3)st

anda

rder

rors

are

repo

rted

inpa

rent

hese

s.

Bu

sEq

Ch

ems

Du

rbl

En

rgy

Hlt

hM

anu

fM

ines

Mo

ney

No

Du

rO

ther

Sh

op

sT

elcm

Uti

ls

ln(B

) ijt

0.46

∗∗∗

0.4∗∗

∗0.

44∗∗

∗0.

43∗∗

∗0.

38∗∗

∗0.

41∗∗

∗0.

46∗∗

∗0.

38∗∗

∗0.

39∗∗

∗0.

43∗∗

∗0.

41∗∗

∗0.

41∗∗

∗0.

3∗∗∗

(0.0

6)(0

.08)

(0.1

1)(0

.03)

(0.0

5)(0

.05)

(0.0

3)(0

.06)

(0.0

5)(0

.06)

(0.0

6)(0

.02)

(0.0

4)ln

(NI+ ij

t)0.

40∗∗

∗0.

36∗∗

∗0.

43∗∗

∗0.

36∗∗

∗0.

48∗∗

∗0.

38∗∗

∗0.

3∗∗∗

0.47

∗∗∗

0.44

0.38

∗∗∗

0.41

∗∗∗

0.44

∗∗∗

0.46

∗∗∗

(0.0

6)(0

.06)

(0.0

9)(0

.05)

(0.0

8)(0

.06)

(0.0

6)(0

.06)

(0.0

7)(0

.07)

(0.0

7)(0

.05)

(0.0

9)I (

<0)

ln(N

I+ ijt)

−0.1

2∗∗∗

−0.1

00.

09−0

.14

−0.4

7∗∗∗

0.15

−0.3

6−0

.23

0.04

−0.1

2−0

.18

0.06

−0.0

2(0

.02)

(0.5

9)(0

.39)

(0.4

2)(0

.17)

(0.1

5)(0

.41)

(0.2

9)(0

.24)

−0.2

2(0

.19)

(0.4

1)(0

.41)

Lev

erag

e−2

.94∗∗

∗−2

.53∗∗

∗−2

.63∗∗

∗−2

.96∗∗

∗−3

.46∗∗

∗−2

.77∗∗

∗−2

.99∗∗

∗−1

.95∗∗

∗−3

.05∗∗

∗−2

.44∗∗

∗−2

.55∗∗

∗−2

.67∗∗

∗−2

.93∗∗

∗

(0.3

2)(0

.45)

(0.4

9)(0

.36)

(0.5

2)(0

.27)

(0.1

9)(0

.23)

(0.3

9)(0

.32)

(0.3

1)(0

.36)

(0.4

6)C

onst

ant

2.84

∗∗∗

3.53

∗∗∗

2.98

∗∗∗

3.44

∗∗∗

3.27

∗∗∗

3.32

∗∗∗

3.18

∗∗∗

2.87

∗∗∗

3.19

∗∗∗

3.06

∗∗∗

3.14

∗∗∗

3.15

∗∗∗

3.96

∗∗∗

(0.2

4)(0

.55)

(0.4

6)(0

.53)

(0.4

)(0

.26)

(0.0

8)(0

.27)

(0.3

7)(0

.32)

(0.3

1)(0

.41)

(0.5

7)O

bs.

284

6968

8612

625

641

462

145

240

210

6786

R2

0.72

0.61

0.68

0.73

0.77

0.63

0.75

0.61

0.69

0.61

0.66

0.75

0.62

∗∗∗ S

igni

fica

ntat

the

0.01

leve

l.


Table III. Correlation between Value Components (Firm Level)

The table reports coefficients of the correlation between the three value components on firm level (i.e.,market-based growth options value (V M

G O ), fundamental growth options value (V FG O ), and excess pricing

(XSP)). We report standard errors in parentheses. All correlations are significant at the 1% level.

(VMGO) VF

GO XSP

V MG O 1

V FG O 0.4231 (0.006) 1

XSP 0.6374 (0.005) −0.3369 (0.006) 1

coefficients link accounting information to firm value, much akin to conventional measuressuch as M /B and price-to-earnings. Therefore, the coefficients have an (conditional) accountingmultiple interpretation. This approach simultaneously considers different accounting measures,so that we link accounting numbers to prices over multiple dimensions of firm value.

Table III reports correlations between the three value components in Equation (3): 1) the market-based growth options value, 2) the fundamental growth options value, and 3) excess pricing. Itcan be observed that the correlation between market-based growth value and fundamental growthvalue, and the correlation between the market-based growth value and mispricing are both positiveand almost add up to one. Additionally, excess pricing and fundamental growth option value movein opposite directions. Table IV reports excess pricing and the mean growth options value byindustry over our firm-level sample, while Table V does so by year. To facilitate comparison withprevious research (e.g., Tong, Alessandri, et al., 2008) and provide a more intuitive explanation ofthe growth options component between industries, we scale the summary statistics by companyvalue. As a first observation of growth value across industries and years, large differences existbetween firms. A standard deviation of approximately 0.66 is common in most years and sectors.This indicates that large differences exist in the appropriation of option values. The marketrewards firms that are better able to capitalize upside potential.

B. Growth Options and Mispricing by Sector

In terms of the market-based growth options value, important differences exist between sectors.The highest market-based growth options value is found in the business equipment (V M

G O =0.96), mining (V M

G O = 0.92), telecom (V MG O = 0.89), and health-care (V M

G O = 0.87) sectors.These industries have more volatile demand and earnings, which increase their option value.Furthermore, these sectors are generally characterized by a large number of small and youngfirms that do not yet have scale economies and consist largely of immaterialized value. This isconsistent with real option theory and previous empirical work, as these industries involve moreunexpected technological changes and moves by competitors. In contrast, utilities (V M

G O = 0.71)and financial services (V M

G O = 0.74) exhibit the lowest growth options value. For these more stable,low-growth income sectors, earnings are more predictable and scale economies are important forevery firm in the sector. This increases the importance of the value of assets in place andproportionally decreases growth value.

Because XSP is estimated for each sector separately, XSP at the sector level has no clearinterpretation of mispricing between sectors. However, by breaking the sample down into bidderand target subsamples within each sector, XSP can be used as a measure of overpricing at the firm


level. We observe that bidders are generally more overpriced than targets, and that this fluctuatesrandomly over industries.

C. Growth Options and Mispricing over Time

From the reported average growth options value by year in Table V, we observe that differ-ences in growth options value are much smaller over time than between industries. The highestsample-wide growth value is observed in 2001, and the lowest in 1995 and 1996. For the

Table IV. Summary Statistics by Sector

Summary statistics are reported per sector from 1996 to 2006. Sector definitions are based on the clas-sification as provided by Kenneth French on his website. Sector definitions are as described in Table II.For each sector, the second column reports the number of observations with nonmissing values for market-based growth options value (V M

G O ), fundamental growth options value (V FG O ), and excess pricing (XSP).

The remaining columns report sample means of V MG O , V F

G O and XSP, respectively, with standard errors inparentheses. V M

G O , V FG O , and XSP are normalized to firm value. To make the XSP mean consistent with

Equation (3), we ignore observations that have nonmissing values for XSP, but not for V MG O and V F

G O .

Sector All

N VMGO VF

GO XSP

Business equipment 3,687 0.96 0.84 0.35(0.73) (0.58) (0.84)

Mining 519 0.92 0.74 0.49(0.75) (0.60) (0.82)

Telecom 1,003 0.89 0.75 0.54(0.7) (0.56) (0.88)

Healthcare 1,679 0.87 0.73 0.39(0.71) (0.56) (0.87)

Durables 922 0.82 0.71 0.54(0.66) (0.49) (0.80)

Chemicals 996 0.82 0.67 0.53(0.65) (0.59) (0.85)

Manufacturing 3,688 0.82 0.72 0.52(0.64) (0.50) (0.78)

Other 3,334 0.82 0.72 0.50(0.64) (0.56) (0.8)

Energy 1,249 0.81 0.66 0.54(0.68) (0.59) (0.82)

Shops 3,124 0.79 0.70 0.49(0.63) (0.53) (0.78)

Nondurables 2,131 0.78 0.65 0.50(0.65) (0.59) (0.84)

Finance 1,043 0.74 0.76 0.39(0.57) (0.46) (0.72)

Utilities 1,216 0.71 0.64 0.60(0.55) (0.43) (0.68)

Total 24,591 0.83 0.72 0.48(0.66) (0.55) (0.81)

(Continued)


Table IV. Summary Statistics by Sector (Continued)

Bidders

N VMGO VF

GO XSP

472 1.01 0.82 0.43(0.76) (0.55) (0.86)

88 0.96 0.72 0.54(0.78) (0.70) (0.89)

168 1.03 0.74 0.74(0.74) (0.64) (0.94)

204 0.99 0.75 0.51(0.73) (0.55) (0.89)

91 0.92 0.77 0.58(0.66) (0.40) (0.76)

93 0.81 0.61 0.64(0.62) (0.86) (0.97)

375 0.83 0.68 0.58(0.63) (0.58) (0.77)

380 0.89 0.72 0.60(0.65) (0.56) (0.82)

143 0.86 0.69 0.63(0.69) (0.51) (0.75)

367 0.88 0.72 0.59(0.65) (0.46) (0.79)

218 0.87 0.65 0.64(0.66) (0.81) (0.91)

107 0.86 0.72 0.63(0.53) (0.51) (0.72)

125 0.88 0.68 0.76(0.59) (0.46) (0.77)

2,831 0.91 0.72 0.58(0.68) (0.58) (0.84)

Targets

N VMGO VF

GO XSP

92 1.05 0.87 0.38(0.72) (0.44) (0.81)

18 1.02 0.86 0.39(0.72) (0.38) (0.73)

31 0.98 0.90 0.50(0.70) (0.48) (0.77)

58 0.87 0.79 0.26(0.63) (0.52) (0.78)

17 0.82 0.62 0.64(0.78) (0.55) (1.01)

24 0.74 0.78 0.18(0.51) (0.45) (0.70)

99 0.81 0.74 0.39(0.72) (0.71) (0.93)

69 0.81 0.81 0.38(0.59) (0.48) (0.70)

28 1.05 0.95 0.44(0.68) (0.74) (0.82)

81 0.57 0.73 0.19(0.56) (0.51) (0.59)

(Continued)


Table IV. Summary Statistics by Sector (Continued)

Targets

N VMGO VF

GO XSP

42 0.56 0.61 0.28(0.64) (0.34) (0.75)

14 0.78 0.90 0.25(0.55) (0.31) (0.82)

54 0.76 0.63 0.69(0.44) (0.43) (0.56)

627 0.82 0.77 0.37(0.65) (0.53) (0.77)

fundamental growth options value, the highest growth value is also observed in 2001 and thelowest in 1995. Compared to the sample average, excess pricing for bidders is higher in everyyear; for targets, it is lower. It can also be seen that the variation in mispricing is slightly less ran-dom between years as between sectors. In most years, the fundamental growth value for biddersis lower than the sample average; for targets, it is mostly higher.

IV. Results

A. Firm-Level Results

This section examines aggregate evidence that real growth options and mispricing play asignificant role in aggregated M&A transactions. Table VI provides a comparison of firm-levelsummary statistics based on whether a firm was involved in a merger or not and, if so, their role aseither bidder or target. Unless mentioned otherwise, results are significant at a 99% confidencelevel.

1. Bidders Are Overvalued

From Table VI, we confirm that bids are placed by “the very highest M /B firms” (M /B =4.94 for bidders; M /B = 4.40 for targets; M /B = 4.20 for nontakeovers) (Rhodes-Kropf et al.,2005). Firms that are involved in takeovers have a higher M /B than firms that are not. Also, themarket-based growth value for bidders is significantly larger than for nontakeovers (V M

G O = 2.69vs. V M

G O = 2.28). Both findings may indicate either growth opportunities or mispricing.However, under the assumption that merging firms are randomly selected across sectors, bidding

firms are typically more overpriced than nonmerging firms (XSP = 0.85 vs. XSP = 0.58). At thesame time, the fundamental growth options value for merging firms is significantly larger thanfor nonmergers. Thus, mispricing and growth options value coexist in takeovers and the high M /Bratios in previous studies cannot be entirely ascribed to mispricing. While mispricing and growthoptions are both significant, the mispricing component slightly dominates.

Both bidders and targets have strong fundamental growth options. Therefore, in terms ofreal option theory, the option to “exchange” its shares for another firm’s shares is an equallyimportant argument for placing a bid as the growth options argument. This confirms our empiricalobservation but also supports Hypotheses 1a and 1b. Bidders have highly priced equity that islargely due to mispricing, and they use it to acquire fundamental growth opportunities that do notdiffer significantly from theirs.

More insight is obtained by examining the difference between market-based growth optionsand the fundamental growth options value. A correction for mispricing decreases the bidders’


Table V. Summary Statistics by Year

In this table, summary statistics are reported per year from 1996 to 2006. The first column of eachsubsample reports the number of observations with nonmissing values for market-based growth optionsvalue (V M

G O ), fundamental growth options value (V FG O ), and excess pricing (XSP) to match the availability

of the SDC database. The second column reports takeover activity per year (Activity), defined as (NBidders

+ NTargets)/NAll. All other columns report sample means, with standard errors in parentheses. V MG O , V F

G O ,and XSP are normalized to firm value. To make the XSP mean consistent with Equation (3), we ignoreobservations that have nonmissing values for XSP, but not for V M

G O and V FG O .

All

Year Activity N VMGO VF

GO XSP

1995 5.3% 1,330 0.75 0.61 0.51(0.65) (0.56) (0.83)

1996 7.6% 2,327 0.79 0.68 0.45(0.64) (0.48) (0.79)

1997 12.3% 2,438 0.83 0.71 0.48(0.65) (0.52) (0.80)

1998 16.6% 1,834 0.85 0.75 0.41(0.67) (0.49) (0.81)

1999 19.8% 2,340 0.85 0.73 0.52(0.67) (0.54) (0.81)

2000 17.8% 2,128 0.79 0.70 0.48(0.67) (0.58) (0.83)

2001 14.3% 2,099 0.90 0.82 0.48(0.67) (0.56) (0.80)

2002 11.0% 2,067 0.89 0.80 0.49(0.67) (0.57) (0.80)

2003 13.0% 2,057 0.82 0.72 0.54(0.67) (0.63) (0.81)

2004 13.3% 1,781 0.82 0.69 0.46(0.68) (0.59) (0.85)

2005 16.7% 2,117 0.83 0.72 0.46(0.66) (0.50) (0.82)

2006 19.0% 2,073 0.84 0.73 0.46(0.65) (0.51) (0.80)

Total 14.6% 24,591 0.83 0.72 0.48(0.66) (0.55) (0.81)

Bidders

N VMGO VF

GO XSP

62 0.70 0.50 0.61(0.69) (0.66) (0.84)

111 0.70 0.50 0.61(0.69) (0.66) (0.84)

183 0.86 0.72 0.55(0.68) (0.50) (0.77)

236 1.00 0.79 0.65(0.70) (0.50) (0.81)

346 0.94 0.75 0.58(0.67) (0.51) (0.84)

321 0.98 0.75 0.69(0.68) (0.50) (0.75)

(Continued)


Table V. Summary Statistics by Year (Continued)

Bidders

N VMGO VF

GO XSP

269 0.90 0.71 0.60(0.69) (0.70) (0.90)

204 0.96 0.82 0.53(0.63) (0.40) (0.75)

234 0.94 0.74 0.60(0.67) (0.76) (0.89)

220 0.81 0.65 0.59(0.67) (0.85) (0.95)

302 0.84 0.69 0.48(0.72) (0.47) (0.80)

343 0.86 0.69 0.53(0.68) (0.54) (0.86)

2,831 0.91 0.72 0.58(0.68) (0.58) (0.84)

Targets

N VMGO VF

GO XSP

11 1.09 0.92 0.48(0.78) (0.68) (0.99)

72 1.09 0.92 0.48(0.78) (0.68) (0.99)

124 0.88 0.71 0.50(0.71) (0.58) (0.88)

56 0.93 0.82 0.47(0.61) (0.45) (0.73)

109 0.78 0.82 0.05(0.74) (0.72) (0.74)

60 0.81 0.78 0.40(0.63) (0.60) (0.69)

33 0.77 0.71 0.44(0.68) (0.51) (0.80)

23 0.91 0.87 0.37(0.69) (0.40) (0.77)

35 0.65 0.91 −0.04(0.59) (0.39) (0.64)

17 0.77 0.73 0.56(0.51) (0.49) (0.69)

45 0.66 0.49 0.26(0.78) (0.63) (1.15)

42 0.74 0.75 0.27(0.65) (0.41) (0.75)

627 0.82 0.77 0.37(0.65) (0.53) (0.77)


Table VI. Characteristics of Merger and Nonmerger Firms (Firm Level)

Summary statistics for size as measured by market value (V ), capital invested (CI), market-to-book (M /B),excess pricing (XSP), market-based growth value (V M

G O ), and fundamental growth value (V FG O ). Active firms

are observed as either bidders or targets in the SDC database from 1996 to 2006, while inactive firms are not.The column t(diff) reports the student t-statistic for the hypotheses: active firm − inactive firm = 0 or bidder− target = 0, correcting for unequal variances across groups using Satterthwaite’s (1946) approximationformula for the degrees of freedom. Observations are required to have M /B ratios between −100 and 100.V M

G O , V FG O , and XSP are normalized to equity value. To make the XSP mean consistent with Equation (3) in

the three bottom rows, we ignore observations that have nonmissing values for XSP, but not for V MG O and

V FG O .

Nonmergers Mergers t(diff) Bidders Targets t(diff)

V Obs. 43,328 7,334 5,862 1,472Mean 2,130 4,051 −9.001∗∗∗ 4,708 1,434 11.328∗∗∗

CI Obs. 40,606 5,250 4,062 1,188Mean 1,866 3,796 −6.741∗∗∗ 4,535 1,269 8.440∗∗∗

M /B Obs. 39,348 5,758 4,873 885Mean 4.203 4.857 −4.384∗∗∗ 4.940 4.403 1.444

Bidders and targets: All (firm level)Obs. 21,214 3,513 2,866 647

XSP Mean 0.583 0.784 −5.591∗∗∗ 0.853 0.477 4.940∗∗∗

V MG O Mean 2.275 2.596 −5.81∗∗∗ 2.689 2.182 4.125∗∗∗

V FG O Mean 1.693 1.812 −2.778∗∗∗ 1.836 1.704 1.353

Bidders and targets: Matched deal pairs (deal level)Obs. 660 660

XSP Mean 1.597 0.842 5.621∗∗∗

V MG O Mean 0.223 0.441 3.536∗∗∗

V FG O Mean 1.655 1.558 0.589

∗∗∗Significant at the 0.01 level.

mean fundamental growth options value by a third. If anything, the corrected value for bidders isnow closer to the value of nonmerging firms (from V M

G O = 2.68 to V FG O = 1.84 vs. V F

G O = 1.69 fornontakeovers). This demonstrates how the conventional growth options measure overestimatesthe growth value for bidders. The correction is smaller for targets, which supports Hypothesis 1b.

2. Bidders Buy Smart

We also confirm that “targets have higher M /B ratios than the average firm” (M /B = 4.40for targets; M /B = 4.20 for nontakeovers) (Rhodes-Kropf et al., 2005), but this difference isinsignificant for our international sample. Also, the market-based growth options value fortargets is not significantly higher than for nontakeovers (resp. V M

G O = 2.18 vs. V MG O = 2.28).

Consistent with Rhodes-Kropf et al. (2005) and Dong et al. (2006), targets have a lower M /B ratiothan bidders (M /B = 4.40 vs. M /B = 4.94). Again, both may indicate either growth opportunitiesor mispricing.

However, under the assumption of random selection across sectors, Table VI demonstrates thattargets are underpriced relative to bidders and nonmerging firms (resp. XSP = 0.48 vs. XSP =0.85 and XSP = 0.58) and possess similar fundamental growth potential compared to bidders andnonmerging firms (resp. V F

G O = 1.70 vs. V FG O = 1.84 and V F

G O = 1.69). At the same time, excesspricing for targets is significantly larger than zero. The growth options value and mispricing


coexist for targets as well. This supports Hypothesis 1b. From a growth options perspective,targets have lower priced equity with similar fundamental value.

According to the conventional measure, targets have less growth options value than bidders(V M

G O = 2.18 vs. V MG O = 2.69), which is counterintuitive. However, controlling for mispricing

decreases the targets’ growth options value to a lesser extent than the bidders’ growth optionsvalue, consistent with Hypothesis 1b.

While mispricing and growth options value are both important, targets may be rationally selectedfor being cheaper while equally desirable from a real options perspective. The conventional growthoptions measure does a better job for targets than for bidders but ignores a significant mispricingcomponent. This also supports Hypothesis 1b.

While bidders, on average, have similar V FG O than targets, it could be that many of the bidders

have higher V FG O in realized deal pairs. To further examine this, we also compare means on the

deal level, for matched deal pairs. The third panel reports that the means differ more for XSP.However, the difference between means remains insignificant for V F

G O . This is consistent withthe results on nondeal pairs.

B. Deal-Level Results

The previous section pooled together observations across countries and industries. Therefore,the results may well be due to industry or country level effects since capital markets in differentcountries have different momentums and idiosyncrasies that may not randomly occur acrosssectors. One may also expect generic macroeconomic factors to play a significant role. Usingdeal-level evidence, this section controls for country, sector, and year effects.6 To examine whethercountry, sector, and year effects have explanatory power, we estimate the following system of twoequations in a setup that will serve as a basis for further inference. We estimate the two-valuecomponents simultaneously since a change in mispricing will also affect fundamental growth andvice versa. Let each equation contain N observations:

Y = β0 + {D} + e, (6)

in which we have compactly written the stacked equations Y = [XSP, V FG O ]′. The estimation of

Equation (6) is done separately for bidders and targets at the deal level. Specifically, for eachdeal, we obtain the target’s XSP and V F

G O and the bidder’s XSP and V FG O , if available.

Each of the disturbances in vectors e have zero mean, equal variance, and are uncorrelated, butcovariances between the two vectors are potentially nonzero for a pair of equations. To allow forsufficient degrees of freedom, we exclude V M

G O from the analysis (since it is a linear combinationof V F

G O and XSP). Two sets of dummies {D} (one for each value component) include the unionof five subsets containing indicator variables for each year and, specific to bidders or targets,each sector and nation.7 As a consequence of including target dummies for bidders and bidderdummies for targets, some observations are lost as not all deals can be matched into pairs.

This specification is appropriate since excess pricing and fundamental growth value may besignificantly correlated and subject to similar shocks. Therefore, it may be difficult to disentanglethe two value components. By using seemingly unrelated regressions (SURs) that allow for

6 We thank the anonymous referee for this suggestion.7 Specifically, the set {D} contains 13 bidder sector dummies, 13 target sector dummies, 84 bidder country dummies, 129target country dummies, and 12-year dummies minus 1. Of the 251 dummies, 87 are dropped for reasons of collinearity.For the sake of brevity, we do not report the remaining 164 dummy results in subsequent analyses.


correlated error terms, we can pick up any overlap in growth opportunities or mispricing anddisentangle them more effectively.

As can be seen from Columns (I) and (III) in Table VII, industry, year, and country effectsexplain a significant portion of variation in our value components. These effects explain 17%(10%) of the variation in V F

G O (XSP) for bidders and 13% (16%) of the variation in V FG O (XSP) for

targets. From the reported R2 statistics, one can observe that country, industry, and year effectsare more important for overpriced targets (R2 = 16%) than for overpriced bidders (R2 = 10%)indicating that the “external shocks” motivation from Andrade et al. (2001) holds more for targetsthan for bidders.

1. Bidders Time Smart

The setup in Equation (6) allows us to test our hypotheses in a multivariate context. For instance,if the perceived market value of acquirers plays an important role when bids are made, weexpect an increase in takeover activity to be strongly related to an increase in bidder mispricing.To further examine correlations with takeover activity, we estimate the following system ofequations:

Y = β0 + β1ACTIV + {D} + e, (7)

in which takeover activity (ACTIV) is defined as the number of deals that a bidder or target isinvolved in during a fiscal year. Specifically, we determine how many bids a bidder places (ora target receives) during each fiscal year, and match the total to the bidder and/or target in thedeal-level data. If a deal is matched, this yields two takeover activity values per deal (i.e., one forthe bidder and one for the target). As stated previously, we estimate Equation (7) separately forbidders and targets with a different takeover variable for bidder and target estimates.

Columns (II) and (IV) in Table VII provide evidence that bidders time their acquisitionswell. After controlling for country, sector, and year effects, overpricing for bidders is positivelyassociated with takeover activity. Hence, when market prices of bidders are high, takeover activityis relatively large. This is evidence in support of Hypothesis 2a.

We also observe that in a multivariate regime, overpricing for targets is strongly negativelycorrelated with takeover activity. This indicates that bids are made for targets that are undervaluedas compared to the bidder. The relationship is significant at the 1% to 5% significance level andprovides significant evidence in favor of Hypothesis 2b.

After controlling for year, sector, and country effects, target growth value demonstrates norelationship with takeover activity. This suggests that buyers also buy smart by bidding for targetswith growth potential that does not depend on the number of takeovers in the economy. Whileinsignificant, the negative sign suggests that target V F

G O is higher when bidder V FG O is low.

When interpreting these findings, it is important to acknowledge that some firms may haveengaged in multiple bids, while other firms may be targeted by multiple acquirers. As this wouldintroduce correlation between residuals, it is possible that standard errors have been calculatedunder inappropriate assumptions. To account for this possibility, we repeat the estimation ofEquation (7) 1,000 times, bootstrapping the standard errors by resampling observations from thedata. We observe in Table VII that the obtained nonparametric standard errors do not affect thevalidity of our results and that the effect of takeover activity becomes slightly stronger for targets.

In addition, the bootstrapped standard errors indicate that some of bidder V FG O is also positively

associated with takeover activity. This suggests that takeover activity is higher when bidders havefundamental growth opportunities. Again, growth options and mispricing seem to coexist.


Tab

leV

II.B

idd

ers

Tim

eS

mar

t(D

ealL

evel

)

Tabl

eV

IIpr

ovid

esth

ere

sult

sfr

omE

quat

ions

(6)

and

(7),

ase

emin

gly

unre

late

dre

gres

sion

ofth

efu

ndam

enta

lgro

wth

opti

ons

valu

e(V

F GO

)an

dex

cess

pric

ing

(XSP

)on

ase

tof

13-s

ecto

rdu

mm

ies,

84bi

dder

coun

try

dum

mie

s,12

9ta

rget

coun

try

dum

mie

s,an

d12

-yea

rdu

mm

ies,

and/

orta

keov

erac

tivit

y(A

CT

IV).

VF G

Oan

dX

SPar

eno

rmal

ized

toeq

uity

valu

e.C

oeff

icie

nts

for

sect

or,n

atio

n,an

dye

ardu

mm

ies

are

notr

epor

ted,

butt

este

dfo

rjo

ints

igni

fica

nce.

Res

ults

for

bidd

ers

and

targ

ets

are

obta

ined

sepa

rate

ly.N

orm

alst

anda

rder

rors

are

inro

und

pare

nthe

ses;

boot

stra

pped

stan

dard

erro

rsar

ein

squa

rebr

acke

ts.

Bid

der

sT

arg

ets

(I)

(II)

(III)

(IV

)

VF G

OX

SP

VF G

OX

SP

VF G

OX

SP

VF G

OX

SP

AC

TIV

0.47

40.

337

−0.1

35−0

.278

(0.3

20)

(0.1

40)∗∗

(0.1

63)

(0.1

15)∗∗

[0.2

48]∗

[0.1

80]∗

[0.1

80]

[0.1

07]∗∗

∗

Sec

tor

YE

S∗

YE

S∗

YE

SY

ES

Cou

ntry

YE

S∗∗

∗Y

ES

∗∗∗

YE

S∗∗

∗Y

ES

∗∗∗

Yea

rY

ES

∗∗∗

YE

S∗∗

∗Y

ES

∗∗∗

YE

S∗∗

∗

Obs

.2,

089

2,08

92,

089

2,08

992

792

792

792

7R

20.

167

0.10

00.

168

0.10

20.

129

0.15

90.

129

0.16

4

∗∗∗ S

igni

fica

ntat

the

0.01

leve

l.∗∗

Sig

nifi

cant

atth

e0.

05le

vel.

∗ Sig

nifi

cant

atth

e0.

10le

vel.


2. Bidders Buy Smart

To demonstrate that our value components measure what they are supposed to measure in amultivariate context, we consider several M&A deal attributes that are desirable for overvaluedbidders when compared to bidders with a relatively high proportion of fundamental growthpotential. First, overvalued bidders gain from paying targets with stock rather than cash suggestingthat the method of payment is an important consideration. Additionally, even if a deal is paid forin cash, an overvalued bidder can finance the acquisition inexpensively by raising the requiredamount of cash through a seasoned equity issuance. Moreover, it is expected that a target’s boardof directors will be unwilling to agree to merge with an overvalued bidder. Consequently, weexpect hostile bids if bidders are overvalued. We estimate the following system of equationsseparately for bidders and targets:

Y = β0 + β0 ACTIV + β2StockPay + β3 1StockFin + β4 1Hostile + {D} + e. (8)

In these equations, Y and {D} are as defined previously. In addition, StockPay represents thepercentage of consideration paid in stock measured as the value paid in stock divided by the totaltransaction value. 1StockFin represents an indicator variable equal to one if the transaction was paidin cash, but financed by issuing new equity, and zero otherwise. The indicator variable 1hostile isequal to one if the target company’s management or board of directors has a hostile attitude towardthe transaction or issues a negative recommendation, and zero otherwise. The SDC database doesnot always have information about the form of payment and, to a lesser extent, the attitude of thetarget. Hence, the number of observations drops further when compared to previous results.

In Table VIII, the left (bidder) columns report that stock payment, stock financing, and a hostileattitude are all significant and positively related to bidder mispricing. Overvalued firms areassociated with a high average stock pay percentage and more often finance M&A transactionswith equity issues. This result indicates that bidders use their overvalued equity to financetheir acquisitions “cheaply” and may aim to protect themselves from future drops in shareprice. Overvalued equity may also explain why many (stock-for-stock) mergers are approved,as found by Bethel and Gillen (2002) and Burch, Morgan, and Wolf (2004). Overvalued firmsalso place hostile bids more often indicating that target firms do not see much value in themerger. Furthermore, year effects significantly affect bidders only. At the same time, biddersare more inclined to hostile takeovers when they have strong growth opportunities. Other dealcharacteristics have no relation to the bidders’ fundamental growth value. These findings providefurther support for Hypothesis 1a.

After controlling for deal characteristics, takeover activity continues to be positively associatedwith bidder overpricing, providing further evidence that bidders with overvalued equity anticipatea future price correction. This finding continues to hold when standard errors are obtainednonparametrically. This adds further support to Hypothesis 2a.

We also examine how these attributes affect each value component for targets. The two right-hand (target) columns of Table VIII present results for target firms. They indicate that stockpayment is associated with lower target V F

G O . This indicates that target growth options are lessvaluable if target shareholders allow the deal to be financed with equity.

M&A transactions that involve equity issuance are associated with targets having higher growthvalue. This implies that equity issuance by bidders is positively related to growth options valuefrom targets. However, nonparametric standard errors are larger than those under the assumptionof i.i.d. residuals suggesting that this result is driven by bidders engaging in multiple bids.


Table VIII. Bidders Buy Smart (Deal Level)

Table VIII presents the results from Equation (8), a seemingly unrelated regression of the fundamentalgrowth options value (V F

G O ) and excess pricing (XSP) on deal payment (StockPay), deal financing (1StockFin),and deal attitude (1hostile), as well as a set of 13-sector dummies, 84 bidder country dummies, 129 targetcountry dummies, and 12-year dummies. V F

G O and XSP are normalized to equity value. Deal coefficients forsector, nation, and year dummies are not reported, but tested for joint significance. The results for biddersand targets are obtained separately. Normal standard errors are in round parentheses; bootstrapped standarderrors are in square brackets. Significance levels are based on bootstrapped standard errors.

Bidders Targets(I) (II)

VFGO XSP VF

GO XSP

ACTIV 1.111 0.618 0.287 0.0413(0.380)∗∗∗ (0.293)∗∗ (0.208) (0.168)[0.409]∗∗∗ [0.346]∗ [0.300] [0.169]

StockPay 0.00383 0.016 −0.012 −0.004(0.0120) (0.0090)∗ (0.0060)∗∗ (0.0040)[0.0090] [0.0070]∗∗ [0.0070]∗ (0.0050)

1StockFin 0.744 6.877 3.490 −1.190(4.2510) (3.2810)∗∗ (1.7020)∗∗ (1.3770)[4.4550] [7.3870] [2.8490] (1.4750)

1hostile 3.349 7.957 −1.00 0.485(1.5750)∗∗ (1.2130)∗∗∗ (0.7250) (0.5850)[1.8620]∗ [4.7930]∗ [0.6940] (0.7210)

Sector YES YESCountry YES YESYear YES∗∗ YESObs. 673 673 298 298R2 0.179 0.170 0.240 0.183

∗∗∗Significant at the 0.01 level.∗∗Significant at the 0.05 level.∗Significant at the 0.10 level.

Finally, for targets, takeover activity is not significantly associated with a target’s overpricingafter controlling for deal characteristics. This indicates that targets are selected that are lesssubject to overpricing, which has a cushioning effect on future price corrections. This is furtherevidence that bidders buy smart by cheaply buying targets that have similar fundamental growthvalue and adds support to Hypothesis 2b. Nonparametric standard errors do not affect the validityof this result.

V. Conclusions

Considering the large variation in takeover activity over the last 50 years, the question as towhat drives takeovers is of interest to a large group of scholars and practitioners. In answering thisquestion for the most recent takeover wave, rational explanations based on growth opportunities,as well as behavioral explanations based on mispricing, have strong theoretical underpinnings andare backed by prior research in strategy and financial economics. We challenge an exclusive claimto wisdom from either strand of literature and demonstrate that mispricing and growth options


explanations coexist. This study explains merger activity by bidders that have high M /B ratiosbecause they are overpriced and because genuine growth potential is present. Simultaneously,we justify target selection by targets that have above average M /B ratios due to their fundamentalgrowth potential, and that are less overpriced.

A. Empirical Implications

First, we find that bidders are overvalued but also have strong growth potential. Bidders typicallyhave similar or higher growth potential and more overpriced equity than targets and nonmergingfirms. When controlling for mispricing, the growth options value decreases more for biddersthan for other firms. Thus, although bidders seem to have growth options value, they are moreoverpriced than targets. Positive correlations with takeover activity indicate that bidders are ableto time their bids smart. When bidders’ mispricing and growth options increase over time, so doesthe frequency of bidding announcements. Finally, our regressions demonstrate that these resultspersist after controlling for country, industry, and year effects.

Additionally, bidders buy smart and time smart. Targets possess similar growth potential butless mispricing than bidders and nonmerging firms. When controlling for mispricing, the growthoptions value decreases to a lesser extent for targets than for other firms. Therefore, although someoverpricing exists, a target’s growth option value is more strongly based on fundamentals. Negativecorrelations of targets with takeover activity demonstrate that bidders time their acquisition smart.Targets are undervalued relative to bidders when the bids are made. Buying fundamental growthmay cushion bidders against a future drop in market value. Furthermore, growth option valueis acquired that does not depend on takeover activity, which allows bidders to retain growthopportunities after bidder growth options have declined with takeover activity.

These results follow only if more precise measures than the conventional M /B or market-basedgrowth options value are used such as the fundamental growth option value. This fundamentalgrowth option measure is straightforward to implement by using an intuitive decompositionof a firm’s share price. The market-based growth options measure (i.e., without correcting formispricing) yields inconsistent results if used for bidders, and the fit can be substantially improvedfor targets by measuring the fundamental growth options value.

B. Contributions

Our decomposition of firm value offers new insights into each of the three value components.With respect to real options theory, this study is the first to provide evidence that takeoversconfer significant embedded growth opportunities. We also demonstrate, however, that the correctapplication of real options theory is contingent upon correct market valuations, takeover activity,involvement in a merger, the role of being a bidder or target and, more generally, time and sectordifferences. With respect to behavioral finance theory, this study validates untested predictions thatfollow from Shleifer and Vishny (2003). That is, high market values induce firms to bid for targetsthat are undervalued when compared to that of bidding firms. Our results are consistent withseveral firm and deal characteristics. We have demonstrated that these firms are likely to pay forthe deal using their overvalued shares, and are more likely to conduct hostile takeovers. Acquirersbid when takeover activity is high and when their future market value is likely to decrease.

The key contribution, however, lies in unifying arguments from both streams of literature andadding nuance to the rationality-irrationality dichotomy. Specifically, while Shleifer and Vishny(2003) argue that bidders select targets merely for being valued relative to the bidder, we confirmthat both bidders and targets may also possess genuine growth potential. Alternatively, while realoptions theory predicts that takeovers are motivated by the growth options value, we demonstrate


that they are, in fact, also triggered by overvalued equity. Thus, by considering the behavioraland growth option components simultaneously, our study indicates that strategic arguments andbehavioral finance are interrelated.

C. Future Research Directions

Integrating the two theories of takeover activity provides interesting directions for futureresearch. We demonstrate the limitations of market-based measures for growth options value andoffer a simple solution. While mispricing is a well-known factor in M&A, distinguishing betweengrowth opportunities and mispricing may add further insight to previous studies in the growth op-tions value of international joint ventures and firm, industry, or country effects. Beyond the M&Acontext, the empirical framework is easily applied to other issues in business strategy. Additionally,our results validate real options explanations for takeovers, but call for analyses that pay specificattention to the irrationality in financial markets. Further theoretical development in real optionstheory may further explore the behavioral opportunities and pitfalls of the option to acquire.

More specific to the complicated nature of M&A transactions, we conjecture that overpricingmay be an important explanation as to why targets go along with overpriced bidders, but futureresearch is needed to address this question. Moreover, we have partially ignored the complexity ofcorporate takeovers. This study does not integrate many previous findings regarding the selectionof targets based on other grounds such as private information and managerial incentives. Sincewe have not controlled for such alternative explanations, future research could examine how thesemotives relate to the three components of market value.

The decomposition in our study can be applied to any setting in which market prices may deviatefrom the fundamental value and simultaneously have good reason to do so. The regression modelthat we use is thin, however, with book value, net income, and leverage being the only covariates.The decomposition of market-based growth options value could depend upon other factors thataffect growth options and mispricing. We have chosen to stay as close as possible to the previousliterature, but acknowledge that the specification leaves room for improvement.

We emphasize that the fundamental growth options value and excess pricing are two residualterms, which we assume additively separable. However, these components are hard to tease apartor demarcate. For instance, we adjust for market sentiment using P̂ijt, but use a single discount rateto take earnings to perpetuity. However, the market-based discount rate may need to be adjustedas well if the markets are irrational.

From a practical perspective, growth options are an important concept for professionals whoneed to select profitable investment opportunities and decide when to do so. From the recent burstof the US real estate bubble and the subsequent condition of the financial markets, the need toadjust for irrational pricing behavior has become abundantly clear. By incorporating irrationalityinto a rational decision framework, our study offers a tool to aid in understanding value creationin a modern, increasingly turbulent world. �

Appendix: Growth Options Value and Economic Profit

The value of the firm (V ) equals equity (P) plus debt (D), and can be expressed as the sum of:1) invested capital that creditors and shareholders have entrusted to the firm over the years (CI),defined as:

total assets − (accounts payable + other current liabilities), (A1)


and 2) the present value (PV ) of all of the firm’s expected economic profit (EP):

V = CI + PV (EP), ( A2)

where PV (EP) consists of a current-level EP component, as well as an EP growth componentthat depends upon the firm’s investments in future growth opportunities:

PV (E P) = PV (Current EP) + PV (EP Growth), ( A3)

Combining these equations, firm value can be rewritten as:

V = C I + PV (Current EP) + PV (EP Growth), ( A4)

where CI and PV (Current EP) are the company-level value of assets in place.Growth measuresthe value of growth options (i.e., VGO), calculated by solving for VGO:

VG O = V − CI − PV (Current EP), ( A5)

where PV (Current EP) is current economic profit, perpetually discounted by the cost of capital.For a single year, Current EP can be expressed as:8

EP = NOPLAT − CI × WACC, ( A6)

where NOPLAT is the firm’s net operating profits less adjusted taxes. It is calculated by deductingall income taxes from net operating profit and adjusting for increases/decreases in deferred taxes(from the balance sheet), which is a source of cash. If deferred taxes from the previous year arenot known, no adjustment is made. WACC is the weighted average cost of capital, defined as:

WACC = D

D + P(1 − T )kd + P

D + Pke, ( A7)

where total debt (D) is the sum of long- and short-term debt, the market value of equity (P) equalsthe share price × the total common shares outstanding, and the income tax (T) is set at 30%.

The cost of debt (kd) is calculated iteratively using interest coverage ratios and default spreadsas in Damodaran (2002). If the earnings are negative, we average earnings over the past five years.When companies are small (i.e., assets worth less than $10 million), we use different spreads. Forfinancial firms, we also use different spreads.The cost of equity (ke) is found using a standardcapital asset pricing model using five-year adjusted betas.9 The most recent date is the same as thatused for determining market value (see Section III). The index used is the S&P 500 Composite

8 EP can be negative if capital CI × WACC is larger than NOPLAT . In economic terms, this means that the investedcapital (or retained earnings invested in capital) will cost a shareholder money and that this investment should be beenpaid out as a dividend. The present value of current level EP is an annuity and negative EP will lead to a negative presentvalue. As a consequence, this will also lead to growth options that exceed firm value.9 Beta is the slope of regressing the security returns on the market index. Therefore, we estimate beta over a rollingwindow of the current year and the four preceding years. Using weekly data, we have approximately 250 observationsper estimate, while daily volatility does not affect the estimates. If less than four years are available, a one- to three-yearestimate of β or ke is used. We use adjusted betas, which yield the most realistic kes of about 10% to 20%, and estimatefuture instead of historical betas. Adjusted betas equal 0.67 × Raw Beta + 0.33 × 1 to adjust for their long-term tendencyto converge toward one.


Index. The market risk premium is assumed to be 8%. For each year, the corresponding average10-year Treasury Bill rate is added to the spreads.

References

Anderson, C.W. and L. Garcia-Feijoo, 2006, “Empirical Evidence on Capital Investment, Growth Options,and Security Returns,” Journal of Finance 61, 171-194.

Andrade, G., M. Mitchell, and E. Stafford, 2001, “New Evidence and Perspectives on Mergers,” Journal ofEconomic Perspectives 15, 103-120.

Baker, M.P., R.S. Ruback, and J.A. Wurgler, 2007, “Behavioral Corporate Finance: A Survey,” in B. EspenEckbo, Ed., Handbook in Corporate Finance: Empirical Corporate Finance, Amsterdam, Elsevier.

Baker, M. and J. Wurgler, 2002, “Market Timing and Capital Structure,” Journal of Finance 57, 1-32.

Barberis, N. and M. Huang, 2001, “Mental Accounting, Loss Aversion and Individual Stock Returns,”Journal of Finance 56, 1247-1292.

Barberis, N. and R. Thaler, 2003, “A Survey of Behavioral Finance,” in G. Constantinides, Rene M. Stulz,and M. Harris, Eds., Handbook of the Economics of Finance, Amsterdam, Elsevier.

Barth, M.E., W.H. Beaver, and W.R. Landsman, 2001, “The Relevance of the Value-Relevance Literaturefor Financial Accounting Standard Setting: Another View,” Journal of Accounting and Economics 31,77-104.

Berk, J.B., R.C. Green, and V. Naik, 1999, “Optimal Investment, Growth Options, and Security Returns,”Journal of Finance 54, 1553-1607.

Bethel, J. and S. Gillan, 2002, “The Impact of the Institutional and Regulatory Environment on ShareholderVoting,” Financial Management 31, 29-54.

Bettis, R. and M. Hitt, 1995, “The New Competitive Landscape,” Strategic Management Journal 16, 7-19.

Bloomfield, R. and R. Michaely, 2004, “Risk or Mispricing? From the Mouths of Professionals,” FinancialManagement 33, 61-81.

Burch, T., W. Christie, and V. Nanda, 2004, “Do Firms Time Equity Offerings? Evidence from the 1930sand 1940s,” Financial Management 33, 5-23.

Burch, T., A. Morgan, and J. Wolf, 2004, “Is Acquiring-Firm Shareholder Approval in Stock-for-StockMergers Perfunctory?” Financial Management 33, 45-69.

Cao, C., T. Simin, and J. Zhao, 2006, “Can Growth Options Explain the Trend in Idiosyncratic Risk?”Review of Financial Studies 21, 2599-2633.

Childs, P.D., S.H. Ott, and A.J. Triantis, 1998, “Capital Budgeting for Interrelated Projects: A Real OptionsApproach,” Journal of Financial and Quantitative Analysis 33, 305-334.

Chung, K.H. and C. Charoenwong, 1991, “Investment Options, Assets in Place and the Risk of Stocks,”Financial Management 20, 21-33.

Chung, K.H., L. Minsheng, and L. Yu, 2005, “Assets in Place, Growth Opportunities and IPO Returns,”Financial Management 34, 65-88.

Copeland, T. and V. Antikarov, 2001, Real Options, New York, Texere Publishing.

Damodaran, A., 2002, Investment Valuation, New York, John Wiley and Sons, Inc.


Daniel, K.D., D. Hirshleifer, and A. Subrahmanyam, 2001, “Overconfidence, Arbitrage, and EquilibriumAsset Pricing,” Journal of Finance 56, 921-965.

Dixit, A.K. and R.S. Pindyck, 1994, Investment Under Uncertainty, Princeton, NJ, Princeton UniversityPress.

Dong, M., D. Hirshleifer, S.H. Teoh, and S. Richardson, 2006, “Does Investor Misvaluation Drive theTakeover Market?” Journal of Finance 61, 725-762.

Fama, E.F. and J.D. MacBeth, 1973, “Risk, Return, and Equilibrium: Empirical Tests,” Journal of PoliticalEconomy 81, 607-636.

Francis, J. and K. Schipper, 1999, “Have Financial Statements Lost Their Relevance?” Journal of AccountingResearch 37, 319-352.

Fu, F., 2010, “Overinvestment and the Operating Performance of SEO Firms,” Financial Management 39,249-272.

Gao, H., 2010. “Market Misvaluation, Managerial Horizon, and Acquisitions,” Financial Management 39,833-850.

Heaton, J., 2002, “Managerial Optimism and Corporate Finance,” Financial Management 31, 33-45.

Hietala, P., S. Kaplan, and D. Robinson, 2003, “What Is the Price of Hubris? Using Takeover Battles to InferOverpayments and Synergies,” Financial Management 32, 5-31.

Holmstrom, B. and S. Kaplan, 2001, “Corporate Governance and Merger Activity in the US: Making Senseof the 80s and the 90s,” Journal of Economic Perspectives 15, 121-144.

Jegadeesh, N., 2000, “Long-Term Performance of Seasoned Equity Offerings: Benchmark Errors and Biasesin Expectations,” Financial Management 29, 5-30.

Jensen, M., 2005, “Agency Costs of Overvalued Equity,” Financial Management 34, 5-19.

Jenter, D., 2004, “Market Timing and Managerial Portfolio Decisions,” Journal of Finance 60, 1903-1949.

Kester, W.C., 1984, “Today’s Options for Tomorrow’s Growth,” Harvard Business Review 62, 153-160.

Klasa, S. and M. Stegemoller, 2007, “Takeover Activity as a Response to Time-Varying Changes inInvestment Opportunity Sets: Evidence from Takeover Sequences,” Financial Management 36, 1-25.

Lambrecht, B.M., 2004, “The Timing and Terms of Mergers Motivated by Economies of Scale,” Journal ofFinancial Economics 72, 41-62.

Larwood, L. and W. Whittaker, 1977, “Managerial Myopia: Self-Serving Biases in Organizational Planning,”Journal of Applied Psychology 62, 194-198.

Longstaff, F.A. and E.S. Schwartz, 2001, “Valuing American Options by Simulation: A Simple Least-SquaresApproach,” Review of Financial Studies 14, 113-147.

Loughran, T. and J. Ritter, 1995, “The New Issues Puzzle,” Journal of Finance 50, 23-51.

Malmendier, U. and G. Tate, 2005, “CEO Overconfidence and Corporate Investment,” Journal of Finance60, 2261-2700.

March, J.G. and Z. Shapira, 1987, “Managerial Perspectives on Risk and Risk Taking,” Management Science33, 1404-1418.

Margrabe, W., 1978, “The Value of an Option to Exchange One Asset for Another,” Journal of Finance 33,177-186.

Miller, K.D. and T.B. Folta, 2002, “Option Value and Entry Timing,” Strategic Management Journal 23,655-665.


Miller, M.H. and F. Modigliani, 1961, “Dividend Policy, Growth, and the Valuation of Shares,” Journal ofBusiness 34, 235-264.

Mitchell, M. and J.H. Mulherrin, 1996, “The Impact of Industry Shocks on Takeover and RestructuringActivity,” Journal of Financial Economics 41, 193-229.

Morck, R., A. Shleifer, and R.W. Vishny, 1990, “Do Managerial Objectives Drive Bad Acquisitions?”Journal of Finance 45, 31-48.

Morellec, E. and A. Zhdanov, 2005, “The Dynamics of Mergers and Acquisitions,” Journal of FinancialEconomics 77, 649-672.

Muelbroek, L., 1992, “An Empirical Analysis of Illegal Insider Trading,” Journal of Finance 47, 1661-1699.

Myers, S.C., 1976, Modern Developments in Financial Management, New York, Praeger.

Myers, S.C., 1977, “Determinants of Corporate Borrowing,” Journal of Financial Economics 5, 147-176.

Penman, S., 1998, “Combining Earnings and Book Value in Equity Valuation,” Contemporary AccountingResearch 15, 291-324.

Rhodes-Kropf, M., D.T. Robinson, and S. Viswanathan, 2005, “Valuation Waves and Merger Activity: TheEmpirical Evidence,” Journal of Financial Economics 77, 561-603.

Rhodes-Kropf, M. and S. Viswanathan, 2004, “Market Valuation and Merger Waves,” Journal of Finance59, 2685-2718.

Ritter, J.R., 1991, “The Long-Run Performance of Initial Public Offerings,” Journal of Finance 46, 3-27.

Roll, R., 1986, “The Hubris Hypothesis of Corporate Takeovers,” Journal of Business 59, 197-216.

Satterthwaite, F.E. 1946, “An Approximate Distribution of Estimates of Variance Components,” BiometricsBulletin 2, 110-114.

Seyhun, H.N., 1992, “Why Does Aggregate Insider Trading Predict Future Stock Returns?” QuarterlyJournal of Economics 107, 1303-1331.

Shiller, R.J., 2000, Irrational Exuberance, Princeton, NJ, Princeton University Press.

Shleifer, A., 2000, Inefficient Markets: An Introduction to Behavioral Finance, Oxford, UK, Oxford Uni-versity Press.

Shleifer, A. and R.W. Vishny, 2003, “Stock Market Driven Acquisitions,” Journal of Financial Economics70, 295-311.

Smit, H.T.J., 2001, “Acquisition Strategies as Option Games,” Journal of Applied Corporate Finance 14,79-89.

Smit, H.T.J. and T. Moraitis, 2010a, “Playing at Serial Acquisitions,” California Management Review 53,56-89.

Smit, H.T.J. and Moraitis, T. 2010b. “Serial Acquisitions Options,” Long Range Planning 43, 85-103.

Smith, W.S. and A.J. Triantis, 1995, “The Value of Options in Strategic Acquisitions,” in L. Trigeorgis, Ed.,Real Options in Capital Investment, Westport, CT, Greenwood Publishing Group, Praeger, 135-149.

Smit H.T.J. and S. van Bekkum, 2011, “Consistent Analysis of Valuation Ratios: An Application to FutureGrowth Opportunities,” Erasmus University Working Paper.

Thijssen, J., 2008, “Optimal and Strategic Timing of Mergers and Acquisitions Motivated by Synergies andRisk Diversification,” Journal of Economic Dynamics and Control 32, 1701-1720.


Tong, T.W., T.M. Alessandri, J. Reuer, and A. Chintakanda, 2008, “How Much Does Country Matter? AnAnalysis of Firms’ Growth Options,” Journal of International Business Studies 29, 387-405.

Tong, T.W. and J.J. Reuer, 2008, “Real Options in Multinational Corporations: Organizational Challengesand Risk Implications,” Journal of International Business Studies 38, 215-230.

Tong, T.W., J. Reuer, and M.W. Peng, 2008, “International Joint Ventures and the Value of Growth Options,”Academy of Management Journal 51, 1014-1029.

Trigeorgis, L. 1996, Real Options: Managerial Flexibility and Strategy in Resource Allocation, Cambridge,MA, MIT Press.

Documents

Buy Smart, Time Smart: Are Takeovers Driven by Growth