Signalling with Dividends, Stock Repurchases, and Equity Issues

Author(s): Paul Asquith and David W. Mullins, Jr.

Source: Financial Management, Vol. 15, No. 3 (Autumn, 1986), pp. 27-44
Published by: Blackwell Publishing on behalf of the Financial Management Association International
Stable URL: http://www.jstor.org/stable/3664842
Accessed: 01/11/2010 10:06
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=black.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Financial Management Association International and Blackwell Publishing are collaborating with JSTOR to
digitize, preserve and extend access to Financial Management.

http://www.jstor.org

Common

Equity

Signalling

Transactions

with

Repurchases,and

Dividends,

Equity

Stock
Issues

Paul Asquith and David W. Mullins, Jr.

Paul Asquithand David W. Mullins, Jr., are membersof thefaculty

of Harvard University.

* Decisions concerningequity cash flows - dividends, stock repurchases,and equity issues - have
long been the focus of controversyand confusion
amongacademicsandfinancialpractitioners.Thepurpose of this paperis to provideinsightinto the capital
markets'reactionto equity cash flow decisions. The
natureof the markets'reactionsuggestsa framework
for relatingandinterpretingthesedecisions.Although
the immediateobjectiveof this researchis to measure
and understandthe markets'reactionto managerial
decisions,the ultimateobjectiveis to improvecorporatefinancialdecisionmaking.

researchshedslighton why financialmanagerscontinue to ignorethe advice of professorsand consultants

andshowerinvestorswith cash. These studiesclarify
the efficacy of dividendsand repurchasesas vehicles
for communicatinginformationto shareholders.They
suggestthatdividendsand repurchasesare perceived
by investorsas signalsof management'sassessmentof
a company'sperformanceand prospects.

A. The Dividend Puzzle

of
Summingup our currentstate of understanding
dividend
Martin
Feldstein
and
corporate
policy,
Jerry
Green[11] conclude:"Thenearlyuniversalpolicy of
payingsubstantialdividendsis the primarypuzzle in
the economicsof corporatefinance."
The puzzle was first posed in a classic paperpublishedin 1961by MertonMillerandFrancoModigliani [18]. Their work demonstratedthat, with a given
investmentand financing policy, a firm's dividend
policy shouldnot affectthe value of its shares(ignoring all imperfections).Thelogic underlyingtheMiller-

I. The Controversy Surrounding

Dividends and Stock Repurchases
A well knownfinancialconsultantis fondof saying,
"Any distributionof cash by a companyrepresents
failure."By this definitionNew YorkStockExchange
companiesfailedin 1984to thetuneof over$68 billion
paid out in cash dividendsand about$77 billion distributedthroughrepurchasesof commonstock.Recent
27

28

Modiglianitheoremcan be seen in the behaviorof

stock prices on ex-dividenddates.
On the day a stockgoes ex-dividend,its pricedrops
by the amountof the dividend.Thisdropis permanent
in the sensethatthe priceshouldalwaysremainbelow
what it would have been were no dividendpaid.
Werethisdropon the ex-dividenddatenotto occur,
profitmotivatedspeculatorscould buy the stock the
day before, sell it on the ex-dividendday, andpocket
the dividend.Inspiredby this attractiveone-day return,the actionsof speculatorsshoulddrive a wedge
between the cum-dividendand ex-dividend stock
price,andthisprofitopportunitydisappearsonly when
the reductionin stockpriceequalsthe dividend.Since
the dividendreceived is exactly offset by the capital
loss on ex-dividendday, investorsshouldonly break
even on dividends.
Millerand Modigliani'swork raises the following
question:How can investorsbenefit from a dividend
when it is, in effect, paid dollarfor dollarout of the
valueof theirshares?The answermaylie in imperfectionspresentin ourworldandabsentfromthe worldof
economictheory.
The prevalenceof dividendsmay be explainedby
their capacity to convey managerialassessmentsto
investors,institutionalrestrictionsin the capitalmarkets and/or"irrational"
investorpreferences.On the
otherhand,thehighordinaryincometaxratesimposed
on dividendincomeandthe costs incurredby firmsin
financing dividends both seem to argue against
dividends.
In the 20 years since the publicationof the Miller
and Modiglianipaper, these competing hypotheses
havebeen subjectedto exhaustivetheoreticalandempiricalanalysis.The fruitsof thisresearcharesummarizedby FischerBlack [4]. Whatshouldinvestorsand
firmsdo aboutdividends?Black'sansweris "Wedon't
know." Despite the amusing befuddlementof their
academiccounterparts,corporateexecutivescontinue
to flood the world with cash dividendsas forcefully
and consistentlyas the tide poundsthe beach.
B. Are Dividends Hazardous
to Your Wealth?
The simplequestionposed in the above headingis
the focus of severalrecentstudies.While not solving
the dividendpuzzle, the studies have suppliedadditionalpieces whichmove us in the directionof a solution. Theyexaminethe impacton stockpricesof dividend announcements.The use of daily stock market
datapermitsexplicitidentificationandcontrolof oth-

FINANCIALMANAGEMENT/AUTUMN
1986

er, roughly contemporaneousinformation.The announcementeffectsof earningsreportsandthelikecan

be isolatedandeliminated,allowinga clearerview of
the separateimpactof dividendannouncements.This
methodologyavertsa seriousflaw inherentin earlier
attemptsto answerthe questionposed in the heading.
C. The Birth of a Dividend Policy
We completed a study examiningthe impact on
stock prices of establishinga policy of paying cash
dividends[3]. Duringthe period 1964-1980, all the
firms in our sample either paid the first dividendin
theircorporatehistoriesorresumeddividendpayments
aftera hiatusof at leastten years.An advantageof this
approachis thatinvestorspresumablyhave littleprior
expectationof the initiationof dividend payments.
Theyhaveno recentdividendhistoryon whichto base
accurateforecastsof futuredividendpolicy. In studies
of ongoingdividendpolicies, the impactof dividends
may be obscured by the effects of investors'
expectations.
Further,investorsin our samplefirmshave for the
precedingten yearsreceivedreturnssolely in the form
of capitalgains. Since they arelikely high tax bracket
investors,ourresultsshouldbe biasedagainsta favorable dividendeffect. Dividendinitiationmay be expectedto inducea changein clienteleswith attendant
costs. A positiveimpactis possibleonly if the benefits
of dividendsoutweighthe tax burdenand othercosts
unexpectedlyimposedby them.
Therefore,focusingon the unexpectedinitiationof
dividendspurgesour sampleof investorexpectations
in subsequentdividends.This shouldalincorporated
low us to see the net impactof the severalhypotheses
concerningdividends.
For80 of the 168firmsin oursample,othernews in
additionto the dividendannouncementwas published
in The Wall Street Journal during the period immedi-

date. In
the dividendannouncement
atelysurrounding
theresultsreportedsubsequentlythesefirmshavebeen
deletedto eliminatethe effectsof othercontemporaneous information.For the remaining88 firmsno other
informationwas madepublicin the most widely read
financialnews source.This insuresas far as possible
thatany impactwe observecan be attributedsolely to
the dividendannouncement.'
To examinethe stockpriceimpactof initiatingdividendswe first calculatedabnormalreturns(ARs)the stockreturnsexcludingthe movementof the mar'An analysis of the complete sample is presentedin [3].

29

WITH EQUITYTRANSACTIONS
ASQUITHAND MULLINS/SIGNALLING

Exhibit 1. Abnormal Stock Returns for Initial Dividend Announcements

Dividend
Announcement
Day
I

Cumulative 7%/ Average

Abnormal 6% Return
5%

CAR,InitialDividendAnnouncements

4%
3%.
2%

0/o

-1%
1

IX

-12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 0' 1 2 3 4 5 6 7

TradingDay Relativeto DividendAnnouncementDate

10 11 12

*Day 0 is the publicationdate in The WallStreetJournal. Since The WallStreetJournal is a morningnewspaper,the informationin articles is often
made public before the end of tradingon the day before publication.For this reasonthe abnormalreturnsfor days - 1 and 0 have been aggregatedto
arriveat the announcementday returnreportedabove.

ket in general and adjusted for the sensitivity of the

firm's stock price relative to market movements (i.e.,
adjusted for its beta). The ARs are then cumulated
beginning a number of days prior to the dividend announcement and continuing until well after the news
date. The resulting cumulative abnormal returns
(CARs) provide a clear view of the price behavior of
our sample firms during the announcement period.
The results reported in Exhibit 1 demonstrate that
for our sample initiating the payment of cash dividends
generated a positive abnormal return to shareholders
averaging almost 5% on announcement day. This return is in addition to the return accruing to owners of
other stocks of the same risk level and is not the result
of other contemporaneous announcements.
The dividend effect is concentrated primarily on the
announcement date. The CAR picks up about 1% in
the 12 days following the announcement. Although the
CAR methodology does not allow unequivocal assessment of the permanence of the effect, the increase in
shareholders' wealth is at least retained many days
after the dividend is announced. For example, 90 trad-

ing days after the announcement, the CAR is a little

more than 6%.
An average abnormal return of only 5% may seem
uninspiring. It is, however, highly significant statistically and much larger than the average returnof roughly 1% reported in previous studies.2 The results are
also reasonably consistent across firms. On the upside,
72% of our 88 sample firms experienced a positive
impact, led by six firms with abnormal returns in excess of 20%. Of the 28% which suffered a reduction in
stock price, only six firms had returnsless than -4%.
To understandthe fate of these unfortunate firms will
require detailed field research.3
We also found that the magnitude of the benefit
2Otherstudies which also report daily abnormalreturnsfor dividend
announcementsinclude Aharonyand Swary [1] and Charest[5]. These
papers only look at changes in the level of already established dividends, however, while our result is for first time dividends.
3One hypothesis is that for these firms investors were anticipatingthe
initiationof dividendsandwere disappointedby the amountof the initial
dividend. However, thereare many other hypothesesthatcould explain
the results for this small sub-sample.

30

FINANCIALMANAGEMENT/AUTUMN1986

Exhibit 2. The Estimated Relationship Between Initial Dividend Announcement Day Returnand Initial Dividend
Yield
Announcement
Day Average
Abnormal Return

10%
9%,
8%

slope =

1.45

5%.
Average Initial Return .
4%,
3%,
2%,

3.02%
Average
Initial
Yield

4%

5%

6%

Initial Dividend Yield

accruing to shareholders is directly proportionalto the

size of the dividend - measured by either the initial
dividend yield or the payout ratio. If investors perceive
dividends as positive signals by management, the abnormal return should be related to the signal size.
This relationship can be estimated statistically by
linear regression. In Exhibit 2 we have depicted the
estimated relationship between the announcement
day's abnormal return and the initial dividend yield.
Announcing an initial dividend with a yield of 1%
generates an abnormal capital gain of almost 2%.
Consistent with the Miller-Modigliani analysis, investors may only break even on ex-dividend day. But
the announcement day capital gain insures that they
come out ahead overall. Indeed, the dividend effect
appears large enough to offset any investor tax differential.4 We should note that the specific results reflect
only the average experience of our sample of firms.

D. Subsequent
Dividends
We repeated the heretofore described procedure for
a sample of subsequent dividend increases within three
years of the initial dividend. Focusing on the largest
increases, we found 66 firms in our sample with no
other announcements roughly concurrent with the announcement of the largest subsequent dividend
increase.
Comparedwith initial dividends, the results presented in Exhibit 3 for subsequent increases look much less
convincing. Nonetheless, the 1% average abnormal

4For example, suppose an investor is subject to a 100% tax on the

dividend and the stock price falls by the full amount of the (pre-tax)
dividend on ex-dividend day. The regressions demonstratethat the
abnormalreturnon announcementday is at least as largeas the dividend
yield. Thus, the capitalgain on announcementday is at least as large as

the capitalloss on ex-dividendday. An investor would not lose even if

the entiredividendis taxed away. Furthermore,the yields employed in
the regressionsare annualizedalthoughmanyof the dividendpayments
were actuallyquarterly.This suggests that dividends may producenet
gains even for high tax bracket investors.

Nonetheless, the relationships they illustrate may be

valid for the larger population of firms.
Thus, for our sample, initiating dividends increased
shareholders' wealth. Any negative tax burden and/or
financing cost associated with dividends is more than
offset by the benefits of dividends.

31

WITH EQUITYTRANSACTIONS
ASQUITHAND MULLINS/SIGNALLING

Exhibit 3. Abnormal Stock Returns for Largest Subsequent Dividend Increase

Cumulative 7% Average
Abnormal 6%Return

Dividend
Announcement
Day

5%-*

I
I

4%

I
I

CAR,SubsequentDividendIncreases

3%
2%1%

-2%

-12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 O' 1 2 3 4 5 6 7

TradingDay Relativeto DividendAnnouncementDate

10 11 12

*Day 0 is the publicationdate in The WallStreet Journal. Since The Wall StreetJournal is a morningnewspaper,the informationin articles is often
made public before the end of tradingon the day before publication.For this reasonthe abnormalreturnsfor days - 1 and 0 have been aggregatedto
arrive at the announcementday returnreportedabove.

returnon announcement day is statistically significant.

In contrast we cannot reject the hypothesis that the
other daily perturbations are merely random
fluctuations.
The impact of subsequent dividend increases appears to be far smaller than the initiation impact. However, our analysis suggests that this may not be the
case. The true impact of a subsequent increase cannot
be assessed simply by observing average returnson the
announcement day.
To evaluate this true impact we need to explore two
factors: (i) the relationship between the abnormal return and the size of the dividend and (ii) investors'
anticipationof subsequent dividend changes. This latter concern may not be necessary in examining initial
dividends. The initiation of dividend payments should
be largely unexpected by investors. However, once a
dividend policy is established, investors may be more
successful in forecasting future dividends, and these
forecasts may already be incorporated in stock prices
on the dividend announcement date.
Both these factors can be seen in the relationship
between announcement day return and the size of the
dividend increase - measured by the increase in dividend yield. This relationship is plotted in Exhibit 4.

Comparedwith the relationship for initiation, the subsequent relationship is much more steeply sloped. Another difference is that abnormal returns are negative
for small changes. Returns are positive over the full
range of yields for initial dividends.
One reason average subsequent returns are smaller
than initials is readily apparent from Exhibit 4. The
largest subsequent increases in dividend yields are
simply much smaller than the initial yield. Were they
as large, the subsequent relationship predicts that they
would produce returns at least as large as initials. For
example, plugging the average initial yield of 3% into
the subsequentrelationship would predict a subsequent
return of almost 7%.
The second factor explaining the observed smaller
impact of subsequent dividend changes concerns the
negative portion of relationship. We interpret this as
confirming that investors forecast subsequent dividends. For this sample their forecast, on average, is a
0.67% change in yield - the yield at which the abnormal returnis zero.5 If the firm announces an increase in
50f course, investorsdo not always expect an increase. The subsequent
increases reflected in Exhibits 3 and 4 were the largest subsequent
increases within three years of the initial dividend. Investorspartially
predictedthese increases.

32

FINANCIALMANAGEMENT/AUTUMN1986

Exhibit 4. The Estimated Relationship Between Announcement Day Return and Increase in Dividend Yield for

Both InitialDividendsand LargestSubsequentDividendIncreases

Announcement

Day Average
Abnormal Return

10% -

d
/
\C

9%8%,/

/. 7%
o?e/
6%- -\;,,
5% ------Average Initial Return 4.75%

= 2.94

slope

slope = 1.45

</
-

4% 3% 2%-

Average
Subsequent

Return 1.17%___
1%'

.67%/

Return From
Anticipated (
Increase

1.07%

2%

/AntiAverage
c
Subsequent
/ipated
Subsequent Increase
Increase

yield of 0.67%, the stockpriceis not affected.In this

case the increaseis fully anticipatedand alreadyreflected in marketprices.
If the announcedincrease is greaterthan 0.67%,
investorsare pleasantlysurprised,resultingin a positive return.If the increaseis less thaninvestorsanticipate, they expresstheirdisappointmentthrougha reductionin the stockprice. A positiveabnormalreturn
on subsequentdividendannouncementday is visible
only if the dividend is greater than forecast by
investors.
It is importantnotto confusethisunexpectedimpact
of subsequentdividends(portrayedby the regression
line in Exhibit4) with their total impact. The total
impactincludesboth the unanticipatedreturnand the
anticipatedreturn.Note thatif the firmfails to increase
its dividendat all (a zero changein yield), it suffersa
negativereturnof about 2%. It avoids this loss by
announcingthe anticipatedchangeof 0.67%. Thus,on
announcement
day a returnof 2% is alreadyincorpo-

3.02%

Average
Average

4%

5%

6%

Increase in Dividend Yield

Yield

ratedin stock pricesreflectinginvestors'anticipation

of a 0.67% increasein yield.
For example, we would estimatethata subsequent
increasein yieldof 1%wouldresultin a totalabnormal
capitalgain of about3% - the anticipatedreturnof
2%alreadyreflectedin stockpricesplus the unanticipatedannouncement
day returnof 1%estimatedfrom
the relationshipdepictedin Exhibit4. Only the latter
componentis easily observable,andthis has led other
researchersto underestimatethe true impactof dividends.Again, the specificsof theseresultsapplyonly
to the firmsin our sample.However,they illustratea
relationshipthatmay exist for the widerpopulationof
firms.

Therefore,we find that subsequentdividendsalso

benefitshareholders.Comparedto an initialdividend,
a later increaseof the same magnitudeproducesan
abnormalreturnthat is at least as large and probably
largerthanthe initial return.Our study also demonstratesthata firm's dividendpolicy leads investorsto

WITH EQUITYTRANSACTIONS
ASQUITHAND MULLINS/SIGNALLING

anticipatefuturedividends. As a result some of the

benefitsof a subsequentdividendincreasearealready
builtintothe stockpriceon the announcement
day, and
the announcementday effect understatesthe total
benefitof the dividend.Onthe announcement
day, the
firm must fulfill investors' expectationsor suffer a
reductionin stock prices.

II. What's Puzzling about Repurchases?

If dividendssufferfromthe burdenof highordinary
income tax rates, stock repurchasesshould provide
relief. WithincertainIRS constraints,these distributions to shareholdersaretaxed at morefavorablecapital gains rates. Nonetheless,academicsandfinancial
consultantsare only marginallymore charitabletowardsrepurchasesthan dividends.
Thereasonis simple.Virtuallyanycashdistribution
generatestax liabilitiesfor some investors.Further,if
the fundsareneededsome time in the future,transactionscosts (flotationfees andthe like) mustbe paidto
"reclaim"the cash from the capitalmarkets.A final
cost to repurchasesis the premiumusually paid to
investorstenderingtheir shares.
Thesedisadvantagescan be avoidedsimplyby neverpayingout anythingto shareholders.Firmscanconstructbalancedportfoliosof businesses. Cash cows
can acquirecompanieswith large, profitableinvestmentopportunities.Fundscan be recycledinternally,
bypassingthe externalcapitalmarketsand the attendantfinancingcosts and investortaxes.
Why, then, do some of ourlargest,most successful
firmsrepurchasetheirstock?Academicstendto favor
threeexplanations.First,the investortax argumentarenotperfect,butareanimprovementon
repurchases
cashdividends.The secondis the leveragehypothesis.
Throughrepurchasesof common stock, a firm can
radicallyalter its capital structure,increase its debt
ratio,andreapthe benefitsof higherleverage.Third,
some have suggestedthatrepurchasesare merelythe
productof recommendations
by a vestedinterestgroup
investment bankers.

Many managershave a simpler answer. Why do

theybuybacktheirstock?Becauseit is underpriced,of
course.If this is theirmotive, repurchasesshouldconvey valuableinformationto investors.A repurchaseis
a signal that managers, who possess an insider's
knowledgeof the firm,areconvincedthattheirstockis
worthmore than its currentprice. In addition,their
conviction is strong enough to lead them to pay a
premiumfor the stock despite the risk of dilutionif
they are wrong.

33

A. Research on Repurchases
Theo Vermaelenin 1981 publisheda study [22] of
He employedthe CAR methodologyin
repurchases.6
a
analyzing sampleof tenderoffers and open-market
repurchasesexecutedduringthe period 1962-1977.

B. Tender Offers
In a tenderoffera firmoffersto purchaseits stockat
a specified price, usually a premiumto the market
price. The offer remainsin effect for a specifiedtime
period,typicallythreeor four weeks. In Vermaelen's
sampleof 131 tenderoffers the averagepremiumwas
23%and, on average, 15%of the firm's outstanding
shareswere repurchased.
Repurchasesby tenderoffer increasestock prices.
The detailsare portrayedin Exhibit5. The abnormal
return during the announcementperiod averages
roughly17%.As theoffersexpireduringthefollowing
few weeks, these stocks sell off by only about4%.
Thereis no additionaldecline in CAR for at least a
year. Thus, Vermaelenconcludesthat the remaining
gain of 13%is permanent.
In his analysis of the underlyingcauses of these
gains, Vermaelenrejectsthe tax advantageof repurchasesover dividends.He attributesthe stock market
reactionto an informationeffect, ratherthanthe leverage hypothesis.
His analysis arguesthat the repurchaseconvinces
investorsthat the stock was undervaluedpriorto the
tenderoffer. Themagnitudeof thebenefitsto investors
is positivelyrelatedto the premiumpaid, the percentageof outstandingsharesrepurchased,andthe fraction
of the firm'ssharesownedby insiders.This is consistentwith a signallingexplanation.These threefactors
shouldbe positivelyrelatedto the market'sperception
of the strengthof managers' conviction that their
sharesare underpriced.
Managers'faithin the futureprospectsof theircompanies is confirmedby subsequentearningsperformance.Samplefirmsexhibitedabnormallyhigh earnings duringthe five years followingthe tenderoffer.

C. Open-Market Repurchases
Firmssometimesrepurchaserelativelysmallquantities of stock in the open market.The purchasesare
executedthroughbrokersat normalcommissionrates
andno premiumis paid. Vermaelen'sstudyexamines
243 open-marketannouncements.The resultsof his
6A varietyof otherstudies, includingDann [8] and Masulis [16], verify
Vermaelen's findings.

34

FINANCIALMANAGEMENT/AUTUMN1986

Exhibit 5. AbnormalStock Returnsfor Stock RepurchaseAnnouncements

z

0.20 -

FH

cr
IJ

(^'

-"

DTENDER OFFERS

0.15
0.10 -

z
w

O
cr
w

0.05 '
0

--t-

_I

-__ ""-"^-v
_^
"-'"'"'w

v^OPEN

-0.05-

MAR---KET
PURCHASES
OPEN MARKET PURCHASES

X_.Jl
F-

0i

I~~~~~~~~~
1
I~~ I~~~~~~~~~~
I~
1
1 ~~II1I
I
-60
_

I
-50

I
-40

I
-30

I
-20

I
-10

I
I
0

10

20

30

40

50

60

DAYS RELATIVETO ANNOUNCEMENT

Source: "CommonStock Repurchasesand MarketSignalling"by Theo Vermaelen, Journal of Financial Economics, Volume 9 (1981), p. 149.

analysisare also reportedin Exhibit5.

The firms in this sample have been experiencing
negativeabnormalstockpriceperformancepriorto the
open-marketrepurchase.In the threemonthspreceding the repurchase,their stock prices have underperformedthe marketby about7%. The repurchaseproducesa gainof a littlemorethan3%, butpricesretreat
about1%duringthe following threemonths.The result is an apparentlypermanentgain of 2%.
As one mightexpect, open-marketrepurchasesare
less powerful than tender offers. For Vermaelen's
the market,
sample,which has been underperforming
theserepurchasesaresuccessfulin haltingthe slideand
of theresults
producinga smallgain. His interpretation
is consistentwith his views on tenderoffers.
Both types of repurchasesbenefit shareholders.If
managersbelievetheirsharesareunderpriced,a repurchase communicatestheirconvictionto shareholders.
Thiscommunication
is issuedby knowledgeableinsiders. It is backedby cash or securitiesand, for tender
offers, the willingness to pay a premiumabove the
currentprice.
This rationaleis supportedby KarlF. Slacik, chief
financialofficer for Levi Strauss& Company,a firm
thattenderedfor 15%of its own shares.ExplainsMr.

Slacik:"Thereis no greaterexpressionof confidence

thanto repurchaseyourown shares.It lookedto us like
the best investmentwe could makeat this time and it
shouldspeakaboutourmanagement'sconfidence.We
wouldn'ttake$150 millionof ourresourcesanduse it
this way if we were concernedaboutthe futureof our
business."7 After most repurchases, the stock price

does not fall backto its pre-announcement

level. Apparently,repurchasesare successfulin convincinginvestorsof the validityof managers'assessments,and
the gain to shareholdersis permanent.

IiI. Signalling with Dividends

and Repurchases
Distributionsto shareholdersare received as good
news. Repurchasesconvincethe stockmarketto price
a firm's stock substantiallyhigherthanthe pre-tender
price. Establishingand pursuinga policy of paying
cash dividendsappearsto producesmall, thoughconsistent, gains. Any investortax burdenor financing
costsareoutweighedby thesebenefits.As faras canbe
determinedwith the CAR methodology, the gains
fromboth dividendsand repurchasesare permanent.
7As quoted in The New YorkTimes, see [23].

WITH EQUITYTRANSACTIONS
ASQUITHAND MULLINS/SIGNALLING

Theseresultsareconsistentwiththe view thatdistributions are signals.

Thebasicideabehinddividendsignallingis simple.
Payingdividendsprovesthata firmis able to generate
cash, ratherthanjust accountingnumbers.By demonstratingits profitabilityin this manner,the firm can
differentiateitself from less profitablefirms. Further,
managementthen has an incentive to performwell
enoughto maintainits dividendandavoidthe adverse
consequencesof a dividendcut or any equityissue to
replacethe fundspaidout. Viewed in this framework,
dividendchangesrevealmanagement'sassessmentof
a firm's profitability.The researchreviewedheretofore suggests that investorsperceivethese signals as
credible,persuasive,and valuable.

35

also help to keep managementhonest. Vermaelen's

findingof abnormallyhigh earningsperformancefollowingrepurchasessuggeststhathis samplefirmsdid
not lie.
False signallingmay misleadthe marketfor a short
time. Thoughmarketvengeancemay not be immediate, it shouldbe unavoidable.Thiswouldalso compromise the credibilityof any futureattemptsat signalling. As a result, managementwould lose tools
valuablein creatingvalue for shareholders.

A. Can Signals Be Trusted?

The precedingevidence does not demonstratethat
managerscan manipulatestock pricesby consistently
misleadinginvestors.For both types of distributions,
false signallingis punished.
Ouranalysisof subsequentdividendsillustratesthat
dividendsare habit-forming.If the marketdoes not
receiveits expecteddosage, the stockpricewill suffer
withdrawalsymptoms.
Establishmentof a dividendprogramgeneratesexpectationsof futuredividends.Managementis forced
to submitto investors'anticipationof a periodicsignal.
Whenevermanagementis unableor unwillingto fulfill
these expectations,the stock pricewill fall. Although
not examinedin this article,our studyandothersfind
substantialreductionswhen dividendsare cut.8These
reductionsare generallygreaterthan the gains from
initiatingandincreasingdividends.Moreover,if managementpaysoutexcessive dividends,it couldreplace
the fundswith a new equityissue. As we shallexplore
laterin this paper,equity issues reducestock prices,
andthis negativereactionencouragesmanagementto
limit dividend payout to a sustainablelevel. Other
costs to false signalling with dividends include the
possible adverse effects of altering investmentand
capital structurepolicy in an attemptto sustain an
excessive dividendpayout.
Similarretributionshouldbe sufferedin responseto
false signalling with repurchases. As subsequent
eventsinevitablyrevealthe truth,the stockshouldfall
below its pre-announcement
price, reflectingthe premiumgiven awayin the tenderoffer. The costs associated with replacingthe funds paid out in repurchases

B. Wouldn't It Be Cheaper
to Send a Postcard?
Thatdividendsarenews is not news. We have long
knownthat managerspossess valuable, inside informationand that dividendsconvey informationto the
market.Criticsof signallingthroughdividendsraisea
simplequestion.In view of the tax burdenand other
costs associatedwith dividends, aren'tthere equally
effective, less costly ways to convey information?
Shouldn'ta candid letter to shareholdersserve the
same purposemore cheaply? Other alternativesincludefinancialandaccountingstatements,managerial
forecasts,and other statementsby management.
Recentevidencedoes not supportthis view. It demonstratesthat dividendannouncementsconvey informationover and above that containedin alternative
announcements.For example, when quarterlyearnings and dividendsare announcedon differentdays,
the two announcementsproduceseparate,significant
marketreactions.Regardlessof whetherthe earnings
announcement
precedesor follows it, the dividendimpact is roughlythe same. The converseis also true.9
There are reasonsfor the efficacy of dividendsas
arebackedby hard,
signals.Dividendannouncements
cold cash. The firmmustgeneratethis cash internally
or convincethe capitalmarketsto supplyit. Alternative communicationsmay lack the credibility that
comes from"sayingit with cash."Investorsmay suspectthatstatementsby managementarebackedby the
ghostwritingof well paid publicrelationsspecialists.
They may feel that financial statementshave been
skillfullymassagedby the financialstaff. In addition,
dividenddecisions tend to be futureorientedas opposed to accountingstatementswhich documentpast
performance.
Besides credibility,dividendsalso have the advantages of simplicity and visibility. Many other announcementsare, at the same time, complex and de-

8In particular,see Charest [5].

9See Aharony and Swary [1].

36

tailed in focus. They require time and expertise to

decipher. In contrast, few investors fail to notice and
understanda check in the mail. An empty mailbox is
also easily interpreted. As simple numerical signals,
dividends facilitate comparative analyses unlike statements by management which may be difficult to calibrate. Simplicity is especially advantageous for investors holding many firms' shares to achieve the benefits
of diversification. Further, dividend signals convey
information without releasing sensitive details that
may be useful to competitors. A firm can reap some of
the benefits of disseminating sensitive information
safely coded as dividend signals, but avoid the competitive damage resulting from the release of detailed
forecasts.
Thus, dividends have a number of unique advantages over alternative managerial communications.
The empirical evidence confirms that alternatives are
not perfect substitutes for dividends. Dividends serve
as a simple, comprehensive signal of management's
interpretationof the firm's recent performance and its
future prospects.
C. But, Shouldn't Repurchases be Better?
They share equally with dividends the advantages of
credibility, visibility, and simplicity. Open-marketrepurchases avoid the cost of the premium usually paid
in tender offers. In addition, they have one substantial
advantage over dividends. Distributions through repurchasesmay be subject only to capital gains tax rates
which are lower than the ordinary income rates applicable to dividend income.
For a repurchaseprogramof the same frequency and
magnitude as dividend payments, this advantage is,
unfortunately, illusory. Under applicable U.S. tax
code, repurchases qualify for capital gains treatment
only if the distributionis "essentially not equivalent"to
paying a dividend. If many firms institutedlarge, quarterly or even annual programs, the IRS would likely
construe these repurchasedistributionsto be dividends
for tax purposes. If it did not, the U.S. Treasurywould
suffer an enormous revenue loss from no longer collecting taxes on cash payments by U.S. firms.
If perfect substitutionof repurchasesfor dividends is
not feasible (in the aggregate at least), what about
relatively large, infrequent programs? Some firms
have implemented such a policy, and it appearsto be a
viable alternative to quarterly dividends. However,
lack of frequency does entail a disadvantage. Shareholders benefit from the frequent communication of
valuable information, and to some extent, a reduction

FINANCIALMANAGEMENT/AUTUMN1986

in frequency reduces the benefits.

Repurchases appear more suited for episodic signalling with timing at the discretion of management.
Vermaelen's results suggest that they may be appropriate whenever a firm's management is convinced that
its stock is undervalued. To be most effective, management must be willing to back its conviction by
paying a premium for a significant percentage of its
shares.
Dividends appear to be the appropriatevehicle for
regular, relatively frequentcommunication of management's ongoing assessment of a firm's prospects. The
tax burden imposed on both types of distributions are
simply costs associated with these signalling mechanisms. The research reviewed in this article provides
evidence that the benefits of employing these mechanisms outweigh these costs.
- Still
IV. Dividends
and Repurchases
after All these Years
Puzzling
The findings reviewed in this paper appear to be
consistent with the way managers and investors have
traditionally viewed dividends and repurchases. This
suggests that managers and investors are not as unintelligent as some academics think. While we would hope
the converse is also true, the evidence is less persuasive on this point.
We still have no precise answer as to how managers
should determine distributions to shareholders
no
definitive guidance on optimal dividend and repurchase policy. We now know more about how such
distributionsaffect stock prices. However, this is only
one input to managers' decisions.
We have not dealt with other importantinputs. Perhaps chief among these is the financing side of dividends and repurchases. Further, the effects we have
described may not apply to all firms. For example,
relatively closely held firms may not need dividends to
communicate with shareholders. The structureof the
shareholderconstituency is still an importantconcern.
Even ignoring other decision inputs, progress on the
information content of dividends has yet to yield specific prescriptions for managers. There is some evidence that suggests dividend signalling may be more
effective for lower risk firms.'0 However, we do not
know precisely how dividends should be set to maximize the value of information communicated.
'Eades[10] finds thatdividendyield and a firm's stock returnvariance
are negatively correlatedand that ceteris paribus, largerinformationis
gained throughdividend changes for low-risk firms.

ASQUITHAND MULLINS/SIGNALLING
WITH EQUITYTRANSACTIONS

37

New theoryand evidence on signallingdoes seem

roughlyconsistentwith time honored(if hidebound)
heuristics:(i) set dividendsto reflect management's
estimateof the lower bound of future intermediateterm earnings, (ii) do not increasedividendsunless
confidentthatthe higherlevel can be maintained,(iii)
reducedividendsonly if absolutelynecessary.A dividendpolicythatconsistentlyreflectsforecastsof intermediate-term,sustainableearnings should produce
signalswhich are simple, credible, and valuable.
Thus,ourreviewof recentresearchdoes notprovide
a comprehensiveanalysis of how managersshould
makedividendandrepurchasedecisions.Nonetheless,
these findings have important implications for
managers.
They suggest that dividends and repurchases,
though similar, play somewhat different signalling
roles. Dividendsappearto be perceivedby the stock
marketas regularlyschedulednews releasesconveying management'songoing assessment of a firm's
prospects.A repurchaseis viewed as an "extra"- a
newsbulletinjustifiedwhenmanagementis convinced
its stock is substantiallyundervalued.The market's
reactiondemonstratesthat in both cases, the news is
credible and, apparently,investors are persuaded.
Moreover,bothtypes of distributionscreatean incentive for managementto fulfill the promiseimplicitin
the signal, andthe integrityof bothtypes of signalsis
policedby substantialcosts associatedwith false signalling. Both dividends and repurchasesare useful
tools that managerscan employ to create value for
shareholders.

kets or can be constructedwith combinationsof existing securities. The availability of many close
substitutesimpliesthatthe demandcurvefor a firm's
sharesis essentiallyhorizontal.A stockpricereduction
shouldnotbe requiredto induceinvestorsto absorban
increasedsupplyof shares,andfirmsshouldbe ableto
issuelargeamountsof equityatthecurrentstockprice.
Neithershoulddilutionof currentearningspershare
reducestock prices when firms issue equity. In efficientcapitalmarketsinvestorsshouldsee throughcurrentearningsdilutionand price a firm's sharesbased
uponexpectedfuturecashflows. As long as a firmcan
earna competitivereturnon the fundsraised,anequity
issue should be a fair deal. The value of the equity
issuedshouldbe exactlyequalto the valuecreatedby
thefirm'sinvestmentof the proceedsleavingthe stock
priceunchanged.Theactionsof profit-motivated
speculators should insure that currentearnings dilution
shouldnot producea price reductionassociatedwith
new equity issues.
Thus, with large, efficient capital marketsfirms
shouldbe able to issue largequantitiesof stock at the
currentprice to finance worthwhileinvestmentprojects. Firms' self-imposedconstraintto finance their
operationsprimarilywith internallygeneratedequity
fundsis consideredan anomalyin academicfinance.
Thisbehavioris less anomalousto financialpractitioners. Financialexecutives, investmentbankers, and
otherpractitionerscontendthatequityissues resultin
depressedstock prices that correspondinglydepress
financialexecutiveswho are consideringexternalequity financing.

V. A Final Puzzle - Equity Issues

Anotherenduringanomalyin financialeconomicsis
the relianceof firms on internallygeneratedfunds as
theirchief sourceof equityfinancingand theircorrespondingreluctanceto issue commonstock." According to academicfinance,firmsshouldnot be reluctant
to issue equity in the large, efficient U.S. capital
markets.
Efficiency means that investors are pricing the
firm'sstockcorrectlybasedupontheriskandexpected
returnassociatedwithits futurecashflows. Onefirm's
shares representonly a small fraction of all assets
availableto investors.Close substitutesfor any firm's
shares,securitieswith similarrisk and returncharacteristics,eitheraredirectlyavailablein thecapitalmar-

A. Equity Issues and Stock Prices

In a recentstudy, we analyzedthe impactof equity
issues on stock prices.12Includedin our samplewere
128 offeringsof seasonedequity by industrialfirms
during the period 1963-1981. As in our dividend
study, we calculatedabnormalreturns(ARs) and cumulatedthe ARs to producecumulativeaveragereturns (CARs) centeredon the announcementof the
stockissue. Theresultis a view of thepricebehaviorof
our samplefirmsaroundthe time of the equityissue.
The resultsreportedin Exhibit6 demonstratethat
foroursample,equityissuesreducedstockprices.The
reductionaveraged3% on the day the offering was
announced.13
Although the stock price reductionis

"For empiricalevidence on firms' financing practices see Donaldson

[9], Lintner [14], and Sametz [21].

'2Fora detailed descriptionof the study, see Asquith and Mullins [2].
'3Masulisand Korwar[17] reporta similarresult on a differentsample
of firms.

38

FINANCIALMANAGEMENT/AUTUMN1986

Exhibit 6. Abnormal Stock Returns for Seasoned Equity Issues

EquityIssue
Annournceent
Day

Cumulative 4%Average
Abnormal 30,oReturn
(CAR)
20o -

I
I

00o0

I
I

es
CAR,EquityIssu4

1/%-

- - - - - - - -

__

-1 0%-

-3 o-40ooi

-10 -9

-8

-7

I I

I1

5 6
3 4
-6 -5 -4 -3 -2 0' 1 2
Announcement
to
Issue
Relative
Day
Equity
TradingDay

10

*Day 0 is the publicationdate in The WallStreetJournal. Since The WallStreetJournal is a morningnewspaper,the informationin articles is often
madepublic before the end of tradingon the day before publication.For this reasonthe abnormalreturnsfor days - 1 and 0 have been aggregatedto
arrive at the announcementday returnreportedabove.

concentratedprimarily on the announcement date, the

CAR falls by about 1% in the five days preceding the
announcement. Moreover, sample stock prices fall by
about 1% surrounding the issue date - the date the
offering is sold. While the CAR methodology does not
allow unequivocal determinationof the permanence of
the price decline, the reduction persists for many trading days after the announcement. For example, one
month after the announcement, the CAR is -3.5%.
This should provide ample time for profit motivated
tradersto capitalize on and thereby correct any unjustified price reductions.
An average abnormal return of only 3% may not
seem to be reason for concern. Nonetheless, it is highly significant statistically. The result is also pervasive
among sample offerings. Over 80% of our sample experienced price reductions associated with the equity
issue announcement. Moreover, most equity issues
represent a relatively small fraction of the shares then
outstanding. Regression analyses of our results confirm that the size of the price reduction is directly
proportional to the size of the equity offering. Addi-

tional insight into the magnitude of the price reduction

can be gained by relating the reduction in the aggregate
marketvalue of the equity to the aggregate proceeds of
the equity issue.
Exhibit 7 presents the distributionof the reduction in
aggregate market value as a percentage of the funds
raised in the equity issue. Although the percentage
reduction in stock price may appear small, the aggregate loss in shareholders' value is a large fraction of the
proceeds of the offerings. On average, when an equity
issue is announced, the loss in market value is 31% of
the funds raised. This means that, to raise $100 million
in new equity, existing shareholders in our sample
gave up an average of $31 million in current market
value. Almost 25% of the sample offerings produced
reductions in market value greater than 50% of the
proceeds of the issue, and for 6% of the offerings,
shareholderslost more on announcement day than the
total proceeds of the equity offering.
A graphic example of the latter result occurred on
February28, 1983. American Telephone & Telegraph
announced its intention to raise $1 billion in new com-

WITH EQUITYTRANSACTIONS
ASQUITHAND MULLINS/SIGNALLING

39

Exhibit 7. Distribution of the Reduction in Aggregate Market Value as a Percentage of the Proceeds of an Equity Issue
Offerings within this Range:
Loss in MarketValue
as a Percentageof the
Cumulative
Proceeds of the
Percentage Percentage
Number of sample of Offerings
Equity Issue

Negative MarketReactionto EquityIssue Greater than 200%

200%
120%
100%
80%
70%
60%
50%
40%
30%
20%
10%

Positive MarketReaction to Equity Issue

to
to
to
to
to
to
to
to
to
to
to

120%
100%
80%
70%
60%
50%
40%
30%
20%
10%
0%

0% to - 10%
-10% to -20%
-20% to -30%
-30% to -60%
less than -60%

1
2
4
6
4
6

0.8%
1.7%
3.3%
5.0%
3.3%

5
6
25

4.1%

11
9
20

9
7
3
2
1
121

5.0%
5.0%
20.7%
9.1%
7.4%
16.5%

7.4%
5.8%
2.4%
1.7%
0.8%

0.8%

2.5%
5.8%
10.8%
14.1%
19.1%
23.2%
28.2%
48.9%

58.0%
65.4%
81.9%

89.3%
95.1%
97.5%
99.2%
100.0%

100%

Average loss in marketvalue as a percentageof the proceeds of the equity issue = 31%.

mon equity. The market greeted AT&T's announcement with a reduction in its aggregate market value of
over $2 billion.14
The results reported in Exhibit 7 can be interpreted
in a number of ways - none of which are very comforting to executives who are considering issuing new
equity. For example, one may view the price reduction
as dilution of existing shareholders' value.
Consider an aggregate reduction in market value of
100% of the proceeds of the issue. After the issue is
executed, the firm's aggregate marketvalue will be the
same as before the issue was announced. However,
this post-issue market value will be divided by the
largernumberof shares outstanding leading to a reduction in stock price. The result is a purely dilutive equity
issue - the price reduction is exactly proportional to
the increase in shares outstanding. This same result
would occur if the market value of the firm remained
unchanged on announcement day but management
simply gave away the new shares receiving nothing in

return. In our sample the average stock issue is highly

dilutive but less than purely dilutive. On average,
shares outstanding increase by about 10% and the
stock price falls by 3%.
There is another way to view the loss in current
shareholders'wealth associated with raising new equity funds. Our results imply that a substantialportion of
the proceeds of an equity issue, in effect, comes out of
the pockets of current shareholders. On average, after
$100 million in new equity is raised, the market value
of the firm has increased by only $69 million. The
other $31 million is "donated"by existing shareholders
in the announcement period price reduction.
Regardless of the interpretation, the reductions in
market value reported in Exhibit 7 represent a hefty
haircutassociated with bringing new equity funds into
the firm from the external capital markets. The reductions may also be viewed as a barrierto external equity
financing - a cost which might be avoided by relying
on internally generated equity funds.

'4The proposed issue representedabout 1.6% of the shares then outstanding,while AT&T's stock price fell by 3.5%. AT&T's stock price
reboundedsomewhat the day after the announcement,but retreatedto
the post-announcementlow by the end of the week.

B. Equity Issues and Timing

Because a firm's stock is always correctly priced in
efficient markets, financial economists argue that

40

FINANCIALMANAGEMENT/AUTUMN1986

Exhibit 8. The Timing of Equity Issues: MarketWide Performance and Market-AdjustedPerformance

(CAR) for Two Years Before and After the Equity
Issue
TradingDay
Relative to the
Announcementof
a New Equity
Issue

CumulativeAverage
Returnon the
Market(S&P 500)

CumulativeAverage
AbnormalReturn
(CAR) for Issuing
Firms

-480
- 280
-80
-40
-20
-10

-0.4%
0.4%
6.0%
9.4%
11.6%

12.3%

0.9%
12.2%
33.5%
37.7%
39.5%
40.4%

Announcement Day
+10
+20
+40
+ 80
+280

12.7%
13.1%
13.2%
13.2%
13.5%
23.0%

35.5%
36.2%
35.6%
37.1%
39.0%
38.9%

+480

26.5%

32.4%

firms should always be willing to issue equity to finance worthwhile projects at the current stock price.
Managers should not be concerned about timing equity
issues to take advantage of high stock prices. Managers should be just as willing to sell equity when the
firm's stock price is near an historic low as when the
price is scaling new heights. In contrast, financial executives considering an equity issue appear to be very
concerned with the level of stock prices. Their enthusiasm for issuing equity seems to be directly proportional to the level of their firm's stock price.
Exhibit 8 provides insight into the timing of equity
issues. The results presented include both the cumulative movement of the market in general and the performance of sample firms relative to the market (i.e.,
the CAR) for selected time periods surrounding the
stock issue.
The market-wide results confirm that firms issue
equity when stock prices in general are rising. Nonetheless, the results reveal no ability to time the general
level of stock prices. Market returnsare positive in the
two years preceding the equity issue, and the market
continues to rise during the two years following the
offering.
A different picture emerges from the results on market-adjusted performance of issuing firms' stock
prices. In the two years prior to the equity offering,
sample firms on average outperform the market by

about 40%. In the ten days concluding with the announcement of the stock issue, sample firms' stock
prices fall by about 5%.15 Market-adjusted performance following the issue is at first slightly positive
then negative.
In addition, the price reduction associated with issuing equity is related to the firm's market-adjustedperformance priorto the issue. Regression analyses of our
results demonstrate that poorer stock price performance in the months preceding the offering is associated with larger price reductions.16This is consistent
with the notion that it is more difficult and costly to sell
equity when a firm's stock has not been performing
well. Investors are apparently more concerned about
the implications of managers' willingness to issue equity in the face of poor recent stock price performance.
Thus, firms issue equity following a period in which
the stock outperforms the market. Subsequent to the
issue, superior performance ceases and average or
below average performance is observed. In our sample
this timing pattern is observed only for the performance of a stock relative to the market in general, and
no systematic differences are apparentin market-wide
returns prior to and following stock issues.

VI. Why Are Stock Issues Bad News?

Our findings confirm financial executives' concerns
about issuing equity. Equity issues reduce stock
prices. The percentage reduction in stock price is
small, but the aggregate loss in market value is a large
fraction of the funds raised in the offering. While sample firms have been outperforming the market as a
whole, the announcement of an equity issue is associated with the cessation of superior stock price
performance.
There are several possible explanations for the negative stock price impact of equity issues. The negative
market reaction may result from earnings per share
dilution, supply-demand imbalance (or price pressure
effects), and/or the valuation effects of a change in
debt-to-equity ratio. Our analysis does not support
these explanations for three principal reasons. First,
the magnitudeof the price reduction is generally incon'5Thisis somewhatgreaterthanreportedin Exhibit6 due to differences
in the sample. Stock price datafor two years before and afterthe equity
issue are available for only 80 of the sample of 128 representedin
Exhibit 6.
'6Anaverageissuing firmoutperformedthe marketby about24%in the
eleven months precedingthe month of the offering and experienceda
3% reductionon announcementday. In contrast, a firm that underperformedthe marketby 24% priorto the issue experienceda price reduction of about 4.3% on the announcementdate.

WITH EQUITYTRANSACTIONS
ASQUITHAND MULLINS/SIGNALLING

sistentwith predictionsbased upon the threehypotheses presentedheretofore.Second, the negativemarket reactionvarieswidely acrossfirmsand,for sample
firmsthatexecutedmorethanone issue, the negative
marketreactionvaries widely for differentofferings.
The magnitudeand natureof this variabilityis not
consistentwith these explanations.Third, these hypotheses are not supportedby the reactionof stock
pricesto othertypes of equity offeringssuch as secondaryofferings and stock sales by knowledgeable
insiders.This final point warrantselaboration.
In additionto primarysales of new equityby corporations,ourstudyanalyzedregisteredsecondarydistrisalesof largeblocksof existbutions,theunderwritten
ing shares. The announcement of secondary
distributionsof equityis associatedwith a stockprice
reductiondespitethe factthatthis producesno dilution
in earningsper shareandno changein the firm'sdebt
ratio. Otherstudieshave documentedthatthe sale of
stockby insidersreducesstock prices.17Afteradjusting forthe size of the sale, secondarydistributionsand
insider sales produce price reductionssubstantially
largerthanthose accompanyingprimarysales of equity by firms.The pricereductionappearstoo largeto be
explainedby a supply-demandimbalance.The preponderanceof evidencesupportsan alternativeexplanationfor the negativemarketreactionto stockissues.
A. Equity Issues as Negative Signals
The decision to sell equity is made by executives
who possess an insider'sknowledgeof the firm, its
currentperformanceand futureprospects.When the
currentstockpriceis highrelativeto managers'assessmentof the firm'sprospects,thereis a powerfulincentive to sell stock to benefit the firm and its existing
shareholders.This incentiveis, of course, simplythe
mirrorimageof the incentiveto repurchasestockwhen
managersview theirstockas underpriced.Conversely,
when managementbelieves the firm's sharesare underpriced,thereis an incentiveto avoidissuingequity
even if the firmhas worthwhileprojectsto finance.To
protectthemselvesagainstthe risk of buyingovervalued shares, investorsmarkdown the stock price in
response to the announcementthat managementis
willingto sell equity.Indeed,this sortof pricehedging
is commonin anytradingsituationwheresomeparticipantsare viewed as having superiorinformation.18
'7For example, see Jaffee [13], and Finnerty [12].
'8For example, block traders routinely mark down the price when
buying securitiesfrom sellers whom they fear possess superiorinforma-

41

Of course, the firm selling equity may simply be

raisingfunds to finance a very profitableinvestment
project. Because of the informationimbalancebetween investorsand managersand investors'vulnerabilityto this imbalance,theremaybe no credibleway
to convince investorsof management'slaudablemotive for issuingequity.19Moreover,new equityissues
aretypicallya relativelysmallpercentageof the existing sharesoutstanding.New shareholdersare investing primarilyin the valuationof the firm's existing
assetsratherthanthe specificinvestmentprojectfunded by the sale. Regardlessof the outcomeof the project, new investors'returnswill be determinedprimarily by the futureperformanceof the firm's existing
businesses.Investorshave little recourseif they purchase overvaluedshares.
Thus, an equity issue is viewed by the marketas a
negativesignal. The stockpricereductionis produced
by investorshedging againstthe risk that, in selling
stock,informedmanagersarerespondingto the incentive to capitalizeon a favorablemarketvaluation.A
more benign interpretationis that the information
availableto managementis not so favorableas to preclude selling stock at the going price, and thus the
decisionto issue equity is a negativesignal.
This signallingexplanationis consistentwith our
empiricalfindings.The size of the equityissue representsthe size of the signal. Investorsfearthatmanagement'swillingnessto sell a largefractionof the firm's
equityreflectstheirassessmentthatthe stock price is
especiallyfavorablerelativeto theirsuperiorinformation. The variabilityof the negativemarketreactionto
equityissuesthroughtimeandacrossfirmsreflectsthe
varyinginformationcontentof equityissue decisions.
Negativereactionsto secondarydistributionsand insidersales suggest that whethermanagerssell equity
for theirown accountor for the firm'saccount,investors are concerned about the implications of the
decision.

The signallingrationaleis also consistentwith the

firm-specifictimingpatternobservedin our empirical
work. The decision to sell stock follows a periodof
superiorstockpriceperformance.The decisionto sell
equitynow, ratherthanwaitfor additionalpriceappreciation, suggests that managementdoes not foresee
continuedsuperiorperformance.The post-issuecessation. Price increases are associated with purchases by investors who
specialize in speculatingon takeover targets. The same is true of real
estate purchasesby buyers who are thought to have informationconcerning future real estate development.
19Myersand Majluf [20] discuss this issue in more detail.

42

FINANCIALMANAGEMENT/AUTUMN1986

Exhibit 9. The Capital Markets' Reaction to (Unanticipated) Equity Cash Flow Decisions
Capital Markets

Firm

Equity Cash Outflows

~~~~~~~~~~-400
Increase in Dividends, Stock Repurchases

Equity Cash Inflows

-

-dD

Decrease in Dividends,EquityIssues

Information
Imbalance
and
Signaling:

Management

Possessing
Superior
Information

Equity Cash Flows as Signals

Which Communicate and Reveal

Outside Investors
With Inferior

Management'sAssessment of
the Firm'sCurrentPerformance
and FutureProspects

Information

tion of superiorstock price performancereflectsboth

the validity of management'sassessmentand investors' responseto the equity issue signal.
Finally, the signallingstory is consistentwith anotheraspect of management'sattitudetowardequity
issues. When queriedabouttheir reluctanceto issue
equity, managersexplain that this reluctancestems
from the inappropriatelylow valuationplaced upon
their sharesby the market.20With this attitudeas a
backdrop,it is not surprisingthatinvestorsfearthata
decisionto sell equityreflectsthe temporaryreversal
of this assessment.
B. Is Issuing Equity a Bad Idea?
The answerto the questionposed in the headingis
"notreally."First, of course, selling stock when the
marketis overlyoptimisticmaybenefitthe firmandits
existing shareholders.Second, an equity issue may
makesenseif the firmhas worthwhileinvestmentprojects, insufficientinternalfundsflow, and insufficient
debtcapacity.In this case the benefitsof pursuingthe
projectmay outweighthe negative impactassociated
with externalequity financing.
20Despite the impressive advance in stock prices over the past two
years, a recentstudyby Louis Harris& Associates, Inc., foundthat60%
of the executives polled felt their stock was valued too low and about a
thirdof the sample executives felt their stock was seriously undervalued. Only 2%felt thattheirstock was overvaluedand 32%believed that
their companies' shares were correctly priced. See [23].

However, the negative signal inherentin issuing

equity will requirethe firm to forgo some profitable
investmentopportunities.This occurswhen the benefits of the projectdo not outweighthe negativeimpact
of issuingequity.In this case theremay be no credible
way managementcan persuadeinvestorsthatthe equity issueis motivatedby a worthwhileinvestmentrather
than the opportunityto take advantageof an overly
favorablestockprice. Investors'vulnerabilityto management'sincentiveto capitalizeon its superiorinformationcreatesa barrierto equityfinancing,a cost that
constrainsfirms from pursuingall worthwhileinvestmentopportunities.Firmscan avoid this difficultyby
designingfinancialpolicies to insureampleavailability of internalfunds to finance all worthwhileinvestmentopportunities.

VII. The Firm as a Black Box

Emergingfrom the empiricalstudies reviewed in
thisarticleis an interestingpattern.Despitethe associated tax burden, increasesin cash dividendsare received by investorsas favorablesignals. The same is
trueof stockrepurchases.The downsideis thatwhena
firmrequiresrefundingfromthe equitymarket,this is
viewed as a negative signal. Similarly,a cut in cash
dividendsis greetedwitha reductionin the stockprice.
As illustratedin Exhibit9, these studiesof equity
cashflows have documentedthatunanticipated
equity
cashflows andstockpricesarepositivelyrelated.This

WITH EQUITYTRANSACTIONS
ASQUITHAND MULLINS/SIGNALLING

suggestsa model of the firm based uponthe superior

informationpossessed by managersvis-a-vis outside
investors.2"
These findingsare consistentwith a view
of thefirmas a "blackbox,"whereunanticipated
equity cash flows communicateinformationto investors.
Cash outflows, stock repurchases,and increases in
dividends are positive signals accompaniedby increasesin stock prices. Conversely,if a firmrequires
cash inflows from the equity market,throughequity
issuesor a reductionin dividends,the resultis a negative signal and a reductionin the stock price.
The informationimbalancearises because of the
separationof ownership and management.Indeed,
thereare theoriesof economicorganizationthatview
efficiencyin informationprocessingas the factorthat
definesfirmsas economicentities.Efficiencyin informationprocessinghelpsdeterminewhichactivitiesare
undertakenwithinorganizationsandwhich are left to
markettransactions.22
Moreover,there are organizationalbehaviormodelsof the firmbaseduponinformation processing.23
The separationof managementand
investorsendowsmanagerswith superiorinformation
concerninga firm's currentperformanceand future
prospects.Equitycashflows serveas signalsthatcommunicatemanagerialinformationto investors.
Of course, it is unrealisticto portraythe firm as an
impenetrableblack box. Empirical studies demonstratethat accountinginformation,managerialstatements,andsecurityanalysts'reportsall providevaluable informationto investors.24Nonetheless, equity
cashflows appearto have valueindependentfromand
in additionto other informativesignals.
Whyis thereresidualsignallingvalueto equitycash
flows over andaboveotherinformativesignals?First,
equityis the residualclaimto the firm'scashflows and
carrieswith it no fixed promiseof return.Equityreturnsare highly volatile, and equity is simply very
difficultto value. Second, the informationimbalance
betweenmanagementandinvestorsis apparentlyquite
severe. Third,as notedearlierin this paper,cash has
uniqueadvantagesas a signal. Equitycash flows constitutesimple, highlyvisible, andcrediblemanagerial
signals. The reactionof the capitalmarketsto equity
cashflows illustratesthat,if not an impenetrable
black
2'Miller and Rock [19] develop such a model.
22Fora survey of relevant literature,see Marrisand Muellar [15].
23See for example Cyert and March [7].
24Fora review of some of this literature,see Copelandand Weston [6].

43

box, the firm is viewed by investorsas a "beigebox"

- an opaqueentityaboutwhichinvestorsdrawinferences from equity cash flow signals.

VIII. Interrelated Corporate

Financial Decisions
An importantsourceof the credibilityof dividend
andrepurchasesignalsis the negativemarketreaction
producedby equity issues and dividendreductions.
This negativeimpacton stock prices associatedwith
equitycash inflows imposesa cash flow constrainton
firms. Even thoughdividendincreasesand stock repurchasesare received as good news, firms that pay
out excessive equitycash flows may laterhave to replacethe fundspaidout withnew equityfinancingor a
reductionin dividends.The negativeimpacton stock
pricesof equityissues anddividendreductionsconstitutes a substantial"cost to false signalling,"which
keepsmanagementhonestandaddscredibilityto dividend and repurchasesignals.
The constraintsimposedby the informationimbalance betweenfirms and investorshave importantimplicationsfor corporatefinancialdecisions. It should
be apparentthat decisions concerningdividends,repurchases,and equity issues are interrelated.These
decisionsmustbe determinedjointly to avoid paying
the cost inherentin violatingthe cash flow constraint
and reducingdividendsand/orissuing equity.
Moregenerally,the information-induced
barrierbetweenthe firm andthe capitalmarketshelps bindthe
firmas an entity separatefromthe capitalmarkets.It
also binds the firm's major financial decisions and
forces the simultaneousdeterminationof investment
policy, capitalstructurepolicy, and dividendpolicy.
The necessityof jointly determiningfinancialpolicies
is mandatedby the constraintimposedby the negative
marketreaction to external equity financing. This
leads to policies that differ from those predicatedon
the assumptionthat a firm can always issue equityat
the currentstock price. Were this assumptionvalid,
decisions could be determined incrementallyand
independently.
The findingsexploredin this paperexplainfirms'
self-imposedequity capital rationing.The desire to
avoidthe negativeinformationimpactof havingto go
to the equitymarketfor funds(or reducingdividends)
encouragesfirmsto limit theirgrowthandinvestment
to that sustainablewith internallygeneratedequity
funds. This explainswhy so many firms use the sustainablegrowthparadigmas an integrativeplanning
frameworkin determiningfinancialpolicies.

44

IX. Future Research on

Equity Cash Flows
Althoughthe resultsreviewedin this paperprovide
useful insightto financialdecisionmakers,this work
does not constituteenoughprogresson how managers
shouldmakeequitycash flow decisions. More progresson this frontrequiresgoing behindcorporatedecisions to investigatehow and why decisionsare made
andwhy some decisionsarefavorablyreceivedby the
capitalmarketsandotherspoorlyreceived.Ourresults
suggestthatcorporatepolicies shouldnotbe studiedas
separatedecisions. Futureresearchneeds to focus on
the interrelatednatureof majorfinancialdecisions
how financialpolicies are reconciledwithinthe constraintsthatbind firms' decisions.
This future research will be more difficult than
measuringthe capital markets'reactionto corporate
decisions. But, the payoff promises to be correspondinglygreateras well. The work to date does
constituteimportantprogresstowardsolvingthe equity cash flow puzzles and provides a foundationfor
futureresearchdesignedto improvecorporatefinancial decisionmaking.

References
1. J. Aharonyand I. Swary, "QuarterlyDividend and Earnings Announcementsand Stockholders'Returns:An Empirical Analysis," Journal of Finance (March 1980), pp.
1-12.
2. P. Asquith and D. W. Mullins, Jr., "Equity Issues and
OfferingDilution,"Journal of Financial Economics(January 1986), pp. 61-90.
3.
"The Impact of Initiating Dividend Payments on
Shareholders'Wealth,"Journalof Business (January1983)
pp. 77-96.
4. F. Black, "The Dividend Puzzle," Journal of Portfolio
Management(Winter 1976), pp. 5-8.
5. G. Charest, "Dividend Information, Stock Returns and
Market Efficiency. II" Journal of Financial Economics
(June 1978), pp. 297-330.
6. T. E. Copeland and F. J. Weston, Financial Theoryand
CorporatePolicy, 2nd ed., Reading, MA, Addison Wesley, 1983.
7. R. M. Cyert and J. March, A Behaviorial Theoryof The
Firm, Englewood Cliffs, NJ, Prentice-Hall, 1963.

FINANCIALMANAGEMENT/AUTUMN1986

8. L. Dann, "Common Stock Repurchases:An Analysis of

Returnsto Bondholdersand Stockholders,"Journal of Financial Economics (June 1981), pp. 113-138.
9. G. Donaldson, CorporateDebt Capacity, Homewood, IL,
RichardD. Irwin, 1971.
10. K. M. Eades, "EmpiricalEvidence on Dividends as a Signal of Firm Value,"Journal of Financial and Quantitative
Analysis (November 1982), pp. 471-500.
11. M. FeldsteinandJ. Green, "WhyDo CompaniesPay Dividends," American Economic Review (March 1983), pp.
17-30.
12. J. E. Finnerty,"Insidersand MarketEfficiency," Journal
of Finance (September 1976), pp. 1141-1148.
13. J. Jaffe, "SpecialInformationand InsiderTrading,"Journal of Business (July 1974), pp. 410-428.
14. J. Lintner,"TheFinancingof Corporations,"in TheCorporation in Modern Society, edited by Edward S. Mason,
Cambridge,MA, HarvardUniversity Press, 1960.
15. R. Marrisand D. C. Muellar, "TheCorporation,Competition, and the Invisible Hand,"Journal of EconomicLiterature (March 1980), pp. 32-63.
16. R. Masulis, "StockRepurchaseby TenderOffer:An Analysis of the Causesof CommonStock PriceChanges,"Journal of Finance (May 1980), pp. 305-319.
17. R. Masulis and A. Korwar, "Seasoned Equity Offerings:
An Empirical Investigation,"Journal of Financial Economics (January1986) pp. 91-118.
18. M. Miller and F. Modigliani, "Dividend Policy, Growth
andthe Valuationof Shares,"Journalof Business (October
1961), pp. 411-433.
19. M. Miller and K. Rock, "Dividend Policy Under Asymmetric Information," Journal of Finance (September
1985), pp. 1031-1051.
20. S. C. Myers and N. S. Majluf, "CorporateFinancingand
InvestmentDecisions When Firms Have Informationthat
InvestorsDo Not Have," Journal of Financial Economics
(June 1984), pp. 187-221.
21. A. W. Sametz, "Trendsin the Volume and Compositionof
EquityFinancing,"Journal of Finance (September1964),
pp. 450-469.
22. T. Vermaelen, "CommonStock Repurchasesand Market
Signalling,"Journal of Financial Economics (June 1981),
pp. 139-183.
23. L. Wayne, "A Look at New CorporateTactics," The New
YorkTimes, Sunday, February26, 1984.
24. S. Wittebort, "Do Investors Really Care about Dividends?", Institutional Investor (March 1981), pp. 213220.