Journal of Financial Economics 11 (1983) 275-300.

North-Holland Publishing Company

STANDSTILL AGREEMENTS, PRIVATELY NEGOTIATED STOCK

REPURCHASES, AND THE MARKET FOR
CORPORATE CONTROL

Larry Y. DANN*
University of Oregon, Eugene, OR 97403, USA

Harry DeANGELO*
University of Rochester, Rochester, NY 14627, USA

Received September 1981, final version received February 1982

Standstill agreements are voluntary contracts which limit a substantial stockholder’i ownership
interest in a corporation for a specified number of years. They are often accompanied by
repurchase of the substantial stockholder’s shares at a premium above the market price.
Standstills and premium buybacks reduce competition for corporate control and provide
differential treatment of large block stockholders. The analysis indicates a statistically significant
negative average effect on non-participating stockholder wealth associated with standstill
agreements. Negotiated premium repurchases are also associated with negative, but less
significant, stockholder returns. The evidence is inconsistent with the hypothesis that these
management actions are in the best interests of non-participating stockholders.

1. Introduction

An important premise of corporate finance theory is that markets

discipline managers to maximize all stockholders’ wealth. Competitive forces
in two markets, the market for corporate control and the market for
managerial labor services, are widely viewed as providing complementary
enforcement of the stockholder wealth maximization rule. Standstill
agreements and negotiated common stock repurchases at a premium above
market price are transactions which (i) reduce competition for corporate
control and (ii) provide for differential treatment of large block stockholders.
*This paper was completed while DeAngelo was at the University of Pennsylvania. We would
like to acknowledge the helpful comments of F. Easterbrook, M.C. Jensen, P. Linneman, R.
Masulis, W. Mikkelson, and the referee, J. Warner. Special thanks are due to L.E. DeAngelo and
E.M. Rice for many helpful discussions and to J. Sington for able research assistance. D.
Whitehurst and C. McNeil1 provided invaluable assistance in preparing this manuscript. We are
grateful to the Managerial Economics Research Center at the University of Rochester for partial
financial support. Responsibility for any errors is ours. After this paper was completed, we
received a copy of a related paper by Michael Bradley and Lee Wakeman. Readers interested in
the issues discussed here should also see the Bradley and Wakeman paper which is published
elsewhere in this volume.

030~05x/83/.$03.00 0 Elsevier Science Publishers

276 L.Y Dann and H. DeAngelo, Standstill agreements and stock repurchases

Consequently, these transactions represent ostensible departures from the

view that markets discipline managers to maximize all stockholders’ wealth.
A standstill agreement is a voluntary contract between an issuing
corporation and a substantial stockholder which limits the stockholder’s
ownership of voting shares to some maximum (less than controlling)
percentage for a stipulated number of years. In a typical standstill, NVF
Company and affiliates and City Investing recently agreed that the former’s
ownership position in City Investing be limited to a maximum of 21% for the
next five years. The agreement stipulates that NVF will not solicit proxies or
otherwise participate in a control contest against the incumbent management
of City Investing. Also, should NVF decide to sell its large block, City
Investing has the right of first refusal to repurchase the shares. The
ownership limitation agreement is not binding if a third party makes a
tender offer for more than 21% of City Investing or if City Investing
proposes a merger with a third party.’ Standstill agreements sometimes
include a stronger transferability restriction which specifies that the issuing
firm’s approval must be obtained before the large block holding can be sold.
A standstill agreement is frequently accompanied by a repurchase of the
substantial stockholder’s block at a premium above the open-market price.
For example, Bliss and Laughlin Industries Inc. recently repurchased Solar
Sportsystems Inc’s 7.4% ownership stake at a 17.6% premium above market
in return for Solar’s (standstill) promise not to own B and L shares for two
and one-half years.’ In numerous other cases, a firm privately negotiates a
repurchase of a large block holding at a premium, but does not obtain an
explicit standstill promise from the seller. The often quite substantial
premium paid in a privately negotiated repurchase is, in contrast to the
typical stock repurchase by tender offer, not made available to other
stockholders.
The favorable treatment accorded large block stockholders in negotiated
premium repurchases is legally valid under extant corporation laws. The
courts have consistently allowed differential payments to large block
stockholders as long as firms demonstrate a valid business purpose for the
transaction, Moreover, the courts have given broad interpretation to factors
which represent a valid business purpose. According to Nathan and Sobel
(1980, fn. 2), court-accepted purposes include (1) a ‘. . . difference in business
philosophies between the corporation’s management and the selling
shareholders’ or (2) ‘. . . eliminating what appeared to be a threat to the future
of the business and preserving an established management’s business policy’.3
Standstill agreements and negotiated premium repurchases often culminate

‘The Wall Street Journal, August 6, 1980, p. 3.

‘The Wall Street Journal, December 15, 1980, p. 30.
3Nathan and Sobel provide an extensive discussion of the legal issues surrounding the
relationship between share repurchases and takeover bids.
 L.Y. Dam and El. DeAngelo, Standstill agreements and stock repurchuses 217

a public control battle between incumbent management and a substantial

stockholder. Regardless of whether a control contest had previously
materialized, the effect of these transactions is to reduce the threat of
takeover posed by a substantial stockholder. Indeed, for corporation finance
theory, the important common elements of standstill agreements and
negotiated premium repurchases are that both reduce potential competition
for control and both entail differential treatment of large block stockholders.
Two competing hypotheses capable of explaining standstill agreements and
negotiated premium buybacks have received attention in the literature.
Briefly summarized, the managerial entrenchment hypothesis predicts that
non-participating stockholders suffer wealth losses when incumbent
management acts to deter a credible threat of control transfer posed by a
substantial stockholder. The competing stockholder interests prediction is
that non-participating stockholders gain because the reduced threat of
competition for control leads to real resource savings associated with the
(assumed costly) process of competition.4
This paper presents a theoretical and empirical analysis of standstill
agreements and negotiated premium repurchases. We begin by outlining the
managerial entrenchment and stockholder interests explanations for these
transactions (section 2). We then test the competing hypotheses by examining
the impact on (non-participating) stockholder wealth of standstill agreements
and negotiated premium repurchases for a sample of firms which engaged in
these transactions during the period 1977-1980 (sections 3-5). A brief
summary concludes the paper (section 6).

2. Costly contracting and corporate control

When contracting is costless, managers and stockholders will resolve all

conflicts of interest by designing contracts which ensure that managers follow
a policy of stockholder wealth maximization. When contracting is costly, the
possibility of conflicts of interest remains and a fundamental question arises:
to what extent do managers deviate from stockholder wealth maximization?
According to the stockholder interests hypothesis, sufficient constraints
exist to limit severely the extent of management’s deviation from stockholder
wealth maximization. For example, Manne (1965, p. 113) has argued that the
threat of takeover is an important constraint against managerial inefticiency
(departures from stockholder wealth maximization) which ‘. . . affords strong
4The managerial entrenchment hypothesis has been advanced by Cary (1969) and Williamson
(1975) and further elaborated by DeAngelo and Rice (1983). DeAngelo and Rice have also
specified a variant of the stockholder interests hypothesis wherein takeover defenses (in
particular, antitakeover charter amendments) benefit common stockholders by more firmly
defining the property rights to effect control transfer. Related arguments supporting the
stockholder interests hypothesis have been advanced by Bradley (1980, p. 348) Grossman and
Hurt (1980) and Jarrel and Bradley (1980, p. 386).
278 L.Y. Dann and H. DeAngelo, Standstill aqeements and stock repurchases

protection to the interests of vast numbers of small, non-controlling

shareholders’. A critical implicit assumption of this argument is that
managers cannot (do not) take unilateral actions which materially blunt the
takeover threat to the detriment of stockholders. The competing managerial
entrenchment view is that incumbent managers have considerable latitude to
consume perquisites. Furthermore, as Williamson (1975, p. 160) has
hypothesized, managers can and do opportunistically act to reduce the
takeover threat and thereby expand their perquisite consumption possibilities
at stockholder expense. Standstill agreements and privately negotiated
premium repurchases are actions taken by incumbent management which
reduce the threat of takeover by a large block stockholder. The predicted
effects on non-participating stockholder wealth of these actions are different
under the managerial entrenchment and stockholder interests hypotheses.

2.1. The managerial entrenchment hypothesis

The managerial entrenchment hypothesis predicts that the dominant effect

of standstill agreements and negotiated premium repurchases is to reduce the
wealth of non-participating stockholders. Two complementary factors lead to
this wealth reduction. First, managerial behavior is imperfectly constrained
by the threat of takeover in the capital market so that a reduction in this
threat induces greater managerial shirking (or perquisite consumption) at
stockholder expense.’ Second, a large block stockholder has a comparative
advantage relative to small stockholders in either initiating a takeover
attempt or in transferring this ability to a third party. Moreover, a large
block holder has a greater incentive to incur the costs of a takeover attempt
since he stands to capture a greater share of any resulting improvements in
managerial efficiency.
The potential personal costs associated with a change in corporate control
provide incumbent management with the incentive to use corporate resources
to blunt a takeover threat. Hence, when a substantial stockholder does pose
a significant threat, management is willing and able to sacrifice the interests
of small stockholders to mitigate the greater takeover threat posed by the
large block holder. The assumption here is that significant costs of
mobilization limit the extent to which dispersed shareholders oppose
management’s differential treatment of a large block stockholder.
Consequently, in at least some cases, a substantial stockholder with a
credible threat is able to extract favorable treatment from management in
exchange for a reduced takeover threat.
‘An implicit assumption here is that substitutes for the takeover mechanism do not operate
costlessly to eliminate all managerial inefficiencies in that (i) managerial compensation contracts
which would align managers’ incentives closely with those of stockholders are prohibitively
costly to write and enforce and (ii) the managerial labor market does not operate frictionlessly.
For elaboration, see Fama (1980) and DeAngelo and Rice (1983, section 3).
 L.Y. Dann and H. DeAngelo, Standstill agreements and stock repurcha.ses 219

One managerial response to a credible takeover threat is to pay an above-

market price to repurchase the large block holding and thereby reduce the
likelihood of control transfer. The obvious benefit to the large block
stockholder is direct cash payment in an amount exceeding his share of
benefits according to strictly proportional stock ownership. These payments
are, of course, inconsistent with the conventional view that each share of a
firm’s common stock provides identical rights and commands the same unit
price regardless of the quantity owned by a particular individual. (We return
to this point in section 5.2 below.) Indeed, the payment of a premium in a
given negotiated repurchase indicates that this particular large block
stockholder has superior ability or power to extract more than his
proportional share from the tirm.‘j
Given a credible takeover threat, another managerial response is to
negotiate a standstill agreement wherein the substantial stockholder promises
not to exceed a stipulated (non-controlling) ownership fraction for some
number of years. Clearly, large block stockholders benefit from voluntarily-
agreed-to standstills although the form of compensation is sometimes not
evident. Frequently observed forms of compensation are representation on
the board of directors and settlements wherein the firm drops litigation
pending against the stockholder (e.g. claims of antitrust or disclosure law
violations). Less frequently observed forms of compensation are joint-venture
participation rights and reimbursement for proxy fight expenses.’
Thus, according to the entrenchment hypothesis, both incumbent
management and the participating stockholder benefit when they voluntarily
engage in negotiated premium repurchases and/or standstill agreements.

6An important unanswered question is why this premium extraction capability is held by
some large block stockholders but not others. We suspect the answer turns fundamentally on
the degree of competition from potential large block stockholders and hence on advantages to
incumbency associated with an established block holding. We conjecture that the extent of
competition among incumbent and potential large block stockholders also depends at least on (i)
relative abilities in monitoring and/or managing the firm and (ii) relative technical abilities and
associated capital market reputations to effect a control transfer via tender offer or proxy tight.
While a detailed analysis of the determinants of premium extraction ability is a worthwhile
endeavor, such an inquiry is beyond the scope of this paper. Consequently, we will take the
payment of a premium as indication of a situation where a current substantial stockholder has
superior ability and leave to future research the interesting question of the determinants of this
ability.
‘From the perspective of finance theory, it is puzzling that a large block stockholder would
enter into a standstill agreement without an accompanying direct payoff (such as a premium
repurchase). This is especially true given the central role typically accorded the threat of
complete takeover in disciplining managerial behavior. However, complete takeover is only one
managerial disciplining strategy which cannot be claimed on a priori grounds to represent the
optimal strategy in all cases. It is possible that informal monitoring and private negotiations
between incumbent management and a substantial minority stockholder (e.g., but not
necessarily, via board representation) may sometimes represent the optimal monitoring
alternative available to the large block stockholder. Indeed, it is difficult to argue otherwise
given that the large block stockholder has a substantial equity interest in the firm and
voluntarily participates in the transaction.
 280 L.Y Dunn and H. DeAngelo, Standstill agreements and stock repurchases

These takeover defenses reduce the wealth of non-participating stockholders

who suffer the valuation consequences of greater managerial inefficiency (due
to reduced takeover threat) and direct differential payments to large block
holders in a negotiated premium buyback.

2.2. The stockholder interests hypothesis

The competing stockholder interests hypothesis predicts that the dominant

effect of standstill agreements and negotiated premium repurchases is to
benefit non-participating stockholders. Substitute mechanisms governing
managerial behavior [e.g. managerial labor market forces as discussed by
Fama (1980)] are assumed to be sufficiently powerful to ensure that any
reduction in the threat of takeover (due to a standstill/buyback) does not
lead to material increases in managerial inefficiency. Furthermore,
competition for control’ between incumbent management and potential
acquirers is assumed to be a costly process so that a reduction in this
competition generates a substantial cost savings to stockholders.
Current stockholders bear (assumed) significant costs of incumbent
management’s competition to maintain control. These costs include direct
expenditures on information production about managerial performance, e.g.,
via ongoing stockholder relations departments, press releases and direct
mailings to stockholders, detailed annual reports, etc. If a control battle
actually materializes, firm resources are used for financial press
,advertisements, to hire proxy solicitation firms, and to obtain legal and

‘At least at this stage of development, the managerial entrenchment hypothesis, if correct,
does not imply that negotiated settlements of control contests should be prohibited on grounds
of economic efficiency. An alternative view - and one that can be neither refuted nor supported
with our evidence - is that differential treatment of large block stockholders can actually
increase economic efficiency. In particular, recognize that a large block stockholder will bear a
disproportionately large share of the costs of monitoring the firm (measured relative to his claim
on benefits under strictly proportional sharing) because each small stockholder has an incentive
to free ride on the monitoring activities of others. A large block stockholder will increase his
monitoring expenditures when such activity potentially allows him to capture more than his
strictly proportional share of benefits. Consequently, as Easterbrook and Fischel (1982) have
argued, all stockholders potentially benefit ex-ante from corporation laws and charters which
allow deviations from proportional sharing, e.g. which allow management the flexibility to buy
out a large block stockholder at a premium or to negotiate a standstill agreement. In other
words, even if the managerial entrenchment hypothesis holds in that we periodically observe
management sacrificing small stockholders’ interests to those of large block stockholders, it may
be that the potential for such future sacrifices induces more monitoring by large block
stockholders and hence greater efficiency than would occur under strict enforcement of
proportional sharing. Whether this efftciency-increasing view is correct is an unresolved (and
indeed, unaddressed) empirical question.
‘The existence of competition for control is consistent with, but does not require, that
incumbent management be shirking or consuming perquisities at stockholders’ expense.
Competition can arise even when incumbent management is taking all feasible actions to
maximize stockholder welfare. In particular, competition for control can arise when (i) current or
potential stockholders perceive the potential for superior performance by new management and
(ii) there are material information costs of evaluating relative managerial abilities.
 L.Y Dann and H. DeAngelo, Standstill agreements and stock repurchases 281

investment banking advice. Firm resources are also frequently used to retain
the services of takeover defense specialists in anticipation of a control battle.
In addition to these information costs, management time is another firm
resource which is directly consumed in competition for control. Management
time directed toward securing control is time not spent generating profit for
stockholders. Thus, the direct costs of the competitive process (to
stockholders) properly include the productive opportunities foregone when
managerial effort is allocated to maintaining the right to direct firm
resources.
There are clear potential benefits (cost savings) to stockholders from
reductions in competition for corporate control. Standstill agreements and
negotiated premium repurchases are, in essence, contracts wherein large
block stockholders agree to limit their competition to direct the firm’s
resources. Non-participating stockholders benefit from these non-competitive
agreements because the reduced threat of takeover encourages a substitution
of firm resources into profit-generating endeavors and away from activities
whose purpose is to discourage attempts to transfer control.” In the
stockholder interests view, these cost savings outweigh any incremental
inefficiencies induced by standstills and premium buybacks so that (non-
participating) stockholders are predicted to gain on net from these
transactions.
In the remainder of this paper, we test this (wealth-increasing) prediction
against the competing (wealth-decreasing) prediction of the managerial
entrenchment hypothesis by examining the average equity value impact of
standstill agreements and/or negotiated premium repurchases for a sample of
firms engaging in these transactions. Observation of a negative average price
impact would be consistent with the hypothesis that the average effect of
these transactions is to entrench incumbent management at the expense of
non-participating stockholders. Observation of a positive price impact is
consistent with the hypothesis that the average effect of standstills and
negotiated premium repurchases is to benefit non-participating stockholders.
The interpretation of the tests presented below is subject to an important
caveat. A significantly negative average price impact does not indicate that
managerial entrenchment is the dominant effect of each standstill and
premium buyback. It is possible that, for some firms within the sample, non-
participating stockholders actually benefited from these transactions.
Conversely, a significantly positive average price impact does not rule out the

“‘DeAngel and Rice (1983) also note that takeover defenses can benefit target stockholders
by avoiding costs associated with competition for control. One could also argue that target
stockholders benefit from reduced competition for control by reducing any managerial incentives
to placate voting stockholders by taking less profitable projects with more easily observed
payoffs.
282 L.Y. Dann and H. DeAngelo, Standstill agreements and stock repurchases

possibility that some standstills/buybacks actually entrenched incumbent

management at stockholder expense.‘l

3. Data sources and sample design

Standstill agreements and privately negotiated stock repurchases for firms

listed on the daily CRSP tapeI were originally identified by direct
inspection of each issue of The Wall Street Journal for 1977-1980. For each
identified event, The Wall Street Journal Index was consulted to verify that
direct inspection had yielded the first report of the transaction. The day of
the first report in The Wall Street Journal was taken to be the date of first
public announcement of the standstill or buyback.13
This search process yielded 142 standstill agreements and/or privately
negotiated stock repurchases. Of these 142 observations initially identified, 61
were excluded from the final sample either because (1) potentially
confounding events were contemporaneously reported in The Wall Street
Journal (41 cases)14, or (2) the consideration paid in a negotiated repurchase
not accompanied by a standstill agreement consisted (at least in part) of
assets other than cash (20 cases). We restricted our analysis of negotiated
repurchases to cash buybacks in order to facilitate determination of whether
the repurchase occurred at a premium relative to the pre-announcement

“Signalling and personal&corporate tax saving explanations for share repurchase have been
investigated by Dann (1981), Masulis (1980b), and Vermaelen (1981). The signalling and tax
models do not consider the differential treatment of large block stockholders which concerns us
here. We discuss the relationship between our findings and those of prior share repurchase
studies in section 5.2 below.
“Centerfor Research on Security Prices, University of Chicago, 1981.
13To minimize the chance that direct inspection had not revealed a reported standstill or
premium buyback, the sample was cross-checked against the private files of several investment
banking firms and the reacquired shares listing of The Wall Street Journal index which is
available beginning with 1978. In three instances, private sources indicated a standstill agreement,
but The Wall Street Journal report did not. Since we are concerned with the impact of publicly
available information on market prices, an observation was included in the standstill sample
only if The Wall Street Journal explicitly reported a voluntary agreement which limited a
stockholder to a less than controlling ownership position.
140f the 41 observations with potentially confounding events, 9 involved standstill agreements
and 32 involved negotiated cash buybacks without associated standstills. The events judged
potentially confounding were simultaneous announcement regarding stockholders’ opportunities
to sell stock back to firm via tender offer or open-market repurchases (10 cases), simultaneous
earnings announcement (7), simultaneous announced change in firm’s dividend payout policy (6),
public announcement date was not clearly identifiable (5), simultaneous announcement re:
investment project (3), government mandated sale by block holder (2), and one each of the
following simultaneous announcements: divestiture, convertible debt call, convertible debt
issuance, litigation filed on unrelated issue, secondary offering, repurchase of unregistered shares,
seller under investigation for trading irregularities in repurchasing firm’s stock, government
imposed restraining order against further stock purchases.
 L.Y. Dann and H. DeAngelo, Standstill apeemrnts and stock repurchases 283

market price. l5 The set of 81 observations that survived this screening

process contains 19 standstill agreements without negotiated repurchases, 11
standstills accompanied by negotiated repurchases, and 51 negotiated cash
repurchases without standstill agreements.”
Table 1 indicates that the 30 standstill agreements in our sample were
reasonably long-term contracts, lasting on average some 5 years. The average
contract length is approximately the same for standstill agreements
accompanied by a negotiated repurchase and for those not so accompanied.
However, the (voluntarily-agreed-to) maximum allowed ownership stake is
quite different depending on whether or not a repurchase accompanies the
agreement. For the 19 observations without associated repurchase, the
average ownership ceiling is set at slightly more than 20 percent. Here, the

Table 1
Standstill agreements (197771980): Maximum allowed ownership percentage and contract
duration (30 events).

Maximum allowed Length of Percentage of

ownership of standstill outstanding common
voting stock agreement stock repurchased
No. of sample mean sample mean sample mean
Type of transaction obs. (median) (median) (median)

All standstill agreements 30 14.7?:, 5.0 years

(15.2”<,) (5.0 years)
without repurchase 19 22.8u/, 4.4 years
(21.0%) (5.0 years)
with (measurable) 7 o.o;,;, 6.3 years 11.1%
premium repurchase” (0.0%) (5.0 years) (9.4”:,)
with (measurable) 1 0.0:); 10.0 years 5.09’
repurchase not at (O.O”/) (10.0 years) (5.0;)
premium”
with repurchase 3 0.33”/;:, 4.0 years 17.1::,
but compensation (0.07;) (5.0 years) (15.77;)
not measurable

“Premium measured relative to dividend-adjusted market closing price two days prior to The
Wall Street Journal
report of the transactton.
The market closing price (prior to announcement) was obtained from Standard and Poor’s
Stock Price Record. The repurchase price, maximum allowed ownership stake, and length of
agreement were obtained from The Wall Street Journal report of the transaction.

15We employed information in The Wall Street Journal reports to classify the consideration
paid in each repurchase. Three standstills were accompanied by negotiated repurchases that
were not strictly cash transactions. These observations were retained for analysis in the standstill
samples reported in section 5.1 below, but were excluded from subsequent analysis of negotiated
cash repurchases.
“The final sample contains 14 observations with announcement dates in 1977, 16 in 1978, 22
in 1979, and 29 in 1980.
284 L.Y Dann and H. DeAngelo, Standstill agreements and stock repurchases

typical stockholder agrees to limit holdings to a substantial, but clearly less

than controlling, ownership block. In contrast, when a repurchase is
involved, the former substantial stockholder agrees to a ‘hands-off’ stock
ownership policy - i.e., the average contract here is characterized by a zero
percent maximum allowed ownership.
Table 2 indicates that the 58 negotiated cash repurchases in our sample
involved fairly large ownership blocks. The 9.8”/, average stake repurchased
(8.0”/0 median) is smaller than that in the typical tender offer but larger than
that in the iverage open-market buyback. For tender offers, Dann (1981),
Masulis (1980b), and Vermaelen (1981) report that the average fraction of
shares repurchased is 1415%. For open-market buybacks, Vermaelen (1981)
reports that approximately 5’j/, of the outstanding shares were repurchased
on average. Note also from table 2 that negotiated repurchases at a premium
over market price involved larger blocks of stock on average than those not
transacted at a premium. (The mean fraction repurchased increases to 11.0%
when the 17 non-premium repurchases are excluded.)

Table 2
Privately negotiated stock repurchases with strictly cash payments during 1977-1980:
Percentage of outstanding shares repurchased and percentage premium payments
(58 events).

Percentage
of outstanding
common stock
repurchased Repurchase premium”
sample mean sample mean
Type of transaction No. of obs. (median) (median)

All negotiated 58 9.8% 10.2%

repurchases (8.0%) (10.5%)
All negotiated 41 11.0% 16.4%
repurchases at a (9.0%) (15.1%)
premium” over
market price
Negotiated 34 11.0% 14.4%
repurchases at a (8.6%) (14.8%)
premium” without
standstill

“Premium measured as repurchase price relative to dividend-adjusted market

closing price two days prior to The Wall Street Journal report of the
transaction.
The market closing price (prior to announcement) was obtained from Standard and
Poor’s Stock Price Record. The Wall Street Journal report was used to determine the
percentage of shares repurchased, the transaction price, and whether the repurchase
was accompanied by a standstill agreement. In several instances, information in
Standard and Poor’s Security Owners’ Stock Guide was utilized to estimate the
percentage of outstanding shares repurchased.
 L. I: Dunn and H. DeAngelo, Standstill agreements and stock repurchases 285

For the sample partitioning presented in table 2, we define the repurchase

to be at a premium if the repurchase price exceeds the dividend-adjusted
closing market price two days prior to The Wall Street Journal report of the
transaction. We have also measured the repurchase premia relative to the
dividend-adjusted market closing price 5 trading days and 20 trading days
prior to announcement. Our results do not change in any material way
under these alternative measurements of the premia.
Finally, note that for the entire sample of 58 negotiated repurchases, the
average payment to the selling stockholder involves a premium of 10.2% over
the pre-announcement market price. For the 41 premium repurchases, the
transactions price exceeds this market price by 16.4% on average. This
sizable premium is the same order of magnitude as the effective average
premium received by tendering stockholders in repurchase tender offers.17

4. Empirical methodology

To test the managerial entrenchment versus stockholder interests

hypotheses we measure price impacts relative to a benchmark estimated from
the following market model:

rjt = tlj + /3jrmr + Ujt,

where rjt is the dividend-inclusive rate of return on security j over period t,

r is the dividend-inclusive rate of return on an equal-weighted market index
o;er period t,‘* pj = cov (rjt, r,,)/var (r,,& cxj= E(rj) - DjE(r,), and uj, is the
disturbance term of security j over period t with E(uj,)=O.
For each sample observation, calendar time is converted to event time by
defining the calendar date of report of the transaction in The Wall Street
Journal as event day 0. The immediately prior market trading day is denoted
event day - 1, the immediately subsequent trading day is denoted event day
+ 1, and so forth.

“The offer price exceeds the prior market price by 22%23%;; in the average repurchase by
tender bid. This overstates the effective premium received by tendering stockholders since
tender bids are subject to pro-rating in the event of over-subscription. The expected pro-rating
effect is capitalized in the approximately 16% average market price impact of a repurchase
tender bid. This overstates the effective premium received by tendering stockholders since
stockholders and is quite close to the 16.4% average paid in negotiated premium repurchases.
The tender offer statistics discussed here are based on the discussions in Dann (1981), Masulis
(1980b), and Vermaelen (1981).
“Since we report as our measure of the average price impact an equal-weighted mean of the
individual security price impacts, there is some logical consistency in using a similar weighting
scheme for our index of market returns. Moreover, Brown and Warner (1980) report that using
an equal-weighted market index offers no systematic disadvantages and perhaps slight
advantages in detecting price impacts relative to alternative specifications.

I.F.E. K
286 L.Y Dann and H. DeAngelo, Standstill agreements and stock repurchases

In applying the market model, we designate two time periods relevant to

the measurement of security price performance. One time period, which we
refer to as the analysis period, represents the period of time surrounding the
initial announcement of a standstill agreement and/or a negotiated
repurchase. Detection of abnormal security price performance is confined to
the analysis period. A second (and distinct) time period, herein labeled the
estimation pkriod, constitutes the period of time from which estimates of
market model parameters aj and /Ij are obtained. The estimation period is
generally selected as a period of time ‘close’ to the analysis period, but one in
which no effects on security prices of the economic event under study are
expected to occur. For the results presented below, our analysis period spans
the period from 40 trading days before the announcement date (day 0)
though 40 trading days following announcement. The estimation period
consists of the 60 trading days before and 60 trading days after the analysis
period.”
Given market model parameter estimates ij and flj from the estimation
period, we calculate prediction errors, pej,, for each firm j over each period t
within the analysis period as follows:

pejt = rjt-ij- tjrmt.

The prediction error pej, is an estimate of the abnormal return to owners of

common stock j for period t. The cross-sectional average prediction error for
event day t is

where N, is the number of sample firms which have prediction errors on

day t. The average prediction error (pe,) measures the mean abnormal
performance on event day t, and can be interpreted as the abnormal return
to an equal-weighted portfolio of the sample securities that is formed in
event time.
Although we define the date that the announcement is reported in The
Wall Street Journal as day 0, the announcement generally will have been
made on the prior day and reported with a one day lag. In a semi-strong
form efficient market, the price impact of standstills and/or premium
buybacks should be registered on day - 1 if public announcement of the
transaction came before the close of trading or on day 0 if announcement
came after the close. Because the available data does not allow us to

“For 10 events, fewer than 100 post-announcement daily returns were available on the 1980
CRSP daily returns file. For these events, the estimation period consists of the 60 days
immediately preceding the analysis period and as many days beyond the analysis period as there
were returns available.
 L. Y. Dann and H. DeAngelo, Standstill agreements and stock repurchases 287
r

ascertain the within-day timing of public announcement, the return over both
days - 1 and 0 offers the greatest feasible precision in assessing the price
impact of public announcement of standstills and/or buybacks. Thus, for
purposes of statistical testing, we utilize two day rates of return.
Two test procedures are employed to assess the statistical significance of
the price impacts at announcement. Our principal statistical procedure is a t-
test. To describe this two-day test-statistic, first let PE, represent the portfolio
two-day prediction error which is calculated from the daily prediction errors
for event days t- 1 and t,

PE,=(l +pe,_,)(l +$,)- 1.

Here, the date subscript t in PE, indicates the end point of each two-day
period for calculating prediction errors. For example, PE, = [( 1 + $- 1)
x (1 + jGo) - l] is the two-day announcement period return over days - 1 and
0. This portfolio prediction error PE, estimates the average impact on non-
participating stockholder wealth of standstill agreements and/or negotiated
premium repurchases. 21 Under the null hypothesis of no abnormal return,
this announcement period prediction error is expected to be zero.
Let D represent the set of dates of two-day returns used to estimate the
standard deviation of the average (portfolio) prediction error,

D=i-39, -37,...,-5, -3, +4, $6 ,..., $38, +40j.

This specification excludes the portfolio prediction errors immediately

surrounding announcement (event days -2 through -+ 2) from D, and
produces 38 non-overlapping two-day prediction errors PE, with t ED. The
estimated standard deviation of two-day portfolio prediction errors is given
by

“To the extent that there is prior leakage, the return over days - 1 and 0 will be an
attenuated measure of the total wealth impact of the standstill/buyback transaction. While prior
leakage will affect the estimated magnitude of the price impact, it should not impart a
directional bias to the observed effect (assuming semi-strong form market efficiency).
Consequently, the possibility of leakage does not bias the tests presented below in favor of either
the managerial entrenchment or stockholder interests hypotheses.
“The portfolio two-day prediction error,, PE,,, is the compound abnormal rate of return of
the daily means of individual security predtction errors on days - 1 and 0. Using the portfolio
two-day prediction error in our statistical tests allows us to incorporate the returns information
for the firms that traded on only one of the two days (one standstill firm and one negotiated
premium repurchase firm).
288 L.Y. Dann and H. DeAngelo, Standstill agreements and stock repurchases

where $ is given by

fi= 38.

The following statistic is used to determine whether the average

announcement date price impact of a standstill agreement and/or premium
negotiated repurchase is significantly different from zero:

On the assumption that the two-day portfolio prediction errors are

independent drawings from a stationary normal distribution, this statistic is
Student t-distributed with thirty-seven degrees of freedom.
As a supplemental significance test, we employ the non-parametric
Wilcoxon signed-ranks test to test the null hypothesis that the median two-
day prediction error at announcement is equal to zero. In essence, this
statistic tests whether the location parameter of the cross-sectional
distribution of two-day prediction errors is zero. While there is some
evidence indicating that the Wilcoxon non-parametric test is inferior to the
parameteric t-test, we report the non-parametric statistic simply as a
complementary check on our parametric test of the average price impact at
announcement.”
Finally, as a robustness check of our use of the prediction error
methodology, we replicated our tests using a variant of the mean adjusted
returns methodology employed by Masulis (1980a) and Dann (1981). Our
findings under this alternative methodology (not reported here) regarding the
managerial entrenchment versus stockholder interests hypotheses are
qualitatively identical.

5. Empirical results

5.1. Market reaction to the first public announcement of a standstill agreement

The daily average (pe) and the cumulative average (cpe) prediction errors
of returns to stockholders of firms negotiating a standstill agreement with a
substantial minority shareholder are presented in table 3 for the 40 days
“Brown and Warner (1980, pp. 218-222) report evidence suggesting that the Wilcoxon
signed-ranks test is misspecified for significance tests of monthly security returns or prediction
errors. They attribute the misspecification to the right skewness in security specific performance
measures noted by Fama, Fisher, Jensen and Roll (1969), which contrasts with the Wilcoxon test
assumption of a symmetrically distributed random variable. But Fama (1976, p. 30) indicates
that whereas monthly returns are slightly right-skewed, there is almost no skewness in daily
returns. Consequently, there is some reason to believe that the misspecilication of the Wilcoxon
test for monthly security specific performance measures does not carry over to daily measures.
L.Y Dunn and H. DeAngelo, Standstill agreements and stock repurchases 289

Table 3

Percentage daily average (pe) and cumulative average (cpe) prediction

errors for the Il-day period centered around announcement of a
standstill agreement.

All standstill agreements Standstill agreements not

(including those accompanied accompanied by a
by a negotiated stock negotiated stock repurchase ~
repurchase) ~ 30 companies 19 companies

(1) (2) (3) (4) (5) (6)

Trading Trading
day P” cp’ day Pe c’pe

-40 - 0.07 -0.07 -40 0.87 0.87

-35 -0.35 -0.50 -35 -0.51 0.65
-30 -0.57 -0.39 -30 -0.49 0.20
-25 -0.20 - I.54 -25 - 0.07 -0.97
-20 0.43 -3.11 -20 0.46 -2.37
-19 -0.19 -3.30 -19 0.43 - 1.94
-18 1.05 -2.25 -18 1.36 -0.58
-17 0.26 - 1.99 - 17 0.37 -0.21
-16 0.25 - 1.74 -16 0.48 0.27
-15 0.34 - 1.40 -15 0.94 1,21
-14 0.22 - 1.18 - 14 -0.20 1.01
-13 0.11 -1.07 -13 0.31 1.32
-12 0.98 - 0.09 -12 0.59 1.91
-11 -0.53 -0.62 -11 -0.71 1.20
-10 - 0.07 -0.68 -10 -0.05 1.15
-9 1.05 0.37 -9 1.27 2.42
-8 0.08 0.45 -8 0.37 2.79
-7 -0.24 0.20 -7 0.09 2.88
-6 -0.58 -0.37 -6 -0.91 1.97
-5 -0.52 -0.90 -5 -0.91 1.06
-4 -0.50 - 1.40 -4 0.27 1.34
-3 -0.17 - 1.57 -3 -0.19 1.14
-2 0.51 - 1.06 -2 0.88 2.03
-1 -2.98 -4.04 -1 -2.81 -0.78
0 - 1.58 -5.62 0 - 1.27 - 2.05
1 0.17 - 5.45 1 0.36 ~ 1.68
2 -0.18 - 5.63 2 -0.43 -2.12
3 0.59 - 5.05 3 0.96 -1.16
4 -0.35 - 5.40 4 -0.16 -1.32
5 -1.04 - 6.43 5 ~ 1.26 -2.57
6 -0.89 -7.32 6 -0.35 -2.92
7 0.41 - 6.91 7 0.88 - 2.04
8 -0.25 -7.16 8 -0.82 -2.86
9 -0.37 -7.53 9 -0.06 -2.91
10 - 0.56 - 8.09 10 - 0.02 - 2.93
11 0.11 - 7.99 11 0.69 -2.24
12 - 0.48 - 8.46 12 0.05 -2.19
13 1.95 -6.51 13 0.03 -2.16
14 -0.66 -7.17 14 -0.97 -3.13
15 -0.17 - 7.34 15 -0.65 -3.78
290 L.Y. Dann and H. DeAngelo, Standstill agreements and stock repurchases

Table 3 (continued)

All standstill agreements Standstill agreements not

(including those accompanied accompanied by a
by a negotiated stock negotiated stock repurchase -
repurchase) - 30 companies 19 companies

‘1’ (2) (3) (4) (5) (6)

Trading Trading
day pe cpe day F cpe

16 - 0.40 -7.74 16 -0.83 -4.60

17 -0.79 -8.52 17 -1.18 - 5.79
18 -0.24 -8.71 18 -0.03 - 5.82
19 -0.63 -9.39 19 -0.77 - 6.60
20 0.23 -9.16 20 0.88 -5.71
25 -0.21 -9.36 25 -0.76 -6.64
30 -0.31 -11.01 30 -0.47 -8.62
35 -0.01 - 12.14 35 -0.26 -9.78
40 - 1.01 - 12.13 40 -0.80 -8.19

before and after the announcement date of the standstill agreement. Two
samples of firms undertaking a standstill agreement are reported in table 3.
In columns 1-3, the daily average prediction errors and cumulative average
prediction errors are portrayed for the sample of all standstills meeting the
sampling criteria set forth in section 3. This sample includes 11 firms
undertaking standstills accompanied by a negotiated stock repurchase. To
isolate the average effect of standstill agreements per se, prediction errors and
cumulative prediction errors for the subsample consisting of 19 firms
negotiating a standstill agreement not accompanied by a stock repurchase are
reported in columns 4-6.
To assess the average market price impact of standstill agreements, we
designate day - 1 and day 0 as the announcement period. For the sample of
all standstill firms, the average (equal-weighted portfolio) prediction error is
-2.98% on day - 1 and - 1.58% on day 0. The two-day portfolio prediction
error for the announcement period is -4.52%, and the median is -3.84%.
Negative announcement returns are also observed for the ‘pure’ (no
repurchase) sample of standstill firms. Portfolio prediction errors for day - 1
and day 0 are -2.81% and -1.27x, respectively, the two-day portfolio
prediction error is -4.04x, and the median prediction error is - 2.66%.
Statistical significance tests indicate that the announcement returns are
inconsistent with the hypothesis that standstill agreements are in the best
interests of non-participating stockholders. The t-statistics (described in
section 4) for the announcement period prediction errors are -5.72 for the
portfolio of all standstill firms and -4.49 for the sample of ‘pure’ standstill
firms, both of which are different from zero at the 0.01 level of significance.
These t-statistics are consistent with the prediction of the managerial
 L.Y Dunn and H. DeAngelo, Standstill agreements and stock repurchases 291

entrenchment hypothesis that, on average, standstill agreements reduce the

wealth of non-participating shareholders.
The Wilcoxon signed-ranks test also indicates that price impacts of
standstill agreements are negative. For the sample of all standstills, the
Wilcoxon test rejects at the 0.005 significance level the null hypothesis of
zero average abnormal returns. For the sample of ‘pure’ standstills the
Wilcoxon test is less definitive, allowing rejection of the null hypothesis at
only the 0.099 level of significance.
Table 4 provides additional support for the interpretation that the negative
average returns are representative of the impact of standstill agreements. The
table presents the frequency distribution of two-day announcement period
prediction errors for the ‘all’ and ‘pure’ standstill samples. The qualitative
picture suggested by these frequency distributions is in accord with the

Table 4
Distribution of two-day announcement period prediction errors for the sample of firms
undertaking a standstill agreement.

Number of observed prediction errors

Magnitude of All standstill agreements Standstill agreements not

two-day announcement (including those accompanied accompanied by a
period prediction by a negotiated stock negotiated stock repurchase -
error (PE,)” repurchase) ~ 30 events 19 events

- 227; < PE, < - 20% 1 1

-2O”‘<PE,< -18”/ 0 0
-18;<PE,<-16; 1 1
-16%<PE,<-14% 3 2
- 14%<PE,< - 12;/, 0 0
- 12%<PE,< - 10% 2 I
-lO%<PE,< -8M 5 1
-S%<PE,< -6% 0 0
-6%<PE,< -4% 2 1
-4%<PE,< -2x, 5 4
-2U/,<PE,< 0% 2 2
O”/,<PE,< 2% 2 0
2%<PE,< 4% 1 1
4%<PE,< 6% 2 2
6% < PE, < 8% 3 2
Number of events for
which no trading
occurred on day
- 1 or day 0 1 1

Total 30 19
Minimum value of PE, -21.84% -21.84%
Median value of PE, - 3.84% - 2.66%
Maximum value of Pi?, 7.03% 7.03%;;

“Here PE, = (1 + pe_ 1)(1+ pe,) - 1, where pe, is the single-day prediction error on event day 1.
292 L.Y Dann and H. DeAngelo, Standstill agreements and stock repurchases

statistical results presented above insofar as the distributions appear centered

below zero. For each of the two standstill samples, 72% of the firms had
negative announcement period prediction errors. Furthermore, while some
firms had positive estimated abnormal returns, the largest positive
observation is +7.03% and, for the ‘all’ standstill sample, there are twelve
negative return observations which are greater in absolute value. In sum, we
interpret the weight of the standstill evidence as indicating a negative average
wealth impact for non-participating stockholders and thus as consistent with
the managerial entrenchment hypothesis.
Within the current literature, our standstill findings seem most closely
related to results reported by Dodd (1980) in his recent investigation of
managerial responses to merger proposals. Dodd found a statistically
significant -5.57% average price decline for a sample of 26 merger proposals
which were vetoed by target management - i.e., which target management
did not approve and therefore chose not to pass on to stockholder vote (pp.
129-130). Merger proposals vetoed by management are qualitatively similar
to standstill agreements insofar as both represent unilateral actions taken by
incumbent management which reduce the likelihood of control transfer. Like
our standstill findings, Dodd’s managerial veto evidence is also consistent
with the hypothesis that some incumbent managements impose significant
wealth losses on stockholders when acting to deter a transfer of control.
A final point (not directly related to our tests of the managerial
entrenchment versus stockholder interests hypotheses) should be noted.
Beyond day 0, we observe from table 3 a sizeable downward drift in the
cumulative average prediction errors. In contrast, no comparable drift occurs
over the pre-announcement period. The cumulative average prediction error
-
(cpe) from day + 1 though day +40 is significantly different from zero at the -
0.10 level (two-tailed test) for the sample of all standstill firms, but the cpe
over the same time period for the sample of pure standstills is not significant
at the 0.10 level.23 At least for the sample of all standstill firms, this evidence
suggests the possibility that (i) additional negative news was disclosed
following the initial standstill announcement and that (ii) the expected value
of any such news was not fully capitalized at initial announcement. However,
an examination of the Wall Street Journal Index listings for these 30 firms
failed to detect any systematic commonalities or pattern of events in the

23The following t-test procedure was employed to assess the significance of the @Z. The
cumulative average prediction errors for days 1 through 40 are - 6.51% and -6.74% for the ‘all’
standstill and ‘pure’ standstill samples, respectively. Over this period, the estimated daily sample
standard deviations are 0.544% and 0.643’/“,, respectively. Assuming that the average prediction
errors for each sample over this time period are independent and identically distributed, then the
standard deviations of the sum of 40 average prediction errors are 3.441% and 4.067%
respectively. The t-values (39 degrees of freedom) for the post-announcement c? of the two
samples are - 1.89 and - 1.66, respectively. This t-test procedure is equivalent to testing whether
the mean daily average prediction error over this period is equal to zero.
 L.Y. Dann and H. DeAngelo, Standstill agreements and stock repurchases 293

immediate post-announcement period. Consequently, we cannot identify a

systematic determinant of these post-announcement negative returns.24

5.2. Market reaction to the first public announcement of a privately negotiated

premium repurchase

For stock repurchases privately negotiated at a premium relative to the

prevailing market price, daily average and cumulative average prediction
errors of returns to non-participating stockholders are presented in table 5.
The format of table 5 is similar to that of table 3. Columns 1-3 portray the’
event day, average prediction error (pe) and cumulative average prediction
-
error (cpe), respectively, for the sample of all firms repurchasing a portion of
their outstanding common shares at a (cash) premium in a privately
negotiated transaction. This sample includes 7 firms for which The Wall
Street Journal reported a standstill agreement with an accompanying cash
repurchase at a premium over market. Columns 4-6 present comparable data
for the subsample of firms undertaking a premium repurchase when The
Wall Street Journal did not report an accompanying standstill
agreement.25,26
To assess the premium repurchase evidence for the managerial
entrenchment versus stockholder interests hypotheses, we again designate day
- 1 and day 0 as the announcement period. Table 5 indicates that, for the
sample of all negotiated premium repurchases, the day - 1 and 0 portfolio
prediction errors are -0.72’://, and - l.OSo/,, respectively. The two-day
portfolio prediction error is - 1.76%, and the median prediction error is
- 1.59%. For the sample of negotiated premium repurchases without
standstills, the day - 1 and day 0 prediction errors are -0.16’% and - 1.00%
respectively, the two-day prediction error is - 1.16’1/,, and the median
prediction error is -0.54oj The t-statistic for the two-day portfolio

24We find virtually identical results for average prediction errors based upon beta estimates
obtained from the preannouncement period only. Thus, the explanation does not appear to lie
with possibly biased estimates of market model parameters.
“The possibility remains that some transactions in the ‘pure’ premium repurchase sample
were actually accompanied by standstill agreements which (i) were not publicly revealed with the
repurchase announcement and/or (ii) were not reported in The Wall Street Journal.
Consequently, the label ‘pure’ premium repurchase may not be a strictly accurate description of
all events in the subsample reported in columns 4 through 6 of table 5.
*‘In contrast with the time series of average prediction errors surrounding standstill
announcements, the negotiated premium repurchase samples do not exhibit significant post-
announcement negative returns. In the pre-announcement period (day -40 through day -2),
each repurchase sample has a negative c? in excess of 4% (absolute magnitude). For the sample
of all premium buybacks, the decline is significantly different from zero at the 0.10 level based
on the test procedure detailed in footnote 23. The pre-announcement performance of the pure
premium repurchase sample is not significant at the 0.10 level. We find no readily apparent
explanation for this negative pre-announcement performance, but we note that virtually all of
the negative c- occurs between days -40 and -20. Thus, it seems doubtful that the negative
performance is attributable to leakage of information about the negotiated buybacks.
294 L.Y. Dann and H. DeAngelo, Standstill agreements and stock repurchases

Table 5

Percentage daily average (p) and cumulative average (EjE) prediction errors for
the Il-day period centered around announcement of a privately negotiated stock
repurchase at a cash premium.”

All premium negotiated repurchases Premium negotiated repurchases

(including those accompanied by a not accompanied by a standstill
standstill agreement) - 41 events agreement - 34 events

(1) (2) (3) (4) (5) (6)

Trading Trading
day w qz day F cpe

-40 0.17 0.17 -40 0.59 0.59

-35 -0.46 - 1.12 -35 -0.39 -0.68
-30 -0.72 - 1.77 -30 -0.73 - 1.87
-25 0.3 1 -3.00 -25 0.45 -3.27
-20 - 0.09 -4.19 -20 -0.02 -4.00
-19 -0.26 -4.45 -19 -0.07 -4.07
-18 -0.34 -4.79 -18 -0.50 -4.57
- 17 0.07 -4.72 -17 0.16 -4.41
-16 0.30 -4.41 - 16 0.25 -4.16
-15 -0.41 -4.82 -15 -0.38 -4.55
-14 0.03 -4.80 -14 - 0.08 -4.62
-13 0.43 -4.37 -13 0.63 -3.99
- 12 I .02 -3.35 -12 0.76 -3.23
-11 -0.26 - 3.62 -11 - 0.05 -3.28
-10 0.66 - 2.96 -10 0.82 - 2.46
-9 -0.34 -3.30 -9 -0.92 - 3.38
-8 -0.26 - 3.57 -8 -0.19 - 3.57
-7 -0.79 -4.36 -7 -0.71 -4.28
-6 0.21 -4.14 -6 0.12 -4.16
-5 -0.15 -4.29 -5 -0.13 - 4.29
-4 -0.12 -4.41 -4 0.25 -4.04
-3 - 0.03 -4.44 -3 0.01 -4.03
-2 -0.31 -4.75 -2 -0.18 -4.22
-1 -0.72 - 5.47 -1 -0.16 -4.38
0 - 1.05 -6.52 0 -1.00 - 5.38
1 -0.60 -7.12 1 -0.82 - 6.20
2 0.10 - 7.01 2 0.39 - 5.82
3 0.06 - 6.96 3 - 5.84
4 -0.60 -7.56 4 -0.51 - 6.36
5 0.45 -7.11 5 0.72 - 5.64
6 -0.20 -7.31 6 -0.01 - 5.65
7 -0.31 - 7.62 7 -0.36 -6.01
8 0.11 -7.51 8 -0.01 - 6.02
9 -0.21 - 7.12 9 -0.13 -6.14
10 -0.19 -7.91 10 0.12 -6.02
11 -0.17 - 8.08 11 -0.22 - 6.24
12 -0.86 - 8.94 12 -0.53 - 6.77
13 1.15 - 7.19 13 0.12 - 6.66
14 -0.88 - 8.67 14 -1.08 - 7.74
15 - 0.47 -9.14 15 -0.56 - 8.30
16 -0.46 -9.60 16 -0.34 -8.63
17 - 0.06 -9.66 17 -0.15 -8.78
 L.Y Dam and H. DeAngelo, Standstill agreements and stock repurchases 295

Table 5 (continued)

All premium negotiated repurchases Premium negotiated repurchases

(including those accompanied by a not accompanied by a standstill
standstill agreement) - 41 events agreements - 34 events

(1) (2) (3) (4) (5) (6)

Trading Trading
day pe cpe day pe cpe

1X 0.14 -9.53 18 0.31 x.47

19 -0.09 -9.62 19 - 0.05 ~ 8.52
20 -0.57 - 10.18 20 -0.56 - 9.09
25 -0.07 - 10.15 25 -0.36 - 8.97
30 -0.64 ~ 11.07 30 -0.88 - 10.30
35 0.56 ~ 10.10 35 0.63 - 8.79
40 -0.16 - 10.22 40 - 0.01 -8.70

“Stock repurchase at a premium is defined here as the case where the negotiated
cash repurchase price per share exceeds the closing market price per share on
day -2 (dividend-adjusted).

prediction errors at announcement is -3.59 for the full sample of premium

buybacks, which is statistically different from zero at the 0.01 level of
significance. For the sample of ‘pure’ premium repurchases, the t-statistic for
the two-day prediction error is - 2.15, which is different from zero at the
0.05 significance level. Results of the signed-ranks test are less definitive,
allowing rejection of the null hypothesis at the 0.057 level for the sample of
all premium repurchases and only the 0.223 significance level for the ‘pure’
premium repurchase sample.
A qualitatively similar picture emerges from inspection of table 6 which
presents the frequency distribution of announcement period two-day
prediction errors for the sample of ‘all’ and ‘pure’ negotiated premium
repurchases. For each sample, approximately 60% of the individual firm two-
day prediction errors are negative. While both distributions appear to be
centered in the -2% to 0% range, the tendency toward negative returns is
not as strong as in the standstill distribution presented in table 4.
The principal conclusion we draw from our statistical tests is that the
negative average price impact associated with negotiated premium
repurchases is inconsistent with the prediction of the stockholder interests
hypothesis that these transactions benefit non-participating stockholders on
average. We base this conclusion on two factors. First, none of our tests
reveals any indication of a significantly positive average wealth impact
associated with premium repurchases. Second, while the negative returns for
‘pure’ premium repurchases are significant at the 0.05 level under the t-test,
they are not different from zero at any conventional level under the
Wilcoxon signed-ranks test. Consequently, our conservative interpretation is
that these repurchase findings principally provide evidence inconsistent with
296 L.Y. Dann and H. DeAngelo, Standstill agreements and stock repurchases

Table 6
Distribution of two-day announcement period prediction errors for the sample of privately
negotiated stock repurchases at a premium.”

Number of observed prediction errors

Magnitude of All premium negotiated Premium negotiated

two-day announcement repurchases (including those repurchases not accompanied
period prediction accompanied by a standstill by a standstill
error (PE$ agreement) - 41 events agreement ~ 34 events

-18%<PE,< -16% I
-16%<PE,<-14% 0
-14”/,<PE,< -12%
-12%<PE,<-10%
-lO%<PE,< -8%
-8%<PE,< -6%
-6%<PE,< -4%
-4%<PE,< -2%
-2%<PE,< 0%
O%<PE,< 2% 4 3
2%<PE,< 4% 6 6
4%<PE,< 6% 4 4
6%<PE,< 8% I 0
Number of events
for which no trading
occurred on day
- 1 or day 0 2 2

Total 41 34

Minimum value of PE, - 16.54% - 16.54%

Median value of PE, - 1.59% -0.54%
Maximum value of PE, 6.75% 4.80%

“Stock repurchase at a premium is defined here as the case where the negotiated cash
repurchase price per share exceeds the closing market price per share on day -2 (dividend-
adjusted).
“Here P&,=(1 +pe_,)(l +pe,)- 1, where pe, is the single-day prediction error on event day t.

the stockholder interests hypothesis and, at best, weak support for the
managerial entrenchment view.
Support for the entrenchment hypothesis must be tempered further when
the results of the following test are considered. In particular, recall that the
managerial entrenchment hypothesis predicts not only a wealth decrease for
non-participating stockholders, but also a decline in total equity value due to
greater potential for managerial inefficiency after a negotiated premium
buyback. We tested this prediction using the following approach which, for
reasons to be explained momentarily, should be interpreted with caution. For
each firm in the sample of ‘pure’ premium repurchases, we estimated the
percentage change in total equity value by taking the sum of (i) the
percentage premium paid to the block‘ holder, TC, times the fraction of
 L.Y. Dunn and H. DeAngelo, Standstill agreements and stock repurchases 297

outstanding shares repurchased, IX,and (ii) the two-day announcement period

rate of return, R,, times the fraction of outstanding shares not repurchased,
(1 -a). ” The cross-sectional mean change in total equity value [rnx+
R,( 1 -II)] is 0.38% and the median is 0.89%. Under the Wilcoxon signed-ranks
test, the median change is significantly different from zero only at the 0.323
level. The t-test which, for tractability, utilizes the two-day prediction error
standard deviation as the dispersion estimate, implies a r-value of 0.70 which
is also insignificant at conventional levels.28,29 Thus, on average, the sum of
the value changes for the selling stockholder and the non-participating
stockholders is negligibly different from zero. These results do not support
the prediction of the entrenchment hypothesis that total equity value will
decline at the time of a negotiated premium repurchase.30
However, these results should be interpreted with caution since there is
reason to believe that our measure of the total change in equity value is
biased upward. The basic problem is that it is not possible to observe
directly the unexpected change in value of the repurchased holding. In
principle, this value change is equal to the difference between (1) the total
payment for shares repurchased (which is observable) and (2) the
pre-repurchase value of the block (which we assert is not observable). The
pre-repurchase value (2) should capitalize the expected portion of the
differential treatment accorded the blockholder and hence, with rational
expectations pricing, the open market price understates the (per share) value
27Note that (ii) presumes that all shares not repurchased are properly valued at open market
prices.
28More precisely the test statistic is the mean change in total equity value at announcement
(0.38%) divided by’the time series estimate of dispersion in prediction errors based on open-
market CRSP returns (d=0.54). This test statistic assumes that the dispersion in open-market
returns appropriately reflects the dispersion in total equity returns, In principle, the dispersion in
the percentage change in total equity value depends on the dispersion in returns to the large
block stockholder, the dispersion in returns to other stockholders, and the correlation between
these returns. The dispersion of returns to other stockholders is reasonably estimated from a
time series of observed open-market returns. However, since we cannot observe rates of return
to large block stockholders, we do not have time series data to estimate either (1) the dispersion
in large block stockholder returns or (2) the correlation in returns to large block and other
stockholders. The test statistic’s dispersion estimate may be biased upward or downward,
depending upon the (unobservable) size of (1) and (2) relative to the dispersion in open market
returns.
*9Virtually identical results are obtained when the prediction error PE, is substituted for R,
as an estimate of the (unexpected) return to non-participating stockholders.
“Nor does the data support the view that pure premium repurchases represent simple wealth
transfers from non-participating stockholders to the selling blockholder. If this view were
correct, we should observe a significant negative cross-sectional relationship between KU and
R,(l -G(). However, both linear regression and rank-order correlation tests reveal no significant
relationship between Z(G(and R,(l -_Go at any conventional significance level.
However, these simple tests do reveal a significantly positive (at the 0.05 level) cross-sectional
relationship between I[, the repurchase premium paid, and a, the fraction of shares repurchased.
In very loose terms, this finding suggests a relationship between blockholder ‘power’ as proxied
by fractional ownership and premium extraction capability (see footnote 6). We view this as pure
conjecture since we cannot, at this point, specify the theoretical conditions under which such a
cross-sectional relationship should obtain.
298 L.Y Dann and H. DeAngelo, Standstill agreements and stock repurchases

of the large block holding prior to repurchase.31 Consequently, in contrast to

the observed (CRSP-reported open-market) announcement period return, R,,
which estimates the unexpected wealth change of non-participating
stockholders, the repurchase premium, X, is an over-estimate of the
unexpected wealth change of the selling blockholder. The important point for
this test is that our estimate of the total unexpected change in equity value
[ncr + R,(l -a)] is biased upward and hence the test includes a bias against
the entrenchment hypothesis3’
While our repurchase findings are not conclusive with respect to the
managerial entrenchment hypothesis, an additional stronger conclusion can
be drawn from our evidence: the average equity value impact of negotiated
buybacks is quite different from that of repurchase tender offers and open
market repurchases. For tender offer repurchases Dann (1980, 1981), Masulis
(1980b), and Vermaelen (1981) report average common stock returns of
approximately 16% at announcement. Furthermore, a positive price impact
persists, on average, after tender offer expiration. For open-market
repurchases, Dann and Vermaelen report average stockholder returns of 3%
at announcement. Moreover, a positive impact is observed for firms
announcing the completion of (a previously unannounced) open-market
repurchase. For both tender offer and open-market repurchases, the
announcement date returns are all statistically positive at the 0.05 (and in
most cases, the 0.01) level of significance, and contrast markedly with the
negative announcement date returns we find for privately negotiated
premium repurchases.
Furthermore, even in the case where no premium is paid to the large block
stockholder, privately negotiated repurchases do not appear to have positive
valuation impacts for non-participating stockholders. In 17 instances, our
sampling procedure revealed privately negotiated stock repurchases occurring
at a cash price at or below the closing market price on (event) day -2. For
this sample of 17 events, the two-day portfolio prediction error is -0.34x
with a t-statistic of -0.33, which is not statistically different from zero at
conventional levels of significance. Again, these findings for negotiated
buybacks differ markedly from the significantly positive price impacts
observed in other types of repurchases.

31The frequency distribution in table 5 indicates that, for the sample of ‘pure’ premium
repurchases, the largest observed abnormal return to non-participating stockholders is +4.80X.
Table 2 reveals that the average selling (participating) stockholder received a premium of 14.4%
above the pre-announcement open-market price. Thus, it is difftcult to argue that participating
and non-participating stockholders are receiving even approximately homogeneous treatment in
these transactions. For related evidence of premium pricing of control blocks in closely held
firms, see Meeker and Joy (1980).
3ZThe more general (and radical) implication is that there are situations where it is
inappropriate to value a block of shares at the open market price which presumably represents
the per share value of dispersed holdings. The issue of the extent of differential pricing of share
blocks is quite clearly an important area for future research.
 L.Y. Dann and H. DeAngelo. Standstill aqeements and stock repurchases 299

Perhaps the most striking difference is the one between negotiated

premium buybacks and repurchase tender offers. The descriptive statistics
presented in section 3 indicate that both forms of repurchase are qualitatively
similar insofar as, on average, both involve (1) the purchase of substantial
equity share fractions and (2) the payment of substantial premiums to selling
stockholders. These structural similarities indicate that both forms of
repurchase involve attributes considered in the earlier cited repurchase
literature. In particular, both involve potential corporate tax effects via
material leverage increases and positive information signals in the form of
management’s demonstrated willingness to pay an above market price for
shares [see Vermaelen (1981, p. 140)].33 Two principal structural differences
are (1) a large block holding is removed in a negotiated repurchase whereas
no such block removal is necessarily present in a tender offer and (2)
dispersed stockholders are not eligible to receive a premium in a negotiated
buyback as they are in a tender offer. The factors underlying these two
structural differences may ultimately help to explain the substantially
different equity value impacts associated with different types of stock
repurchase.

6. Summary

Standstill agreements and negotiated premium repurchases are transactions

between a corporation and a substantial stockholder which (i) reduce
competition for control over the firm’s resources and which (ii) provide for
differential treatment of the large block stockholder. In this paper, we
examined two alternative explanations for standstills and premium buybacks.
According to the managerial entrenchment hypothesis, the dominant effect of
standstill agreements and negotiated premium repurchases is that incumbent
management removes a credible threat of control transfer and thereby
obtains job protection at the expense of non-participating stockholders. The
competing stockholder interests hypothesis predicts that non-participating
stockholders gain from these transactions because they lead to a more
efficient (less costly) process of competition for control of the firm.
We tested the competing hypotheses by examining the common stock price
impact associated with standstill agreements and negotiated premium
repurchases for a sample of firms engaging in these transactions during the
period 1977-1980. Standstill agreements were associated with statistically
significant negative average returns to non-participating stockholders.
Negotiated premium repurchases were associated with negative, but less
significant, average returns to non-participating stockholders. We interpret
the standstill and premium repurchase evidence as inconsistent with the
33To the extent they are present, any such tax or information effects (or, for that matter,
bondholder expropriation effects) will impart an upward bias to the stock return at
announcement and thus will tend to mask any managerial entrenchment effect associated with
negotiated premium repurchases.
300 L.Y. Dam and H. DeAngelo, Standstill agreements and stock repurchases

stockholder interests hypothesis and the standstill evidence as providing

support for the entrenchment view.
In sum, while the issue is far from settled, we interpret our findings as
raising a serious question about the extent to which some target
management actions in the market for corporate control are in the best
interests of all stockholders. This, in turn, raises a question regarding the
extent to which the market for corporate control and the managerial labor
market discipline incumbent management to maximize all stockholders’
wealth.

Bradley, M., 1980, Intertirm tender offers and the market for corporate control, Journal of
Business, Oct., 345-376.
Brown, S.J. and J.B. Warner, 1980, Measuring security price performance, Journal of Financial
Economics, Sept., 205-258.
Cary, W., 1969970, Corporate devices used to insulate management from attack, Antitrust Law
Journal 39, 318-333.
Dann, L.Y., 1980, The effect of common stock repurchase on securityholder returns, Ph.D.
dissertation (University of California, Los Angeles, CA).
Dann, L.Y., 1981, Common stock repurchases: An analysis of the returns to stockholders and
bondholders, Journal of Financial Economics, June, 113-l 38.
DeAngelo, H. and E.M. Rice, 1983, Antitakeover charter amendments and stockholder wealth,
Journal of Financial Economics 11, this issue.
Dodd, P., 1980, Merger proposals, management discretion and stockholder wealth, Journal of
Financial Economics, June, 1055137.
Easterbrook, F. and D. Fischel, 1982, Corporate control transactions, Yale Law Journal,
forthcoming.
Fama, E., 1976, Foundations of finance (Basic Books, New York).
Fama, E., 1980, Agency problems and the theory of the firm, Journal of Political Economy,
April, 2888307.
Fama, E., L. Fisher, M. Jensen and R. Roll, 1969, The adjustment of stock prices to new
information, International Economic Review, Feb., 1-21.
Grossman, S.J. and 0. Hart, 1980, Takeover bids, the free-rider problem, and the theory of the
corporation, Bell Journal of Economics, Spring, 42-64.
Jarrell, G.A. and M. Bradley, 1980, The economic effects of federal and state regulations of cash
tender offers, Journal of Law and Economics, April, 371-407.
Manne, H., 1965, Mergers and the market for corporate control, Journal of Political Economy,
April, 1 lG120.
Masulis, R.W., 1980a, The effects of capital structure change on security prices: A study of
exchange offers, Journal of Financial Economics, June, 139-l 77.
Masulis,R.W., 1980b, Stock repurchase by tender offer: An analysis of the cause of common
stock price increases, Journal of Finance, May 305-319.
Meeker, L. and 0. Joy, 1980, Price premiums for shares of closely held stock, Journal of
Business, July, 297-314.
Nathan, C.M. and M. Sobel, 1980, Corporate stock repurchases in the context of unsolicited
takeover bids, Business Lawyer, July, 1545-1566.
Standard and Poor’s Daily Stock Price Record.
Standard and Poor’s Security Owners Stock Guide.
The Wall Street Journal.
The Wall Street Journal Index.
Vermaelen, T., 1981, Common stock repurchases and market signalling, Journal of Financial
Economics, June, 139-l 83.
Williamson, O., 1975, Markets and hierarchies: Analysis and antitrust implications (The Free
Press, New York).