Wealth reduction in white knight bids.
The existing literature on takeover activity is inconsistent regarding the extent and direction of change in the value of the bidding firm at the time of the bid.(1) A number of empirical studies document the negative investor response to bidders upon announcement of opening bids. Yet, a wide variation in the evidence exists, and seems to be based on sample and time interval considered.(2) This variation may imply that there are subsets of bidders exhibiting considerable within-group homogeneity, but which differ from other subsets in the pattern of abnormal returns during takeover bids. If a number of such subsets exist with varying abnormal return patterns, an overall description of bidders is nothing more than a weighted average of the heterogeneous collection of bidders from different subsets. Some such subsets have indeed been identified by researchers. Mitchell and Lehn |15~ ascribe the negative market reaction to "bad" bidders: those who subsequently become hostile takeover targets. Travlos |22~, and Amihud, Lev and Travlos |1~ show that negative bidder returns are caused by stock financing of acquisitions and low managerial ownership of bidder's stock.
We look at a subset of bidders, called white knights. Our study investigates:
(i) Equity value changes for all three participants in a corporate control contest involving white knights over a number of variable length intervals of transaction-specific length during the contest; and
(ii) The differing welfare implications of a subset of corporate control contests where a participating market group is likely to lose.
The paper is organized in the following manner. Section I defines white knight bidding and hypothesizes that the nature of their involvement in corporate control contests might lead to negative valuation consequences for this subset of bidders. Section II provides an overview of the related literature, and Section III considers the welfare implications of multibidder corporate control contests. The sample selection procedures, data sources and methodology for the analysis are considered in Section IV. The results are presented and interpreted in Section V. The final section includes a summary and conclusion.
I. White Knight Contests
In a corporate control contest for the acquisition of a target, the white knight bid is different from those made by hostile bidders. As defined in this study, the white knight bid has the following three characteristics:
(i) It is a subsequent bid.
(ii) It is a friendly bid.
(iii) It follows a hostile bid.
The above features can imply that the white knight bid is made in relative haste in order to gain control of the target. As Schipper and Thompson |19~ indicate, bidding firms make acquisition plans long before their actual bid; and, the market responds to the plan rather than the bid. However, evidence exists that the market also responds to the specific bid. The average time interval between the first hostile bid and the white knight bid is only 27 trading days. If the white knight and the target had prior plans of merging their operations in a friendly manner, at the very least, the white knight firm would probably have revealed its acquisition plans to investors well in advance, without specifying a particular target. We can then hypothesize that in the absence of any such prior indication, the market may reasonably assume that the white knight enters the corporate control contest in an unplanned and hasty manner, even though it may have the advantage of placing a better and more accurate value on the target due to the possible high degree of cooperation from the target firm in divulging detailed internal information. If the market places a greater value on corporate strategic planning for acquisitions and extensive preparatory time, than it does on cooperation between the white knight and the target, the market is apt to respond negatively to the white knight bid, seeing these bids as likely to represent overpayment.(3) Further, by increasing their size through a quick acquisition, the white knight firms may become less susceptible potential targets for future acquisitions by other hostile bidders. This prospective antitakeover strategy of acquisitions by white knights then becomes consistent with the immediate anti-takeover strategy of the target firm of inviting a white knight to fend off the unwelcome hostile bidder. Once again, the market response to such an interpretation of bidder behavior is liable to be negative.(4) Any perception of noncommercial motivations (e.g., emanating out of personal affiliation between target and white knight managers) might raise further questions regarding the value implications for white knight shareholders. The white knights could thus comprise a largely homogeneous subset of bidders in the sense described earlier.
The above perspectives motivate an examination of the dollar value changes associated with white knight bids to assess whether the amounts are consequential in the calculation of overall synergistic dollar gains in corporate control contests. This is especially relevant in the context of the existing literature on synergistic gains in takeovers, which indicates that low absolute percentage gains for bidders can translate to large dollar gains, because bidders are usually much larger than targets.(5) Large dollar losses for white knights may point to the need for a reappraisal of synergistic gains from the subset of takeover activity which involves the white knight as one of the bidders. We also consider the welfare consequences of these corporate control contests.
II. A Review of the Literature
A. Dollar Losses/Gains
The research on dollar gains and losses for the participants of a corporate control contest has not been as extensive as that on percentage gains and losses for the relevant market groups during the acquisition. Malatesta |14~ estimated the dollar value of abnormal returns to the common stocks of acquiring and acquired firms in mergers. Over a four-month interval prior to and including the month of board approval of the merger, he found that the average dollar gain to stockholders of acquired companies was significant at $19.67 million, and the dollar loss to stockholders of acquiring companies was significant at $27.65 million. Dennis and McConnell |5~ studied the dollar gains to acquiring and acquired companies, where both the companies had traded senior securities. Over a 40-day interval equally divided before and after the merger announcement date, the stockholders of the acquiring company gained, on the average, $34.0 million under the dollar loss due to price fall method (DLPF) and $52.4 million under the market-adjusted abnormal return method (MAAR).(6) The stockholders of the acquired companies gained, on the average, $28.5 million and $31.8 million under DLPF and MAAR, respectively.(7) Bradley, Desai and Kim |4~ studied successful tender offers (including multiple bids). They constructed a variable interval window from five days prior to the first bid to five days after the last bid, and found that acquiring companies gained an average of $17.30 million, whereas target (acquired) companies gained an average of $107.08 million. For the period 1981 to 1984, however, the acquiring companies lost, over the defined interval, an average of $27.28 million. During the same period, the acquired targets gained, on average, $233.53 million which was more than a 200% increase over previous years. Black and Grundfest |3~ estimated the aggregate dollar gains to targets to amount to $134.4 billion for the period 1983 to 1986.
Not surprisingly, the evidence of dollar gains to targets is overwhelming. The gains to bidders are less clear and vary with the sample examined. There is some indication of dollar losses to acquiring firm stockholders in the 1980s, and over extended intervals covering the entire range of control contests.
B. White Knights
Smiley and Stewart |21~ looked at a sample of white knights tender offers that involved "white knights" for the period 1972 to 1978. By inference, their "white knights" seem to be second bidders who were inducted by investment bankers into the control contest. Based on an analysis with monthly stock returns, they did not find any significant abnormal returns associated with the white knight bids. Niden |16~ looks at white knight corporate takeovers for the period 1974 to 1984, with a view of ascertaining whether these takeovers result in overpayment or synergy. She investigates security price reaction involved in these takeovers over event intervals of (-1 from date of first bid, date of last bid) and (-41 from date of first bid, date of last bid) for bidders and targets respectively. Her definition of white knights is substantially similar to ours, and she documents a significant market reaction of -3.6% to white knight bids over her defined bidder interval indicated above.
III. Welfare Implications of Control Contests
In terms of the criteria for welfare gain, a Pareto improvement is a change from which at least one person gains and nobody loses. The Kaldor-Hicks (K-H) compensation principle considers a change as an improvement where the gainers have the potential for compensating the losers while preserving a part of the gain for themselves, such that everyone can be made better off; and the losers are not able to bribe the gainers not to undertake the change (Scitovsky reversal test). The K-H principle is weaker than the Pareto principle since it stipulates no potential losers instead of no actual losers. Unlike the Pareto criterion, which considers both efficiency and equity to be a requirement of welfare gain, the compensation principle deals with the issue of efficiency only and leaves the issue of equity (actual payment of compensation to the losers) to the policymakers. In this manner, the K-H principle, by focusing on the creation of gain and ignoring its distribution, can enable a comparison to be made between a significantly larger number of states than could be done with the Pareto principle.(8)
An adequate analysis of the effects of corporate control contests, in terms of the above welfare concepts, should determine its impact on each of the market groups.(9) Generally, such two-party contests represent Pareto improvements in the sense that at least one market group (target firms) is better off and, on average, no group (bidding firms) is materially worse off. Further, societal benefit, in terms of both efficiency and equity, is derived from the fact that the postcontest state is a Pareto superior state. However, we have reasoned that in three-party contests (i.e., target firm, hostile bidder and white knight), the white knights might experience value decreases. In such a situation, the state represented by the stage of the contest just prior to the entry of the white knight seems to be Pareto optimal since movement from this state does not appear to be feasible without making at least one of the market groups (white knights) worse off. In addition, the friendly nature of the white knight multiple bid will probably have the economic connotation of resulting in a socially undesirable outcome on acquisition. Thus, the entry of the white knight into the corporate control contest has negative welfare implications. In any case, the precontest state and the postcontest state are Pareto noncomparable since one of the participating market groups (white knights) likely loses in the movement between the two states.
The possible failure of the Pareto criterion to evaluate the welfare consequences of control contests involving white knights would require an appraisal of the contest through the weaker criterion of the K-H principle. If we find empirically that the gainers in the contest could potentially compensate the losers so that all market groups would be better off, these contests could be termed as efficient and have a weaker, but positive, welfare implication.
IV. Sample, Data and Methodology
The sample sources for the study are the Mergerstat Review and the Dow Jones News Retrieval Service,(10) which provided a universe of 100 white knights for the ten-year period 1978-1987. Our definition of the white knight required the firm to fulfill both of the following criteria:
(i) It must make a subsequent bid to acquire the target, or to acquire a substantial nonmajority interest in the target after: (a) one or more hostile bids; or (b) initiation of a proxy contest; or (c) acquisition of an equity position with hostile intent.
(ii) There must be evidence of "collaboration" between the white knight and the target prior to the white knight bid. This required an explicit reference in the responsible financial press to the bidder as a white knight.
In the financial press, the term "white knight" is used somewhat imprecisely, and some of the circumstances do not meet the "late-entry, collaborative" criteria above. For example, the Mergerstat Review includes within their list of "white knights" those firms which preemptively take over a target firm expecting a hostile bid. In such situations, our definition of a white knight is not met since the white knight bid is not subsequent. In addition, both of the sample sources include "white knights" which are subsequent bidders, but there is no indication of collaboration. Such firms were identified through scrutiny of the Wall Street Journal Index and relevant articles in the Wall Street Journal, and then deleted, providing a residual of 62 white knight firms of which 57 were exchange-listed. Forty-seven white knight firms were successful in acquiring their targets, ten were not. The market value of equity of the white knight firms ranged from less than $100 million to more than $5 billion.
In addition to examining the wealth effects for the white knights, we sought to estimate the wealth effects for the other market groups participating in the corporate control contests involving white knights to examine the welfare consequences of the contest using the Kaldor-Hicks compensation principle. For this, we needed matched-trios (white knights, hostile bidder, target) where data were available for all parties. There were 33 such matched-trios where all the firms in each individual trio were exchange-listed at the time of the contest.(11)
We identified three sequential event dates of interest in the corporate control contests involving white knights, and we have denoted these as follows:
|D.sub.1~ = date of first bid by hostile bidder.
|D.sub.2~ = date of first bid by white knight.
|D.sub.3~ = date of exit of hostile bidder or unsuccessful white knight.
1. Abnormal Returns
The event time methodology is used for calculating the cumulative abnormal returns (CARs) from the CRSP daily return files, and follows the methodology described in Dodd and Warner |6~, and Hite and Owers |7~. The test statistic Z, described by Dodd and Warner |6~, is the mean standardized cumulative abnormal return. To compute this statistic, the abnormal return A|R.sub.jt~ is standardized by its estimated standard deviation |S.sub.jt~,(12) to give
SA|R.sub.jt~ = A|R.sub.jt~/|S.sub.jt~. (1)
Each SA|R.sub.jt~ is assumed to be distributed unit normal in the absence of abnormal performance. Under this assumption, the test statistic Z(13) for a sample of N securities is also unit normal.
2. Valuation Consequences
To get a measure of dollar losses, we follow the methodology adopted by Dennis and McConnell |5~. The relevant window for abnormal return analysis is from |T.sub.lj~ to |T.sub.2j~, and that interval is defined to have a length of n (= |T.sub.2j~ - |T.sub.1j~ + 1) days. The dollar amounts of such abnormal returns can then be ascertained both in terms of n day market-adjusted cumulative abnormal returns (MAAR), as well as dollar change due to price fall in absolute terms over the n day window (DLPF). For each firm, the dollar loss is thus calculated under each procedure as follows:
Dollar loss = n day CAR x price on day (|T.sub.1~-1) x # of shares.(2)
Dollar loss = n day price change x # of shares.(3)
The firm-specific losses under MAAR are then summed up to obtain aggregate dollar losses under the respective methods; the same goes for DLPF.
A. Abnormal Returns for White Knights
Exhibit 1 reports the results of the event study around |D.sub.2~, the date of the white knight bid. The results indicate that the average excess return is negative at t = -1 (AR = -1.59%) and t = 0 (AR = -1.72%), as expected. The t-statistics for t = -1 and t = 0 are significant at the 0.01 level, thereby rejecting the null hypothesis that the abnormal returns on the event date are zero. The Z-scores for the CARs for intervals from t = +1 to t = +5 and t = +1 to t = +10 are insignificant, as are the t-statistics for ARs prior to t = -1. This shows that:
(i) There appears to be little, if any, leakage of information regarding the identity of the white knight.
(ii) There is no recovery of the losses immediately after the white knight bid, with the average CAR being insignificant at +0.3% from (|D.sub.2~ + 1) to (|D.sub.2~ + 10).
Further strong evidence of negative returns for the white knight is apparent from the division of positive and negative returns for the different intervals, as shown in Exhibit 2. Over the time interval (|D.sub.2~ - 1, |D.sub.2~), 78.95% of the CARs among firms are negative, and 40.35% of the CARs are negative and significant at p |is less than~ 0.05. In contrast, none of the 21.05% positive CARs are significant at p |is less than~ 0.05. We also note that over the entire period of the contest, a significant average negative CAR for the portfolio appears only in time intervals which contain the two-day window (|D.sub.2~ - 1, |D.sub.2~). Additionally, these longer intervals show significant negative average CARs for the portfolio larger than the CAR for the two-day window ending on |D.sub.2~, implying that there is no prior gain or subsequent recovery of the losses incurred between |D.sub.2~ - 1 and |D.sub.2.~.(14) Statistically, we can thus assert that shareholders of white knight firms do suffer wealth reductions on announcement of the white knight bid, which is not a loss of prior gains, and is not recovered up to the completion of the contest. These findings enable us to focus primarily on the two-day window from |D.sub.2~ - 1 to |D.sub.2~.
Exhibit 1. Daily Average Market-Adjusted Abnormal Returns (AR), Cumulative Sum of the Daily Average Abnormal Return (CAR'), Mean Cumulative Abnormal Return (CAR) and Test Statistics for Specified Intervals From 10 Days Before to 10 Days After the First White Knight Bid |D.sub.2~ for 57 Exchange-Listed White Knights During the Period 1978-1987 Panel A Date in t-statistic Event Time AR (%) CAR (%) for AR -10 -9 0.08 0.08 0.25 -8 -0.02 0.06 -0.07 -7 0.05 0.11 0.15 -6 0.13 0.26 0.39 -5 -0.03 0.23 -0.10 -4 0.11 0.34 0.32 -3 0.06 0.40 0.18 -2 -0.18 0.22 -0.54 -1 -1.59 -1.37 -5.41(*) 0(|D.sub.2~) -1.72 -3.09 -5.97(*) Panel B Period in Event Time CAR (%) z-statistic -1 to 0 -3.31 -11.75(*) +1 to +5 -0.11 -0.12 +1 to +10 0.33 0.52 Note: *p |is less than~ 0.01.
B. Outcomes of White Knight Contests
It is reasonable to conjecture that the negative CAR profile for white knights could be driven by the unsuccessful white knights. Perhaps these white knights, on observing the negative market reaction to their bid, were motivated to withdraw from the contest. A positive market reaction might then be observed around the date of withdrawal of the unsuccessful white knight. To test for the empirical validity of this proposition, we investigated the subsamples of successful and unsuccessful white knights. There were no consequential differences between the CAR profiles of the successful and unsuccessful white knights, even when the unsuccessful white knight withdrew from the contest. For the two-day interval ending on |D.sub.2~, the date of the white knight bid, the successful white knights had an average CAR of -3.2% (p |is less than~ 0.01), and the unsuccessful white knights had an average CAR of -3.5% (p |is less than~ 0.01). For the unsuccessful white knights, there was no recovery of losses at any time after their first bid.
Exhibit 2. Mean Cumulative Daily Abnormal Returns (CAR) for 57 Exchange-Listed White Knights for the Period 1978-1987 for Various Specified Intervals Relative to the Date of the First Hostile Bid |D.sub.1~, the Date of the First White Knight Bid |D.sub.2~, and the Date of the Hostile Bidder or White Knight Exit |D.sub.3~ Intervals CAR (%.) % Negative CAR Ending With the First Hostile Bid |D.sub.1~ (|D.sub.1~ - 50, |D.sub.1~) -1.9 54.39 (|D.sub.1~ - 4, |D.sub.1~) 0.7 50.88 (|D.sub.1~ - 1, |D.sub.1~) -0.2 52.63 Ending With the First White Knight Bid |D.sub.2~ (|D.sub.1~ - 50, |D.sub.2~) -4.5(**) 68.42(*) (|D.sub.1~ + 1, |D.sub.2~) -2.6(*) 68.42(*) (|D.sub.2~ - 4, |D.sub.2~) -3.0(*) 75.44(*) (|D.sub.2~ - 1, |D.sub.2~) -3.3(*) 78.95(*) Ending With the Hostile Bidder/White Knight Exit |D.sub.3~ (|D.sub.1~ - 50, |D.sub.3~) -5.7(*) 71.18(*) (|D.sub.1~ + 1, |D.sub.3~) -3.8(*) 71.18(*) (|D.sub.2~ - 1, |D.sub.3~) -4.2(*) 75.44(*) (|D.sub.2~ + 1, |D.sub.3~) -1.1 48.98 Notes: *p |is less than~ 0.01. **p |is less than~ 0.05.
The small subsample of ten contests involving unsuccessful white knights had diverse outcomes after the first white knight bid. In four contests, the hostile bidder prevailed by outbidding the white knight. In six contests, the target was subsequently acquired by a third bidder. In these six contests, the ultimate acquirer was another white knight in two instances and a management-led buyout team in a third instance. In all the control contests, the target finally ceased to exist as an independent entity irrespective of whether the white knight prevailed or not.
Exhibit 3. Dollar Losses in Time Interval (|D.sub.2~ - 1, |D.sub.2~) for 57 Exchange-Listed White Knights, 1978-1987 MAAR DLPF ($/million) ($/million) Total number of firms 57 57 Two-day dollar loss -4606.0(*) -4290.0(*) Mean -80.8 -75.3 Median -17.7 -22.9 Standard deviation 218.3 216.2 Maximum 735.0 765.0 Minimum -1003.5 -984.6 Note: *p |is less than~ 0.01.
C. Dollar Losses to White Knights
The predominant importance of the interval |D.sub.2~ - 1 to |D.sub.2~ for the white knight was established when reviewing patterns of abnormal return accumulation. This enables us to use |T.sub.1~ = -1 (relative to |D.sub.2~) in the methodology outlined in Section IV.B to arrive at the estimates of dollar losses by white knights. These are shown in Exhibit 3. It reports that white knights suffer an aggregate loss of $4.606 billion under MAAR ($4.290 billion under DLPF) over the two-day interval ending on |D.sub.2~. The average dollar loss for this period is $80.8 million under MAAR and $75.3 million under DLPF. Using the MAAR metric, 77% of the firms in our sample suffer wealth reductions aggregating $5.5 billion, and 23% of the firms gain $0.9 billion under MAAR.
The evidence of white knight loss of value is further strengthened by the fact that the firm having the largest maximum wealth reduction over this interval accounts for only 18% of the aggregate wealth reductions; being $1,003,500,000 out of a total of $5,476,500,000 of wealth losses suffered by the 45 white knights with negative outcomes.(15) In contrast, the firm having the highest wealth increase comprises 85% of the aggregate wealth gains; being $735,000,000 out of a total of $870,800,000 realized by the 12 white knights with positive outcomes.(16) These additional distributional details emphasize the finding that losses are typical and gains are unusual. Omission of these two extreme observations for positive and negative outcome white knights does not cause any material change in the results, with the aggregate two-day dollar loss and the average dollar loss per firm decreasing by 5.8% and 2.4%, respectively, under MAAR. The nonparametric binomial test continues to yield significance at p |is less than~ 0.01. It is noteworthy that of the 12 firms that show dollar gains, seven white knights follow a different bidding sequence by entering the bidding process after two successive bids by the hostile bidder, and two white knights are unsuccessful hostile bidders in prior contests involving other white knights, presumably with acquisition plans awaiting implementation. For the firms having negative dollar returns, there are no white knights who are unsuccessful hostile bidders in prior control contests involving white knights.
In the light of this evidence, we conclude that, in general, white knight firms suffer significant wealth reductions while making bids for targets. This feature also makes it impossible for us to posit welfare gains from the contest in terms of the Pareto criterion, and requires an assessment of the valuation consequences on all the participating market groups for an application of the Kaldor-Hicks compensation principle.
D. Subsequent White Knight Bids
It is further observed that some white knights bid for the target more than once (after subsequent hostile bids), until they finally acquire the target. Though the number of such white knights is small (only nine such contests), they suffer further wealth reductions through their subsequent bids, with a significant CAR of -7.6% from |D.sub.2~ + 1 to |D.sub.3~, the date of withdrawal of the hostile bidder.(17) In contrast, no further significant changes were observed after |D.sub.2~, the date of the first white knight bid, for white knights who bid only once. The further losses to white knights who bid more than once amount to $0.9 billion under MAAR ($1.3 billion under DLPF). The aggregate wealth reductions to white knights would thus amount to $5.5 billion under the MAAR metric. Repeated bidding by white knights may have enabled them to prevail, but they did not "win" in terms of value creation.
TABULAR DATA OMITTED
E. Valuation Impact for All Participants: Matched-Trios
When assessing the dollar changes in value of the various participants as a result of the contest, only those variable time intervals which had a direct connection with the bidding process were selected. The first set of such intervals, for which equity value changes are reported in Exhibit 4, ended with |D.sub.2~, the date of announcement of the white knight bid. The second set, for which equity value changes are reported in Exhibit 5, ended with |D.sub.3~, the date when the hostile bidder or the white knight exited from the contest. In each of the sets, three separate bid-related dates formed the beginning of the intervals: 50 days prior to the first hostile bid, immediately following the first hostile bid, and one day prior to the white knight bid. These intervals generated three columns in each exhibit. The six variable intervals provided a reasonably complete indication of the impact of the contest on the various participants.
TABULAR DATA OMITTED
An analysis of the average length of the intervals revealed that they ranged from a constant two days for the interval (|D.sub.2~ - 1, |D.sub.2~), to 83 days for the interval (|D.sub.1~ - 50, |D.sub.3~). The two methods for calculating dollar losses yielded similar results for the variable intervals of shorter average length. The divergence increased, however, with longer intervals. This was anticipated since, under the DLPF method, changes in the stock price were not corrected for variations due to changes in the market index. Such changes might, in the context of historical market return patterns, be large over longer intervals. Accordingly, we chose to adopt the MAAR method for all the intervals.(18)
1. White Knights
A review of Exhibits 4 and 5 indicates that over all the defined intervals covering the contest, the portfolio of white knight firms suffered large dollar losses, ranging from a minimum of $3128.70 million over the interval (|D.sub.2~ - 1, |D.sub.3~) of average 13-day length to a maximum of $5242.09 million over the interval (|D.sub.1~ - 50, |D.sub.2~) of an average 71-day length. The corresponding average losses of white knight firms thus ranged from $92.02 million to $154.18 million. In both the sets of intervals ending on |D.sub.2~ and |D.sub.3~, respectively, the dollar losses increased as the length of the intervals increased. Since the intervals spanned the entire contest and even incorporated a 50-day window prior to the contest, there is strong support for the view that the white knight bid was, on average, an equity-value-reducing transaction.
2. Hostile Bidders
In contrast to the investor reaction to white knight bids, the portfolio of hostile bidders posted dollar gains over all the relevant intervals. The magnitude of such gains over equivalent intervals was, however, considerably less than the losses of the white knights, ranging from a minimum of $128.51 million over the interval (|D.sub.2~ - 1, |D.sub.3~) of an average 13-day length to a maximum of $2085.28 million over the interval (|D.sub.1~ + 1, |D.sub.2~) of an average 25-day length. The average gains of hostile bidders thus ranged from $3.89 million to $63.19 million. The gains for hostile bidders were higher for intervals prior to the white knight bid at |D.sub.2~, and fell thereafter, though the average hostile bidder retained a smaller dollar gain by the end of the contest at |D.sub.3~ for all the time intervals.
The wealth consequences for bidders in general were observed to be negative in corporate control contests involving white knights. The overall portfolio of bidders suffered large dollar losses ranging from $1859.92 million over the interval (|D.sub.1~ + 1, |D.sub.2~) of average 21-day length to $4225.82 million over the interval (|D.sub.1~ - 50, |D.sub.3~) of average 83-day length. The average dollar loss for all bidders in aggregation ranged from $27.76 million to $63.07 million.
In view of the above result, it would seem to be inappropriate to evaluate the wealth consequences of combined portfolios of bidders in situations such as corporate control contests involving white knights, due to the opposite nature of effects for different subsets of bidders. In this instance, equity values reduce for the white knights and rise for the hostile bidders. Further evidence for the varying value consequences of bidding by white knights and hostile bidders in the contests can be seen from the fact that in averaging over all the six intervals, 71% of the white knights incurred dollar losses. The equivalent number for hostile bidders was 50%.
3. Target Firms
As expected, target firms posted large dollar gains over all intervals, ranging from $4525.38 million over the shortest two-day interval (|D.sub.2~ - 1, |D.sub.2~) to $13920.18 million over the longest interval (|D.sub.1~ - 50, |D.sub.3~) of average 83 days length. The average target gains thus ranged from $137.13 million to $421.82 million. When combined with bidders, the average firm in these matched trios had a net gain ranging from a minimum of $19.95 million for the interval (|D.sub.2~ - 1, |D.sub.2~) of two days length to a maximum of $99.78 million for the interval (|D.sub.1~ - 50, |D.sub.2~) which had an average length of 71 days. For the entire length of the contest, from 50 days prior to the start of the bidding process (|D.sub.1~ - 50) to the end of the bidding process (|D.sub.3~), the average participating firm had a net gain of $96.94 million. This encompasses very large gains to targets and, on average, small gains to hostile bidders and large losses to white knights.
The gains by the target firms were more than adequate to offset not only the net bidder losses but also the losses suffered by the portfolio of white knights. Thus, the corporate control contests where white knights participate do involve a transfer of wealth from the white knight to the target and the hostile bidder, and result in a net overall average dollar gain to the participants in the contest. From the economic welfare viewpoint, gainers (targets and hostile bidders) can potentially compensate the losers (white knights), and the contest does have a weaker (when compared with two-party contests), but nevertheless positive, economic welfare implication. It is, however, obvious that no actual transfer of resources occurs from the target (or the hostile bidder) to the white knight for attaining Pareto optimality. The market also does not expect such a transfer to take place on average, given the fall in the white knight stock price on announcement, which persists for the length of the contest.
VI. Summary and Conclusion
This paper investigates corporate control contests involving the participation of late-entry, collaborative bidders. We reason that this "white knight" subset of bidders might have different motivations from the hostile bidders. If their participation is reflexive and not in the context of strategic initiatives for value creation, then any benefits emanating from collaboration with the target may be more than offset by other considerations. In this context, we hypothesize that white knights might have negative value implications associated with their involvement in three-party corporate control contests. We find this to be the case, and we interpret our findings to provide additional insights into the range of outcomes for bidders that have been observed in studies of acquisitions employing general samples. The experience for late-entry, collaborative bidders is predominantly negative. For the total sample of 57 white knights, aggregate equity value decrease is $4.606 billion at the initial announcement of their involvement. Total losses for the 45 firms with negative outcomes are $5.476 billion, while the 12 firms with positive outcomes experience aggregate gains of 0.870 billion, 85% of which accumulates to one outlier. The profile of white knight losses is confirmed in the analysis of matched-trios in the 33 control contests where each party is exchange-listed. On average, losses to white knights are larger than gains experienced by hostile bidders. The net bidder losses are more than offset by target gains. We conclude that corporate control contests involving white knight bidders meet the Hicks-Kaldor criterion for welfare enhancement, but not the more stringent efficiency and distributional requirement of Pareto optimality.
1 Jensen and Ruback |11~, Jarrell, Brickley and Netter |8~, Black |2~, and Jarrell and Poulsen |9~ summarize various studies on the impact of takeover activity on the sample specific abnormal return profiles of bidders and targets.
2 Roll |18~ states, while discussing bidder returns, that "depending on the paper, the sample, the period, and the biases of the reader, widely differing conclusions can be reached" |18, p. 243~.
3 The extent of preparatory time prior to a bid can be looked upon as a proxy for collection and assimilation of complex information regarding possible targets and their strategic fit with the acquiring firm. The quick availability of internal information from a cooperative target, while highly desirable, is only one side of the equation. It is equally important for the bidding firm to be prepared for acquisitions in terms of their growth and value creation strategies.
4 The explanations offered for white knight behavior, which can result in negative market response, are consistent with the size maximization hypothesis of Shleifer and Vishny |20~ and the "free" cash flow hypothesis of Jensen |10~. It may be noted that both of these theories imply manager deviation from the goal of shareholder wealth maximization, and point to the persistence of negative market response. There are other theories, viz. the "hubris" theory of Roll |17~ and "winner's curse" theory of Varaiya and Ferris |23~, which ascribe overpayments by bidders as unintentional, caused by overenthusiasm or poor judgement on the part of the managers. Even in the context of these theories, the market reaction to the white knight bid is likely to be negative.
5 See, for example, Dennis and McConnell |5~.
6 For an explanation of DLPF and MAAR, please see Section IV.B. Note that Dennis and McConnell |5~ obtained a negative CAR for their sample of bidders for the two-day interval ending with the date of announcement of the bid, which was not converted to dollar values and reported (see Roll |18, pp. 242, 243~).
7 The studies by Malatesta |14~ and Dennis and McConnell |5~ are not directly comparable as the event dates were the month of board/management approval of the merger for the former and the merger announcement date for the latter.
8 Layard and Walters |13~ critique the K-H principle on the grounds that the efficiency and equity aspects of public policy are not independent of one another. Yet, they acknowledge the usefulness of the principle, since the final choice can always be made dependent on trade-offs between effects of the change on efficiency and equity, respectively |13, p. 38~. We are grateful to an anonymous reviewer for drawing our attention to this analysis.
9 While the concepts of welfare economics were originally developed for single producers and consumers, they are applicable to markets as well in a partial equilibrium framework. A detailed discussion of welfare measures in aggregation is available in Chapter 8 of Just, Hueth and Schmitz |12~.
10 The Dow Jones News Retrieval Service provides information from June 1979 onwards. We conducted the search using key expressions relating to white knight transactions. The announcement dates in the Wall Street Journal were cross-checked with those in the Dow Jones News Retrieval Service wherever available, and a minimum two-day window was used for event analyses.
11 In one contest, there were two white knights, with the first being unsuccessful. The market reaction to both the white knight bids was negative. As such, the number of white knights in the matched sample went up to 34, while the number of targets and hostile bidders remained at 33 each. There were no white knights who repeated their white knight role for a different target.
12 The value of |S.sup.2.sub.jt~ is given by
|Mathematical Expression Omitted~
|Mathematical Expression Omitted~ = residual variance for security from the market model regression;
|D.sub.j~ = number of observations during the estimation period;
|R.sub.mt~ = rate of return on the market index for day t of the event period;
|Mathematical Expression Omitted~ = mean rate of return on the market index during the estimation period; and
|R.sub.m|tau~~ = rate of return on the market for day |tau~ of the estimation period.
13 The standardized cumulative abnormal return |SCAR.sub.j~ over the interval t = |T.sub.ij~,..., |T.sub.2j~ is
|Mathematical Expression Omitted~
The test statistic for a sample of N securities is
|Mathematical Expression Omitted~
14 For perspectives on the persistent nature of the losses, see footnote 4.
15 Gulf Oil Corp. lost $1,003,500,000 over the two days ending with the report of its announcement that it was going to be the white knight in the control contest resulting from the hostile bid of Mesa Petroleum Co. for acquiring Cities Service Co. in 1982.
16 Dow Chemical Co. had a dollar gain of $735,000,000 over the two-day period ending with the report of its announcement that it was going to be the white knight in the control contest resulting from the hostile bid of Alberto Culver Co. for La Maur Inc., in 1987.
17 A revealing feature of the subsequent bids by white knights was their difference from the subsequent bids of the hostile bidder. While each subsequent bid of the white knight attracted a significant negative reaction from the market, the market remained indifferent to subsequent bids by the hostile bidder.
18 Results under the DLPF method are available from the authors.
1. Y. Amihud, B. Lev, and N.G. Travlos, "Corporate Control and the Choice of Investment Financing: The Case of Corporate Acquisitions," Journal of Finance (June 1990), pp. 823-837.
2. B.S. Black, "Bidder Overpayment in Takeovers," Stanford Law Review (March 1989), pp. 597-660.
3. B.S. Black and J.A. Grundfest, "Shareholder Gains From Takeovers and Restructurings Between 1981 and 1986: $162 Billion is a Lot of Money," Continental Bank Journal of Applied Corporate Finance (Spring 1988), pp. 5-15.
4. M. Bradley, A. Desai, and E.H. Kim, "Synergistic Gains from Corporate Acquisitions and Their Division Between the Stockholders of Target and Acquiring Firms," Journal of Financial Economics (May 1988), pp. 183-206.
5. D.K. Dennis and J.J. McConnell, "Corporate Mergers and Security Returns," Journal of Financial Economics (June 1986), pp. 143-187.
6. P. Dodd and J. Warner, "On Corporate Governance: A Study of Proxy Contests," Journal of Financial Economics (April 1983), pp. 401-438.
7. G. Hite and J.E. Owers, "Security Price Reactions Around Corporate Spin-Off Announcements," Journal of Financial Economics (December 1983), pp. 409-436.
8. G.A. Jarrell, J.A. Brickley, and J.M. Netter, "The Market for Corporate Control: The Empirical Evidence Since 1980," Journal of Economic Perspectives (Winter 1988), pp. 49-68.
9. G.A. Jarrell and A. Poulsen, "The Returns to Acquiring Firms in Tender Offers: Evidence From Three Decades," Financial Management (Autumn 1989), pp. 12-19.
10. M.C. Jensen, "Agency Costs of Free Cash Flow, Corporate Finance and Takeovers," American Economic Review (May 1986), pp. 323-329.
11. M.C. Jensen and R.S. Ruback, "The Market for Corporate Control: The Scientific Evidence," Journal of Financial Economics (April 1983), pp. 5-50.
12. R.E. Just, D.L. Hueth, and A. Schmitz, Applied Welfare Economics and Public Policy, Englewood Cliffs, NJ, Prentice Hall Inc., 1982.
13. P.R.G. Layard and A.A. Walters, Microeconomic, Theory, New York, McGraw-Hill Book Co., 1978.
14. P.H. Malatesta, "The Wealth Effect of Merger Activity and the Objective Function of Merging Firms," Journal of Financial Economics (April 1983), pp. 155-181.
15. M. Mitchell and K. Lehn, "Do Bad Bidders Make Good Targets?" Journal of Political Economy (April 1990), pp. 372-398.
16. C. Niden, "An Empirical Analysis of Gains to White Knight Corporate Takeovers: Synergy or Overpayment?" Working Paper, University of Pittsburgh, 1989.
17. R. Roll, "The Hubris Theory of Corporate Takeovers," Journal of Business (April 1986), pp. 197-216.
18. R. Roll, "Empirical Evidence on Takeover Activity and Shareholder Wealth" in Knights, Raiders and Targets, J.C. Coffee, Jr., L. Lowenstein, and S. Rose-Ackerman (eds.), Oxford, Oxford University Press, 1988.
19. K. Schipper and R. Thompson, "Evidence on the Capitalized Value of Merger Activity for Acquiring Firms," Journal of Financial Economics (April 1983), pp. 85-119.
20. A. Shleifer and R. Vishny, "Value Maximization and the Acquisition Process," Journal of Economic Perspectives (Winter 1988), pp. 7-20.
21. R.H. Smiley and S.D. Stewart, "White Knights and Takeover Bids," Financial Analysts Journal (January-February 1985), pp. 19-26.
22. N. Travlos, "Corporate Takeover Bids, Method of Payment, and Bidding Firms' Stock Returns," Journal of Finance (September 1987), pp. 943-963.
23. N. Varaiya and K. Ferris, "Overpaying in Corporate Takeovers: The Winner's Curse," Financial Analysts Journal (May-June 1987), pp. 64-70.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Special Issue: Corporate Control|
|Author:||Banerjee, Ajeyo; Owers, James E.|
|Date:||Sep 22, 1992|
|Previous Article:||Substitutes for voting rights: evidence from dual class recapitalizations.|
|Next Article:||Foreign takeover activity in the U.S. and wealth effects for target firm shareholders.|