Opinion
C.A. No. 20367.
Submitted: October 2, 2003.
Decided: October 7, 2003.
Gregory P. Varallo, Esquire, Richard P. Rollo, Esquire, RICHARDS, LAYTON FINGER, P.A., Wilmington, Delaware; Peter R. Bray, Esquire, BRAY, CHIOCCA MILLER, LLC, Parsippany, New Jersey, Attorneys for the Plaintiffs.
William M. Lafferty, Esquire, Charles D. Reed, Esquire, MORRIS, NICHOLS, ARSHT TUNNELL, Wilmington, Delaware; Jeffrey J. Bresch, Esquire, REED SMITH, LLP, Pittsburgh, Pennsylvania, Attorneys for the Defendants.
MEMORANDUM OPINION AND ORDER
I.
The plaintiffs in this case mounted a proxy contest to unseat the chairman of the board of directors of a publicly held Delaware corporation at its 2003 annual meeting. The outcome of the election was close, but the independent inspector of elections preliminary reported that the insurgents' candidate lost by over 190,000 votes. In reaching that conclusion, the inspector disqualified two proxy cards submitted by a bank representing a total of 232,376 shares after concluding that those proxy cards represented an overvote of the bank's position. Although 203,800 of those shares were voted in favor of the insurgent, 29,400 were voted in favor of the incumbent. Thus, the exclusion of those cards did not affect the outcome of the election.
In this lawsuit, the insurgents argue that the inspector of elections improperly defined the "overvote" at issue by excluding all of the shares covered by two other "omnibus" proxies given by the same bank in favor of two other banks holding shares as fiduciaries for several company-sponsored employee compensation plans. According to the insurgents, the inspector of elections should have defined the "overvote" to include all of the proxies given by the first bank because (they say) the inspector was never able to obtain adequate reliable information to form a judgment as to the validity of any of the proxies. Thus, the insurgents argue for the exclusion of all 859,430 shares voted either by the first bank or pursuant to the authority transferred in its omnibus proxies. If all of those votes are disregarded, the insurgent nominee will have gained a plurality of the votes cast and will have been elected.
The issues presented are whether the inspector of elections properly discharged its duties in defining the "overvote" in a way that disqualified some but not all of the proxy cards given by the first bank. If the answer to that question is "no," the court must then determine whether it may now take note of the facts adduced in discovery in this matter to validate the proxy cards reflecting the votes of the employee participants in those company-sponsored plans since the record shows that those proxy cards voted exactly the right number of shares.
II.
A. The Parties And The Background of The Dispute
The parties presented this matter to the court for its final decision on the basis of the Stipulated Record for § 225 Challenge ("Stipulated Record"). All of the facts discussed in this opinion are taken from that stipulation.
Defendant G.A. Financial, Inc. ("GAF") is a Delaware corporation with its principal place of business in Pittsburgh, Pennsylvania. Defendant John Kish is the Chief Executive Officer and Chairman of the Board of Directors of GAF. Shares of GAF common stock are registered under the Securities and Exchange Act of 1934 and are listed for trading on the American Stock Exchange. On or about January 3, 2003, GAF issued a press release announcing that its 2003 Annual Meeting of shareholders would take place on April 23, 2003 (the "Annual Meeting"). The record date for the Annual Meeting was March 10, 2003 (the "Record Date"). In connection with the call of the Annual Meeting, the GAF board of directors nominated Kish and another incumbent director, Hess, for re-election as directors to three-year terms.
Plaintiff Seidman and Associates L.L.C. is a New Jersey limited liability company with offices in Parsippany, New Jersey. Seidman and Associates is a stockholder of GAF. Plaintiff Lawrence B. Seidman is the managing member of Seidman and Associates. Seidman and Associates and other stockholders aligned with it formed the GA Financial Committee to Preserve Shareholder Value and made the requisite filings with the Securities and Exchange Commission to solicit proxies for the election of Seidman.
Because the Committee was only seeking one seat (of the two up for election) on the GAF Board of Directors, its proxy cards also included Hess. Hess's re-election is not challenged in this action.
Before the Annual Meeting, GAF appointed Corporate Election Services, Inc. ("CES") to serve as the inspector of elections for the Annual Meeting. In correspondence with GAF, CES described, inter alia, the procedures it would follow in the case of an "overvote," as follows:
If overvoting occurs, we will attempt to resolve the overvotes. If we are unable to contact the appropriate bank or broker, we will duly note the overvote and our treatment. In all ambiguous or unresolved cases, we will look to both sides for a mutually agreed upon resolution. Lacking this resolution, we will exclude the overvote from the tabulation, but note it in the certification.
Stipulated Record, Ex. B at 2.
In addition, GAF retained Georgesen Shareholder Communications, Inc. and the Committee retained D. F. King Co., Inc. to serve as their respective proxy solicitors.
B. The Overvote
On or about April 23, 2003, CES issued its Preliminary Tabulation of the votes cast at the Annual Meeting, reporting the vote with respect to the contested board seat, as follows:
Kish 2,117,179 votes Seidman 1,926,903 votes
In conjunction with the issuance of the Preliminary Tabulation, CES issued an Overvote Report disclosing a voting discrepancy attributable to The Bank of New York ("BONY") position. According to that report, CES preliminary determined that two proxies given by BONY to Automated Data Processing ("ADP") (which acts as agent for many banks and brokers in connection with proxy solicitations), attempting to vote 233,376 shares, overvoted the BONY position by 824 shares. Because CES had been unable to resolve the overvote issue in its communications with ADP and BONY, it invalidated those two proxies.
When it issued its Overvote Report, CES knew the following information about the source of the overvote.
• Cede Co., the nominee name of the Depository Trust Company ("DTC"), was the record holder of 4,422,673 shares of GAF common stock as of the Record Date.
• In connection with the Annual Meeting, Cede issued an "omnibus" proxy in favor of BONY for 859,647 shares of GAF common stock, reflecting BONY' s entire Record Date position at DTC. This proxy is dated as of the Record Date. This "omnibus" proxy was executed for the purpose of granting to BONY voting power over that number of shares.
• BONY in turn executed (and CES received) two further "omnibus proxies" in favor of two other banks, First Bankers Trust Company and First Bank of Clayton (Missouri). These proxies were for 625,771 shares and 1,500 shares, respectively, and were also dated as of the Record Date, although the Stipulated Record describes them as having been issued "[pirior to the closing of the polls."
Stipulated Record at para. 34 and 43.
• BONY also "voted" 233,200 shares through two proxies submitted through ADP, splitting the "vote" 203,800 for Seidman and 29,400 for Kish.
Id. at para. 49, 50 and 60.
• CES held two other proxies that bore on the issue. First was a proxy given by First Bankers Trust voting only 625,554 of the shares covered by the omnibus proxy given by BONY in its favor. The second was a proxy given by First Bank of Clayton (Missouri) voting all of the 1,500 shares covered by the omnibus proxy given by BONY in its favor.
Before issuing its Overvote Report, CES took certain measures to resolve the overvote. On April 23, 2003, CES contacted the proxy clerk at ADP and explained that there was an overvote. That clerk stated that he would contact BONY and get back to CES. On April 28, 2003, the proxy clerk at ADP reported that BONY was unable to ascertain the reason or reasons for the overvote and suggested that CES contact BONY directly. CES contacted a representative of BONY the same day. According to the Stipulated Record, the BONY representative "acknowledged the issuance of" the two omnibus proxies in favor of the two banks and reported that "she had checked her system, the votes looked fine, she could not see any obvious overvote and would have to further research the issue." CES informed her that "a review and challenge session was scheduled for May 2, 2003 and she was asked to contact CES if she discovered anything with respect to the overvote." This BONY representative did not further contact CES.
Id.
Id. at para. 42.
Before the May 2, 2003 review and challenge session, CES remained in contact with ADP and BONY but was unable to resolve the overvote. CES reported these facts to the parties. CES did not contact First Bankers Trust (or, presumably, the First Bank of Clayton) to ascertain or verify the number of shares in its position. At the May 2, 2003 challenge session, based on the information it then knew, CES determined to include the 625,554 votes cast by First Bankers Trust through ADP and the 1,500 votes cast by First Bank of Clayton through ADP and determined that First Bankers Trust had voted 217 fewer shares than it was authorized to vote pursuant to BONY's 625,771 share Omnibus Proxy. CES also determined to disqualify the 233,376 shares voted by BONY on the two proxy cards it gave to ADP, based on its view that the "overvote" was properly limited to those two cards. D.F. King, in its capacity as Seidman and Associates representative made a timely objection to this conclusion, arguing that "the entire BONY position should be disqualified, that it was improper to pick and choose [among] the BONY votes cast, and improper to `carve-out' part of the vote, allowing some votes to stand and disqualify others."
Id. at para. 63.
CES then certified the results of the election, reporting that Seidman lost by 190,276 votes. As mentioned earlier, the results of the election are not changed by including the 233,376 share proxies, as Seidman gains fewer net votes on those cards than necessary to overcome the vote differential. However, if all of the BONY shares are disqualified, the result is different. In that case, Seidman would win the contest by 347,826 votes.
C. Additional Information Learned In Discovery
As a result of discovery taken in this action, the parties learned that the proxy card actually given by First Bankers Trust to ADP (625,554) voted exactly the right number of shares. In other words, the omnibus proxy BONY gave in favor of First Bankers Trust for 625,771 shares overstated the latter bank's Record Date position by 217 shares. Discovery also established that the omnibus proxy for 1,500 shares given by BONY in favor of First Bank of Clayton (Missouri) was correct. There is nothing in the Stipulated Record that explains the remaining overvote found in the two proxy cards BONY gave to ADP voting the 233,376 shares.
These facts lead to several conclusions:
• CES was actually mistaken in concluding that First Bankers Trust had undervoted its position. Instead, the error lay in the omnibus proxy BONY gave in favor of First Bankers Trust.
• BONY's overstatement of First Bankers Trust's position by 217 shares in that proxy did not fully account for the overvote of 824 shares identified by CES.
• Even if CES had learned the correct information regarding the First Bankers Trust position as of the Record Date, CES would not have been able to resolve the remaining overvote on the two proxies BONY gave to ADP voting 233,376 shares.
III.
In Delaware, "[a] stockholder's ability to participate in corporate governance through the election of directors is a fundamental part of our corporate law." Under our law, "there is a `general policy against disenfranchisement.'" At the same time, Delaware law recognizes the need for certainty and finality in corporate elections, in order to avoid prolonged periods of turmoil. "The necessity for an expeditious conclusion of corporate elections that is consistent with the stockholders' franchise has resulted in the development of practical and certain rules."
Preston v. Allison, 650 A.2d 646, 649 (Del. 1994).
Id. (quoting Centaur Partners v. National Intergroup, Inc., 582 A.2d 923, 927 (1990)); Concord Fin. Group, Inc. v. Tri-State Motor Transit Co., 567 A.2d 1, 5 (Del.Ch. 1989) ("Delaware courts have vigilantly guarded against stockholder disenfranchisement in contested corporate elections").
Concord Fin. Group. Inc., 567 A.2d at 6.
Id.
The Delaware General Corporation Law directly addresses itself to the situation of an apparent overvote by a bank, broker or similar person. Section 231(d) clearly permitted the inspector of elections look beyond the proxies and the corporate records in an effort to reconcile that overvote:
In determining the validity and counting of proxies and ballots, the inspectors shall be limited to an examination of the proxies, any envelopes submitted with those proxies, any information provided in accordance with § 211(e) or § 212(c)(2) of this title, or any information provided pursuant to § 211(a)(2)(B)(i) or (iii) of this title, ballots and the regular books and records of the corporation, except that the inspectors may consider other reliable information for the limited purpose of reconciling proxies and ballots submitted by or on behalf of banks, brokers, their nominees or similar persons which represent more votes than the holder of a proxy is authorized by the record owner to cast or more votes than the stockholder holds of record.
8 Del. C. § 231(d) (emphasis added).
Section 231, adopted in 1990, codified the general rule of common law limiting the information that an inspector is permitted to review but carved out an exception for bank and broker overvotes. The Official Synopsis explains the changes to the common law as follows:
Subsection (d) specifies the information the inspectors may consider in determining the validity and counting proxies and ballots. This subsection is intended to be a codification of pre-existing common law with two exceptions. One change from the pre-existing common law is that inspectors are permitted to examine "reliable information" other than the proxies, ballots and books and records of the corporation, but only for the limited purpose of reconciling bank and broker "over votes" viz., proxies and ballots which represent more votes than the holder of the proxy is authorized by the record owner to cast or move votes than the stockholder holds of record. . . .
See Official Synopsis, 67 Del. Laws Ch. 376, § 9 (1990) (emphasis added).
The Delaware General Assembly adopted the overvote rule in response to the decision of the Delaware Supreme Court in Concord Financial, in which that court, following established precedent, invalidated the decision of an inspector to rely upon extrinsic evidence to resolve an overvote. Under the ruling in Concord Financial, clerical errors, even if readily resolvable, could not be rectified by the inspector of elections even though they might result in the disenfranchisement of large blocks of stock held in street name. The Delaware General Assembly adopted Section 231(d) to avoid this harsh result:
This provision [§ 231(d)] was intended in part to address the restrictions previously imposed by the Courts upon the examination by inspectors of extrinsic evidence to resolve overvotes. In cases where a record holder voted more shares than it actually held, "the error could not be corrected by looking at the face of the proxy or the books and records of [the company and therefore the proxy was] disregarded with no votes attributable to that proxy being cast" for either contestant. Concord Fin. Group, Inc. v. Fri-State Motor Transit Co., note 210 supra, 567 A.2d at 16-17. The adoption of Section 231 permits inspectors to examine extrinsic evidence in appropriate circumstances to resolve such overvotes.
R. Balotti J. Finkelstein, 2 The Delaware Law of Corporations Business Organizations § 7.33 n. 415 (3d ed. 1998 Supp. 2003).
The authors of another treatise explain that Section 231(d) "allows inspectors some additional latitude" in "resolving what are commonly referred to as "broker overvotes"':
Prior to the enactment of Section 231, inspectors of election had developed the practice of making telephone inquiries to brokerage houses in order to attempt to resolve overvotes. In Concord Financial, Inc. v. Tri-State Motor Transit Co., the Delaware Court of Chancery held this practice to be contrary to Delaware law. The adoption of Section 231 effectively conforms Delaware Law to the practice of inspectors in resolving broker overvotes and allows them to consider information the inspectors believe to be "accurate and reliable."
R. Balotti, J. Finkeistein, and G. Williams, Meetings of Stockholders, § 10.3. at 10-9 to 10-10 (1987 Supp. 2003) (footnotes omitted).
Other commentators similarly have explained the rationale behind the exception in Section 231(d) for broker overvotes:
This exception for broker overvotes was deemed necessary because, without extrinsic evidence, it is often impossible in contested situations to ascertain which proxies are revocations of proxies earlier submitted by the broker or nominee holder and which are new votes submitted on behalf of other beneficial holders. Hence, where the total vote of a broker or nominee holder exceeds the number of shares actually registered in its name, and some votes are cast for each slate, some form of communication with the broker or record holder is the only way of determining how the proxies should be voted.
D. Drexier, L. Black A. Sparks, Delaware Corporation Law Practice § 24.05[5] at 24-19 (1998 Supp. 2002) (emphasis added).
IV.
The first question presented on the record in this case is whether CES had a reasonable basis in fact upon which to conclude that the First Bankers Trust and First Bank of Clayton proxies were not part of the overvote. The court is satisfied that the information reported by BONY to CES provided a reasonable basis for CES's decision to limit the "overvote" issue to the 233,376 shares BONY attempted to vote. Because CES was not able to obtain reliable information to resolve that overvote, it properly excluded the two proxy cards reflecting those votes.CES specifically inquired of BONY about both of the omnibus proxies given by BONY in favor of the other two banks. In both cases, according to the Stipulated Record, BONY specifically "acknowledged the issuance of" those omnibus proxies and reported that, after a check of BONY's system, "the votes looked fine." In the case of omnibus proxies, this is the sort of information that CES could ordinarily be expected to regard as reliable. Unlike a proxy card that purports to "vote" shares, an omnibus proxy merely creates a paper record of a transfer of voting power to another bank or broker down the chain of ownership or title. The single piece of relevant information was the number of shares held by that other bank or broker as of the Record Date. That information should have been readily obtainable by the BONY proxy clerk from a review of BONY's internal records. Thus, it was reasonable of CES to rely upon information from BONY verifying the accuracy of the omnibus proxies.
See generally, R. Thomas and C. Dixon, Aranow Finhorn on Proxy Contests for Corporate Control § 15.05 at 15-35.
Further, there were no other circumstances present that should have caused CES to inquire further or demand documentary proof. In particular, the fact that First Bankers Trust appeared to have undervoted its position did not give cause for concern. In the case of proxy contests, banks and brokers are precluded from voting the shares they hold on behalf of others without explicit instructions. Thus, the fact that a position is undervoted would not be expected to raise a red flag in the mind of an inspector of elections. Undervotes are common and to be expected. They are ordinarily accepted without any need for inquiry. It is only overvotes that give rise to counting problems. Moreover, CES was able to observe that the 217 share discrepancy between the number of shares reported on the omnibus proxy given in favor of First Bankers Trust and the number of shares voted was 607 shares less than the total overvote observed. Thus, even if CES had had reason to question the reliability of the information received from BONY, it would not have had reason to conclude that there was anything wrong with either the 625,554 share proxy card or the 1,500 share proxy card given to ADP by First Bankers Trust and First Bank of Clayton, respectively. This is entirely consistent with the general policy of Delaware law favoring enfranchisement.
Id. § 8.03(D) at 8-57 to 8-59.
Preston, 650 A.2d at 649.
V.
An additional reason to uphold the decision reached by CES is found in the decision of the Delaware Supreme Court in Preston v. Allison . In that case, the plan trustee submitted "mirror image" proxy cards (one voting in favor of the management slate and one voting against the insurgent slate) that were meant to confirm the plan participants' support of management. Keeping with established rules, however, the inspector of elections determined that the two proxies cards were inconsistent on their face, could not be reconciled without resort to extrinsic evidence, and, thus, excluded both. The Delaware Supreme Court affirmed a decision of this court holding that, notwithstanding the general rule to the contrary, ERISA plan "shares should be voted in accordance with the participants' intent, even though the intent could not be discerned from the two irreconcilable proxies submitted on their behalf."The Preston court first explained the general rule as follows:
[A]lthough many investors choose to hold their stock through brokers or depositary companies, the corporation generally is entitled to rely on its own stock list and recognize votes or other stockholder action only when initiated by the stockholder of record. . . . If a beneficial stockholder is disenfranchised because of the record stockholder's failure to follow instructions, no relief is afforded in the usual case.
Id. at 649. Perhaps because the inspector of elections in Preston treated the problem as one involving irreconcilable proxy cards, not an overvote, the opinion does not discuss the then recent amendments to Section 23 1(d) discussed above. It should be noted, however, that the same result could have been reached by reference to that statute.
Nevertheless, the Delaware Supreme Court recognized that, in the case of plan participants who would be disenfranchised by a voting error committed by the plan administrator or some similar fiduciary, an exception to the general rule of law is appropriate. "Where stockholders are forced, by statute, to give up [control over the registration of shares], we conclude that they should not bear the risk of the trustee's mistake." Thus, the Supreme Court affirmed a decision in a Section 225 action that gave effect to the apparently irreconcilable proxies in a manner designed to give effect to the actual intent of the stockholders discovered during the course of the litigation.
Id.
Here, the votes that Seidman and Associates seeks to disqualify were cast by participants in company-sponsored benefit plans governed by the Employee Retirement Income Security Act of 1974. These plans included GAF's Employee Stock Ownership Plan ("ESOP") (540,029 shares) and its 1996 Stock-Based Incentive Compensation Plan ("Stock Plan") (85,525 shares). The participants in those plans were required by law to hold their shares through those plans, with First Bankers Trust acting as trustee. Moreover, there is no question about the voting intent of those plan participants. The Stipulated Record clearly shows that the proxy given by First Bankers Trust to ADP voting 625,554 shares accurately reflected the combined intended vote in the ESOP and the Stock Plan. Moreover, it was established in discovery and reported in the Stipulated Record that the total voting position of those two plans as of the Record Date was exactly 625,554 shares. Following the rationale of Preston v. Allision, those votes (as well as the 1,500 shares voted by First Bank of Clayton) must be counted.
Apparently, First Bank of Clayton acted as plan trustee in a much smaller plan holding only 1,500 shares.
The court notes, but rejects, Seidman and Associates's efforts to distinguish the decision in Preston. In particular, the court is persuaded that it makes no difference that this case involves an overvote and Preston did not. Nor should it matter that the error in Preston was actually committed by the plan fiduciary, whereas here the error, if any, was committed by BONY either by overstating First Bankers Trust's Record Date position in the omnibus proxy or in voting the 233,376 shares. The rationale of Preston is simply not so limited.