Opinion
Civil Action No. 19444-NC.
Submitted: July 7, 2004.
Decided: November 24, 2004.
Elizabeth M. McGeever, Esquire and Paul A. Fioravanti, Jr., Esquire of PRICKETT, JONES ELLIOTT, P.A., Wilmington, Delaware and Thomas E. Jamison, Esquire of FRUTH, JAMISON ELSASS, P.A., Minneapolis, Minnesota, Attorneys for Petitioners.
Arthur L. Dent, Esquire and Erica L. Niezgoda, Esquire of POTTER ANDERSON CORROON LLP, Wilmington, Delaware, Attorneys for Respondent.
MEMORANDUM OPINION
This is an appraisal action pursuant to 8 Del. C. § 262. Petitioners were stockholders of eMachines, Inc. ("eMachines" or the "Company") until the Company cashed-out stockholders pursuant to a tender offer and short-form merger at $1.06 per share. At trial, both parties presented expert testimony on the fair value of the shares as of December 31, 2001 (the "Merger Date"). For the reasons set forth below, the Court concludes that the per share fair value of eMachines as of the Merger Date was $1.64.
I. BACKGROUND A. The Parties
The Company was formed in September 1998 as a joint venture between computer parts manufacturers TriGem Computer, Inc. and Korea Data Systems Co., Ltd. ("KDS Ltd.") and their wholly owned subsidiaries, TriGem Corporation and Korea Data Systems America, respectively. John Hui, president of computer monitor company KDS USA, Inc., an entity related to KDS Ltd., was among the founders of eMachines, a member of its board of directors (the "Board"), and the owner of approximately 1% of eMachines' outstanding stock.
John Hui sometimes is referred to as Lap Shun Hui in the exhibits.
Petitioners, Charles T. Gholl and Michelle L. Gholl ("Petitioners"), were the beneficial owners of 339,000 shares of eMachines common stock at the time of the merger, December 31, 2001. Petitioners did not vote their shares in favor of the merger and duly exercised their statutory right to have this Court determine the "fair value" of their shares. Other shareholders holding 1,005,600 shares also have exercised their appraisal rights and are entitled to receive the appraised value for their shares as determined in this appraisal action.
See 8 Del. C. § 262.
B. The Company
The Company's initial business strategy was to sell low cost desktop personal computers ("PCs") through retail channels. It planned to keep its fixed costs low by outsourcing nearly all functions, including the design and manufacture of its PCs and monitors, customer service, and technical support. Although other computer manufacturers, such as Apple, Dell, and Gateway, also outsourced portions of their business functions, eMachines' fully outsourced business model distinguished it from its competitors. In fact, in some ways eMachines' structure was more akin to an electronics wholesaler — acting as a middleman between overseas computer manufacturers and domestic consumers. This structure allowed eMachines to generate more revenue, both per dollar of assets employed and per employee, than rival computer manufacturers. In other ways, however, eMachines operated much like any other computer manufacturer. It had to develop a brand, differentiate its products, and advertise to consumers. If successful, product differentiation strategies can allow a company to achieve higher margins. In contrast, wholesalers largely avoid these costly initiatives and instead compete primarily on price, with relatively small margins. Thus, eMachines' financial structure and market position made it somewhat atypical.In March 2000, eMachines completed an initial public offering ("IPO") at $9.00 per share, raising a net $165 million. At the same time, eMachines modified its business strategy to focus on deriving revenue from initiating Internet-based consumer relationships. Under this plan, the Company endeavored to sell its computers at break-even prices (or less), while seeking additional revenue from inducing customers to sign up for long-term Internet service contracts and from placing Internet companies' links on its computers' desktops.
The Company continued, however, its strategy of outsourcing most corporate functions, and actually decreased its number of employees from 162 in early 2000 to only 113 by the end of 2001. Trial Ex. 10 at 10; Trial Ex. 101 at 19.
The Internet-based revenue model proved to be a colossal failure, or in eMachines' words, a "financial disaster." By the end of 2000 revenue had declined 16%, gross margins had declined from 4.1% to a negative 3.1%, and eMachines' stock was trading below $.50 per share. Not only had the Internet revenue not materialized sufficiently, but also the Company had a poor reputation for quality and substantial operating problems. Thus, in early 2001, only one year after instituting the Internet-revenue business plan, eMachines hired a new management team to change the Company's course again.
eMachines' PTOB at 9. Citations to the parties' post-trial opening and answering briefs are in the form "Petitioners' PTOB or PTAB at ___" or "eMachines' PTOB or PTAB at ___," respectively.
Trial Ex. 3 at 3.
Tr. at 218-21 (Anderson). Citations in this form are to pages of the trial transcript with a parenthetical indication, where appropriate, of the witness who gave the cited testimony.
The Company recruited Wayne Inouye from Best Buy, eMachines' largest retail channel, to be its new CEO and President. Inouye, in turn, recruited another Best Buy executive, Adam Anderson, to become the firm's new COO and a Senior Vice President. The new management team announced a new business plan to eMachines' employees on May 18, 2001. The new business plan sought to return eMachines to a more traditional model whereby eMachines would earn a profit selling its hardware and provide much improved customer service. Specifically, the Company sought to: (1) have the highest Customer Satisfaction Index in the industry; (2) increase sales; (3) increase inventory turns from 6.5 to 22; (4) reduce returns by 50%; (5) reduce SGA to 2%; and (6) do all these things in a dollar neutral fashion. The Company hoped for considerable cost savings, and hence margin improvement, particularly as a result of reducing returns and increasing inventory turnover. It sought to reduce returns because of the high cost of refurbishment and the subsequent reduction in resale price due to obsolescence and the machines' status as refurbished goods. Similarly, increasing inventory turnover was important because rapid obsolescence of PC's caused a continuous decrease in the value of the inventory eMachines owned.
Trial Ex. 24.
Id. at EM009373.
Tr. at 281 (Anderson); Trial Ex. 16 at 7 (Inouye explained that "a large part of our inability to make money was based on the devaluation of products that we received back from the channel").
As eMachines' new management was implementing its turn around strategy, the perception in the marketplace was that the Company would not survive. Retailers began withholding payment of invoices fearing that eMachines would not be around to support returns. On May 24, 2001, eMachines was delisted from the NASDAQ, and its listing was transferred to the Over-the-Counter Bulletin Board market. Nonetheless, the Company continued to have sizeable liquid reserves from its IPO and even increased its cash and cash equivalents position in the second quarter of 2001 to $153 million ($1.05 per share) by June 30.
Tr. at 237-38 (Anderson).
Trial Ex. 11 at R002460.
The Company's new business plan began to produce results almost immediately; in only a few months, improvements were well under way. A July 25, 2001 press release highlighted the Company's implementation of new inventory management initiatives, the initial development of new customer care programs (to begin during the fourth quarter of 2001), and a reduction in salary expense of over 50%. At the same time, eMachines' market share increased. The Company's business continued to improve during the third quarter. During a quarterly conference call on October 23, Inouye explained that eMachines: (1) had its costs "well under control," (2) was gaining market share, and (3) had improved its inventory turns from 4 in 2000 to 16 in the third quarter of 2001. During the conference call, Inouye also declared that he didn't "understand why [eMachines'] stock is not at least a $2 stock."
Id. at R002461.
Trial Ex. 16 at 5, 16. By the end of 2001 inventory turns had more than doubled to 38 turns per year. Trial Ex. 2 at 8.
Trial Ex. 16 at 21.
C. The Company Explores Strategic Alternatives
On April 17, 2001, the eMachines Board formed an advisory committee to consider strategic alternatives, including a possible sale of the Company. On May 17, the day before the new business plan was unveiled to employees, the Company announced the retention of investment bank Credit Suisse First Boston ("CSFB") to assist in exploring such alternatives. CSFB contacted approximately 55 potential suitors and by August had identified two parties at least somewhat interested in pursuing a transaction: Gores Technologies Group and Medion AG. Gores proceeded to submit a nonbinding indication of interest at $.71 per share, or $110 million. On July 30, 2001, eMachines indicated to Gores that an acceptable offer would have to be at least equal to its cash value, approximately $170 million as of July 31, 2001. On or about August 8, Gores increased its proposal to $.83 per share, or $125 million. eMachines rejected the Gores offer as inadequate. The Company, however, never disclosed the Gores offer to the public. On September 17, eMachines announced that it had terminated the bidding process and CSFB's engagement.Less than two months later, on October 30, 2001, director Hui submitted an offer to purchase eMachines through an acquisition entity, EM Holdings, at a price of $.78 per share, or $118 million. In light of Hui's offer, the Board reconvened its advisory committee and re-engaged CSFB. The Board asked CSFB (1) to contact the parties that previously had expressed interest in a transaction with the Company and (2) to render a fairness opinion on the offer presented. On November 9, Hui's offer was announced to the public. The only companies CSFB contacted again were Gores and Medion. Once contacted, Gores re-entered the bidding process on November 11, this time offering $.97 per share, or $150 million. On the heels of the Gores offer, Hui submitted an offer of $1.00 per share, or $151.7 million.
The Board then made a surprising move on November 16: it set a tight deadline of November 18, 2001 for final bids. Within a few days, Hui won the bidding with an offer of $1.06 per share, or $161 million, outbidding Gores' latest offer of $1.01 per share.
D. CSFB's Fairness Opinion
On November 18 and 19, 2001, CSFB presented its analysis of the outstanding offers to the Board. CSFB opined in its written report (the "Fairness Opinion") that Hui's final offer of $1.06 was "fair, from a financial point of view." CSFB used several valuation methods, including two different discounted cash flow ("DCF") analyses — one based on management projections for 2002 furnished in mid November (the "Management Case") and one based on CSFB's own, much gloomier projections (the "Sensitivity Case"). CSFB's Management Case DCF produced a valuation range of $215-$340 million ($1.41-$2.22 per share), while its Sensitivity Case DCF indicated a valuation range of $138-$182 million ($.91-$1.20 per share). CSFB never explained which valuation method or methods it actually relied upon in finding the transaction fair. Because its Management Case valuation range far exceeded the deal price, however, the Court infers that CSFB did not base its Fairness Opinion on that set of projections.
Trial Ex. 21 at R000211. Originally CSFB suggested that eMachines management had prepared both the Management Case and the Sensitivity Case projections, and that they were both five year projections. Id. At trial, however, the evidence showed that eMachines "never budgeted beyond one year." Tr. at 248, 263-64 (Anderson).
Trial Ex. 21 at R000213, R000215. CSFB calculated the per share amounts using 154.1 million fully diluted shares and $2.4 million in proceeds from option exercise. Id. at R000174.
According to eMachines, CSFB felt the Management Case projections were too optimistic. eMachines' 14d-9 filing states: "[CSFB] believed that, in light of the historical performance and current market conditions, eMachines' ability to achieve the management case projections was subject to significant risks." Trial Ex. 6 at 24.
The management projections relied upon by CSFB in its Management Case came from the early stages of a budgetary process eMachines started at the beginning of November 2001. The process consisted of reconciling two different components: a "top-down" budget model created by senior management and a series of "bottoms-up" departmental budgets. The end result was the Company's baseline projection for 2002 (the "2002 Budget"). As the budgetary process evolved, the projections remained fluid and eMachines' forecasted net income for 2002 steadily increased from $1.8 million (mid November), to $6.6 million (late November), to $21.8 million by the end of the process in mid February 2002. There is no evidence that eMachines shared these increasing projections with CSFB or announced them to the public before the Merger Date. According to Anderson, however, the assumptions for the 2002 Budget were generally set by the end of 2001 (and the Merger Date), even though the actual budget, which projected net income of $21.8 million in 2002, was not finalized until February 2002.
Anderson at 181-82. Citations in this form are to the deposition transcript of the individual named. Trial Ex. 20 at EM007091.
Trial Ex. 31 (shows EBIT forecast for 2002 of $6.6 million excluding interest income).
Trial Ex. 20 at EM007091.
Anderson at 182.
E. Consummation of the Merger
On November 19, 2001, the Board unanimously approved the Merger (with Hui abstaining) and agreed to recommend it to eMachines stockholders. The tender offer commenced on November 27, and by its close on December 27, Hui had purchased 87% of the Company's outstanding stock. On December 28, 2001, Hui exercised a "top-up" option, signed along with the Merger Agreement. By exercising the top-up option, Hui was able to purchase directly from eMachines a number of shares sufficient to give him the 90% required to do a short-form merger under section 253 of the DGCL. The merger was completed December 31, 2001.F. The Domestic PC Industry
eMachines made most of its sales in the domestic PC industry. It also had a relationship with one foreign retailer, the sales to which amounted to approximately 10% of revenues in 2001. Tr. at 236 (Anderson).
By 2001 the domestic market for PCs was showing signs of maturing. In August of 2001, domestic unit shipments were forecasted to grow at a 4.67% annualized rate through 2005. The increasing acceptance of laptop computers had further depressed the outlook for desktop PCs specifically, where growth was forecasted at only 2.6% per year between 2000 and 2005. Additionally, the average selling price of a desktop PC in the U.S. had fallen from $1,847 in 1998 to $1,529 in 2000, and was forecasted to fall to $1,067 by 2005.
This mid single digit growth is contrasted with 17% and 23.8% growth for 1998 and 1999, respectively. Trial Ex. 21 at R000179: IDC Reports "Worldwide PC Forecast Update, 1999-2005" and "Worldwide PC Forecast Update, 1999-2004."
The PC market historically has been divided into two groups of vendors: Tier One vendors that were able to leverage their scale to achieve higher levels of profitability, and Tier Two vendors. CSFB, in its Fairness Opinion, stated that, as a consequence of slowing growth and declining average selling prices, only those vendors with the strongest profitability structures would be able to survive. As support for its opinion, CSFB pointed to several large computer manufacturers that had recently shut down or sold their operations due to continuing losses.
Trial Ex. 21 at R000177.
II. APPRAISAL
Under 8 Del. C. § 262, Petitioners are entitled to their pro rata share of the "fair value" of eMachines common stock as of the Merger Date. Fair value, as used in an appraisal setting, is defined as "the value of the Company to the stockholder as a going concern, rather than its value to a third party as an acquisition." Moreover, section 262(h) requires the court to determine the going concern value "exclusive of any element of value arising from the accomplishment or expectation of the merger or consolidation."
Gray v. Cytokine Pharmasciences, Inc., 2002 WL 853549, at *6 (Del.Ch. Apr. 25, 2002).
M.P.M. Enter., Inc. v. Gilbert, 731 A.2d 790, 795 (Del. 1999).
In an appraisal, both sides have the burden of proving their respective valuations by a preponderance of the evidence. Any "techniques or methods which are generally considered acceptable in the financial community and otherwise admissible in court" may be used. If neither party satisfies its burden, however, the court must then use its own independent judgment to determine fair value.
M.G. Bancorporation, Inc. v. LeBeau, 737 A.2d 513, 520 (Del. 1999).
Weinberger v. UOP, Inc., 457 A.2d 701, 713 (Del. 1983).
See Gonsalves v. Straight Arrow Publ'rs, 701 A.2d 357, 361 (Del. 1997) (noting this Court's responsibility to "independently determine the value of the shares that are the subject of the appraisal action") (emphasis added).
A. The Experts' Valuations
Petitioners called as their expert Daniel Larson, an Accredited Senior Appraiser and principal at the business and consulting firm EaglePoint Group. Before founding EaglePoint, Larson received an MBA from the Stanford School of Business and served as a Managing Director of Corporate Finance at Dain Rauscher, where he led its valuation services group. Larson performed both a DCF and a market analysis (also known as a comparable company analysis). He based his DCF on the 2002 Budget. Applying a 75/25% weighting to the respective models, Larson arrived at an equity value of $2.48 per share, or $379.6 million, as of December 31, 2001.The Company's expert was Gregg Jarrell, a professor of economics and finance at the William E. Simon Graduate School of Business Administration of the University of Rochester in New York. Professor Jarrell received both an MBA and a PhD in economics and finance from the University of Chicago. He also served three years as Chief Economist of the SEC. Jarrell first prepared a DCF based on the Management Case projections, arriving at a value of between $.93 and $1.03 per share, or $148.7 and $159.4 million. In addition, in response to Petitioners' position that the 2002 Budget contained the most useful projections, Jarrell performed an alternate DCF using those projections. That analysis indicated that eMachines' value on December 31, 2001 was within a range of $1.00 to $1.22 per share.
Jarrell also performed two market based analyses, including a control premium analysis and an analysis of the "controlled auction" that resulted in the sale to EM Holdings. Jarrell, however, did not rely on either of these market-based analyses to determine the fair value of eMachines. Instead, he averred that such evidence corroborated his fair value determination. The Court therefore will consider Jarrell's market based analyses only for a "reality check" of fair value. eMachines' PTOB at 24.
B. Valuation Methodologies
The parties' respective experts have presented remarkably divergent valuations. As is often the case in appraisals, the Court does not find either party to have fully satisfied its burden of persuasion regarding its valuation. The Court therefore must conduct its own independent valuation. Both experts relied primarily on the same valuation technique, DCF analysis, and agree on the basic methodology. A DCF analysis involves: (1) projecting operating cash flows out to a valuation horizon, (2) determining a terminal value which represents the business' value at the horizon, and (3) discounting all cash flows to their present value. In other words, a DCF model appraises the value of a company by looking at the present value of all its future cash flows. This method is widely accepted in the financial community and has frequently been relied upon by this Court in appraisal actions.Larson averaged his DCF with a market analysis in which he compared eMachines to Gateway. In comparing eMachines to Gateway, Larson applied a 25% discount to Gateway's multiples in recognition of the differences between the two companies. The valuation multiples Larson used were: (1) price-earnings of projected 2002 and 2003 earnings, (2) aggregate value-to-2001 revenue, (3) price-to-net cash, and (4) price-to-tangible book value. This analysis produced equity values ranging from $220 million to $820 million, with a median value of approximately $301 million.
Trial Ex. 3 at 24 and Ex. IX.
A comparable company analysis "depends on the similarity between the company [being valued] and the companies used for comparison." Because no two companies are identical, the court is always forced to determine at what point the differences between the companies "become so large that the comparable company method becomes meaningless for valuation purposes." In this case, the Court finds that Gateway is not sufficiently similar to eMachines to yield a useful comparable company valuation. In general, this is due to eMachines' unique market position and structural make-up. CSFB reached the same conclusion, noting the lack of "appropriate comparables."
In re Radiology Assoc., Inc., 611 A.2d 485, 490 (Del. Ch. 1991). See also Prescott Group Small Cap, L.P. v. Coleman Co., C.A. No. 17802, slip op. at 60 (Del.Ch. Sept. 8, 2004).
In re Radiology Assoc., 611 A.2d at 490.
Trial Ex. 23 at R000934.
Additionally, as eMachines points out and Larson admits, there were important differences between eMachines and Gateway at the time of the Merger. Whereas eMachines outsourced nearly all corporate functions, Gateway performed many tasks in-house and even managed its own retail stores. Gateway's far larger overhead gave it a markedly different cost structure than eMachines. Gateway's product lines were far broader than eMachines and included the faster growing laptop segment. More importantly, Gateway had invested many millions of dollars in building its brand in the marketplace whereas eMachines' brand had been associated with poor quality and service. This allowed Gateway to reap significantly larger gross margins than eMachines. Finally, the two companies' respective sizes were not comparable as Gateway's 2001 revenues exceeded eMachines' by roughly 12 times. In a rapidly consolidating industry where scale was at a premium, this size differential constituted a key difference between the two companies.
Trial Ex. 3 at Ex. VIII.
Furthermore, Larson's comparison of eMachines to Gateway yielded a widely divergent set of values. When a market analysis is based on only one "comparable" company and yields such a wide range of results, the Court seriously questions its usefulness. Specifically, Larson's earnings valuation multiples resulted in valuations above $600 million dollars, while his other approaches, such as price-to-net cash and price-to-tangible book value, resulted in valuations at or below $301 million dollars. This wide range of values implicitly violates the law of one price which holds that similar assets should sell for a similar price. The dichotomy in values shows that the two companies were structured and viewed by the marketplace quite differently. Larson's multiples analysis implies that an investor comparing eMachines' earnings to Gateway's would be willing to pay for the Company more than twice what an investor more concerned with balance sheet ratios such as price-to-net cash and price-to-tangible book would pay. This shows that some or all of the multiples Larson relied on are inapplicable and the Court is in no position to determine which, if any, are useful.
See Emerald Partners v. Berlin, 2003 WL 21003437, at *36 (Del.Ch. Apr. 28, 2003).
See Cede Co. v. JRC Acquisition Corp., 2004 WL 286963, *3 n. 28 (Del.Ch. Feb. 10, 2004).
Second, the substantial impact of Larson's use of the median value, as opposed to the mean, causes the Court to question the reliability of his analysis. The decision to use the median of the data set, $301 million, instead of the mean, $454 million, decreased the resulting fair value by 33%. As Larson testified, the effect of this approach is to "throw out the two highest numbers and the two lowest numbers. . . ." The wide variation between the median and mean values and the lack of a clearly articulated rationale for using one over the other further undermines the credibility of Larson's comparable company analysis.
Tr. at 55 (Larson).
The resulting dependence of Larson's comparable company analysis on one single ratio, price-to-net cash, is exemplified by a dispute regarding Gateway's net cash as of December 31, 2001. Larson used a figure of $1,026 million. Trial Ex. 2 at Ex. VIII. Jarrell, however, contends that Gateway's December 31, 2001 10-K reports a cash balance of $1,166 million. Trial Ex. 101 at 19 n. 37. The Court need not determine the correct figure, but notes that substitution of the larger net cash number into Larson's analysis decreases eMachines' fair value by $0.28 per share.
For the foregoing reasons, the Court will not apply a comparable company analysis to determine eMachines' fair value.
C. The Court's Analysis
Having determined not to utilize Petitioner's comparable company analysis, the Court still faces a difficult task in appraising the fair value of eMachines stock as of December 31, 2001. The Court has reservations about applying a DCF model in the present circumstances, where projections for only one year are available and the relevant operating history is equally limited. Both parties, however, have relied on DCF valuations in their presentations to this Court and appear to accept the efficacy of the method, although they disagree on many of the assumptions used therein. Further, the analytical difficulties facing the Court stem largely from the timing and the attendant process of the merger, which were controlled by eMachines. Thus, the Court is compelled in these circumstances to proceed cautiously using a DCF model, keeping in mind the limitations of the approach and the source of much of the uncertainty.
As discussed above, a DCF valuation involves three main components: (1) cash flows out to a valuation horizon, (2) terminal value at the horizon, and (3) the discounting back of all cash flows to their present value. In determining the first step, the value of cash flows during the initial period, the Court must establish projected revenues, operating margins, and taxes owed. To determine the second component, terminal value, the Court must establish the method of computation and a discount rate. Finally, to accomplish the third basic component of a DCF, the Court must choose a discount rate with which to reduce the cash flows from the first two steps to present value.
1. Management projections
An important threshold issue in conducting the Court's DCF valuation is the choice of projections for the Company's near term cash flows. Each side contends that a different set of projections by management best represents what was known or knowable as of the Merger Date. As a matter of law, the Court must consider "all relevant factors" involving the value of the Company. In doing so, all facts which "were known or which could be ascertained as of the date of the merger" are to be considered. Specifically, this Court recognized in Cede Co. v. Technicolor, Inc., that "[c]ontemporary pre-merger management projections are particularly useful in the appraisal context because management projections, by definition, are not tainted by post-merger hindsight and are usually created by an impartial body." Thus, as a starting point, "[w]hen management projections are made in the ordinary course of business, they are generally deemed reliable." In Technicolor, projections that management began developing two months before the merger date, but did not finalize until after that date, were held to contain information that was known or knowable as of the merger date.
Tri-Continental Corp. v. Battye, 74 A.2d 71, 72 (Del. 1950).
2003 WL 23700218, at *7 (Del.Ch. Dec. 31, 2003).
Id.
Petitioners urge use of the 2002 Budget because it was prepared contemporaneously to the Merger Date. They contend that even though the 2002 Budget was not finalized until mid February 2002, it was based on facts known or knowable as of the December 31, 2001 Merger Date. The Company disagrees and urges the Court to base its valuation on the Management Case projections. The Company advances three reasons for not using the 2002 Budget. First, it claims that the 2002 Budget was still "a work in progress" as of the Merger Date, and thus did not represent projections reliable enough to be considered known or knowable at the time. Second, eMachines contends that the 2002 Budget did not reflect a realistic forecast for 2002, but rather the Company's aspirational goals for the coming year. Finally, eMachines contends that if the 2002 Budget is used, it must be discounted in light of the speculative nature of the projections.
eMachines' PTAB at 15-16 and PTOB at 36.
The Company's reliance on the early November Management Case projections is misplaced. Because the 2002 Budget was not finalized until mid February, eMachines argues that the underlying information was not known or knowable on December 31, 2001. The Court rejected similar reasoning on similar facts in Technicolor. As in this case, preparation of the budget at issue in Technicolor began before, but was not completed until after, the merger. In finding that budget contained the most useful forecast, notwithstanding its date of completion, Chancellor Chandler stated, "[a]lthough March 1983 is post-merger, the CY 1983 Plan contains information that only validates what was known or knowable and susceptible of proof on or about January 24, 1983 [the merger date]."
See also Coleman, C.A. No. 17802, slip op. at 22 (applying projections not completed until nearly a month after the merger date because "most of the work . . . had been completed before or at the time of the merger").
Technicolor, 2003 WL 23700218, at *7. See also In re Cysive, Inc. S'holders Litig., 836 A.2d 531, 544 (Del.Ch. 2003) (in an entire fairness claim, rejecting respondent's argument that informally-kept budget numbers were of no use for valuation purposes).
As in Technicolor, eMachines' 2002 Budget process began two months before and ended a number of weeks after the Merger Date. Thus, unless there were material changes in the business during the weeks after the merger, all of the information underlying the 2002 Budget would have been known or knowable by the Merger Date. Anderson's testimony confirms that the assumptions underlying the 2002 Budget were developed before the Merger date and "pretty much stayed the same." Further, Anderson testified that the 2002 Budget numbers, despite not having been finalized until February, were arrived at by the end of 2001 and that "[t]here probably was not a lot of difference between the iterations of the forecast . . . from the end of the year until January [meaning the end of the budget process]. . . ." The Court therefore finds that the facts underlying the 2002 Budget were known or could have been ascertained as of the Merger Date.
Anderson at 183 (referring to the 2002 Budget assumptions as outlined in Trial Ex. 20 at EM007090).
Id. at 182.
Even if eMachines' argument that the 2002 Budget was still in flux as of the Merger Date were correct, it still would not justify acceptance of the Management Case projections. Instead, the Court would be inclined to use projections somewhere between the last concrete projections known to have been made before December 31, 2001 — i.e., the November 27, 2001 $6.6 million projection — and the final 2002 Budget. Having concluded that the 2002 Budget is reliable, however, the Court need not analyze this issue further.
In arguing for use of the Management Case projections, eMachines suggests that the two sets of projections were created for different purposes and were based on wholly different assumptions. It touts the Management Case projections as "conservative, historically-based forecasts" prepared by managers only six weeks before the merger. In contrast, eMachines describes the 2002 Budget projections as "aspirational," "highly aspirational," and "highly aggressive." Indeed, eMachines repeatedly cited in its post-trial briefs to the testimony of its COO, Anderson, analogizing the 2002 Budget to a football coach diagramming a "hail Mary, a triple reverse, flea-flicker with a pass to the quarterback in the end zone." In other words, the 2002 Budget was a true long shot, "prepared on the assumption that all of the Company's new business initiatives would be wildly successful." For the reasons stated below, the Court finds Anderson's characterization and eMachines' adoption of it a gross exaggeration that undermines the credibility of their entire presentation.
eMachines' PTAB at 2. The exact date of delivery of the Management Case is not clear, but it was after November 13, 2001 (when CSFB requested delivery of the projections) (Trial Ex. 22) and before November 18 (the date CSFB presented its Fairness Opinion to the Board) (Trial Ex. 6 at 9).
eMachines' PTAB at 2, 16.
Tr. at 288 (Anderson).
eMachines' PTOB at 36.
First, and most importantly, the Court notes that the two sets of projections were merely iterations of the same set of continuously evolving projections (the 2002 Budget representing the final iteration). The budget process began with management meeting for two days in early November to complete their "top-down" goals. These goals were then reconciled with "bottoms-up" departmental budgets, and this process continued from early November until the 2002 Budget was finalized just prior to presentment in February. The beginning of this process yielded the Management Case projections. The Company never created a forecast of any kind specifically for CSFB; they merely provided CSFB with their budget figures for 2002 as of early November. The 2002 Budget represents the culmination of that budgetary process. There is no evidence that the process began with conservative assumptions and that management later ramped them up to create a pie-in-the-sky document in the form of the 2002 Budget. Thus, because of the nature of the process, there is no reason to believe that projections developed in the very beginning were more reliable than later projections, since each passing iteration represented the most current and refined thinking of management at that particular time. Nor is there any evidence to suggest that the February 2002 Budget reflected any "element of value arising from the accomplishment or expectation of the merger," as proscribed by 8 Del. C. § 262(h).
Tr. at 263, 331 (Anderson). A third iteration, based on data from a period between the Management Case and the 2002 Budget, was contained in an email from Anderson to Hui and Inouye on November 27. At this juncture, less than two weeks after delivery of the Management Case, the then-current budget showed an increase in projected net profit for 2002 from $1.8 million to $6.6 million. Trial Ex. 31.
The Company's dismissal of the 2002 Budget as "aspirational" is also unconvincing since it represented eMachines' only operating budget. As in Technicolor, the 2002 Budget was produced in the ordinary course of business, for operational purposes. Thus, all of the Company's planning, on both company-wide and departmental levels, was keyed to it. The manner in which the 2002 Budget was presented to employees at eMachines' Town Hall meeting in February 2002 corroborates this conclusion. At that time, eMachines called the 2002 Budget its "baseline budget." Anderson explained that the baseline budget represented "business as usual" without new projects, and testified:
Trial Ex. 20 at EM007092.
Anderson at 183.
Well at the very macro level we do what's called baseline budgeting. We essentially look at the business, the way it was in 2001, and say if we didn't change — fundamentally change anything, no new markets, no new products, no significant differences in the business, what do we think we could achieve.
Tr. at 267 (Anderson).
That hardly sounds like a "hail Mary" or a "triple reverse, flea-flicker with a pass to the quarterback in the end zone."
The Company's characterization of the 2002 Budget as "aspirational" is further undermined by the bonus plan management presented to its employees along with the 2002 Budget (the "Bonus Plan"). The Bonus Plan was based on achieving operating profit of $25 million, nearly 15% more than the 2002 Budget projections. The mere existence of the Bonus Plan contradicts eMachines' characterization of the 2002 Budget as "aspirational." If the 2002 Budget represented management's wildest dreams come true, it would be illogical and callous to key the Bonus Plan to even higher targets that were not achievable — a point Anderson confirmed. Thus, far from supporting eMachines' claim that the 2002 Budget was "aspirational," the evidence shows that it represented management's best projections of performance for 2002 on a "business as usual" basis.
Trial Ex. 20 at EM007087.
Id. at EM007102. Development of the Bonus Plan did not begin until after the Merger Date (T. Tr. at 296-97 (Anderson)); thus, the Court relies on the Bonus Plan only as impeaching eMachines' "aspirational" argument, and not as a basis upon which to project future operating income.
Anderson at 192.
The Company's actual results for 2002 certainly were not known or knowable as of the Merger Date. In light of eMachines repeated protestations that the 2002 Budget was merely "aspirational" and a "hail Mary," the Court concludes that the evidence of its actual 2002 results is admissible for the limited purpose of testing the veracity of those protestations, and solely for that purpose. The actual results were net income of $25.5 million, surpassing even the Bonus Plan figures and besting the 2002 Budget by nearly 17%. Trial Ex. 19.
The Company also contends Larson's model is unreliable because it is based on the one-year 2002 Budget. They seek to contrast that with the allegedly five-year Management Case created by eMachines management. In actuality, however, eMachines never prepared five-year projections for CSFB, the Management Case, or any other purpose. CSFB itself created the five-year projections referred to in its opinion by extrapolating from the one-year Management Case numbers supplied by eMachines. Because this Court frequently has expressed a preference for contemporaneous projections of management prepared for business use, as opposed to those of financial experts, the Management Case rightly can be considered to refer only to the one-year projections actually supplied by management. Thus, both sets of projections were only one-year projections, requiring the experts to forecast beyond 2002 regardless of which projections they used.
See, e.g., eMachines' PTOB at 27 and PTAB at 16, 27-28.
Tr. at 248 (Anderson); Anderson at 110; Tr. at 519 (Jarrell).
Tr. at 248 (Anderson).
See, e.g., Technicolor, 2003 WL 23700218, at *7; In re Radiology Assoc., 611 A.2d at 490-91; Harris v. Rapid-American Corp., 1990 WL 146488, at *7 (Del.Ch. Oct. 2, 1990), aff'd in part, rev'd in part on other grounds, 603 A.2d 796 (Del. 1992).
The Company also argues that the 2002 Budget was unsupported by the Company's past performance. That criticism, however, would apply to any forecast of eMachines' earnings at the Merger Date since the Merger took place less than a year after the new management team was put in place and only seven months after it implemented an entirely new business plan. Moreover, the Company's prior history reflected the results of its ill — fated Internet — based strategy. Because there is no dispute that strategy was abandoned as a colossal failure, it offers little in the way of relevant history to the Merger at issue here.
In sum, the Court does not doubt that the Management Case projections reflected management's knowledge at the time they were prepared in early November 2001. The rapid and substantial evolution of the budget during November and December, however, illustrates the fluidity of the process. The rapid progression of estimates and the continued success of eMachines' new business plan in late 2001 cause the Court to find the Management Case projections less reliable. In contrast, the 2002 Budget was primarily completed by the Merger Date and represented facts known or knowable on the Merger Date. It is, of course, regrettable that more relevant historical information from which to make projections is not available. The Court is required, however, to appraise the stock in light of all the circumstances presented. The Court, therefore, views the 2002 Budget as the best available projection of net profit for 2002 based on what was known or knowable on December 31, 2001.
The Court notes that in comparing the respective experts' valuation assumptions, it will examine only the set of assumptions Jarrell applied in his valuation based on the 2002 Budget. The assumptions used in Jarrell's alternative valuation based on the Management Case understandably differed significantly from those based on the 2002 Budget and are not germane.
2. The Court's DCF
Having decided that initial cash flow projections should be based on the 2002 Budget, the Court now turns to the individual assumptions to be used in its DCF.During the many rounds of valuation reports, revisions, and rebuttal reports, the parties managed to establish some common ground. When using the 2002 Budget figures, both parties agreed on the following points: (1) utilization of a five-year valuation horizon, (2) an assumption of 5% annual revenue growth during the initial five-year period, (3) the applicability of taxes and net operating loss ("NOL") carry forwards, (4) expected litigation expense, (5) proceeds from stock option exercise, and (6) the number of fully diluted common shares outstanding. The areas of contention relate to the operating margins, the method for determining terminal value, the discount rate, and the treatment of excess cash. Each of those areas is discussed below.
3. Operating margins
The 2002 Budget forecasted an operating margin of 2.7%. Unfortunately, as noted above, management made no forecasts beyond 2002, leaving eMachines' future margins an area of heated debate. Larson increased margins 10 basis points per year from 2.7% in 2002 to 3.1% in 2006. He opined that eMachines' margins were likely to increase because it would benefit from economies of scale as its sales volume grew. Larson could not point to any evidence, however, to support his assumption of a continual increase over five years. To the contrary, at least one expert in the field asserts that economies of scale should not be presumed. Further, there are reasons specific to eMachines to doubt the impact of economies of scale in this case. Economies of scale are created when increased volume leverages fixed costs, thus decreasing average costs and increasing margins. By design, eMachines' structure sought to limit fixed costs by outsourcing nearly all corporate functions. With only 113 employees and no factories, it had few fixed costs to leverage into larger margins. The Court therefore finds Larson's contention that eMachines' margins were likely to continue increasing by 10 basis points per year unpersuasive.
See Cornell, at 127.
Jarrell, on the other hand, linearly ratcheted eMachines' margins down from the 2.7% projected for 2002 to 1.3% by 2006. He then assumed 1.3% would remain eMachines' margin rate in perpetuity. In support of this position, Jarrell posited that 2.7% margins would result in return on investment sufficient to attract competition, thus driving margins down. Jarrell based his opinion on information contained in Ibbotson's Cost of Capital 2001 Yearbook, a treatise often used in business valuations. He focused on SIC codes 3571 and 5045, representing computer manufacturers and computer and computer related product wholesalers, respectively, noting that companies in the two industries vary greatly in their operating margins and sales-to-investment ratios. While the median wholesaler (5045) had a sales-to-capital ratio on the order of 11 to 1, the ratio for the median computer manufacturer (3571) was in the range of 0.5 to 1. Conversely, SIC code 5045's five-year average operating margin was only 1.87%, compared to 9.31% for SIC code 3571. According to Jarrell, this reflected the fact that industries with relatively small capital needs necessarily must be low-margin businesses. Otherwise, the argument goes, capital would flow into the low-investment/high-profit business, driving margins down to an equilibrium level. Based on eMachines' large sales-to-capital ratio and the lack of barriers to entry in the bargain PC industry, Jarrell assumed it would have low margins of 1.3%, the figure used in CSFB's Fairness Opinion, for 2006 and beyond.
Tr. at 469-72 (Jarrell).
Trial Ex. 111; Tr. at 473-78 (Jarrell). Jarrell testified at trial that eMachines is in SIC Code 5045 although its computer manufacturer competitors are classified in SIC code 3571.
Trial Ex. 111.
Id.
Trial Ex. 101 at 17.
The Court accepts Jarrell's basic economic premise, that businesses without large barriers to entry and with low capital investment needs can be expected to have relatively small margins. Moreover, eMachines appears to fit that description. This principle alone, however, sheds little light on whether 1.3%, 2.7% or 3.1% more accurately reflects the probable long-term margins for eMachines, since all three figures are "small" when compared to the 9.31% margins of the median computer manufacturer.
Jarrell relies on the 1.3% figure because it was the projected margin for 2006 in CSFB's Fairness Opinion based on the Management Case. The Court finds this unreasonable, since the Management Case projections forecasted only 0.3% margins for 2002. From there, CSFB (not management) predicted eMachines' margins would improve only modestly each year until they reached 1.3% in 2006, presumably through execution of the new business plan. The prospects for the Company reflected in the more representative 2002 Budget, however, were much brighter, with margins forecasted to jump to 2.7% on expected net income of $21.8 million in 2002. Thus, the 2002 Budget not only predicted success in accomplishing the Company's operational goals, but also showed the profitability anticipated under the new business plan. The Court therefore finds that an operating margin of 2.7% for eMachines was both realistic and sustainable.
Faced with a paucity of relevant data, especially in terms of the limited operating history as of the Merger Date for eMachines' new business model and the absence of any long-term projections by management, the Court is not convinced of the reasonableness of either side's predictions regarding future margins. In such a case, the Court finds it most reasonable to assume that a combination of the factors described by the parties, including economies of scale and competition, would roughly counterbalance each other. Thus, for purposes of its valuation, the Court will assume that eMachines' future margins would remain at the same level as projected for 2002, 2.7%, indefinitely into the future.
4. Discount rate
A discount rate is used to calculate the present value of future cash flows. To determine an appropriate discount rate, both parties, as well as CSFB, relied on the well-accepted weighted average cost of capital methodology ("WACC"). The results were as follows: CSFB 13.7-14.8%, Larson 18%, and Jarrell 18-22%. In the circumstances of this case, unlike some other cases, use of different WACCs does not have a major impact on eMachines' fair value. In fact, using the two extremes of the proposed WACC values (14 vs. 22%) in the Court's final valuation produces a difference of less than 9%.
The formula used to derive the WACC is: WACC = (Cost of Equity × Equity-to-Total Capital Ratio) + (Cost of debt × (1 — Tax Rate) × Debt-to-Total Capital Ratio). See Lane v. Cancer Treatment Ctrs. of Am., Inc., 2004 WL 1752847, at *30 (Del.Ch., July 30, 2004).
Larson used the "build-up" method to estimate eMachines' cost of equity capital, arriving at a figure of 20%. Larson then estimated eMachines' cost of debt at 7%, leaving an after-tax cost of debt of 4%. Finally, Larson examined the capital weightings of technology firms and decided that a 10% debt and 90% equity structure was typical. Jarrell, on the other hand, used the capital asset pricing model ("CAPM") to estimate eMachines' cost of equity, arriving at a range of 19.39 to 23.42%. To estimate eMachines' cost of debt, he looked to Hui's 8% cost of financing the transaction (which utilized the Company as security). Jarrell did not discount the 8% cost of debt for tax savings because he believed interest expense deductions would not benefit the Company since it already paid no taxes due to its NOLs. Finally, assuming a debt ratio of between 5 and 15%, Jarrell calculated eMachines' WACC to be between 18% and 22%.
The build-up method begins with the risk-free rate as a base. Additional rates representing the security's unique risks are then added to the risk-free base. Shannon Pratt, The Lawyer's Business Valuation Handbook 122 (2000). Larson began with a risk-free rate of 5.7% before adding an equity risk premium of 10.2% and an industry risk premium of 4%. Trial Ex. 3 at 21-22.
In other words, a 10% "debt ratio."
Tr. at 437 (Jarrell); Trial Ex. 115. Jarrell revised his original report, Trial Ex. 100, at trial to arrive at the 8% figure in Trial Ex. 115.
Both experts agreed that eMachines' debt ratio should be in the 10% range. Larson offered no explanation for his derivation of a 7% before-tax cost of debt, while, Jarrell based his 8% cost of debt on an actual financing in place before the Merger Date. The Court therefore will use 8%. The Court disagrees, however, with eMachines' argument that its NOLs render potential tax deductions stemming from interest payments worthless. Section 172 of the Internal Revenue Code defines NOLs as "the excess of the deductions allowed by this chapter over the gross income." Thus, NOLs are to be calculated after applying the Code's other deductions, and any deductions for interest payments would allow the Company to save its NOLs for subsequent years. Based on the foregoing, the Court will apply a 4.8% after-tax cost of debt.
Turning to the cost of equity figures, Larson's 20% figure corresponds to the lower end of Jarrell's 19.39 to 23.42% range. In its Fairness Opinion, CSFB used a 14.8% cost of equity, derived using the CAPM. In view of this range of figures, and without making a determination as to the relative merits of the parties' methodologies, the Court will use a 20% cost of equity. Thus, using a 10% debt ratio, a 20% cost of equity, and a 4.8% after-tax cost of debt, the Court concludes that eMachines' WACC is 18.5%.
The calculation is as follows: WACC = (Cost of Equity × Equity-to-Total Capital Ratio) + (Cost of debt × (1 — Tax Rate) × Debt-to-Total Capital Ratio) = (.20 × .90) + (.08 × (1 — .40) × .10) = 18.5%.
5. Terminal value
The two established methods for computing terminal value are the exit multiples model (a market approach) and the growth in perpetuity model. In this case, Larson used the former, based on the price to earnings ratios of purportedly comparable companies, while Jarrell used the latter. Both approaches have been accepted by this court in the past.Jarrell attacks Larson's use of the exit-multiples approach because of the 14% growth in perpetuity his calculation implies. This Court agrees that a 14% growth rate is too high, when compared to the economy's long-term growth rate of 3-4% and the domestic PC industry's projected growth of 4.67%. Larson's exit-multiples approach also results in eMachines' terminal value representing over 70% of its estimated total present value. This large reliance on eMachines' value in 2007 raises a red flag in the Court's mind and suggests that the exit-multiples approach is not appropriate in this case. Further, since a multiples approach is a market-based approach, its reliability depends on being able to identify comparable companies. As discussed above, eMachines' rather unique structure does not lend itself to being compared with its computer manufacturer rivals. For these reasons, the Court will use a perpetuity growth model to calculate eMachines' terminal value.
Tr. at 481 (Jarrell).
Trial Ex. 3 at Ex. VII(b).
See Cytokine Pharmasciences, 2002 WL 853549, at *9 (noting that results of a DCF must be regarded suspiciously where the terminal value accounts for over 75% of total present value).
The two main components of a growth in perpetuity model are the discount rate and the growth rate of cash flows beyond the valuation horizon. The Court will apply an 18.5% discount rate as derived above. Jarrell assumed 5% growth in perpetuity, or steady growth before and after the valuation horizon. Larson's exit-multiples model does not require a growth rate as a necessary input, but it implies 14% growth in perpetuity. Use of a 14% growth rate would represent a large increase from the 5% growth Larson forecasted for years 2002-2006. Because Larson offered no evidence justifying his large growth rate in the terminal period, the Court finds it more reasonable to assume continued growth of 5% into the terminal period. Moreover, assuming a 3-4% inflation rate, 5% growth implies less than 200 basis points of real growth into the future, and is roughly in-line with the forecasted trend for the domestic PC industry.
The Company argues that the WACC rate used after the valuation horizon should be lower than that used at the Merger Date to represent the decreased risk associated with the then more mature company. The Court rejects this assertion on two grounds. First, eMachines' own expert's models applied a stable WACC across all time periods. Trial Ex. 100 at Ex. 16; Trial Ex. 101 at Ex. 28. Jarrell did not use a lower WACC in later periods in either his independent DCFs or in his rebuttal adjustments to Larson's opinion. Second, in Valuing a Business, Shannon Pratt explains that varying the discount rate over time to reflect changes in the riskiness of cash flows "is a highly judgmental (and usually quite subjective) matter." Shannon Pratt, Valuing a Business 159 (4th ed. 2000).
Both parties assume eMachines' margins will remain constant after the valuation horizon, and that by then it will pay taxes at a full 40% rate. Thus, unlike the period before the valuation horizon, in the terminal period eMachines' cash flows are assumed to grow at the same rate as its revenues.
Trial Ex. 101 at 11.
6. Excess cash
Generally, a business needs cash to fund ongoing operations. Such capital is reflected in a DCF valuation just as any operating asset is. Nevertheless, any cash in excess of that necessary to fund the ongoing operations of the business is considered "excess cash" and is not reflected in a DCF. Thus, in determining the fair value of a corporation, excess cash must be added to the result of the DCF valuation.
See Ala. By-Products, 1990 WL 109243, at *16.
Id.; In re Radiology Assoc., 611 A.2d at 495.
As of the Merger Date, eMachines possessed $180.5 million of cash. Neither party claims eMachines needed the entire $180.5 million as operating capital. The Company insists that it needed between $50 and $75 million of cash to sustain operations. In support of this claim, eMachines relies on Anderson's testimony that a "significant portion" of the $180.5 million was necessary to operate on a day-to-day basis. The Company also notes that its working capital balance was wholly attributable to cash, and that its noncash working capital balance was actually a negative $11.5 million as of the Merger Date. Petitioners argue that the entire $180.5 million should be considered excess. They point out that during eMachines' third quarter conference call, Inouye told investors that the Company didn't plan to consume any cash (or other working capital) during the fourth quarter. Also, Larson points out that negative noncash working capital balances are common in the PC industry, and are held by other companies such as Gateway, Apple, and Dell.
Tr. at 309-10 (Anderson).
Trial Ex. 100 at 20; Trial Ex. 106 at R000132. Here a brief accounting primer may be useful. Working capital is derived by subtracting current liabilities from current assets and represents the capital the business has at its disposal to fund operations. See Shannon Pratt, The Lawyer's Business Valuation Handbook 422 (2000). The designation "current assets" includes categories such as short-term investments, accounts receivable, and inventories, as well as cash. Trial Ex. 106 at R000132. Thus, Jarrell's argument is that if all of eMachines' cash were removed, it not only would have no cash for operations, but also would have short term liabilities that exceeded its current assets by over $10 million.
The Court is not convinced by Petitioners' claims that the entire $180.5 million should be considered excess cash. First, Petitioners have mischaracterized Inouye's comments during the conference call. There is an appreciable difference between stating that eMachines didn't plan to consume cash during the fourth quarter and implying that eMachines did not need to carry any cash at all. Taking into account its customers', and the market's, doubts about its financial stability, eMachines likely needed cash to assuage concerns regarding its continued viability. Additionally, Larson's assertion that negative noncash working capital is common in eMachines' industry is beside the point. Larson suggests that fact means that removing all $180.5 million of eMachines' cash and leaving it with a negative working capital balance would not cause a problem. The issue is not, however, whether PC companies carry negative noncash working capital balances, but what size their overall working capital balances are. In 2001, Gateway, Apple, and Dell had working capital balances of $977 million, $3.69 billion, and $483 million, respectively. This implies that each of those companies had a large cash position. Thus, whether they also had negative noncash balances is immaterial to what the appropriate amount of excess cash is in this case.
Trial Ex. 4 at Ex. III.
That being said, there is little support for eMachines' contention that it needed $50 to $75 million in cash to support operations. Anderson testified that a "significant portion" of the $180.5 million was necessary for the Company to operate on a day-to-day basis and pay for inventory before those costs were recouped from retailers. The Company offered no proof, however, that its cash balance decreased substantially at any point during the preceding year, or, for that matter, that its estimation of its cash needs was anything other than litigation driven. Thus, the Court finds that eMachines' $50-$75 million estimation of its cash needs is not supported by the evidence.
Tr. at 309-10 (Anderson).
This Court prefers projections prepared by management in the course of business. The Company presented no contemporaneous projections of the amount of cash needed to run its business as of the Merger Date. Yet, there is evidence that one eMachines director, the prospective owner, Hui, did consider that issue in late 2001. In purchasing the Company, Hui planned to use approximately $145 million of eMachines' cash to pay deal-related expenses and monies borrowed to fund the Merger, and $21 million to reimburse himself for cash contributed to the Merger. Thus, Hui planned to use $166 million of eMachines' cash in consummating the transaction, leaving only roughly $15 million available for operations on a going-forward basis. Recognizing the position this would put eMachines in, Hui planned to renegotiate eMachines' letter of credit and terms with vendors. Because these plans contemplated Hui putting some of his own money on the line, he had a strong incentive to consider eMachines' capital requirements carefully. Additionally, Hui's lender requested that he obtain a solvency opinion, which he did, from the investment bank of Houlihan Lokey Howard Zukin. Because Hui's plan to remove much of eMachines' cash reflected careful analysis by an insider contemporaneous with the Merger Date, the Court finds it highly probative. Thus, the Court finds that eMachines needed only $15 million cash for working capital as of the Merger Date, and therefore will add $165.2 million to the DCF valuation as excess cash.
Trial Ex. 51; Tr. at 320 (Anderson).
Trial Ex. 31 at R000340; Tr. at 320-21 (Anderson).
D. The Company's Reliance on Various Market Indicators
The Company urges the Court to use various market indicators of value to perform a "reality check" on Larson's DCF and impliedly, the Court's DCF. It contends the large control premium the Merger Price represented over the pre-Merger price of eMachines stock ($0.43-0.54) confirms that the Merger Price represented fair value. The Company also contends that what it believes was a "thorough and fair" auction for eMachines further corroborates that conclusion. Finally, eMachines contends that sophisticated holders of significant positions in eMachines stock had representatives on the Board and voted to approve the transaction at $1.06 per share. These holders included TriGem, America Online ("AOL"), and Idealab founder Bill Gross.
The Court embraces the concept of reasonableness checks consistent with the discussion in Technicolor. In many cases market indicia of value such as control premiums, approval by stockholding directors, and competitive and fair auctions can provide informative checks on the reasonableness of financial valuation techniques. In the circumstances of this case, however, the indicia cited by eMachines have limited utility.
2003 WL 23700218, at *43-44.
Id.
The Court gives little weight, for example, to eMachines' control premium argument. Jarrell notes that the Merger Price represented a 96% control premium and was considerably larger than the 30% average premium found in CSFB's study of 643 acquisitions of greater than $10 million consummated during 2001. Certainly the control premium paid was large. In the context of eMachines, however, the Court considers a control premium of this size more indicative of the inefficiencies in the market for the Company's stock than reflective of its value as a going concern. As of the time of the auction, only limited information regarding eMachines' turnaround had been disseminated to the market. In fact, just prior to Hui's bid, on October 23, 2001, Inouye refused to give any guidance for 2002 during eMachines' third quarter conference call. Information about the increasing 2002 revenue projections, as ultimately reflected in the 2002 Budget, were not available to the market before the Merger. The importance of this fact is magnified by the very limited, relevant operating history of eMachines at the time of the Merger. In these circumstances, eMachines' arguments are unconvincing.
Likewise, the Court finds that the auction completed in this case was not sufficiently open to ensure an adequately level playing field to warrant giving great weight to the resulting price. Price derived from a competitive and fair auction is strong evidence of fair value. The manner in which the eMachines' auction was conducted, however, causes the Court to question its effectiveness. In particular, the "auction" after Hui's bid on October 30, 2001 likely did not include all potential bidders, was conducted quite hastily, and probably reduced the likelihood that all bidders would be fully apprised of the Company's current prospects.
Union Ill. 1995 Inv. Ltd. P'ship v. Union Fin. Group, 847 A.2d 340, 357 (Del.Ch. 2004).
First, CSFB, as the Company's agent, failed to cast a wide net in attempting to attract potential buyers to the post-October 30 auction. After Hui's initial bid, CSFB contacted only the two parties that had shown interest during its initial engagement in the summer of 2001, Gores and Medion. When CSFB first shopped eMachines, however, it had a reputation for poor quality and customer service. Furthermore, it had not yet proven that it could improve its operating efficiency by increasing inventory turns and cutting SGA, as called for in its new business plan. By the end of 2001 eMachines had improved greatly in all three areas. Thus, it was at least possible, if not likely, that other suitors would have been interested in eMachines if presented with the rapidly improving state of the Company and the fact that two separate entities had expressed interest in acquiring the Company.
Gores' earlier bids for eMachines had not been disclosed to the stockholders or the public.
An auction was created, however, once Gores re-entered the bidding on November 11, 2001. The Gores bid was 24% higher than Hui's initial bid and prompted Hui to raise his bid a full 28% over his initial bid. At this relatively early stage in the bidding process, eMachines immediately set a deadline for additional bids of November 18. The deadline was only seven days after Gores re-entered the bidding and only nine days after the initial Hui offer was announced to the public. The Company never explained the rationale for this decision, which resulted in the immediate termination of an otherwise active auction. The quick termination of the bidding may have kept the price down and reduced the chances of another bidder entering the auction. Although it is impossible to know what price a longer auction might have wrung from the bidders had it been allowed to go on, the Court finds that the auction as conducted is inconclusive as to the fair value of eMachines.
In examining the fairness of an auction, the Court looks to whether the outside bidders were provided with confidential information on par with insiders or bidders favored by management. The Court assumes that as a co-founder, insider, and director, Hui had access to internal projections and other inside information. This fact is illustrated by the discussion between Hui, Inouye, and Anderson, regarding the ever-changing and nonpublic budget figures for 2002 memorialized in the November 27 email. On the other hand, it is unclear how the information regarding eMachines' improving fortunes that was made available to Gores or Medion compares to that available to Hui. The evidence suggests that Gores had access to information comparable to that provided to CSFB, but nothing more. For the reasons stated previously, the Court considers that information less than optimal for purposes of the valuation.
Gores "reviewed [the] data room the week of June 18" and was provided with a balance sheet current to July 31. Trial Ex. 23 at R000931.
These factors, taken together, prevent the Court from relying on the auction price as strong evidence of fair value and leave the Court with the difficult task of valuing the Company by other means.
Approval of a merger by sophisticated independent directors also can be strong evidence of fair value. In this case, however, the strategic relationships between eMachines and the sophisticated holders of stock it relies upon cast doubt on the relevance of their reasons for approving the Merger to the question of valuation. Both TriGem and AOL, for example, were past strategic partners of eMachines. TriGem was being pressured by its biggest customer, and eMachines competitor, Hewlett Packard to distance itself from eMachines. AOL had been a partner of eMachines under its old Internet-revenue business model, but was by late 2001, not a part of eMachines' future plans. These facts suggest that these stockholders may well have had a number of reasons for approving the Merger and thereby creating a liquidity event for themselves, other than a careful and reliable valuation analysis.
Technicolor, 2003 WL 23700218, at *45.
Even assuming conflicting motivations of AOL and TriGem (and their representatives on eMachines' Board), their conduct does cast doubt on the significantly higher value of $2.48 per share ascribed to eMachines by Larson. At the same time, the Court does not view the conduct of these former stockholders of eMachines' as inconsistent with the Court's valuation of the Company. Applying the Court's DCF model to the Management Case projections available at the time the board approved the merger produces a value of eMachines only pennies greater than the approved Merger Price. Only the application of the 2002 Budget projections, which the Court has found to be known or knowable as of the Merger Date, causes the value of eMachines to increase appreciably beyond the Merger Price. Therefore, the Court finds that, in light of all the circumstances, the acceptance of the Merger Price by sophisticated insiders, using the information they had been provided, does not undermine the reasonableness of the Court's valuation.
The Court also notes that CSFB's DCF using the Management Case yielded equity values between $1.41 and $2.22 per share. Trial Ex. 21 at R000213.
E. Petitioners' Glassman Argument
Petitioners strenuously argue that the fair value of their shares should take into account elements of damages stemming from allegedly unfair practices by eMachines. In Glassman v. Unocal Exploration Corp., the court held that a minority stockholder's only recourse in challenging a short-form merger under 8 Del. C. § 253 was appraisal, but that "fair value must be based on all relevant factors, including damages and elements of future value, where appropriate." The court further explained that if, for example, "the merger was timed to take advantage of a depressed market, or a low point in the company's cyclical earnings, or to precede an anticipated positive development, the appraised value may be adjusted to account for those factors." "[W]e have held that claims for unfair dealing may not be litigated in an appraisal. . . . [S]tockholders may not receive rescissionary relief in an appraisal."
Petitioners' PTOB at 43-47. Petitioners cite the following as instances of unfair conduct: (1) representations to shareholders and the SEC that CSFB's sensitivity case was based on management forecasts when in fact management knew it was not, (2) concealment of the repeatedly increasing 2002 projections, and (3) timing the Merger to coincide with the post-9/11 period.
777 A.2d 242, 248 (Del. 2001) (reaffirming Weinberger's interpretation of the appraisal statute's scope).
Id.
Id.
In any event, however, consistent with Glassman, this Court has considered all relevant factors that were known or knowable on the Merger Date. In this case, "all relevant factors" include the timing of the purchase just as the Company's new business model was beginning to take hold, but before much actual operating history was available or the public had much information as to the extent of its projected success. If Hui had waited until the end of 2002 to launch his bid, for example, the new business model would have had an 18 month, as opposed to an 8 month, track record on which to be judged. This led to eMachines' arguments that the 2002 Budget was "aspirational" or too speculative to be of use. Since Hui's and eMachines' own timing and evident preference for an abbreviated and quick auction created this situation, their arguments against the Court's acceptance of the assumption that the new business model would be as successful as forecasted in the 2002 Budget ring hollow. In sum, the Court has based its determination of fair value on a consideration of all relevant factors.
Although there is no need to decide the issue here, the Court notes that in the context of this case, which involves a two-step going private transaction that includes a tender offer and does not involve a majority stockholder, Glassman is arguably inapposite.
F. Valuation Conclusion
The Court concludes that the valuation of eMachines should be based on a DCF analysis. The Court's DCF reflecting the above findings yields an equity value of eMachines' operations of $89.5 million. Adding $165.2 million in cash, plus other uncontested adjustments to the DCF value, the Court holds that eMachines' total equity value is $253 million. Using the agreed upon figure of 154,612,560 outstanding shares, eMachines' fair value per share is determined to be $1.64.
III. INTEREST
In addition to determining the fair value of Petitioners' shares, the Court must determine under 8 Del. C. § 262(h), a fair rate of interest to be paid upon the amount determined to be the fair value. An award of interest serves two purposes. It compensates the petitioner for the loss of use of its capital during the pendency of the appraisal process and causes the disgorgement of the benefit respondent has enjoyed during the same period. These twin purposes can be accomplished by applying a rate of interest derived from equal weightings of the prudent investor rate and the company's cost of debt. Interest may be simple or compound, but this Court's trend has been to award compound interest because it "best reflects the purposes of Delaware's appraisal statute." Finally, it is well established that "[e]ach party bears the burden of proving an appropriate rate under the circumstances." If each party fails to fulfill its burden of proof, the court may look to the legal rate of interest for guidance, but where the record is sufficiently developed, the legal interest rate generally is irrelevant.
Gonsalves v. Straight Arrow Publ'rs, Inc., 793 A.2d 312, 327 (Del.Ch. 1998).
See, e.g., Hintmann v. Fred Weber, Inc., 1998 WL 83052, at *13 (Del.Ch. Feb. 17, 1998); Gonsalves, 793 A.2d at 327.
Taylor, 2003 WL 21753752, at *12.
Grimes v. Vitalink Communications Corp., 1997 WL 538676, at *9 (Del.Ch. Aug. 26, 1997).
Gonsalves, 2002 WL 31057465, at *10.
Gonsalves, 793 A.2d at 327.
In this case, Petitioners seek compound interest at the statutory rate of 5% over the discount rate as of the Merger Date, or 6.5%. Petitioners have, however, failed to prove that the legal rate should apply in this circumstance, arguing only that they "could reasonably seek a much higher interest rate."
Petitioners' PTOB at 47.
The Company relies on the framework elucidated in Gonsalves, to support its contention that a fair rate of interest is 6.21%. In Gonsalves, the court gave an equal weight to a prudent investor rate and defendant's cost of debt to compute a rate consistent with the dual purposes of an interest award in an appraisal case. Jarrell calculated 6.21% by averaging the prudent investor rate and eMachines' cost of debt on the date of the Merger. To find the prudent investor rate, Jarrell used a weighted average consisting of 20% stocks, 20% corporate bonds, 20% US Treasuries, and 40% money market returns. The resulting prudent investor rate was 4.5%. For eMachines' cost of debt, Jarrell began by subtracting the Prime Rate as of the Merger Date from the interest rate eMachines obtained for its Merger financing. The remainder, 3.25%, represented the risk premium associated with eMachines' debt. He then added the 3.25% risk premium back to the Prime Rate as it stood on average during the period from December 31, 2001 to February 10, 2004. The Court finds Jarrell's methodology in accordance with this Court's precedent and his proposed interest rate of 6.21% reasonable in this situation. Accordingly, the Court will award interest to Petitioners at the rate of 6.21% compounded monthly.
793 A.2d at 327 (applying the legal rate as an estimate of defendant's cost of debt because the record contained insufficient evidence to calculate it more accurately).
In Gonsalves the court used a mix of three asset classes to derive its short-term component — 90-day T-bills, 90-day commercial paper, and one-year CDs. See id. The Court, however, views the money market rate, alone, as sufficiently representative of short-term yields.
IV. CONCLUSION
For the reasons stated, the Court concludes that the fair value of eMachines common stock as of the Merger Date is $1.64 per share. Petitioners and all those having appraisal rights in eMachines shares are entitled to that amount per share plus interest at the rate of 6.21% compounded monthly from December 31, 2001 to the present.
Counsel for Petitioners shall prepare an appropriate form of Final Judgment and submit it, upon notice, for the Court's consideration.