The Delaware Court of Chancery has been on a slow but clear path toward increased reliance on the merger price in determining fair value in appraisal cases. While the court’s reliance on the merger price as the best indicator of fair value has been more frequent recently, the court appears to continue to struggle with a fundamental conflict in its approach to determining fair value. On the one hand, the court has explicitly recognized that, when a merger price has been derived through an arm’s-length sale process, involving parties with a real economic stake in the outcome and during which the target company has been actively shopped, the merger price likely represents a better indication of the target company’s value than would the results of financial valuation analyses conducted by (using the court’s words, repeated in numerous opinions) “litigation-driven” parties in appraisal proceedings or “law-trained judges”. On the other hand, the court’s willingness to rely on the merger price has been limited by the court’s interpretation of the statutory mandates that, in an appraisal proceeding, the court must (i) consider “all relevant factors” in determining fair value and (ii) exclude from fair value any value that is expected to arise from the merger itself (such as, we note, certain synergies or, it would seem, a control premium). The court has dealt with this conflict in numerous recent cases (including the latest one, Merion v. BMC Software (Oct. 21, 2015)), by: first, finding that the merger price is the best indication of value when it has been derived through an arm’s-length process with active shopping of the target company; second, conducting a discounted cash flow (DCF) analysis (because, as Vice Chancellor Glasscock stated in BMC Software, “I am required to do [so]”)—and finding (a) that the DCF result is close to the merger price and/or (b) that the DCF result is not as reliable as the merger price as an indication of value; and, third, considering what adjustment must be made to the merger price to exclude value arising from the merger—and, after finding that there is insufficient evidence to determine the amount of an appropriate adjustment, declining to make any adjustment. In this memorandum, we review the statutory and common law context for the court’s mandate to determine fair value; summarize the court’s thinking to date (as reflected in its appraisal opinions) on the requirement to consider all relevant factors; and offer our view that the court’s latest appraisal decision, Fried Frank M&A Briefing 2 BMC Software, represents an expansion of the court’s reliance on the merger price in appraisal cases (and portends likely further expansion). We also note DCF analysis-related practice points arising from BMC Software. (We will discuss the court’s approach to the requirement to exclude merger-specific value in a separate memorandum that will be available on the Fried Frank website next week.) Delaware Appraisal Statute Stockholders of Delaware corporations who dissent to a merger and perfect their appraisal rights are entitled to an appraisal hearing before the Court of Chancery, at which the court, after hearing presentations by the parties’ respective experts, will make an independent determination of the “fair value” of the shares subject to appraisal. These stockholders (and, unlike in fiduciary duty litigation, not any of the other stockholders) will be entitled to receive the court-determined “fair value” of their shares in lieu of the merger consideration. The fair value as determined by the court may be the same, higher or lower than the price per share paid in the merger. The Delaware statute provides the court with wide discretion in determining fair value. The statute prescribes no specific methodology to be used, directing only that “the Court shall determine the fair value of the shares exclusive of any element of value arising from the accomplishment or expectation of the merger,” and that “[i]n determining such fair value, the Court shall take into account all relevant factors.” Pursuant to well-established judicial precedent, fair value in the context of an appraisal proceeding is the value to a stockholder of the firm as a going concern, as opposed to the firm’s value in the context of an acquisition or other transaction. The underlying theory is that a dissenting stockholder is entitled to receive the value of what he relinquished in the merger to which he objected—i.e., his proportionate share of the corporation as a going concern. Accordingly, and as implicitly mandated by the appraisal statute, going concern value does not include, for example, synergies that are expected from the merger itself. In summary, the court must consider all data and use any valuation methodologies that the court deems appropriate under the circumstances to value the company as an independent going concern. Historical Context—Court’s Resistance (until 2013) to Using the Merger Price to Determine Fair Value Golden Telecom. Although prior to 2010 the Chancery Court had looked at the deal price as a factor to be considered in the fair value determination (at least where the price was based on an effective arm’slength sale process), the court generally did not do so after the 2010 Delaware Supreme Court decision in Golden Telecom v. Global GT. The Supreme Court held in Golden Telecom that the Chancery Court could not defer to the merger price to establish fair value because the court is charged by statute with using its independent judgment to make the determination, taking into account all relevant factors. Then Chief Justice Steele wrote: [The statute] neither dictates nor even contemplates that the Court of Chancery should consider the transactional market price of the underlying company.…[The statute] controls appraisal proceedings, and there is little room for this Court to graft common law gloss on the statute even were we so inclined.…Requiring the Court of Chancery to defer—conclusively or presumptively—to the merger price, Fried Frank M&A Briefing 3 even in the face of a pristine, unchallenged transactional process, would contravene the unambiguous language of the statutes and the reasoned holdings of our precedent. Therefore, … we reject [petitioner’s] call to establish a rule requiring the Court of Chancery to defer to the merger price in any appraisal proceeding. Orchard Enterprises. That approach was followed by the Chancery Court in In re Appraisal of Orchard Enterprises (2012), in which then Chancellor Strine, declining to give weight to the merger price, stated that the court “must use its own independent judgment to determine the fair value of the [dissenting] shares.” Strine rejected the notion that the go-shop provision of the merger agreement gave validity to the merger price as a basis for the fair value determination, stating that an appraisal proceeding (unlike a fiduciary duty case) must be focused simply on the going concern value of the company, without regard to the sale process (and the price produced by that process). He wrote: [Respondent] makes some rhetorical hay out of its search for other buyers. But this is an appraisal action, not a fiduciary duty case, and although I have little reason to doubt [Respondent’s] assertion that no buyer was willing to pay…[more than the deal price], an appraisal must be focused on [Respondent’s] going concern value. CKx—a new regime. In a departure from the usual pattern of rejecting use of the merger price, in late 2013, in Huff v. CKx (Nov. 1, 2013, affirmed in full by the Delaware Supreme Court, without an opinion, Feb. 13, 2015), Vice Chancellor Glasscock held that the court can and should look to the deal price as an important indication of fair value (subject to adjustment to exclude merger synergies), at least where: [T]here has been a thorough and effective sales process, there are no truly comparable companies or transactions to form the basis of a comparables analysis, and the revenue projections are too unreliable to form the basis of a discounted cash-flow analysis. The court found that no other valuation method was available to be used in CKx. The court concluded that the CKx projections (a key input to a DCF analysis) were unreliable because they had not been created in the ordinary course of business but, rather, according to the respondent company’s own testimony at trial, had been prepared with the objective in mind of trying to generate the highest price possible from potential buyers of the company. The court also deemed the projections to be too speculative, as the company’s interest in the American Idol television show (which accounted for 75% of the company’s cash flow) was about to expire and it was impossible to foretell whether, and if so on what terms, those rights might be renewed. The court also deemed the parties’ respective comparables analyses as unreliable because the companies and transactions utilized were not sufficiently comparable. Under these circumstances, Vice Chancellor Glasscock opined, a deal price generated by an arm’slength negotiating process may be “the best and most reliable indicator” of fair value. The Vice Chancellor explicitly interpreted the Golden Telecom decision as having rejected only an automatic presumption in favor of the use of the merger price and not as having rejected its use as one of many potentially relevant factors (or its use as the most, or even only, relevant factor) that the court may consider to determine fair value in any given case. Unacknowledged influence of the merger price. We note that, in every appraisal decision from 2010 through Huff in late 2013, while the court expressly declined to rely at all on the merger price, professing the irrelevance of the merger price in an appraisal proceeding as one of the indications of fair value, it appears that the court was influenced substantially by the merger price in cases in which the merger price had been derived through an arm’s-length third-party process (i.e., in “disinterested” transactions), that Fried Frank M&A Briefing 4 included active shopping of the company as part of the sale process. In the disinterested transactions during this period, the appraisal awards represented only modest premiums above the merger price (a 16% premium in the highest case), while in interested transactions (i.e., transactions involving a controller, or a going private transaction, so that the merger price implicitly was not derived at arm’slength), the appraisal awards represented higher, and often very significant, premiums above the merger price (ranging from 20% to 149%). (Our calculation of premiums above the merger price does not include the statutory interest included in the appraisal award.) Court’s Post-2013 Use of the Merger Price to Determine Fair Value—Two-Prong Test for Reliance on the Merger Price Of the six appraisal decisions since CKx, two involved “interested” transactions (i.e., transactions with a controller) and four involved “disinterested” arm’s-length transactions. In both of the interested transactions, the court did not use the merger price to determine fair value, relied on a DCF analysis, and found fair value to be significantly higher than the merger price. In all of the disinterested transactions, the two-prong test articulated in CKx was satisfied, and, as in CKx, the court expressly relied primarily on the merger price to determine fair value and found fair value to be at (or very near) the merger price. Interested transactions--Hesco and Cannon. In Laidler v. Hesco (June 2014), the first appraisal decision to follow CKx, Vice Chancellor Glasscock rejected use of the merger price on the basis that the transaction there, by contrast with CKx, had not involved an arm’s-length competitive process. Under those circumstances, the Vice Chancellor explained, the merger price, which had been dictated by the 90% parent, was not a reliable indicator of value. The amount determined to be fair value in this case, based on a DCF analysis, represented an 87% premium to the merger price. In Owen v. Cannon, an interested transaction involving a squeeze-out merger (with no market check, and at a price far below the value that had been indicated in third party valuations received by the company), Chancellor Bouchard, relying on a DCF analysis, determined fair value to be an amount that represented a 60% premium above the merger price. Disinterested Transactions—Ancestry, AutoInfo, Ramtron and BMC Software. In Ancestry (January 2015), Vice Chancellor Glasscock determined fair value to be equal to the merger price, relying on the fact that the merger price had been established in a competitive bidding situation (and a “check” on the merger price based on a DCF analysis that yielded a result within a few cents of the merger price). In AutoInfo (April 2015), a disinterested transaction involving a competitive public auction, Vice Chancellor Noble, relying on the merger price, held that the fair value was equal to the merger price. In Ramtron (June 2015), a disinterested transaction involving a hostile takeover bid and a thorough search for a white knight buyer (although no white knight or other competing bidder emerged), Vice Chancellor Parsons, relying on the merger price, determined fair value to be just slightly (three cents) below the merger price (after making a nominal adjustment to the merger price to exclude merger-specific synergies). Relevance of Latest Decision—BMC Software: Possibly Expanded Use of Merger Price In BMC Software, the latest appraisal decision, Vice Chancellor Glasscock determined that the appraisal amount payable to the stockholder petitioners was equal to the merger price ($46.25). The petitioners were dissenting from the merger in which BMC Software, Inc. had been taken private by a consortium of investment firms after an auction of the company with multiple competing bidders. The petitioners and the respondent company each had relied primarily or exclusively on a DCF analysis. Their respective Fried Frank M&A Briefing 5 analyses yielded widely divergent results (the petitioners’ result was $67.08; the respondent’s was $37.88). The court’s own DCF analysis yielded a result that was between the two and just above the merger price ($48). The court concluded that the merger price was the “best indication of fair value” due to the robust, arm’s-length, competitive sale process, and the court determined fair value to be equal to the merger price. (The court considered making, but found that there was not sufficient evidence to support, an adjustment to the merger price to deduct the value of merger-specific synergies.) Does BMC Software mean that the court will rely on the merger price when it is regarded as reliable, without regard to the reliability of financial valuations? The answer is uncertain. In finding fair value to be equal to the merger price in BMC Software, Vice Chancellor Glasscock emphasized that the price had been established through a “pristine” sales process with multiple rounds of auctions involving multiple bidders. Notably, the two-prong test—which, as noted above, was expressly utilized by the court in all of the appraisal decisions in the recent past in which the court relied primarily on the merger price—was not even mentioned in the BCM Software opinion. While the test was, as a substantive matter (albeit not expressly) applied, and the opinion does reflect that the court found the merger price to be reliable and the DCF analysis to be unreliable, the court’s focus was on the reliability of the merger price and there was far less discussion about the unreliability of the DCF analysis. Most importantly in our view, of the three factors cited by the court as the reasons the court was “reluctant” to rely on the result of its DCF analysis, at least two (and possibly all three) appear to have been generic concerns about the reliability of inputs to the analysis that would be applicable in any (i.e., every) case in which a DCF is utilized. The generic nature of the court’s concerns about inputs to the DCF analysis in BMC Software. In the recent past, the court has viewed the result of a DCF analysis as too uncertain to be reliable only in cases in which unusual factors have made the analysis more uncertain than is typical. In BMC Software, however, the court’s reluctance to rely on its own DCF analysis seems to have been based, at least to a large extent, on more generic concerns that would be applicable in many (or even most) appraisal cases. Discount rate. The court expressed concern about the reliability of the inputs the court had selected to determine the appropriate discount rate. The court’s concern was based on (a) a “meaningful debate” within the financial community about whether a supply side or a historical equity risk premium was generally preferable and (b) the court’s not having “complete confidence in the reliability of taking the midpoint between inflation and GDP as the Company’s expected growth rate.” These concerns go to the fundamental reliability of DCF analysis in general and would appear to be applicable in any case in which a DCF analysis is utilized. Widely divergent DCF results. The court noted a number of times in the opinion the “wildly” and “dismayingly” divergent results of the DCF valuations by the parties’ respective experts—with one result being substantially above, and the other substantially below, the merger price. As noted, significantly different results from the parties’ respective litigation-driven DCF analyses have been the norm in appraisal cases. Thus, this concern also would appear to be applicable in most or all appraisal cases. Projections. It is unclear whether the court’s concern about the reliability of the projections in BMC Software was or was not a generic concern about projections; however, at a minimum, the court’s concern has expanded the list of issues relating to projections that the court has previously found problematic. In recent past cases, the reliability of projections has been of concern to the court when the projections have not been prepared by management in the Fried Frank M&A Briefing 6 ordinary course of business (but, rather, for the sale process or for the appraisal proceeding); the management had no previous experience preparing projections; the projections were unduly biased; and/or the business was of a nature (or faced a particular issue) that made it especially difficult to make forecasts. By contrast, in BMC Software, the court found that the projections had been prepared “reasonably”; that they did not reflect undue bias; and that they were possibly “particularly reliable” due to the fact that the company had a subscription-based business. The court’s concern, instead, was that the projections had not been historically accurate as compared to actual results. The court provided little detail or discussion on this point—stating that the respondent company had “demonstrated” that its projections had historically “fallen short” of actual results on the revenue side. Thus, it is not clear whether or not the court viewed the extent, frequency and/or type of inaccuracy as greater in this case than might be expected in many other cases. In summary, to the extent that the court’s concerns about its DCF analysis in BMC Software are generic in nature (or, in the case of the projections, to the extent the court has, at a minimum, expanded the types of concerns that make it reluctant to rely on a DCF analysis), these concerns would appear to be applicable in many or most cases in which a DCF analysis is utilized. Thus, in our view, BMC Software suggests that the court may be moving toward a reluctance to rely primarily on a DCF analysis in many appraisal cases, even when the DCF analysis is not regarded as particularly unreliable. BMC Software does not seem to indicate that the court will determine fair value without regard to financial valuation methods such as a DCF analysis; indeed, the court stated that it was “required” to conduct a DCF analysis in order to take all relevant factors into account. However, given that, in BMC Software, the court did not mention the two-prong test; the court emphasized the reliability of the merger price due to the competitive auction process; the court expanded the types of concerns it has about projections; and the court’s specified concerns about the DCF analysis were generic in nature, BMC Software appears to suggest that the court has expanded even further its openness to primary reliance on the merger price. In our view, the court may be moving toward primary reliance on the merger price in any case in which the target company has been actively shopped, whether or not the financial valuation analyses are regarded as reliable. The court may now be expected to rely primarily on the merger price (with adjustment for merger-specific synergies) at least in the factual context that was applicable in BMC Software, which was: a merger price “generated by the market through a thorough and vigorous sale process” (in BMC Software, multiple rounds of a competitive auction); the parties’ respective DCF analyses yielding “dismayingly divergent” results, with one result far below and the other far above the merger price; the court’s having a “lack of confidence” in the inputs the court chose for its own DCF analysis (in BCM Software, with respect to the reliability of the projections and the appropriate discount rate); and the court’s DCF analysis yielding a result that was not far from the merger price (although this (potentially important) factor was not addressed by the court). Open issues Fried Frank M&A Briefing 7 A number of open issues remain after BMC Software with respect to the court’s reliance on the merger price to determine appraised fair value: Will the court’s increased inclination to use the merger price as the most important factor in determining fair value expand to include any transaction with active shopping (or, possibly, any effective market check), whether or not there were reliable inputs for a financial analysis? Reliance on the merger price whenever there has been a sale process that provides confidence in the merger price as a reliable indicator of value—thus, whether or not the company projections and other DCF inputs are regarded as reliable—no doubt holds allure. That approach would short-cut the considerable difficulties inherent in a DCF analysis (particularly that many of the inputs are inherently uncertain and that the analysis can yield widely varying results with even small changes to the inputs). We note (as the court has) that, even when a DCF analysis is conducted in a cooperative, non-litigation context, there may be considerable uncertainty about the reliability of the results. Further, the court has often expressed frustration with “litigation-driven valuations” submitted by the parties’ experts and the inadequacy of “lawtrained judges” evaluating and conducting DCF analyses. Moreover, as discussed above, as a practical matter, in the case of disinterested transactions, the court’s results (i.e., fair value equal or close to the merger price) would not necessarily differ from those that have pertained historically when the DCF analysis has been used. Will the court resolve the conflict presented by the statutory mandate to consider all relevant factors in a way that permits it to rely primarily on the merger price? Might the court, for example, determine that, in a pristine market process, the merger parties themselves, through the sale process, already have taken all relevant factors about value into account (including DCF, comparables and other valuation analyses)—with the merger price being a reflection of that process, conducted by parties with more information about the target company and greater financial sophistication than the court, and with a real economic stake in the determination? If so, would the mandate that the court consider all relevant factors in determining fair value then be satisfied by the court’s considering all relevant factors relating to the sale process to determine whether it had been inclusive and informative enough so that court was confident that the merger price itself reflected all relevant factors? (We note that this would be a far different standard than would apply in the context of the court’s inquiry with respect to the sale process in the concededly very different context of fiduciary duty litigation.) To the extent that the court continues or expands further its reliance on the merger price, it remains to be seen what the impact will be on the statutory mandate to exclude from the merger price value that arises from the merger itself. Our view, consistent with the court’s record so far, is that it will be the rare case in which the court makes an adjustment to the merger price to exclude merger-specific value. (For a discussion of this issue, please see our memorandum that will be available on the Fried Frank website next week.) Could the shopping process ever be strong enough for the court to use the merger price to determine fair value in an interested transaction? To date, the appraisal cases involving interested transactions have included, in the court’s view, no (or weak) shopping of the company. Moreover, the amount by which the court’s fair value determinations (using financial valuation analyses) have exceeded the merger price has generally corresponded to the apparent strength of the sale process. It remains to be seen whether there are circumstances under which the sale Fried Frank M&A Briefing 8 process could be robust enough in an interested transaction for the court to rely primarily on the merger price to determine fair value. Would the merger price or the financial valuation take precedence in the case of a transaction in which there was active shopping, but also the inputs for the financial analyses were reliable and they substantially exceeded the merger price? How would the court respond if the financial valuations—if based on reliable company projections and/or sufficiently comparable companies and transactions—led to values materially in excess of the merger price? In such a case, either one would have to question whether what was viewed as an effective shopping process had indeed been “effective”, or one would have to assume that the financial analysis inputs that were viewed as “reliable” had, for whatever reason, not reflected the market’s selection of inputs. Presumably, the course the court would choose in these circumstances would depend on the court’s view of the reason for the discrepancy between the merger price and the financial valuations. DCF-Related Practice Points Arising from BMC Software Petitioners should consider the game theory of arguing for a DCF result that is less dramatically above the merger price than has typically been the case. Based on the court’s continued and increased reliance on the merger price in appraisal cases, petitioners should consider whether, in cases in which the sale process was arm’s-length and included an effective shopping process (or, possibly, other effective market check), their expert’s DCF analysis might be accorded greater credence by the court if the result were less, rather than more, dramatically in excess of the merger price than has typically been the case. Possible new basis for challenging reliability of company projections. In BMC Software, unlike previous appraisal cases, the court questioned the reliability of projections based on the fact that, while they had been prepared reasonably and did not reflect undue bias, they historically had turned out to be inaccurate as compared to the company’s actual results. Accordingly, the historical accuracy of the company’s projections should be considered as a basis for a challenge to their reliability in a DCF analysis—although the court offered no guidelines as to the extent of historical inaccuracy necessary for the concern. Comparables analysis. Neither expert in BMC Software relied on a comparables analysis. One of the experts conducted a comparables analysis as a “check” on the DCF analysis, but did not even submit the comparables analysis to the court. We note that the parties in appraisal cases now rarely rely on comparables analyses (undoubtedly due to the court’s consistent concerns in past cases about their reliability, based on uncertainty as to the extent to which the companies and transactions utilized in the analysis were actually comparable to the company and transaction at issue). Selection of DCF inputs. In conducting its own DCF analysis in BMC Software, the court indicated its view on the following: o Equity risk premium. The court adopted a supply-side equity risk premium on the basis that that has been the court’s “practice of the recent past” (citing its use in two cases, Golden Telecom (2009) and Orchard Enterprises (2014)). The court expressed a preference for using forward-looking data (as opposed to the historical or the supply-side approach), but noted— Fried Frank M&A Briefing 9 and expressing its reluctance to rely on the DCF analysis because of—the continuing academic debate about which method is more reliable. o Terminal growth rate. The court reaffirmed the view that inflation is generally the floor for a terminal value in a DCF analysis. In the absence of evidence that suggested that the growth rate should be limited to inflation, the court chose the midpoint between inflation and GDP as the growth rate (3.25%)—although, again, the court expressed reluctance to rely on the DCF analysis because of the uncertainty inherent in simply choosing the midpoint. o Excess cash. The court found credible the testimony at trial that the company was preserving its cash balance in contemplation of the merger closing and that, but for the merger, the company would not have conserved an extra $127 million in cash. The court agreed with the respondent’s adjustment in its DCF analysis to excess cash for the expense associated with repatriating cash held abroad, despite the petitioners’ pointing out that the company’s 10-K stated the company’s intent to maintain cash balances overseas indefinitely. The court found: “These funds, however, represent an opportunity for the Company either in terms of investment or in repatriating those funds for use in the United States, which would likely trigger a taxable event.” Therefore, it was appropriate to include a “reasonable offset” in the DCF analysis for the tax associated with repatriating those funds, the court found. o Stock-based compensation. The court found it reasonable that the company treated stockbased compensation (SBC) as an expense in the DCF analysis. The court noted the importance of SBC in a DCF analysis of a technology company (as BMC was). While the petitioner argued that this approach vastly overstated the cost of SBC, the court noted that the company’s history of buying back stock awarded to employees to prevent dilution was “in line with a cash expense” and that, in any event, the petitioner’s methodology was not an acceptable alternative because it failed to account for future SBC. o M&A expense. The court found that BMC Software intended to continue with its inorganic growth strategy of M&A transactions and that “management’s projections incorporated M&A in their forecast of future performance.” Therefore, the court concluded, in this case, in the DCF analysis, “the expenses of that M&A must be deducted from income to calculate free cash flow.” The court noted that the multiple potential buyers through the course of the sale process “must have similarly determined that tuck-in M&A was embedded in the company’s growth projections, or else those buyers would have been foregoing up to $1.8 billion in value by not topping the Buyer Group’s winning bid” (emphasis in original).