LongPath Capital, LLC v. Ramtron International Corp., 2015 Del. Ch. LEXIS 177 (June 30, 2015)

The Delaware Court of Chancery was a hive of activity during summer 2015. In two months, the court issued a number of key valuation decisions—all long and deep. The following statutory appraisal ruling falls in line with a suite of other decisions in which the court held the merger price was a more accurate indicator of value than the values achieved with other techniques, including the often-favored discounted cash flow (DCF) model. In this case, irreparably compromised cash flow projections rendered the DCF meaningless.

Hostile deal. The petitioner was an investment vehicle that pursued an appraisal arbitrage strategy. The respondent was a fabless semiconductor company that produced F-RAM. “Fabless” means the company outsourced the manufacture of the silicon wafers used in its products. RAM stands for random access memory. The company’s F-RAM was a relatively unique product; it had fast read and write speeds, could be written to a high number of times, and used low power. Also, it retained memory if power were lost.

The company was the target of a hostile bid by a third party, Cypress Semiconductor Corp. (Cypress), which was not a party to the litigation. In March 2011, when the subject was experiencing serious cash management problems, Cypress made a nonpublic written offer for $3.01 per share, which the company rejected as inadequate. The offer was a 37% premium over the then closing price of the company’s stock.

In mid-June 2012, Cypress issued a bear hug letter offering to buy the company for $2.48 per share. Again, the company rebuffed the bid. About a week later, Cypress commenced a hostile tender offer at $2.68 per share. This offer, too, represented a 37% premium over the company’s stock price, which by then had fallen. The company’s board again found the offer inadequate and announced it was exploring strategic alternatives.

Meanwhile, in July 2012, the company acknowledged that its second-quarter 2012 earnings were below expectations and there was a good possibility that the company would come in below the full-year 2012 estimate by at least $10 million. That same month, Merriman Capital, the only analyst covering the company, stated that, if Cypress were to pull its offer, the company’s shares might return to the $2.00 range or even lower.

While Cypress pursued its hostile bid, the company engaged a financial advisor to find other potential buyers and explore other strategic options. The advisor contacted over 20 possible suitors, including Cypress. Six parties were sufficiently interested to sign nondisclosure agreements, but, ultimately, none made a firm bid, except for Cypress. Two serious contenders expressed concern over the company’s cost structure.

Cypress, on the other hand, increased the offer to $2.88 per share and extended the term of the tender offer. In fall 2012, Cypress and the company were negotiating in earnest, settling on $3.10 per share as the final transaction price. The parties signed the merger agreement in September 2012, and stockholders approved it in November 2012.

The merger price was 25% higher than Cypress’s starting bid. It represented a 71% premium over the company’s unaffected stock price of $1.81.

A month after the merger announcement, the petitioner bought shares in the company. Three weeks after the transaction closed, it filed suit, claiming the market undervalued the shares and asking the Delaware Court of Chancery for a fair value determination.

Company’s operational problems. Because the company did not manufacturer its products, its life and success depended on its relationship with its foundry. At the time of the merger, the company had a contract with Texas Instruments (TI). Before that, it had a different primary foundry. Transitioning from the first foundry to TI was a difficult process that took seven years, causing problems in day-to-day business, related to the company’s sales and revenue recognition models.

The company sold mostly to distributors that sold to end-users. And it used a point-of-purchase model that recognized revenue when the product shipped rather than when the product sold to the end-user. At trial, witnesses explained that this approach made it difficult to forecast future sales because distributors served as buffers and could mask actual demand. Specifically, the difficult transition to a new foundry forced the company to place its customers—mostly distributors—on allocation. Concerned about supply shortages, the distributors reacted by overordering. By 2011, the company faced a “massive inventory bubble,” overrecognition of revenue, and a cash crunch since the company had to pay for the extra inventory it ordered. During 2011 and 2012, the company either missed or had to renegotiate its loan contracts with its primary lender. Also, in July 2011, the company undertook a secondary public offering of about 20% of its outstanding shares. This transaction occurred at $2 per share. The company used the proceeds to pay off its excess inventory. However, by spring 2012, the company faced another cash crisis, which continued until the signing of the merger agreement.

Another “pitfall” of the point-of-purchase system was that it allowed for channel stuffing. The record showed that company management, concerned over making the revenue forecast in the first quarter of 2012, stuffed excess inventory into its distribution channels. As a result, the second-quarter numbers suffered.

Management projections. Prior to 2012, the company had created five-quarter forecasts but never multiyear forecasts. Witnesses explained that the semiconductor industry made forecasting difficult because of the stiff competition among players, technological change, and “macroeconomic factors.” But, in June 2012, just days after Cypress announced its public bid, the company’s new CEO ordered an equally new management team (most of the executives had been in their jobs for less than two years) to create long-term projections. He specified he wanted a “product by product build up, with assumptions, for it to hold water in the event of a subsequent dispute.” He also wanted the projections done by using a point-of-sales approach, as opposed to the company’s customary point-of-purchase method. At trial, the company’s CFO explained the projections had a dual purpose: “marketing the company to a white knight and creating inputs for a DCF analysis.” The head of sales, at the time, noted that even the car industry was more predictable than the semiconductor industry and proposed to “simply plug in 30% CAGR” for “out years.”

Polar opposite views of the company. In trial, the parties presented two polar opposite views of the company—which found expressions in the “widely disparate” value conclusions the parties’ appraisers reached. According to the petitioner, the company was “on the verge of taking off like a rocket.” It had strong intellectual property protections for core products and a successful new management team; also management projections reflected the company’s excellent business prospects, the petitioner contended. According to the company, however, it then faced a bleak future and might have folded if the Cypress deal had not ended well.

The petitioner’s expert claimed the company’s stock price was $4.96 per share. This price, the court pointed out, was more than 274% of the company’s unaffected stock price. The expert used a combination of the DCF analysis and a comparable transaction analysis to arrive at this value conclusion. He weighted the former at 80% and the latter at 20%.

The company’s expert arrived at a fair value of $2.76 per share. He rejected a DCF analysis, finding the management projections were overly optimistic and said there were no comparable transactions. Instead, he used the transaction price and backed out synergies. Despite his misgivings about the management projections, he performed a DCF as a check, which produced a $3.08-per-share value.

Both experts agreed that there were no comparable companies.

DCF lacks sound foundation. “Much has been said of litigation-driven valuations, none of it favorable,” the court said in reaction to the parties’ positions. It emphasized that the petitioner rejected the merger price but bought its shares after the merger’s announcement, “thereby effectively purchasing an appraisal lawsuit.” Such arbitrage can be lucrative when the market in fact undervalued the company, the court said. But, in this case, the petitioner invested so little that, even if it prevailed in this action, it would probably not break even because of the litigation expenses. As for the company, the court stressed that it “submitted an eyebrow-raising DCF that, based on projections its expert presumed were overly optimistic, still returns a ‘fair’ value two cents below the Merger price.”

The court agreed with the company’s expert that the DCF model was inappropriate in this case. “The foundational inputs of a DCF are the company’s cash flows,” the court said. Here, there were numerous reasons to question the accuracy of the cash flow projections underlying the petitioner’s DCF study. Most obvious, they were prepared by a new management team, not in the ordinary course of business, and using a methodology the company had never used before. What’s more, the record showed that, even using traditional methods, management’s ability to forecast its own business more than two quarters was “of middling quality,” the court pointed out. At trial, the company’s CEO described the quality of the projections he saw when he assumed his position, in early 2011, as “probably mediocre.” When asked whether he succeeded in making improvements, he said, “No.” Part of the difficulty stemmed from a lack of understanding the market, he suggested.

The court also noted that the projections at issue were based on unrealistic assumptions regarding the company’s ability to transition to another foundry that was to serve as a second foundry. The company had reached a management agreement with that foundry in late July 2012; the projections assumed the transition would take place in early 2013 even though the first transition took seven years and went so badly that it caused the company to place distributors on allocation, the court said with emphasis. Moreover, establishing a foundry requires substantial financial resources, which the company did not have in July 2012, the court pointed out. To assume the company could expect to start commercial production at the second foundry in 60 days and start experiencing cost savings within six months, as the projections did, was not reasonable.

Moreover, the management projections relied on inaccurate 2011 and 2012 revenue figures. Because of customer allocation issues and channel stuffing, they did not reflect actual demand for the company’s products. Also, the projections did not align with the company’s historical performance; they stated revenue growth and gross margins that required dramatic improvements to the business when there were no changes. The court said that the final nail in the coffin for the management projections was that they were not prepared in the ordinary course of business. There was testimony that the company used other sets of projections for managing the company’s finances, including providing estimates of revenue and cash flow numbers to the company’s bank. The contested projections were in anticipation of litigation as well as to shop the company to potential white knights, the court concluded.

‘Dearth of data points.’ The court also rejected the comparable transactions analysis the petitioner’s expert had done and by which he arrived at an implied value of $3.99 per share. He used two transactions involving companies that produced memory products and that also did not have their own foundry. He determined multiples based on three metrics: (1) equity value/last 12 months’ revenue (EV/LTM); (2) equity value/next 12 months’ forecasted revenue (EV/NTM); and (3) EV/NTM + 1. Averaging the multiples, he generated an implied value for the subject company.

The company’s expert noted that one of the companies featured in the transactions might not be comparable because the proxy statement for the transaction did not show the subject company on the list of comparable companies. But he also observed that the multiples for that company indicated the merger price in the transaction at issue was fair value.

A “dearth of data points” raises doubt about the reliability of the comparable transactions approach, the court noted. In the past, the court had rejected analyses that used as few as five and two transactions. Here, there were only two data points and the multiples differed considerably, the court said. The EV/LTM multiple produced a synergy-adjusted per-share value range of $3.99, with per-share values of $2.74 to $6.13. The EV/NTM multiple indicated equity values of $2.72 to $4.55, a range of $1.83. The EV/NTM + 1 multiple generated a closer range of $3.27 to $4.53.

These concerns and the fact that the petitioner’s expert himself accorded little weight to the value stemming from the approach militated against using it to determine the company’s fair value, the court concluded.

Transaction produces reliable value. Under case law, when a sound transaction process likely results in an accurate valuation of the acquired company, the court may give the merger price 100% weight, the court stated. It determined that this was the situation here, even though only one company, Cypress, actually made a bid and even though the merger was a hostile deal. For one, the company repeatedly resisted the bidder’s offer causing Cypress to raise its price five separate times, the court said with emphasis. Ultimately, management succeeded in extracting a substantial premium—25% higher than Cypress’s starting offer—from the bidder.

Also, during the three months in which Cypress pursued its hostile bid, the company tried hard to sell itself “to anyone but Cypress,” the court said. But even though the company’s financial advisor contacted about two dozen potential buyers, six of whom signed nondisclosure agreements, other than Cypress, no one bid. The lack of interest had to do with the company’s operative reality, not with flaws in the merger process, the court concluded. Therefore, the merger price was a reliable indicator of value.

Synergy adjustment. The final step in the fair value analysis was excluding any merger-specific value, the court noted. Here, the company’s expert, who had advocated for the merger price as the indicator of value, deducted $0.34 per share to account for synergies. In contrast, the petitioner’s expert claimed the synergistic value was only $0.03 per share. He noted that negative revenue synergies and transaction costs had to be added back to the value the opposing expert determined for synergies. And the petitioner contended the deduction was applicable to the value Cypress attributed to the subject company, somewhere between $3.90 and $5.44, rather than the merger price.

The court disagreed with the latter proposition, noting it ran counter to the applicable statute, which required the court to determine fair value exclusive of any element of value attributable to the merger. The court also rejected the figure—over 10% of the transaction price—the company’s expert suggested to account for synergies. It was based on a marketwide analysis of premiums paid by financial as opposed to strategic buyers. That analysis, the court said, “does not tell me anything about this specific transaction,” which was the focus in a statutory appraisal action. Here, there was evidence that the acquiring company expected negative synergies of between 10% and 15% of revenue. Accordingly, the court used the petitioner expert’s $0.03 synergy figure to arrive at a fair market price of $3.07 per share.