We’ve written before about the SWS appraisal case, decided in mid 2017. After the ruling, petitioners appealed to the Delaware Supreme Court. On Wednesday, February 21, the Delaware Supreme Court held oral argument (which you can watch on this site). Part of the argument focused on the concept of size premium – a primer on which is available here – and which is being contested in the SWS appeal. For more on the SWS oral argument, see Law360 [$$$].
Further to our prior post about Delaware’s two new appraisal decisions, SWS Group was a small, struggling bank holding company that merged on January 1, 2015 into one of its own substantial creditors, Hilltop Holdings. Stockholders of SWS received a mix of cash and Hilltop stock worth $6.92 at closing. Vice Chancellor Glasscock rejected the sale price as an unreliable indicator of fair value and performed his own DCF analysis, setting the fair value at $6.38, a price 7.8% below the merger price.
At trial, the petitioners persuaded the court that its critiques of the sale process were valid. However, the stockholders failed to persuade the court that SWS was on the verge of a turnaround, as the court instead determined that SWS consistently underperformed management projections and, given its structural problems, a turnaround was simply unlikely.
Problems with SWS’s Projections and Performance
As was true in PetSmart, SWS had not historically performed long-term projections, but only created annual budgets that aggregated projections from individual business sectors. Those single-year projections were then extrapolated into three-year “strategic plans” that assumed the annual budgets would be met. The court found, however, that SWS never met those budgets between 2011 and 2014. Also, despite straight-line growth assumptions in the management forecasts, SWS failed to hit its targets and continued to lose money on declining revenues. The various problems facing the company led the court to embrace the respondents’ theory that SWS would continue to face an uphill climb given its relatively small size, which prevented it from scaling its substantial regulatory, technological, and back-office costs.
Hilltop’s Influence on the Sale Process Rendered Merger Price Unreliable
Even before SWS launched its sale process, Hilltop was interested in buying the company (unbeknownst to SWS). Also, since Hilltop had observer status on SWS’s board, it had unique access to SWS’s board meetings and management not available to others. The court found that Hilltop’s acquisition theses were driven by synergies, as it viewed its acquisition of SWS as resulting mainly in cost savings by reduction of overhead. A Special Committee was formed after Hilltop made its initial offer in January 2014, and the court found that even though the committee engaged legal and financial advisors, the management projections that evolved in the sale process were still overly optimistic and unrealistic about SWS’s projected growth. Only two other bidders emerged, one of which was found not to be credible and the other continued to bid through March 2014 despite apparent pressure by Hilltop to proceed with its deal. Finally, in response to Hilltop’s unilateral March 31, 2014 deadline, the board decided to accept its offer, which at that time was valued at $7.75, consisting of 75% Hilltop stock and 25% cash. As of closing on January 1, 2015, the value dropped to $6.92 per share based on a reduction in Hilltop’s own stock price.
Another factor making the deal price unreliable was that Hilltop was a creditor of SWS pursuant to a Credit Agreement. That agreement contained a covenant prohibiting SWS from undergoing a “Fundamental Change,” which was defined to include the sale of SWS. The agreement thus conferred upon Hilltop a veto right over any competing offers, which right Hilltop refused to waive during the sale process.
Valuation Model & DCF Inputs
The court undertook its own DCF analysis, on which it relied exclusively. The court refused to put any weight on petitioners’ comparable companies analysis, finding that the comp set diverged too much from SWS in terms of size, business lines, and performance to be meaningful. The court held that SWS’s unique structure, size, and business model – particularly its composition of a broker-dealer business alongside its banking line – rendered the stockholders’ selected peers not truly comparable.
In performing its DCF valuation, the court used the existing three-year projection period in the management projections, rejecting the stockholders’ argument that SWS had not yet reached a “steady state” and that an additional two years was needed to normalize SWS’s financial performance. The court found that SWS’s declining revenues in the period leading up to the merger deprived it of any basis to assume (unprecedented) straight-line growth beyond the existing three-year projection period. In addition, the court found that the exercise of warrants three months prior to the merger pursuant to the Credit Agreement, which resulted in a change to SWS’s capital structure by cancelling debt in exchange for new shares, was part of SWS’s “operative reality” for purposes of the fair-value determination. This ruling differed from other cases, such as BMC Software and Gearreald, where changes to the company’s balance sheet resulting from actions by the company solely in expectation of the merger – like the company paying off its debt – was not considered to be within the company’s operative reality.
As to the other DCF inputs, the court adopted the respondents’ perpetuity growth rate of 3.35%, which was the midpoint between the long-term inflation rate of 2.3%, and the long-term economic growth rate of 4.4%. In selecting the appropriate equity risk premium, the court observed that whether to use supply-side or historical ERP should be determined on a case-by-case basis. Nevertheless, it found supply-side ERP appropriate as the “default” method in recent Delaware chancery cases, unless a party provided a compelling reason to use historical ERP. With regard to beta, the court found fault with both side’s approach. The respondents’ expert looked at two years of SWS weekly stock returns, which measurement period included a “merger froth” and too much volatility to be reliable. The petitioners’ expert, in contrast, surveyed multiple betas and used a blended median; even though the court found that this approach relied on comparable companies that were not truly comparable, it nevertheless adopted this beta – despite its apparent drawbacks – as the one more closely in line with the record evidence. Finally, to determine size premium, the court took the midpoint of both side’s decile (which was 3.46%), finding that using market capitalization is generally appropriate for public companies (the respondents’ approach), and yet SWS’s capital structure, including its substantial in-the-money warrants and the outsized influence of its major creditor, made it more like a private company and not susceptible to a market cap approach (the petitioners’ argument).
In reaching its final determination of $6.38, the court said that a sub-merger price award was not surprising here given the synergistic nature of the transaction. Also, given the award of statutory interest, which runs from the January 1, 2015 consummation date, it appears that the petitioners will ultimately recover more than the merger price after all.
In Farmers & Merchants Bancorp, an appraisal case involving a small closely-held community bank that was sold in a stock-for-stock deal valued at $83 per share, Chancellor Bouchard disregarded merger price, as well as the “wildly divergent valuations” of both sides’ experts. He arrived at an independent valuation of $91.90 per share based on his own discounted net income analysis (which is similar to a discounted cash flow but does not adjust net income for non-cash income and expenses and does not consider cash outflows for capital goods).
The court rejected the merger price as a reliable indicator of fair value because the merger was not the product of an auction and was not conditioned on obtaining the approval of a majority of the minority shareholders. Rather, the sale was driven by the same family that controlled both the target and the acquiror; even though the target formed a special committee to negotiate on behalf of the minority stockholders, the court was not confident that the negotiations were truly arms-length.
Likewise, the court rejected the comparable transaction analysis that the petitioner’s expert put forth. The petitioner’s expert calculated the median P/E ratio from the eight most comparable companies out of a pool of 160 community bank acquisitions that had taken place over the prior two years. The court found that the eight selected banks were good comparables but rejected the analysis because the expert failed to adjust for synergies that were potentially incorporated into the merger price of those banks, despite strong evidence from several witnesses that community bank mergers typically do include synergies. As to the respondent’s expert, the court rejected its comparable transaction multiples because the choice of selected comparables was suspect – it excluded 15 regional banks that would have raised the average P/E ratio significantly – and found that its guideline public company valuation was unreliable since most community banks were not publicly traded, and even the publicly traded shares were less liquid than non-community banks.
In light of these facts, the court gave no weight to the merger price or either expert’s analysis, relying entirely on its own discounted net income analysis, which projected a stream of income using a single year of earnings and applied a long-term growth rate, while using a discount rate calculated under CAPM. In determining the equity risk premium, the court chose the long-term supply-side as opposed to the historical premium, citing then-Vice Chancellor Strine’s ruling in Golden Telecom that the professional and academic valuation literature favored that approach.
Finally, the court adopted the respondent expert’s 3.0% terminal growth rate over the 4.375% rate suggested by the petitioner. Citing Owen v. Cannon, the court recognized that Delaware precedent favored a perpetuity growth rate that is a premium, such as 100 basis points, over inflation. The court also found that the 3.0% rate was consistent with the annual growth rate projected in the target’s strategic plan, which reflected its limited growth potential in a county with a declining population and stagnant economy. That rate likewise comported with the perpetual growth rates used to value mature companies in several recent cases.
It remains to be seen how much precedential value a decision like this will have on the public M&A that is more commonly the subject of Delaware appraisals, but once again the court has rejected merger price and undertaken a truly independent valuation.
The July 2015 article “Appraisal Arbitrage – Is there a Delaware Advantage?” by Gaurav Jetley and Xinyu Ji of the Analysis Group analyzes the extent to which economic incentives have improved for appraisal arbitrageurs in recent years, which the authors believe helps explain the “observed increase” in appraisal activity. The article concludes that appraisal arbitrageurs enjoy an economic benefit by delaying their investment until after the record date, as they are privy to better information about the target’s value while minimizing their exposure to the risk of deal failure. The study also finds that the Delaware Chancery Court utilizes a lower equity risk premium than do the financial advisors handling the deal itself, resulting in another benefit to arbitrageurs (and, one would think, historical holders as well). Finally, the authors conclude that the statutory interest rate more than compensates appraisal petitioners for the time value of money.
Based on these findings, the authors propose several policy recommendations, including (i) a limitation on the arbitrage strategy by reducing stockholders’ ability to seek appraisal for shares acquired after the record date, and (ii) enactment of the proposal by the Council of the Delaware Bar Association’s Corporation Law Section (which the Delaware legislature has not enacted) to allow appraisal respondents to prepay claimants some portion of the merger consideration in order to limit the interest accruing during the pendency of the case. On this latter point, the authors likewise acknowledge that allowing such prepayment is tantamount to funding claimants’ appraisal actions, thus potentially spurring on funds to increase their arbitrage strategy as they can redeploy such prepaid capital to the next case.
Vice Chancellor Glasscock issued his valuation decision this week in the BMC Software case, which we have previously blogged about concerning its threshold ruling rejecting any share-tracing requirements and thus allowing appraisal arbitrageurs to proceed with a valuation case. As we have previously reported, Merion Capital was seeking a 45% premium to the merger price, while BMC Software argued that the fair value was actually far below the merger price.
Merger Price Reflected Fair Value in This Deal
In rendering his decision, the Vice Chancellor once again looked to merger price as a measure of fair value, as he did in the Ancestry.com decision. Perhaps most interesting was his discussion of synergies toward the end of the opinion, in which he hypothesized that if Company B, holding a patent on the bow, found it advantageous to acquire Company A, a manufacturer of arrows, “synergies could result from the combination that would not have composed part of the going concern or the market value of Company A, pre-merger,” in which case Company B might value Company A more highly than the market ordinarily would. In this situation the court would be required to deduct any such synergistic value from the merger price. However, in this particular case, the acquisition was not strategic but financial, and while Respondent pointed to tax savings and other cost savings that it claims it would have realized as a private entity, the court refused to discount the merger price for any synergies.
The court further noted that receiving fair value is not necessarily the same as going concern value, and that any tax or other take-private savings may not be subject to exclusion from the awarded price as synergistic, but those savings could indeed be excludable from the going-concern value. The evidence in this case did not demonstrate any specific dollar-per-share value attributable to such savings, so the court refused to so discount the merger price. In a footnote, Vice Chancellor Glasscock said that the requirement to reduce from fair value any “non-speculative increases in value requiring a change in corporate form” was an “artifact of the policy decision to engraft ‘going concern’ valuation onto the explicit language of the appraisal statute itself,” citing Union Illinois.
The court also rejected Respondent’s request to deduct synergies based on take-private cost savings because they required a 23% internal rate of return in their business model to justify the acquisition, which raised the question of “whether the synergies present in a going-private sale represent a true premium to the alternatives of selling to a public company or remaining independent.” Thus, it was unclear whether any alleged going-private savings outweighed the buyer’s rate of return that was required “to justify the leverage presumably used to generate those savings.”
The Court’s DCF Exceeded Merger Price
The parties relied exclusively on their own DCF models, finding a comparable companies or precedent transactions analysis unreliable. Likewise, the court did its own DCF analysis and came up with a $48 per share price, exceeding the $46.25 merger price that it ultimately found to be fair value. The Vice Chancellor said that he was reluctant to use his own valuation and instead deferred to merger price, given the optimism inherent in the management projections; the raging debate within the academic community over the proper equity-risk premium to apply (thus undermining the reliability of his discount rate); and the difficulties in predicting the accurate terminal growth rate, which could be anywhere in between the floor of inflation and GDP (here, the court had picked the midpoint of those two measures in setting its own terminal rate). In this respect, the court’s DCF valuation, which exceeded merger price, was different from that in Ancestry.com, where the court’s own DCF came up just short of the merger price.
In examining the merger process itself, the court found the process to be robust insofar as there were two auctions conducted for several months each, and there were a total of five financial sponsors and eight strategic entities considering the acquisition. There was a go-shop clause, as a result of which the company contacted sixteen bidders — seven financial and nine strategic — which resulted in no alternative offers. And finally, in an arguable blurring of the line between fiduciary duty actions and appraisal rights (a subject on which we’ve posted several times before), the court explicitly looked to the settlement of the class action fiduciary duty litigation as an indication that the process was found to be free of any irregularities or fiduciary duty violations. In particular, the court found significant that the company’s activist investor, Elliot Associates, who had pressured the company to sell, was also forced to conclude that the auction itself was “a fair process.”
Other DCF Valuation Metrics
As to some of the basic valuation metrics (which were used to calculate the court’s own DCF that it ultimately refused to utilize in favor of the merger price):
- The court adopted a supply-side equity risk premium, finding that it had already been done in Golden Telecom and Orchard Enterprises and is thus indicative of the “Court’s practice of the recent past.” The court found a preference for using forward-looking data as opposed to the historical or the supply-side approach, notwithstanding the continuing debate within academe concerning the more reliable method.
- The court reaffirmed that inflation is generally the floor for a terminal value, and here, since there was no evidence to suggest that the growth rate should be limited to inflation, the court ultimately chose a growth rate at the midpoint of inflation and GDP, which was 3.25%.
- The court found it appropriate to include a reasonable offset for the tax associated with repatriating offshore cash, rejecting Petitioner’s argument that the company’s plan to keep that money offshore indefinitely should translate to no offset at all. We have seen this same point argued (and now pending before the court) in the Dell appraisal case as well.
- As is true of many tech companies, BMC had a sizable stock-based compensation (SBC) policy, which the court found was required to be accounted for, and further found reasonable to treat as an expense, particularly because this practice was expected to continue into the future. The court thus agreed with Respondent’s expert, who treated SBC as a cash expense, as opposed to Petitioner’s expert, who didn’t account for future SBC at all.
While the court seemed to take a jab at Petitioner for being “arbitrageurs who bought, not into an ongoing concern, but instead into this lawsuit,” nothing in this opinion referred back to or otherwise altered its prior ruling allowing appraisal arbitrage to proceed unfettered by a constraint such as the share-tracing requirement that BMC had asked the court to impose.
Prior posts in our “Valuation Basics” series have examined the various components of the cost of equity capital under the Capital Asset Pricing Model (“CAPM”). In this post we continue our discussion of those components, focusing on the equity risk premium and its modifying coefficient, the beta.
The CAPM has become the Delaware Court of Chancery’s preferred method for calculating a company’s cost of equity (i.e., the rate of return an investor would demand in order to purchase the company’s stock). A company’s cost of equity under the CAPM is generally the sum of (1) a risk-free rate, plus (2) the equity risk premium adjusted by a beta, plus (3) a size risk premium.
The “equity risk premium” is the difference between the risk-free rate and the expected return from the market. That is, the equity risk premium predicts how a stock index will perform compared to a risk-free investment, such as a treasury bond. Because not all stocks listed on a particular index perform alike, an appraiser valuing a specific company typically adjusts the equity risk premium by a volatility metric called a “beta.” A company with a beta of 1.0 will have an equity risk premium in line with the market. A company with a beta higher than 1.0 will be more volatile than the market, and a company with a beta of less than 1.0 will be less volatile than the market.
Calculating the Equity Risk Premium
The Ibbotson SBBI Valuation Yearbook provides two methods for calculating the equity risk premium: historic and supply-side. The historic equity risk premium looks at stock market returns against risk-free returns dating back to 1926. The supply-side equity risk premium modifies the historic equity risk premium by adjusting the historic equity risk premium for any inflation included in the price-to-earnings ratio. The supply-side method thus produces a slightly lower equity risk premium than the historic method.
Although the historic equity risk premium is the more traditional method, in its recent appraisal opinions the Delaware Court of Chancery has embraced the supply-side equity risk premium as the prevailing methodology. In Global GT LP v. Golden Telecom, Inc., for example, the court adopted the supply-side method over the historic method because the weight of authority supported a rate of return that was closer to the supply-side equity risk premium.
Small variances in beta can lead to large discrepancies in the overall valuation of a company. For example, suppose an appraiser determines that, as of the merger date, the equity risk premium for Company X was 6%. A beta of 1.5 would increase that number to 9%. A beta of 0.5 would decrease that number to 3%. Assuming Company X had very little debt, this could lead to an almost 6% swing in the weighted average cost of capital. Not surprisingly, therefore, beta calculations are frequently contested in appraisal actions.
Although the historical market beta of a publicly traded company can be calculated by examining the covariance between the stock’s historical performance and that of the S&P 500, this method is often unreliable when calculating the beta of smaller public companies, where the stock may not trade in an efficient market. An alternative method for calculating beta is to use the published betas of guideline companies to select a beta for the subject company. Because the guideline companies have their own unique capital structures, however, the appraiser must “unlever” the guideline betas to remove the impact that the guideline company’s debt has on its beta. An unlevered beta is calculated using the following equation:
where LB is the levered beta of the guideline company; T is the tax rate of the guideline company; D is the percentage of the guideline company’s capital structure that is debt; and E is the percentage of the guideline company’s capital structure that is equity.
After selecting an appropriate unlevered beta for the subject company based on the unlevered betas of the guideline companies, the appraiser must “relever” the selected beta based on the capital structure of the subject company, using the following equation:
UB*[1 + (1 – T)*(D/E)]
where UB is the selected unlevered beta for the subject company; T is the tax rate of the subject company; D is the percentage of the subject company’s capital structure that is debt; and E is the percentage of the subject company’s capital structure that is equity. This levered beta is then applied to the equity risk premium as part of the calculation of the subject company’s cost of equity capital. This is a generally accepted method for calculating beta under the CAPM, although it is not the only method.