Bandera Master Fund LP v. Boardwalk Pipeline Partners, LP

Opinion

IN THE COURT OF CHANCERY OF THE STATE OF DELAWARE

BANDERA MASTER FUND LP, BANDERA ) VALUE FUND LLC, BANDERA OFFSHORE ) VALUE FUND LTD., LEE-WAY FINANCIAL ) SERVICES, INC., and JAMES R. MCBRIDE, ) on behalf of themselves and similarly situated ) BOARDWALK PIPELINE PARTNERS, LP ) UNITHOLDERS, )

)

Plaintiffs, )

)

v. ) C.A. No. 2018-0372-JTL )

BOARDWALK PIPELINE PARTNERS, LP, ) BOARDWALK PIPELINES HOLDING ) CORP., BOARDWALK GP, LP, ) BOARDWALK GP, LLC, and LOEWS CORP., )

)

Defendants. )

MEMORANDUM OPINION

Date Submitted: July 14, 2021 Date Decided: November 12, 2021

A. Thompson Bayliss, J. Peter Shindel, Jr., Daniel G. Paterno, Eric A. Veres, Samuel D. Cordle, ABRAMS & BAYLISS LLP, Wilmington, Delaware; Attorneys for Plaintiffs.

Srinivas M. Raju, Blake Rohrbacher, Matthew D. Perri, John M. O’Toole, RICHARDS, LAYTON & FINGER, P.A., Wilmington, Delaware; Rolin P. Bissell, YOUNG CONAWAY STARGATT & TAYLOR LLP, Wilmington, Delaware; Daniel A. Mason, PAUL, WEISS, RIFKIND, WHARTON & GARRISON LLP, Wilmington, Delaware; Stephen P. Lamb, Andrew G. Gordon, Harris Fischman, Robert N. Kravitz, Carter E. Greenbaum, PAUL, WEISS, RIFKIND, WHARTON & GARRISON LLP, New York, New York; Lawrence Portnoy, Charles S. Duggan, Gina Cora, DAVIS POLK & WARDWELL LLP, New York, New York; Attorneys for Defendants.

LASTER, V.C.

In 2005, Loews Corporation formed Boardwalk Pipeline Partners, LP (“Boardwalk”

or the “Partnership”). Loews controlled Boardwalk by controlling Boardwalk’s general partner. From 2005 until 2018, Boardwalk was a master limited partnership (“MLP”), meaning that the common units representing its limited partner interests traded on an exchange.

Throughout its existence, Boardwalk has served as a holding company for subsidiaries that operate interstate pipeline systems for the transportation and storage of natural gas. The Federal Energy Regulatory Commission (“FERC” or the “Commission”) regulates interstate pipelines. Loews took Boardwalk public in 2005 after FERC implemented a regulatory policy that made MLPs a highly attractive investment vehicle for pipeline companies.

As a business matter, Loews wanted to be able to take Boardwalk private again if FERC took regulatory action that would have a material adverse effect on Boardwalk. To address that business issue, the lawyers who drafted Boardwalk’s partnership agreement included a provision that gave Boardwalk’s general partner the right to acquire the limited partners’ interests if certain conditions were met (the “Call Right”). Two conditions are front and center in this case.

The first condition required that the general partner receive “an Opinion of Counsel that the Partnership’s status as an association not taxable as a corporation and not otherwise subject to an entity-level tax for federal, state or local income tax purposes has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers” (respectively, the “Opinion,” and the “Opinion

Condition”). The Opinion Condition required counsel to address a mixed question of fact and law: whether an event had or was reasonably likely in the future to have a material adverse effect on the maximum applicable rate that Boardwalk could charge its customers. By focusing on a rate that could be charged to customers, the Opinion Condition meshed imperfectly with Loews’ business goal of protecting against future regulatory action that would have a material adverse effect on Boardwalk. And as this decision details, the Opinion Condition used language that presented a host of interpretive difficulties.

The second condition required that the general partner determine that the Opinion was acceptable (the “Acceptability Condition”). Boardwalk’s general partner was itself a limited partnership. The general partner of that limited partnership was a limited liability company, and it had both a board of directors and a sole member, each of which had authority to make certain decisions regarding the Partnership. Boardwalk’s partnership agreement did not specify which decision-maker in this structure would determine whether the Opinion was acceptable. Other agreements did not clearly answer the question either. Reading the agreements in combination led to at least two possible answers. Under one interpretation, the LLC’s board of directors would make the acceptability determination. That made sense from a governance perspective, because the LLC’s board of directors included outside directors who could inject a measure of independence into the determination. Under another interpretation, the LLC’s sole member would make the determination. The LLC’s sole member was a subsidiary of Loews, and all of the decision- makers at that entity were Loews insiders. That interpretation enjoyed more textual support, but it rendered the Acceptability Condition surplusage, because Loews always had

3

the ability to make a de facto acceptability determination when deciding whether or not to exercise the Call Right.

In March 2018, FERC proposed a package of regulatory policies that could have made MLPs an unattractive investment vehicle for pipeline companies. Everyone recognized that the proposals were not final, and industry players lobbied vigorously to change them. One of the major questions surrounding the proposals was how FERC would treat a pipeline’s outstanding balance for accumulated deferred income taxes (“ADIT”). Boardwalk made clear in its public comments to FERC that it was impossible to determine the effect of FERC’s proposals on Boardwalk’s rates until FERC made a decision on the treatment of ADIT.

Boardwalk and other industry participants expected FERC to provide further insight at its July 2018 meeting. At that meeting, FERC implemented its proposals in conjunction with a determination that pipelines could eliminate their outstanding ADIT balances. Rather than making MLPs a less attractive investment vehicle for pipeline companies, that regulatory result made MLPs even more attractive.

In the interim, Loews seized on the period of maximum uncertainty that existed after FERC announced the proposed changes but before FERC implemented the actual changes. Loews caused Boardwalk’s general partner to exercise the Call Right, and the acquisition closed just one day before FERC announced the final package of regulatory measures.

By acquiring the limited partner interest, Loews generated what its management team described euphemistically as $1.5 billion in “Value Creation”—much of which would be characterized more aptly as value expropriation. And Loews was able to acquire the

4

limited partners’ interest at a highly attractive price even though the regulatory changes ultimately did not have any negative effect on Boardwalk.

Loews achieved this remarkable result because its in-house legal team and outside counsel worked hard to generate a contrived Opinion. The Opinion that outside counsel provided did not satisfy the Opinion Condition because outside counsel did not render it in good faith. Outside counsel knowingly made unrealistic and counterfactual assumptions, knowingly relied on an artificial factual predicate, and consistently engaged in goal- directed reasoning to get to the result that Loews wanted. Among other noteworthy decisions detailed in this opinion, outside counsel determined that the regulatory proposals were sufficiently final to trigger the Call Right, even though everyone knew the proposals were not final. And outside counsel determined that the proposals were reasonably likely to have a material adverse effect on Boardwalk’s rates, even as Boardwalk stated in its comments to FERC that it was impossible to determine the effect on Boardwalk’s rates until FERC made a decision on the treatment of ADIT. To address the issue that management deemed impossible to assess, outside counsel examined hypothetical indicative rates, failed to incorporate the admittedly low chance that Boardwalk’s rates actually would change, and derived the magnitude of the assumed change from a simple syllogism. Viewed as a whole, outside counsel’s conduct went too far to constitute a good faith effort to render a legal opinion.

Loews locked in its ability to exercise the Call Right by having the sole member of the LLC that served as the general partner of Boardwalk’s general partner pronounce the Opinion acceptable. That determination did not satisfy the Acceptability Condition because

5

the partnership agreement is ambiguous. Under the doctrine of contra proferentem, the resulting ambiguity must be resolved against the general partner, not in favor of the general partner. In this case, the doctrine requires interpreting the partnership agreement so that only the board of directors of the LLC could pronounce the Opinion acceptable. Four of the eight members of that board of directors were outsiders. Vesting the decision in that decision-maker is more favorable to the limited partners than an interpretation that gives sole authority over the decision to the sole member of the LLC, where all of the decision- makers were Loews insiders.

A bevy of lawyers strived to paper the record so that the Opinion Condition and the Acceptability Condition would appear satisfied. In reality, they were not. The general partner therefore breached the partnership agreement by exercising the Call Right and acquiring the limited partners’ interests.

At this point in the analysis, the general partner argues that it is nevertheless insulated against liability by two protective provisions in the partnership agreement. The first provision generally exculpates the general partner against liability, but contains an exception for willful misconduct. Because the general partner acted intentionally and opportunistically, the general partner’s contractual breach constituted willful misconduct, and the general partner is not exculpated from liability. The second provision protects the general partner if it relies on opinions, reports, or other statements provided by someone that the general partner reasonably believes to be an expert. Here, the general partner participated knowingly in the efforts to create the contrived Opinion and provided the propulsive force that led the outside lawyers to reach the conclusions that Loews wanted.

6

The general partner therefore cannot claim to have relied on the Opinion, and the defense is unavailable.

The general partner is liable for damages in the amount of $689,827,343.38, plus pre- and post-judgment interest on that amount from July 18, 2018, through the date of payment. The plaintiffs are entitled to an award of costs as the prevailing party.

I. FACTUAL BACKGROUND Trial took place over four days using the zoom videoconferencing platform. Eight fact witnesses and six experts testified live. The parties introduced 1,978 exhibits, including twenty deposition transcripts.

In the pre-trial order, the parties commendably agreed to nearly 400 stipulations of fact. The court thanks litigation counsel for their efforts as officers of the court in preparing those detailed stipulations. This decision relies on them when applicable.1 The stipulations do not address all of the factual issues, and they do not determine the inferences to be drawn from the stipulated facts when evaluated in conjunction with the evidence.

1 Citations in the form “PTO ¶ ––” refer to stipulated facts in the pre-trial order. See Dkt. 173. Citations in the form “[Name] Tr.” refer to witness testimony from the trial transcript. Citations in the form “[Name] Dep.” refer to witness testimony from a deposition transcript. Citations in the form “JX –– at ––” refer to a trial exhibit, with the page designated by the internal page number. If a trial exhibit used paragraph numbers or sections, then references are by paragraph or section. Citations in the form “PDX –– at –– ” refer to the plaintiffs’ demonstrative exhibits that summarized information appearing of record in other sources.

7

The court has evaluated the credibility of the witnesses and carefully weighed the evidence. The court has placed the burden of proof on the plaintiffs for all contested issues. The plaintiffs proved the following factual account by a preponderance of the evidence. A. The Partnership Boardwalk is a limited partnership organized under the laws of the State of Delaware. During the period relevant to this litigation, Boardwalk owned three principal subsidiaries, each of which operated an interstate pipeline and storage system for natural gas: Texas Gas Transmission, LLC (“Texas Gas”); Gulf South Pipeline Company, LP (“Gulf South”); and Gulf Crossing Pipeline Company LLC (“Gulf Crossing”).

Loews formed Boardwalk in August 2005. At all times since Boardwalk’s formation, Loews has controlled Boardwalk. Loews is a diversified conglomerate whose shares trade on the New York Stock Exchange under the symbol “L.” Loews is controlled and managed by members of the Tisch family.

FERC regulates interstate pipeline companies, including the rates that pipelines can charge for cost-based services. PTO ¶ 61. Loews took Boardwalk public as an MLP after FERC implemented a regulatory policy that made MLPs a highly attractive investment vehicle for pipeline companies. Thirteen years later, Loews exercised the Call Right after FERC proposed a package of regulatory policies that could have made MLPs an unattractive investment vehicle for pipeline companies. As it turned out, the package of policies that FERC actually implemented made MLPs an even more attractive investment vehicle for pipeline companies. Because of the importance of the potential and actual

8

regulatory changes to the case, a basic understanding of the regulatory landscape is necessary to make sense of what transpired.

1. The Regulation Of Pipeline Rates As part of its regulatory mandate, FERC determines the maximum rates—also known as “recourse rates”—that a pipeline can charge the firms who pay the pipeline to transport and store their product—known as “shippers.” PTO ¶¶ 61, 80, 111. Under the Natural Gas Act (“NGA”), a pipeline’s recourse rates must be “just and reasonable.” PTO ¶ 88.

FERC establishes a pipeline’s recourse rates through a litigated administrative proceeding known as a “rate case.” Id. ¶ 81; JX 89 at 7–8. If a pipeline believes its recourse rates are too low, then it can file a rate case under Section 4 of the NGA to obtain new, higher rates. JX 89 at 7. If FERC or a shipper believes the pipeline’s recourse rates are too high, they can file a rate case under Section 5 of the NGA to challenge the rates. See id. at 7–8.

Recourse rates remain in effect until FERC approves new rates in a subsequent rate case. PTO ¶ 88. Once approved, a pipeline’s recourse rates are listed publicly in a schedule known as a “tariff.” As a result, they are sometimes called “tariff rates.” See JX 1744 (Webb Report) ¶ 89.

Recourse rates are not mandatory rates. FERC generally grants pipelines the authority to contract with shippers to provide services at agreed-upon rates. PTO ¶ 97. The resulting “negotiated rates” are “not bound by the maximum and minimum recourse rates in the pipeline’s tariff.” Id. FERC also allows pipelines “to selectively discount their rates,”

9

resulting in what are referred to as, unsurprisingly, “discounted rates.” Id. Negotiated and discounted rates are alternatives to recourse rates. The term “recourse rate” reflects the fact that a shipper always has recourse to the rates specified in the tariff and cannot be forced to pay a different rate. PTO ¶ 97.

A rate case is a complex affair that involves a five-step process, known as “cost-of-

service ratemaking.” JX 89 at 7, 10. Cost-of-service ratemaking aims to “establish just and reasonable rates” that will provide the pipeline with the opportunity to recover all components of its cost of service and to generate a reasonable rate of return that will adequately compensate its investors. PTO ¶ 93.

What follows is a high-level overview of each of the five steps. Those curious about cost-of-service ratemaking may consult FERC’s 106-page Cost-of-Service Rates Manual, which includes much more detail on each of the five steps and an example of the five steps as applied to a fictional pipeline company. See generally JX 89.

The first step in the ratemaking process is to determine the pipeline’s cost-of-service requirement, which represents the total revenue that the pipeline needs both to cover its expenses and to provide a reasonable rate of return on its invested capital. Id. at 12. The total investment in a pipeline is known as its rate base. Id. at 14. To arrive at a pipeline’s cost-of-service requirement, FERC (1) multiplies a pipeline’s rate base by its overall rate of return, then (2) adds a pipeline’s operating and maintenance expenses, administrative and general expenses, depreciation expenses, and non-income and income taxes, and (3) subtracts any revenue credits. Id. at 12–13. The pipeline’s overall rate of return is a function of the pipeline’s capitalization ratio, its cost of debt, and an allowed rate of return on equity

10

(“ROE”). Id. at 20. In 2018, to calculate a pipeline’s allowed ROE, FERC used a discounted cash flow model. Webb Report ¶ 67.

As noted, a pipeline’s rate base “represents the total investment of the pipeline,”

determined using a formula specified by FERC. JX 89 at 14. Among other things, the formula accounts for ADIT, discussed in greater detail below. Id. at 14, 17–18.

After determining the pipeline’s cost-of-service requirement, the analysis moves to step two. That phase involves computing a “functionalized cost-of-service” by allocating the expenses associated with a pipeline system between its two main functions: transmission and storage. JX 89 at 29–30. There are two main categories of expenses: operation and maintenance expenses, and administrative and general expenses. Id. at 30. Assigning operation and maintenance expenses to one function or another is relatively easy because of existing pipeline accounting requirements. Id. Assigning administrative and general expenses is less straightforward, and FERC prefers to allocate those expenses using a four-step process known as the Kansas-Nebraska Method. Id. at 30–31. FERC then functionalizes any remaining expenses, costs, or credits. Id. at 31–32. At the end of step two, the analysis has generated a functionalized cost of service for both the transmission and storage functions.

Step three is itself a two-step process. Id. at 34. Each of the functionalized costs is “classified as either fixed or variable.” Id. A functionalized cost is fixed if it “remain[s] constant regardless of the volume of throughput” and typically is “associated with capital investment in the pipeline system.” Id. Variable costs, unsurprisingly, are those that “vary with the volume of throughput.” Id. The fixed and variable costs are then further designated

11

as either reservation (demand) costs or usage (commodity) costs. Id. at 35. Whether a cost is classified as a demand or a commodity cost can have an effect on the rate. Id. Generally, variable costs are designated as commodity costs. Id. There is no similar consensus on fixed costs, which require a case-by-case assessment. Id.

Step four splits the functionalized and classified costs derived in steps two and three “between jurisdictional and non-jurisdictional services, among zones and among jurisdictional services.” Id. at 39. FERC uses volume metrics to allocate costs between jurisdictional and non-jurisdictional services, but the importance of that distinction has waned over time. Id. at 42–43. When a pipeline is divided into geographic regions, FERC uses distance metrics to allocate costs among zones. Id. at 43.

The final step of rate design “directly translate[s] the costs allocated to the jurisdictional customers into unit charges or rates.” Id. at 45. The goal of this phase is to design rates that enable the pipeline to “recover the jurisdictional cost-of-service.” Id. Rate design includes both a “firm service rate,” which is made up of a “reservation charge” and a “usage charge,” and an “interruptible service rate,” which is “charged per unit of gas transported.” Id. at 45–46. Calculating the “interruptible service rate” requires a separate multi-step analysis. Id. at 47–48.

The accuracy of a rate design is determined by running a revenue check. Id. at 49.

A rate is accurate if the product of the rates for each service and its accompanying billing determinant (for example the volume of gas transported over a given contractual period) equals the cost of service calculated at step one. Id. The numbers need not be exactly equal, but they must be within 1/100th of a percent of each other. Id.

12

2. The Income Tax Allowance And ADIT One component of a pipeline’s cost of service is the income taxes that the pipeline pays. In the years before 1995, FERC allowed all pipelines to include an “income tax allowance” in their cost-of-service calculations, regardless of how they were organized as entities. As a general rule, including the income tax allowance increases the total cost of service, which in turn supports a higher rate base and a greater revenue requirement. JX 89 at 12. A higher cost of service generally (but not always) leads to higher recourse rates. That result favors pipelines, who could therefore charge shippers higher rates.

A related component of a pipeline’s cost of service is ADIT, which is an accounting concept that arises because various tax provisions authorize pipelines to depreciate their assets on an accelerated basis. PTO ¶ 98. When calculating recourse rates, however, FERC uses straight-line depreciation. Because a pipeline can claim depreciation more quickly for tax purposes than for rate setting, the pipeline pays lower income taxes in the years when accelerated depreciation applies, resulting in greater cash flows than FERC’s rate-setting calculations contemplate. Id. ¶ 99. Once the period of accelerated depreciation ends, the process reverses, and the pipeline ends up paying higher taxes than FERC’s rate-setting calculations contemplate. Id. ¶ 100.

By accelerating depreciation and deferring taxes, the pipeline benefits from the time-value of money. To reflect the fact that the taxes ultimately must be paid, the pipeline records the accumulated value of the tax deferral on its balance sheet as ADIT. During the years when the pipeline benefits from accelerated depreciation and pays lower taxes, the

13

ADIT balance builds up. After the period of accelerated depreciation, once the pipeline begins paying higher taxes, the ADIT balance declines. Id. ¶¶ 99–100.

In substance, the accelerated depreciation acts as an interest-free loan from the government that the pipeline eventually must repay. The balance on the pipeline’s balance sheet is therefore referred to as an “ADIT liability.” Id. ¶ 99. More importantly for present purposes, FERC historically treated a positive ADIT balance as a cost-free source of capital. Id. ¶ 98. FERC therefore subtracted the ADIT balance from the pipeline’s rate base for purposes of the cost-of-service calculation.

As a general rule, subtracting ADIT decreases the total cost of service, which in turn supports a lower rate base and a lower revenue requirement. A lower cost of service thus generally (but not always) leads to lower recourse rates. See id. ¶¶ 98, 101. That result favors shippers, who have recourse to lower rates.

The foregoing discussion makes explicit an obvious economic reality: pipelines and shippers have opposing interests in setting recourse rates. As a general rule, pipelines want higher recourse rates, and they advocate for regulatory approaches that tend to generate higher rates. Shippers want lower recourse rates, and they advocate for regulatory approaches that tend to generate lower rates.

3. Changes In Cost Of Service Do Not Necessarily Lead To Changes In Recourse Rates.

Although the cost-of-service calculation is a core part of the ultimate determination of recourse rates, a change in a pipeline’s cost of service is not the same as a change in its recourse rates. The two ideas reflect “different things.” Wagner Tr. 286. A pipeline’s cost

14

of service changes over time, but those changes do not automatically trigger changes in recourse rates. See id. at 265. As a result, it is improper to equate a change in cost of service with a change in recourse rates. See McMahon Tr. 547–48; JX 575 at 2; JX 1139 at 30–31.

Instead, there must be a “vehicle” for a rate change, namely a rate case under Section 4 or 5 of the NGA. See, e.g., McMahon Tr. 481; Wagner Tr. 264–66; Webb Tr. 936–37. If there is no rate case, then there cannot be a change in recourse rates. If a rate case is unlikely, then a change in recourse rates is unlikely. Wagner Tr. 266.

If a rate case is filed, and if the evidence shows that one cost-of-service input has changed, then rates still might end up increasing, decreasing, or staying the same. As described above, the complex five-step analysis in a rate case looks to all of the cost-of- service inputs and applies principles of rate design. It does not simply adjust a single cost- of-service variable (such as the income tax allowance) to generate a change in recourse rates. See Wagner Tr. 274–75; Webb Tr. 914. That type of approach is called “single-issue ratemaking,” and FERC has a general policy against it.2 There is also a longstanding legal prohibition against FERC engaging in “retroactive ratemaking.” That term refers to any effort to adjust a pipeline’s current rates to make up

2 See JX 1743 (Court Report) ¶ 39 (“[A]lthough one component of the cost-of-

service calculation may have increased, others may have declined[,] . . . . and any decreases in an individual component may be offset against increases in other cost components.”); McMahon Tr. 548 (same); Johnson Tr. 663 (agreeing that “if you change one variable in a rate calculation, you have to revisit all the other variables as well”); id. at 614–16 (same); Sullivan Dep. 102 (agreeing that changing one cost-of-service element does not provide “meaningful information” regarding recourse rates).

15

for over- or under-collection in prior periods. See Court Tr. 854–55. Put another way, “FERC’s regulation of rates has to be prospective only.” Johnson Tr. 662. In a decision from 1990, the United States Court of Appeals for the District of Columbia (the “DC Circuit”)—the final court of appeal as of right from FERC determinations—applied the prohibition against retroactive ratemaking to an ADIT balance. Pub. Utils. Comm’n of Cal. v. FERC, 894 F.2d 1372 (D.C. Cir. 1990). The case involved a pipeline changing how it priced its services such that it would no longer draw on an accumulated ADIT balance to fund future tax liability. See id. at 1375–76. The pipeline’s customers sought a refund of the ADIT balance, but the court rejected that request. Among other reasons, the court stated that refunding ADIT would violate the prohibition on retroactive ratemaking by forcing the pipeline to return a portion of the rates that FERC had approved and the pipeline had collected during prior periods. Id. at 1383.

4. FERC’s 2005 Policy Because cost-of-service calculations ultimately affect rates, and because pipelines and shippers have opposing interests when it comes to rates, FERC’s regulations and policies regarding cost-of-service calculations are subject to constant challenge. Pipelines and shippers engage relentlessly in litigation and lobbying to advance their competing interests.

One perennial debate concerns the extent to which a pipeline organized as a pass-

through entity for tax purposes, and which therefore does not pay taxes at the entity level, can nevertheless claim an income tax allowance for purposes of its cost-of-service calculation. The prevailing pass-through entity in the pipeline industry is the limited

16

partnership, so the debate has been framed in terms of the extent to which a pipeline organized as a limited partnership can claim an income tax allowance.

In 1995, FERC issued a ruling that permitted a pipeline organized as a limited partnership to claim an income tax allowance when calculating its cost of service, but only to the extent that its partnership interests were held by a corporation. FERC announced that ruling in a decision involving the Lakehead Pipeline Company, so the ruling became known as the Lakehead policy. See Lakehead Pipeline Co., Ltd. P’ship, 71 FERC ¶ 61,338 (1995), abrogated by SFPP, L.P. v. FERC, 967 F.3d 788 (D.C. Cir. 2020).

When adopting the Lakehead policy, the Commission focused on the existence of two potential levels of taxation before returns from the pipeline reached investors. The Commission noted that for the partnership interests owned by the corporation, the corporation would have to pay corporate-level tax before distributing any returns to its investors. The Commission reasoned that the pipeline should be able to take into account the corporate-level tax when determining the level of return that those investors would require. By contrast, the Commission noted that for the partnership interests owned by individual investors, there would not be an intervening level of tax; those investors would receive the returns from the pipeline directly. Accordingly, the Commission reasoned that because the individuals would not pay corporate-level tax, the pipeline should not receive a tax allowance for those individuals. Otherwise, the Commission concluded, the pipeline would be able to claim an unrealistically large cost-of-service requirement and provide its

17

investors with a rate of return greater than warranted. See Lakehead, 71 FERC ¶ 62,313– 15, 62,329.3 Nine years later, in 2004, the DC Circuit abrogated the Lakehead policy. The case involved challenges to the Commission’s determinations in a rate case involving SFPP, L.P., an oil pipeline organized as a limited partnership. See BP West Coast Products, LLC v. FERC, 374 F.3d 1263 (D.C. Cir. 2004). The Commission had applied the Lakehead policy to SFPP, ruling that SFPP could claim a tax allowance for the taxes paid by its corporate parent, which owned a 42.7% interest in the partnership. The Commission had determined that SFPP could not claim a tax allowance for any of the interests held by its public investors. The DC Circuit rejected that analysis and the Lakehead policy in general, finding that the Commission had not provided any grounds for distinguishing between the tax liability of the corporate partner and the tax liability of other partners. Id. at 1290. The DC Circuit explained that the regulated entity was entitled to include its own costs of service in its rate base, including taxes, but not costs incurred by its investors, again including taxes. Id. The DC Circuit held squarely that “no such [tax] allowance should be included.” Id. at 1291.

3 An example illustrates how the Lakehead policy operates. Assume that a pipeline is organized as an MLP, that its corporate general partner owns 50% of the partnership interests, and that public investors own the rest. If the corporation paid taxes at a rate of 35%, then the pipeline could claim a tax allowance of 17.5%, reflecting the taxes paid at the corporate level. Rosenwasser Tr. 42. The pipeline could not, however, claim a tax allowance for taxes paid by the individual investors.

18

In 2005, FERC responded to the BP West decision by heading in the opposite direction. Rather than concluding that a pipeline organized as a partnership could not claim an income tax allowance, as BP West held, FERC announced that it would “return to its pre-Lakehead policy” and permit a pipeline organized as a partnership to claim an income tax allowance for all of its partners. PTO ¶ 104; see JX 205 (the “2005 Policy”). In reaching this conclusion, the Commission took the view that all partners pay income taxes and that their taxes should be imputed to the pipeline for purposes of determining the pipeline’s cost of service. PTO ¶ 104. Because a pipeline organized as a limited partnership does not actually pay entity-level income taxes, the 2005 Policy made pipelines organized as limited partnerships a highly attractive investment vehicle. See id. ¶ 106; Rosenwasser Tr. 39–40. B. Loews Forms Boardwalk.

To take advantage of the 2005 Policy, Loews formed Boardwalk in August 2005.

PTO ¶ 106. Loews planned to take Boardwalk public through an initial public offering (“IPO”) later that year. Id. Loews retained Michael Rosenwasser, then a partner at Vinson & Elkins LLP, to lead the legal team that prepared Boardwalk’s organizational documents and supported the IPO. Id. ¶ 51.

1. Boardwalk’s Structure Loews organized Boardwalk as a Delaware limited partnership. As a result, its internal affairs were (and are) governed by its partnership agreement. By the time of the events giving rise to this litigation, the operative version was the Third Amended and Restated Agreement of Limited Partnership dated June 17, 2008. JX 352 (the “Partnership Agreement” or “PA”).

19

The Partnership’s general partner was another Delaware limited partnership, defendant Boardwalk GP, LP (the “General Partner”). The General Partner held a 2% general partner interest in the Partnership and owned all of its incentive distribution rights. JX 256 at 14. The General Partner did not have a board of directors. Id.

The sole general partner of the General Partner was defendant Boardwalk GP, LLC (“the GPGP”). Id. The GPGP was a Delaware limited liability company, so its internal affairs were governed by its limited liability company agreement. By the time of the events giving rise to this litigation, the operative version was the First Amended and Restated Limited Liability Company Agreement dated November 15, 2005. JX 235 (the “LLC Agreement” or “LLCA”).

The sole member of the GPGP was defendant Boardwalk Pipelines Holding Corp.

(“Holdings,” or the “Sole Member”). Id. § 1.1 at 7. At all relevant times, Holdings was a wholly owned subsidiary of Loews. Through Holdings, Loews controlled the GPGP. Through the GPGP, Loews controlled the General Partner. Through the General Partner, Loews controlled Boardwalk and its subsidiaries.

In addition to having Holdings as its Sole Member, the GPGP had a board of directors (the “GPGP Board”). The LLC Agreement generally assigned authority over the business and affairs of the GPGP and the Partnership to the GPGP Board. PTO ¶ 76. The LLC Agreement granted the Sole Member “exclusive authority over the business and affairs of [the GPGP] that do not relate to management and control of [the Partnership].” LLCA § 5.6.

20

For the vast majority of the Partnership’s existence as an MLP, the GPGP Board had eight members. Four were outside directors whose only affiliation with Boardwalk or Loews was their status as directors on the GPGP Board. The other four members were:

• Kenneth I. Siegel, Senior Vice President of Loews and Chairman of the GPGP Board;

• Andrew H. Tisch, the Co-Chairman of the board of directors of Loews, the Chairman of the Executive Committee of Loews, and member of the Office of the President of Loews.

• Peter W. Keegan, a Senior Advisor to Loews; and • Stanley C. Horton, the President and Chief Executive Officer of Boardwalk. During the period relevant to this litigation, the Holdings board of directors (the “Holdings Board”) consisted of Siegel, Keegan, and Jane Wang, Vice President of Loews.

The different composition of the GPGP Board and the Holdings Board meant that if Holdings made a decision for the GPGP as its Sole Member, then Loews controlled the decision. By contrast, if the GPGP Board made the decision for the GPGP, then the outside directors would participate in the decision. If the four outside directors unanimously opposed the Loews and Boardwalk representatives, then they could prevent the GPGP from taking the action that Loews wanted.

The following diagram depicts Boardwalk’s organizational structure and its principal pipeline subsidiaries.

21

2. The Call Right The provision at the heart of this case is the Call Right, which granted the General Partner the right to acquire the common units that the General Partner and its affiliates did not already own as long as certain conditions were met. The Call Right came to be included in the Partnership Agreement because the 2005 Policy was contentious. It favored pipelines over shippers, and shippers challenged it immediately. See, e.g., ExxonMobil Oil Corp. v. FERC, 487 F.3d 945 (D.C. Cir. 2007) (addressing shipper challenge to 2005 Policy). Loews was concerned that FERC might change course. McMahon Dep. 62, 160–61.

Loews wanted a mechanism for taking Boardwalk private again if the 2005 Policy changed in a manner that was materially adverse to Boardwalk. See Rosenwasser Tr. 41– 44; McMahon Tr. 480, 544–45. Rosenwasser recalled these matters vividly. He testified that Loews was not “going to go forward with [Boardwalk’s IPO] unless [Rosenwasser and

22

his team] were able to include a provision in [the Partnership Agreement] which would allow them quickly, easily and without dispute, to go private if there was an adverse change in that tax policy or the way it was implemented.” Rosenwasser Dep. 34–35; see id. at 39 (“Loews . . . wanted a mechanism that would allow them to go private in a simple, clear manner without dispute if, in fact, there was a change in FERC policy that would be adverse to maximum applicable rates.”). He testified at trial that Loews told the underwriter for the IPO that it would not take Boardwalk public unless it could guard against the risk of “los[ing] any substantial portion of the tax allowance if there was a reversion to Lakehead.” Rosenwasser Tr. 42. Early drafts of the Partnership Agreement referred to the call right as a “Lakehead call.” PTO ¶ 109. Referring to the 2005 Policy, the IPO prospectus and Boardwalk’s subsequent annual reports informed investors that “[i]f the FERC policy is reversed . . . our general partner’s call right may be triggered.” JX 256 at 31; accord JX 285 at 11.

Critically, however, no one intended the Call Right to be triggered by a change that “wasn’t substantive, wasn’t meaningful.” Rosenwasser Tr. 46. Loews “wanted an off-ramp if FERC reverse[d] its policy” in a way that materially threatened revenues. McMahon Tr. 480, 545. Rosenwasser and his team attempted to draft the Call Right to achieve that business objective. Rosenwasser Dep. 39. It was a “business point,” not a “legal point.” Id. at 40.

In an effort to implement this business point, Rosenwasser included language stating that the General Partner could exercise the Call Right if three conditions were met. First, the General Partner and its affiliates had to own “more than 50% of the total Limited

23

Partner Interests of all classes then Outstanding.” PA § 15.1(b)(i). Second, the General Partner had to satisfy the Opinion Condition by receiving an “Opinion of Counsel” that Boardwalk’s status as a pass-through entity for tax purposes “has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers.” Id. § 15.1(b)(ii). Third, the General Partner had to satisfy the Acceptability Condition by determining that the Opinion was “acceptable to the General Partner.” Id. § 1.1 at 24.

As long as these conditions were met, then the General Partner could decide whether to exercise the Call Right. When making that decision, the General Partner could act in its sole discretion, free of any fiduciary duty or express contractual standard, with the express right to consider its self-interest, and constrained only by its obligation to comply with the non-waivable implied covenant of good faith and fair dealing. Id. § 7.1(b)(iii).

The Partnership Agreement did not impose any timeline for obtaining the Opinion, but once the Opinion Condition was satisfied, the General Partner had ninety days to exercise the Call Right. Id. §15.1(b). The Partnership Agreement did not require that independent counsel render the Opinion. The term “Opinion of Counsel” was not specific to the Opinion Condition and appeared in multiple provisions in the Partnership Agreement; the agreement defined it as “a written opinion of counsel (who may be regular counsel to the Partnership or the General Partner or any of its Affiliates).” Id. § 1.1 at 24.

If the General Partner exercised the Call Right, then the General Partner was obligated to send notice by mail to that effect to the limited partners. Id. § 15.1(c). The General Partner was then obligated to purchase all of the outstanding limited partner

24

interests that it did not already own “at a purchase price . . . equal to the average of the daily Closing Prices . . . for the 180 consecutive Trading Days immediately prior to the date three days prior to the date that the notice described in Section 15.1(c) is mailed.” Id. § 15.1(b) (the “Purchase Price”).

3. The IPO On November 8, 2005, Boardwalk offered common units to the public at a price of $19.50 per unit. JX 260 at 1. Until the General Partner acquired the public units at a price of $12.06 per unit on July 18, 2018, Boardwalk’s common units traded on the New York Stock Exchange under the symbol “BWP.”

During the intervening years, Loews caused Boardwalk to issue additional units at prices well above $12.06 per unit. Loews also sold units to the public in secondary offerings at values well above $12.06 per unit. The following table summarizes those offerings:

25

PDX 6 at 1 (footnotes omitted).

When Loews exercised the Call Right, public investors held approximately 49% of Boardwalk’s common units. PTO ¶ 48. It is undisputed for purposes of this litigation that the General Partner and its affiliates held a sufficient percentage of the total limited partnership interests to satisfy the first condition for exercising the Call Right. C. The United Airlines Decision For purposes of the current litigation, the next significant development took place in 2016. The initial efforts by shippers to challenge the 2005 Policy failed when the DC Circuit held in 2007 that the 2005 Policy was “not unreasonable” and hence entitled to deference. ExxonMobil, 487 F.3d at 953. Nine years later, however, the shippers prevailed in United Airlines, Inc. v. FERC, 827 F.3d 122 (D.C. Cir. 2016).

Despite its name, the United Airlines case was an appeal from FERC’s determinations in a rate case involving SFPP. Advancing a different argument than the theory the DC Circuit had rejected in 2007, the shippers contended that by permitting MLP pipelines to claim an allowance for partner-level taxes, the 2005 Policy “permit[ted] [the] partners in a partnership pipeline to ‘double recover’ their taxes.” Id. at 127.

FERC rejected that contention, but the DC Circuit endorsed it. In vacating the Commission’s order and ruling in favor of the shippers, the DC Circuit cited the following undisputed facts:

First, unlike a corporate pipeline, a partnership pipeline incurs no taxes, except those imputed from its partners, at the entity level. Second, the discounted cash flow return on equity determines the pre-tax investor return required to attract investment, irrespective of whether the regulated entity is a partnership or a corporate pipeline. Third, with a tax allowance, a partner

26

in a partnership pipeline will receive a higher after-tax return than a shareholder in a corporate pipeline, at least in the short term before adjustments can occur in the investment market.

Id. at 136 (internal citations omitted). Based on these undisputed facts, the DC Circuit concluded that “granting a tax allowance to partnership pipelines results in inequitable returns for partners in those pipelines as compared to shareholders in corporate pipelines.” Id. at 137. The DC Circuit remanded the case with instructions for the Commission to determine whether it could eliminate the double-recovery problem, such as by changing the calculation of the ROE. The DC Circuit also noted that “prior to ExxonMobil, FERC considered the possibility of eliminating all income tax allowances and setting rates based on pre-tax returns,” and that none of the court’s precedents “foreclos[ed] that option.” Id.

In December 2016, FERC responded to the United Airlines decision by issuing a notice of inquiry requesting “comment[s] regarding the double-recovery concern.” JX 579 ¶ 1. Before FERC announced the results of that inquiry, Congress enacted the Tax Cuts and Jobs Act (the “Tax Act”). Among other things, the Tax Act lowered the federal corporate income tax rate from 35% to 21%, effective January 1, 2018. Pub. L. No. 115- 97, 131 Stat. 2054 (2017). D. The March 15 FERC Actions At its regularly scheduled meeting on March 15, 2018, FERC took four interrelated actions to address the implications of the United Airlines decision and the Tax Act (the “March 15 FERC Actions”). In presenting the March 15 FERC Actions, the Commission explained that it was “addressing these issues concurrently” to “ensure[] administrative efficiencies by reducing the number of filings required of regulated entities.” JX 554 at 49.

27

1. The Revised Policy The first of the March 15 FERC Actions was the issuance of a revised policy statement on the treatment of income taxes. JX 579 (the “Revised Policy Statement” or “Revised Policy”). In the Revised Policy, FERC stated that it would no longer permit pipelines organized as MLPs to recover both an income tax allowance and a ROE determined by the discounted cash flow methodology in their cost-of-service calculations. See id. ¶ 8. FERC stated in a concurrently issued notice of proposed rulemaking that it would promulgate regulations to address the effects of the Revised Policy “on the rates of interstate natural gas pipelines organized as MLPs.” Id.; see JX 580.

During the March 15 meeting, in response to a question about when “FERC Jurisdictional Rates [would] actually change,” FERC staff stated that “the NOPR anticipates that the deadlines for pipeline filings will be late summer or early fall [2018]. We obviously have to go to a final rule first.” PTO ¶ 117. The Revised Policy thus had no impact on Boardwalk’s rates. Court Report ¶¶ 102–12.

At the same time, FERC signaled that pipelines would have answers on the regulatory issues soon—in “late summer or early fall”—which would allow them to make anticipated regulatory filings. PTO ¶ 117. Boardwalk anticipated that FERC would address the March 15 FERC Actions further in connection with its regularly scheduled meeting on July 19, 2018. See JX 1152 at 2.

2. The Notice Of Proposed Rulemaking The second of the March 15 FERC Actions was the issuance of a notice of proposed rulemaking titled Interstate and Intrastate Natural Gas Pipelines; Rate Changes Relating

28

to Federal Income Tax Rate (the “NOPR”). JX 580. The NOPR was not an actual rule and did not have any immediate effect on Boardwalk or other industry participants. It was a notice of a proposed rule that invited comment.

In the NOPR, FERC proposed to require interstate natural gas pipelines to make a one-time informational filing on a proposed Form 501-G so that FERC could evaluate the impact of the Tax Act and the change in income tax policy on pipelines’ revenue requirements. JX 580 ¶ 32. FERC explained that the purpose of the Form 501-G was to provide information “regarding the continued justness and reasonableness of the pipeline’s rates after the income tax reduction and elimination of MLP income tax allowances.” Id. ¶ 26. The Form 501-G therefore would call for “an abbreviated cost and revenue study in a format similar to the cost and revenue studies the Commission has attached to its orders initiating NGA section 5 rate investigations in recent years.” Id. ¶ 32.

FERC proposed that when completing the Form 501-G, a pipeline would use data from its 2017 FERC Form No. 2, which provided information on the major components of its cost of service for that year. Id. Using that information, the pipeline would estimate (1) the percentage change in its cost of service resulting from the Tax Act’s reduction of the corporate income tax from 35% to 21% and the Revised Policy’s reduction of the corporate income tax allowance for MLPs from 35% to 0% and (2) the pipeline’s ROE both before and after those developments. Id.; see also PTO ¶ 120. To derive the cost-of-service component associated with the return to equity investors, FERC proposed that pipelines use an ROE of 10.55%. JX 580 ¶ 34.

29

FERC intended for resulting calculations to indicate whether the pipeline’s rate base could have decreased as a result of the elimination of the income tax allowance. The resulting calculations also would indicate whether, based on the pipeline’s actual historical revenues, the pipeline was over-recovering its rate base in a manner that might warrant a rate case.

The NOPR proposed that a pipeline would have four options to consider in connection with its Form 501-G:

• The pipeline could make a limited filing under Section 4 of the NGA to reduce the pipeline’s recourse rates to reflect a decrease in its revenue requirements.

• The pipeline could commit to file a general rate case under Section 4 of the NGA in the near future to establish new recourse rates.

• The pipeline could file a statement explaining why a rate adjustment was not needed.

• The pipeline could take no action other than filing the Form 501-G. PTO ¶ 121; JX 580 ¶¶ 41–51. If a pipeline chose the third or fourth option, the Commission anticipated that it would consider, based on information in the Form 501-G, whether to issue an order to show cause to the pipeline requiring a reduction in its rates. PTO ¶ 121.

FERC recognized that even with a lower tax rate and the elimination of the income tax allowance, “a rate reduction may not be justified for a significant number of pipelines.” JX 580 ¶ 48. As an example, FERC noted that “a number of pipelines may currently have rates that do not fully recover their overall cost of service,” such that a reduction in tax costs “may not cause their rates to be excessive.” Id. Typically, a pipeline would be under- recovering its costs if it operated in a competitive market and hence had to offer discounted

30

rates to shippers. See JX 1139 at 11. FERC also cited other possibilities that would obviate the need to adjust rates, such as “an existing rate settlement [that] provides for a rate moratorium” or the existence of contracts providing for negotiated rates. See JX 580 ¶¶ 45, 48–49.

3. The Notice Of Inquiry The third of the March 15 FERC Actions was a notice of inquiry that sought industry comment on the effect of the Tax Act and the Revised Policy on recourse rates. In particular, FERC sought comment on how it should address ADIT. See JX 576 (the “ADIT NOI”).

In requesting comment on ADIT, FERC distinguished between the “[t]reatment of ADIT for [p]artnerships” and the treatment of ADIT for other regulated entities. Id. ¶¶ 24– 25. For partnerships, FERC specifically asked that “commenters . . . address whether previously accumulated sums in ADIT should be eliminated altogether from cost of service or whether those previously accumulated sums should be placed in a regulatory liability account and returned to ratepayers.” Id. ¶ 25.

4. The Order On Remand The fourth and final of the March 15 FERC Actions was the issuance of an order implementing the United Airlines decision for the ongoing proceeding involving SFPP. JX 553 (the “Order on Remand”). The Order on Remand required SFPP to revise its rate filing consistent with the Revised Policy and prohibited SFPP from claiming an income tax allowance. Id. ¶¶ 28, 58(B). That was the only binding and immediately applicable component of the March 15 FERC Actions, and it did not affect Boardwalk.

31

Also on March 15, 2018, FERC initiated two proceedings under Section 5 of the NGA against interstate natural gas pipelines. FERC initiated one proceeding against Dominion Energy Overthrust Pipeline, a natural gas pipeline owned by an MLP, based on an estimated calculation that the pipeline achieved ROEs for calendar years 2015 and 2016 of 23.4% and 19.9%, respectively. The order initiating the proceeding noted that “[i]f Overthrust’s ROEs for 2015 and 2016 were recalculated consistent with the Revised Policy Statement, its ROEs would have been 36.4 percent and 30.9 percent, respectively.” PTO ¶ 133.

FERC also initiated a proceeding against Midwestern Gas Transmission Company, a natural gas pipeline, based on an estimated calculation that Midwestern had achieved ROEs for calendar years 2015 and 2016 of 15.8% and 16.6%, respectively The order initiating the proceeding noted that “if the reduced 21 percent corporate income tax rate had been in effect during 2015 and 2016, Midwestern’s ROE for those years would have been 19.2 percent and 20.2 percent, respectively.” PTO ¶ 133. E. The Reaction To The March 15 FERC Actions The March 15 FERC Actions triggered a flurry of activity from industry participants. Over the next four months, shippers, pipelines, trade associations, and others filed thirteen requests for rehearing, 108 comments, sixteen reply comments, and numerous other submissions. See PDX 9 at 12; Court Tr. 858. Each participant sought to persuade FERC to adopt its preferred outcome. Matters were very much in flux.

The resulting uncertainty generated market reactions. The trading price of Boardwalk’s units dropped by more than 7% from its closing price on March 14, 2018, the

32

day before the March 15 FERC Actions. PTO ¶ 135; see JX 1802 at 1. The Alerian Index, which tracks an index of MLPs in the oil and gas industry, fell by 4.6%. Collectively, MLPs lost $15.8 billion in market capitalization. Plaintiff James McBride tweeted, “Blood in the street. Where’s the buying opportunity?” JX 1839 at 3. Barry Sullivan, a respected FERC consultant who worked for Boardwalk, emailed its executives saying, “I hope you guys are still breathing. That was unbelievable. Sorry.” JX 546 at 1.

Several MLPs issued press releases stating that they did not anticipate that the March 15 FERC Actions would have a material impact on their rates, primarily because their customers were locked into negotiated rate agreements. McMahon Tr. 498–99; Siegel 735–36; see, e.g., JX 592 (Spectra Energy Partners press release stating that it “anticipates no immediate impact to its current gas pipeline cost of service rates as a result of the revised policy”). One industry analyst report stated that although “FERC dropped a bombshell on the industry,” stock prices were rebounding “as companies issued statements saying minimal impact.” JX 624 at 4, 6. Horton, Boardwalk’s CEO, told Loews’ senior management that the analyst report offered “a pretty good summary” of what had happened. Id. at 1.

F. Boardwalk’s Initial Assessment: No Material Effect On Rates, But A Chance For Loews To Exercise The Call Right.

After the announcement of the March 15 FERC Actions, Horton instructed Ben Johnson, Boardwalk’s Vice President of Rates and Tariffs, to conduct an expedited analysis of the possible impact on Boardwalk’s three interstate pipelines. JX 565 at 1. The day’s events prompted questions that Boardwalk’s management team needed to answer. Siegel

33

and Thomas Hyland, an outside director on the GPGP Board, asked Jamie Buskill, Boardwalk’s Chief Financial Officer, for his “thoughts on the economic impact on [Boardwalk].” JX 567 at 1; see also JX 548. Molly Whitaker, Boardwalk’s Director of Investor Relations and Corporate Communications, fielded similar inquiries from approximately a dozen investors and analysts. JX 550 at 1.

To answer these questions, Johnson used an analysis that Boardwalk had performed in early February 2018 to project the effect of the March 15 FERC Actions on the rates that each of the three pipelines could charge. By that evening, he had preliminary answers.

Johnson viewed Gulf Crossing as “relatively protected” from any impact on its rates.

JX 572 at 1. Almost all of Gulf Crossing’s contracted volumes were subject to negotiated rates, meaning that a change in cost-of-service-based rates would not affect the pipeline. Id. at 2. Johnson also viewed Gulf South as “relatively protected.” Id. at 1. A majority of its contracts provided for negotiated or discounted rates, and Gulf South was also subject to a rate case moratorium until May 2023. See PTO ¶ 409; JX 604; JX 1139 at 6.

Texas Gas was the only pipeline that had potential exposure to a rate case, but it too had factors that would help in defending against any challenge to its rates. Among other things, Texas Gas served highly competitive markets, and a majority of its contracts with shippers provided for negotiated or discounted rates. See JX 1139 at 6. Assuming a rate case was filed, Johnson estimated that the downside impact of eliminating the income tax allowance would be about $20.5 million. See JX 572 at 1–2.

Importantly, Johnson characterized his estimate of the downside as a floor, because it “ignores any bounce from rate base increase associated with removal of ADIT.” Id.

34

Elaborating in a later email, he explained that “it’s unclear on what they [FERC] would do with [Boardwalk’s] current ADIT” balance, and he observed that FERC could decide that the ADIT balance should be “zeroed out because there’s no income taxes (because there would be no difference between book and tax depreciation).” JX 602 at 1. Johnson thus recognized at the outset that the treatment of ADIT would be critical for understanding the implications of the March 15 FERC Actions. For purposes of his analysis, Johnson “assume[d] that [the ADIT balance] would just remain until it’s amortized off.” Id.

Having reached the conclusion that the March 15 FERC Actions would not have a materially adverse impact on the rates that Boardwalk’s subsidiaries could charge, Boardwalk’s management team noted that other MLP pipelines had issued press releases expressing similar views about their own rates. Boardwalk’s management team worried that if Boardwalk did not issue a similar statement, then the market participants would infer the March 15 FERC Actions would have an adverse effect on Boardwalk’s rates, which Boardwalk had determined not to be the case. See McMahon Tr. 498–99; Alpert Tr. 322.

Horton therefore instructed Michael McMahon, Boardwalk’s General Counsel, to draft a short press release that described the extent to which Boardwalk’s pipelines were protected from any impact on their rates. JX 568 at 1. In his first draft, McMahon pointed out that FERC had invited pipelines to “file statements explaining why an adjustment to rates to reflect the impact of the Commission’s decisions is not required.” JX 571 at 7. McMahon noted that this path seemed tailor-made for Boardwalk’s pipelines. As he put it, “[t]his option recognizes the unique competitive circumstances of each pipeline, for example, essentially all of the contracts on our Gulf Crossing and a number of the contracts

35

on Texas Gas are negotiated rate agreements and Gulf South is currently under a rate moratorium until 2023 . . . .” JX 571 at 7.

Buskill proposed making the release stronger by stating that the overall impact to Boardwalk and its rates would not be material. JX 571 at 1. McMahon agreed that “the elimination of the income tax allowance will not result in a material impact.” Id. Neither Buskill nor McMahon addressed the possible upside of eliminating ADIT. See id.

By late evening on March 15, 2018, Boardwalk management was satisfied with the language of the release. But as discussed below, the draft would go through a series of revisions once Loews’ personnel got involved.

In the meantime, Buskill responded to the inquiries about the effect of the March 15 FERC Actions by explaining that they would not have a material impact on Boardwalk. During the evening of March 15, 2018, Buskill told Hyland, the outside director on the GPGP Board, that virtually all of the shippers at Gulf Crossing and Gulf South were under negotiated or discounted rate agreements, that Gulf South was under a rate moratorium until 2023, and that only about 20% of Texas Gas’ revenues were from tariff rates. JX 548 at 1. Buskill concluded: “Based on our interpretation of the rules, we don’t think it will have a material impact to Boardwalk.” Id.

Buskill conveyed similar information to Siegel, who immediately forwarded the information to Jim Tisch, the CEO of Loews, and Ben Tisch, another senior officer of Loews. JX 566 at 1. The Loews executives quickly focused on ADIT. JX 601 at 2. At Ben Tisch’s request, a Loews employee analyzed the March 15 FERC Actions and reported that “the loss of 100 percent of taxes in calculating allowed ROE’s would be a flesh wound for

36

the long haul pipes like . . . [Boardwalk].” Id. at 1. But if FERC required that pipelines return their ADIT balances to ratepayers, then that “would be the a-bomb outcome” and would be “extremely painful.” Id. The treatment of ADIT dominated the analysis.

1. A Chance To Exercise The Call Right When the March 15 FERC Actions took place, Buskill and McMahon were each angling to succeed Horton as CEO of Boardwalk. Both immediately realized that the March 15 FERC Actions might give Loews the ability to exercise the Call Right. That course of action could be attractive to Loews because the Purchase Price was calculated using a trailing market average.

In addition to the stock drop resulting from the March 15 FERC Actions, there was reason to believe that Boardwalk’s market price continued to reflect a shock that Boardwalk had delivered by slashing its distributions in 2014. As an asset class, common units in MLPs are a yield-based investment, and MLPs generally make regular quarterly distributions to their investors. In 2014, Boardwalk stunned investors by cutting its quarterly distribution from $0.5325 to $0.10 per unit, making Boardwalk one of the lowest yielding MLPs in the industry. Boardwalk’s trading price fell from the low $30s to the low $10s almost overnight. The unit price never again approached its former levels. See Horton Dep. 52; PDX 11 at 9.

Between 2014 and 2017, Boardwalk spent $2.077 billion on capital expenditures, including $1.6 billion in growth capital expenditures. PTO ¶ 85–86. During the same period, Boardwalk distributed $405.1 million to unitholders. Id. ¶ 85. There is evidence

37

that investors were unsure about how to value the growth capital expenditures. See PTO ¶ 87.4 On March 15, 2018, Buskill and McMahon each made a point of flagging the Call Right for Loews. Buskill emailed Siegel and described the opportunity presented by the Call Right as “compelling” because Loews could “buy back all units when the units are trading well below book value.” JX 567 at 1. Siegel told Buskill that he “need[ed] to better understand the deferred taxes,” namely ADIT. Id.

McMahon contacted Marc A. Alpert, Loews’ Senior Vice President and General Counsel. He told Alpert that “FERC’s actions might have triggered the call.” McMahon Tr. 552; PTO ¶ 136–37. McMahon recommended that Alpert contact Rosenwasser, who had since joined Baker Botts LLP, to ask whether he could issue the Opinion that would enable the General Partner to exercise the Call Right. McMahon Tr. 552–53. McMahon told Alpert that while practicing at Vinson & Elkins, Rosenwasser was “one of the principal draftspersons of the [C]all [R]ight.” Alpert Tr. 325, 330; see Rosenwasser Tr. 39–40; McMahon Tr. 503; McMahon Dep. 31–32.

Alpert liked the idea of hiring Baker Botts and Rosenwasser. Baker Botts had ten nationally ranked practice groups, including groups providing regulatory, litigation, and

4 As this court has observed in other settings, an advantageous time for a controller to acquire a controlled company is when the controlled company has invested capital in net-positive-value projects, but when minority investors have not yet received the benefit of those investments. See, e.g., In re Dole Food Co., Inc. S’holder Litig., 2015 WL 5052214, at *36 (Del. Ch. Aug. 27, 2015); Del. Open MRI Radiology Assocs., P.A. v. Kessler, 898 A.2d 290, 315–16 (Del. Ch. 2006).

38

transactional advice to the oil and gas sector. JX 1498 at 149. Rosenwasser was highly regarded and considered the “[D]ean of the MLP Bar.” Alpert Tr. 325. And although Rosenwasser was a principal drafter of the Call Right, Baker Botts as a firm had never done any work for Boardwalk, which Loews and Boardwalk viewed as a helpful fact. See Rosenwasser Tr. 54–55; Alpert Tr. 324–25; McMahon Tr. 503.

2. Alpert Calls Rosenwasser.

On March 16, 2018, Alpert called Rosenwasser. PTO ¶ 137. Rosenwasser’s secretary transcribed Alpert’s message as saying there was “something urgent that he needs to speak with you about.” Id. At trial, Rosenwasser recalled a brief and measured conversation in which Alpert described the assignment as whether Baker Botts could advise one way or the other about whether it could give the Opinion. Rosenwasser recalled saying only that he would “look into it.” Rosenwasser Tr. 55.

Consistent with an urgent and significant assignment, Rosenwasser quickly assembled a team within Baker Botts. He brought in a group of senior Baker Botts attorneys to act as an ad hoc opinion committee. Rosenwasser had to assemble an ad hoc opinion committee because Baker Botts does not typically utilize opinion committees and does not have a standing committee. Its members were: • Andy Baker, the Chair of the firm;

• Mike Bengtson, the Chair of the firm’s corporate practice group and a member of the Executive Committee;

• Michael Bresson, the leader of the firm’s energy capital markets tax practice; • Joshua Davidson, the leader of the firm’s capital markets practice;

39

• Richard Husseini, a partner focused on tax litigation; and • Julia Guttman, the firm’s General Counsel.

To perform the substantive work, Rosenwasser recruited three other Baker Botts partners:

• Greg Wagner, a FERC practitioner who was representing shippers in their rate disputes with SFPP, including in the United Airlines case;

• Michael Swidler, a transactional partner and longtime colleague of Rosenwasser who previously had worked at Vinson & Elkins as part of the team that drafted the Call Right; and

• Seth Taube, a former federal prosecutor and SEC official whose practice includes securities and commercial litigation.

Rosenwasser and his colleagues spent the weekend reviewing a package of documents from Alpert.

3. The Loews-Approved Press Release Meanwhile, Loews weighed in on the press release about the March 15 FERC Actions. Loews delayed its publication and edited it heavily, admittedly with an eye to the potential exercise of the Call Right. Alpert Dep. 36 (“I certainly had [the Call Right] in my mind when I looked at the press release.”). Boardwalk issued the Loews-approved draft on the morning of March 19.

Cognizant of the Call Right, Loews changed the wording of the release to address revenues rather than rates. Recall that the General Partner’s ability to exercise the Call Right turned on whether a law firm could opine that Boardwalk’s status as a pass-through entity for tax purposes “has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers.” PA § 15.1(b)

40

(emphasis added). In changing the language of the press release, Loews focused on the fact that the language of the Call Right did not mention revenues.

The draft press release prepared by Boardwalk’s management explained that the March 15 FERC Actions were unlikely to have a negative impact on Boardwalk’s rates. See JX 607. Other pipeline companies likewise issued press releases that focused on rates. See, e.g., JX 592 (Spectra Energy Press Release: “Any future impacts would only take effect upon the execution and settlement of a rate case. In the event of a rate case, all cost of service framework components would be taken into consideration which we expect to offset a significant portion of any impacts related to the new FERC policy.”).

As prepared by Boardwalk’s management, the draft press release contained three sentences identifying the factors FERC had cited as mitigating the need for any rate adjustment and explaining how they applied to Boardwalk’s pipelines. Loews struck those statements. See JX 607 at 3. Loews also drafted the headline to focus on revenue rather than rates.

After the Loews edits, the press release read, “Boardwalk Does Not Expect FERC’s Proposed Policy Revisions To Have A Material Impact On Revenues.” JX 615. The body of the press release elaborated on the effect on revenues:

Based on a preliminary assessment, Boardwalk does not expect FERC’s proposed policy revisions to have a material impact on the company’s revenues. All of the firm contracts on Boardwalk’s Gulf Crossing Pipeline and the majority of contracts on Texas Gas Transmission are negotiated or discounted rate agreements, which are not ordinarily affected by FERC’s policy revisions. Gulf South Pipeline currently has a rate moratorium in place with its customers until 2023. Boardwalk will continue to evaluate the potential impact these proceedings could have on its interstate pipelines, and the company plans to submit comments to FERC.

41

Id.

At his deposition, Rosenwasser tried to distance himself from the press release. He speculated that “somebody was pressured at Boardwalk to get something out quickly” and issued the press release “with just . . . thoughts and without analysis.” Rosenwasser Dep. 97. This was not accurate: Rosenwasser’s speculation notwithstanding, Boardwalk had analyzed the effect on its subsidiaries’ rates, and Loews was thinking about the Call Right when its personnel revised the language of the release. Implicitly recognizing that the release was problematic for the exercise of the Call Right, Rosenwasser testified adamantly that he “had nothing to do with this disclosure[]. And if [he] had, it wouldn’t have said this.” Rosenwasser Dep. 95; see also id. at 95–98.

4. The Post-Press Release Call With Baker Botts Several hours after Boardwalk issued the Loews-approved press release, Alpert convened a call with Rosenwasser and other members of the Baker Botts team. Loews wanted answers to two questions. First, had the contents of the press release affected Baker Botts’ ability to issue the Opinion? Second, were the March 15 FERC Actions sufficiently concrete to enable Baker Botts to issue the Opinion?

The next day, Baker Botts answered both questions. On the press release, Loews got the answer it wanted. Baker Botts advised that, “[g]iven [the press release’s] focus on [Boardwalk’s] revenues, and not on the maximum applicable rate that can be charged by [Boardwalk’s] interstate gas pipelines, we are not concerned that the release precludes any strategic analysis or action of the type that we were discussing.” JX 627 at 1. Loews’ edits

42

had paid off, and Alpert quickly forwarded the response to members of Loews’ senior management. JX 632 at 1.

Baker Botts also addressed whether the March 15 FERC Actions constituted a sufficient triggering event. On this issue, the answer did not meet Loews’ expectations.

Wagner explained that there were “two FERC actions that directly affect the analysis: the Revised Policy and the Notice of Inquiry.” JX 626 at 1. Absent further regulatory developments, neither would have an effect on Boardwalk’s rates:

The Revised Policy Statement, in which FERC announced its new policy prohibiting MLP-owned gas pipelines from including an income tax allowance in their cost of service, is effective now as a statement of FERC policy. Standing alone, it does not require pipelines to take any action but it announces how FERC intends to treat the issue on a going-forward basis.

The Revised Policy Statement will be implemented through the proposed regulations, which when adopted, will require all interstate gas pipelines to make informational filings revising their cost of service, which may lead to rate challenges. These regulations would be administrative in that they will not announce new policy. I expect that any litigated rate challenges would not be resolved and therefore result in decreased rates until 2020 at the soonest.

The second action is the Notice of Inquiry in which FERC is seeking comment on how to address overfunded deferred tax balances held by MLP pipelines. Comments will be due in late May, 60 days after the notice is published in the Federal Register. Any policy emerging from this proceeding would have the potential to further reduce gas pipelines’ cost of service.

Unlike the proposed rulemaking, FERC is simply gathering information and there is no proposed timetable for action. FERC may issue a Policy Statement on Deferred Taxes announcing a generally applicable policy or it may determine that it will address the issue in individual litigation. My best judgment is that FERC should act in this proceeding by the end of 2018. Any FERC decision is not likely to be self-implementing and would require additional proceedings to reflect the policy in pipeline rates.

Id. In simple terms, Wagner recognized that the March 15 FERC Actions did not have any immediate effect. The Revised Policy did not require any action, and nothing would happen

43

until FERC issued regulations. The same was true for the ADIT NOI. Even then, there would not be any effect on rates absent litigated rate cases.

Four minutes later, Alpert requested a second call with Baker Botts. JX 626 at 1.

During the call, Alpert criticized Wagner’s analysis as having “[t]oo much nuance.” JX 646 at 5. Alpert wanted a direct answer addressing when Loews could get the Opinion. Id. (“When do we can [sic] get [the] opinion? When [would it be] prudent to act?”).

Rosenwasser told Alpert what Loews wanted to hear. He said that the “most important thing has happened” so that “we’re already there.” JX 646 at 5. But because Wagner had provided a well-reasoned explanation supporting a different conclusion, Alpert asked Baker Botts to confirm Rosenwasser’s view that “we’re already there.” Id. at 7 (“2x check that we think issuance of [the Revised Policy Statement] is appropriate triggering event for issuing opinion.”). After the call ended, Alpert updated Loews’ senior leadership. Copying Rosenwasser, Wagner, and Swidler, Alpert reported that Baker Botts would analyze whether the Revised Policy was a sufficient trigger “in the context of all the facts and the likelihood of future actions changing materially the outcome of the conclusions that would support any opinion of counsel.” JX 625 at 1. Alpert also cautioned Loews’ executives to “address [all emails on this matter] to me and cc others so we can best argue communications are privileged.” Id. G. Baker Botts Reframes The Analysis.

Alpert scheduled a follow-up call with Baker Botts and Boardwalk for March 29, 2018. That gave Baker Botts just over a week to take a position on rendering the Opinion. To get to the outcome Loews wanted, Rosenwasser crafted a syllogism.

44

1. Rosenwasser’s Syllogism Rosenwasser knew that the Call Right was intended to address a business problem.

He was, after all, the one who drafted it. Rosenwasser Dep. 40 (characterizing Section 15.1(b) of the Partnership Agreement as “a business point . . . not a legal point”). The Call Right sought to protect Loews against a regulatory change that would have a materially adverse effect on Boardwalk. The provision referred to rates because rates generate revenue. The Call Right was not intended to create a trapdoor that Loews could open based on a regulatory change that had no real-world effect. Rosenwasser Dep. 45 (describing the Call Right as not “easy to trigger” as indicated by the fact that the “[O]pinion takes lots of thought and it takes lots of analysis to make certain that the [O]pinion could be given”).

But the Call Right’s reference to “rates,” combined with Loews’ careful parsing of that distinction when editing the March 19 press release, gave Rosenwasser an opening. Rosenwasser decided to take the view that the Call Right was not concerned with the actual economic impact on Boardwalk; it was only concerned with the abstract concept of “maximum applicable rates.” See JX 679 at 5, 8. If a regulatory change could have a materially adverse effect on the abstract concept of “maximum applicable rates,” then the Call Right could be exercised. And because a tax allowance had been part of the cost-of- service calculation, a policy change eliminating the tax allowance could be said to lead ineluctably to a change in that abstract concept.

On March 21, 2018, Rosenwasser explained his approach to Wagner, who took contemporaneous notes. JX 637. Wagner’s transcription memorializes the Rosenwasser syllogism:

45

1 – A pipeline charges COS [cost-of-service] rates 2 – COS includes ITA [income tax allowance]

[No] ITA → material effect No examination of FERC actions/shipper actions COS/over/under-recovery

Just saying [no] ITA = lower COS = MAE on

max applicable rates

JX 639 at 1. As Wagner correctly and immediately perceived, Baker Botts was “[j]ust saying” that no income tax allowance meant a lower cost of service, which would equate to a material adverse effect on maximum applicable rates. Id.

For Baker Botts, the beauty of Rosenwasser’s syllogism was that it did not require any type of predictive exercise about when an actual rate case might be brought or what the outcome of a full-blown, litigated, cost-of-service proceeding might be. See JX 639 at 1 (“No examination of FERC actions/shipper actions” or Boardwalk’s “over/under- recovery” of its pipelines’ costs of service). Indeed, the syllogism did not require any real factual analysis about the effect of the March 15 FERC Actions. The principal step involved elementary subtraction.

To implement Rosenwasser’s syllogism, Baker Botts asked Boardwalk “what would FERC allow them to charge” in a hypothetical world that assumed “there was a full mkt for services.” JX 646 at 3. Swidler found reassurance for this approach in the fact that the Call Right did not contain any language addressing “the commercial conditions that might prevail in setting rates (e.g., whether or not the pipeline’s capacity is in high demand).” JX 645 at 1.

46

Rosenwasser’s syllogism did not account for ADIT. No one knew what would happen with ADIT. See JX 644 at 1 (“[G]iven the lack of clarity on FERC’s eventual policy on this [ADIT] issue, [McMahon] had no estimates” concerning “the potential effect of a return of ADIT to ratepayers”). But Baker Botts knew that FERC’s treatment of ADIT could “affect the rate impact on the pipelines substantially.” JX 619 at 1. The known unknown of ADIT defeated Rosenwasser’s syllogism, but Baker Botts went ahead anyway.

2. The March 29 Memorandum In preparation for the scheduled meeting with Loews and Boardwalk on March 29, 2018, Baker Botts prepared a memorandum that worked through the issues that had to be resolved before Baker Botts could render the Opinion. JX 679 (the “March 29 Memorandum”). There were many, and Baker Botts resolved them all in Loews’ favor.

One issue was the Call Right’s use of the term “maximum applicable rates,” which had no established meaning in FERC regulatory parlance. The FERC lexicon equates the terms “maximum rates,” “tariff rates,” “cost-of-service rates,” and “recourse rates.” Only in the context of its capacity release regulations had FERC used a similar phrase— “applicable maximum rate.” PTO ¶ 89. An investor or a court might interpret the idiosyncratic insertion of the word “applicable” to refer to the actual rates applicable to a particular pipeline’s customers, including discounted rates or negotiated rates. Without an established meaning, the term could be regarded as ambiguous, and under the doctrine of contra proferentem, a court applying Delaware law would interpret the term against the general partner and its affiliates and in favor of the limited partners.

47

To solve this problem, the March 29 Memorandum interpreted “maximum applicable rates” as synonymous with “the maximum rates Boardwalk can charge, as a legal matter, not as an economic matter.” JX 679 at 5. Baker Botts asserted that the Call Right’s drafters would not have used the words “maximum” and “can be charged to customers” if they had meant for the Call Right to focus on the rates that Boardwalk actually charged its customers. Id. at 5–6. Without explanation, the March 29 Memorandum concluded that the word “applicable” “certainly does not mean actual.” Id. at 6.

To support its interpretation, Baker Botts looked to extrinsic evidence in the form of references in Boardwalk’s Form S-1 from its IPO. That document indeed contained passages that seem to equate “maximum applicable rates” with recourse rates. See PTO ¶¶ 90–91. Other Boardwalk filings, such as its Form 10-Ks, use the term in similar ways. See id. ¶ 92. Baker Botts also found orders that FERC issued in rate cases involving Boardwalk, where Boardwalk seemed to have used the term as a substitute for recourse rates. See JX 637 at 1. Baker Botts could not identify any broader uses of the term. Id.

Another issue was the need for an analysis of Boardwalk’s rates. One of the ostensible justifications for Rosenwasser’s syllogism was that the legal opinion addressed a question of law that did not require predicting the outcome of a rate case. The March 29 Memorandum could not keep up that pretense. Recognizing that factual analysis was required, the March 29 Memorandum stated, “Boardwalk will need to prepare an analysis of each pipeline’s regulatory cost of service” and counsel would need “certificates from Boardwalk’s officers” so that counsel could rely on it. JX 679 at 6. Recognizing that the

48

ratemaking principles would be implicated, the March 29 Memorandum stressed “[c]ounsel will need to review that analysis in detail to confirm that the analysis is being prepared consistent with counsel’s understanding of federal regulatory rate making requirements.” Id.

Yet another problem was how to interpret the term “material adverse effect.” If interpreted consistent with Delaware cases like In re IBP, Inc. Shareholders Litigation, 789 A.2d 14 (Del. Ch. 2001), and its progeny, then that standard would be difficult to meet. The Baker Botts team acknowledged that the drafters “did not want to make it easy” for there to be a sufficient effect. JX 679 at 7. But even though the term appeared in a partnership agreement governed by Delaware law, the Baker Botts team found “no reason to think the drafters of Section 15.1(b) intended to incorporate the meaning the Delaware courts have applied to merger and acquisition MAC clauses to the words ‘material adverse effect.’” Id. Instead, Baker Botts planned to interpret the phrase by looking to federal securities law, where “something is material if an investor would consider it important in making an investment decision.” Id.; see id. at 6 (“Those rates [that Boardwalk’s subsidiaries charge] are regulated by federal law. The opinion requested therefore involves an analysis of federal law.”). Baker Botts also asserted that the doctrine of contra proferentem would permit the Call Right to be interpreted in favor of its drafter—contrary to what the doctrine contemplates. See id. at 7. Once again, the March 29 Memorandum could not keep up the pretense that the analysis was purely a legal question. The memorandum concluded: “Materiality is not, however, a fundamentally [] legal concept.

49

Therefore, in giving any opinion required by Section 15.1(b), counsel will need to rely heavily on Loews and Boardwalk.” JX 679 at 7.

The March 29 Memorandum also flagged an issue raised by the Acceptability Condition: Who would determine on behalf of the General Partner whether the Opinion was “acceptable”? Would that determination be made by Holdings, the Sole Member of the GPGP, where all the decision-makers were Loews insiders, or would the decision be made by the GPGP Board, which included outside directors? JX 679 at 7–8. Baker Botts concluded that Holdings was the correct decision-maker. As Baker Botts saw it, because the General Partner could exercise the Call Right “at its option” and in its individual capacity, it did not make sense for there to be any constraint on the General Partner’s ability to determine in its own interest that the Opinion was acceptable. JX 679 at 7–8.

3. The March 29 Meeting On March 29, 2018, Rosenwasser and Wagner spoke with Alpert and McMahon as planned. They agreed on the outcome that favored Loews: The March 15 FERC Actions “met the procedural predicate” for the exercise of the Call Right. JX 688 at 1; see id. (“Policy Statement sets up factual predicate for [t]he P[artnership] [contract] [.]”). Even though the March 15 FERC Actions were not final, and despite the known unknown of ADIT, they decided that enough had happened for Baker Botts to proceed with the Opinion that could enable Loews to exercise the Call Right. H. The Financial Data To generate the Opinion, Baker Botts needed what the Opinion would refer to as “Financial Data.” Johnson took charge of providing it. On April 4, 2018, Johnson reported

50

that he had numbers that “should get us where we need to go.” JX 713 at 1. He sent McMahon an email attaching two analyses for use by Baker Botts, a “Form 501-G Analysis” and a “Rate Model Analysis.” JX 727 at 4.

The Form 501-G Analysis contained the information that Boardwalk would include in a Form 501-G filing if FERC adopted regulations consistent with the NOPR. The proposed Form 501-G contemplated that each pipeline would disclose its cost-of-service requirement for 2017 and how much revenue the pipeline actually collected. Each pipeline then would recalculate those figures using a tax allowance based on the lower tax rate of 21% established by the Tax Act and a hypothetical tax rate of 0% to reflect the absence of any tax allowance. JX 580 ¶ 32. The Form 501-G also included lines for amortization of ADIT, but it did not specify a methodology for treating ADIT. JX 558.

The following table summarizes Johnson’s Form 501-G Analysis:

Form 501-G Analysis:

35% Tax 21% Tax 0% Tax 35% COS 35% COS 21% COS 21% COS COS COS COS Delta % Change Delta % Change Texas Gas Pipeline 2017 $431.09 $406.47 $362.23 $68.86 15.97% $44.24 10.88% Gulf South Pipeline 2017 $640.21 $601.93 $534.50 $105.71 16.51% $67.43 11.20% Gulf Crossing Pipeline 2017 $275.50 $259.88 $232.30 $43.20 15.68% $27.59 10.62%

JX 727 at 4. In reaching these results, Johnson assumed that each pipeline’s ADIT balance would be returned to ratepayers through amortization over the life of each pipeline, an approach known as the “Reverse South Georgia Method.” Id. at 1. At that time, FERC had not decided how to treat ADIT balances. One option, which pipelines favored, would be to eliminate the ADIT balance entirely. Another option, which shippers favored, would be to

51

require a cash refund of the ADIT balance. Intermediate options involved amortizing the ADIT balance over various periods. The Baker Botts attorneys and Boardwalk executives knew that FERC could handle ADIT in a number of ways, each of which would result in a different outcome. Yet because they believed the Reverse South Georgia Method was the most likely, that was the only one they analyzed.

The Form 501-G Analysis did not include the actual revenue calculations that the Form 501-G contemplated. If Johnson had performed them, they would have shown that both Gulf South and Gulf Crossing were under-recovering their cost of service, generating ROEs that would not warrant a rate case, and were in no danger of having their rates lowered. See JX 644 at 1. Boardwalk’s actual Form 501-G submissions, filed in late 2018, confirmed that fact: Gulf South’s ROE in 2017 was 4.9%, and Gulf Crossing’s was 4.7%. Webb Report Ex. 16 at 6765 (Gulf Crossing ROE); id. Ex. 17 at 6770 (Gulf South’s ROE). Texas Gas, on the other hand, faced some risk of a rate case, because its indicative ROE was 24.3%, and historically FERC would file a rate case if a pipeline’s ROE was above 20%. JX 1064; Sullivan Dep. 168. Nevertheless, Wagner and Sullivan “share[d] the opinion that there is a low probability that Texas Gas would face a section 5 case in the next 1–2 years.” JX 1064 at 1. “Beyond that time frame,” they concluded, “there are too many variables to make a prediction with any confidence.” Id.

Johnson’s Rate Model Analysis followed the same basic steps as the Form 501-G Analysis. JX 727 at 2. But unlike the Form 501-G Analysis, which used FERC’s indicative ROE of 10.55%, Johnson performed the calculations in the Rate Model Analysis using an ROE of 12.0%. That decision increased the cost-of-service requirement. In his cover email,

52

Johnson explained that his choice of an ROE of 12.0% was “[t]he biggest driver as to the difference in Cost of Service from the Form 501-G analysis.” JX 727 at 2. At trial, Johnson testified that he found that rate in an annual report issued by a shipper-side advocacy group that lobbies FERC to pursue rate cases against pipelines. Johnson Tr. 617, 658–59. It was not an unreasonable selection, but it also was not a pro-pipeline selection. It is, however, another indication that Loews and Boardwalk did not think that the March 15 FERC Actions necessarily would be implemented as proposed.

The following tables summarize the results of Johnson’s Rate Model Analysis:

Rate Model Analysis:

35% Tax 21% Tax 0% Tax 35% Tax 35% Tax 21% COS 21% COS COS COS COS Delta % Change Delta % Change Texas Gas Pipeline-Overall System 2017 $424.34 $393.66 $346.57 $77.77 18.33% $47.09 11.96% Gulf South Pipeline 2017 $491.05 $457.04 $403.55 $87.50 17.82% $53.49 11.70% Gulf Crossing Pipeline 2017 $220.29 $198.62 $167.60 $52.70 23.92% $31.03 15.62%

35% Tax 21% Tax 0% Tax Indicative Indicative Indicative 35% Rate 35% Rate 21% Rate 21% Rate Rate Rate Rate Delta % Change Delta % Change Texas Gas Pipeline 2017 $0.2337 $0.2168 $0.1909 $0.0428 18.33% $0.0259 11.98% Gulf South Pipeline 2017 $0.3698 $0.3442 $0.3040 $0.0658 17.80% $0.0402 11.68% Gulf Crossing Pipeline 2017 $0.3549 $0.3200 $0.2700 $0.0849 23.92% $0.0500 15.62%

JX 727 at 4. The Rate Model Analysis thus resulted in a bigger percentage change than the Form 501-G Analysis.

Baker Botts used the Rate Model Analysis to render the Opinion. No one on the Boardwalk team prepared any sensitivity analysis using different treatments of ADIT or

53

different ROE calculations. See Webb Report ¶¶ 128, 134–35; JX 1757 (Webb Rebuttal) ¶¶ 29–30.

Although Johnson claimed at trial to have followed all of the steps of cost-of-service ratemaking in his analysis, he plainly did not. The Rate Model Analysis presented a hypothetical cost-of-service calculation, subtracted the income tax allowance, and concluded that the new total was lower. FERC does not calculate rates by changing a single element in a cost-of-service calculation. Instead, FERC evaluates all elements of a pipeline’s cost of service when calculating a pipeline’s rates. Court Report ¶¶ 146–48; Webb Report ¶¶ 129–33.

Sullivan, the rate expert that Baker Botts hired to assist with the Opinion, testified that the Rate Model Analysis was “not a recourse rate calculation.” Sullivan Dep. 151. While Johnson attempted to justify his approach by contending that the Rate Model Analysis generated an “indicative rate” for each of Boardwalk’s pipelines, Sullivan made clear that “an indicative rate doesn’t mean anything.” Id. 168–69. Sullivan explained that

[t]o really find out what the true rate reduction is, you have to do the billing determinant adjustments . . . where you take into account how much of the billing determinants are discounted, how much are negotiated discounted rates, how much is [interruptible transportation], how much are firm recourse rates. You have to do all those calculations to properly calculate a rate reduction.

Id. at 120. The Rate Model Analysis did not do that. Id.

In his cover email circulating the Rate Model Analysis, Johnson explained the limitations of the exercise he conducted. JX 727 at 2. As pertinent here, he stated:

In order to provide a comparable rate assessment for each of the assets to assist in business decision-making, we have provided indicative rates that are

54

postage stamp (i.e., every shipper pays the same maximum rate for each molecule) and unadjusted (i.e., does not adjust the maximum tariff rate for any under-recoveries of cost associated with either discounted or negotiated rate capacity that is below the maximum tariff rate). This provides the cleanest approach to understanding the relative rate impact of changes in the income tax rate and income tax policy within each of the three pipes and removes any argument as to subjective adjustments to volumes tied to a calculated rate reflected on the summary.

Id.

In reality, Boardwalk’s pipelines do not have just one rate. In April 2018, they had 167 total recourse rates on file with FERC.5 Those rates covered nine different pipeline zones and incorporated forty-six different rate schedules. Webb Report ¶¶ 91–93. In the real world, the “postage stamp” approach does not work for assessing the rates charged by Boardwalk’s subsidiaries.

The abbreviated analysis that Johnson conducted contrasts with the voluminous record generated for a rate case. In their most recent rate cases, Texas Gas and Gulf South submitted hundreds of pages of complex calculations to determine cost-based recourse rates. See Johnson Tr. 652–53. In stark contrast, the Rate Model Analysis contained approximately five pages of calculations for each pipeline. Id. at 640. The Rate Model Analysis gave no consideration to issues of competition, discounting, or other adjustments that would affect the determination of recourse rates in a FERC rate case.6 By assuming

5 Webb Report Ex. 1 at 8–26 (Texas Gas, 114 recourse rates); Webb Report Ex. 2 at 10–37 (Gulf South, 42 recourse rates); Webb Report Ex. 3 at 3–5 (Gulf Crossing, 11 recourse rates).

6 When competition pressures a gas pipeline to provide services at a discount to applicable recourse rates, the pipeline is no longer recovering its full cost of service. In its

55

that a change in cost of service would translate directly into a change in recourse rates, the Rate Model Analysis ignored critical elements of rate design. Johnson effectively admitted as much. Johnson Tr. 648–49, 651–52.

Perhaps most significantly, the Rate Model Analysis ignored the reality that rate changes are not self-implementing. Even if a pipeline’s cost of service changes, recourse rates do not change unless and until there is a litigated rate case.7 If a pipeline is unlikely to face a rate case, then it is all the more unlikely that its recourse rates will change.

The Rate Model Analysis made no effort to incorporate the risk of a rate case. It easily could have. The NOPR contemplated using the Form 501-G to assess the need for a rate case. FERC also identified factors that could obviate the need to change a pipeline’s rates, all of which applied to Boardwalk’s subsidiaries. Gulf South and Gulf Crossing faced

next rate case, the portion of the pipeline’s cost of service that would have been allocated to the discounted services is reduced, and the difference is reallocated to the pipeline’s less price sensitive customers. Webb Report ¶ 177. That way, FERC permits the pipeline to raise its remaining undiscounted recourse rates so that it can recover its full cost of service. All three of Boardwalk’s pipelines have emphasized in FERC filings that they face significant competition. The actual recourse rates of Texas Gas and Gulf South reflect that competition. Webb Report ¶ 95 (Texas Gas); id. ¶ 96 (Gulf South); id. Ex. 4 at 443; id. Ex. 6 at 626. Texas Gas and Gulf South earn less than a third of their revenue from recourse rates; Gulf Crossing earns essentially none. Id. ¶ 194. McMahon’s statement that Gulf Crossing will be undersubscribed by the time its contracts expire in 2023 evidences a likelihood that discount adjustments will figure prominently in its next rate case. Finally, in transmittal letters attached to the Form 501-G filings that Boardwalk submitted on behalf of its pipelines in late 2018, Johnson identified significant and apparently increasing competition as a reason FERC should not require them to lower their rates. See Webb Report ¶ 198 & n.175.

7 Wagner Dep. 77–79; Sullivan Dep. 79–80; see McMahon Tr. 507–08, 512–13.

56

no risk of a rate case in the foreseeable future. For Texas Gulf, the rate case risk was low through April 2020; beyond that, it was impossible to predict the likelihood of a rate case “with any confidence.”8 Yet the Rate Model Analysis implicitly assumed a 100% likelihood that all three pipelines would face a rate case immediately, lose the rate case, and each have their rates reduced by an amount determined by singe-issue ratemaking. I. Alpert Adds Skadden To The Team.

Shortly after hiring Baker Botts, Alpert hired Skadden, Arps, Slate, Meagher & Flom LLP (“Skadden”) to supplement the legal team. Alpert had considerable experience working with Skadden, and the firm had a deep bench in FERC matters, extensive experience with MLPs, and expertise in Delaware law. Rosenwasser Tr. 61–62; Alpert Tr. 326–27. Richard Grossman, a corporate partner, led the Skadden team. Jennifer Voss, a litigation partner in Skadden’s Delaware office, provided advice on Delaware issues.

Alpert hired Skadden after Rosenwasser suggested that bringing in another law firm to advise on whether the Opinion was acceptable might further protect Loews from liability. See JX 975 at 1. The Partnership Agreement contains language exculpating the

8 See JX 1064 (Wagner advising Loews that Texas Gas had a low rate-case risk for “the next 1–2 years”); see Wagner Tr. 245, 248 (Wagner testifying that there was some risk of a rate case at Texas Gas due to its ROE, but that because of FERC’s workload, a rate case was unlikely in the next one to two years); Johnson Tr. 632–34 (testifying that Texas Gas faced some risk of a rate case); see also JX 1807 at 6 (Wagner noting that Sullivan believed FERC would use an ROE of 20–30% to screen for rate cases). The defendants’ FERC expert testified at trial that Texas Gas would have an ROE of between 17.5% and 24.3%, which was high enough to create some risk of a rate case. See Kelly Tr. 1104. The plaintiffs’ rate expert agreed that FERC historically pursued rate cases when pipelines had ROEs in this range. Webb Tr. 1007–08.

57

General Partner and its Affiliates from monetary liability unless it engages in fraud, bad faith acts, or willful misconduct. PA § 7.8(a). The Partnership Agreement also states that the General Partner will be “conclusively presumed” to have acted in good faith if it “reli[ed] upon the advice or opinion [of legal counsel] (including an Opinion of Counsel).” Id. § 7.10(b). Rosenwasser and Alpert thought that if Skadden advised the General Partner that Baker Botts was qualified to render the Opinion and that the Opinion was acceptable, then those additional protections would apply.9 At the time, Alpert also thought that Skadden would handle any litigation challenging the exercise of the Call Right. See Alpert Tr. 445–46; JX 1136. He later would decide not to use Skadden for any litigation after Skadden balked at giving Alpert the advice he wanted.

The first issue that Skadden looked at was Baker Botts’ assertion that Holdings was the proper entity to decide whether the Opinion was “acceptable to the General Partner.” See JX 679 at 7–8. Rosenwasser had struggled with this question, which the Partnership Agreement did not plainly address. See JX 596 (Rosenwasser’s handwritten notes on the Partnership Agreement); Rosenwasser Dep. 65. By late March, Rosenwasser had taken the position that Holdings, rather than the GPGP Board, would determine acceptability. See JX 679 at 8. On March 27, Alpert suggested that Skadden “confirm” that “the redemption was the sole decision of the [General Partner]—such that the [GPGP] [B]oard will not need to act.” JX 669 at 1.

9 Rosenwasser Tr. 61; Alpert Tr. 325–26, 407; JX 1100 (Skadden engagement letter dated April 23, 2018).

58

Instead of confirming Rosenwasser’s position, Voss reached the exact opposite conclusion. In an insightful internal email that carefully worked through the issues, she expressed the view that “the MLP Agreement likely requires that the [GPGP] Board make the determination to accept the Opinion of Counsel. Or, at a minimum, it is ambiguous.” JX 747 at 1.

Skadden subsequently prepared a memorandum for Alpert, where Skadden framed its concerns in more lawyerly and less direct language. Skadden began by noting that the Call Right

is atypical and, to the best of our knowledge, notwithstanding the many MLP cases (and MLP contract terms) that have been litigated, no Delaware court has interpreted such a provision. . . . [I]t’s also fair to say that courts generally dislike the interpretive difficulties often inherent in MLP agreements. . . .

And, here, we think that any “question marks” or ambiguities likely would be decided against the “sophisticated drafter” and not the minority unitholders.

JX 773 at 1. Skadden also flagged arguments that a plaintiff could make about the circumstances surrounding the exercise of the Call Right, such as “purported efforts to depress the price of the units prior to the exercise of the right by, for example, increasing capital expenditure” or “purported partnership ‘admissions’ about the ‘lack of materiality’ of the FERC’s March 15 policy statement.” Id.

Setting aside those issues, Skadden agreed with Baker Botts that Holdings had the right to exercise the Call Right in its individual capacity. But Skadden perceived that to be a different question than who had the ability to determine whether the conditions for exercising the Call Right were met. Skadden noted the following:

59

• “[T]he ‘right to purchase’ . . . does not seem to arise unless and until certain preconditions exist, including acceptance by the General Partner of a specified ‘Opinion of Counsel.’” Id. at 2.

• “A plaintiff could argue that this Opinion of Counsel must be acceptable to the General Partner in its capacity as general partner and not in its individual capacity.”

Id.

• “[T]he words ‘exercisable at its option’ (indicating ‘individual capacity’) do not appear in the ‘precondition’ portion of the provision.” Id.

• “At a minimum, the matter is arguably ambiguous.” Id.

Skadden also discussed the structure of the Partnership Agreement. Skadden observed that if the Acceptability Condition existed to benefit the General Partner in its individual capacity, then it followed that an affiliate of the self-interested General Partner could determine acceptability. But if the Acceptability Condition was intended to introduce some check on the quality of the Opinion for the benefit of the limited partners, then enabling the self-interested General Partner to make the decision did not make sense. It was “akin to permitting the fox to guard the henhouse.” Id. at 3. Instead, “the added ‘layer’ of [GPGP] Board involvement serves a purpose and must occur before the right to call arises.” Id. Skadden reiterated that “at a minimum, there is arguable ambiguity here.” Id. To address the resulting litigation risk, Skadden recommended that the GPGP Board determine whether the Acceptability Condition had been met. Id. at 2. Skadden also recommended that the outside directors on the GPGP Board participate in and not abstain from the determination. Id. at 4.

Skadden plastered its analysis with caveats about its views being “preliminary” and “for discussion purposes only.” Id. at 1. Skadden also downplayed its internal conclusion

60

regarding ambiguity by adding the adjective “arguable” in the memorandum it provided to Alpert. Id. at 2. But the overall tenor of Skadden’s memo was clear, and Skadden presented its advice with the understanding that Loews would rely on it.

Loews begrudgingly did just that. Alpert and McMahon found Skadden’s recommendation “frustrating” and viewed the firm as a “pain in the ass.” See JX 874 at 1 (Layne handwritten notes); Layne Dep. 111–12. But consistent with Skadden’s reasoned analysis, Loews initially decided to have the GPGP Board make the acceptability determination. See JX 948 at 2; JX 979 at 1. J. Baker Botts Struggles With The Material Adverse Effect Inquiry.

By the second week of April 2018, Baker Botts was struggling with the need to conclude that the March 15 FERC Actions would have an effect that was both material and adverse. They wanted Skadden’s help. See JX 770 at 1; JX 772. But as a matter of firm policy, Skadden does not render opinions on whether an event constitutes a material adverse effect, and Grossman was not willing to give Baker Botts any analysis that might be construed as expressing an opinion on it. See JX 771 at 1.

For its part, Skadden was skeptical about the claim that a 10–15% change in a maximum applicable rate could be deemed in the abstract to qualify as a material adverse effect. JX 772 at 1. The Skadden attorneys believed that an 11% change in the maximum applicable rate was “likely insufficient” under Delaware law, although they acknowledged that the duration of the change would be a pertinent consideration. See id. The Skadden attorneys did not think anyone could assess whether a change in the range of 10–15% constituted a material adverse effect without delving into the facts. Id.

61

Alpert wanted Grossman to support Baker Botts. But during a call with Alpert, Grossman held the line on not providing any analysis that might be construed as an opinion on the existence of a material adverse effect. Alpert emailed his colleague, Tom Watson, that Grossman was “pissing [him] off.” JX 798 at 1. Watson’s response was more telling:

Yes, these calls are getting really annoying. Too many lawyers doing nothing but muddying the waters on what is a clear question (to me). If people think the language says that the relevant test is what is the real world effect, then we have an issue. I think it’s crystal clear that we’re talking hypothetical future max FERC rates.

Id. In other words, Watson understood that the material adverse effect analysis only worked under Rosenwasser’s syllogism based on “hypothetical future max FERC rates.” Under Rosenwasser’s syllogism, the answer was baked into the assumptions. But in the real world, the March 15 FERC Actions did not have any meaningful effect, much less a material and adverse effect.

Grossman ultimately agreed to provide Baker Botts with a description of the key cases “so that they did not miss a key case or an important factor looked at by the Delaware courts.” JX 777 at 1. Grossman also had Mike Naeve, a Skadden partner and former FERC Commissioner, speak with Wagner, Alpert, and McMahon about the various issues presented by the Opinion. See JX 790 at 2. Going into a call on April 10, 2018, Naeve had doubts about what “maximum applicable rates” meant. But after talking it over with the group, he thought that “recourse rates” was a more reasonable reading of “maximum applicable rates” than “the maximum rate that can be charged a specific customer under a negotiated or discounted rate agreement.” Id. To get Naeve “more comfortable” with the Baker Botts position, Wagner sent Naeve over 500 pages culled from Boardwalk’s Form

62

S-1 and the FERC orders involving Boardwalk’s pipelines that used the term “maximum applicable rates” as a synonym for recourse rates. Id.

After speaking with the Baker Botts team, Naeve identified a number of issues surrounding the material adverse effect analysis in discussions with Grossman and other Skadden partners. Naeve immediately flagged the question of whether any of Boardwalk’s pipelines actually faced a risk of a rate case. As Naeve explained,

[t]he risk that a customer will ask for a new rate case and that FERC will agree to grant that request will depend on whether there is substantial evidence that a new rate case will result in materially lower rates. A reduction in the revenue requirement to take out taxes would suggest lower rates, but it is possible that any reduction might be offset by other factors such as recent facility investments expenditures or changes in allowed ROE.

JX 800 at 1. In other words, Naeve recognized that whether the March 15 FERC Actions would have a material adverse effect on recourse rates depended on both the risk of a rate case and on the full ratemaking exercise that would take place in a rate case. It was much more than just a function of Rosenwasser’s syllogism and its subtraction of a tax allowance.

Naeve and his Skadden colleagues also discussed whether the inquiry into a material adverse effect needed to account for Boardwalk’s existing contracts for negotiated rates and discounted rates or any rate-case moratoriums at its pipelines. See JX 800 at 2. Those were real-world factors with real-world impacts, and FERC had cited them as reasons why a change in rates might not be warranted. But Baker Botts had no intention of taking those issues into account. Baker Botts instead was taking the position that

because pipelines are long-lived assets, and because the relevant language refers to the potential for material adverse rate effects in the future, their analysis need not be affected by discounts or moratoria that will be lifted within the next several years.

63

JX 800 at 2. K. Baker Botts Works Towards A “Preliminary” Opinion.

Rosenwasser wanted to be in a position to provide Loews with a “preliminary”

version of the Opinion by the end of April 2018. See JX 1956. The preliminary version would turn out to be an all-but-signed version that Baker Botts could render formally if and when Loews requested it.

Rosenwasser and his drafting team prepared an initial draft of the Opinion dated April 4, 2018. See JX 726 (the “April 4 Draft”). Like the preliminary Opinion and the final Opinion, the April 4 Draft was a non-explained opinion that identified background information, flagged assumptions, and stated a conclusion, but did not provide reasoning or cite authority to support the conclusion.

Throughout April, Rosenwasser and his drafting team worked with the senior Baker Botts lawyers comprising the ad hoc opinion committee. The senior lawyers raised a number of concerns that highlight how difficult it was for Baker Botts to reach the outcome necessary to render the Opinion.

A persistent problem was the meaning of “maximum applicable rates.” The April 4 Draft simply stated that it addressed “maximum applicable rates” without explaining how Baker Botts interpreted that term. JX 726 at 2. The next significant draft, dated April 17, 2018, sought to address the ambiguity inherent in the term by stating,

Based on the wording of Section 15.l(b)(ii) and supported by disclosure in the Registration Statement and discussions with representatives of the Partnership who assisted in preparing the Registration Statement, it is our judgment that . . . we should not consider the impact of negotiated rates, discounted rates, contractual rates, settlement rates, market-based rates, rate

64

moratoria, or other market-related factors when interpreting the term “maximum applicable rates that can be charged to customers.”

JX 935 at 2. That language telegraphed all the market-based, real-world considerations that Baker Botts was leaving out, and subsequent drafts continued to dispense with any analysis of the real-world impact of facts that would affect the actual “maximum applicable rates that can be charged to customers.” Rosenwasser continued to claim that the Opinion would not look at real-world effects, which he characterized as “speculation about real market conditions and their impact on rates.” JX 879 at 1.

Another persistent problem was that the March 15 FERC Actions would not have any effect on Boardwalk’s recourse rates unless those rates changed through a rate case. The April 4 Draft addressed that issue head on by expressly assuming that Boardwalk’s pipelines would file rate cases and take any other actions necessary to permit them to charge the reduced recourse rates that would generate a material adverse effect. See JX 726 at 2 (“[W]e have requested that the Partnership assume that the Subsidiaries will file rate cases and take any other appropriate and legal action to be permitted to charge the maximum rates permitted under the applicable cost of service rules and regulations regardless of competitive conditions or any other non-legal factor.”). But by including this explicit assumption, the April 4 Draft both highlighted the role of rate-case risk and openly assumed that Boardwalk and its subsidiaries would act contrary to their own interests. By April 17, Baker Botts had deleted this language and substituted an assumption that Boardwalk’s pipelines would charge customers their new recourse rates, without addressing how those rates would come about. The new assumption reached the same

65

result, but without advertising the counterintuitive premise. See JX 935 (omitting reference to Boardwalk’s subsidiaries filing rate cases).

Yet another problem was the fact that the March 15 FERC Actions were not final, could be revised significantly, and required clarification. The April 4 Draft contained language recognizing that reality, while assuming that the March 15 FERC Actions would not be revised. See JX 726 at 2 (acknowledging that “[i]mportant details of implementing the Revised Policy require clarification”). By April 17, Baker Botts had eliminated that acknowledgment of uncertainty. See JX 935 at 2. That draft instead sought to strengthen the assumption that the March 15 FERC Actions would not be revised, would be implemented as written, and would be applied by FERC in individual regulatory proceedings. See id. Subsequent drafts took the same approach. See id.

The senior Baker Botts lawyers also flagged other issues with the language of the Call Right. One debate concerned the reference to Boardwalk’s “status as an association not taxable as a corporation.” See JX 1958 at 1, 8; see also JX 878 at 2; JX 939 at 1. That phrase seemed to refer to Boardwalk’s status as an entity taxed as a partnership, but that created an issue for the Opinion because the Revised Policy did not affect all entities taxed as partnerships. It was thus difficult to say that Boardwalk’s status as an entity taxed as a partnership had a causal effect on the rates it could charge. See JX 1958 at 1; JX 1957 at 5.

The senior Baker Botts lawyers also questioned whether Baker Botts should be giving an opinion under Delaware law about the existence of a material adverse effect. See JX 878 at 4. The April 4 Draft only addressed federal law, and it did not contain any discussion of the term “material adverse effect.” See JX 726 at 2.

66

Once Baker Botts came to grips with the fact that the existence of a material adverse effect under the Partnership Agreement was a question of Delaware law, the firm was out of its depth. Baker Botts generally rendered enforceability opinions under the Delaware Revised Uniform Limited Partnership Act, but that was it. The firm did not render opinions more broadly on Delaware issues. See JX 878 at 4. By April 17, the draft included language which noted that the term “material adverse effect” was “not defined in the Partnership Agreement” and stated that Baker Botts had considered “what we believe to be relevant law.” JX 935 at 3. As Grossman had anticipated, the senior Baker Botts lawyers wanted to rely on Skadden’s work product on this issue. See JX 878 at 4–5; JX 892 at 2.

The senior Baker Botts lawyers also wanted reassurance on the Financial Data. The April 4 Draft referred only to information provided by the Partnership about its “cost of service . . . , and the related maximum rates that can be charged.” JX 726 at 2. By April 17, the draft contained language discussing the Financial Data and containing assumptions that it was “prepared in a reasonable manner and in good faith.” JX 935 at 3. By April 19, the extent of the assumptions regarding the Financial Data had grown further. See JX 1005 at 3.

Rosenwasser was concerned that the Financial Data alone might not be enough. He sought to bolster the case for a material adverse effect by asking Johnson to expand his analysis beyond the Financial Data to include projections for 2020 and add “DCF, EBIDTA, and EBIT (Operating Income) comparisons.” See JX 775 at 1; see also JX 797. He thus sought to include the real-world effects of changed rates when considering their

67

effect on Boardwalk, despite persisting in refusing to consider real-world effects when evaluating whether the March 15 FERC Actions would have any effect on rates.

During this timeframe, Johnson simplified the presentation of the Financial Data by dropping the scenarios that involved a tax rate of 35%. JX 775 at 3–4; JX 785 at 1–2. A version of the Financial Data from April 10 presented the information as follows:

Form 501-G Analysis:

21% Tax COS 0% Tax COS 21% COS Delta 21% COS % Change Texas Gas Pipeline

2017 $406.47 $362.23 $44.24 10.88% Gulf South Pipeline

2017 $601.93 $534.50 $67.43 11.20% Gulf Crossing Pipeline

2017 $259.88 $232.30 $27.59 10.62%

Rate Model Analysis:

21% Tax COS 0% Tax COS 21% COS Delta 21% COS % Change Texas Gas Pipeline-Overall System 2017 $386.21 $339.12 $47.09 12.19% Gulf South Pipeline 2017 $457.04 $403.55 $53.49 11.70% Gulf Crossing Pipeline 2017 $198.62 $167.60 $31.03 15.62%

21% Tax 0% Tax

Indicative Rate Indicative Rate 21% Rate Delta 21% Rate % Change Texas Gas Pipeline 2017 $0.2129 $0.1871 $0.0258 12.12% Gulf South Pipeline 2017 $0.3442 $0.3040 $0.0402 11.68% Gulf Crossing Pipeline 2017 $0.3200 $0.2700 $0.0500 15.62%

JX 775 at 3–4; JX 785 at 1–2. Compared to the April 4 figures, the percentages for Texas Gas in the Rate Model Analysis had creeped up from 11.96% (cost of service) and 11.91%

68

(indicative rate) to 12.19% (cost of service) and 12.12% (indicative rate). Otherwise, the figures remained the same as in the information Johnson had provided on April 4.

By this point, however, Boardwalk’s management team was preparing comments in response to the ADIT NOI and was focused on the implications of ADIT. Horton expressed concern that the Financial Data gave up Boardwalk’s argument that “the [Revised Policy] essenti8ally [sic] eliminates ADIT,” meaning that Boardwalk’s pipelines “do not have a reduction of rate base.” JX 797 at 1. He wanted to caveat Johnson’s analysis to make clear “that it does not include any impact from adjusting the ADIT balances to account for the reduction or the elimination of income taxes.” Id.

Like Boardwalk’s management, the Baker Botts lawyers knew that the treatment of ADIT would have a significant effect on the Financial Data. Wagner was representing the shippers on remand in the United Airlines case, so he understood that different industry participants were arguing for different outcomes.

The Baker Botts team had retained Sullivan as a rate expert,10 and Wagner asked Sullivan to examine how the Financial Data treated ADIT:

It seems to us that different assumptions on how to handle that issue could affect the calculations. Have they assumed that they will flow back the ADIT over the remaining life of the assets (with the corresponding reversals of the reduction to rate base)? Or is there another method used here?

10 Sullivan had thirty-eight years of experience working in the oil and gas industry, including twenty-five years working at FERC, and he had testified in FERC proceedings more than fifty times. PTO ¶ 154; JX 1498 at 151. His expertise is unchallenged.

69

JX 868 at 2. Sullivan reported that Johnson was using the Reverse South Georgia Method, which Sullivan thought was appropriate. See JX 868 at 1. Boardwalk’s executives and the Baker Botts lawyers thought that was the most likely regulatory outcome. But they also understood that the approach FERC took on ADIT would have a big effect. Wagner’s handwritten notes show him regularly wrestling with the uncertainty generated by how FERC would treat ADIT. See JX 646 at 8; JX 1400 at 1; JX 1807 at 3–4 (“[T]he effect on ADIT is unknown & unknowable.”). In one set of notes, he commented, “Will want to run scenarios on ADIT flowback.” JX 1807 at 12. Another set of notes stated: “ADIT NOI – Policy Statement w/ no immediate effect. 501-G filings do not acct for ADIT. No idea what they’ll do w/ ADIT. If there’s litigation coming from 501-Gs, ADIT policy will prob factor in there.” JX 1216 at 3.

After conducting further review of the Financial Data, Sullivan advised Wagner that “the spreadsheet work done by Boardwalk appropriately represents the cost of service for each Boardwalk interstate pipeline, the federal income tax impact at 21%, and the potential reduction in the cost of service for each pipeline if FERC reduces the income tax allowance to 0.” JX 960 at 2 (emphasis added). Wagner did not think that a statement about a cost of service analysis was sufficient. He asked Sullivan to let him know “[o]nce you’re able to state definitively that you agree with their rate analyses.” Id. (emphasis added).

On April 18, 2018, Sullivan told Wagner that he had finished his review. He did not provide the representation that Wagner wanted. Instead, Sullivan stated:

I have confirmed that Boardwalk has properly used the correct financial and accounting entries in the calculated cost of service for each of its pipelines.

In my expert judgment Boardwalk’s spreadsheets provide an accurate

70

presentation of the cost of service impact of the January 2018 federal income tax change from 35% to 21%. Boardwalk’s spreadsheets also provide an accurate presentation of the cost of service impact of the potential reduction in the cost of service for each pipeline if FERC eliminates the federal income tax allowance for MLP owned interstate pipelines as proposed in Docket No.

PL17-1.

JX 960 at 1 (emphases added). In his deposition, Sullivan explained persuasively that the Financial Data did not attempt to engage with principles of rate design and did not address the risk of a rate case. See Sullivan Dep. 101, 126, 149, 150–51.

In a separate call with Loews, Sullivan addressed the risk of a rate case at Texas Gulf, where the Financial Data indicated an ROE of approximately 24.3% after the elimination of the tax allowance and using the Reverse South Georgia Method for ADIT. Although returns at that level had caused FERC to initiate rate cases in the past, Sullivan thought that resource constraints on the agency meant that the probability was low that Texas Gas would face a rate case in the next one to two years. JX 1064 at 1. The likelihood of a shipper filing a rate case was also low. See id. No one thought that the risk of a rate case at Gulf Crossing or Gulf South was worth discussing.

Sullivan’s work confirmed what everyone knew. In the real world, any potential effect on Boardwalk’s rates could not be understood without a FERC determination regarding ADIT. And even if FERC implemented the March 15 FERC Actions, the regulations would not have a material adverse effect on Boardwalk’s rates because there was no risk of a rate case at Gulf Crossing or Gulf South and only a low risk of a rate case at Texas Gas. The March 15 FERC Actions only had an effect in the hypothetical world of

71

Rosenwasser’s syllogism, and only if supported by a coterie of assumptions necessary to generate the result that Loews wanted. L. Baker Botts Calls On Richards Layton.

As noted previously, the senior Baker Botts lawyers wanted to be able to rely on Skadden’s work product for purposes of the material adverse effect issue. See JX 878 at 4– 5; JX 892 at 2. When they received Skadden’s description of the Delaware cases, it fell short of their expectations. See JX 913 at 1 (Baker Botts attorney David Kirkland telling Rosenwasser, “I was expecting more analysis than this”); see also JX 936 at 1. Rather than analyzing the Call Right, Skadden’s memorandum explicitly disclaimed any intent to do so. JX 900 at 2.

Seeking to reassure his partners that Baker Botts still should render the Opinion, Rosenwasser reported that Loews only would exercise the Call Right if Skadden advised that Baker Botts’ Opinion met the Acceptability Condition. JX 913 at 1. Rosenwasser’s partners wanted that condition built into the Opinion, so the Baker Botts attorneys added language to the preliminary draft which stated that Baker Botts’ Opinion was “based on,” and its delivery “conditioned on,” the fact that “other counsel has advised [the General Partner] that [its] reliance on this opinion when delivered should provide the benefits set forth in Section 7.10(b) of the Partnership Agreement.” JX 1955 at 6 (draft from April 17, 2018); JX 1959 at 7 (draft from April 18, 2018). Perhaps anticipating pushback from Skadden, Baker Botts subsequently eliminated the “based on” and “conditioned on” language. See JX 1960 (draft from April 19, 2018).

72

The senior Baker Botts lawyers also wanted reassurance on the analysis of a “material adverse effect.” And Baker Botts was on a deadline, because Loews had made clear that it wanted an indication from Baker Botts that it could deliver the Opinion by Friday, April 20, 2018. Rosenwasser knew that Boardwalk and Loews had quarterly security filings to make and that Loews’ CEO, Jim Tisch, was planning to hold board meetings before the end of month to approve those filings. Rosenwasser understood that Tisch wanted to know where Baker Botts stood going into those meetings. See JX 914 at 1.

To satisfy his partners, Rosenwasser contacted Srinivas Raju, a partner at Richards, Layton & Finger, P.C. See JX 957 at 1; JX 975 at 1. In a call on Wednesday, April 18, 2018, Rosenwasser told Raju that a FERC rate expert had modeled a “decrease of 12.19% on top line revenue” for Texas Gas, an “11.70% decrease” for Gulf South, and a “15.62% decrease” for Gulf Crossing. JX 975 at 1; see also id. (“top line revenue impact – excess of 10% impact”). In reality, those figures referred to the percentage changes in cost of service and indicative rates under the Rate Model Analysis that Johnson prepared. JX 775 at 3; JX 785 at 2. Those figures would only translate into a comparable effect on topline revenue if Boardwalk’s subsidiaries charged recourse rates for a high percentage of their volumes. They did not.

Rosenwasser also told Raju that the FERC rate expert had projected that EBIT would decrease by 21–22% and distributable cash flow would decrease by “closer to 25%.” JX 975 at 1; see also id. (“21% decline in net income” and “even higher in distribution”). Sullivan had not addressed the effect on EBIT or distributable cash flow. Sullivan Dep.

73

140–42 (discussing final Financial Data in JX 1398); see also id. at 141 (Q: “Did you offer an opinion regarding the calculation of DCF, EBITDA or EBIT?”; A: “I do not believe I did specifically cite to EBITDA, EBIT or the DCF.”). Rosenwasser told Raju about those factors because he wanted to be able to consider real-world effects on Boardwalk’s business, as well as real-world stock market reactions, when determining whether a material adverse effect had occurred.11 Yet he continued to want to ignore the real-world reasons why the March 15 FERC Actions would not have any material effect on the rates that Boardwalk’s pipelines could charge.

Having provided these representations, Rosenwasser asked Raju to consider whether “material adverse effect” is “only measured based on the effects on the ‘maximum rate’ or is . . . measured by the effect on the business as a result of the decline in the maximum rate.” JX 975 at 1; see also JX 957 at 2. He also asked whether Richards Layton

11 Rosenwasser’s back-up memorandum offers further insight into what he wanted to consider to reach a conclusion that the effect on Boardwalk was “not immaterial.” PTO ¶ 161. There, he wrote:

The fact that so many regulated pipelines have requested that the FERC reconsider the Revised Policy is an indication they considered the changes caused by the Revised Policy are not immaterial. The magnitude of the adverse effect that the Revised Policy had on the trading market for many MLPs that own regulated pipelines is an indication that the matter is not immaterial. The fact that several MLPs that owned regulated pipelines have indicated that they are converting to corporate tax status is an indication that the matter is not immaterial.

Id.

74

could support the assertion that an adverse effect in “excess of 10%” would be sufficient under Delaware law. JX 1502 at 21.

Less than twenty-four hours later, Raju and his team gave advice orally to Baker Botts via teleconference. JX 956 at 1. Raju advised that the “[b]etter [r]eading” was to “look [at] rates more, not effects.” JX 1007 at 1. He also cautioned that a Delaware court would “construe ambig[uity] ag[ai]nst [the] drafter.” Id.

In response, the Baker Botts team clarified that their rate expert had not analyzed the Revised Policy’s effect on Boardwalk’s rates. Instead, the analysis considered “Hypothetical Rates.” JX 1007 at 1. Notes taken by a Baker Botts partner reveal that everyone focused on the core issue: There would be “no actual change—no effect yet screw min[ority].” Id. That was obviously a “challenging fact.” Id.

Turning to the magnitude of the change in rates that would be necessary for a material adverse effect, Raju advised that he would have a “hard time saying [12% in perpetuity is] not material.” Id. at 2. Raju noted that there was “not a lot of precedent” and,

75

in any event, “no cases against us” because “MAC cases [are] different” and the rate change was assumed to have an effect in “perpetuity.” Id.

Raju agreed to put his advice into an email. But he cautioned that it would be caveated by “assumptions and carve-outs” and say “[n]othing stronger” than that existence of a material adverse effect based on a change of 12–13% to rates in perpetuity represented the “better argument.” JX 975 at 1. Raju also stressed that Baker Botts could not reference his advice in the Opinion. Id.; see Raju Dep. 113–14.

Raju’s advice reassured Rosenwasser’s partners. After the call, Rosenwasser emailed Raju, telling him “[y]ou are so good.” JX 1003 at 1. Baker Botts sent Richards Layton a copy of their preliminary opinion. The next day, Raju told Baker Botts, “We stand by what was discussed on the call yesterday, and nothing in the draft opinion changes our thinking.” JX 1031 at 1. M. Baker Botts Makes Clear That It Can Deliver The Opinion.

As noted, Loews had been pushing Baker Botts to provide an indication that it could deliver the Opinion, and Loews wanted an answer by Friday, April 20, 2018. See JX 914 at 1. After his call with Raju, Rosenwasser told Alpert and Siegel that there was “no show stopper yet,” but that Baker Botts still needed to secure internal approvals. See JX 1006 at 1. Alpert and Siegel were not pleased. Id.

The internal approval that Rosenwasser needed was signoff from the firm’s chairman, Andy Baker. Baker could not provide the signoff by Friday because he was in the United Kingdom attending his daughter’s wedding. Rosenwasser told Loews that because of Baker’s absence, Baker Botts would not be able to get his signoff until Monday.

76

JX 1019 at 2. That did not sit well with Loews. Siegel wanted to know why Baker Botts had not raised this issue earlier, since “[t]hey must have known for weeks that Baker would be in London.” Id. Jim Tisch wanted Alpert to ask Rosenwasser “why they didn’t anticipate this problem, and whether this is an indication that there may be a problem with the opinion committee.” JX 1020 at 1.

Alpert told Tisch and Siegel that Rosenwasser was just “trying to be emotionally intelligent with his partners in an effort to obtain the desired result.” Id. at 1. But he nevertheless pressed Rosenwasser “to make absolutely sure” that there was no way to reach Baker on April 20. JX 1033 at 3. On April 20, 2018, at 6:47 a.m., Alpert asked Rosenwasser for a call that morning. JX 1059. One hour later, at 7:51 a.m., Alpert sent a follow-up email. He told Rosenwasser that “[y]our timing affects many things, especially our disclosure, [Siegel’s] conversations with board members and Loews special board meeting being held next week.” JX 1033 at 3. He also conveyed that the senior Loews executives did not understand why no one anticipated the issues created by Baker’s absence. Id.

Eleven minutes after the second email, Rosenwasser emailed his partners, telling them that Tisch “need[ed] board support for his plans” and “need[ed] to tell [the] board this afternoon” about whether Baker Botts could issue the Opinion. JX 1032 at 1. In response to Rosenwasser’s email, Baker Botts attorneys David Kirkland and Mike Bengtson separately considered whether to try to reach Baker. Id. Kirkland told Bengtson that he had “already been lobbied by Mike R[osenwasser] this morning to let him give Jim T[isch] the thumbs up this morning.” Id.

77

Rosenwasser’s lobbying was successful. Around 11:00 a.m., Rosenwasser emailed Alpert that “we are still working but believe at this point that we will be able to give the General Partner the Opinion of Counsel if and when requested.” JX 1065.

At 12:09 p.m., Rosenwasser sent Alpert a draft of the Opinion. JX 1045 at 1 (the “Preliminary Opinion”).12 The Preliminary Opinion was in substantially the same form as the final Opinion delivered more than two months later on June 29. Compare JX 1045 (Preliminary Opinion) with JX 1522 (Opinion). N. Skadden Makes Clear That It Will Say That The Opinion Is Acceptable.

After securing a “thumbs up” from Baker Botts, Alpert sought confirmation from Skadden that, if and when asked, it would advise the GPGP Board that the Opinion was “acceptable.” Alpert anticipated that Skadden’s advice would protect the GPGP Board when determining that the Opinion was acceptable for purposes of the Acceptability Condition. Everything would be buttoned down.

After receiving the draft from Baker Botts, Alpert forwarded it to Grossman and asked for an answer by the afternoon of Tuesday, April 24 “at the latest.” JX 1121 at 1.

12 At his deposition, Rosenwasser denied that Baker Botts provided Loews any commitment on April 20. Instead, he claimed that Baker Botts gave Loews an indication that it was “more likely than not” that Baker Botts could deliver the Opinion. Rosenwasser Dep. 122, 129, 257–82. That testimony was not credible. Baker Botts made clear that it was prepared to deliver the Opinion if asked. See JX 1234 at 2; Grossman Dep. 76–77. Loews did not want to receive the formal Opinion at the end of April because it would create a disclosure issue and start a ninety-day clock for Loews to exercise the Call Right. Loews wanted to control the timing of the issuance of the Opinion, which would start the clock for exercising the Call Right.

78

Skadden objected to the language in the draft stating that other counsel “has advised you that your reliance on this opinion when delivered should provide the benefits set forth in Section 7.10(b) of the Partnership Agreement.” JX 1056 at 5. Skadden had feared that Baker Botts would try to rely on its work, and the Skadden attorneys viewed this language as a backdoor attempt to do that. See JX 1094 at 1. Skadden asked to strike language. JX 1126 at 1.

Alpert was furious, and he “threatened to fire Skadden.” JX 1116 (“I told Skadden tell me today if [they] can’t get there or I’ll hire other counsel.”). Alpert told Rosenwasser he was “in no mood to negotiate with [Skadden]” and that he had “senior management back-up to move to another firm if [Skadden] is not reasonable.” JX 1113 at 1. In an email to Skadden, Alpert made his expectations “absolutely clear.” Id.

I thought we were absolutely clear on the following, but if not, we need to be. I need to know that if we ask for the opinion from Baker Botts, that Skadden can and will advise the [GPGP] [B]oard that based on Baker Bott’s [sic] experience, the diligence and process they conducted, the wording of the opinion and other factors, it is reasonable for the board to accept the Baker Botts opinion.

Id. at 1–2.

Skadden relented. Alpert told his colleagues that Skadden “fell into line,” but that he “[r]eally had to beat on them.” JX 1136 at 1. Alpert had planned to use Skadden for any litigation challenging the exercise of the Call Right. Now he decided that he would “look to other firms re potential litigation.” Id.

79

O. Boardwalk’s Public Comments On The NOPR While Baker Botts was working on a legal opinion that treated the NOPR and other March 15 FERC Actions as final, Boardwalk’s management team filed public comments on the NOPR, consistent with the fact that it was not final. It was the eventual regulations, not the NOPR, that would matter. Indeed, Naeve, the former FERC commissioner, noted that “If I were Baker Botts I would prefer to wait until FERC acts on the comments.” JX 1076 at 1.

On April 25, 2018, Boardwalk filed its public comments on the NOPR. JX 1139.

Rosenwasser printed out a physical copy of the comments and made handwritten annotations. See JX 1130. A section that addressed the treatment of ADIT caught his attention, and he underlined and double-starred key text:

Id. at 14. That, of course, was exactly what Baker Botts was doing in the Opinion— purporting to correctly assess the impact of FERC’s actions on its pipelines’ costs of service. And Baker Botts was relying on a Rate Model Analysis that largely paralleled the Form 501-G analysis, which Boardwalk said “will be misleading and inaccurate” unless and until FERC had addressed ADIT. And Baker Botts was going further. Baker Botts was not just addressing cost of service. Under Rosenwasser’s syllogism, Baker Botts was

80

claiming that eliminating one component of the cost of service—the income tax allowance—would have a material adverse effect on maximum applicable rates.

When Skadden saw the comments the next day, Voss focused on the same passage.

She noted dryly, “this seems to be relatively unhelpful.” JX 1207 at 2. Another Skadden attorney asked if the comment “could be problematic.” Id. at 1.

The passage that Rosenwasser double starred appeared within the following larger section that Rosenwasser annotated:

2. The Commission Must Align the Timing of Its Actions Under This NOPR and the ADIT NOI.

Contemporaneous with the NOPR, the Commission has issued the ADIT NOI, which seeks comment on how the Commission should address changes related to ADIT as a result of the Revised Policy Statement. ADIT is a critical issue in analyzing a pipeline’s maximum recourse rates.

Although ADIT is a non-cash item—merely the function of the timing

☆

difference between book depreciation and tax depreciation—certain shippers have and will continue to argue that ADIT should be treated in a manner that results in a large and immediate cash refund from the pipelines.

Significant dollars and the validity of certain portions of the Form No. 501-

G are at stake. The Commission should not sideline the ADIT issue while it attempts to rush the Form No. 501-G NOPR to be ready for decision by its July meeting.

ADIT is a key element of the proposed Form No. 501-G, and the ADIT NOI raises a number of questions fundamental to the treatment of this rate component under the Revised Policy Statement. For example, will the Commission adhere to normalization methodologies? The uncertainty surrounding how to handle ADIT is particularly problematic for an MLP like Boardwalk, which, as a result of the Revised Policy Statement, owns pipelines that are no longer allowed to collect income taxes in their rates but still have large ADIT balances on their FERC books. Boardwalk intends to address these and other questions in more detail in response to the ADIT NOI.

81

Until the Commission provides a final decision on the treatment of ADIT,

☆☆

Boardwalk cannot correctly assess the impact of the Revised Policy Statement and ADIT on its pipelines’ costs of service, and any response in the Form No. 501-G will be misleading and inaccurate.

The comment date for the ADIT NOI is not until May 21, 2018 (approximately thirty days after comments are due in this NOPR proceeding), and the date of final Commission action on the ADIT NOI is unknown. It is improper for the Commission to require the industry to complete a new form, a key element of which is directly tied to the cost of service intended to be addressed by the Form No. 501-G, and which is still under review. Without resolution of the ADIT issues, the Form No. 501-G will be misleading and inaccurate, and will substantially hamper a pipeline’s ability to have meaningful settlement discussions with its customers, since the calculation of a key element of rate base will be subject to change. Pipelines may also be discouraged from selecting the option to file a limited section 4 rate case with the potential to face additional risk regarding ADIT in a subsequent proceeding which would render that proposed option in the NOPR moot.

The Commission must resolve the issues raised in the ADIT NOI at the same time or before it issues a final rule in this proceeding to ensure that pipelines have the necessary information to complete the Form No. 501-G accurately, select the appropriate filing option, and/or to engage in meaningful settlement discussions with their customers.

JX 1130 at 13–15 (underlining and annotations in original) (footnotes omitted).13 In this passage, Boardwalk explained that without a determination on ADIT, matters were so unsettled that pipelines could not even have meaningful discussions with shippers about rates. Yet Baker Botts was claiming for purposes of its Opinion that matters were so

13 At trial, Rosenwasser claimed that he was not “reading it that closely” and that he starred or double-starred passages so that he could “go back and read it again.” Rosenwasser Tr. 82. That testimony was not credible. Rosenwasser underlined, starred, and double-starred aspects of Boardwalk’s comments because they fatally undermined the syllogism that drove the Opinion. Revealing that he was reading the comments for problematic language, Rosenwasser wrote “nothing bad here” next to a passage reciting the procedural history of the ADIT NOI. JX 1130 at 9.

82

settled that the firm could opine as a matter of law that the March 15 FERC Actions would have a material adverse effect on Boardwalk’s recourse rates.

Other aspects of the comments were equally problematic for purposes of the Opinion. For example:

• Boardwalk pointed out that the Policy Statement was “not a binding rule” and that FERC had not justified its application. JX 1139 at 2. The Opinion treated the Policy Statement as a binding rule. Rosenwasser drew a line next to this paragraph, and also made an unintelligible note. JX 1130 at 2.

• Boardwalk objected to FERC instructing pipelines to complete the Form 501-G that evaluated changes in cost-of-service requirements based solely on changes in income taxes, then using the revised cost-of-service requirements to identify an “Indicative Rate Reduction.” Boardwalk explained that using that procedure to establish rates constituted improper “single-issue rulemaking.” JX 1139 at 12, 30– 31; see JX 1296 at 9. The Rate Model Analysis on which the Opinion depended took the same approach that Boardwalk criticized.

• Boardwalk made clear that the Commission’s treatment of ADIT was not known and that different outcomes were possible. See JX 1139 at 13–14. Yet the Rate Model Analysis operated as if the treatment of ADIT under the Reverse South Georgia Method was a known fact.

• Boardwalk asserted that its “fixed negotiated rate agreements—almost all of which expressly state that they will apply ‘without regard’ to the pipeline’s maximum or minimum applicable rates—should not be affected by any potential impact to recourse rates.” JX 1139 at 16. The Opinion ignored the existence of Boardwalk’s fixed negotiated rate agreements.

• Boardwalk asserted that there is no impact on Gulf South’s revenue requirements due to the rate case moratorium that extended through May 1, 2023. JX 1139 at 20.

The Opinion ignored the existence of the rate moratorium and assumed a rate impact at Gulf South.

What Boardwalk conspicuously did not argue in its comments was that FERC should eliminate the ADIT balance entirely as a natural consequence of removing the income tax allowance. Boardwalk instead argued that FERC should instruct pipelines to

83

amortize the ADIT balances over the remaining depreciation life of the asset, using the Reverse South Georgia Method. That was the method that Boardwalk was using in the Rate Model Analysis, and it was where Boardwalk management and Sullivan thought FERC ultimately would come out.

Many other pipelines, however, argued explicitly that FERC should eliminate the ADIT balance entirely. PTO ¶ 337. Shippers generally took the opposite side of the issue, arguing that FERC should require pipelines to pay a cash refund of the ADIT balance or require amortization on an accelerated schedule. Id. ¶ 339. P. Loews Prepares To Make The Potential Exercise Disclosures.

Well before Baker Botts gave Loews the “thumbs up” that it could issue the Opinion if and when asked, Loews took a number of steps in anticipation of exercising the Call Right.

One task involved preparing the disclosures that Boardwalk and Loews would issue in their quarterly reports on their respective Form 10-Qs, assuming Baker Botts gave the anticipated “thumbs up.” Those discussions involved Loews, Boardwalk, Baker Botts, and Skadden, as well as Loews’ outside securities counsel Davis Polk & Wardell LLP, and lawyers from Vinson & Elkins. Id. ¶ 229.

The evolution of Boardwalk’s Form 10-Q reveals at least two things. First, there was a widespread understanding that the March 15 FERC Actions were not final, that their effects could not be predicted, and that they would not be likely to have a material adverse impact on Boardwalk. Second, despite that widespread understanding, Loews pushed the disclosures in a contrary direction that would facilitate the exercise of the Call Right.

84

On April 4, 2018, Baker Botts sent Loews a first draft of the Boardwalk Form 10-

Q. Id. ¶ 230. The draft contained relatively nuanced disclosures about the March 15 FERC Actions, including that “[i]mportant details of implementing the new policy statement require clarification and the Company will continue to assess the financial impacts as more information becomes available.” Id. Similar statements about the lack of finality surrounding the March 15 FERC Actions did not appear in the final Form 10-Q.

On April 4, 2018, Vinson & Elkins sent Boardwalk a first draft of the Form 10-Q.

Id. ¶ 232. Like the Baker Botts draft, it flagged that the March 15 FERC Actions were not final and noted that “[r]equests for rehearing or clarification of the Revised Policy Statement may change the outcome of the FERC’s decision on these requests.” Id. It stated that as a result, the “impacts that such changes may have on the rates we can charge for natural gas transportation and storage services are unknown at this time.” Id. The draft likewise observed that the NOPR proposed a new rule, that rule was not final, and that, as a consequence, “[a]t this time, we cannot predict the outcome of the NOPR, but adoption of the regulation in its proposed form could impact the rates we are permitted to charge our customers.” Id. ¶ 233. The Vinson & Elkins draft also recognized that the treatment of ADIT was an open issue and that there was no necessary connection between the elimination of the income tax allowance and a change in the treatment of ADIT and a reduction in rates, explaining that “[a]lthough changes in these two tax related components may decrease, other components in the cost-of-service rate calculation may increase and result in a newly calculated cost-of-service rate that is the same as or greater than the prior

85

cost-of-service rate . . . .” Id. ¶ 236. Similar statements did not appear in the final Form 10- Q.

On April 10, 2018, McMahon circulated his draft, using the Vinson & Elkins draft as a starting point. Id. ¶ 237.

• McMahon retained the statement that “requests for rehearing or clarification of the Revised Policy Statement may change the outcome of the FERC’s decision on this issue” and stated that the “ultimate outcome regarding the Revised Policy Statement could impact the maximum rates we are permitted to charge.” Id. ¶ 238.

• McMahon retained the statement that “any potential impacts from final rules or policy statements issued following the NOI on the rates we can charge for transportation services are unknown at this time.” Id. ¶ 239.

• McMahon added language stating that Boardwalk “cannot predict the outcome of the NOPR, but adoption of the regulation in its proposed form could ultimately impact the rates we are permitted to charge our customers.” Id. ¶ 240.

The Boardwalk draft was thus relatively neutral and balanced.

Later on April 10, 2018, Alpert circulated Loews’ comments, which took a different approach.

• The Loews draft stated, “we do not expect the FERC to reverse [the Revised Policy Statement] or otherwise revise the policy in a manner favorable to master limited partnerships.” Id. ¶ 245.

• Loews deleted the language stating that “[a]t this time, we cannot predict the outcome of the NOPR, but adoption of the regulation in its proposed form could ultimately impact the rates we are permitted to charge our customers.” Id. ¶ 246.

• Loews added language stating, “[a]s we do not expect FERC’s Revised Policy Statement to be reversed or modified in a manner favorable to master limited partnerships, we believe that our status as a pass-through entity for tax purposes will reasonably likely in the future have a material adverse effect on the maximum applicable rates” that Boardwalk’s subsidiaries could charge. Id. ¶ 247.

• Loews added language stating, “[i]n addition, the ultimate outcomes of the NOI and NOPR may have further material adverse effects.” Id.

86

Viewed charitably, Loews sought to characterize events in a way that would facilitate Loews’ exercise of the Call Right.

McMahon and Horton objected to aspects of the Loews draft. Horton believed that Loews’ language resulted in Boardwalk “rendering an opinion on the materiality issue.” Id. ¶ 249. McMahon regarded the draft as tilted in favor of Loews. Id.

A push and pull ensued over the disclosures. See id. ¶¶ 250–64. Loews took a particular interest in eliminating the language which stated that “[a]lthough changes in these two tax-related components may decrease, other components in the cost of service rate calculation may increase and could result in a newly calculated cost of service rate that is the same as or greater than the prior cost of service rate.” See id. ¶ 254. Loews also pushed for language focusing on the effects on Boardwalk’s rates, rather than on revenue or other aspects of Boardwalk’s business. See id. ¶¶ 263–64.

In addition to editing Boardwalk’s disclosures, Loews analyzed the effect of the disclosures on the trading price of Boardwalk’s common units with the assistance of investment bankers from Barclays. See id. ¶¶ 271–74. The analyses projected a short-term bump in the trading price, followed by a steady decline over time. See JX 822; JX 882; JX 915; see also JX 1051 at 3. Barclays attributed the decline in part to “[u]ncertainty regarding timeline” and the “[p]robability Loews doesn’t” exercise the Call Right. JX 915 at 15–16. Because the lower trading price would feed back into the formula for the Call Right, Loews would pay a lower exercise price the longer it waited.

Loews also began lining up the members of the GPGP Board to make the determination that the Opinion was acceptable. In the leadup to a meeting of the GPGP

87

Board on April 26, 2018, Siegel contacted each director. See Siegel Dep. 232–34; Alpert Dep. 172. Siegel reported that Loews had “retained Baker Botts to determine whether it can give the opinion.” JX 1069 at 2; see also Alpert Dep. 90. He also explained that although Holdings would determine whether to exercise the Call Right, “the [GPGP] Board would be required to make a narrow determination as to whether the opinion is an acceptable opinion.” JX 1069 at 2. The outside directors had a “hostile reaction” and asked “shouldn’t we have independent counsel[?]” JX 874 at 5; see Layne Dep. 160.

Alpert and Siegel had approached the GPGP Board based on Skadden’s advice in the hope of eliminating any litigation risk posed by the uncertainty over which decision- maker would make the acceptability determination. With the solution creating additional problems, Loews reversed course. See JX 874 at 5 (“→ Alpert’s view – getting board involved was to take an issue off the table = probably not going to the directors, & L[oews] will exercise”).

Around April 27, 2018, Alpert asked Richards Layton to “take a fresh look” at whether the GPGP Board’s involvement was necessary. Alpert Dep. 224; JX 1340 at 5. Alpert did not tell Richards Layton about Skadden’s prior advice or the GPGP Board’s reaction. Raju Tr. 809, 843. The question of who would determine the acceptability of the Opinion would play out over the ensuing days. Q. Boardwalk And Loews Issue The Potential Exercise Disclosures.

On April 30, 2018, Boardwalk and Loews each filed their Form 10-Qs. PTO ¶ 222.

As discussed in the prior section, Boardwalk and Loews coordinated their filings in advance to ensure that the disclosures were consistent. See id.

88

After the push-and-pull of the prior month, Boardwalk’s Form 10-Q contained disclosures regarding the March 15 FERC Actions that were largely consistent with the initial press release Boardwalk had issued on March 18, 2018. The Form 10-Q stated:

While we are continuing to review FERC’s Revised Policy Statement, [Notice of Inquiry,] and NOPR, based on a preliminary assessment, we do not expect them to have a material impact on our revenues in the near term.

All of the firm contracts on Gulf Crossing and the majority of contracts on Texas Gas Transmission, LLC are negotiated or discounted rate agreements, which are not ordinarily affected by FERC’s policy revisions. Gulf South currently has a rate moratorium in place with its customers until 2023, which we believe will be unaffected by these actions.

JX 1201 at 40. The only addition was the reference to the absence of any material effect on revenue “in the near term.” Boardwalk’s initial press release had not limited the absence of a material impact to the near term, and the record does not suggest any additional analysis that would have shortened the time horizon of any effect. In reality, Boardwalk did not anticipate any material impact on revenue for the foreseeable future.

Despite this reassuring language, the Form 10-Q went on to disclose that in light of FERC’s actions, Boardwalk’s General Partner was evaluating the potential exercise of the Call Right (the “Potential Exercise Disclosures”). See JX 1201 at 40–42, 48. The Form 10- Q stated flatly: “[O]ur general partner has a call right that may become exercisable because of recent FERC action. Any such transaction or exercise may require you to dispose of your common units at an undesirable time or price, and may be taxable to you.” Id. at 48. Continuing, the Form 10-Q explained:

[A]s has been described in our SEC filings since our initial public offering, our general partner has the right under our partnership agreement to call and purchase all of our common units if (i) it and its affiliates own more than 50% in the aggregate of our outstanding common units and (ii) it receives an

89

opinion of legal counsel to the effect that our being a pass-through entity for tax purposes has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers by our subsidiaries that are regulated interstate natural gas pipelines. Because our general partner and its affiliates hold more than 50% of our outstanding common units, this call right would become exercisable if our general partner receives the specified opinion of legal counsel.

The magnitude of the effect of the FERC’s Revised Policy Statement may result in our general partner being able to exercise this call right. Any exercise by our general partner of its call right is permitted to be made in our general partner’s individual, rather than representative, capacity; meaning that under the terms of our partnership agreement our general partner is entitled to exercise such right free of any fiduciary duty or obligation to any limited partner and it is not required to act in good faith or pursuant to any other standard imposed by our partnership agreement. Any decision by our general partner to exercise such call right will be made by [Holdings], the sole member of [GPGP], rather than by our Board. . . . We have been informed by [Holdings] that it is analyzing the FERC’s recent actions and seriously considering its purchase right under our partnership agreement in connection therewith.

Id. at 48 (emphasis added).

In its Form 10-Q, Loews made similar disclosures. PTO ¶ 223. In addition to issuing its Form 10-Q, Loews amended its previously filed Schedule 13-D to state as follows:

In light of the FERC announcement, the General Partner is analyzing the FERC’s recent actions and seriously considering its purchase right under the Limited Partnership Agreement in connection therewith. The exercise of the purchase right would be subject to the approval of the Board of Directors of Loews. There is no assurance that the Loews Board will authorize the purchase or that the pre-conditions to the exercise of the purchase right under the Limited Partnership Agreement will be satisfied, and even if such pre-

conditions are met, there is no assurance that there will be a determination by the General Partner to exercise the purchase right discussed herein or the timing thereof.

Id. ¶ 224.

90

Later on April 30, 2018, Boardwalk held an earnings call. Id. ¶ 225. During the call, Horton explained the formula for calculating the exercise price for the Call Right. Id. ¶ 226. He noted that the decision on the Call Right was for Loews to make and stated that “given where we are in this process, we need to rely on the disclosures and the relevant SEC filings and are unable to answer questions concerning the decision-making process or the possible timing of any such decision.” Id. ¶ 227.

Loews made similar statements during its earnings call later that day. Jim Tisch informed investors that the FERC actions “may result in Loews being able to exercise a call right under the terms of the Boardwalk partnership agreement.” Id. ¶ 228. He added:

We at Loews are exploring all our options regarding these developments.

Although we expect to be able to make a decision sometime this year, no decisions have yet been made. As you can imagine, we’ll have to let our documents speak for themselves since we are constrained from answering any questions on this topic.

Id.

The initial market reaction to the announcements tracked the bump that Barclays anticipated. Boardwalk’s units had closed at price of $11.04 per unit on April 27, 2018, the last trading day before the Potential Exercise Disclosures. Id. ¶ 277. On the day of the disclosures, Boardwalk’s units traded up to a high of $12.70, before trading down to close at $11.37. Id.; JX 1774 at 2. Internally, Barclays bankers observed that the units were “up ~8.3% right now - firmly within the 7–10% estimate to which we guided.” JX 1174 at 1.

After the initial market reaction, however, the implications of the Call Right began to sink in. On May 1, 2018, U.S. Capital Advisors downgraded Boardwalk “to Hold from Buy” and reduced its price target from $20 to $11. JX 1222 at 1. The report explained that

91

any purchase by Loews “would be at a formula-derived price, which, if a deal were consummated, would likely result in limited upside on the price of BWP units.” Id. at 2. McMahon was impressed by the analysis: “[a]mazing how good they are.” Id. at 1. Alpert circulated the note to Loews management. JX 1232.

Subjected to the overhang of the pricing formula, Boardwalk’s trading price declined steadily. The units closed at $10.94 on May 1, then at $10.88 on May 2. On May 3, the price fell to $10.01. On May 4, it fell to $9.56. On May 7, the units closed at $9.26. PTO ¶ 277.

Fund managers and traders working for Bandera Partners LLC, one of the plaintiffs, initially viewed the price trend as a buying opportunity. On May 8, 2018, a fund manager emailed a colleague that “we should buy heavily at this price.” Id. ¶ 278. On May 9, the colleague reported that “we bought with both hands today . . . [and] we will likely get more stock tomorrow.” Id.

But as investors began to understand the effect of the Call Right, they became outraged. TAM Capital Management published an open letter criticizing Loews. See JX 1915. After seeing that letter, the Bandera representatives began drilling down into the mechanics of the Call Right. See PTO ¶ 279.

On May 6, 2018, Deutsche Bank explained the “Prisoner’s Dilemma” that Loews had created. JX 1270 at 2.

Stakeholders could expect no higher price for shares of BWP than $11.50 unless Loews chose voluntarily to tender at a higher share price (or chose not exercise at all). Given that the probable “best” the stakeholders could do seemed to be around $11.50 in August 2017, there seemed to be little incentive to hold onto BWP shares above that price. And so the stock has

92

begun to fall. However, as the stock falls, so too does the 180-average price for which Loews can demand tender. This has engendered a real-time game theory practice known as “the prisoner’s dilemma.” By this logic, the stakeholders assume the worst of their fellow stakeholders and aim to sell first in order to arguably . . . get a better price than those who wait. This has created a pile-on where stakeholders are willing to part with their shares below what some might argue is fair value. And no shareholder has the incentive to pay more than this price if Loews has the option to tender below that price level.

Id.

On May 10, 2018, Barclays issued a research report that expressed concern about the potential exercise of the Call Right. The report noted that

[w]hile the FERC actions could change the max rates the pipelines could charge, we note that Gulf Crossing is 100% negotiated rates while cost of service only makes up ~25% on Gulf South and Texas Gas, making it a bit difficult to see how [Boardwalk’s] cash flows would be materially impacted later on as the FERC changes primarily impact cost of service contracts.

PTO ¶ 284. Barclays suggested that Loews appeared was using a “loophole” in the Partnership Agreement “to buy in the assets for what we believe is an extremely attractive price.” Id. The report explained that

the more appropriate thing for Loews to have done, if they were going to indeed buy in [Boardwalk], was to get the legal opinion and then just announce it would be buying in the MLP rather than just tease the market that they were “seriously considering” it, putting pressure on the stock and in essence, trying to time the potential purchase at a time that would be most favorable to them.

Id.

Loews’ management did not like the report, particularly since Loews had used Barclays for advice on the Call Right. Siegel contacted Gary Posternack, Head of Global M&A at Barclays, to express his “dissatisfaction” with the report. Id. ¶ 285. Posternack

93

emailed Jes Staley, Barclay’s then-CEO, with a heads up that he might be “getting a call in the next day from Jim Tisch at Loews, who is very upset about some equity research commentary that our analyst put out. I should brief you before you speak.” Id. The call ultimately did not take place.

On May 15, 2018, with Boardwalk’s trading price continuing to fall, JP Morgan issued an analyst report that described it as “fundamentally undervalued at this juncture.” Id. ¶ 288. JP Morgan expressed the view that Loews should exercise the Call Right “at least at the ~$13/unit 180 trading VWAP leading up to the April 30 announcement should the company seek to avoid the perception of securities manipulation.” Id. ¶ 289. JP Morgan subsequently issued a clarification stating that its report included “certain wording [which] could have inadvertently been construed as implying a legal conclusion.” Id. ¶ 290.

By May 21, 2018, Bandera’s views about Loews’ actions had changed. Bandera issued a public letter addressed to Loews asserting that its actions had caused a “catastrophic collapse in the market price of Boardwalk’s units” and that the “[t]he units’ 180 consecutive trading day average, which sets the purchase price, is considerably lower than it would have been without this announcement.” Id. ¶ 291. Bandera also cited Boardwalk’s decision in 2014 to cut its distributions and the implications for the unit price:

We believe that you, as stewards of Boardwalk’s capital, made a tough but wise decision to slash the partnership’s cash distribution, and invest substantial funds into the existing base of assets. While these strategic actions depressed unit prices, they were implemented to drive meaningful long-term returns for investors. We estimate that Boardwalk has raised over $3 billion from its limited partners to execute this long-term strategy. The benefits of these investments should accrue to all of the partnership’s investors, not just Loews. This is why we believe the best outcome for unitholders would be for Loews to pass on its purchase right altogether. If Loews does exercise its

94

option, we think that, at a minimum, it must do so only at a fair price and in accordance with straightforward procedures that accord with unitholders’

reasonable expectations of fairness.

Id. ¶ 293. R. Loews Ties Off The Acceptability Issue.

While the Potential Exercise Disclosures were having their effect on the market, Loews was tying off the loose ends created when the outside members of the GPGP Board had a hostile reaction to determining whether the Opinion satisfied the Acceptability Condition. In response to Alpert’s appeal for expedited advice, Richards Layton had advised orally that it felt the “far better view” was that Holdings had the authority to make both the acceptability determination and the exercise decision. Raju Tr. 809, 842. Richards Layton “did not know Skadden had been asked to analyze this issue until after [Richards Layton] had given [its] oral advice to Loews.” Id. at 809. It was only after Richards Layton provided this advice that Alpert sent over Skadden’s analysis. See JX 1197. He then asked Richards Layton to speak with Skadden to see “if they can get on the same page.” Alpert Dep. 224. When the firms connected on May 1, Skadden’s “main point” was that “there is ambiguity and ambiguity is construed against the General Partner.” JX 1228 at 1.

On May 1, 2018, Richards Layton sent Alpert an email memorializing their advice.

JX 1225. The email stated that “[w]hile there is some ambiguity and arguments can certainly be made to the contrary, we think that the better view is that the [acceptability determination] is within the sole authority of the Sole Member [Holdings] pursuant to Section 5.6 of the LLC Agreement.” Id. at 2–3 (emphasis added). The email included the following caveat:

95

[I]f the Board of Directors is approached and declines to determine that the Opinion of Counsel is acceptable and the Section 15.1(b) call right is exercised by the Sole Member anyway, that would be a difficult fact to overcome in any future litigation regarding the exercise of the Section 15.1(b) call right.

Id. at 3 (emphasis added). It was not until discovery in this litigation that Richards Layton learned that Loews had approached the GPGP Board and that the outside directors had reacted negatively to making the determination. See Raju Tr. 843.

Less than two hours after receiving the email, Alpert drafted and circulated new talking points for Siegel to deliver to the GPGP directors. JX 1213 at 1. Alpert’s talking points represented that “[w]e and outside counsel agree that the documents provide that [Holdings’] authority to exercise the call right includes the ability to determine that the opinion of counsel is acceptable.” Id. at 2 (emphasis added). That description did not match Skadden’s view, so when the Skadden lawyers saw the talking points, they struck the “[w]e and outside counsel agree” and substituted “[w]e believe the better reading . . . is.” See JX 1863; 1864 at 1. At first, Alpert accepted the change. See JX 1852; 1853. But five minutes later, he reintroduced the reference to “outside counsel” and added: “—as we are confident that the sole member has the ability and authority to make the determination of an acceptable opinion.” See JX 1850; 1851 at 1. Alpert sent the revisions to Baker Botts and Richards Layton but not to Skadden. See JX 1850.

That evening, Alpert, McMahon, Rosenwasser, Layne, and Richards Layton had a call “to get on the same page” about the acceptability determination. See McMahon Tr. 576–77; JX 1237 at 1. Alpert told Richards Layton that its email was too “measured” and did not reflect the strength of their oral advice. Alpert Dep. 214. After the call, Richards

96

Layton sent Alpert a revised email saying that it was the “far better view” that Holdings could make the acceptability determination, but otherwise maintained its comments about ambiguity. See JX 1265 at 4.

Siegel then held follow-up calls with the members of the GPGP Board and told them that their involvement was not required after all. PTO ¶ 323; Siegel Dep. 235–36. The reversal of position worried the outside directors, who requested “a board call to discuss the partnership agreement and [their] obligations under that agreement.” PTO ¶ 325; JX 1319 at 1.

On May 14, 2018, the GPGP Board met telephonically. JX 1318; see also JX 1435 at 1. Instead of having Skadden or Richards Layton lead the discussion, Alpert tapped Layne of Vinson & Elkins. Alpert knew Layne “was of the firm view . . . even stronger than Rosenwasser, that the proper entity was the sole member, [Holdings].” Alpert Tr. 394. Unlike Skadden and Richards Layton, Layne never prepared a written analysis of the acceptability issue, and contemporaneous documents suggest that when presenting to the Board, he lumped together the question of authority to exercise with the determination of acceptability.14

14 Documents created on or around May 14 suggest that Layne only discussed who had the authority to exercise the Call Right and did not separately address the question of acceptability. See JX 1325 at 1; JX 1331 at 2; JX 1343 at 1; JX 1812 at 1. Two weeks later, Layne, McMahon, and trial lawyers at Vinson & Elkins and Foley & Lardner LLP signed off on minutes which only documented Layne addressing the Call Right’s exercise. JX 1435 at 1, 3. It was not until May 31, 2018, that McMahon revised the minutes to add a reference to the question of acceptability. JX 1444 at 1, 3. In pertinent part, McMahon revised the minutes to read, “Layne stated that if the 15.1(b) right is exercised, the outside directors would not approve that decision or the appropriateness of the Opinion of

97

In any event, the GPGP’s outside directors were “pleased that they did not have to be part of this very awkward process.” Siegel Tr. 739. With the GPGP Board out of the picture and Baker Botts and Skadden prepared to deliver their opinions when asked, Loews was ready to exercise the Call Right. S. The ADIT Issue Gets Worse.

With Loews preparing to exercise the Call Right, the uncertainty regarding the known unknown of ADIT grew worse. On May 14, 2018, SFPP submitted a compliance filing in response to the Order on Remand that FERC had issued in response to the United Airlines decision. The Order on Remand had directed SFPP to revise its filings in accordance with the Revised Policy. SFPP not only removed the income tax allowance, but also eliminated ADIT. See JX 1330 at 185–86. If SFPP had treated ADIT correctly, then the result would be a boon for Boardwalk, but fatal to the Opinion.

Loews, Boardwalk, and their advisors immediately focused on this development.

See Johnson Tr. 684. Baker Botts was particularly attuned to the news, because Wagner was representing BP West Coast Products LLC and ExxonMobil Oil Corporation in the proceedings involving SFPP and filed a submission on their behalf that opposed SFPP’s filing. See Wagner Tr. 304–05; Wagner Dep. 387–89; JX 1465 at 37.

Counsel.” Id. at 3 (emphasis added). In his cover email, McMahon explained that “some changes” were “suggested by certain of the outside directors,” and requested that Layne and the litigators call him if they “ha[d] any questions about them.” JX 1444 at 1. Layne testified that he told the GPGP Board that, “under the LLC [A]greement, the board of directors did not have authority with respect to exercise of the call or acceptability of the opinion.” Layne Dep. 216 (emphasis added).

98

The ADIT NOI process was also unfolding. Between May 21 and June 20, sixty industry participants filed comments, reply comments, or both in FERC’s ongoing NOI proceeding. Court Report ¶ 74. The vast majority of comments from shippers and organizations aligned with their interests took the position that ADIT balances should be refunded or amortized on an accelerated basis. The vast majority of comments from pipelines and organizations aligned with their interests took the position that ADIT balances should be eliminated. See Webb Rebuttal ¶ 40 n.46 & Ex. 25; JX 1549 ¶ 9.

Van Ness Feldman and Vinson & Elkins, two of Boardwalk’s go-to law firms, argued in favor of eliminating the ADIT balances on behalf of multiple pipeline clients. See, e.g., JX 1382 at 2, 18–23; JX 1460 at 4, 6–7, 18. They did not make that argument on behalf of Boardwalk, even though its subsidiaries had accumulated ADIT balances totaling at least $750 million. See JX 644 at 1. The reality was that Boardwalk could not advocate publicly to eliminate its ADIT balance without undercutting the Rate Model Analysis and the assumptions driving the Opinion. Instead, Boardwalk publicly advocated for a middle ground—either the Reverse South Georgia Method or the Average Rate Assumption Method. JX 1388 at 11 (NOI Comments).

Privately, however, Boardwalk wanted FERC to eliminate ADIT. See JX 797 at 1 (Boardwalk not wanting to “give up [the] argument” that “the Policy Statement essenti8ally [sic] eliminates ADIT”). To advance that position, Boardwalk management lobbied FERC through the Interstate Natural Gas Association of America (“INGAA”). See JX 1457 (INGAA NOI Comments) at 7; Horton Dep. 183. Boardwalk has been a member of INGAA for almost three decades. McMahon Tr. 565–66. McMahon, Boardwalk’s general

99

counsel, is the Chairman of INGAA’s Legal and Rates Committee, serves on INGAA’s Board of Directors, and served as the Chair of INGAA’s Board of Directors in 2016. See McMahon Dep. 23–24. Johnson also serves on INGAA’s Legal and Rates Committee, along with two other Boardwalk executives. See Johnson Dep. 78; McMahon Tr. 566–67. Attorneys at Van Ness Feldman reviewed the comments, and the defendants’ privilege log reveals Boardwalk executives and Van Ness Feldman were heavily involved. McMahon Tr. 572–73; McMahon Dep. 29–30; JX 1881 (Privilege Log) at Rows 3527–44, 3565–76, 3580–83.15 McMahon and Johnson also met with FERC staff on June 12, 2018, “as part of a group from [INGAA]” to discuss “Taxes and ADIT.” JX 1680 at 45; JX 1464 at 1. McMahon and Johnson had helped INGAA prepare a supporting presentation, but FERC ended up prohibiting the discussion of ADIT at the last minute. See JX 1463 at 1; JX 1471 at 2. The presentation nevertheless made a compelling case that ADIT represented “a cost- free form of financial capital,” was “not a loan from customers but from the federal government,” and should be handled in accordance with IRS normalization rules, all of which were premises for the argument that ADIT balances should be eliminated. See JX 1476 at 6–8.

15 Despite this evidence, at their depositions and at trial, McMahon and Johnson attempted to distance themselves from INGAA’s comments. See McMahon Dep. 30–31; Johnson Dep. 183–84.

100

Even though it was obvious that ADIT was an unsettled issue, and even though everyone knew that different outcomes for ADIT were possible, Baker Botts did not update its analysis. No one prepared sensitivity analyses for different outcomes regarding ADIT. The Preliminary Opinion provided the answer Loews wanted, and developments in the real world were not going to change that. T. This Litigation And The Original Settlement On May 24, 2018, two holders of common units (the “Original Plaintiffs”) filed this action and moved for expedited proceedings. The Original Plaintiffs wanted to prevent the General Partner from exercising the Call Right using a 180-day measurement window that included trading days that had been affected by the Potential Exercise Disclosures. The defendants opposed the motion, arguing that the dispute was not ripe because the General Partner had not yet elected to exercise the Call Right.

Five days after the action was filed, the court held a hearing on the motion to expedite. The court agreed with the defendants and denied the motion.

Having defeated the motion to expedite on the theory that the claims were not yet ripe, defense counsel contacted the lawyers for the Original Plaintiffs the very next day to explore settling the non-justiciable claims. A settlement in this litigation would give the defendants the ultimate protection: a global release of claims relating to the exercise of the Call Right.

The lawyers for the Original Plaintiffs understood that Loews wanted to exercise the Call Right. They offered up a settlement, including a global release, if Loews did what it wanted to do. As part of the negotiations with the defendants, lead counsel made precisely

101

that argument, telling defense counsel, “Your clients want to make this purchase. Getting a release on a deal they want to make anyway is actually an amazing outcome for them.” Dkt. 56 Ex. 1.

The Original Plaintiffs initially proposed settling if the General Partner agreed to exercise the Call Right using June 1, 2018, as the end date for the 180-day measurement period, which would have included twenty-four trading days after the issuance of the Potential Exercise Disclosures in the calculation of the Purchase Price. The defendants countered with an end date of September 1, 2018, which would have included sixty-four trading days after the issuance of the Potential Exercise Disclosures in the calculation.

On June 11, 2018, eighteen days after the lawsuit was filed, the parties agreed that Loews would exercise the Call Right on or before June 29, 2018. The resulting period included forty-four affected days in the pricing formula. Using that end date, the formula yielded a Purchase Price of $12.06 per unit.

On June 22, 2018, the parties informed the court by email that they had reached an agreement in principle and asked the court to review the settlement papers in camera. JX 1487. The court rejected that request as seeking a non-public advisory opinion. Dkt. 26.

That night, the parties filed a stipulation of settlement. JX 1496 (the “Original Settlement”). Under its terms, the defendants would receive a global release as long as the General Partner exercised the Call Right on or before June 29, 2018—the day that Barclays had projected Loews might exercise. Id. at 15–16; JX 915 at 15. That date was optimal for the defendants because it ensured that purchases under the Call Right would close before FERC’s regularly scheduled meeting on July 19, when FERC was expected to make

102

additional announcements regarding the subject matter of the March 15 FERC Actions. See PA §15.1(c) (governing timing of exercise, notice and purchase date); JX 793 at 1 (“[T]he Commission indicated its desire to issue an order on the [NOPR] in its July meeting which will take place on July 19.”). The Original Settlement contemplated a fee award for plaintiffs’ counsel “in an amount not to exceed $1.8 million.” JX 1496 at 20. U. Baker Botts Renders The Opinion.

Believing that they had secured a settlement that would extinguish and release any challenges to the exercise of the Call Right, Loews asked its advisors to finalize their work product. See JX 1489 at 1 (Richards Layton email thread reporting that “Loews is likely to settle its litigation this evening and is likely to exercise the purchase right on Friday”).

On June 29, 2018, Baker Botts delivered the Opinion. JX 1522. It was substantially unchanged from the Preliminary Opinion that Baker Botts had provided on April 29.

The Opinion resembled a closing opinion in that it expressed a conclusion, without supporting reasoning or citations to legal authority. The Opinion did not reference a single case or statute, much less provide any discussion or application. The Opinion thus proceeded as if Baker Botts were opining on a routine issue, such as the due formation of an entity, its good standing, or its authority to enter into an agreement.

As is customary in a closing opinion, the Opinion began by listing the materials that Baker Botts had consulted. The Opinion next provided its conclusion, consisting of the following statement:

On the basis of the foregoing, and subject to the assumptions, limitations, and qualifications set forth herein, we are of the opinion that the status of the Partnership as an association not taxable as a corporation and not otherwise

103

subject to an entity-level tax for federal, state or local income tax purposes has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers by subsidiaries of the Partnership that are regulated interstate natural gas pipelines (the “Subsidiaries”).

Id. at 2.

The Opinion then provided two paragraphs summarizing the Financial Data. Those paragraphs stated:

In rendering this opinion, we have requested and received from the Partnership cost, rate and other financial information, including projections, estimates and pro forma information (“Financial Data”) relating to the Partnership and the Subsidiaries, which we have relied upon. We have been assisted in our review of the Financial Data by a consultant engaged by us with expertise in the calculation of the cost of service of regulated interstate natural gas pipelines. The Financial Data includes a calculation of the estimated cost of service of each of the Subsidiaries under two scenarios. In preparing Financial Data pertaining to both scenarios, the Partnership made several assumptions, including that each Subsidiary would charge all its customers the maximum applicable rate, and as a result, each Subsidiary would recover its entire cost of service. The first scenario included in the cost of service of each Subsidiary an income tax allowance derived from the current federal, state and local income tax rates. The second scenario excluded an income tax allowance from the cost of service of each Subsidiary. We have participated in conferences with officers and other representatives of the Partnership, [the General Partner] and [the GPGP] in which the Financial Data, as well as other matters, were discussed. The purpose of our engagement, however, was not to establish or confirm the accuracy of factual matters or the reasonableness of projections, estimates or pro forma information provided to us or reviewed by us. Therefore, we have assumed that the Financial Data is correct in all material respects, that all calculations were performed accurately in all material respects and that the Financial Data was prepared in a reasonable manner and in good faith.

With regard to the Financial Data, in rendering our opinion referred to above, we relied substantially on the fact that the Financial Data indicated that the removal of the income tax allowance derived from the current federal, state and local income tax rates from the cost of service of the Subsidiaries would result, in the case of each Subsidiary, in an estimated reduction in excess of ten percent in the maximum applicable rates that can be charged to the

104

customers of each of the Subsidiaries on a long-term basis. The Financial Data included a “Rate Model Analysis for 2017,” which compared an estimate of (a) the maximum applicable rate that each Subsidiary could charge its customers, based on the development of a system wide rate for each Subsidiary and assuming each Subsidiary could include an income tax allowance derived from the current federal, state and local income tax rates in its cost of service with (b) the maximum applicable rate that each Subsidiary could charge its customers, based on the development of a system wide rate for each Subsidiary and assuming that each Subsidiary could not include any income tax allowance in its cost of service. The Rate Model Analysis indicates that elimination of an income tax allowance from the cost of service would result in an estimated 12.12% decline in the maximum applicable rate for Texas Gas Transmission, LLC, an estimated 11.68% decline in the maximum applicable rate for Gulf South Pipeline Company, LP, and an estimated 15.62% decline in the maximum applicable rate for Gulf Crossing Pipeline Company LLC. We also took notice that, because these reductions in the maximum applicable rates would not be offset by any reduction in costs incurred by the Subsidiaries, the reductions in the maximum applicable rates would have a substantially larger percentage impact on the earnings before interest and taxes and on the cash available for distribution of each of the Subsidiaries assuming each Subsidiary could actually charge and collect its maximum applicable rate.

Id. at 3.

The remainder of the Opinion consisted of a series of assumptions. Id. at 3–5. They included the following:

• “[T]he Revised Policy will not be revised, reversed, overturned, vacated, modified or abrogated in any relevant manner by any court or administrative or executive body, including the FERC, or by an act of Congress;” and

• “[T]he Revised Policy will be applied to individual FERC regulatory proceedings involving the Subsidiaries in accordance with its terms . . . .”

Id. at 4.

This section of the Opinion also included descriptions of how Baker Botts interpreted the terms “maximum applicable rate” and “material adverse effect.” On the issue of “maximum applicable rate,” the Opinion stated:

105

Based on the wording of Section 15.l(b)(ii) of the Partnership Agreement, other provisions of the Partnership Agreement and support in the Registration Statement (particularly the final prospectus included therein), in rendering the opinion set forth above, we have, in using our judgment, interpreted the words (a) “maximum applicable rate that can be charged to customers by subsidiaries that are regulated interstate natural gas pipelines of the Partnership,” to mean the recourse rates of the Subsidiaries now and in the future as that term is used by the FERC in its regulations, rulings and decisions, and (b) “status as an association not taxable as a corporation,” to mean status as an entity not taxable as a corporation.

Id. at 4.

On the issue of “material adverse effect,” the Opinion stated:

The term “material adverse effect” as used in Section 15.1(b)(ii) of the Partnership Agreement is not defined in the Partnership Agreement or in the Delaware Revised Uniform Limited Partnership Act. In rendering the opinion set forth above, we have considered Delaware case law construing such term. Our analysis leads us to the conclusion that there is no case directly applicable to this situation and no bright-line test regarding what is a “material adverse effect,” although the case law has provided us some guidance.

Id. at 4.

Baker Botts limited its Opinion to “applicable federal law of the United States, the Delaware Revised Uniform Limited Partnership Act, the Delaware Limited Liability Company Act, and, only to the extent relevant, in our judgment, to the opinion set forth above, Delaware law as it applies to the interpretation of contracts.” Id. at 5. A non- Delaware law firm thus rendered a non-explained opinion on the existence of a material adverse effect, a subject on which both a Delaware law firm (Richards Layton) and a national law firm with a Delaware office (Skadden) would not opine.

On the same day that Baker Botts rendered the Opinion, the firm’s rate expert—

Sullivan—testified in a proceeding before FERC that it was impossible to assess the effects

106

of changing the income tax allowance without a determination on the treatment of ADIT. Webb. Tr. 949. V. The General Partner Exercises The Call Right.

After receiving the Opinion, Loews management recommended that Loews cause the General Partner to exercise the Call Right. See JX 1515 at 2; JX 1523 at 2–3. In their “Updated Base Case,” management estimated that the transaction would generate more than $1.5 billion in “Value Creation” for Loews. JX 1515 at 9. After discussion, the Loews board of directors adopted resolutions authorizing Holdings to exercise the Call Right on behalf of the General Partner. JX 1523 at 4.

The Holdings Board met afterwards. See JX 1509. Skadden made a presentation concluding that “it would be within the reasonable judgment of [Holdings] to find that” the Opinion was acceptable. JX 1518 at 23. Comprised of three Loews insiders, the Holdings Board approved resolutions deeming the Opinion acceptable and exercising the Call Right. See JX 1509 at 5–9.

Later that day, Boardwalk announced that the General Partner had elected to purchase all outstanding units at a price of $12.06 per common unit, for approximately $1.5 billion in total consideration. JX 1526 at 1. Ten days later, on July 18, 2018, the transaction closed on schedule. See JX 1547 at 2. W. FERC Makes Its Determinations.

Hours after the closing, FERC issued an order on rehearing of the Revised Policy and a final rule in response to the NOPR. JX 1549 (the “Order on Rehearing”); JX 1546 (the “Final Rule”). In the Order on Rehearing, FERC reiterated that its policy would not

107

automatically permit MLP pipelines to recover an income tax allowance in their cost of service, but MLPs would not be precluded from arguing in a rate case that they were entitled to an income tax allowance based on an evidentiary record. JX 1549 ¶ 8.

Critically, FERC stated that MLPs that were no longer entitled to an income tax allowance could eliminate their overfunded ADIT balances without returning the balances to rate payers (whether by refund or amortization). See id. ¶ 10. The Commission based its ADIT decision on the arguments raised by pipeline-side commenters in the NOI docket and INGAA’s presentation to FERC staff. Id. ¶ 13.

In its Final Rule, FERC adopted the procedures proposed in the NOPR with certain modifications, required all interstate natural gas pipelines to file a Form 501-G, provided options for each pipeline to address the recovery of tax costs (including filing a statement explaining why an adjustment to rates was not needed), and reiterated that a rate reduction might not be justified for a significant number of pipelines for several reasons. See JX 1546. Consistent with the Order on Rehearing, FERC “modifie[d] the proposed Form 501- G so that, if a pass-through entity state[d] that it d[id] not pay taxes, the form w[ould] not only eliminate its income tax allowance but also eliminate ADIT.” Id. ¶ 132 (emphasis added). The Commission reasoned that doing otherwise would violate the prohibition against retroactive ratemaking. Id. ¶¶ 133–34. The DC Circuit ultimately agreed that returning ADIT to shippers would violate the prohibition against retroactive ratemaking. SFPP, 967 F.3d at 801 (dismissing shippers’ contrary arguments as “non-starters”).

The Final Rule meant there would be no effect on Boardwalk’s recourse rates. When one of his colleagues who had worked on the Opinion commented that the news “sounds

108

pretty good for MLPs,” Rosenwasser responded: “Seems all mitigates adverse effect without changing policy. Loews buy in of [B]oardwalk closed day before order came out.” JX 1569 at 1.

Johnson circulated the news within Boardwalk. One of the executives responded, “Maybe I wish we were still publically [sic] traded….. [sic].” JX 1532 at 1.

On August 3, 2018, Wagner sent McMahon a summary of FERC’s actions, copying Rosenwasser and Alpert. Confirming that the March 15 FERC Actions had opened the door to changes in ADIT, he explained that “FERC’s March 2018 Revised Policy Statement created an issue of first impression by prohibiting MLP-owned pipelines from collecting a tax allowance, which raised the issue of how to treat the ADIT.” JX 1578 at 1. He also confirmed the effect of the Order on Rehearing: “FERC announced that MLP-owned pipelines may reduce the balance to zero without providing any refunds or rate reductions. This has the net effect of reducing the pipeline’s exposure to rate reductions.” Id. (emphasis added). X. The Current Plaintiffs Pursue The Litigation.

The current plaintiffs objected to the Original Settlement. On September 28, 2018, the court declined to approve the Original Settlement. Because the current plaintiffs had prevailed on their objections, the court permitted them to take over the litigation.

The court subsequently certified a plaintiffs’ class consisting of:

Any natural person or entity who held Boardwalk limited partnership units on July 18, 2018 and whose units were purchased on that date by Boardwalk GP, LP, together with their heirs, assigns, transferees, and successors in interest, but excluding Defendants, their successors in interest and assigns,

109

and any natural person or entity that is a director, officer or affiliate of any of the foregoing

Dkt. 194 ¶ 1. The case proceeded through discovery and to trial.

II. LEGAL ANALYSIS

The plaintiffs proved that the General Partner breached the Partnership Agreement by exercising the Call Right without first satisfying the Opinion Condition or the Acceptability Condition. By acting manipulatively and opportunistically, the General Partner engaged in willful misconduct when it exercised the Call Right, and the exculpatory provisions in the Partnership Agreement therefore do not protect the General Partner from liability. This decision does not reach the plaintiffs’ other claims. A. Governing Principles Of Contract Law The plaintiffs’ principal claim asserts that the General Partner breached the Partnership Agreement, which is a contract governed by Delaware law. Delaware law therefore governs the claim for breach of the Partnership Agreement.

Under Delaware law, the elements of a breach of contract claim are (i) a contractual obligation, (ii) a breach of that obligation by the defendant; and (iii) causally related harm to the plaintiffs. WaveDivision Hldgs. v. Millennium Digit. Media Sys., L.L.C., 2010 WL 3706624, at *13 (Del. Ch. Sept. 17, 2010). No one disputes the status of the Partnership Agreement as a binding contract. No one disputes that the General Partner exercised the Call Right and acquired the publicly held units, thereby causing the resulting effects on the plaintiffs. The central issue is the question of breach. If the General Partner breached the

110

Partnership Agreement, then the court must determine the quantum of harm, which also logically will serve as the measure of damages.

To determine the scope of a contractual obligation, “the role of a court is to effectuate the parties’ intent.” Lorillard Tobacco Co. v. Am. Legacy Found., 903 A.2d 728, 739 (Del. 2006). “If a writing is plain and clear on its face, i.e., its language conveys an unmistakable meaning, the writing itself is the sole source for gaining an understanding of intent.” City Investing Co. Liquidating Tr. v. Cont’l Cas. Co., 624 A.2d 1191, 1198 (Del. 1993). A writing is plain and clear on its face “[w]hen the plain, common, and ordinary meaning of the words lends itself to only one reasonable interpretation . . . .” Sassano v. CIBC World Mkts. Corp., 948 A.2d 453, 462 (Del. Ch. 2008). When a writing is plain and clear, the court “will give priority to the parties’ intentions as reflected in the four corners of the agreement, construing the agreement as a whole and giving effect to all its provisions.” In re Viking Pump, Inc., 148 A.3d 633, 648 (Del. 2016) (internal quotations omitted).

A writing that is ambiguous is not plain and clear on its face, and the text of the agreement therefore cannot be the exclusive source of contractual meaning. “[A] contract is ambiguous only when the provisions in controversy are reasonably or fairly susceptible of different interpretations or may have two or more different meanings.” Rhone-Poulenc Basic Chems. Co. v. Am. Motorists Ins. Co., 616 A.2d 1192, 1196 (Del. 1992). “A contract is not rendered ambiguous simply because the parties do not agree upon its proper construction.” Id.

111

If the language of a contract is ambiguous, then a court may look beyond the contract itself to determine the parties’ shared intent. Under appropriate circumstances, extrinsic evidence sheds light on “the expectations of contracting parties” and can “reveal[] . . . the way contract terms were articulated by those parties.” SI Mgmt. L.P. v. Wininger, 707 A.2d 37, 43 (Del. 1998). Because its purpose is to elucidate “the expectations of contracting parties,” extrinsic evidence is only relevant when it “can speak to the intent of all parties to a contract.” Id. “Thus, it is proper to consider extrinsic evidence of bilateral negotiations when there is an ambiguous contract that was the product of those negotiations . . . .” Id. It follows that if there have not been “bilateral negotiations,” then “extrinsic evidence is irrelevant to the intent of all parties at the time they entered into the agreement.” Id. at 43– 44.

A partnership agreement for an MLP is not the product of bilateral negotiations; the limited partners do not negotiate the agreement’s terms. Extrinsic evidence therefore cannot speak to the intent of all parties to the agreement. In that setting, Delaware courts apply the doctrine of contra proferentem and “construe ambiguous provisions of the partnership agreement against the general partner.” Martin I. Lubaroff et al., Lubaroff & Altman on Delaware Limited Partnerships § 14.02[B], at 14-39 (2d ed. 2021 Supp.); see Dieckman v. Regency GP LP, 155 A.3d 358, 366 n.18 (Del. 2017); Norton v. K-Sea Transp. P’rs L.P., 67 A.3d 354, 360 (Del. 2013). In addition to recognizing that extrinsic evidence is unhelpful in that setting, the doctrine of contra proferentem “protects the reasonable expectations of people who join a partnership or other entity after it was formed and must rely on the face of the [entity] agreement to understand their rights and obligations when

112

making the decision to join.” Stockman v. Heartland Indus. P’rs, L.P., 2009 WL 2096213, at *5 (Del. Ch. July 14, 2009). B. The Failure To Satisfy The Opinion Condition Before the General Partner could exercise the Call Right, the General Partner had to satisfy the Opinion Condition. For that condition to be satisfied, the General Partner had to receive “an Opinion of Counsel that the Partnership’s status as an association not taxable as a corporation and not otherwise subject to an entity-level tax for federal, state or local income tax purposes has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers.” PA § 15.1(b)(ii). If the General Partner exercised the Call Right without satisfying the Opinion Condition, then the exercise of the Call Right breached the Partnership Agreement. The General Partner obtained the Opinion, but the plaintiffs proved at trial that the Opinion was not a bona fide “Opinion of Counsel” that could satisfy the Opinion Condition. The General Partner therefore breached the Partnership Agreement.

When parties to a contract agree that the delivery of an opinion of counsel is necessary to satisfy a condition precedent, “it is [counsel]’s subjective good-faith determination that is the condition precedent.” Williams Cos., Inc. v. Energy Transfer Equity, L.P., 2016 WL 3576682, at *11 (Del. Ch. June 24, 2016), aff’d, 159 A.3d 264 (Del. 2017). Counsel renders on opinion in subjective good faith by applying expertise to the facts in an exercise of professional judgment. Id.

Beyond that foundational principle, Delaware decisions have not expounded on what it means for an opinion giver to act in subjective good faith. In a related setting, the

113

Delaware Supreme Court has held that a general partner violated the implied covenant of good faith and fair dealing by relying on an opinion “that did not fulfill its basic function.” Gerber v. Enter. Prod. Hldgs., LLC, 67 A.3d 400, 422 (Del. 2013), overruled on other grounds by Winshall v. Viacom Int’l, Inc., 76 A.3d 808 (Del. 2013). That holding implies that an opinion giver cannot render an opinion in good faith if the opinion giver knows that the opinion does not fulfill its basic function.

Authorities on the rendering of closing opinions confirm related and self-evident propositions about what it means for an opinion giver to render an opinion in good faith.16 For example, an opinion giver plainly must have competence in the particular area of law. See Glazer et al., supra, § 2.7.1 at 61–62. An opinion giver who knowingly lacks competence in the area of law and nevertheless proceeds is not acting in good faith. In that setting, the opinion giver must look elsewhere for the relevant experience, and an opinion giver who lacks the competence to opine on an area of law may rely on an opinion from counsel with competence in that area. See id.; TriBar Report, supra, § 5.1 at 637–39.

16 See, e.g., Restatement (Third) of the Law Governing Lawyers §§ 50, 51, 95, Westlaw (Am. L. Inst. database updated Oct. 2021) [hereinafter Restatement]; Donald W. Glazer et al., Glazer & FitzGibbon on Legal Opinions (2d ed. 2001); Legal Ops. Comm. of the ABA Section of Bus. L., Legal Opinion Principles, 53 Bus. Law. 831 (1998) [hereinafter Opinion Principles]; TriBar Op. Comm., Third-Party “Closing” Opinions: A Report of the Tribar Opinion Committee, 53 Bus. Law. 591 (1998) [hereinafter TriBar Report] see also Legal Ops. Comm. of the ABA Section of Bus. L., Third-Party Legal Opinion Report, Including the Legal Opinion Accord, of the Section of Business Law, American Bar Association, 47 Bus. Law. 167 (1991) [hereinafter ABA Accord]. As stated in the text, this decision regards the principles it articulates as self-evident manifestations of what it means for an opinion giver to act in subjective good faith. This decision cites the authorities as providing illustrative support for those principles.

114

These principles apply equally to the rendering of opinions on matters of Delaware entity law, where it is nevertheless customary for sophisticated law firms to provide third- party closing opinions on routine matters, such as due formation. Glazer et al., supra, § 2.7.1 at 94.

Non-Delaware lawyers, however, normally do not render opinions on more difficult questions of Delaware corporation law or on questions arising under Delaware commercial law. In those circumstances, they usually rely on an opinion of Delaware counsel or deal with the issue in some other way, for example by relying on an express assumption.

Id. § 2.7.3 at 64–65. Although the quoted passage discusses Delaware corporate law, those same principles apply to opinions involving other types of Delaware entities. See id. § 2.7.3 at 65.

It is also self-evident that an opinion giver must act in good faith when establishing the factual basis for an opinion, including when making assumptions. Legal opinions “do not address the law in the abstract. Rather, they apply the law to real companies in real transactions.” Id. § 4.1 at 82. Legal opinions accordingly “require grounding in the facts as well as the law.” Id. The opinion giver usually will have firsthand knowledge of some of the facts necessary to render the opinion, but rarely will the opinion giver have firsthand knowledge of all of the necessary facts.17

17 See Glazer et al., supra, § 4.1, at 83, 85–86; Opinion Principles, supra, § III.A at 833; Tribar Report, supra, §§ 2.1.1 to .1.2 at 608–09.

115

To establish the factual basis for an opinion, the opinion giver can rely in good faith on factual information provided by others.18 An opinion giver cannot act in good faith by relying on information known to be untrue or which has been provided under circumstances that would make reliance unreasonable.19 For example, an opinion giver could not rely in good faith on information if the opinion giver knew that the person providing the information had not done the work required to support it. See Glazer et al., supra, § 4.2.3.6 at 105. An opinion giver also could not rely in good faith on factual representations that effectively establish the legal conclusion being expressed. See Opinion Principles, supra, § III.C at 833. If the factual representations are “tantamount to the legal conclusions being expressed,” then the opinion giver is regurgitating facts, not giving an opinion in good faith. See id.

In lieu of factual representations, an opinion giver may establish the factual predicate for an opinion by making assumptions that certain facts are true. See Glazer et al., supra, §§ 4.1, 4.3.1 at 83, 109; Restatement, supra, § 95 cmt. c. Whether the opinion giver can make an assumption in good faith depends on the nature of the opinion. If an assumption or set of assumptions effectively establishes the legal conclusion being expressed, then the opinion giver cannot properly rely on those assumptions, as doing so vitiates the opinion. See Opinion Principles, supra, §§ III.C–D at 833; ABA Accord, supra,

18 See Glazer et al., supra, § 4.1 at 83; Restatement, supra, § 95 cmt. c.

19 See Glazer et al., supra, §§ 4.1, 4.2.3 at 83, 95–96; Restatement, supra, § 95 cmt.

c.; Opinion Principles, supra, §§ I.F, III.A at 832–33; TriBar Report, supra, § 2.1.4 at 610.

116

¶ 4.6 at 189–90. As with factual representations, if the assumptions establish the legal conclusions being expressed, then the opinion giver is simply making assumptions, not giving an opinion in good faith.

Although an opinion giver cannot rely on factual information known to be untrue, an opinion giver can base an opinion in good faith on an assumption that is contrary to existing fact. The flexibility to rely on a counterfactual assumption enables an opinion giver to render an opinion based on facts that do not exist on the date of the opinion but that the giver and recipient are confident will exist in the future. See Glazer et al., supra, § 4.3.6 at 119. For example, an opinion giver might assume that stock will be duly authorized after the closing of a transaction once necessary filings are made. Id. Or the opinion giver may use counterfactual assumptions to address situations that are not expected to arise, but which the recipient wants the opinion giver to address, such as the possibility that the law of a particular jurisdiction may govern the transaction. Id.

To rely in good faith on a counterfactual assumption, the opinion giver must identify the assumption explicitly. The opinion giver cannot rely in good faith on an unstated factual assumption that is known to be untrue. See Glazer et al., supra, § 4.3.4 at 115; Restatement, supra, § 95 cmt. c; TriBar Report, supra, § 2.3(c) at 616.

In this case, the Opinion Condition limited the ability of the opinion giver to rely on assumptions. To satisfy the Opinion Condition, the opinion giver had to conclude that Boardwalk’s status as a pass-through entity for tax purposes “has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers.” PA § 15.1(b)(ii). The Opinion Condition required an opinion about

117

an actual event (“has . . . a material adverse effect”) or a future event (“will reasonably likely in the future have a material adverse effect”). The opinion giver thus was not being asked to opine on a counterfactual event. To render that Opinion, the opinion giver could make good faith predictions about what would happen in the future, but the opinion giver could not assume what would happen in the future. In particular, the opinion giver could not construct a set of assumptions about the existence of future facts that would generate the conclusion that the Opinion Condition required.

The plaintiffs proved that the Opinion did not reflect a good faith effort to discern the actual facts and apply professional judgment. Instead, Baker Botts made a series of counterfactual assumptions that were designed to generate the conclusion that Baker Botts wanted to reach. Baker Botts then deployed those assumptions as part of a syllogism that turned on elementary subtraction. In the process, Baker Botts stretched its analysis in myriad other ways. The Opinion was a contrived effort to reach the result that the General Partner wanted.

1. The Assumptions In the Opinion, Baker Botts made a series of counterfactual assumptions. One was explicit. The rest were not. Baker Botts did not make those assumptions legitimately because its client asked for a hypothetical opinion about a set of alternative facts. Instead, Baker Botts made those assumptions because Baker Botts knew they were the only way that the firm could purport to reach the outcome that its client wanted. By making those assumptions, Baker Botts did not address whether an event had occurred that “has or will

118

reasonably likely in the future have a material adverse effect.” Baker Botts addressed an imaginary scenario that was never reasonably likely to come to pass.

a. Counterfactual Assumption: The Revised Policy Was Final.

To facilitate the exercise of the Call Right, Baker Botts assumed that the Revised Policy was final such that FERC had “revers[ed] its prior policy of allowing interstate natural gas pipelines owned by publicly traded partnerships . . . to include an income tax allowance in their cost of service.” JX 1522 at 1. Baker Botts also assumed that “the Revised Policy will be applied to individual FERC regulatory proceedings involving the Subsidiaries in accordance with its terms and will not be directly or indirectly revised to allow any of the Subsidiaries to recover an income tax allowance in its cost-of-service rates.” Id. at 4. Those assumptions were contrary to known facts.

An agency’s statement of policy “is not finally determinative of the issues or rights to which it is addressed,” but rather, only “announces the agency’s tentative intentions for the future.” Pac. Gas & Electric Co. v. FPC, 506 F.2d 33, 38 (D.C. Circ. 1974); see Consol. Edison Co. of NY, Inc. v. FERC, 315 F.3d 316, 323 (D.C. Cir. 2003) (“‘Policy statements’ differ from substantive rules that carry the ‘force of law,’ because they lack ‘present binding effect’ on the agency.” (quoting Interstate Nat. Gas Ass’n v. FERC, 285 F.3d 18, 59 (D.C. Cir. 2002))). Because of these attributes, “when [an] agency applies [a general statement of] policy in a particular situation, it must be prepared to support the policy just as if the policy statement had never been issued.” Pac. Gas, 506 F.2d at 38.

Those principles of law applied with greater force to the Revised Policy, which was subject to further regulatory proceedings. Court Tr. 861. FERC stated in the concurrently

119

issued NOPR that it intended to promulgate regulations to address the effects of the Revised Policy “on the rates of interstate natural gas pipelines organized as MLPs.” JX 579 ¶ 8. When announcing the March 15 FERC Actions, Commission personnel responded to a question asking when “FERC Jurisdictional Rates [would] actually change,” by saying that “the NOPR anticipates that the deadlines for pipeline filings will be late summer or early fall [2018]. We obviously have to go to a final rule first.” PTO ¶ 117 (emphasis added). Absent a final rule and the filing of a rate case, jurisdictional rates, i.e. recourse rates, would not change.

Over the next four months, Boardwalk joined other pipelines, shippers, trade associations, and other industry participants in seeking to change the Revised Policy. Collectively, they filed thirteen requests for rehearing, 108 comments, sixteen reply comments, and numerous other submissions in response to the March 15 FERC Actions. See PDX 9 at 12; Court Tr. 858. And while the regulatory process was unfolding, members of Congress were “grill[ing]” the FERC commissioners about whether they were pursuing an appropriate policy. See JX 1076 at 1. The regulatory situation was in flux, and no one could predict where matters would end up. See JX 1525 at 67 (Sullivan testifying that “FERC’s income tax allowance policy for ‘pass through entities’ is still being determined”).

Baker Botts understood that reality, and Wagner explained those facts to Alpert in an email on March 20, 2018. JX 626. Wagner observed that “[s]tanding alone, [the Revised Policy] does not require pipelines to take any action.” Id. at 1. He noted that by issuing the NOPR, FERC had made clear that it would implement the policy through regulations. Id.

120

He added that if the final regulations called for the contemplated Form 501(g) filing, then those filings “may lead to rate challenges,” but that those challenges would not be resolved until 2020 at the earliest. Id. (emphasis added). Alpert, however, pushed Baker Botts to take the position that the March 15 FERC Actions were sufficiently final to render the Opinion. In a call that Alpert convened shortly after receiving Wagner’s email, Rosenwasser told Alpert what Loews wanted to hear. Rosenwasser agreed that the “most important thing has happened” and that “we’re already there.” JX 646 at 5.

Rosenwasser knew that was not true. He knew about and understood Wagner’s analysis. Later, he acknowledged in his backup memorandum that “FERC could choose in its discretion to change the Revised Policy.” JX 1502 at 10. In the April 4 Draft, Baker Botts recognized that “[i]mportant details of implementing the Revised Policy require clarification, and as a result our understanding regarding the implementation of the Revised Policy could prove to be incorrect.” See JX 1949 at 2. That candid language did not appear in the final Opinion.

Boardwalk’s executives did not believe that the Revised Policy was final. In Boardwalk’s comments on the NOPR, they pointed out that the Revised Policy “is not a binding rule.” JX 1139 at 2. They asked FERC to modify the Revised Policy by “eliminat[ing] issues related to the MLP income tax allowance from the proposed rule,” and they asserted that the Revised Policy was “arbitrary and capricious and not the product of reasoned decisionmaking.” Id. They also cautioned that any determination by the Commission to implement the Revised Policy needed to take into account the related issue of ADIT. Id. at 5.

121

Rosenwasser reviewed and marked up Boardwalk’s comments on the NOPR, and he double-starred Boardwalk’s statement that “[u]ntil the Commission provides a final decision on the treatment of ADIT, Boardwalk cannot correctly assess the impact of the Revised Policy Statement and ADIT on its pipelines’ costs of service.” JX 1138 at 14. That was exactly what Baker Botts was purporting to do in the Opinion. And Baker Botts was going further by assuming that the Revised Policy was final not only for the purpose of determining Boardwalk’s cost of service but also for purposes of assessing an effect on rates.

If Baker Botts had been asked to render an opinion for a client about what might happen in the hypothetical event that the Revised Policy became final, then these assumptions would not have been problematic. But the Opinion Condition required that Baker Botts express a legal opinion based on a set of facts: whether there had been a regulatory development that “has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers.” The assumption that a sufficient trigger had happened drove the result.

In finding that Baker Botts improperly assumed that the Revised Policy was final, this decision clarifies an aspect of its ruling on the defendants’ motion to dismiss, which the defendants invoke to support their arguments. In the complaint, the plaintiffs asserted that Baker Botts “relied on assumptions that Defendants knew to be false,” including the assumption that the Revised Policy would not be changed, and argued that “the defendants purportedly ‘knew on June 29, 2018[,] that FERC’s March 15 Proposed [sic] Policy Statement would soon be ‘revised, reversed, [or] modified.’” Bandera Master Fund LP v.

122

Boardwalk Pipeline Pr’s, LP, 2019 WL 4927053, at *20 (Del. Ch. Oct. 7, 2019). The court rejected that allegation, explaining:

This assumption was not false. FERC did not revise, reverse, or modify the Revised Policy Statement. FERC issued an order on July 18, 2018, in which it declined to reconsider the Revised Policy Statement and reaffirmed that FERC “will generally not permit MLP pipelines . . . to recover an income tax allowance in their cost of service.” The Final Rule addressed other aspects of FERC’s new rate-setting policies, including the treatment of ADIT balances, but it did not revise, reverse, or modify the Revised Policy Statement.

The plaintiffs’ allegations do not support a reasonable inference that Baker Botts failed to exercise its independent judgment when it assumed that the Revised Policy Statement would not be revised, reversed, or modified. The motion to dismiss this aspect of Count II is granted.

Id. at *20 (citations omitted). The court understood the plaintiffs’ argument at the motion to dismiss stage to be that the defendants knew that the Revised Policy in fact would be changed, rendering the assumption false. Because the Revised Policy was not changed, that allegation could not support a claim on which relief can be granted.

The trial record establishes that when Baker Botts rendered the Opinion, Baker Botts and the defendants knew that the policy could be changed. The policy on the tax allowance was not changed, but the related decision on the treatment of ADIT was so substantial as to operate as a change. By assuming that the policy was final when issued on March 15, Baker Botts accelerated the date when it could render the Opinion. That decision meant that Loews did not have to wait until the terms of the Revised Policy and the related treatment of ADIT were known. Instead, Loews could exercise the Call Right during a period of maximum market uncertainty, thereby benefitting itself.

123

The record presented at trial demonstrates that the Revised Policy was not final. The fact that the lawyers who wanted the General Partner to be able to exercise the Call Right convinced themselves over time that the Revised Policy was sufficiently final to render the Opinion—and testified to that belief at trial20—does not mean that it was final. The Opinion started from a counterfactual premise that Baker Botts knew was untrue.

b. Counterfactual Assumption: Recourse Rates Would Change Without A Rate Case.

To facilitate the exercise of the Call Right, Baker Botts assumed that the rates that Boardwalk’s subsidiaries could charge would change to the subsidiaries’ detriment without a rate case. Unlike its first assumption, Baker Botts did not make this second assumption explicitly. Without that unstated counterfactual assumption, Baker Botts could not have rendered the Opinion.

To satisfy the Opinion Condition, Baker Botts had to conclude in good faith that Boardwalk’s status as a pass-through entity for tax purposes “has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers.” PA § 15.1(b)(ii). A threshold question was the meaning of “maximum applicable rates.”

If “maximum applicable rates” meant the real-world rates applicable to the shippers who purchased capacity on the subsidiaries’ pipelines, then the March 15 FERC Actions— even if they became final—would not have a meaningful effect, because the majority of

20 See Rosenwasser Tr. 65; Wagner Tr. 207; Alpert Tr. 335; McMahon Tr. 525.

124

the shippers on Boardwalk’s pipelines paid negotiated or discounted rates. As discussed in greater detail below, Baker Botts sidestepped that issue by interpreting “maximum applicable rates” to mean “recourse rates.” But that solution created another problem: Recourse rates do not change without a rate case. Assessing whether there would be a material adverse effect on recourse rates therefore required evaluating the risk that someone would bring a rate case against one of Boardwalk’s Subsidiaries. See JX 1138 at 2; JX 1307 at 7; Court Tr. 860. It also required assessing whether Boardwalk’s rates would change if a rate case was brought. See Court Tr. 861–65.

Baker Botts assumed away these issues. The Opinion did not address either the risk that someone would bring a rate case or the risk that Boardwalk’s rates would change as a result of a rate case. Instead, the Opinion implicitly made the counterfactual assumption that each of Boardwalk’s subsidiaries would be involved in a rate case and lose. See Court Report ¶¶ 113–14.

The April 4 Draft made that assumption openly, stating: “[W]e have requested that the Partnership assume that the Subsidiaries will file rate cases and take any other appropriate and legal action to be permitted to charge the maximum rates permitted under the applicable cost of service rules and regulations regardless of competitive conditions or any other non-legal factor.” See JX 1949 at 2. The April 4 Draft thus made clear that Baker Botts was assuming that the subsidiaries would act contrary to their own interests, file rate cases seeking to lower their rates, and eschew any arguments that might enable them to maintain or raise their rates.

125

The Opinion dropped the clear language from the April 4 Draft and omitted any reference to rate cases. In its place, the Opinion substituted the more laconic assumption “that each Subsidiary would charge all of its customers the maximum applicable rate.” JX 1522 at 3. That outcome only could happen if someone filed rate cases in which Boardwalk’s subsidiaries lost. The assumption from the April 4 Draft thus remained, but was now unstated. See Wagner Tr. 273–74; see also Rosenwasser Tr. 91.

Two of Boardwalk’s subsidiaries did not face any rate case risk, and a third faced only low risk. See JX 571 at 7; JX 1064; JX 1521 at 16. When issuing the NOPR, FERC made clear that many pipelines had characteristics that would obviate the need for a rate adjustment, including (i) rate moratoria, (ii) negotiated rates, or (iii) under-recovery of costs. See JX 580 ¶¶ 45, 48–49. Typically, a pipeline under-recovers its costs because it operates in a competitive market and must offer discounted rates to capture business. See JX 1139 at 11.

Those criteria mapped onto Boardwalk’s pipelines. See JX 571 at 1.

• Virtually all of Gulf Crossing’s contracted volumes were subject to negotiated rates.

PTO ¶ 139; JX 572 at 2–3. Gulf Crossing also operated in highly competitive markets, was under-recovering its cost of service and would be “highly undersubscribed ” as its negotiated-rate contracts rolled off. See JX 644 at 1; JX 676 at 8.

• A majority of Gulf South’s contracts provided for negotiated or discounted rates.

Gulf South was also subject to a rate case moratorium until May 2023. And Gulf South operated in highly competitive markets and thus was under-recovering its cost of service. See PTO ¶ 139; JX 604 at 1; JX 644 at 1; JX 676 at 8; JX 1139 at 6; JX 1521 at 16.

• A majority of Texas Gas’ contracts with shippers provided for negotiated or discounted rates. See JX 1139 at 6. Only 20% of its volumes were shipped at recourse rates and potentially subject to any effect. See JX 548 at 1. It too served highly competitive markets. Id.

126

Loews, Boardwalk and their advisors concluded there was “[n]o expected near-term rate case risk for Gulf South or Gulf Crossing” and that over the long-term, rate case risk was minimal because “current RoE [was] likely to be below allowable RoE.” JX 1521 at 16; see Wagner Tr. 269. After some initial concern about the rate case risk at Texas Gas, Baker Botts and its rate expert assured Loews that the rate case risk at Texas Gas was “low” through April 2020. JX 1064 at 1. Beyond that, Baker Botts and its rate expert believed it was impossible to “make a prediction with any confidence.” Id.; see JX 1078.

The Opinion rested on an unstated counterfactual assumption about the inevitability of an adverse decision in a rate case. If Baker Botts had been asked to render an opinion for a client about what might happen in a hypothetical world where all three subsidiaries faced rate cases and lost, then an opinion based on explicit assumptions to that effect would have been acceptable. But the Opinion Condition required that Baker Botts express a legal opinion about whether Boardwalk’s status as a pass-through entity for federal tax purposes has or “will reasonably likely in the future have a material adverse effect on maximum applicable rates.” Rendering that opinion required assessing the risk of a material adverse effect on rates, not making the unstated counterfactual assumption that each subsidiary would face and lose a rate case.

c. Counterfactual Assumption: Hypothetical Indicative Rates Are The Same As Recourse Rates.

To facilitate the exercise of the Call Right, Baker Botts made yet another counterfactual assumption: Recourse rates are the same as hypothetical indicative rates. Like the second counterfactual assumption, the third assumption was unstated.

127

As discussed previously, the Opinion Condition required that the Opinion address whether there has been or will reasonably likely be a material adverse effect on the “maximum applicable rate that can be charged to customers” The Partnership Agreement did not define “maximum applicable rate,” and FERC has not defined it either. See Court Report ¶¶ 152–55; Rosenwasser Dep. 365. None of defendants’ advisors, nor their FERC expert in this litigation, identified a FERC order or ruling that defined or explained that phrase. See Court Report ¶¶ 157–69; JX 1756 (Court Rebuttal) ¶¶ 11–17. At trial, Rosenwasser conceded that the meaning of “maximum applicable rate” was a “key” question his team “had to grapple with.” Rosenwasser Tr. 64.

Multiple law firms generated analyses of the phrase, in part because Baker Botts was unable to identify any settled meaning of the term in its first attempt. See JX 637 (email from Baker Botts interpreting the term); JX 781 at 1 (same); JX 800 at 2 (notes from Skadden interpreting the term); JX 1375 (memorandum from Baker Botts interpreting the term); JX 1437 (email from Van Ness Feldman interpreting the term). Naeve, the Skadden partner and former FERC Commissioner, believed that the phrase reasonably could mean either (1) “the maximum rate applicable to customers taking into consideration discounted contracts that have been filed at FERC,” or (2) “the maximum rate contained in the tariff which the pipeline could have charged and is free to charge other customers[.]”21 Layne,

21 JX 800 at 2. Naeve discussed this concern with Alpert. See id.; Alpert Tr. 421.

When Grossman raised the same point, Alpert was furious. See JX 798 at 1 (“Rich is pissing me off.”). Baker Botts had to send Skadden a copy of Boardwalk’s Form S-1 to “get [them] more comfortable” with the interpretation that Baker Botts needed to use. See JX 790 at 2.

128

the Vinson & Elkins transactional partner, similarly observed that there were multiple reasonable interpretations.22 The Opinion implicitly conceded that the term “maximum applicable rates” was ambiguous. Rather than asserting that the claim had a plain meaning, Baker Botts stated that

we have, in using our judgment, interpreted the words . . . “maximum applicable rate that can be charged to customers by subsidiaries that are regulated interstate natural gas pipelines of the Partnership,” to mean the recourse rates of the Subsidiaries now and in the future as that term is used by the FERC in its regulations, rulings and decisions . . . .

JX 1522 at 4 (emphasis added).

Everyone knew that a Delaware court would apply the doctrine of contra proferentem and construe ambiguous language against the General Partner and in favor of the minority unitholders. Yet to reach the conclusion that the phrase meant “recourse rates,” Baker Botts declined to apply the doctrine of contra proferentem and looked to two sources of extrinsic evidence: (i) Boardwalk’s own use of the phrase in its public filings, and (ii)

Baker Botts thus turned to extrinsic evidence to support its reading of “maximum applicable rates.”

22 See JX 733 at 1 (“One interpretation is that that means the maximum rates that could be charged, assuming the customers were paying maximum cost of service rates. On the other hand, because of the discounts, market based rates and negotiated rates (and presumably the possibility of all this getting changed by FERC again), REVENUE won’t take a hit…even though theoretical maximum rates (if we could charge them) would be materially adversely effected [sic].”).

129

FERC’s use of the phrase in orders in proceedings involving Boardwalk, where FERC was commenting on Boardwalk’s filings.

If Baker Botts had reached that interpretive judgment, assessed each pipeline’s risk of a rate case, relied on a full ratemaking analysis, and rendered opinions about the reasonably likely effect on recourse rates, then Baker Botts’ decision to interpret “maximum applicable rates” as “recourse rates” would not have fatally undermined the Opinion. Although relying on extrinsic evidence to interpret ambiguous language runs contrary to how Delaware courts interpret MLP agreements, the Delaware Supreme Court has looked on occasion to the surrounding transactional context, including by considering language in an issuer’s public filings, to give meaning to a disputed phrase. See, e.g., Airgas, Inc. v. Air Prods. & Chems., Inc., 8 A.3d 1182, 1189 (Del. 2010). Baker Botts thus could have reached a reasoned conclusion that it was appropriate under the circumstances to consider extrinsic evidence in the form of Boardwalk’s Form S-1, and Baker Botts could have concluded in good faith, based on that broader transactional context, that when drafting the Call Right, Rosenwasser meant to refer to recourse rates. In other words, to the extent that extrinsic evidence and judgment enter the picture, reading “maximum applicable rates” to mean recourse rates is a more persuasive reading than other possibilities.

But Baker Botts did not do those things. Baker Botts made an unstated assumption that resulted in the Opinion not actually interpreting the phrase “maximum applicable rate” as “recourse rates.” Baker Botts instead considered the highest rates that FERC would allow Boardwalk to charge in a hypothetical world that assumed there was a full market

130

for the pipelines’ services. JX 646 at 3. As Wagner wrote in his contemporaneous notes: “‘Max hypothetical rate.’ This is not the recourse rate.” JX 646 at 4. Other contemporaneous writings refer to the rates that Baker Botts examined as “indicative rates,” “theoretical maximum rates,” and “maximum hypothetical rates.” See JX 727 at 2 (“indicative rates”); JX 733 at 1 (“theoretical maximum rates”); JX 798 (“[I]t’s crystal clear that we’re talking hypothetical future max FERC rates.”); JX 1007 at 1 (“hypothetical rates”).

In reality, the Opinion examined indicative rates, and Baker Botts’ conclusion rested on the unstated counterfactual assumption that indicative rates were the same as recourse rates. If Baker Botts had been asked to render an opinion for a client about what might happen to “hypothetical future max FERC rates,” then equating indicative rates with recourse rates would not have been problematic. The Opinion Condition, however, did not turn on “hypothetical future max FERC rates.” The Opinion Condition required that Baker Botts express a legal opinion about whether Boardwalk’s status as a pass-through entity for tax purposes “has or will reasonably likely in the future have a material adverse effect on maximum applicable rates.” Once Baker Botts expressly assumed that “maximum applicable rates” were the same as “recourse rates,” Baker Botts had to stick with that assumption. Instead, Baker Botts made an additional, unstated, and counterfactual assumption that recourse rates were the same as “hypothetical future max FERC rates.”

131

d. Counterfactual Assumption: The Treatment Of ADIT Was Known.

To facilitate the exercise of the Call Right, Baker Botts made a fourth counterfactual assumption. Like the second and third assumptions, it too was implicit. This time, Baker Botts assumed that the open question of how FERC would treat ADIT was a known fact and that FERC would use the Reverse South Georgia Method. In reality, no one knew how FERC would treat ADIT, and it was impossible to determine what effect the March 15 FERC Actions would have on rates without knowing how FERC would treat ADIT.

In the March 15 FERC Actions, FERC made clear that the treatment of ADIT was an open issue. The ADIT NOI sought industry input on that very question. See JX 576 ¶ 25. FERC staff specifically flagged whether an MLP’s accumulated ADIT balance should be eliminated from cost of service or whether those previously accumulated sums should be placed in a regulatory liability account and returned to ratepayers. See id.

Boardwalk understood that the treatment of ADIT was an open issue. In Johnson’s initial analysis of the impact of the March 15 FERC Actions, he characterized his estimate of the downside as a floor, because it “ignores any bounce from rate base increase associated with removal of ADIT.” JX 572 at 1–2. Elaborating in a later email, he explained that “it’s unclear on what they [FERC] would do with [Boardwalk’s] current ADIT” balance. JX 602 at 1. He further observed that FERC could decide that the ADIT balance should be “zeroed out because there’s no income taxes (because there would be no difference between book and tax depreciation).” Id.

132

When the Loews executives examined Johnson’s analysis, they likewise recognized that ADIT was the critical issue. JX 601 at 2. A Loews employee determined that losing the income tax allowance was “a flesh wound for the long haul pipes like . . . [Boardwalk].” Id. at 1. But if FERC required that pipelines return their ADIT balances to ratepayers, then that “would be the a-bomb outcome” and would be “extremely painful.” Id.

Baker Botts knew that the future treatment of ADIT was an open issue. Just four days into Baker Botts’ engagement, Wagner acknowledged that “FERC has not stated how to treat ADIT balances” and “[t]his can affect the rate impact on the pipelines substantially.” JX 619 at 1. Wagner explained to Alpert in an email on March 20, 2018, that the ADIT NOI did not have a time frame for resolution but could be resolved by the end of 2018. JX 626 at 1. He noted that any regulation was “not likely to be self- implementing and would require additional proceedings to affect pipeline rates.” Id.

During a call on March 22, 2018, Boardwalk executives and Baker Botts lawyers discussed whether they could estimate the effect of ADIT, concluding that they had “[n]o idea [because we] don’t know rules.” JX 646 at 1; see JX 644 at 1 (noting the “lack of clarity on FERC’s eventual policy on” the treatment of ADIT and characterizing any possible effects as “highly speculative at this point”); JX 740 at 1 (“[W]e may want to see the results under a few different scenarios.”); JX 868 at 2 (“[D]ifferent assumptions on how to handle [the ADIT] issue could affect the calculations.”); see also JX 1525 at 67 (Sullivan testifying that FERC was still determining “how [ADIT] balances will be treated”). The Loews executives likewise understood that they did not have the answer on ADIT. See JX 567; JX 601 at 1–2.

133

A chorus of defense witnesses testified at trial that they believed that FERC would instruct pipelines to amortize ADIT using the Reverse South Georgia Method. 23 That was indeed one reasonable method, and the witnesses’ testimony about their belief seemed convincing. The problem is that the Reverse South Georgia Method was only one possibility, and no one knew what FERC actually would do.

Without knowing how FERC would treat ADIT, it was impossible to determine what effects the March 15 FERC Actions would have. In its public comments on the NOPR, Boardwalk emphasized that, “[u]ntil the Commission provides a final decision on the treatment of ADIT, Boardwalk cannot correctly assess the impact of the Revised Policy Statement and ADIT on its pipelines’ costs of service . . . .” JX 1130 at 14. Skadden understood what that meant for the Opinion. In a model of understatement, Voss described the language as “relatively unhelpful.” JX 1164 at 1. Rosenwasser also knew the language posed a problem. In his personal notes on Boardwalk’s NOPR comments, Rosenwasser underlined the text and double-starred it. See JX 1138 at 14.

Boardwalk’s comment was more than just unhelpful. It established that Baker Botts had no basis for the Opinion.

The Opinion thus rested on the unstated counterfactual assumption that the treatment of ADIT was known and would follow the Reverse South Georgia Method. If Baker Botts had been asked to render an opinion for a client about what the effect on rates

23 See Rosenwasser Tr. 78; Wagner Tr. 217–18, 223; Alpert Tr. 347; McMahon Tr.

497, 517; Johnson Tr. 619.

134

would be if FERC required amortization of ADIT using the Reverse South Georgia Method, then making that counterfactual assumption would have been fine. But the Opinion Condition required an opinion based on fact. Instead, Baker Botts assumed its way to a conclusion that a sufficient regulatory development had occurred.

2. The Factual Inputs The foregoing assumptions formed the basis for Rosenwasser’s syllogism. That exercise dictated the result of the Opinion by deploying elementary subtraction. Baker Botts then obtained information from Boardwalk to make the syllogism work.

a. Rosenwasser’s Syllogism As described in the Factual Background, Rosenwasser developed his syllogism so that Baker Botts could render the Opinion. Rosenwasser knew that the Call Right was intended to address a business issue by protecting Loews against a regulatory change that would have a materially adverse effect on Boardwalk. Rosenwasser Dep. 39–40. Rates were relevant because they led to revenue. McMahon Tr. 545. The Call Right was not intended to create a regulatory trapdoor that could be triggered by a change that “wasn’t substantive, wasn’t meaningful.” Rosenwasser Tr. 46. In fact, Rosenwasser did not believe that “rates” were what the Call Right was designed to protect. JX 1502 at 34 (“Rates themselves are not what is being protected. It must be the entities charging the rates.”). The Call Right was intended to provide Loews with an “off-ramp” if FERC changed its policy in a way that materially threatened Boardwalk as an entity. McMahon Tr. 480, 545.

That understanding comported with how Delaware cases approach the concept of a material adverse effect. Determining whether a material adverse effect is reasonably likely

135

to occur involves forecasting, not fantasizing. “There must be some showing that there is a basis in law and in fact for the serious adverse consequences prophesied by the party claiming the MAE.” Akorn, Inc. v. Fresenius Kabi AG, 2018 WL 4719347, at *65 (Del. Ch. Oct. 1, 2018) (quoting Frontier Oil v. Holly Corp., 2005 WL 1039027, at *36 n.224 (Del. Ch. Apr. 29, 2005)), aff’d, 198 A.3d 724 (Del. 2018). Simply “proclaiming that bad things can happen” is insufficient to establishing that a material adverse effect is reasonably likely to occur. See Frontier Oil, 2005 WL 1039027, at *36 n.224.

The March 15 FERC Actions were not reasonably likely to have a material adverse effect on Boardwalk. The Boardwalk management team determined immediately that the March 15 FERC Actions were not reasonably likely to have a material adverse effect on Boardwalk’s revenue. See JX 615 at 1; JX 733 at 1. The March 15 FERC Actions also were not reasonably likely to have a material adverse effect on recourse rates. Two of Boardwalk’s pipelines had characteristics which meant that if the March 15 FERC Actions became final, they would not face a rate proceeding. For the third pipeline—Texas Gas— the risk of a rate proceeding was low, and any effect on revenue would be small.

To deliver the Opinion, Rosenwasser needed to shift from the real world into an imaginary one. He therefore took the position that the Call Right was not concerned with the actual economic impact; it was only concerned with the abstract concept of “maximum applicable rates.” See JX 645 at 1; JX 679 at 5, 8. If a regulatory policy affected that abstract concept, then the Call Right could be exercised. And because a tax allowance had been built into the cost-of-service calculation, a policy change eliminating the allowance would lead ineluctably to a change in the maximum applicable rate, as Baker Botts was defining

136

that term. When Wagner heard Rosenwasser’s reasoning, he immediately understood what they were doing: “Just saying” that eliminating the tax allowance led to a lower cost of service and therefore a material adverse effect. JX 639.

The resulting syllogism turned on elementary subtraction, and it was fundamentally flawed. Boardwalk knew that. During a discussion of the March 15 FERC Actions, Jonathon Taylor from the FERC Office of General Counsel foreshadowed what would become Rosenwasser’s syllogism when he explained that “when a tax expense decreases, so does the cost of service.” JX 588 at 22. At the time, McMahon and his outside counsel ridiculed that line of reasoning. McMahon wrote to Gregory Junge, a regulatory lawyer: “That was a priceless statement[.] [T]axes go down[.] COS goes down[.] This is going to be a train wreck.” JX 575 at 2. Junge responded: “That is . . . just [the] type of 1:1 thinking that we were trying to explain is not the case.” Id. And in its comments to FERC on the NOPR, Boardwalk rejected that simplistic approach. Boardwalk asserted that it was “misleading” to equate a change in the cost of service stemming from the removal of the income tax allowance with a “rate reduction,” because a cost-of-service change has “little bearing” on whether or not a rate reduction will occur. JX 1138 at 30 (NOPR Comments). If FERC tried it, then it would violate its policy against single-issue ratemaking. JX 1307 at 7; see Johnson Tr. 663.

Grasping for grounds to confirm that this approach was nevertheless justified, Rosenwasser relied on the fact that the Opinion called for a legal opinion from counsel, not a factual opinion from some other type of professional like a rate expert or an investment banker. See JX 646 at 3 (“This is a legal opinion, independent of what’s happening in mkt.

137

Not a primarily factual analysis.”); see also JX 686 at 1; Rosenwasser Tr. 49–51. That is nonsensical; the notion that the Partnership Agreement called for a legal opinion did not mean that the opinion could ignore facts. Lawyers (and law-trained judges) apply the law to facts. Legal opinions turn on facts. See Glazer et al., supra, § 4.1 at 82.

Not surprisingly, Rosenwasser and Baker Botts could not maintain the pretense that the Opinion did not require considering real-world facts. Uncertain about whether it could opine that the effect on indicative rates was sufficiently material and adverse, Baker Botts wanted to consider other indications of materiality, such as the effect that a comparable reduction in revenue would have on Boardwalk’s EBIT, EBIDTA, and distributable cash flow. See PTO ¶ 182; JX 775 at 1. Rosenwasser sought reassurance from Richards Layton that Baker Botts could consider these other effects, but Richards Layton advised the “[b]etter [r]eading” was to “look [at] rates more, not effects.” JX 1007 at 1. Even then, Baker Botts referred to the pass-through effect in the Opinion, stating that

[w]e also took notice that, because these reductions in the maximum applicable rates would not be offset by any reduction in costs incurred by the Subsidiaries, the reductions in the maximum applicable rates would have a substantially larger percentage impact on the earnings before interest and taxes and on the cash available for distribution of each of the Subsidiaries assuming each Subsidiary could actually charge and collect its maximum applicable rate.

JX 1522 at 3. Baker Botts thus considered real-world effects when doing so helped reach the result that its client wanted, but not when doing so might cut in the opposite direction.

Rosenwasser’s syllogism ignored that the Call Right was drafted to address a business issue, not an abstract legal question. The syllogism ignored the absence of any real-world effect on revenue in favor of focusing on recourse rates. It ignored the question

138

of rate case risk and the real-world events that would have to take place before there was any effect on recourse rates. The syllogism was a contrived exercise designed to achieve a particular result.

b. The Rate Model Analysis To provide the factual basis for the Opinion, Baker Botts had Boardwalk prepare the Rate Model Analysis. That analysis implemented Rosenwasser’s syllogism and was designed to “get us where we need to go.” JX 713 at 1. The exercise generated declines in hypothetical indicative rates of 11.68%, 12.12%, and 15.62% under circumstances where the rates that shippers actually paid had not changed at all and where recourse rates were unlikely to change for the foreseeable future.

The Rate Model Analysis departed from ratemaking principles. The Rate Model Analysis calculated a single, hypothetical, indicative rate for each of Boardwalk’s three pipeline subsidiaries. See JX 1415 at 3. It then projected that the indicative rate would drop as a result of the removal of income tax allowance. See id. In other words, the Rate Model Analysis changed only the income tax allowance variable while holding all else constant. See, e.g., JX 639 at 1; Wagner Tr. 258; Webb Tr. 938. That is single-issue ratemaking.

Through single-issue ratemaking, the Rate Model Analysis avoided any meaningful assessment of how, if at all, a change in the cost of service might impact any of the 167 recourse rates that Boardwalk had on file with FERC. Sullivan, the rate expert hired by Baker Botts, testified in his deposition that FERC would not focus on an indicative rate because it does not “mean anything.” Sullivan Dep. 169. He confirmed that the Rate Model Analysis calculated a cost-of-service reduction, not a rate reduction. Id. at 118. He

139

explained that deriving an indicative rate reduction by changing one cost-of-service variable was “kind of meaningless” because a rate change does not depend on one cost-of- service variable. Id. at 101. He observed that the Rate Model Analysis could not be used to calculate the change to Boardwalk’s actual recourse rates. Id. at 150. At trial, the plaintiffs’ rate expert testified persuasively on these same points. See Webb Tr. 913–14 (describing indicative rates as “meaningless” and “hypothetical”).

Because the Rate Model Analysis employed a simple syllogism, it only contained a few pages of analysis. The calculations for the purported rate impact at Texas Gas took only five pages. Johnson Tr. 640, 652. By contrast, the rate models used in actual rate cases involve hundreds of pages of complex calculations to determine cost of service and, ultimately, recourse rates. See Webb Report ¶ 174; see also Johnson Tr. 653 (conceding that Gulf South’s initial submission in a recent rate case spanned 3,844 pages). The Rate Model Analysis was much shorter because it skipped essential steps in the ratemaking process. See, e.g., Johnson Tr. 651–52 (conceding that the Rate Model Analysis did not calculate discount adjustments); id. at 648–49 (conceding that FERC requires use of zone- based rate design where pipelines employ zones but the Rate Model Analysis failed to do so). At the same time, the Rate Model Analysis applied a de-functionalizing step that is not part of ratemaking process. Webb Tr. 967.

The resulting simplified calculation was highly sensitive to assumptions about ADIT and ROE. The Rate Model Analysis thus confirms that Baker Botts could not opine on the effect of the March 15 FERC Actions on rates without knowing more about the regulations that FERC intended to adopt.

140

The Rate Model Analysis assumed that FERC would require amortization of ADIT using the Reverse South Georgia Method, which was one possibility. Virtually all of the pipelines (other than Boardwalk) publicly advocated for FERC to eliminate ADIT. Changing from the Reverse South Georgia Method to the elimination of ADIT would have eliminated Baker Botts’ ability to claim a material adverse effect on indicative rates.

Baker Botts

Percentage BB % Change with Subsidiary Change ADIT Adjustment Texas Gas 12.12% 2.58% Gulf South 11.68% 1.80% Gulf Crossing 15.62% -0.85%

Webb Report ¶ 128 fig. 6. The changes at Texas Gas and Gulf South become minimal, and Gulf Crossing’s rates move in the opposite direction.

The Rate Model Analysis was also sensitive to assumptions about ROE. While Baker Botts was working on the Opinion, some industry participants thought that FERC might permit pipelines to calculate their cost-of-service requirements using higher ROEs to offset the effect of the lost income tax allowance. See, e.g., JX 910 at 9 (“Guggenheim [Partners, LLC] thinks . . . . the change to the tax allowance might not be material, as the increased ROE could recover the cost lost by losing the tax allowance.”). While he was acting as Baker Botts’ rate expert, Sullivan gave testimony in which he advocated for increased ROEs. See Webb Report ¶¶ 132–33 (collecting Sullivan’s advocacy); Sullivan Dep. 55 (conceding that he would have used a 13.5–14% ROE in a rate case).

Increasing the ROE in the Rate Model Analysis from 12% to 14% lowers the percentage change in rates by approximately five percent:

141

Baker Botts

Percentage BB % Change with Subsidiary Change ROE Adjustment Texas Gas 12.12% 7.14% Gulf South 11.68% 6.91% Gulf Crossing 15.62% 9.32%

See Webb Report ¶ 134 fig. 7. The changes at all three pipelines fall below the level that Baker Botts opined could give rise to a material adverse effect.

Changing both variables in the Rate Model Analysis—eliminating ADIT and increasing the permissible ROE—reverses the direction of the change in indicative rates.

Baker Botts BB % Change with Percentage Both ROE Correction Subsidiary Change and ADIT Adjustment Texas Gas 12.12% -3.33% Gulf South 11.68% -3.95% Gulf Crossing 15.62% -8.66%

See Webb Report ¶ 136 fig. 8. Instead of a projected decrease (which Baker Botts reports as a positive percentage), there is a projected increase (reflected as a negative percentage). That means that indicative rates would increase, resulting in a beneficial effect rather than an adverse effect. The plaintiffs concede that these outputs do not mean that Boardwalk’s recourse rates were reasonably likely to rise. See Webb Tr. 959. What they demonstrate is that the Rate Model Analysis depended heavily on assumptions, including an answer on the treatment of ADIT that no one knew when Baker Botts rendered its Opinion.

The Rate Model Analysis could not provide an adequate factual basis for the Opinion. The Rate Model Analysis simply implemented Rosenwasser’s syllogism, which ignored real world effects but allowed Baker Botts to reach the conclusion its client wanted.

142

3. Other Efforts To Reach The Desired Conclusion After making all of the foregoing efforts to create a structure that would permit the issuance of the Opinion, Baker Botts still had to stretch to render the Opinion. Those strained conclusions are signs of motivated reasoning.

Most notably, Baker Botts stretched on what constituted a material adverse effect.

Richards Layton advised that “the better argument” was that a reduction in rates of 12– 13%, in perpetuity, would suffice for a material adverse effect.24 The Skadden attorneys believed that an 11% change was “likely insufficient” under Delaware law, although the duration of the change would be a pertinent consideration. See JX 772 at 1. In the Opinion, Baker Botts went further and took the position that a material adverse effect would result from “an estimated reduction in excess of ten percent in the maximum applicable rates that can be charged to the customers of each of the Subsidiaries on a long-term basis.” JX 1522 at 3 (emphasis added); see Rosenwasser Tr. 96–98. Baker Botts had to dip below 12% because the Rate Model Analysis generated a decline of 11.68% in the hypothetical indicative rates that Texas Gas could charge. See JX 1522 at 3.

And Baker Botts stretched on other issues as well:

• Baker Botts was not sure what standard to use for “reasonably likely to have a material adverse effect.” Rosenwasser decided to “call it more likely than not.” JX 1807 at 12; accord Rosenwasser Tr. 98–99.

24 JX 975 at 1; JX 1507 at 1–2. At trial, Raju testified that Richards Layton thought the “better argument” was that “a 10 percent or greater adverse effect into perpetuity on the rates metric would constitute an MAE.” Raju Tr. 800–01. The contemporaneous documents do not provide that additional color.

143

• Baker Botts viewed the reference to the Partnership’s “status as an association not taxable as a corporation” as incorrect terminology. JX 939. Baker Botts decided to “tear off the band-aid and substitute ‘entity’ for ‘association’ in our statement of our opinion.” Id. Thus, the real issue, as Baker Botts saw it, was the Partnership’s status as an MLP. JX 733 at 1.

In substance, Baker Botts rewrote the Call Right so that it could render the Opinion.

As written, the Call Right required an opinion that

the Partnership’s status as an association not taxable as a corporation and not otherwise subject to an entity-level tax for federal, state or local income tax purposes has or will reasonably likely in the future have a material adverse effect on the maximum applicable rate that can be charged to customers by subsidiaries of the Partnership that are regulated interstate natural gas pipelines.

PA § 15.1(b).

As rewritten by Baker Botts, the Call Right called for an opinion that

a notice of a proposed regulation about whether a regulated interstate natural gas pipeline organized as an MLP can claim an income tax allowance in its cost of service the Partnership’s status as an association not taxable as a corporation and not otherwise subject to an entity-level tax for federal, state or local income tax purposes has or will reasonably likely more likely than not in the future will have a material 10% or more adverse effect on the maximum applicable hypothetical indicative rates that can be charged to customers by subsidiaries of the Partnership that are regulated interstate natural gas pipelines if each subsidiary faces and loses a rate case in which FERC (i) removes only the income tax allowance from the pipeline’s cost of service, (ii) requires amortization of ADIT using the Reverse South Georgia method, (iii) does not conduct the other steps in the ratemaking process, (iv)

does not consider rate moratoria, the effects of competition, or other factors that FERC considers when determining rates, and (v) thereby violates the policy against single-issue ratemaking.

Baker Botts chose to give the latter opinion. It could not have given the former opinion.

144

4. Knowingly Going Where Others Would Not Tread In addition to counterfactual assumptions, in addition to Rosenwasser’s syllogism, and in addition to stretching on a series of issues that amounted to rewriting the Call Right, at least two other dimensions of Baker Botts’ conduct support a finding of bad faith. Baker Botts rendered a non-explained opinion on a complex issue of Delaware law that the two Delaware law firms who were consulted would not formally address. And Baker Botts did so in the face of fatal uncertainty that could have been mitigated simply by waiting.

Baker Botts is a sophisticated law firm, but it is not a Delaware law firm. Baker Botts is also a leader in transactions involving MLPs, but it is not in the habit of opining on complex issues of Delaware limited partnership law. Many sophisticated firms render closing opinions on routine issues of Delaware entity law, such as the due formation of an entity or the due authorization of a contract. Baker Botts generally rendered enforceability opinions under the Delaware Revised Uniform Limited Partnership Act, but the firm did not render opinions more broadly on other Delaware issues. See JX 878 at 4.

In this case, Baker Botts took on one of the most difficult issues under Delaware law: determining the existence of a material adverse effect. Neither of the Delaware firms in this case would render such an opinion. Skadden has a policy against rendering an opinion on whether an event constitutes a material adverse effect, and Grossman was not willing to give Baker Botts any work product that might be construed as expressing an opinion. See JX 771 at 1. Richards Layton gave oral advice about what was the “better argument” and was willing to memorialize its advice in an email, but it would not go further

145

than that and would not let Baker Botts reference its views. See JX 975 at 1; see Raju Dep. 113–14.

Internally, Baker Botts appropriately questioned its ability to render this opinion under Delaware law. Initially, Baker Botts sought to recast the matter as an issue of federal law. See JX 679 at 7. After accepting that it was a Delaware law question, Baker Botts looked to Skadden for help. See JX 770 at 1; JX 772. Skadden, however, only provided a summary of the main Delaware authorities and disclaimed any intent to analyze the Call Right. JX 900 at 2. That fell short of what Baker Botts wanted. See JX 913 at 1; see also JX 936 at 1. Facing a deadline from Loews, Rosenwasser turned to Richards Layton, but in an effort to obtain advice that would reassure his partners, Rosenwasser provided the Richards Layton attorneys with a misleading description of the factual record. See Part I.L, supra. Rosenwasser’s query resulted in Richards Layton’s oral advice that the firm would have a “hard time saying [a decline of 12% in perpetuity is] not material.” JX 1007 at 2. Richards Layton later stated that subject to assumptions and carveouts, it would regard as the “better argument” the contention that a 12–13% change in rates in perpetuity was sufficiently material and adverse, but Richards Layton would not let Baker Botts reference its advice in the Opinion. JX 975 at 1; see Raju Dep. 113–14.

Baker Botts nevertheless rendered a non-explained opinion to the effect that a 10% decline in indicative rates was reasonably likely to constitute a material adverse effect. Baker Botts, a non-Delaware firm that did not regularly render opinions on complex Delaware issues, did not explain how it reached that conclusion. It did not identify any indicators of materiality that would justify that threshold. It did not discuss and distinguish

146

the well-known and (at that point) unbroken line of transactional cases which had failed to find a material adverse effects, such as In re IBP, Inc. Shareholders Litigation, 789 A.2d 14 (Del. Ch. 2001), Frontier Oil Corp. v. Holly Corp., 2005 WL 1039027 (Del. Ch. Apr. 29, 2005), Hexion Specialty Chemicals, Inc. v. Huntsman Corp., 965 A.2d 715 (Del. Ch. 2008), or Mrs. Fields Brand, Inc. v. Interbake Foods LLC, 2017 WL 2729860 (Del. Ch. June 26, 2017). Baker Botts acted as if it was rendering a third-party closing opinion on a routine issue, which it plainly was not. The fact that Baker Botts rendered a non-explained opinion on the existence of a material adverse effect itself suggests that Baker Botts was serving Loews’ interests.

The timing of the Opinion points in the same direction. Given the non-final nature of the Revised Policy, the avalanche of comments that FERC received, the direct linkage between the Revised Policy and the ADIT NOI that Boardwalk itself identified, and the uncertainty regarding the treatment of ADIT, Baker Botts could not have believed in good faith that it could render the Opinion before FERC provided further guidance. There were too many known unknowns. And an opportunity for clarity on these unknowns was on the horizon: FERC was likely to provide more guidance at its meeting on July 19, 2018. Baker Botts needed to wait.

Naeve, the Skadden partner and former FERC Commissioner, candidly observed in real time that Baker Botts should have waited. He wrote to a colleague, “If I were Baker Botts I would prefer to wait until FERC acts on the comments.” JX 1076 at 1. Among other things, Naeve noted that the Revised Policy was a “blunt instrument that ignore[d]” the fact that some MLPs (including Boardwalk) were “predominately owned by C-corps that

147

pay federal income taxes.” Id. Naeve described how “the 5 FERC Commissioners testified before a House Subcommittee and were grilled on this issue and others.” Id. According to Naeve, “at least one Commissioner appeared to be having second thoughts about whether the Commission had fully considered industry input before acting.” Id.

Yet Baker Botts pushed ahead. In doing so, Baker Botts gave Loews the ability to exercise the Call Right to maximum effect, during a fleeting period of maximum uncertainty before FERC provided additional information on its future decisions. Rather than exercising reasoned judgment, Baker Botts knowingly served Loews’ interests.

5. The Human Dynamics In the course of evaluating whether the Opinion was rendered in good faith, the court has taken account of the professional and personal incentives that Baker Botts faced. Throughout its work on the Opinion, Baker Botts approached the assignment with an advocate’s mindset. “Lawyers by nature tend to be loyal to their clients. This is sort of baked into our professional rules.” Williams Cos., 159 A.3d at 280 (Strine, CJ., dissenting). Baker Botts strived to conclude that the General Partner could exercise the Call Right because that is what its client wanted.

Rosenwasser had an additional, personal incentive to push the limits. He drafted the Call Right, and he understandably wanted that provision to accomplish what his client thought it should do.25 And Loews was a forceful client. Throughout the events giving rise

25 Loews and Baker Botts recognized that Rosenwasser’s prior representation of Boardwalk in connection with its IPO and the drafting of the Partnership Agreement created a conflict of interest, and they called it out in Baker Botts’ engagement letter. In an

148

to this litigation, Alpert demonstrated that he knew how to manipulate his outside counsel so that counsel would deliver the answers that he wanted to receive. Sometimes he did so subtly, as when he called for an immediate teleconference after receiving Wagner’s email about the March 15 FERC Actions not being final.26 Sometimes, he was less subtle, as

effort to neutralize it, they included the following statement: “We [Baker Botts] believe, and you have agreed, that the prior work by [Rosenwasser and other lawyers] while at Vinson & Elkins LLP for Boardwalk, is not substantially related to the Matter.” JX 906 at 2.

That was not true. Under any reasonable understanding of the term, the two matters were “substantially related.”

Beyond switching sides in the same matter, the concept of substantial relationship applies to later developments out of the original matter. A matter is substantially related if it involves the work the lawyer performed for the former client. For example, a lawyer may not on behalf of a later client attack the validity of a document that the lawyer drafted if doing so would materially and adversely affect the former client.

Restatement, supra, § 132 cmt. d(ii); see J.E. Rhoads & Sons, Inc. v. Wooters, 1996 WL 41162, at *4 (Del. Ch. Jan. 26, 1996) (applying rule to disqualify a firm from litigating a case that involved an employment agreement that was part of a transaction that the firm helped negotiate and document).

This court expresses no view regarding Baker Botts’ compliance with the ethical rules, both because in most circumstances any resulting conflict can be waived, and because any ethical issue did not affect the fairness of these proceedings. Cf. In re Appeal of Infotechnology, Inc., 582 A.2d 215, 220 (Del. 1990) (holding that a trial court has no authority to rule on ethical issues involving Delaware lawyers, because that subject falls within the exclusive jurisdiction of the Delaware Supreme Court). The point is rather that the issue created by Rosenwasser’s former representation was front and center for everyone. A related point is that the General Partner and Baker Botts attempted to deal with the issue by agreeing to something that was untrue.

26 See JX 616 at 1; e.g., Rosenwasser Tr. 183–84 (testifying about obvious pressure from Alpert and Loews to give a “thumbs up”); JX 1225 (obtaining advice from Richards Layton to push back on Skadden without informing Richards Layton that Loews had

149

when he “really beat on Skadden” until they “fell in line,” but nevertheless decided to impose a consequence on Skadden by “look[ing] to other firms re potential litigation.” JX 1136 at 1.

It is also contextually relevant that the Opinion was rendered for an interested transaction involving an MLP. In the MLP ecosystem, interested transactions abound and become routinized. Governance practices are frequently suboptimal, and the Delaware courts have had cause to question opinions rendered to facilitate transactions (albeit by financial advisors rather than lawyers).27 The court recognizes that a parade of lawyers testified that they subjectively acted in good faith. Where, as here, witnesses testify about their intent, the trial judge must “make credibility determinations about [each] defendant’s subjective beliefs by weighing witness testimony against objective facts.” Allen v. Encore Energy P’rs, L.P., 72 A.3d 93, 106 (Del. 2013). The credibility determination turns in part on “the demeanor of the witnesses whose states of mind are at issue.” Johnson v. Shapiro, 2002 WL 31438477, at *4 (Del. Ch. Oct.

already consulted the independent directors); JX 1262 at 1 (bringing in Davis Polk to address what Alpert described as “unusual language” in the Opinion).

27 See, e.g., Gerber, 67 A.3d at 422 (holding that plaintiffs stated a claim because a fairness opinion “did not fulfill its basic function”); In re El Paso Pipeline P’rs, L.P. Deriv. Litig., 2015 WL 1815846, at *21–22 (Del. Ch. Apr. 20, 2015) (“[The financial advisor’s] work product further undermined any possible confidence in the Committee. . . . [the financial advisor’s] actions demonstrated that the firm sought to justify Parent’s asking price and collect its fee.”); cf. Allen v. El Paso Pipeline GP Co., 113 A.3d 167, 188 (Del. Ch. 2014) (denying a motion for summary judgment on an implied covenant of good faith and fair dealing claim where a fairness opinion did not take into account the possibility of excessive dilution), aff’d, 2015 WL 803053 (Del. Feb. 26, 2015) (TABLE).

150

18, 2002). A finding that a witness’ account is not credible does not mean that the witness lied. Human recall is not like playing a video tape. The act of remembering shapes recollection, as does the context in which the remembering takes place. A wide range of situational and subjective factors prime and shape first-hand accounts. When a witness’ conduct is at issue, and as the witness strives to recall what happened in a setting where a particular set of recollections both supports the witness’ self-image and generates a favorable outcome in the case, it is understandable that the witness could come to believe in a personally favorable account, while failing to recall or discounting contrary beliefs or disconfirming evidence.

A finding that a party did not act in good faith does not require a confession. It requires that the plaintiff prove by a preponderance of the evidence that the party in question knew it was not acting legitimately when it performed the actions in question. That finding can be made even if the human actors for that party convince themselves after the fact that they acted properly.

6. The Court’s Finding Based on the foregoing confluence of factors, and the more detailed recitation set forth in the Factual Background, the plaintiffs proved the Opinion did not reflect a good faith effort to discern the facts and apply professional judgment. The Opinion therefore failed to satisfy the Opinion Condition.

The analysis of the Opinion is necessarily holistic. Although this decision has discussed various aspects of the Opinion individually, it is the totality of the evidence that results in the finding that the Opinion did not reflect a good faith effort.

151

If Baker Botts had only stretched once or twice, or made an isolated counterfactual assumption, then it would not be possible to reject the Opinion. Under those circumstances, the court might have disagreed with Baker Botts’ assessments, but those disagreements would not have been sufficient to support a lack of good faith. But here, the record as a whole depicts a contrived effort to generate the client’s desired result when the real-world facts would not support it. Baker Botts produced a simulacrum of an opinion, and that flawed imitation did not satisfy the Opinion Condition. C. The Failure To Satisfy The Acceptability Condition Before the General Partner could exercise the Call Right, the General Partner also had to satisfy the Acceptability Condition. PA §§ 1.1 at 24, 15.1(b)(ii). The Opinion Condition derives directly from Section 15.1. The definition of “Opinion of Counsel” adds the Acceptability Condition. If the Opinion was not acceptable, then the Acceptability Condition could not be met and the General Partner could not exercise the Call Right.

The General Partner purported to satisfy the Acceptability Condition by having Holdings determine in its capacity as Sole Member of the GPGP that the Opinion was acceptable. But the language of the operative agreements is ambiguous as to whether Holdings or the GPGP Board has the authority to make that determination. One reading of the relevant agreements would recognize Holdings as having that authority. That reading rests on textual hooks in the Partnership Agreement and the LLC Agreement, but it renders the Acceptability Condition surplusage. Another reading of the relevant agreements would recognize the GPGP Board as having the authority to make the acceptability determination.

152

That reading has fewer textual supports but meshes better with the overall structure of the agreements. Both readings are reasonable.

As this decision has discussed, the doctrine of contra proferentem applies when a partnership agreement governing an MLP is ambiguous. That doctrine calls for the court to apply the reading that is more favorable to the limited partners. The reading that the GPGP Board had authority to make the acceptability determination is more favorable to the limited partners than a reading in which Holdings, an entity where all of the decision- makers were Loews insiders, had authority to make the acceptability determination in its own interests. Under the contra proferentem doctrine, the GPGP Board had the authority to make the acceptability determination. Because it did not, the Acceptability Condition was not satisfied.

1. The Contractual Language The Acceptability Condition exists because the Call Right uses the defined term, “Opinion of Counsel.” PA § 15.1(b). The Partnership Agreement defines “Opinion of Counsel” simply as “a written opinion of counsel . . . acceptable to the General Partner.” Id. § 1.1 at 24. The Partnership Agreement defines “General Partner” to mean “Boardwalk GP, LP . . . except as the context otherwise requires.” Id. § 1.1 at 18 (punctuation omitted).

The Partnership Agreement does not go further in defining who determines whether an Opinion of Counsel is acceptable. It does not discuss the internal governance structure of the General Partner or identify what organ within the General Partner would make the acceptability determination. Traditionally, a general partner would be a natural person or an entity with a single governing body, such as a corporation with a board of directors. In

153

that scenario, it would be clear who would make the determination. But Loews chose a more complicated structure. When Loews created Boardwalk, it structured the General Partner as another limited partnership, then installed the GPGP as its general partner. The GPGP is a limited liability company with both a board of directors (the GPGP Board) and a sole member (Holdings). The GPGP Board has general authority to act on behalf of the GPGP. The Sole Member has specific authority to make certain decisions on behalf of the GPGP.

The Partnership Agreement did not attempt to allocate authority for the acceptability determination among the multiple entities and decision-makers that Loews created. The Partnership Agreement only spoke in terms of action by the General Partner. To the extent that the Partnership Agreement considered the internal structure of the General Partner, it contemplated that the General Partner would have a board of directors. See, e.g., PA § 7.9(a). From a structural standpoint, the Partnership Agreement implied that the General Partner would make decisions through a board of directors.

Rather than assigning authority over different decisions to different actors, the Partnership Agreement distinguished between actions that the General Partner took in an individual capacity and actions that the General Partner took in an official capacity. The Partnership Agreement explains “[b]y way of illustration and not of limitation,” that if a provision uses “the phrase ‘at the option of the General Partner,’ or some variation of that phrase,” then that language “indicates that the General Partner is acting in its individual capacity.” PA § 7.9(c). The Call Right contains that type of signaling language, so the decision whether to exercise the Call Right is a decision that the General Partner makes in

154

its individual capacity. See id. § 15.1(b) (stating that General Partner has the “right . . . exercisable at its option . . . to purchase” all the outstanding limited partner interests so long as it satisfies the preconditions). The Opinion of Counsel definition does not have that signaling language. See id.§ 1.1 at 24.

Notably, whether the General Partner is acting in an individual capacity or an official capacity does not imply that a different decision-maker makes the decision. If the general partner was a natural person or an entity with a single governing body, such as a corporation with a board of directors, then the same decision-maker would make the decision regardless of whether the general partner was acting in an individual capacity or an official capacity. What would change is the contractual standard of review that would apply to the resulting decision.28 For present purposes, the issue is not what standard of review to apply to the General Partner’s decision to exercise the Call Right. The issue is whether the proper decision-maker made the decision.

28 Compare PA § 7.9(b) (providing the standard of review for a decision made by the General Partner “in its capacity as the general partner of the Partnership as opposed to in its individual capacity”), with id. § 7.9(c) (providing the standard of review for a decision made by the General Partner “in its individual capacity as opposed to in its capacity as the general partner of the Partnership”); see also JX 1201 at 48 (“Any exercise by our general partner of its call right is permitted to be made in our general partner’s individual, rather than representative, capacity; meaning that under the terms of our partnership agreement our general partner is entitled to exercise such right free of any fiduciary duty or obligation to any limited partner and it is not required to act in good faith or pursuant to any other standard imposed by our partnership agreement.”).

155

A limited partner thus could not readily determine from the Partnership Agreement who would make the acceptability determination on behalf of the General Partner. The Partnership Agreement is silent and ambiguous.

Lacking guidance, a limited partner might turn to other sources. A logical next step would be to look to the partnership agreement governing the internal affairs of the General Partner, but no one has suggested that any provision in that agreement would be pertinent.

Still lacking guidance, a limited partner might search further. A sophisticated limited partner might realize that the General Partner was itself a limited partnership with the GPGP as its general partner. A diligent limited partner who pressed on might thus end up at a third agreement: the LLC Agreement governing the internal affairs of the GPGP.

The LLC Agreement also does not clearly address what decisionmaker would make the acceptability determination. The LLC Agreement provides generally that, “[e]xcept as otherwise specifically provided in this Agreement, the business and affairs of the Company shall be managed under the direction of the Board.” LLCA § 5.2(a). Section 5.6 creates an exception that gives Holdings “exclusive authority over the business and affairs of the Company that do not relate to management and control of the [Partnership].” Id. § 5.6. The LLC Agreement adds that Holdings “shall have exclusive authority to cause the Company to exercise the rights of the Company and those of the MLP General Partner . . . provided in . . . Section 15.1.” Id. § 5.6(xi) (the “Authority Provision”).

The LLC Agreement thus divides the world of possible decisions into two categories. Unlike in the Partnership Agreement, those two categories do not depend on whether the General Partner is acting in an individual capacity or an official capacity.

156

Rather, the categories in the LLC Agreement divide the world into decisions relating to “the business and affairs of the Company,” where the GPGP Board has authority, and decisions “that do not relate to management and control of the [Partnership],” where Holdings has authority. The LLC Agreement then adds the Authority Provision to confirm that Holdings has authority over the rights provided in Section 15.1 of the Partnership Agreement. That addition suggests that without the Authority Provision it would be unclear whether the decision to exercise the Call Right fell within the purview of the GPGP Board or Holdings. The lack of clarity that would exist without the Authority Provision is also consistent with the fact that whether an action is done in the General Partner’s “individual” or “official” capacity only dictates the applicable standard of review, not which decision- maker makes the decision.

The LLC Agreement also contains a definition of “Opinion of Counsel” that expressly refers to the Sole Member. Unlike the definition of “Opinion of Counsel” that appears in the Partnership Agreement, the definition in the LLC Agreement defines the term as “a written opinion of counsel (which may be regular counsel to the Company or the MLP or any of their respective Affiliates) acceptable to the Sole Member.” LLCA § 1.1 at 7. But the LLC Agreement never uses the term “Opinion of Counsel” in any substantive provision. It is a stray definition.

As this discussion shows, none of the constitutive agreements gives a clear answer as to which entity makes the acceptability determination. Instead, the agreements divvy up decisions into categories, including (i) the difference between determining the acceptability of the Opinion and exercising the Call Right, (ii) the difference between action in an official

157

capacity and action in an individual capacity, and (iii) the difference between decisions that relate to “the business and affairs of the Company” and those “that do not relate to management and control of the [Partnership].” When mixed and matched, the three pairs could generate eight combinatorial outcomes.

2. The Competing Arguments One reasonable reading of the provisions is that Holdings makes the acceptability determination. From a textual perspective, that reading treats the phrase “Opinion of Counsel” as the linguistic version of an equivalency formula, like “X = [the definitional text].” Under this reading, the definitional text is substituted algebraically wherever the “X” appears, such that the full language of the “Opinion of Counsel” definition would be substituted wherever the term “Opinion of Counsel” appears in the Partnership Agreement, including in the Call Right in Section 15.1. The Call Right thus would state that if the General Partner held “more than 50% of the total Limited Partner Interests of all classes then Outstanding” and had received “a written opinion of counsel . . . acceptable to the General Partner” then the General Partner could exercise the Call Right, assuming the Opinion of Counsel satisfied the Opinion Condition. See PA §§ 1.1, 15.1(b). At that point, the argument goes, the Authority Provision in the LLC Agreement specifies that Holdings makes decisions regarding the General Partner’s rights under Section 15.1, so Holdings has the authority to make the decision as the Sole Member. This reading has an added benefit of giving some purpose to the stray definition of “Opinion of Counsel” in the LLC

158

Agreement. Although the definition is never used, it does refer to Holdings making the determination as Sole Member.29 The problem with this analysis is that Holdings always and inherently had the right to determine whether the Opinion is acceptable. Holdings possessed that authority as part of its ability to decide whether or not to exercise the Call Right. If Holdings did not think that the Opinion was acceptable, then Holdings could simply decide not to exercise. Because Holdings always had the ability to make a de facto acceptability determination, assigning the acceptability determination to Holdings renders the Acceptability Condition surplusage. Under standard principles of contract interpretation, a Delaware court generally eschews an interpretation that would result in surplusage. See Sunline Com. Carriers, Inc. v. CITGO Petroleum Corp., 206 A.3d 836, 846 (Del. 2019).

When viewed as a whole, the language of the Partnership Agreement suggests that rather than serving as a redundant condition for the benefit of Holdings, the Acceptability Condition exists to protect the Partnership. Both the Opinion Condition and the

29 Note that the argument in favor of Holdings making the acceptability determination is not advanced by equating (i) the Partnership Agreement’s reference to the General Partner taking action in an official capacity with the LLC Agreement’s reference to the GPGP Board having authority over decisions that relate to “the business and affairs of the Company,” and (ii) the Partnership Agreement’s reference to the General Partner taking action in an individual capacity with the LLC Agreement’s reference to Holdings having authority over decisions “that do not relate to management and control of the [Partnership].” Aligning the categories in that way leads to the conclusion that Holdings exercises the Call Right, which is consistent with the Authority Provision. But that conclusion does not address whether the acceptability determination is part of the exercise of the Call Right.

159

Acceptability Condition ensure that the General Partner cannot exercise the Call Right arbitrarily without satisfying an up-front test. The Opinion Condition establishes the basic hurdle that the General Partner must clear, and the Acceptability Condition ensures that the General Partner cannot obtain a contrived opinion. The Acceptability Condition is thus not a protection for Holdings, which can always protect itself by deciding not to exercise the Call Right. It is instead a protection for the minority partners. In this regard, the Call Right at issue in this case contrasts with a second call right that the General Partner can exercise without satisfying either the Opinion Condition or the Acceptability Condition, as long as the General Partner owns 80% or more of the common units. See PA § 15.1(a). The difference between the two call rights indicates that the Opinion Condition and the Acceptability Condition were intended as meaningful limitations on the General Partner’s ability to exercise the Call Right at the lower ownership level.

Viewed within this structure, the acceptability determination logically belongs to the GPGP Board. Only the GPGP Board has outside directors, and only the GPGP Board can inject a measure of independence into the determination of acceptability. The need for some measure of independence becomes critical for the Call Right, because otherwise the General Partner can exercise that right in its individual capacity, free of any duty or constraint whatsoever. The defendants’ interpretation would make the General Partner the “judge in [its] own cause.” See Dr. Bonham’s Case, 8 Co. Rep. 107a, 114a, 118a, 77 Eng. Rep. 638, 646, 652 (C.P. 1610) (“[O]ne should not be judge in his own cause, indeed it is unjust for one to be a judge of his own matter; and one cannot be Judge and attorney for any of the parties . . . .”).

160

Against this backdrop, the textual arguments for treating the acceptability determination as a decision for Holdings to make as Sole Member are weaker than they initially seem. To reiterate, the distinction between the General Partner acting in an individual capacity as opposed to an official capacity does not shed light on who makes the acceptability determination. That distinction only determines the standard of review that applies to a decision made by the General Partner, not which entity within the General Partner makes the decision. See PA § 7.9(b), (c).

The distinction between the two definitions of “Opinion of Counsel,” one in the Partnership Agreement and the other in the LLC Agreement, also appears in a different light. The fact that the LLC Agreement contains a reference to the Sole Member confirms the obvious: the drafters could have included a similar reference in the Partnership Agreement. The fact that they did not implies that the Partnership Agreement did not intend to confer the authority to make the acceptability determination on the Sole Member. See Int’l Rail P’rs LLC v. Am. Rail P’rs, LLC, 2020 WL 6882105, at *9 (Del. Ch. Nov. 24, 2020) (explaining that evidence of specific language in one agreement but not in a distinct yet related agreement “reflects that the drafters knew how to craft” and include the specific language at issue if they so desired).

Finally, the notion that the acceptability determination becomes part of the exercise of the Call Right also becomes suspect. The Call Right is structured as a conditional option. It first identifies conditions that the General Partner must meet, including receiving an Opinion of Counsel that both addresses the substantive issue identified in the Call Right and does so in an acceptable way. PA § 15.1(b)(ii). The second part of the Call Right

161

provides that if the General Partner satisfies those conditions, “then the General Partner shall then have the right . . . exercisable at its option within 90 days of receipt of such opinion to purchase all, but not less than all, of all Limited Partner Interests then Outstanding held by Persons other than the General Partner and its Affiliates.” Id. (emphasis added). The conditions for exercise must be satisfied before the General Partner can determine whether to exercise it.

As noted previously, the reading that gives Holdings authority over the acceptability determination requires replacing “Opinion of Counsel” with the definitional language for that term. That move does not change the structure of the Call Right. It merely introduces the definitional language into the conditions that must be met before the General Partner can decide whether to exercise the Call Right. It does not change the fact that the condition must be met before the General Partner can act, and it does not address who has authority over evaluating the condition.

The term “Opinion of Counsel” was not drafted specifically for the Call Right. The Partnership Agreement uses it in many substantive provisions. See, e.g., PA §§ 4.6(c), 4.8(b), 7.10(b), 11.1(b), 12.1(a), 12.2(iii), 13.1 (f), 13.3(d), 13.11, 14.3(d),(e). It requires consideration of context to determine who would make the resulting determination. For purposes of the Call Right, the Acceptability Condition remains part of the conditions that must be satisfied before the General Partner can exercise the Call Right. It is not part of the decision to exercise the Call Right. It follows that the General Partner’s authority to exercise the Call Right in its individual capacity does not mean that it can determine acceptability in its individual capacity. For similar reasons, the Authority Provision does

162

not clearly give the Sole Member the ability to make the acceptability determination as part of the “rights . . . of the General Partner . . . provided in . . . Section 15.1.” LLCA § 5.6(xi). The Acceptability Condition is not a right of the General Partner; it is a condition that must be satisfied before the General Partner can exercise its rights.

Ultimately, the path to understand who makes the acceptability determination ends in the marshy distinction that the LLC Agreement makes between an issue that relates to the “business and affairs” of the Partnership, which is conferred to the GPGP Board, and an issue that does “not relate to [the] management and control of the [Partnership],” which is left to the Sole Member. LLCA § 5.6. At first blush, that distinction might seem to track the distinction in the Partnership Agreement between official capacity decisions and individual capacity decisions, but the language is different. To the extent the two concepts do align, there are no textual signals relating to the Acceptability Condition that would suggest that the General Partner makes the acceptability determination in an individual capacity, such that the decision would “not relate to [the] management and control of the [Partnership].”

Instead, the concepts of “business and affairs” and “management and control”

hearken to Section 141(a) of the Delaware General Corporation Law, which establishes the capacious scope of authority possessed by a board of directors. See 8 Del. C. § 141(a). Landmark Delaware Supreme Court cases establish that decisions about whether a public entity’s shares are acquired relate to the business and affairs of the enterprise; a purchase of shares is not exclusively an investor-level transaction between a buyer and seller that

163

falls outside the board’s purview.30 Elsewhere in Section 5.6, the LLC Agreement expressly invokes corporate law principles by stating that “[e]xcept as otherwise specifically provided in this Agreement, the authority and functions of the Board, on the one hand, and the Officers, on the other hand, shall be identical to the authority and functions of the board of directors and officers, respectively, of a corporation organized under the General Corporation Law of the State of Delaware.” LLCA § 5.6. Under corporate law principles, a decision that would affect the success of a take-private transaction would relate to the business and affairs of the corporation and fall within the authority of the board of directors. Even without the backdrop of Delaware corporate law, the exercise of the Call Right would “relate to” the management of the Partnership. If the Call Right cannot be exercised, then the General Partner will continue to manage the Partnership as an MLP with minority investors, making regular public filings with the SEC, complying with listing requirements, and experiencing all of the other costs and benefits of public status. If the Call Right is exercised, then the Partnership will no longer be an MLP, and the General Partner can manage the Partnership’s affairs solely in the interest of Loews and without the accoutrements of public status. Making the acceptability determination therefore “relate[s] to [the] management and control of the [Partnership].”

30 See, e.g., Moran v. Household Int’l, Inc., 500 A.2d 1346, 1353 (Del. 1985) (“[W]e note the inherent powers of the [b]oard conferred by 8 Del. C. § 141(a), concerning the management of the corporation’s ‘business and affairs’ . . . also provides the [b]oard additional authority upon which to enact the [r]ights [p]lan.” (emphasis removed) (citing Unocal Corp. v. Mesa Petroleum Co., 493 A.2d 946, 953 (Del. 1985))).

164

There is thus a reasonable reading of the pertinent agreements under which the GPGP Board has the authority to make the acceptability determination. Recognizing the potential merit in that argument, Loews initially intended to have the GPGP Board make the acceptability determination. But the outside directors had a “hostile reaction,” and they asked “shouldn’t we have independent counsel[?]” JX 874 at 5; see Layne Dep. 160. The outside directors recognized the importance of the acceptability determination, and they did not want to be treated as a speedbump on Loews’ path to the take-private. The outside directors’ reaction shows why the Acceptability Condition exists, viz., it could provide an external check.31 3. Counsel’s Contemporaneous Recognitions Of Ambiguity Contrary to the defendants’ assertions, all of the lawyers acknowledged the ambiguity that Loews created for the acceptability determination by establishing Boardwalk’s complex entity structure. Within Skadden, Voss conducted the most thorough and detailed analysis. After reasoning through the various issues, she expressed the view that “the MLP Agreement likely requires that the [GPGP] Board make the determination to accept the Opinion of Counsel. Or, at a minimum, it is ambiguous.” JX 747 at 1.

Skadden later prepared a memorandum for Alpert that framed the analysis more conservatively and with additional caveats and qualifications. The memorandum

31 At trial, two defense witnesses disputed whether the outside directors had a “hostile” reaction. McMahon Tr. 535; Siegel Tr. 738. It is not clear why the witnesses quibbled over this point. They agreed that the outside directors were uncomfortable with the determination and did not want to be involved. McMahon Tr. 535; Siegel Tr. 738.

165

nevertheless made clear that there were ambiguities surrounding the acceptability determination. See JX 773 at 1, 3. And when advising Holdings about whether it could accept the Opinion, Skadden would say only that it was reasonable for Holdings to conclude that it had the authority to make the acceptability determination. See JX 1508 at 3. Even during his deposition, the farthest that Grossman would go in favor of the defendants’ current view is that “the better reading” was for the GPGP Board to make the decision. Grossman Dep. 70–71.

Richards Layton also saw both sides of the interpretive coin. In contrast to Skadden’s more detailed analysis, Richards Layton gave advice orally on a twenty-four hour turnaround, and without knowing that Loews had already received advice from Skadden and contacted the members of the GPGP Board about making the acceptability determination. In the initial call with Alpert, Richards Layton went beyond Grossman by an adverb, saying it was the “far better view” that Holdings could make the acceptability determination. Raju Tr. 808, 842. Only after receiving Richards Layton’s oral advice did Alpert tell Richards Layton about Skadden’s view. No one told Richards Layton about Loews’ outreach to the GPGP Board until this litigation.

After receiving Richards Layton’s oral advice, Alpert asked the firm to memorialize its advice in an email. JX 1225 at 1. The email backed away from the oral advice by removing the adverb, stating: “While there is some ambiguity and arguments can certainly be made to the contrary, we think that the better view is that the [acceptability determination] is within the sole authority of the Sole Member [Holdings] pursuant to

166

Section 5.6 of the LLC Agreement.” Id. at 2–3 (emphasis added). The email included the following caveat:

[I]f the Board of Directors is approached and declines to determine that the Opinion of Counsel is acceptable and the Section 15.1(b) call right is exercised by the Sole Member anyway, that would be a difficult fact to overcome in any future litigation regarding the exercise of the Section 15.1(b) call right.

Id. at 3 (emphasis added). Richards Layton did not know that the GPGP Board had been approached already about making the decision. See Raju Tr. 843. At Alpert’s request, Richards Layton later revised its email to restore the adverb, but it kept the caveats. See JX 1265 at 4.

Even Baker Botts never opined explicitly that the plain language of the Partnership Agreement and the LLC Agreement made clear that Holdings made the acceptability determination. In its initial advice to Alpert, Baker Botts wrote that “[i]t seems that determination of the acceptability of an opinion of counsel in the context of Section 15.1(b) should be made by the Sole Member as opposed to the board of directors of the General Partner.” JX 686 at 4 (emphasis added). After obtaining advice from Skadden and Richards Layton, Baker Botts still only would go so far as to describe that as the “better view,” while noting that “arguments can be made to the contrary.” JX 1508 at 40.

The lawyer who asserted most strongly that the Partnership Agreement gave the General Partner the authority to make the acceptability determination was Layne. He never prepared any written analysis, and he seems originally to have credited the argument that

167

the GPGP Board would determine acceptability.32 After the outside directors on the GPGP Board expressed their displeasure about being involved in the acceptability determination, Alpert tapped Layne to explain why they no longer had to address the issue. At that point, Layne seems to have lumped together the issue of the authority to exercise the Call Right with the issue of the authority to determine acceptability.33 The vacillation in Layne’s views is also consistent with the ambiguity inherent in the Acceptability Condition.

4. Ambiguity Means The GPGP Board Had To Make The Acceptability Determination.

Because the question of who could make the acceptability determination was ambiguous, well-settled interpretive principles require that the court construe the agreement in favor of the limited partners. See Norton, 67 A.3d at 360. Under the interpretation that favors the limited partners, the GPGP Board had the authority to make the acceptability determination. Because the GPGP Board did not make the acceptability determination, the General Partner breached the Partnership Agreement by exercising the Call Right.

32 When reviewing a draft of a memorandum from Richards Layton which explained that Section 5.6 “specifies that the Sole Member has exclusive authority to cause GP LLC to exercise the rights of GP LLC,” Layne commented, “but not to determine applicability.” JX 1810 at 3. Next to another sentence that stated that Holdings decided whether the Opinion of Counsel was acceptable “pursuant to Section 5.6 of the LLC Agreement because the determination to accept the Opinion of Counsel is a part of Section 15.1 of the Partnership Agreement,” Layne wrote “not exercise.” Id.

33 See JX 1325; JX 1331 at 2; JX 1343; JX 1435 at 1, 3; JX 1812.

168

D. Contractual Immunity To Damages The defendants maintain that even if the General Partner breached the Partnership Agreement and otherwise would be responsible for damages, the plaintiffs cannot recover because the defendants immunized themselves contractually against any damages award. There are two relevant provisions in the Partnership Agreement. The first is a true exculpation provision. The second is a provision that establishes a conclusive presumption of good faith if the General Partner or another decision-maker relies on an advisor. The General Partner cannot rely on either of them to escape liability in this case.

1. The Exculpation Provision Section 17-1101(f) of the Delaware Revised Uniform Limited Partnership Act authorizes a partnership agreement to eliminate “any and all liabilities for breach of contract and breach of duties (including fiduciary duties) of a partner or other person to a limited partnership or to another partner or to another person that is a party to or is otherwise bound by a partnership agreement,” other than “any act or omission that constitutes a bad faith violation of the implied contractual covenant of good faith and fair dealing.” 6 Del. C. § 17-1101(f).

The Partnership Agreement takes full advantage of this statutory authority. Section 7.8(a) states:

Notwithstanding anything to the contrary set forth in this Agreement, no Indemnitee shall be liable for monetary damages to the Partnership [or] the Limited Partners . . . for losses sustained or liabilities incurred as a result of any act or omission of an Indemnitee unless there has been a final and nonappealable judgment entered by a court of competent jurisdiction determining that, in respect of the matter in question, the Indemnitee acted in bad faith or engaged in fraud, [or] willful misconduct . . . .

169

PA § 7.8(a). The Partnership Agreement defines “Indemnitee” to include the “General Partner,” “any Person who is or was an Affiliate of the General Partner,” and “any Person who is or was a member, partner, director, officer, fiduciary or trustee of . . . the General Partner or any Affiliate of . . . the General Partner.” Id. § 1.1 at 19.

Under this provision, to recover damages from the General Partner, the plaintiff must prove that the General Partner “acted in bad faith or engaged in fraud [or] willful misconduct.” Id. § 7.8(a). The Partnership Agreement does not define these terms. Under Delaware law, however, all three require a showing of scienter.

The exception for willful misconduct best fits the facts of this case. That term requires a showing of “intentional wrongdoing, not mere negligence, gross negligence or recklessness.” Dieckman v. Regency GP LP, 2021 WL 537325, at *31 (Del. Ch. Feb. 15, 2021) (quoting 12 Del. C. § 3301(g)), aff’d per curiam, No. 92, 2021, slip op. (Del. Nov. 3, 2021); see Willful Misconduct, BLACK’S LAW DICTIONARY (11th ed. 2019) (“Misconduct committed voluntarily and intentionally.”). The concept of misconduct involves “unlawful, dishonest, or improper behavior, esp. by someone in a position of authority or trust.” Misconduct, BLACK’S LAW DICTIONARY (11th ed. 2019).

While serving as a member of this court, Chief Justice Strine described two situations that could support a finding of willful misconduct. See Gotham P’rs, L.P. v. Hallwood Realty P’rs, L.P., 2000 WL 1476663 (Del. Ch. Sept. 27, 2000). A limited partner of an MLP asserted that the general partner “designed” a series of transactions “to entrench its owner [Hallwood Group Incorporated (“HGI”)], by placing a large number of [partnership] units in HGI’s hands at an unfairly low price.” Id. at *3. The limited partner

170

also asserted that the general partner “timed the [t]ransactions so as to enable HGI to grab up a control block at a depressed price.” Id. Chief Justice Strine held on a motion for summary judgment that the plaintiffs could not prove a claim for fraud, but that the ruling did not eliminate the possibility that the plaintiffs could prove willful misconduct. Possible scenarios included if the general partner or its affiliates

(i) purposely misled the [independent directors] about (a) the underlying value of the [p]artnership units or (b) the ability of the [p]artnership to get a higher price for the units than HGI was willing to pay, (ii) in order to induce the [independent directors] to approve a sale to HGI at an unfair price.

Id. at *14. Another possible scenario that would provide evidence of willful misconduct involved the general partner having “a secret plan to snatch up a large number of units that could entrench it at a bargain price before an expected up-turn in the market and did not disclose that plan to the [independent directors].” Id.

Striving to limit the conceptual space available for a finding of willful misconduct, the defendants argue that the court must (i) focus on the three individuals who comprised the Holdings board (Siegel, Keegan, and Wang), (ii) examine their individual states of mind when deciding to exercise the Call Right, and (iii) deny any recovery to the class unless all three acted with scienter. The defendants would have the court ignore all of the other actors in the drama and all of the events leading up to the decision to exercise the Call Right.

If the court were deciding whether to hold Siegel, Keegan, or Wang personally liable for their decision to exercise the Call Right, such as under a tortious interference theory,

171

then that mode of analysis might be warranted. But the plaintiffs are seeking to recover damages from the General Partner, not those three individuals.

“A basic tenet of corporate law, derived from principles of agency law, is that the knowledge and actions of the corporation’s officers and directors, acting within the scope of their authority, are imputed to the corporation itself.” Stewart v. Wilm. Tr. SP Servs., Inc., 112 A.3d 271, 302–03 (Del. Ch. 2015), aff’d, 126 A.3d 1115 (Del. 2015); see Teachers’ Ret. Sys. of La. v. Aidinoff, 900 A.2d 654, 671 n.23 (Del. Ch. 2006); Restatement (Third) of Agency § 5.03 Westlaw (Am. L. Inst. database updated Oct. 2021). That principle extends to alternative entities like the General Partner. See CompoSecure, L.L.C. v. CardUX, LLC, 206 A.3d 807, 823–24 (Del. 2018). “An entity … can only make decisions or take actions through the individuals who govern or manage it.” Dieckman, 2021 WL 537325, at *36 (quoting Gerber v. EPE Hldgs., LLC, 2013 WL 209658, at *13 (Del. Ch. Jan. 18, 2013) (omission in original)).

During the relevant period, numerous individuals acted on behalf of the General Partner in a manner sufficient to impute scienter to the General Partner. During the relevant period, Alpert, Siegel, McMahon and Johnson were management-level officers and agents of Loews, Holdings, the GPGP, the General Partner, and Boardwalk. Their actions and intent were imputed to the General Partner. Together, those individuals orchestrated the sham Opinion, supported the sham Opinion with the inadequate Rate Model Analysis, and diverted the acceptability determination for the sham Opinion from the GPGP Board to Holdings.

172

In addition, Baker Botts acted as counsel to the General Partner in rendering the Opinion. A lawyer acts as an agent for its client, and the lawyer’s knowledge is imputed to the client for matters within the scope of the lawyer’s agency. Vance v. Irwin, 619 A.2d 1163, 1165 (Del. 1993). Ordinarily, an issue would exist about whether to impute an attorney’s knowledge to the client when the attorney did not act in good faith. Here, however, the General Partner wanted Baker Botts to render the Opinion and pushed for the outcome that Baker Botts reached. Under the circumstances, Baker Botts’ scienter in issuing the Opinion can be attributed to the General Partner.

The General Partner engaged in “intentional wrongdoing . . . designed to . . . seek an unconscionable advantage.” Dieckman, 2021 WL 537325, at *36 (quoting 12 Del. C. § 3301(g)). The General Partner and Baker Botts pasted together an Opinion intended to achieve the goal of enabling the General Partner to exercise the Call Right. That conduct is sufficient to render the exculpatory provision inapplicable.34

34 The parties have not addressed who has the burden to prove that the exculpatory provision applies. Authorities demonstrate persuasively that the General Partner should bear this burden. In the analogous context of corporate law exculpation, the director defendants must prove that they fall within the exculpatory provision’s protections. See Emerald P’rs v. Berlin, 726 A.2d 1215, 1223–24 (Del. 1999). For purposes of a breach of contract claim, the exculpatory provision operates as an exception to normal principles of contract liability. As a matter of hornbook law, “[a] party seeking to take advantage of an exception to a contract is charged with the burden of proving facts necessary to come within the exception.” 29 Am. Jur. 2d Evidence § 173, Westlaw (database updated Aug. 2021). This decision has nevertheless analyzed the question of scienter as if the plaintiffs bore the burden of proof.

173

2. The Conclusive Presumption Section 17-407(c) of the Delaware Revised Uniform Limited Partnership Act states that a general partner

shall be fully protected from liability to the limited partnership, its partners or other persons party to or otherwise bound by the partnership agreement in relying in good faith upon . . . opinions, reports or statements presented . . .

by any . . . person as to matters the general partner reasonably believes are within such . . . person’s professional or expert competence . . . .

6 Del. C. § 17-407(c).

The Partnership Agreement supercharges this statutory concept by providing as follows:

The General Partner may consult with legal counsel . . . and any act taken or omitted to be taken in reliance upon the advice or opinion (including an Opinion of Counsel) of such [counsel] . . . shall be conclusively presumed to have been done or omitted in good faith and in accordance with such advice or opinion.

PA § 7.10(b) (the “Reliance Provision”).

The General Partner cannot invoke the Reliance Provision when it knows that the opinion in question was contrived to generate a result. Under those circumstances, the General Partner is not relying on the contrived opinion. The opinion is window dressing to enable the General Partner to take action.

That reality prevents the General Partner from relying on the Opinion for purposes of the Reliance Provision. The General Partner not only knew the Opinion was contrived, but the General Partner’s representatives participated actively in the manufacturing of the Opinion.

174

The General Partner also cannot rely on Skadden’s advice about the acceptability of the Opinion. As a threshold matter, it is not clear that the Reliance Provision envisions opinions like Matryoshka dolls, in which counsel renders an opinion, then another counsel opines on the opinion, and so on, with the breadth of protection expanding at each level. If anything, the procuring of a second opinion can be a tell, implying inadequacies or taints in the original opinion. Boards often retain a second investment banker when they learn that their chosen banker has a conflict of interest that could render its advice suspect. At least in that setting, the second banker addresses the core issue. Here, Skadden refused as a matter of firm policy to opine on the core issue and instead provided an opinion about an opinion.

Regardless, the Reliance Provision only protects the General Partner when it actually relies on the underlying opinion, not when it manufactures the opinion and then gets another opinion to whitewash the first one. No matter what Skadden said about the Opinion, the General Partner knew how the Opinion came about, including that it addressed hypothetical maximum rates in a setting where the regulatory changes were not yet final and were unlikely to have any meaningful real-world effect. Under those circumstances, the General Partner cannot invoke the Reliance Provision.

Finally, the General Partner cannot invoke the Reliance Provision for purposes of the Acceptability Condition because the wrong decisionmaker considered the issue. The General Partner knew about the ambiguity surrounding the acceptability condition. The General Partner opted for the decisionmaker more favorable to its interests rather than the decisionmaker more favorable to the interests of the limited partners. With the wrong

175

decisionmaker having acted, the General Partner cannot claim to have relied validly on Skadden’s advice. E. Damages Having found that the General Partner breached the Partnership Agreement, and having concluded that the General Partner can be held liable for damages, the next step is to determine whether the plaintiffs suffered damages, and if so, the amount of a damages award. The plaintiffs proved that by exercising the Call Right in breach of the Partnership Agreement, the General Partner inflicted damages on the class of $689,827,343.38. Plaintiffs are entitled to pre- and post-judgment interest on that amount. As the prevailing party, the plaintiffs are also entitled to an award of fees.

[T]he standard remedy for breach of contract is based upon the reasonable expectation of the parties ex ante. This principle of expectation damages is measured by the amount of money that would put the promisee in the same position as if the promisor had performed the contract.

Expectation damages thus require the breaching promisor to compensate the promisee for the promisee’s reasonable expectation of the value of the breached contract, and, hence, what the promisee lost.

Duncan v. Theratx, Inc., 775 A.2d 1019, 1022 (Del. 2001).

An injured party “need not establish the amount of damages with precise certainty where the ‘wrong has been proven and injury established.’” Siga Techs., Inc. v. PharmAthene, Inc., 132 A.3d 1108, 1131 (Del. 2015) (quoting Del. Express Shuttle, Inc. v. Older, 2002 WL 31458243, at *15 (Del. Ch. Oct. 23, 2002)). “[D]oubts about the extent of damages are generally resolved against the breaching party.” Id. at 1131. “Public policy has led Delaware courts to show a general willingness to make a wrongdoer ‘bear the risk of uncertainty of a damages calculation where the calculation cannot be mathematically

176

proven.’” Beard Rsch., Inc. v. Kates, 8 A.3d 573, 613 (Del. Ch. 2010) (quoting Great Am. Opportunities, Inc. v. Cherrydale Fundraising, LLC, 2010 WL 338219, at *23 (Del. Ch. Jan. 29, 2010) (collecting cases)). That said, expectation damages “should not act as a windfall.” Paul v. Deloitte & Touche, LLP, 974 A.2d 140, 146 (Del. 2009).

The plaintiffs proved that the General Partner breached the Partnership Agreement by exercising the Call Right without meeting the necessary conditions. By exercising the Call Right improperly, the General Partner deprived the plaintiffs of the stream of distributions that they otherwise would have received as unitholders. The appropriate measure of damages is therefore the difference between the present value of those future distributions and the transaction price. The transaction price is undisputed. The General Partner paid $12.06 per unit when it exercised the Call Right. Unsurprisingly, the parties dispute the present value of the future distributions, and they presented drastically different estimates to the court.

To make their respective cases, both sides presented damages experts. J.T. Atkins submitted a report and testified on behalf of the plaintiffs. Atkins has been involved in numerous M&A financing and restructuring transactions in the energy and MLP sectors, and has acted as an expert witness in thirteen separate litigations involving energy companies or MLPs. Atkins Tr. 1018. R. Glenn Hubbard submitted a report and testified on behalf of the defendants. Hubbard is a professor at Columbia University’s business school and has testified as an expert before this court on matters of valuation on numerous occasions. JX 1745 (Hubbard Report) ¶¶ 2, 5.

177

Atkins measured damages using a discounted distribution model (a “Distribution Model”). He calculated the fair value of the units to be $17.84 at the low end and $19.30 at the high end, resulting in a range of damages from $720 million to $901.6 million. JX 1761 (Atkins Rebuttal Report) ¶ 2(d).

Hubbard also prepared a Distribution Model, but he discarded it in favor of a valuation based on the market price of Boardwalk’s units. Using his market price metric, Hubbard opined that the fair value of the units was $10.74 per unit. Hubbard Report ¶ 9. Because that value was less than the Call Right exercise price, he concluded that the plaintiffs suffered no damages.

Hubbard’s approach was not persuasive. This decision uses Atkins’ model with one modification.

1. Hubbard’s Approach After considering several valuation indicators, Hubbard opined that the best evidence of the value of the units was their unaffected market price. In reaching this conclusion, Hubbard examined various jurisprudential indicators of market efficiency and concluded that when applied to Boardwalk’s units, those indicators were “generally consistent with . . . trading in an efficient market.” Hubbard Report ¶ 71.

To derive a measure of damages based on the unaffected market price, Hubbard could not simply use the market price on the date of the Call Right, because the Potential Exercise Disclosures and the self-referential mechanic in the Purchase Price calculation drove the market price downward. To derive an unaffected market price, Hubbard started with the market price on the last trading day before the issuance of the Potential Exercise

178

Disclosures, then used a regression analysis to bring the market price forward to the date on which Loews exercised the Call Right. See id. ¶ 89. Based on this analysis, Hubbard concluded that the unaffected market price of the units would have been lower than the Purchase Price. He therefore opined that the limited partners did not suffer any damages. Id. ¶ 9.

Hubbard’s analysis is not persuasive because he failed to account for the General Partner’s control over the Partnership and the resulting valuation overhang. A market for a company’s shares “is more likely efficient, or semi-strong efficient, if it has . . . no controlling stockholder.”35 Conversely, a market for a company’s shares is less likely to be efficient if it has a controlling stockholder. The presence of a controlling stockholder matters because “participants will perceive the possibility that the controller will act in its own interests and discount the minority shares accordingly.” In re Appraisal of Regal Ent. Gp., 2021 WL 1916364, at *26 (Del. Ch. May 13, 2021) (emphasis removed) (declining to rely on unaffected trading price given the presence of a controlling stockholder); accord Glob. GT v. Golden Telecom, Inc., 993 A.2d 497, 503, 508–09 (Del. Ch. 2010), aff’d, 11 A.3d 214 (Del. 2010). It is undisputed that Loews controlled the Partnership through the General Partner. Hubbard’s starting point—the supposedly unaffected market price on the

35 Dell, Inc. v. Magnetar Glob. Event Driven Master Fund Ltd., 177 A.3d 1, 25 (Del.

2017); In re Appraisal of Stillwater Mining Co., 2019 WL 3943851, at *51 n.22 (Del. Ch. Aug. 21, 2019) (collecting research supporting the reliability of unaffected trading price in absence of controlling stockholder), aff’d sub nom. Brigade Leveraged Cap. Structures Fund Ltd. v. Stillwater Mining Co., 240 A.3d 3 (Del. 2020).

179

last trading date before the issuance of the Potential Exercise Disclosures—was thus not a reliable estimate of fair value.

Hubbard’s analysis also failed to account for the fact that the market did not possess material information about the level of distributions that Boardwalk could make in the future. “Under the semi-strong form of the efficient capital markets hypothesis, the unaffected market price is not assumed to factor in nonpublic information.” Verition P’rs Master Fund Ltd. v. Aruba Networks, Inc., 210 A.3d 128, 140 (Del. 2019). Consequently, it is inappropriate to rely on the unaffected trading price as a measure of value when there is “material, nonpublic information” which “could not have been baked into the public trading price.” Id. at 139.

In this case, Loews projected internally that the Partnership’s distributions would quadruple in 2023. See JX 1529, “Side Model” tab. Because Loews controlled the Partnership, Loews had the ability to make that happen. The market was not aware of Loews’ internal projections, and the unaffected trading price of the units could not and did not reflect this information. See Dell, 177 A.3d at 25–26 (explaining that “valuation gaps” can occur when “information fail[s] to flow freely or . . . management purposefully temper[s] investors’ expectations for the [c]ompany so that it [can] eventually take over the [c]ompany at a fire-sale price”). By relying on the unaffected trading price, Hubbard’s approach failed to take into account this source of value.

Hubbard’s analysis of the trading price does not provide a reliable damages estimate. This decision therefore declines to use it.

180

2. Atkins’ Approach Atkins provided a damages estimate using a Distribution Model. That methodology is a variant of a discounted cash flow analysis, but instead of discounting future cash flows at the entity level, the Distribution Model discounts the value of expected future distributions at the investor level. Because the Distribution Model only looks at returns to the equity, the discount rate is the company’s cost of equity capital. Atkins Tr. 1022. As Hubbard acknowledged, a Distribution Model is a “customary” method for valuing units in an MLP.36 The principal inputs to a Distribution Model are cash flow projections, the company’s cost of equity capital, and a terminal growth rate. Atkins Tr. 1025–26. The defendants do not dispute Atkins’ cost of equity capital or his terminal growth rate. In both cases, Atkins used more conservative figures than Hubbard used in his competing Distribution Model. See Hubbard Tr. 1195.

36 Hubbard Tr. 1194; see JX 397 at 15 (industry analyst white paper stating that “[t]he methodology we prefer [for valuing MLPs] is the distribution discount model”); JX 423 at 85 (industry analyst white paper stating that “[o]ur primary tool for valuing MLPs is a three-stage distribution (dividend) discount model”); JX 429 at 3 (analyst report valuing the Partnership using a Distribution Model); JX 431 at 10 (same); JX 523 at 4 (same); JX 1223 at 8 (same); see also JX 451 at 29 (analyst white paper using the same methodology but calling it a “Dividend Discount Model”). Hubbard prepared his own Distribution Model to “corroborat[e]” his damages estimate. Hubbard Report ¶¶ 150–51, 155, Ex. 32A.

181

The defendants focused their attack on the cash flow projections that Atkins used.

Thus, the central question is whether the cash flow projections were sufficiently reliable to use for valuation purposes.

“When evaluating the suitability of projections, Delaware cases express a strong preference for management projections prepared in the ordinary course of business and available as of the date of the [transaction].” Regal Ent. Gp., 2021 WL 1916364, at *21 & n.17 (collecting cases). “[L]itigation-driven projections” are less likely to be reliable and therefore are disfavored. Gray v. Cytokine Pharmasciences, Inc., 2002 WL 853549, at *8 (Del. Ch. Apr. 25, 2002). Relying on ex post, litigation-driven projections creates an “untenably high” risk of “hindsight bias and other cognitive distortions.” Agranoff v. Miller, 791 A.2d 880, 891–92 (Del. Ch. 2000); accord Owen v. Cannon, 2015 WL 3819204, at *22 (Del. Ch. June 17, 2015) (finding that “the after-the-fact projections . . . created for purposes of this litigation are tainted by hindsight bias and are not a reliable source to determine the fair value of [the] shares” (footnotes omitted)).

Both experts relied on a model that the Loews management team prepared (the “Loews Model”). The Loews Model started from a five-year plan that Boardwalk’s management team created in the ordinary course of business. Siegel Dep. 115; see Siegel Tr. 754–55. The Loews management team then extended the five-year plan to the year 2029. In the course of assisting Loews senior executives in determining whether to exercise the Call Right, the Loews management team modified and refined their model many times. See, e.g., JX 767; JX 881; JX 1485; JX 1529.

182

Atkins used version ninety-one of the Loews Model. That version was the last one that the Loews management team prepared before the Loews board of directors met on June 29, 2018, and decided to cause the General Partner to exercise the Call Right. See JX 1529. Hubbard used version ninety of the Loews Model, which was the immediately preceding version. See JX 1485. The two versions are virtually identical, and both project the same amount of distributions. Compare JX 1485, “Side model” tab, Row 20, with JX 1529, “Side model” tab, Row 20.

Both experts agreed that the Loews Model was an appropriate starting point for a Distribution Model. The court concurs. The Loews Model started from a five-year plan prepared in the ordinary course of business, and the Loews management team refined it so it could be used in real time to make a $1.5 billion dollar investment. The projections were not created for litigation, nor is there any other reason to doubt their accuracy.

Both experts nonetheless made adjustments to the Loews Model. Hubbard made multiple modifications to the cash flow projections. Atkins kept the cash flow projections in the Loews Model, but he eliminated a reduction in EBITDA from the forecast. This decision declines to adopt any of the adjustments and uses the Loews Model in its original form.

a. Hubbard’s Adjustments To The Loews Model For purposes of his Distribution Model, Hubbard arbitrarily removed the projections for 2028 and 2029 from the Loews Model. See Hubbard Report Ex. 25. By doing so, Hubbard shortened the projection period and changed the cash flows for the terminal period. See id. Ex. 32A. Hubbard did not provide a persuasive explanation for this change.

183

Hubbard was serving as a litigation expert, and he lacked prior experience with MLPs in general and Boardwalk’s business in particular. There is no reason to believe that Hubbard had a better understanding of Boardwalk’s prospects than the Loews management team.

Hubbard also eliminated the distributions in the out-years of the Loews Model.

Hubbard claimed that he reduced the projections “so that the forecasts for the terminal period would reflect a more realistic and sustainable steady state.” JX 1759 (Hubbard Rebuttal Report) ¶ 10. That explanation was conclusory and unpersuasive.

In addition, Hubbard progressively increased the projected capital expenditures for the years 2023–2027. Compare JX 1529, “Side model” tab, with Hubbard Report Ex. 25. Hubbard allocated all capital expenditures to maintenance capital, which reduced the projected distributions during those years. See Hubbard Report ¶ 114; Atkins Rebuttal Report ¶ 26. By the year 2027, Hubbard’s approach resulted in more than double the expenditures of maintenance capital than the Loews management team had projected. See Atkins Rebuttal Report ¶ 26 tbl. 1. That was neither reasonable nor persuasive.

Hubbard’s modifications to the Loews Model caused distributions to decline over time. Hubbard Report Ex. 32A. The high point for distributable cash flow in Hubbard’s model was 2022, the last year before Hubbard’s modifications kicked in. See id. After that, the value of the distributions declined steadily. Atkins explained persuasively that such a result was counterintuitive, both in terms of the underlying business and given Loews’ decision to exercise the Call Right:

[I]nstead of having the normal projections where you have a slow and steady growth in your distributions, [Hubbard’s] assumptions . . . push distributions downward. Why would Loews. . . not just sell the business, get out of this

184

business, if it really believed that [the] distributions would decline as opposed to go up over time?

Atkins Tr. 1057.

Hubbard made these adjustments based on an interview with two Loews executives.

Hubbard Report ¶ 106 n.161. Hubbard claimed that the executives told him that, “Loews focused mostly on the period 2018 through 2022 and [that] their assumptions for 2023 through 2029 were vetted less rigorously.” Hubbard Report ¶ 106 n.161. The executives’ account was self-serving, and the defendants could not produce any documents to support it. See JX 1752. The defendants also did not call either executive at trial to support Hubbard’s assertion. Instead, they called Siegel, who knew next to nothing about the Loews Model.37

37 See Siegel Tr. 755 (“Q: By April 4th your team was up to Version 25 of the model;

right? A: I don’t know.”); id. at 756–57 (“Q: By April 9th, your team had built a switch into the model; correct?” A: I don’t know. Q: You could toggle the switch from base FERC impact to downside FERC impact or to off; correct? A: Don’t know. . . . I never studied the actual model itself and how it was put together, so I can’t comment. Q: If the switch was toggled to downside FERC impact, the model would show a hit to EBITDA from the refund to ADIT from the customers; correct? A: I don’t know. Q: If the switch was off, the model would show no hit to EBITDA; correct? A: I don’t know. Q: On April 9th, your team was at Version 39 of the Loews’ [sic] model; correct? A: Don’t know.”); id. at 758 (“Q: By that point, the model was up to Version 43; correct? A: Again, I don’t know.”); id. at 761–62 (“Q: Barclays gave input to Ms. Wang about the model; correct? A: I don’t know.”); id. at 763 (“Q: First of all, [the Loews Model] initially went out ten years; correct? A: I don’t know. . . . Q: Version 43 of the model goes out 12 years; isn’t that right? A: I have no recollection of seeing that model or many of the models you’ve referred to.”); id. at 764 (Q: “Isn’t it true that the incentive distribution rights kick in in years 11 and 12 of the Loews’ [sic] model? A. I don’t know. I’m not sure I’ve seen the model. Q. That’s why the model goes out 12 years; right, Mr. Siegel? A. I don’t know.”); id. at 765 (“Q: Isn’t it true that there are 91 versions of this model, Mr. Siegel? A. I have no idea.”); id. at 766 (“Q: Isn’t it true that Version 91 of the model was used to prepare the June 29th Loews’ [sic] board deck? A. I don’t know. Q. Isn’t it true that the inputs or the pages of the Loews’

185

This court has rejected expert opinions when the experts downsized management projections for purposes of litigation. While serving as a member of this court, Chief Justice Strine rejected an expert’s opinion that was based “on a substantial negative revision of . . . projections that he came up with after discussions with [the company’s] managers after the valuation date.” Agranoff, 791 A.2d at 891. A party seeking to vary from reliable projections must “proffer legitimate reasons to vary from the projections.” Prescott Gp. Small Cap, L.P. v. Coleman Co., Inc., 2004 WL 2059515, at *21 (Del. Ch. Sept. 8, 2004) (internal quotation marks omitted). To proffer legitimate reasons, a party must offer more than just “reliance on management’s off-the-record denigrations of its own projections.” Id. “Any other result would condone allowing a company’s management or board of directors to disavow their own data in order to justify a lower valuation . . . .” Gray, 2002 WL 853549, at *8. The same reasoning supports rejecting Hubbard’s modifications to the Loews Model.

This court likewise has rejected a valuation opinion when the expert increased capital expenditures without good reason, thereby reducing cash flows. See In re Emerging Commc’ns, Inc. S’holders Litig., 2004 WL 1305745, at *15 (Del. Ch. May 3, 2004). Hubbard did the same thing. As Atkins explained, Hubbard’s changes were inconsistent with “Boardwalk’s actual operational history.” Atkins Rebuttal Report ¶ 27. Maintenance

[sic] June 29th board deck come directly from Version 91 of the model? A. I don’t know. I’m not sure I’ve seen Version 91 of the model. Q. Isn’t it true that your expert in this case uses Version 90 of the model? A: Again, I don’t know.”).

186

capital expenditures for pipelines are “normally significantly less than depreciation,” and Boardwalk’s “maintenance capital expenditures were on average 39.3% of depreciation expense.” Atkins Rebuttal Report ¶¶ 28–29 (quoting Credit Suisse, CS MLP Primer – Part Deux 14 (Nov. 23, 2011)). Hubbard projected that maintenance capital expenditures would increase to 61.7% of depreciation by the terminal year of his Distribution Model. Atkins Rebuttal Report ¶ 29; see id. Ex. B. at 46. That percentage exceeded Boardwalk’s historical levels and the levels at eleven of twelve comparable MLPs. Id. ¶ 29 tbl. 2.

b. Atkins’ Adjustment To The Loews Model Atkins made one modification to the Loews Model. The Loews management team included a “switch” in the Loews Model labeled “FERC Impact,” which enabled a user to toggle between three possible scenarios: “Base FERC Impact,” “Downside FERC Impact,” and “Off,” meaning no FERC impact (the “FERC Switch”) The first two options—Base FERC Impact and Downside FERC Impact—reflected Loews management’s assessment of the potential implications of the March 15 FERC Actions. Johnson Tr. 636. The model built on FERC’s proposed Form 501(g), which instructed MLPs to submit cost-of-service information using an indicative ROE of 10.55%. Because FERC had singled out that figure, the Loews management team was concerned that FERC could use it as a trigger for pursuing a rate case.

Even using these assumptions, Gulf South and Gulf Crossing did not face any risk of a rate case. Texas Gas faced some risk. The Loews management team projected that if Texas Gas filed its Form 501(g) and presented its cost-of-service calculations using the indicative ROE, no income tax allowance, and ADIT amortized using the Reverse South

187

Georgia method, then Texas Gas would show an ROE of 24.3%, which was within the range of ROEs that historically had triggered rate cases. See JX 1071 at 1, 3; accord Wagner Tr. 247. If FERC initiated a rate case and mandated an adjustment in the rates that Texas Gas could charge based on an ROE of 10.55%, then the Loews Model calculated that Texas Gas would face a revenue reduction of $73.9 million per year. See Johnson Tr. 636. The “Base FERC Impact” scenario therefore deducted $73.9 million from Boardwalk’s EBITDA for every year of the discrete projection period, beginning in 2019. See JX 1485, “Side Model” tab, Row 11. Turning the FERC Switch to “Off” removed the negative impact.

Projecting a rate case for Texas Gas based on these assumptions reflected the conservativism that went into the Loews Model. Wagner, the internal FERC expert on the Baker Botts team, believed that there was “a low probability that Texas Gas would face a section 5 case in the next 1–2 years.” JX 1071 at 1. Although an ROE of 24.3% was “the type of return that has caused FERC to initiate a section 5 case” in the past, Wagner believed that FERC’s existing workload, in addition to the influx of Form 501-G filings, made it likely that FERC would “probably be somewhat swamped and not able to begin those investigations.” Wagner Tr. 245; see JX 1071 at 1. Beyond two years, there were “too many variables to make a prediction with any confidence.” JX 1071 at 1. Sullivan, the outside rate expert that Baker Botts hired, thought that it would require an ROE of 20–30% to trigger a rate case for the foreseeable future. See JX 1807 at 6; Sullivan Dep. 168. The plaintiffs’ rate expert also believed that there was a “low risk of a rate case for Texas Gas.” Webb Tr. 1008.

188

Based on Webb’s opinion, Atkins set the FERC Switch to the “Off” position. That was reasonable, and it finds support in the broader record. But it results in an alteration to the Loews Model. The Loews Model adopted a conservative approach on the assumption that the Base FERC Impact scenario would occur. This decision therefore uses the Base FERC Impact scenario.

By using the Base FERC Impact scenario, this decision also adopts a conservative measure of damages compared to the more than $900 million that the court could have awarded under the wrongdoer rule. That rule provides that when the “defendant’s wrongful act” causes uncertainty in estimating damages, “justice and sound public policy alike require that he should bear the risk of the uncertainty thus produced.” Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 565 (1931). The wrongdoer rule is a “corollary to [the] presumption” that “doubts about the extent of damages are generally resolved against the breaching party.” PharmAthene, 132 A.3d at 1131. Under the wrongdoer rule, the court “take[s] into account the willfulness of the breach in deciding whether to require a lesser degree of certainty” about the extent of damages.38

38 See Restatement (Second) of Contracts § 352 cmt. a, Westlaw (Am. L. Inst.

database updated Oct. 2021) (“A party who has, by his breach, forced the injured party to seek compensation in damages should not be allowed to profit from his breach where it is established that a significant loss has occurred. A court may take into account all the circumstances of the breach, including willfulness, in deciding whether to require a lesser degree of certainty, giving greater discretion to the trier of facts.”); see also Agilent Techs., Inc. v. Kirkland, 2010 WL 610725, at *27 (Del. Ch. Feb. 18, 2010) (“[I]n cases where a specific injury to the plaintiff cannot be established, the defendant’s actual gain may be considered.”).

189

In this case, the General Partner breached the Partnership Agreement by exercising the Call Right without first meeting the necessary conditions. The General Partner’s breach was willful. The uncertainty about the FERC Impact switch only existed because of the timing of the willful breach, which resulted in the take-private transaction being completed just before FERC published its final rule. The publication of the final rule “mitigate[d]” the supposed “adverse effect” of the March 15 FERC Actions that formed the basis for the Opinion. JX 1569. The uncertainty embodied in the Base FERC Impact scenario would not have existed but for the opportunistic timing of the exercise of the Call Right. Under the wrongdoer rule, that uncertainty should be resolved against the defendants, meaning the proper measure of damages should use the Loews Model with the FERC Switch in the “Off” position.

This decision nonetheless declines to apply the wrongdoer rule. Because Atkins’

model with the FERC Switch in the Base FERC Impact position results in a persuasive and reliable measure of damages, the court adopts it.

3. The Finding Regarding Damages With the FERC Switch set for the Base FERC Impact Scenario, Atkins’ Distribution Model results in a valuation of $17.60 per unit. The transaction price was $12.06 per unit. The plaintiffs are entitled to damages of $5.54 per unit.

190

When the General Partner exercised the Call Right, there were 124,467,395 units outstanding that were not beneficially owned by Loews or its affiliates. 39 Multiplying 124,467,395 by $5.54 yields total damages of $689,827,343.38.

The resulting damages figure is conservative compared to the more than $900 million that the court could have awarded if it had adopted Atkins’ opinion in full. It is also conservative relative to Loews’ contemporaneous estimate of the $1.557 billion in “Value Creation” that Loews expected to enjoy from exercising the Call Right. JX 1505 at 10.

The plaintiffs are entitled to pre- and post-judgment interest on the damages award from July 18, 2018, until the date of payment. When neither party submits evidence showing the appropriate rate of interest, “the court looks to the legal rate of interest.” Taylor v. Am. Specialty Retailing Gp., Inc., 2003 WL 21753752, at *12 (Del. Ch. July 25, 2003). “The legal rate of interest, as defined by 6 Del. C. § 2301, is 5% over the Federal Reserve

39 See JX 1514 at 3 (June 29, 2018, Schedule 13D filing showing “250,296,782 Common Units Outstanding as of March 31, 2018,” of which “124,710,649 Common Units that may be deemed to be beneficially owned by [Loews] based on the right of the General Partner to acquire voting and investment power over such Common Units on July 18, 2018 as a result of the Transaction”); PTO ¶ 388 (“[T]hrough the exercise of the Call Right, Loews . . . acquired all 124,710,469 of the outstanding common units”). Directors and officers of the Partnership disposed of 243,254 units in the Call-Right Exercise. JX 1561 at 1 (Hyland and Hyland’s spouse disposed of 29,307 units); JX 1562 at 1 (Rebell, Rebell’s spouse, and an affiliated LLC disposed of 60,583 units); JX 1563 at 1 (Shapiro disposed of 33,907 units); JX 1564 at 1 (Tisch disposed of 81,050 units); JX 1565 at 1 (Cordes disposed of 23,407 units); JX 1566 at 1 (Horton’s spouse disposed of 15,000 units). Subtracting 243,254 from 124,710,649 yields 124,467,395, the total number of shares held by the class. See Atkins Report Ex. C at 7.

191

discount rate.” Doft & Co. v. Travelocity.com Inc., 2004 WL 1152338, at *12 (Del. Ch. May 20, 2004). When the court “award[s] the legal rate of interest, the appropriate compounding rate is quarterly.” Id.; accord Taylor, 2003 WL 21753752, at *13. The plaintiffs therefore are entitled to pre- and post-judgment interest at the legal rate, compounded quarterly, from July 18, 2018, until the date of payment, with the legal rate fluctuating with changes in the underlying reference rate. The plaintiffs are additionally entitled to an award of fees as the prevailing party. F. The Implied Covenant Of Good Faith And Fair Dealing As an alternative theory of breach, the plaintiffs contend that the General Partner breached the implied covenant of good faith and fair dealing that inheres in every contract governed by Delaware law. Because the court has held that the General Partner breached the express terms of the Partnership Agreement, there is no need to reach the implied covenant.

The plaintiffs have articulated non-duplicative implied covenant theories about the effect of the Potential Exercise Disclosures and the operation of the Purchase Price formula, but a judgment in the plaintiffs’ favor on those questions would result in a lower damages award than the claim for breach of the Call Right. The plaintiffs are only entitled to one recovery. This decision therefore does not wade into the additional implied covenant issues. G. The Claims Against The Defendants Other Than The General Partner The plaintiffs have asserted theories that would enable them to recover from the GPGP, Holdings, and Loews. Those affiliates of the General Partner directed its actions

192

and caused it to exercise the Call Right, but the affiliates are not parties to the Partnership Agreement and hence are not liable in contract. The plaintiffs maintain that the GPGP, Holdings, and Loews are liable to the class on a claim for tortious interference with contract and under the doctrine of unjust enrichment.

Determining whether the General Partner’s affiliates should be liable for tortious interference will require a complex balancing of different factors. See, e.g., NAMA Hldgs., LLC v. Related WMC LLC, 2014 WL 6436647, at *25–36 (Del. Ch. Nov. 17, 2014). This decision has covered much ground, and it would extend its length significantly to take on the tortious interference claim at this time. Furthermore, as a practical matter, it should be unnecessary to determine whether the General Partner’s affiliates tortiously interfered with the Partnership Agreement. As noted, the plaintiffs are only entitled to a single recovery, and if the General Partner pays the damages award, then the class will have no basis to pursue the other defendants.

The facts of this case make it unlikely that pursuing the other defendants will be necessary to ensure the plaintiffs recover their damages. The General Partner acquired 49% of the limited partner interest by exercising the Call Right. It already possessed a 2% general partner interest and all of Boardwalk’s incentive distribution rights. The General Partner thus has access to substantial cash flows.

The same is true for the plaintiffs’ claim for unjust enrichment, although that claim is comparatively easier to analyze. The General Partner remains the principal wrongdoer. It should satisfy the claim.

193

Given these dynamics, the court will not adjudicate the claims for tortious interference or unjust enrichment at this time. Those claims are severed and stayed. If the General Partner satisfies the judgment, then those claims will be moot. If the General Partner fails to satisfy the judgment, then the claims can be revived.

III. CONCLUSION

The General Partner is liable to the plaintiff class for damages in the amount of $689,827,343.38, plus pre- and post-judgment interest on that amount through the date of payment. The plaintiffs are also entitled to an award of costs as the prevailing party.

The parties will incorporate the court’s rulings into a partial final judgment that has been agreed as to form. The partial final judgment will not extinguish the separate claims for breach of the implied covenant of good faith and fair dealing against the General Partner or for tortious interference and unjust enrichment against the General Partner’s affiliates.

If there are other issues that the court needs to address before such an order can be entered, then the parties will prepare a joint letter that identifies the issues and proposes a procedure for resolving them.