In re Cellular Telephone Partnership Litigation ( 2022 )

IN THE COURT OF CHANCERY OF THE STATE OF DELAWARE
IN RE CELLULAR TELEPHONE ) COORDINATED C.A. No. 6885-VCL
PARTNERSHIP LITIGATION )
THIS FILING APPLIES TO COORDINATED CIVIL ACTIONS 6886 AND 6908
MEMORANDUM OPINION ADDRESSING CLAIMS FOR
BREACH OF FIDUCIARY DUTY IN CONNECTION WITH
FREEZE-OUT OF MINORITY PARTNERS IN
SALEM CELLULAR TELEPHONE COMPANY
Date Submitted: January 11, 2022
Date Decided: March 9, 2022
Carmella P. Keener, COOCH AND TAYLOR, P.A., Wilmington, Delaware; Marcus E.
Montejo, Kevin H. Davenport, John G. Day, PRICKETT, JONES & ELLIOTT, P.A.,
Wilmington, Delaware; Thomas R. Ajamie, David S. Siegel, Ryan van Steenis, AJAMIE
LLP, Houston, Texas; Michael A. Pullara, Houston, Texas; Attorneys for Plaintiffs.
Todd C. Schiltz, FAEGRE DRINKER BIDDLE & REATH LLP, Wilmington, Delaware;
William M. Connolly, FAEGRE DRINKER BIDDLE & REATH LLP, Philadelphia,
Pennsylvania; Zoë K. Wilhelm, FAEGRE DRINKER BIDDLE & REATH LLP, Los
Angeles, California; Maurice L. Brimmage, Jr., Laura P. Warrick, AKIN GUMP
STRAUSS HAUER & FELD LLP, Dallas, Texas; Attorneys for Defendants.
LASTER, V.C.
Salem Cellular Telephone Company (the “Partnership”) was a Delaware general
partnership that held a license to provide cellular telephone services in a geographic area
centered around Salem, Oregon. Defendant AT&T Mobility Wireless Operations Holdings
LLC (“Holdings”) owned 98.119% of the partner interest in the Partnership. Holdings is
an indirect, wholly owned subsidiary of non-party AT&T Inc. Through Holdings and other
affiliates, AT&T controlled the Partnership, directed its business and affairs, and managed
its day-to-day operations.1
In October 2010, AT&T caused the Partnership to transfer all of its assets and
liabilities to defendant New Salem Cellular Telephone Company LLC (“New Salem”), a
recently formed affiliate of AT&T. As consideration, AT&T paid the Partnership $219
million in cash, reflecting the value of the Partnership as determined by a valuation firm
retained by AT&T. The Partnership dissolved after the transaction, and AT&T sent each
partner a payment equal to their pro rata share of the liquidating distribution. After the
transaction, AT&T continued to operate the business of the former Partnership. The
transaction thus functioned as a freeze-out of the minority partners (the “Freeze-Out”).2
1
The entity currently known as AT&T came to control the Partnership through a
complex series of corporate transactions spanning years. The evolution of AT&T as an
entity is not directly relevant to this proceeding. For simplicity, this decision refers to
AT&T, unless the context requires a more specific referent. Interested readers may consult
a prior decision for a description of the evolution of AT&T during the life of the
Partnership. See In re Cellular Tel. P’ship Litig. (Salem Contract Decision), 2021 WL
4438046, at *4 n.4, *9 n.13, *11 n.15, *24 n.27, *50 (Del. Ch. Sept. 28, 2021).
2
Between October 2010 and June 2011, AT&T engaged in similar freeze-out
transactions involving twelve other partnerships. The thirteen transactions resulted in the
filing of fifteen civil actions in this court. The cases were coordinated for purposes of pre-
The plaintiffs were minority partners who collectively owned a 1.881% minority
interest in the Partnership. At the price AT&T paid in the Freeze-Out, they collectively
received approximately $4.1 million for their interest.
The plaintiffs assert that AT&T breached its fiduciary duties by effectuating the
Freeze-Out through an unfair process and by paying an unfair price. The parties agree that
the Freeze-Out is subject to the entire fairness standard of review. As a result, AT&T bore
the burden of proving that when considered holistically, the Freeze-Out was entirely fair
to the minority partners.
AT&T failed to prove that the Freeze-Out was entirely fair. For starters, AT&T
failed to prove that it followed a fair process. AT&T correctly anticipated that over the next
trial discovery under the caption In re Cellular Telephone Partnership Litigation, C.A. No.
6885-VCL (the “Coordinated Action”). By agreement, the parties subsequently conducted
a coordinated trial. The court is therefore issuing this decision in the Coordinated Action.
Five of the other partnerships have histories and governance structures that are
substantially similar to the Partnership’s. Those five are (1) Bremerton Cellular Telephone
Company, (2) Melbourne Cellular Telephone Company, (3) Provo Cellular Telephone
Company, (4) Sarasota Cellular Telephone Company, and (5) Visalia Cellular Telephone
Company.
Seven of the other partnerships have histories and governance structures that differ
to varying degrees from the Partnership. Those seven are (1) Alton CellTelCo, (2)
Bellingham Cellular Partnership, (3) Bloomington Cellular Telephone Company, (4)
Bradenton Cellular Partnership, (5) Galveston Cellular Partnership, (6) Las Cruces Cellular
Telephone Company, and (7) Reno Cellular Telephone Company.
At times, this decision refers to the other partnerships. When referring to a specific
partnership, this decision uses the name of its market. For example, a reference to
“Melbourne” refers to the Melbourne Cellular Telephone Company.
2
decade, an explosion in data usage would lead to profitable new businesses and products,
causing the value of the Partnership to increase significantly and enabling the Partnership
to pay higher distributions. By acquiring the minority partners’ interests, AT&T sought to
capture that value for itself. AT&T did not employ any procedural protections to ensure
fairness to the minority partners; AT&T simply hired an outside valuation firm. Although
AT&T claimed that the firm was independent, the record shows that the lead partner had a
longstanding relationship with AT&T and that internal AT&T personnel influenced the
outcome of the valuation. AT&T thus failed to prove that it dealt fairly with the minority
partners.
AT&T also failed to prove that it paid a fair price. During this litigation, rather than
relying on the work of the valuation firm it chose, AT&T brought in a litigation expert who
conducted her own analyses. Neither the original valuation firm nor AT&T’s litigation
expert used persuasive valuation methodologies.
AT&T therefore breached its duty of loyalty by engaging in an unfair and self-
interested transaction at the minority partners’ expense. As a remedy, this decision awards
the plaintiffs damages equal to the difference between the consideration they received and
a pro rata share of the fair value of the Partnership as determined by the court. Because its
fair value determination is being used for the purpose of remedying a proven breach of the
duty of loyalty, the court has sought to achieve a remedy that eliminates to the extent
possible the ability of AT&T to profit from its breach. On close issues, the court has given
the plaintiffs the benefit of the doubt, resulting in a valuation that favors the plaintiffs.
3
This decision holds that the fair value of the Partnership for purposes of the remedial
award was $714 million. The plaintiffs’ pro rata share of the fair value of the Partnership
was $13.4 million. Subtracting the consideration that the plaintiffs received in the Freeze-
Out results in a damages award of $9,311,965. The plaintiffs are entitled to that amount,
plus pre- and post-judgment interest at the legal rate, compounded monthly, from the date
of the Freeze-Out until the date of payment.
I. FACTUAL BACKGROUND
Trial took place over five days. The parties introduced 3,187 exhibits, including
thirty-nine deposition transcripts. Four fact witnesses—all present or former AT&T
executives—and three experts testified live. The following factual findings represent the
court’s effort to distill this record.3
3
Trial was held in the Coordinated Action and addressed all of the partnerships and
all of the coordinated lawsuits. Unless otherwise noted, citations to docket entries refer to
items filed in the Coordinated Action. Citations in the form “PTO ¶ ––” refer to stipulated
facts in the pre-trial order. Dkt. 600. 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. The parties deposed some witnesses multiple times.
For those witnesses, the citation includes the year of the pertinent deposition. Citations in
the form “JX ––– at –––” refer to a trial exhibit with the page designated by the last three
digits of the control or JX number or, if the document lacked a control or JX number, by
the internal page number. If a trial exhibit used paragraph numbers, then references are by
paragraph.
The court issued a previous decision addressing claims for breach of contract
involving the Partnership. See Salem Contract Decision, 2021 WL 4438046. The court’s
comments in that decision regarding the credibility of witnesses carry over to this decision.
The court has treated the report of the Special Discovery Master in the same manner for
purposes of this decision as it did in the Salem Contract Decision.
4
A. The Formation Of The Partnership
During the 1980s, the Federal Communications Commission (the “FCC”)
conducted lotteries to award the rights to construct cellular telephone networks in particular
geographic areas. If the lottery winner built out the network and complied with other
regulatory requirements, then the FCC granted the lottery winner a license to provide
cellular telephone service in that area. The legacy wireline carrier received a separate
license and was not permitted to participate in the lottery.
One of those geographic areas was centered around Salem, Oregon (the “Salem
Market”). To increase their odds of winning, some of the lottery participants entered into
an arrangement similar to an office pool: If one of them won, then the winner would
contribute its rights to a partnership and receive a 50.01% interest in the partnership. The
other members of the pool would receive shares of the remaining 49.99% interest.
One of the group members won the lottery for the Salem Market. The winning
member sold her rights to a predecessor of AT&T. In 1988, AT&T and the other members
of the pool formed the Partnership and executed a partnership agreement to govern its
affairs. JX 15 (the “Partnership Agreement” or “PA”). As the party contributing the rights
to build, maintain, and operate the network, AT&T received a 50.01% interest in the
Partnership. Over 110 members of the group shared the minority interest, with each
member initially receiving a 0.3424% interest in the Partnership. See id. at ’431–33.
From the outset, AT&T controlled the Partnership. First, AT&T controlled the
Partnership at the partner level. The Partnership Agreement generally authorized the
5
partners to take action by majority vote. Id. § 4.1. As the holder of a majority interest,
AT&T controlled the outcome of any partner-level vote.
Second, AT&T controlled the Partnership’s governing body. The Partnership
Agreement delegated “complete and exclusive power to conduct the business affairs of the
Partnership” to a three-member Executive Committee, with two representatives appointed
by the majority partner and one by the minority partners. Id. § 4.3. As the majority partner,
AT&T appointed two representatives.4 And because the Partnership Agreement authorized
the Executive Committee to act by majority vote, AT&T’s representatives could dictate the
outcome of any vote.5
Third, AT&T controlled the Partnership’s day-to-day operations. The Partnership
did not have its own officers or employees. The Partnership did not even have its own bank
account. AT&T employees performed all of the tasks that the Partnership needed to
accomplish. See Wages Tr. 125, 134.
4
During the period relevant to this case, Eric Wages served as one of AT&T’s
representatives. Wages was an AT&T executive who oversaw AT&T’s Partnership
Accounting Group. The other representative was the director of AT&T’s regional business
unit that included the Salem Market. See Wages Tr. 129.
Wages served as AT&T’s principal witness. As discussed in the Salem Contract
Decision, problems with Wages’ testimony caused the court to approach his assertions with
care. See 2021 WL 4438046, at *5 n.6.
5
In practice, AT&T only acted through the Executive Committee on formal matters,
such as authorizing a distribution to the partners. Wages Tr. 131–32, 276. AT&T generally
ran the business of the Partnership as an integrated part of AT&T’s wireless network,
without seeking or obtaining Executive Committee approval for particular decisions.
6
B. The Subscriber-Based Business Model
From the early days of the cellular industry until the mid-2000s, wireless carriers
pursued a relatively stable business model that depended on “postpaid” wireless voice
plans. Postpaid subscribers entered into long-term contracts (typically one or two years)
and paid fees based on their monthly usage. AT&T employed this business model, and the
Partnership’s primary revenue stream consisted of fees paid by postpaid subscribers that
AT&T allocated to the Partnership.
Wireless carriers tracked subscribers and their usage using a system known as
“NPA-NXX,” a shorthand term for the area code and next three digits of the subscriber’s
phone number.6 For example, in the phone number (999)-555-1234, the NPA-NXX is 999-
555. The last four digits produce a block of 10,000 phone numbers, ranging from 0000 to
9999, associated with that particular NPA-NXX.
Wireless carriers assigned particular NPA-NXX blocks to their market-level entities
based on geography. Within its accounting system, AT&T assigned company codes to its
blocks of NPA-NXX numbers. AT&T then used the codes to attribute revenue and expense
to particular market-level entities, such as the Partnership.
To assign the proper NPA-NXX number to a new subscriber, AT&T asked the
subscriber to identify the phone’s primary place of use. An AT&T employee then assigned
6
The “NPA” referred to the fact that wireless carriers received blocks of 10,000
telephone numbers from the North American Number Plan Administration, a publicly
funded entity run by an FCC contractor.
7
the subscriber an NPA-NXX number based on the subscriber’s reported primary place of
use. If the subscriber identified a primary place of use that corresponded to the area covered
by the Partnership, then the customer received an NPA-NXX number assigned to the
Partnership and was treated thereafter as a subscriber of the Partnership.
By assigning NPA-NXX numbers based on primary place of use and allocating
revenue and expense to the market unit corresponding to the NPA-NXX number, AT&T
sought to connect portions of its network with the revenue and expense those portions
generated. The system broke down if a customer moved to a new market, because AT&T
had no mechanism for assigning the existing NPA-NXX number to the new market.
Instead, the customer’s usage continued to be attributed to the original market. In other
words, if a customer with a number assigned to the Partnership moved from Salem, Oregon,
to Salem, Massachusetts, then the customer’s revenue and expense would continue to be
attributed to the Partnership.
Until the mid-aughts, that major defect was not a significant problem, because other
aspects of the wireless business model resulted in an NPA-NXX number acting as a strong
proxy for primary place of use. During that era, if a subscriber used her cellular phone
outside of her local market, then the carrier charged the subscriber for “roaming.”
Competing carriers entered into agreements that permitted their subscribers to roam
on their competitors’ networks. As a result, there were two types of roaming. Intra-
company or intra-carrier roaming referred to a customer who used her phone outside of her
home area, but still used her provider’s network. Inter-company or inter-carrier roaming
8
referred to a customer who used her phone outside of her home area, but used a different
carrier’s network.
Regardless of type, roaming was expensive. Due to the high cost of roaming, a
customer who relocated outside of her home area had a strong financial incentive to obtain
a new NPA-NXX number. Moreover, until the advent of number portability in 2004, any
subscriber who changed carriers was treated as a new subscriber and received a new NPA-
NXX number. A customer’s NPA-NXX number therefore correlated strongly with the
customer’s primary place of use, and customers holding NPA-NXX numbers associated
with the Partnership were highly likely to be primarily using the Partnership’s portion of
AT&T’s network.7
C. The Advent Of Nationwide Plans And Number Portability
The basic cellular business model remained relatively stable until late 2003. Two
developments brought the stable period to an end: nationwide plans and number portability.
As the cellular industry grew, competition among carriers increased. One
consequence was a shift to nationwide plans that eliminated roaming fees. Those plans in
turn removed the financial incentive for a customer to obtain a new NPA-NXX number
after relocating. Instead, there was a natural disincentive for a customer to avoid that hassle.
7
That said, the system was not foolproof. For example, a college student with an
NPA-NXX number associated with her family’s home market might use her phone in a
different market while attending college. Or a “snowbird” with an NPA-NXX number
associated with New York might use her phone in Florida during the winter. The NPA-
NXX system nevertheless generally operated as a reliable proxy for principal place of use.
9
As more subscribers moved while keeping their original numbers, the linkage between
NPA-NXX and principal place of use became less reliable.
The other major development was number portability. That concept refers to a
subscriber’s ability to keep the same phone number when switching carriers. Before the
advent of number portability, a subscriber who switched wireless carriers had to obtain a
new NPA-NXX number based on primary place of use, so the need to change numbers
helped keep the NPA-NXX numbers current.
In the Telecommunications Act of 1996, Congress mandated that cellular carriers
take steps to enable number portability by 2004. Beginning in 2004, a subscriber no longer
had to receive a new NPA-NXX number when changing carriers.
The combination of nationwide plans and number portability undermined the
association between NPA-NXX and primary place of use. By the time of the Freeze-Out,
the NPA-NXX system had become so unreliable that AT&T could not provide basic
information about its subscribers or the Partnership’s:
• AT&T did not know the number of AT&T subscribers who resided in the
Partnership’s service area but used a non-Partnership NPA-NXX number.
• AT&T did not know the number of AT&T subscribers who resided in a non-
Partnership service area and used an NPA-NXX number assigned to the Partnership.
• AT&T did not know the number of AT&T subscribers who moved to the
Partnership’s service area, changed their billing address and primary place of use to
an address in the Partnership’s service area, yet continued to use a non-Partnership
NPA-NXX number.
• AT&T did not know the percentage of AT&T subscribers nationwide who resided
in AT&T service areas different from the one that issued their NPA-NXX number.
Wages Tr. 364–66.
10
D. The Management Agreement
In 2005, in response to the developments affecting the cellular industry, AT&T
caused the Partnership to enter into a Management and Network Sharing Agreement. JX
217 (the “Management Agreement” or “MNSA”). The terms of the Management
Agreement specified how AT&T would operate the Partnership as part of its national
cellular business.
Despite having written the Management Agreement, AT&T pervasively
disregarded it. See Salem Contract Decision, 2021 WL 4438046, at *19–24. Instead of
using the metrics specified in the Management Agreement to allocate revenue and expense
to the Partnership, AT&T’s accountants allocated revenue and expense using their own
judgment about what would be “fair and reasonable.” Wages 2020 Dep. 189–90.
Most notably for valuation purposes, AT&T disregarded the Management
Agreement’s allocation methodology for “Shared Revenues,” defined as “the aggregate
revenue generated by Subscribers of [the Partnership’s] Business and [AT&T’s] Business
utilizing [the Partnership’s] System and [AT&T’s] System, and any other applicable
revenues generated by utilization of the Entire Network, but excluding Outcollect Roaming
Revenues.” MNSA at ’751. In the Management Agreement, AT&T agreed to aggregate
Shared Revenues and allocate a portion of the resulting revenue to the Partnership based
on the proportion of “Traffic” carried by the Partnership’s system. See id. Ex. A (the
“Shared Revenues Formula”). The Management Agreement defined “Traffic” broadly as
“electronic signals including voice, data, and other associated electronic signals.” Id. at
’752. The Management Agreement called for each “Unit of Traffic” to be allocated based
11
on “the cell site that carries such Traffic, with the convention that the Unit of Traffic from
any one call or transmission shall be assigned to the cell site upon which such call or
transmission is first carried.” Id. § V(G)(1).
AT&T never followed the Shared Revenues Formula. AT&T never established an
aggregate pool of Shared Revenues. AT&T never allocated a proportionate share of Shared
Revenues to the Partnership. Instead, AT&T developed its own internal allocation
methodologies without considering the Management Agreement. Those methodologies
involved identifying items of revenue and expense and assigning them to the Partnership
whenever possible, even if the Management Agreement specified a different allocation
methodology.
AT&T also disregarded the requirement in the Management Agreement that AT&T
apply a premium to Shared Revenues and a discount to the Partnership’s share of “Sales
and Marketing Expenses,” defined as “expenses associated with Sales and Marketing
Services,” which in turn was defined as “marketing, sales, advertising and other
promotional and subscriber acquisition and retention services.” Id. at ’751. The operative
language stated:
[AT&T] may apply a premium to certain revenues and/or a discount to
certain expenses to ensure that the allocation method set forth below is no
less favorable to the [Partnership] than the allocation method used in prior
periods. Initially, [AT&T] will apply a premium of 25% to [the
Partnership’s] share of Shared Revenues and a discount of 10% to [the
Partnership’s] share of Sales and Marketing Expenses.
Id. Ex. A (the “Premium Provision”). Through the Premium Provision, AT&T committed
to treat the Partnership better than its own business units and indisputably better than an
12
arm’s-length third party. Although AT&T reserved the right to apply a different premium
to Shared Revenues and a different discount to Sales and Marketing Expenses, AT&T
agreed that any revised allocation method would be “no less favorable to [the Partnership]
than the allocation method used in prior periods.” Id. AT&T thus committed to add at least
a 25% premium to the Partnership’s allocated share of revenue and to deduct at least a 10%
discount from the Partnership’s share of Sales and Marketing Expenses. AT&T never
applied a premium to the Partnership’s share of revenue and never applied a discount to
the Partnership’s share of Sales and Marketing Expenses.
Furthermore, AT&T failed to follow the Management Agreement’s definition of the
Partnership’s business, which established that the Partnership’s wireless business was
coextensive with AT&T’s wireless business. Instead, AT&T’s accountants viewed the
Partnership’s business as limited to “cellular phones, voice, data, SMS text, those things .
. . we do with our phones.” Wages Tr. 395–97. For activities outside of this narrow
definition, AT&T’s accountants believed that the Partnership only was entitled to be “made
whole” for AT&T’s use of its assets. Id. at 406. AT&T’s definition of “wireless activity”
was not written down, and it was inconsistent with the Management Agreement. Id.
AT&T also continued to use NPA-NXX to assign subscribers to the Partnership. By
the mid-2000s, the arrival of nationwide unlimited roaming and number portability had
undermined the reliability of NPA-NXX as a proxy for primary place of use. The
Management Agreement did not mention NPA-NXX, and the record indicates that AT&T
caused the Partnership and its sister entities to adopt the Management Agreement as part
13
of an effort to move away from NPA-NXX. AT&T nevertheless persisted in using NPA-
NXX to identify subscribers to the Partnership.
Because AT&T pervasively disregarded the Management Agreement, the
Partnership did not receive its full share of revenue. The Partnership also did not receive
the benefits of the Premium Provision, under which the Partnership should have received
at least 25% more revenue than AT&T allocated to it.
Setting aside these problems, AT&T’s accountants strived to allocate revenue and
expense consistently and accurately. They deployed extensive accounting resources to
track and allocate transactions, maintained overlapping systems of oversight, consulted
with AT&T’s outside auditors, and interacted with AT&T executives whose compensation
structures gave them an incentive to ensure that AT&T allocated revenues consistently
across market-level entities. As a result of these careful procedures, AT&T’s accounting
records accurately reflected the revenue and expense that AT&T identified and assigned or
allocated to the Partnership, as well as the methodologies that AT&T used. It is not the
case, for example, that AT&T decided that the Partnership should receive a particular type
of revenue, but its accountants failed to allocate it. Nor is it the case that the accountants
decided to use a particular methodology, then implemented it errantly. The Partnership’s
financial records accurately depict how AT&T treated the Partnership.
E. The Data Revolution
By 2007, the market for wireless voice communications, known as legacy wireless,
was relatively mature. For that business, the double-digit subscriber growth of the early
2000s had leveled off to single digits. Wireless providers began to see stronger growth in
14
data services, driven by demand for downloadable music, games, ring tones, and text and
photo messaging. The growth of data services opened up a vast array of new businesses
and new revenue streams, including mobile internet advertising.
AT&T planned to capitalize on these changes. In June 2007, AT&T announced the
appointment of Randall Stephenson as its new chief executive officer (“CEO”). Stephenson
envisioned AT&T as “a brand-new company, with wireless at the heart of what we do.” JX
282.
A cornerstone of AT&T’s plan was Apple Inc.’s iPhone, which launched on June
29, 2007. AT&T was the exclusive wireless carrier for the iPhone for the first eighteen
months after its release. AT&T later extended its exclusive arrangement with Apple
through early 2011. AT&T correctly predicted that the iPhone would transform how people
approached wireless communication. See Lurie Dep. 31.
At an analyst conference later in 2007, AT&T highlighted that wireless data traffic
was “[g]rowing [e]xponentially,” with “[u]sage [q]uadrupling [e]very [y]ear.” JX 325 at
11; see JX 359 at 9. Annual growth in revenue from data services had “exceeded 60% for
five consecutive quarters,” reaching an annualized run rate of more than $7 billion in the
third quarter of 2007. JX 324 at 10, 22; see JX 325 at 10, 15. The iPhone was the key driver
in the growth of revenue from data services, with subscribers’ data usage nearly doubling
after they purchased an iPhone. JX 325 at 12. AT&T’s partnership with Apple meant that
AT&T was “[e]arly in [the] [a]doption [c]urve for [m]edia-[r]ich [d]evices,” leading to
“[h]uge [p]otential in [w]ireless [d]ata.” Id.
15
Another driver of increased demand for data services came from “Connected
Devices,” which are electronic devices other than cell phones that communicate via
wireless networks. Connected Devices include tablets, electronic readers, personal and
vehicle navigation devices, home security systems, machine-to-machine devices, and
“smart grid” devices. AT&T offered data services for some Connected Devices through
individual data subscriptions. AT&T offered data services for other Connected Devices
through commercial agreements with companies like Amazon, General Motors, and
Garmin (the “Commercial Network Agreements”). For example, Amazon entered into a
Commercial Network Agreement for its Kindle product. General Motors entered into a
Commercial Network Agreement for its On-Star vehicle security system. And Garmin
entered into a Commercial Network Agreement for its navigation devices.
Increased revenue was not the only side of the equation. AT&T would need to invest
in its network to handle the increased demand. To expand its wireless capabilities, AT&T
paid $2.5 billion in October 2007 to purchase additional wireless spectrum. In the press
release announcing the purchase, an AT&T executive explained that “[c]ustomer demand
for mobile services, including voice, data and video, is continually increasing.” JX 319.
With its purchase, AT&T touted that it could capitalize on data revenue growth due to its
dominance in “Average Spectrum Depth” in the top one hundred U.S. markets. JX 324 at
4.
F. AT&T Considers Eliminating The Minority Partners.
With revenue from data services and related businesses expected to balloon, AT&T
began to explore ways to eliminate the minority partners. Chris Reeves, an executive in
16
AT&T’s Corporate Development department, conducted the initial analysis. In August
2007, he gave a presentation to Wages, who oversaw AT&T’s Partnership Accounting
Group, and Philip Teske, an AT&T executive with AT&T Mobility LLC (“Mobility”), the
entity that managed AT&T’s wireless business.
Titled “Partnership Relations—Restructuring Update and Buyout Analysis,”
Reeve’s presentation identified fifty-one market-level entities, including the Partnership,
where AT&T could benefit by acquiring the minority interest. See JX 310 at ’765–67. The
presentation explained that “[a] buyout today will be much less expensive than 1 or 2 years
down the road given OIBDA growth rates.” Id. at ’765.8
The presentation divided the potential acquisitions into three tiers. The “Priority
One” tier consisted of six corporations and “two partnerships where [AT&T] owns more
than a 99% interest.” Id. at ’765. The “Priority 2” tier consisted of twelve “legacy AT&T
wireless partnerships,” including the Partnership. Id. at ’766. For those entities, AT&T
could eliminate the minority using freeze-out mergers or similar transactions. Id. The
“Priority 3” tier consisted of “legacy Cingular partnerships where [AT&T] would need to
approach the partner(s) and negotiate a purchase.” Id. at ’767. Reeves contemplated
focusing first on the entities where minority holders had no ability to block the buyouts.
In early 2008, Reeves gave a similar presentation to Mobility’s chief financial
officer (“CFO”), Pete Ritcher. Titled “Minority Partnership Acquisition Summary,” the
8
“OIBDA” is an acronym for “operating income before depreciation and
amortization.” See Teske Tr. 583. It is “essentially the same” as EBITDA. Id. at 584.
17
presentation stated that “[p]ending asset mergers and accelerating valuations provide an
opportune time to buyout certain minority partnerships.” JX 344 at 2. Consistent with the
2007 presentation, Reeves recommended starting with “the 18 legacy AT&T wireless
partnerships” where transactions could be accomplished “without partner consent.” Id.
Reeves identified multiple reasons for the buyouts. He noted that the minority
partners’ lack of “knowledge of the industry and historically poor relations” had led to
“consistent conflict.” Id. at 3. He also observed that the partnerships required “extensive
use of accounting and field personnel to resolve ongoing issues.” Id. at 4. As a result, the
buyouts would “eliminate value leakage.” Id.
Equally important, Reeves highlighted the value AT&T could capture retaining the
full value of the profits that the partnerships paid out in distributions. Reeves contemplated
buyouts at eighteen partnerships for a total cost of $100 million. By acquiring the minority
interests in those entities, Reeves projected that AT&T could “[e]liminate[] $186M of
Distribution and Dividend payments over 10 years” with a present value of $113 million.
Id. at 2, 4. The buyouts would create “total savings of $243M in perpetuity.” Id. at 7. The
presentation explained that the value of the partnerships was “projected to increase 60%
from 2007 to 2010 based on Mobility OIBDA growth rates,” equivalent to an OIBDA
compound annual growth rate of 18%. Id. at 8. By acquiring the minority interests, AT&T
would “retain [the] lift in value driven by [the] projected growth of the business” and
acquire the interests “at a discount to future growth.” Id. at 8, 9.
The forecast of $243 million in savings projected that distributions to the minority
partners would grow at a rate of 14% in 2009, 8% in 2010, 6.5% in 2011, and 3.5%
18
thereafter. See id. App’x A. The steep drop was questionable, and Teske asked why there
was “such a quick decline to 3.5%” from the high of 14% in 2009. JX 345 at ’254. If the
drop was not so steep, then the buyouts would be even more beneficial to AT&T. Teske
wrote, “I will be keeping my fingers crossed that you get ‘the green light’ to proceed.” Id.
G. AT&T Invests In Its Network To Capitalize On The Demand For Data.
While AT&T was considering eliminating the minority partners, the demand for
data services continued to increase. In its fourth quarter 2007 earnings release, AT&T
announced “record wireless gains” that included 57.5% annual growth in revenue from
wireless data services. JX 342 at 1, 4. AT&T viewed data services as a promising new
source of revenue growth as wireless voice revenue began to slow. See JX 359 at 30.
AT&T ultimately decided not to buy out the minority partners in 2008. AT&T chose
instead to buy additional spectrum and to invest in its network. Teske Tr. 535–36.
As suggested by its spectrum purchases, AT&T remained bullish on the prospects
for data usage. AT&T projected that “[t]otal revenue growth” would be “fueled by growth
in data revenue as more subscribers utilize their handsets for [web] browsing and e-mail
services and data only subscriptions begin to proliferate.” JX 367 at ’324. AT&T forecasted
“[t]op-line growth coming from data” as voice revenue growth slowed and AT&T achieved
“[h]igher penetration of data services on handsets.” Id. at ’322. AT&T expected to “add[]
29.6M new subscribers between 2009 and 2018,” with “[l]aptops and emerging device
growth account[ing] for 17M of the net adds.” Id. at ’323. AT&T expected to gain market
share and forecasted “stable margins” for the foreseeable future. Id. at ’322.
19
AT&T also recognized the need to continue investing in its network to support
increased data usage. For example, during the second quarter of 2008, AT&T deployed a
new 3G mobile telephone protocol to enable AT&T’s network to achieve faster mobile
broadband speeds, boosting the rate and volume of wireless data transmission. See JX 325
at 116; JX 1994 at 6, App’x B ¶ 3. AT&T also planned to invest in 4G technology, which
was expected to achieve “[c]ommercial availability in 2010.” JX 367 at ’322. AT&T took
these steps to capitalize on the data revolution and the revenue streams it would generate.
H. AT&T Develops New Data Services And Businesses.
In October 2008, AT&T announced the appointment of Glenn Lurie as the head of
its Emerging Devices Organization. Describing the Connected Devices business, Lurie
explained that “[h]igh speed wireless broadband service can enhance a huge variety of
gadgets” and announced that there were “a host of exciting new applications – from social
networking to navigation to location-based solutions – being developed that will rely on
wireless connectivity.” JX 401.
In spring 2009, AT&T provided further details about its efforts to capitalize on the
growing demand for data services. At an industry conference in April 2009, Lurie reported
that the Emerging Devices Organization “was talking to a whole range of device makers
— from garage start-ups to billion-dollar companies — along with the major retailers” to
develop new “wireless applications for consumer electronics devices, including game
machines, electronic book readers and video and still cameras.” JX 447 at 1. In April 2009,
AT&T announced partnerships with the Houston Independent School District and Hertz
20
Corporation to provide fleet management and other services for school buses and rental
cars. JX 451; JX 452.
In May 2009, Lurie gave the keynote address at the GoMobile wireless industry
conference. He touted AT&T’s plans to “bring[] wireless connectivity to a host of new
consumer electronics devices and applications — including personal computers, internet
devices, in-car entertainment and navigation systems, cameras, and machine-to-machine
communications solutions.” JX 455 at 3. Also in May 2009, AT&T announced an exclusive
partnership with Jasper Wireless that would provide a platform for the Emerging Devices
Organization “to accelerate market entry for new categories of connected devices on
AT&T’s network, including personal navigation, e-readers, mobile internet devices,
gaming, healthcare, tracking, and in-car navigation systems, among others.” JX 460 at 1.
The announcement quoted Lurie as saying, “This is a significant step for AT&T as we
continue to gain momentum with our emerging device strategy . . . . It’s the ‘technological
underpinnings’ to make our strategy possible.” Id.
At the end of May 2009, AT&T published “The Mobile Enterprise: Moving to the
Next Generation.” JX 471. AT&T stated that “[n]ew devices, new applications and new
uses of wireless connectivity are coming and we’re committed to being the leader.” Id. at
4. In a June 2009 interview, Lurie discussed AT&T’s plans to capitalize on the demand for
data by “embed[ding] wireless access into anything that isn’t a smartphone, netbook or
PC.” JX 485. Lurie also described AT&T’s plans for new data pricing models. See id.
21
I. AT&T Again Considers Eliminating The Minority Partners.
In fall 2009, AT&T again considered eliminating the minority partners. For AT&T,
the economic rationale for buying the minority interest had grown more compelling,
because AT&T’s distributions to the minority partners had nearly doubled since AT&T
first considered a buyout. See JX 616 at 13; JX 3516, “actualDist” tab.
Anticipating lawsuits from the minority partners, AT&T quickly brought litigation
counsel into the process. Initially AT&T treated the buyouts as new projects, referring to
them as “Project Smoothie” and “Project Slim.” But after recognizing the desirability of
having a non-valuation-based justification for the transactions, Stephens and Wages
directed their teams to refer to “[P]roject LESS in all communications.” JX 721 at ’850;
accord Stephens Tr. 31–32; see JX 587 at ’883; JX 616 at ’933; JX 682. Project LESS was
an effort that AT&T had pursued over the preceding decade to reduce the number of entities
in the AT&T corporate family and save administrative costs. Stephens Dep. 153–54.
To support an argument that the buyouts were just another part of Project LESS,
AT&T put the leaders of Project LESS in charge of the process. The point person was
Debbie Dial, an executive who handled special projects for John Stephens, AT&T’s CFO.
Dial directed Teske, Wages, and others to compile information about the savings
that a restructuring could generate. The team based its analysis on the “Minority Entity
Acquisition Summary” that Reeves prepared in January 2008. See JX 518; JX 547.
Contemporaneously, AT&T’s senior executives were reviewing AT&T’s strategy
for mobile applications and Connected Devices. AT&T executives viewed initiatives in
those areas as critical to “provid[ing] AT&T a stronger foothold in the value chain” and
22
enabling AT&T to compete with the likes of Apple, Microsoft Corporation, and Google.
JX 532 at 5. AT&T believed that by leveraging its network, AT&T could “drive a 10%
uplift in AT&T data revenue,” equal to $8 billion, by 2014. Id. at 7.
In November 2009, the Project LESS team presented its findings. JX 616. The
presentation evaluated three structures to create savings for AT&T: a national asset roll-
up, regional asset roll-ups, and minority buyouts. Id. at ’937. The team strongly
recommended the minority buyouts as the “[s]traightfoward/simple alternative” that would
(i) “[a]void[] litigation risks associated with business unit integration,” (ii) provide a
“[c]lear legal path forward,” and (iii) “[e]liminate 360 ownership stakes” associated with
the entities. Id. at ’938.
J. AT&T Retains PwC To Value The Partnership.
After the presentation from the Project LESS team, Stephens directed Teske and
Greg Hall, Mobility’s controller, to engage a valuation firm to value the minority partners’
interests. AT&T contacted PricewaterhouseCoopers LLP (“PwC”) and Deloitte LLP, but
quickly zeroed in on PwC because of its comfort with Aaron Gilcreast, who would lead the
PwC team.
Before joining PwC, Gilcreast had worked on a number of projects for AT&T and
developed relationships with AT&T business executives. While at Standard & Poors,
Gilcreast had worked on AT&T’s acquisition of Cingular. After Duff & Phelps acquired
his practice group, Gilcreast performed accounting and valuation work for AT&T,
including projects that involved valuing market-level entities like the Partnership. See
Gilcreast Dep. 23–25, 39; Kobos Dep. 32–33; Hall Tr. 1077–79; JX 347 at 3. Gilcreast had
23
left Duff & Phelps to join PwC in 2009, and he remained bound by a noncompetition
agreement that prohibited him from working for clients of Duff & Phelps. At AT&T’s
request, Duff & Phelps waived the restriction. JX 722; see Gilcreast Dep. 31–32.
Before AT&T hired PwC, AT&T’s litigators vetted Gilcreast. The litigators also
participated in the initial meeting with Gilcreast’s team, and they instructed Teske to
provide Gilcreast with a set of documents so they could question Gilcreast. JX 678; JX
687. Teske advised Gilcreast that the attorneys would be focusing on his
“capacity/effectiveness as a potential witness.” JX 689; see also JX 691 (Teske anticipating
that Gilcreast would bill by the hour “for any potential testimony”).
In February 2010, after the litigators gave the thumbs up, AT&T formally retained
PwC. AT&T agreed to pay PwC $31,860 to value the Partnership. In total, AT&T agreed
to pay $417,320 for valuations of all thirteen partnerships. To get PwC started, AT&T
directed Duff & Phelps to send PwC valuations for four of AT&T’s partnerships that Duff
& Phelps prepared in 2007. JX 3502.
AT&T provided PwC with two sets of financial forecasts. One was a three-year plan
for AT&T’s wireless business prepared by the Financial Planning Team at Mobility (the
“Three-Year Plan”). The other was a ten-year plan for AT&T as a whole prepared by the
Corporate Financial Planning Team at AT&T (the “Ten-Year Plan”).
1. The Three-Year Plan
Each summer, AT&T Mobility’s Financial Planning Team prepared that year’s
iteration of the Three-Year Plan. It consisted of a three-year forecast and a valuation of
AT&T’s wireless business. AT&T considered the Three-Year Plan “to be an accurate
24
representation of [its] true expectations for the business.” JX 780 at ’841. AT&T rigorously
monitored its performance against the Three-Year Plan. Stephens Dep. 34–35.
AT&T used the Three-Year Plan to draft Mobility’s annual budget. Stephens Tr. 39,
41; Paoletti Dep. 21. AT&T also used the Three-Year Plan when calculating incentive
compensation for its executives. Stephens Tr. 39; Stephens Dep. 32–34; Kobos Dep. 112–
13. And AT&T used the Three-Year Plan when communicating with investors about
targets for subscriber growth, revenue, and earnings per share. Stephens Dep. 32–33;
Kobos Dep. 127, 147–48. More broadly, Mobility used the Three-Year Plan for strategic
planning and to measure performance against the operational and financial goals set by
AT&T’s Financial Planning Team. Paoletti Dep. 37–38.
Dale Paoletti, the Director of Financial Analysis at Mobility, oversaw the
preparation of the Three-Year Plan. His team started by analyzing historical trends and
Mobility’s forecast for the current year. The team then sought input from AT&T subject
matter experts, such as sales, marketing, finance, information technology, and network
specialists, who identified their departments’ needs and provided feedback on trends.
Paoletti’s team typically deferred to the subject matter experts, but would seek clarification
on any input that “looked out of trend.” Paoletti Dep. 26–28; see Stephens Tr. 40. The
process thus began with a “top-down” approach, then applied a “bottoms-up” approach
based on “[i]nput received from across the organization.” JX 780 at ’841.
After preparing the Three-Year Plan, Paoletti’s team handed it off to AT&T’s
Financial Planning Team, who used the first year of the plan to prepare the current year’s
budget for AT&T’s wireless business. Stephens Tr. 39, 41; Paoletti Dep. 21–22. After
25
review by senior AT&T executives, the Three-Year Plan and the annual budget were
presented to and approved by AT&T’s board of directors. JX 1269 at ’176; Stephens Tr.
40; see Paoletti Dep. 42–43.
2. The Ten-Year Plan
Each spring, AT&T’s Financial Planning Team prepared that year’s iteration of the
Ten-Year Plan, which AT&T also referred to as a “10-Year View Sum-of-Parts Valuation.”
JX 1269 at ’176. The Ten-Year Plan was a long-term forecast and valuation for all of
AT&T’s businesses.
AT&T regarded the Ten-Year Plan as a “business as usual” view of the company as
a whole. See Kobos Dep. 26–27; Stephens Tr. 40–41. AT&T used the Ten-Year Plan as a
baseline when considering how new initiatives might affect the company. AT&T also used
the Ten-Year Plan in its annual tests of goodwill impairment.9
Philip Kobos, AT&T’s Director of Corporate Planning, oversaw the preparation of
Mobility’s portion of the Ten-Year Plan. Kobos’ team used a “top-down” process that
involved updating the prior year’s plan with the latest operational results and the
projections from the Three-Year Plan. Kobos Dep. 48; see PTO ¶ 230. The team then
9
Duff & Phelps valued AT&T’s spectrum licenses as a single asset for goodwill
impairment purposes. The valuation process involved a variant of a discounted cash flow
(“DCF”) analysis. JX 3551 at 4. AT&T then compared the Duff & Phelps’ valuation to the
book value of the spectrum licenses. The process thus implicitly valued the spectrum
owned by market-level entities, such as the Partnership. See Stephens Tr. 68–69. AT&T’s
CFO certified the accuracy of the impairment test in AT&T’s securities filings. See id. at
67–68.
26
extended those projections into the future based on historical growth rate trends and
AT&T’s strategic goals. Kobos Dep. 50–51. The final Ten-Year Plan contained forecasts
for key metrics such as wireless service revenue, wireless penetration, AT&T’s market
share, average revenue per user (“ARPU”), capital budget needs, and operating profit. See
JX 362 at 2.
The Ten-Year Plan occasionally was used in presentations for senior AT&T
executives. But unlike the Three-Year Plan, the Ten-Year Plan did not have to be approved
by AT&T’s board of directors. Mobility’s Financial Planning Team did not use the Ten-
Year Plan. Paoletti Dep. 38. AT&T did not use the Ten-Year Plan to set incentive
compensation for directors and officers. Kobos Dep. 111.
3. The Relative Merits Of The Plans
The parties have engaged in vigorous debate over the relative merits of the Three-
Year Plan and the Ten-Year Plan. The plaintiffs prefer the Three-Year Plan, which
projected greater near term growth. AT&T prefers the Ten-Year Plan, which anticipated
more moderate growth.
The Three-Year Plan is generally more credible and reliable than the Ten-Year Plan.
It resulted from a more reliable and detailed process, it was approved by AT&T’s board of
directors, and it was used more extensively in AT&T’s business planning and operations.
AT&T measured its performance against the Three-Year Plan, including for purposes of
executive compensation.
The Ten-Year Plan is relatively less credible and reliable. It primarily resulted from
a top-down process, it was not approved by AT&T’s board of directors, and it was not used
27
as extensively in AT&T’s business. That is not to say that the Ten-Year Plan is unreliable.
It still warrants consideration, but it is less persuasive than the Three-Year Plan.10
When PwC prepared its valuation of the Partnership, only the 2009 versions of the
Three-Year Plan and the Ten-Year Plan were available. See JX 749. AT&T prepared the
2010 Ten-Year Plan and provided it to PwC late in the valuation process. See JX 749; JX
750; JX 2419 at 73 n.198. Gilcreast did not recall using it. See Gilcreast Dep. 225.
K. PwC Completes Its Valuations.
When PwC began work in February 2010, AT&T gave PwC a deadline of March
29, 2010, for its valuations. See JX 867 at ’251–52. During the negotiations over PwC’s
engagement letter, AT&T even asked Gilcreast to reduce the amount of time his team
would spend on the valuations. Gilcreast obliged. JX 706.
On March 10, 2010, the PwC team met with Wages and Suma Gopalan, the regional
business operations director for the Partnership. The meeting focused on basic
characteristics of the Partnership’s business, such as the size of the Salem Market, the rate
10
AT&T has attempted to argue that its actual performance mirrored the Ten-Year
Plan, making the Ten-Year Plan more reliable. That is classic hindsight bias. As the
plaintiffs correctly point out, a wide range of causes and conditions contributed to AT&T’s
real-world performance over that ten year period. At the time of the Freeze-Out, the Three-
Year Plan provided a more reliable projection of AT&T’s future than the Ten-Year Plan.
The plaintiffs ask the court to disregard the Ten-Year Plan entirely. They argue that
the Ten-Year Plan was created for AT&T’s asset impairment testing, which the plaintiffs
contend was an inherently conservative exercise designed to produce stable valuations for
AT&T’s spectrum licenses. That argument was not persuasive. The Ten-Year Plan was a
legitimate planning document that can provide insight into AT&T’s views for the years
beyond 2013.
28
of wireless penetration, the Partnership’s ARPU, and the Partnership’s “churn”—a
measure of the number of the Partnership’s subscribers who disconnected their service
during a given period. See JX 1075. The meeting thus addressed the metrics for valuing
the Partnership as if it were a mini-independent wireless company that only offered
traditional wireless services in the Salem Market, rather than as an integrated part of
AT&T’s nationwide network. Gopalan believed that growth rates for the Partnership’s
inputs should “follow the overall AT&T trend.” Id. at 8–9.
On Friday March 26, 2010, PwC met with employees in AT&T’s Corporate
Development Department. See JX 887; JX 889. AT&T has maintained in this proceeding
that no one instructed or pressured PwC to make changes to its valuation, but the evidence
shows that PwC made changes to its model after the meeting. In fact, over the weekend,
PwC indicated that while it could provide the first draft report on March 30, it would take
time to incorporate all the changes that the Corporate Development Department wanted.
Gilcreast proposed that AT&T review the first report and provide comments so that PwC
could incorporate the input in the other reports. JX 887.
Coincidentally, the first report was for the Partnership. PwC concluded that the
value of the Partnership was $219 million. PwC reached this conclusion by giving equal
weight to a DCF analysis and a comparable companies analysis. The DCF analysis
generated a value of $208 million. The comparable companies analysis generated a value
of $242 million. PwC averaged the two, added the Partnership’s net cash, and subtracted
the value of the AT&T-owned spectrum that the Partnership used free of charge. JX 894.
29
During April and May 2010, PwC sent AT&T its final valuations. Each used March
30, 2010, as the valuation date.
PwC’s final valuation for the Partnership was unchanged at $219 million. AT&T
used that valuation to set the price it paid in the Freeze-Out.
The total valuation for the minority partners’ interests across all the partnerships
was $130 million. That was much lower than the $243 million of avoided distributions that
AT&T had estimated in 2008 that the buyouts would unlock. See JX 1123a.
After receiving the valuations, Stephens sought approval from Stephenson, AT&T’s
Chairman and CEO, to eliminate the minority partners. In connection with the meeting,
AT&T executives sent Stephenson a memorandum that quantified the administrative,
audit, and tax savings that AT&T would achieve. The memorandum also identified a
benefit in the form of a “[d]ecreased distribution payment stream.” JX 3514 at ’575. The
memorandum conspicuously failed to quantify the decreased distributions. AT&T
necessarily knew the value of that component, because AT&T observed that eliminating
the minority partners “will be slightly accretive to AT&T’s earnings and free cash flow per
share.” Id. Knowing that the number would become a focal point of any litigation
challenging the transactions, AT&T omitted any direct reference to the most significant
financial benefit. On June 3, 2010, Stephenson approved the Freeze-Out and the
transactions involving the Partnership’s sister entities.
L. AT&T Continues To See Future Growth In Wireless Data.
While the valuation process was going on, AT&T continued to stress that its future
lay in wireless data. In its annual report for 2009, issued on February 25, 2010, AT&T
30
described mobility services as “the driving force behind our industry’s growth.” JX 3812
at 2. AT&T identified Mobility as its “No. 1 investment priority.” Id. at 4. The report
highlighted AT&T’s leadership in smartphones and emerging devices, which the report
described as “the next wave of growth in wireless.” Id. at 2.
In March 2010, Ralph de la Vega, the CEO of Mobility, predicted that Connected
Devices would “become a billion dollar business” for AT&T “in a few years.” JX 878.
Later in spring 2010, Lurie outlined Mobility’s vision for the future of wireless:
[T]he most key message is that going forward, everything is going to be
wirelessly enabled. Every device, anything that has a current running through
it, is going to have wireless connectivity, which is exactly what people expect
[and] want, and it’s going to raise the value of all of the products and it’s
going to change our lives. The takeaway is that that is not something that’s
going to happen in the future; it’s happening right now.
JX 990 at 0:22–0:40; see also JX 853. AT&T expected robust growth in “data plans and
data usage” and believed its network was uniquely positioned to capitalize on the coming
wave of wireless activity. JX 2432 at ’797.
By summer 2010, consolidation in the wireless industry had resulted in four major
carriers—AT&T, Verizon, Sprint, and T-Mobile—dominating 90% of the U.S. market.
Taylor Report at 10; see JX 1209 at ’044. Regional wireless carriers such as US Cellular,
MetroPCS, and Leap Wireless accounted for the balance. Taylor Report at 10. By this
point, the legacy wireless voice business was relatively mature. See JX 1209 at ’045; JX
1239; JX 1330. Data-related businesses represented the future. See JX 1239 at 11.
Reflecting a decline in voice revenue, AT&T’s earnings for the second quarter of
2010 fell short of expectations for revenue growth and new subscribers. JX 1312 at ’160;
31
see JX 1310; JX 1330 at ’079. AT&T also faced the loss of iPhone exclusivity in 2011. See
JX 1311 at 12. AT&T predicted that overall wireless revenue would “continue[] to grow
but at a slower rate.” Id. at 17.
Reflecting the promising future for data revenue, smartphone penetration continued
to increase. By summer 2010, smartphone penetration had reached 28%, up from 21% in
the fourth quarter of 2009. Taylor Report at 13. AT&T also expected continued growth in
Connected Devices. JX 1312 at ’165. In the second quarter of 2010, AT&T added nearly
900,000 Connected Devices, bringing the total number on AT&T’s network to
approximately 6.7 million. JX 1301 at 1. In a press release, AT&T reported that “[s]ince
forming a dedicated organization to focus on wirelessly connecting new categories of
emerging devices in late 2008, AT&T has emerged as the clear industry leader in one of
the wireless industry’s fastest growing areas.” Id. at 2. The press release also described
AT&T’s plans to expand into new Connected Device categories. Id. at 3.
M. AT&T Changes Its Pricing Models.
With wireless data usage expanding, AT&T took steps to adopt pricing models that
would replace the “all-you-can-eat” plans that its customers were enjoying. At an investor
conference in March 2010, Stephenson previewed a tiered pricing model in which
“intensive Internet users . . . pay higher monthly fees than other customers.” JX 801. In
June 2010, AT&T stopped offering unlimited data plans for iPhone and iPad subscribers,
replacing them with a tiered pricing model under which the customer purchased a base
amount of data and paid for excess data usage. See JX 1057 at ’980–81; JX 1259 at 2–4;
32
JX 1384 at 3–4. AT&T viewed this move as the beginning of a transition to a different
industry pricing model. JX 1057 at ’979. Analysts reacted favorably. See JX 1259 at 2–4.
AT&T also changed how it billed for intercompany data roaming. Starting in July
2010, AT&T paid the Partnership a per-kilobyte data fee if any AT&T customer’s wireless
device pinged off a cellular antenna belonging to the Partnership. AT&T also charged the
Partnership a per-kilobyte fee each time a Partnership subscriber’s wireless device pinged
off an AT&T antenna that did not belong to the Partnership. The amount of the per-kilobyte
fee “was determined on a cost basis by kilobytes used, including network system costs,
interconnect/facilities costs, and network property taxes.” JX 2166 at 16.
By allocating intra-carrier data roaming and expense based on cost, AT&T treated
the Partnership like a third party service provider. The Partnership was a net loser under
the new roaming model, and AT&T determined in late 2010 that the new system had
resulted in a net loss of $300,000 for the Partnership since its implementation in July. See
JX 1872d, “Data Settlement” tab; JX 2166 at 16. Part of the net loss resulted from AT&T’s
decision to systematically charge a rate for incollect data roaming expense that was 0.38%
higher than it paid for outcollect data roaming revenue.11 The unfavorable rate differential
11
Compare JX 1872b, “Incollect Costs - DATA” tab, at Cells G163:H163 (AT&T
charged the Partnership $0.00004276593 and $0.00004251361 per kilobyte in July and
August 2010, respectively, for intra-carrier data roaming), with id., “Outcollect Revenue -
DATA” tab, at Cells G163:H163 (AT&T paid the Partnership $0.0000423005 and
$0.0000423336 in July and August 2010, respectively, for intra-carrier data roaming).
33
was less severe for the Partnership than for its sister entities, where the average rate
differential was -4.87%.
N. AT&T Eliminates The Minority Partners.
In July 2010, AT&T offered to purchase the minority partners’ interests at a 5%
premium over the PwC valuation. AT&T’s offer letter informed the minority partners that
if they did not accept, then AT&T would convene a meeting of partners and vote its interest
in favor of selling the Partnership’s assets and liabilities at the PwC valuation. The letter
explained that after the sale, the Partnership would dissolve and each remaining partner
would receive a payment equal to its pro rata share of the purchase price. In other words,
minority partners who declined the buyout offer would receive their pro rata share, without
the 5% premium.
Some minority partners accepted AT&T’s offer. When AT&T began the freeze-out
process, the Partnership had twelve minority partners holding a total minority interest of
3.078%.12 Seven of the twelve minority partners rejected the offer, and they held an
aggregate interest of 1.881%. That meant that a 58.33% majority of the minority partners
by number and a 61.11% majority of the minority by interest rejected AT&T’s offer, even
though it provided a 5% premium over the Freeze-Out price. See JX 2569 at ’038.
12
These figures are obviously much lower than the 110 minority partners who held
a 49.9% minority interest when the Partnership was formed. The record does not provide
detailed information about the changes.
34
In September 2010, AT&T caused New Salem to offer to purchase all of the
Partnership’s assets and assume all of its liabilities for a cash payment of $219 million. On
October 12, 2010, AT&T convened a special meeting to vote on the offer. At the meeting,
AT&T voted its interest in favor of the Freeze-Out. The minority partners in attendance
voted against the Freeze-Out. JX 1485; see PTO ¶ 66.
Immediately after the vote, AT&T caused the Partnership to enter into an asset
purchase agreement with New Salem. Immediately after that, the Partnership and New
Salem entered into a bill of sale and assignment and assumption agreement that effectuated
the transfer of the Partnership’s assets and liabilities. Wages then executed a “Statement of
Dissolution,” also dated October 12, 2010, which recited that the Partnership had dissolved
as a result of completing a sale of all of its assets. JX 1493. That same day, AT&T sent
checks to the minority partners reflecting their pro rata share of the Freeze-Out price.
Roughly contemporaneous with the Freeze-Out, AT&T engaged in similar
transactions at five other partnerships. Over the following months, AT&T engaged in
similar transactions involving seven additional partnerships, bringing the total number of
freeze-outs to thirteen.
O. Litigation Commences.
On August 26, 2011, the former minority partners in the Partnership who were
eliminated in the Freeze-Out sent AT&T a letter asserting that AT&T had breached the
Partnership Agreement when engaging in the transaction. The former minority partners
gave AT&T thirty days to cure its breach. JX 2030 at 2; see PTO ¶ 70.
35
On September 23, 2011, AT&T filed a lawsuit seeking declaratory relief against the
seven minority partners in the Partnership who were eliminated in the Freeze-Out. C.A.
No. 6886-VCL, Dkt. 1. AT&T’s complaint sought broad declarations absolving AT&T of
any contractual or fiduciary liability for its actions. On October 4, 2011, five of the former
minority partners in the Partnership filed suit against AT&T. C.A. No. 6908-VCL, Dkt. 1.
Either AT&T, the minority partners, or both filed lawsuits addressing the other
twelve freeze-outs. In total, fifteen lawsuits were filed challenging thirteen different
transactions, with minority partners appearing as the plaintiffs in some and as defendants
in others. To achieve a measure of consistency, the court realigned all of the minority
partners as plaintiffs and AT&T and its affiliates as defendants.
As a result of the court’s orders, the following minority partners are the plaintiffs
for purposes of claims relating to the Partnership. The chart identifies their minority interest
in the Partnership at the time of the Freeze-Out.
Salem Minority Partners Interest
Alan R. Bell 0.3420%
Michael T. Bowers 0.1710%
The Ronald J. Gotchall Living Trust (Rosa Lee 0.1710%
Gotchall, Trustee)
The Rosa L. Gotchall Living Trust (Ronald J. 0.3420%
Gotchall, Trustee)
Om Parkash Kalra 0.3420%
Ellen M. Martin 0.1710%
Roam-Tel Partners 0.3420%
PTO ¶ 9. Based on the consideration that AT&T paid for the assets and liabilities of the
Partnership in the Freeze-Out, the plaintiffs received $4,119,390 in the aggregate.
36
P. The Coordination Order
Initially, the fifteen actions proceeded separately, albeit with the parties making
parallel moves across the lawsuits. In June 2012, AT&T filed motions to dismiss the claims
which asserted that AT&T failed to comply with the partnership agreements when
effectuating the freeze-outs. In March 2013, the court issued a series of orders which held
that AT&T had the power to effectuate the freeze-outs.
In June 2013, the court entered a stipulated order coordinating the actions for
purposes of pre-trial discovery. The actions were not formally consolidated. The court
directed the parties to make all filings in the Coordinated Action and designate them as
pertaining to all of the coordinated actions or to specific cases.
Q. Discovery
Discovery unfolded over the better part of eight years. During the process, AT&T
aggressively resisted discovery, even after the court ruled against AT&T on specific issues.
As the court noted on several occasions, AT&T was the most obstructive litigant that this
judge has ever seen, whether in private practice or on the bench. Interested readers may
consult a prior decision for a description of how the litigation unfolded. See Salem Contract
Decision, 2021 WL 4438046, at *32–40.
In December 2020, the case proceeded to trial. The initial coordination order had
provided for coordination only for purposes of pre-trial discovery. In the pre-trial order,
the parties agreed to a coordinated trial.
The parties tried both the plaintiffs’ claims for breach of the partnership agreements
and the plaintiffs’ claims for breach of fiduciary duty. Both sides relied heavily on experts.
37
Carlyn Taylor served as AT&T’s valuation expert. Taylor is a Senior Managing Director
in the Corporate Finance Group of FTI Consulting, Inc., a publicly traded financial and
economic consulting firm. Taylor specializes in the telecommunications and media
industries. J. Armand Musey served as the plaintiffs’ wireless industry expert. He is the
President and founder of Summit Ridge Group, LLC, a consulting firm that serves clients
in the telecommunications, media, and satellite industries. Lorraine Barrick served as the
plaintiffs’ valuation expert. She is a Certified Public Accountant and an accredited
appraiser who operates her own consulting firm.
After considering the post-trial briefing, the court decided to issue two bellwether
decisions, both involving the Partnership. The Salem Contract Decision addressed the
claims for breach of the Partnership Agreement. This decision addresses the claims for
breach of fiduciary duty.
II. WHETHER THE FREEZE-OUT WAS ENTIRELY FAIR
The plaintiffs asserted that AT&T breached its fiduciary duty of loyalty to the
minority partners by effecting the Freeze-Out. AT&T controlled the Partnership, and
AT&T stood on both sides of the Freeze-Out. As a result, AT&T bore the burden of proving
that the terms of the Freeze-Out were entirely fair. AT&T failed to carry its burden. Instead,
AT&T breached its duty of loyalty by failing to follow a fair process and by imposing an
unfair price. The plaintiffs are therefore entitled to an award of damages. See Part III, infra.
38
A. The Entire Fairness Test
When determining whether a fiduciary has breached its duties, Delaware courts
evaluate the fiduciary’s conduct using a standard of review. The parties agree that the
Freeze-Out is subject to the entire fairness standard of review.
When the entire fairness test applies, “the defendant fiduciaries bear the burden of
showing that the challenged decision was entirely fair” to the plaintiffs. In re Energy
Transfer Equity, L.P. Unitholder Litig., 2018 WL 2254706, at *18 (Del. Ch. May 17,
2018). To meet its burden, the defendant must establish “to the court’s satisfaction that the
transaction was the product of both fair dealing and fair price.” Cinerama, Inc. v.
Technicolor, Inc., 663 A.2d 1156, 1163 (Del. 1995) (cleaned up). “Not even an honest
belief that the transaction was entirely fair will be sufficient to establish entire fairness.
Rather, the transaction itself must be objectively fair, independent of the [defendant
fiduciary’s] beliefs.” Gesoff v. IIC Indus., Inc., 902 A.2d 1130, 1145 (Del. Ch. 2006).
“The concept of fairness has two basic aspects: fair dealing and fair price.”
Weinberger v. UOP, Inc., 457 A.2d 701, 711 (Del. 1983). The fair dealing inquiry
“embraces questions of when the transaction was timed, how it was initiated, structured,
and negotiated, and how the transactional approvals were obtained.” Id. The fair price
inquiry focuses on the economic and financial considerations of the challenged transaction.
Id. “[T]he test for fairness is not a bifurcated one as between fair dealing and price. All
aspects of the issue must be examined as a whole since the question is one of entire
fairness.” Id.39
Because the entire fairness test is not bifurcated, “the two aspects of the entire
fairness standard interact.” In re Dole Food Co., Inc. S’holder Litig., 2015 WL 5052214,
at *34 (Del. Ch. Aug. 27, 2015). “A strong record of fair dealing can influence the fair
price inquiry, reinforcing the unitary nature of the entire fairness test. The converse is
equally true: process can infect price.” Reis v. Hazelett Strip-Casting Corp., 28 A.3d 442,
467 (Del. Ch. 2011) (collecting authorities).
B. Fair Dealing
The first dimension of the unitary entire fairness test involves the issue of fair
dealing. AT&T failed to prove that it dealt fairly with the minority partners. The only step
AT&T took towards instantiating a fair process was to hire a financial advisor to value the
Partnership, then use that valuation when setting the price for the Freeze-Out. No special
committee or other independent bargaining agent negotiated on behalf of the minority. No
one had the ability to veto the deal. And contrary to AT&T’s assertions in this litigation,
AT&T acted for the primary purpose of acquiring the minority partners’ interests for less
than the value that the minority partners otherwise would have received in distributions.
1. The Timing And Initiation Of The Freeze-Out
One factor pertinent to the dimension of fair dealing is how the transaction was
timed and initiated. Weinberger, 457 A.2d at 711. AT&T failed to prove that the timing
and initiation of the Freeze-Out were consistent with fair dealing. Instead, AT&T timed
and initiated the Freeze-Out to acquire the minority partner interests before the data
revolution caused the value of the Partnership to increase.
40
AT&T timed the Freeze-Out to take advantage of the data revolution. Although the
market for legacy wireless voice communication was relatively mature by 2007, AT&T
saw the potential for continued growth in data services, driven by new products and
applications.
To capitalize on the data revolution, AT&T had created the Emerging Devices
Organization and was investing heavily in Connected Devices. During the period leading
up to the Freeze-Out, AT&T’s executives consistently stressed the potential of AT&T’s
data services business. On the day after PwC delivered its valuation reports, Lurie gave a
keynote speech in which he described AT&T’s plans to connect “anything that has a
current running through it.” JX 990 at 0:26–0:29. Days later, Lurie highlighted the
“unbelievable change” in wireless penetration rates and expressed his belief that “wireless
penetration could be 5, 6, 700%” by 2013, largely because of Connected Devices and
“wirelessly-enabl[ed] everything.” JX 1011 at 4:19–5:00. In an interview in August 2010,
Stephenson expressed similar statements. When asked about AT&T’s business, he
responded:
We’re in the connectivity business. We like to say that we connect people to
their world. That is basically what we do. And so our objective is to make
sure that you’re connected to your business information needs, to your home
entertainment information needs, to your family, to your associates, that the
machines that you possess that transmit data, that they’re connected to other
machines that are important. And if you think about this, we have this really
terrific world-class global network, [and] hanging off of that network are
well over 1 billion devices that are basically allowing you to connect with
data, machines, and whatnot, and that includes ATMs, and gas pumps, and
so forth. So we’re in the business of connecting people to information. And
data. . . . Our objective . . . is to mobilize everything that you do. Whatever
your connectivity needs are, I want those needs mobilized.
41
JX 1340 at 0:37–1:54.
By freezing out the minority partners, AT&T ensured that it would enjoy 100% of
the anticipated gains from the data-driven businesses and avoid sharing those gains with
the minority partners. AT&T began giving serious consideration to eliminating the
minority partners in late 2007, when the shift towards data services began. After the release
of the iPhone in summer 2007, the cost to AT&T of paying distributions to the minority
partners increased 82% year-over-year. Teske Tr. 611–12; JX 616 at 20. AT&T ultimately
chose not to move forward in 2008, but not because AT&T had any doubts about the data
revolution. Instead, AT&T focused on purchasing additional spectrum and developing new
business lines to take advantage of the shift toward data services. AT&T deferred engaging
in the Freeze-Out and similar transactions involving the Partnership’s sister entities until
2010 and 2011, but AT&T’s principal rationale remained the same: acquire the minority
interest before its value increased.
AT&T’s early analyses of the potential freeze-out transactions were quite candid in
this regard. In a presentation in August 2007, AT&T’s Corporate Development team
explained that “[a] buyout today will be much less expensive than [a buyout] 1 or 2 years
down the road given OIBDA growth rates.” JX 310 at 10. In January 2008, AT&T’s
Corporate Development team again pitched the Freeze-Out as an opportunity to acquire
the minority partners’ interests “at a discount to future growth.” JX 347 at 9.
By the end of 2008, the distributions to the minority partners had nearly doubled
since August of the prior year. See JX 616 at 13; JX 3516, “actualDist” tab. And revenue
from data services continued to grow rapidly. By August 2009, AT&T projected that its
42
investment in the Emerging Devices Organization would deliver an internal rate of return
“over 50% within the 5 year period” from 2010 to 2014. JX 532 at 16. AT&T therefore
again took up the possibility of freeze-out transactions. See JX 587; JX 616.
Recognizing that the early planning documents telegraphed that AT&T wanted to
eliminate the minority partners to capture the value of the minority interests before they
increased, AT&T contends that the 2007–2008 effort was separate and distinct, with no
bearing on the transactions that occurred two years later. See Dkt. 614 at 31. Citing ACP
Master, Ltd. v. Sprint Corp., AT&T claims that the delay between 2007 and 2008 and the
actual freeze-outs “freshened the atmosphere.” Id. at 31–32 (quoting 2017 WL 3421142,
at *29 (Del. Ch. July 21, 2017), aff’d, 184 A.3d 1291 (Del. 2018) (TABLE)). The events
in ACP bear no similarity to this case. There, a controlling stockholder was pursuing a
freeze-out merger when a third party submitted a topping bid for the controlled corporation.
ACP, 2017 WL 3421142, at *9. When the target failed to engage, the third party made a
bid for the controller. Id. at *11. With the controller continuing to pursue the freeze-out, a
group of four large stockholders banded together to oppose it. Id. In the end, the controller
completed the freeze-out but was forced to pay approximately 70% more than the price it
originally proposed. Despite serious problems with the fairness of the initial transaction,
the court found that the third party’s intervention and the stockholder group’s resistance
“freshened the atmosphere and created a competitive dynamic” that resulted in the eventual
transaction being fair. Id. at *29.
Nothing like that happened here. There was no third-party bidder. There was no
ability of minority investors to push back on AT&T. There was no price competition.
43
AT&T simply delayed the buyouts. When AT&T started considering them again, the same
members of the Corporate Development Team were intimately involved. See, e.g., JX 558;
JX 562; JX 568 JX 569; JX 572; JX 582; JX 585; JX 597; JX 611. They followed the same
plan, and they based their analyses on their earlier work. See JX 565; JX 577.
AT&T’s attempt to distance the Freeze-Out from earlier efforts is an invention of
litigation counsel, whom AT&T brought in when it revisited the buyouts to sanitize the
record. As early as October 1, 2009, the Corporate Development Team circulated a review
of “prior court cases” involving partnership freeze-outs. JX 568. AT&T’s litigators
participated in discussions with the Corporate Development team about how to prepare
valuations and move forward. JX 3587. AT&T even had its litigators vet Gilcreast as a
potential witness before hiring PwC. See Gilcreast Dep. 45; JX 689 at ’424; see also JX
691.
As part of its strategy for eventual litigation, AT&T sought to associate the buyouts
with Project LESS, and AT&T has argued in this case that it did not act selfishly to capture
value from the minority partners but merely to simplify its complex corporate structure and
reduce administrative costs. Project LESS was indeed an ongoing effort in that direction,
and the Freeze-Out and its sister transactions did have those effects. But they did not
provide AT&T’s primary motivation. AT&T acted because it anticipated a period of
44
significant growth in data-driven wireless businesses, wanted 100% of the benefits for
itself, and did not want to share the benefits with the minority partners.13
The relative magnitude of the shared administrative savings and the avoided
distributions reveals what really drove AT&T’s decision. In an analysis of a buyout of
twenty-one market-level entities, including the Partnership, AT&T projected that it would
save $725,000 annually in administrative expenses, and AT&T ascribed a present value of
nearly $7 million to those savings. JX 3516, “AdminSavings” tab; accord JX 3514 at ’575.
In the same analysis, AT&T projected that it would save at least $13.43 million annually
in distributions to the minority partners, and AT&T ascribed a present value of $128.4
million to those savings. JX 3516, “actualDist” tab. The avoided distributions were
eighteen times larger than the administrative savings, and the avoided distributions were
based on the actual distributions in 2009, which were lower than the distributions to
minority partners in 2008. In addition, AT&T would gain a tax benefit by eliminating the
minority partners in the form of a step up in basis on their member interests. See JX 3514
at ’580. Across all of the partnerships where AT&T eliminated the minority partners,
AT&T estimated the present value of the tax benefits at between $16 million and $18.9
million, more than double the administrative savings. Id. The non-administrative benefits
13
See, e.g., JX 344 at 11 (buyout presentation slide calculating the “DCF Value of
Avoided Distributions”); JX 616 at ’952 (AT&T analyzing the growth in minority
distribution streams as a benefit of pursuing buyouts); JX 1473 (same); JX 616 at ’951
(valuing the “[e]limination of distribution payment stream”); JX 3514 at ’575 (citing
benefit of “[d]ecreased distribution payment stream”).
45
thus provided approximately twenty times the value of the administrative savings that
AT&T now claims drove its decision.
When planning the freeze-outs, AT&T specifically focused on the Partnership,
sixteen sister partnerships, and four corporations because they were low-hanging fruit. See
Teske Tr. 581–83. The Corporate Development team evaluated buyouts at fifty-one
market-level entities. They targeted the Partnership and its sister entities because the
minority investors could be eliminated unilaterally. JX 2522 at 2; see JX 310 at ’765–66.
In response, AT&T has offered more counterfactual assertions. First, AT&T claims
that no one expected a sudden increase in the value of the wireless business such that AT&T
could not have acted to claim a near-term burst in value. That is true but irrelevant. No one
is suggesting that AT&T spring-loaded an option. As depicted by the following graphic,
AT&T anticipated substantial growth over time:
46
JX 1902 at 9. AT&T anticipated highly profitable growth in its wireless business, and
AT&T sought to secure that value for itself.
Next, AT&T has argued that the wireless business requires significant investment,
including regular upgrades, so the wireless business was not all upside. No one disputes
that, and it is not a meaningful response. AT&T saw significant value in the wireless
business net of the necessary investment. Moreover, AT&T was already funding the
investment because it operated the Partnership and its sister entities as part of its nationwide
wireless network. AT&T was thus sharing the fruits of its investment with the minority
partners. AT&T wanted to capture 100% of the gains.
Finally, AT&T has argued that it could not have benefitted from the data revolution
because most of its wireless customers had purchased “all you can eat” plans that offered
unlimited usage. That is another effort at misdirection. At the time of the freeze-outs,
AT&T was changing its pricing policies so it could benefit from increased data usage. See
JX 801; JX 1057 at ’980–81; JX 1384 at 3–4.
AT&T timed the Freeze-Out to eliminate the minority partners before the data
revolution would increase the value of AT&T’s wireless business and the distributions that
the minority partners would receive. AT&T chose the Partnership and its sister entities so
AT&T could control the transaction timeline. The timing and initiation of the Freeze-Out
are evidence of unfair dealing.
47
2. The Negotiation And Structure Of The Freeze-Out
Fair dealing also examines how the transaction was negotiated and structured.
Weinberger, 457 A.2d at 711. In this case, there was no negotiation, and the structure of
the Freeze-Out was not consistent with fair dealing.
No one negotiated the Freeze-Out. The Partnership had a minority representative on
the Executive Committee, and the Executive Committee could have empowered the
minority representative to negotiate.14 AT&T did not engage with the minority
representative. Instead, AT&T kept the minority representatives at the Partnership and its
sister entities in the dark by claiming during the same period when AT&T was using
confidential information belonging to those entities to plan the freeze-outs that there was
no other business for the Executive Committee to discuss. See, e.g., JX 1080. AT&T did
not involve the minority representatives in the process at all, then justified its actions on
14
Cf. S. Muoio & Co. LLC v. Hallmark Ent. Invs. Co., 2011 WL 863007, at *9–10
& n.73 (Del. Ch. Mar. 9, 2011) (holding process was entirely fair where, among other
things, “the Special Committee was independent, fully informed, and . . . negotiated . . . at
arm’s length.”), aff’d, 35 A.3d 419 (Del. 2011) (TABLE); In re John Q. Hammons Hotels
Inc. S’holder Litig., 2011 WL 227634, at *2–3 (Del. Ch. Jan. 14, 2011) (noting that use of
“independent and disinterested” special committee that had a “duty to reject any offer that
was not fair to the unaffiliated stockholders” supported finding that process was entirely
fair); In re Cysive, Inc. S’holders Litig., 836 A.2d 531, 553–56 (Del. Ch. 2003) (observing
that use of independent committee who “negotiated . . . aggressively” and “retained the
flexibility to abandon the [agreement] in favor of a better deal” supported finding that
process was entirely fair); see also Ams. Mining Corp. v. Theriault, 51 A.3d 1213, 1243–
44 (Del. 2012) (“[J]udicial review for entire fairness of how the transaction was structured,
negotiated, disclosed to the directors, and approved by the directors will be significantly
influenced by the work product of a properly functioning special committee of independent
directors.”).
48
the theory that the representative’s involvement was not required. See, e.g., JX 1623 at
’765.
AT&T also did not condition the Freeze-Out on a majority-of-the-minority vote. A
controller is not required to provide minority investors with a majority-of-the-minority
vote, and a controller may be able to show fair dealing without providing such a vote, but
the absence of a vote still figures into the analysis.15 At best for AT&T, the absence of a
majority-of-the-minority vote is not a negative factor suggesting unfairness. The absence
of a majority-of-the-minority vote is equally not a positive factor suggesting fairness. With
AT&T bearing the burden of proving entire fairness, the absence of a majority-of-the-
minority vote does not move the needle in AT&T’s favor.
In addition to failing to take steps to protect the minority, AT&T engaged in tactics
designed to induce the minority to sell in transactions that would not be subject to fiduciary
review. Before the Freeze-Out, AT&T told the minority partners that either they could
accept an offer from AT&T at a 5% premium to the PwC valuation, or they would be frozen
out at the value set by PwC. AT&T thus created a coercive, two-tier offer that pressured
15
See Hallmark, 2011 WL 863007, at *15 (considering absence of majority-of-the-
minority vote but holding transaction process was fair when special committee “dropped
the majority-of-the-minority condition near the end of the negotiations, in exchange for
other favorable concessions”); Van de Walle v. Unimation, Inc., 1991 WL 29303, at *14
(Del. Ch. Mar. 7, 1991) (considering absence of majority-of-the-minority vote and holding
that process was fair in part because “a majority of the minority shares actually voted” were
voted in favor of transaction); Jedwab v. MGM Grand Hotels, Inc., 509 A.2d 584, 599–
600 (Del. Ch. 1986) (citing absence of majority-of-the-minority vote and independent
committee approval as “pertinent factors in assessing whether fairness was accorded to the
minority” but which were outweighed by other indicia of fairness).
49
the minority to accept the front-end price or else be cashed out at a lower price. A coercive
structure is evidence of unfair dealing. See, e.g., Kahn v. Lynch Commc’n Sys., Inc., 638
A.2d 1110, 1120–21 (Del. 1994) (finding controller’s threat that rejecting offer would lead
to hostile transaction at lower price meant there was no “semblance of arm’s length
bargaining”).
Even with its coercive structure, a majority of the minority partners rejected
AT&T’s offer. As discussed in the Factual Background, a 58.33% majority of the minority
partners by number and a 61.11% majority of the minority by interest rejected AT&T’s
offer, notwithstanding its coercive structure. See JX 2569 at ’038. That is strong evidence
that the offer was unfair, even with the 5% premium.
There also is evidence that AT&T provided false answers and refused to respond to
minority partners’ questions during the special meetings at which AT&T voted its
controlling interest to approve the Freeze-Outs. At the special meeting of the Bradenton
partnership, a minority partner asked if AT&T had other valuations of the partnership. JX
1703 at 2. Wages falsely responded that AT&T did not. Id. At the same meeting, a minority
partner asked if PwC had “other on going [sic] business with AT&T.” Id. Instead of
answering the question, Wages “explained the procedure for selecting PwC.” Id. AT&T
never informed the minority partners about its longstanding relationship with Gilcreast, the
PwC partner who led the valuation team.
AT&T argues that it dealt fairly with the Partnership because it engaged in a
transaction that was permitted by the Partnership Agreement. That argument ignores the
difference between equitable and legal review. Under Delaware law, the actions of a
50
fiduciary are “twice-tested, first for legal authorization, and second for equity.” Bäcker v.
Palisades Growth Cap. II, L.P., 246 A.3d 81, 96 (Del. 2021) (cleaned up). Minority holders
“can entrust [their fiduciaries] with broad legal authority precisely because they know that
that authority must be exercised consistently with equitable principles of fiduciary duty.”
Id. at 97 (cleaned up). “[I]nequitable action does not become permissible simply because
it is legally possible.” Schnell v. Chris-Craft Indus., Inc., 285 A.2d 437, 439 (Del. 1971).
Fair dealing does not turn on whether AT&T did the bare minimum that the law or the
Partnership Agreement required. The fair dealing inquiry looks for steps designed to ensure
fairness to the minority. Here, AT&T did not take any steps to ensure fair dealing.
3. The Use Of A Financial Advisor To Set The Price
The inquiry into fair dealing also examines other factors relating to the setting of
the price, such as the independence of any financial advisor involved in the transaction and
the quality of its valuation opinions. Cases typically evaluate the independence and analysis
of a financial advisor that a special committee has retained to assist in its negotiations with
the controller.16 Cases have not given significant weight to the financial advisor that the
buyer hires to help it set the price. But because AT&T has relied on its retention of PwC
16
See Hallmark, 2011 WL 863007, at *13–15 (independent committee’s use of
valuations and transaction advice by “two independent financial advisors” supported
finding that process was fair); Gesoff, 902 A.2d at 1147 (“As has been repeatedly held,
special committee members should have access to knowledgeable and independent
advisors, including legal and financial advisors.”); In re Tele-Commc’ns, Inc. S’holders
Litig., 2005 WL 3642727, at *10 (Del. Ch. Dec. 21, 2005) (“The effectiveness of a Special
Committee often lies in the quality of the advice its members receive from their legal and
financial advisors.”).
51
as its principal evidence of fair dealing, this decision must focus on that aspect of the
process.
AT&T’s retaining of PwC is hardly a significant step. Buyers routinely hire a
financial advisor for assistance. It would have been striking if AT&T had not retained a
financial advisor to help with the transaction.
AT&T emphasizes that PwC was an independent firm. But AT&T hired and paid
PwC, and PwC knew that it would be creating a valuation for a transaction in which AT&T
was the buyer. As this court’s experience with appraisal cases demonstrates plainly,
valuation professionals reach outcomes that are influenced by the interests of the party that
retains them, even when ostensibly acting as disinterested experts.17 Scholars have reached
the same conclusions.18
17
See In re Appraisal of Dole Food Co., Inc., 114 A.3d 541, 557 & n.10 (Del. Ch.
2014) (collecting cases recognizing the omnipresent “widely divergent, litigation-driven
expert valuations” this court sees in appraisal proceedings).
18
See Dan Elnathan et al., An Analysis of Private Versus Public Firm Valuations
and the Contribution of Financial Experts, 45 Int’l J. Acct. 387, 406–07, 409 (2010)
[hereinafter Elnathan, Contribution of Financial Experts] (presenting evidence that “expert
valuations of private as well as public firms are affected by the identity of the commissioner
[of the valuation]. Specifically, as expected, they seem to coincide with the interests of the
side to the transaction that commissioned the valuation” such that “valuation multiples for
private firms, as well as those for public firms, are lower when the valuation is
commissioned by the buyer than when it is commissioned by the seller”); Dan Elnathan et
al., On the Added Value of Firm Valuation by Financial Experts, 4 Int’l J. Bus. & Mgmt.
70, 71 (2009) (presenting evidence that ostensibly impartial financial experts retained to
value shares “are not impartial; while they are supposed to provide an independent expert
opinion, they seem to be, in fact, biased towards majority shareholders who hired them to
value the firm”). The bias is more pronounced in private firms than in public firms.
Elnathan, Contribution of Financial Experts, supra, at 407.
52
The PwC team had prior relationships with AT&T. Gilcreast led the team, and he
previously worked on AT&T’s acquisition of Cingular while with Standard and Poor’s,
then continued to perform accounting and valuation work for AT&T after Duff & Phelps
acquired his practice group. Gilcreast had developed business relationships with several
AT&T executives through his prior work. After Gilcreast moved to PwC, he was prohibited
from working for AT&T by a noncompetition agreement, but AT&T secured a waiver from
Duff & Phelps so they could use him. During PwC’s engagement, Gilcreast continued to
solicit further business from AT&T. See JX 1814. In light of these ties and events, AT&T’s
hiring of PwC looks less like the engagement of a truly independent outside advisor and
A recent study found “clear evidence for the existence of . . . engagement bias” in
purportedly neutral valuation professionals who were assigned randomly to perform
valuation tasks on behalf of a buyer or a seller. Marc J. R. Broekema et al., Are Business
Valuators Biased? A Psychological Perspective on the Causes of Valuation Disputes, 23
J. Behav. Fin. 23, 34 (2022). The authors found that
[v]aluators appear to be affected by their clients’ interests, such that they
indicate that a valuation should be adjusted in accordance with their clients’
interests. Specifically, when they represent a buyer and therefore have an
incentive to lower the value of the shares, they also indicate the valuation
should be adjusted downwards more heavily and also indicate a lower value
range for the true value of the company. The opposite is the case when they
represented the seller.
Id. The purportedly neutral experts also exhibited a “blind spot” for their own bias:
“Whereas 58.7% believed the valuator representing the opposing party was biased, only
25.1% believed they themselves were biased.” Id. In their original study, the professors
identified reactive devaluation as an alternative explanation; a companion study reported
in the same article rejected that alternative.
53
more like the continuation of a longstanding business relationship with an individual who
knew how to deliver the answer AT&T wanted.
When the analysis moves to what PwC did, the evidence is mixed. To the good, the
PwC team developed a generally sound model. Also to the good, AT&T generally gave
PwC the information it requested. And PwC did not take all of AT&T’s suggestions about
how to value the Partnership. For example, Gopalan recommended that PwC set population
growth for the Salem Market “equal to 1% for all years.” JX 1075 at 7. PwC performed its
own internal analysis and set population growth at 1.5%. JX 1209 at ’052; accord JX 1075
at 7. PwC also declined to accept Teske’s suggestion that the cash flow estimates should
deduct a royalty equal to 4% of the revenue on the theory that an independent entity would
have to pay AT&T for its brand. JX 1209 at ’220.
To the bad, AT&T withheld important pieces of information from PwC, such as
board presentations about the buyouts and information about revenue from Connected
Devices. PwC sought copies of board presentations in its initial data request. JX 728 at
’369; JX 760. Teske claimed that he could not ask for the presentations because they were
prepared for the board of AT&T, while he worked for Mobility. Teske Tr. 552–53, 630–
33. That explanation was not credible. Teske was the designated point of contact for PwC,
and PwC’s job was to create valuations that Teske and his colleagues would use to seek
authorization to spend millions of dollars of AT&T’s capital on buying out the minority
partners. Teske could have obtained the presentations, just as he obtained the Ten-Year
Plan. At a minimum, he could have asked.
54
The logical inference is that Teske did not provide PwC with the presentations
because the information they contained would have been problematic for AT&T, such as
AT&T’s plan to acquire the minority interests “at a discount to future growth.” JX 347 at
9; see JX 310 at 10. AT&T only provided PwC with valuations performed by Duff &
Phelps, whose previous valuation work was “in line” with the price AT&T wanted to pay.
See JX 347 at 9. AT&T thus steered PwC towards its preferred valuation.
Despite PwC’s requests, AT&T failed to provide any detailed information about its
projections for Connected Devices. The 2009 version of the Ten-Year Plan that AT&T sent
to PwC contained estimates for subscribers, but it did not break out revenue. See JX 553.
On March 22, 2010, PwC asked for “details on the updated 10 yr plan, especially as it
relates to connected devices.” JX 850. Gilcreast reiterated the request the next day. JX 867
at ’251. Teske responded that AT&T “[a]greed that it is something to consider, but [AT&T]
doesn’t want there to be any delay.” Id. He added that “there appears to be . . . an
overwhelming focus on timeline right now.” Id. In its final valuation report for the
Partnership, PwC stated AT&T had failed to provide any “data or basis for including the
impact of connected devices in the forecasts.” JX 1209 at ’223–24.
By not giving PwC detailed information about Connected Devices, AT&T limited
PwC’s ability to value a business that AT&T regarded as central to Mobility’s future. In
his deposition, Gilcreast posited that AT&T rebuffed PwC’s requests because “the industry
as a whole was still trying to get its head around what the potential for [Connected Devices]
was.” Gilcreast Dep. 294. But AT&T had the data, and AT&T executives were touting the
55
value of Connected Devices in their public statements. The logical inference is that AT&T
withheld the data to get the transactions done at a lower price.
AT&T also dictated that PwC use a higher tax rate for the Partnership than PwC
otherwise would have applied. PwC suggested that because the partnerships were pass-
through entities, it should use the “synthetic” tax rate developed by this court in Delaware
Open MRI Radiology Associates, P.A. v. Kessler, 898 A.2d 290 (Del. Ch. 2006). JX 892;
see Gilcreast Dep. 242–44. AT&T instructed PwC to use AT&T’s higher corporate tax rate
of 38.5% without providing any explanation. See Teske Tr. 668–70; Gilcreast Dep. 243–
45; JX 893. Using a lower tax rate would have increased PwC’s valuation conclusion.
In addition to these specific items, AT&T’s Corporate Development Team met with
PwC and reviewed the valuations in detail, resulting in changes to the model. Not
surprisingly, PwC ended up valuing the Partnership at “the lower end” of the range that the
Corporate Development Team had developed. See JX 1323.
The record regarding PwC’s involvement is mixed. On the whole, however,
AT&T’s interactions with PwC provide additional evidence of an unfair process.
4. The J&J Celcom Case
AT&T finally argues that it followed a fair process because it took steps that
mirrored a partnership buyout process from an earlier case in which a federal district court
found no violation of the duty of loyalty under Washington law. See J&J Celcom v. AT&T
Wireless Servs., Inc., 2005 WL 1126924 (W.D. Wash. May 10, 2005), aff’d in part, 215 F.
App’x 616 (9th Cir. 2006). AT&T’s reliance on J&J Celcom does not support the fairness
of the process in this case.
56
First, the court in J&J Celcom applied Washington law. See J&J Celcom v. AT&T
Wireless Servs., Inc. (J&J Appeal), 215 F. App’x 616 (9th Cir. 2006). Washington
partnership law imposes a duty of loyalty that “is limited in scope.” J&J Celcom, 2005 WL
1126924, at *12. Under Washington law, “a partner may engage in conduct that would
violate the duty of loyalty if the partner discloses all the material facts and all of the partners
or a number or percentage specified in the partnership agreement authorize the conduct.”
Id. In that scenario, the partner must only show that all material facts were disclosed and
that the price paid “was fair.” Id. at *13. In other words, there is no requirement of fair
dealing, only full disclosure and fair price.
By contrast, this case requires the court to apply Delaware law, which requires fair
dealing. And unlike Washington’s test, Delaware’s test is not a bifurcated one. Regardless,
this decision has already found that AT&T did not disclose all material facts.
The outcome of J&J Celcom also appears to have turned largely on ineffective
advocacy by the plaintiffs in that case. Counsel for the plaintiffs “represented that he was
unaware of any conflict of law” that would warrant applying the law of a different state,
even though three of the partnerships were Delaware entities. J&J Appeal, 215 F. App’x
at 618, 620. The court also noted that “Plaintiffs’ argument regarding Defendants’ alleged
breach of fiduciary duty is difficult to follow; their response is incomprehensible at points
and their citation to the record lacks clarity and precision.” J&J Celcom, 2005 WL
1126924, at *13. The plaintiffs apparently attacked the fairness of the transaction price by
submitting an expert report challenging AT&T’s methods for allocating revenue and
expense, but the court struck the report “because it was not filed timely.” Id. at *14. The
57
court also found that the plaintiffs had presented “no coherent opinion of what the fair
value of the partnerships . . . should have been” and “no coherent evidence that the values
should have been materially different.” Id. Given these serial failures, the court granted
AT&T’s motion for summary judgment on the plaintiffs’ claim for breach of the duty of
loyalty. In this case, the plaintiffs have successfully developed a deep record that reveals
numerous ways in which AT&T failed to deal fairly with the minority partners.19
5. The Finding Regarding Fair Dealing
AT&T failed to prove that the process that led to the Freeze-Out was fair. AT&T
timed the Freeze-Out to take advantage of the data revolution with the goal of eliminating
its obligation to pay the minority partners their fair share of future distributions. AT&T
employed an outside valuation firm with prior connections to AT&T and negatively
influenced the valuation outcome. AT&T then constructed a coercive offer and imposed
the Freeze-Out unilaterally on the minority partners who did not accept it.
19
AT&T also relies on the United States Court of Appeals for the Ninth Circuit’s
affirmance of the J&J Celcom decision, and on the Supreme Court of Washington’s
confirmation that the federal district court correctly applied Washington partnership law.
The Ninth Circuit certified the question of “whether, under the Revised Uniformed
Partnership Act . . ., a controlling partner violates the duty of loyalty where the controlling
partner causes the partnership to sell its assets to an affiliated party” to the Supreme Court
of Washington. J&J Celcom v. AT&T Wireless Servs., Inc., 169 P.3d 823, 823 (Wash.
2007). The Supreme Court of Washington held that the district court’s factual findings
were dispositive. Id. at 824–25. With that answer in hand, the Ninth Circuit affirmed the
district court’s decision. J&J Celcom v. AT&T Wireless Servs., Inc., 508 F.3d 1177 (9th
Cir. 2007) (Mem.).
58
C. Fair Price
The second dimension of the unitary entire fairness test involves the question of fair
price. AT&T set the Freeze-Out price at $219 million, equal to PwC’s valuation. To derive
that value, PwC prepared a DCF analysis and a comparable companies analysis and gave
50% weight to each approach. AT&T failed to prove that the Freeze-Out price was fair.
When assessing the aspect of fair price in a parent-subsidiary merger, the proper
“test of fairness” is whether “the minority stockholder shall receive the substantial
equivalent in value of what he had before.” Sterling v. Mayflower Hotel Corp., 93 A.2d
107, 114 (Del. 1952); accord Rosenblatt v. Getty Oil Co., 493 A.2d 929, 940 (Del. 1985).
When applying this standard in Weinberger, the Delaware Supreme Court distinguished
between the valuation standard to be applied in the fair price dimension of the entire
fairness inquiry, which is part of the standard of review used to determine whether a
fiduciary is liable for breach, and the potential remedy available in the event of breach,
where the cestui que trust is not limited to an award of fair value. The Delaware Supreme
Court made clear that if the transaction failed to pass muster under the entire fairness test,
then the Court of Chancery would be able “to fashion any form of equitable and monetary
relief as may be appropriate, including rescissory damages.” Weinberger, 457 A.2d at 714.
To prove the fairness of the Freeze-Out price, AT&T did not rely directly on PwC’s
valuation. AT&T did not even call Gilcreast at trial. Instead, AT&T relied on expert
testimony from Taylor, an expert that AT&T has used repeatedly in cases involving freeze-
59
outs.20 The plaintiffs successfully impeached the credibility of the valuation judgments
Taylor made in this case. In addition to the specific issues discussed below, they showed
that when presented with similar valuation issues in a prior case, she reached a different
conclusion. See JX 3552 at 54–55 (valuing step-up in basis in similar freeze-out; affording
no value to step-up in basis in this case). They also showed that in three different cases
involving similar wireless company freeze-outs, Taylor has used three different weighting
schemes.21
Taylor constructed a comparable companies analysis, a comparable transaction
analysis, and a DCF analysis. From these analyses, she generated a valuation range for the
Partnership of $171.34 million to $224.1 million. For purposes of this case, she gave 50%
weight to the DCF analysis and 25% weight to each of the other models. Noting that the
Freeze-Out price of $219 million fell towards the high end of her valuation range, Taylor
opined that the price “represent[ed] at least [the] Fair Value of the Partnership equity
interests.” Taylor Report at 8–9.
20
Taylor appeared as AT&T’s expert in J&J Celcom. 2005 WL 1126924, at *14.
She also served as AT&T’s expert in two cases before this court involving freeze-outs. See
B&L Cellular v. USCOC of Greater Iowa, LLC, 2014 WL 6882207, at *3 (Del. Ch. Dec.
8, 2014); In re AT&T Mobility Wireless Operations Hldgs. Appraisal Litig. (AMWOH),
2013 WL 3865099, at *2 (Del. Ch. June 24, 2013) (ORDER).
21
See Taylor Tr. 968–69; JX 3552 at 5, Ex. 1 (assigning equal weight to a DCF
model and a comparable companies model); JX 3554 at 8, Ex. 1 (assigning 50% weight to
a DCF analysis, 30% to a comparable companies analysis, and 20% to a comparable
transactions analysis); Taylor Report at 121–22, Ex. 49 (assigning 50% weight to a DCF
analysis, 25% to comparable companies analysis, and 25% to comparable transactions
analysis).
60
AT&T’s reliance on its litigation expert failed to overcome the evidence
demonstrating that the Freeze-Out price was unfair. First, the price failed to account for the
value to which the Partnership was contractually entitled under the Management
Agreement, as well as the litigation asset based on AT&T’s past breaches of the
Management Agreement. Second, contemporaneous documents generated by AT&T
indicate that the Partnership was worth considerably more than the Freeze-Out price. Third,
AT&T relied on unpersuasive valuation methodologies.
1. The Failure To Account For The Management Agreement
The first reason why AT&T did not prove that the Freeze-Out price was fair is its
failure to account for the Management Agreement. That agreement contained the Premium
Provision, which obligated AT&T to add a premium of at least 25% when allocating Shared
Revenues to the Partnership. The Premium Provision also obligated AT&T to subtract a
discount of at least 10% when allocating Sales and Marketing Expenses to the Partnership.
AT&T never complied with the Premium Provision.
AT&T’s pervasive breach of the Management Agreement had two consequences
for the fairness of the Freeze-Out. First, the financial information PwC used did not
incorporate either the 25% premium for Shared Revenues or the 10% discount for Sales
and Marketing Expenses. PwC’s analyses comported with AT&T’s historical practice of
ignoring the Management Agreement, including the Premium Provision, but that does not
mean it was accurate. The Partnership’s operative reality included the right to receive the
benefit of the Premium Provision. By failing to account for these contractual rights, PwC’s
analyses undervalued the Partnership.
61
Second, PwC’s analyses did not afford any value to the Partnership’s claim against
AT&T for breach of the Management Agreement. As Chancellor Allen explained, “If the
company has substantial and valuable derivative claims, they, like any asset of the
company, may be valued in an appraisal.”22 The same is true for purposes of the fair price
dimension of the entire fairness test.23 This court has held that AT&T breached the
22
Porter v. Tex. Com. Bancshares, Inc., 1989 WL 120358, at *5 (Del. Ch. Oct. 12,
1989) (Allen, C.); accord In re Cox Radio, Inc. S’holders Litig., 2010 WL 1806616, at *14
(Del. Ch. May 6, 2010) (“Under Delaware law, breach of fiduciary duty claims that do not
arise from the merger are corporate assets that may be included in the determination of fair
value in an appraisal proceeding. Thus, even though the Appraisal Objectors’ claims
related to the propriety of the Transaction are released by the Settlement, any fiduciary
duty claim they may have that is not related to the Transaction, including their claim
challenging the stock repurchase program, is not subject to the settlement’s release and,
thus, can be valued at appraisal.” (cleaned up)), aff’d, 9 A.3d 475 (Del. 2010) (TABLE);
Bomarko, Inc. v. Int’l Telecharge, Inc., 1994 WL 198726, at *3 (Del. Ch. May 16, 1994)
(rejecting argument that litigation of derivative “claims would be an impermissible
expansion of the statutory appraisal remedy” and holding that “breach of fiduciary claims
that do not arise from the merger are corporate assets that may be included in the
determination of fair value”); In re Radiology Assocs., Inc. Litig., 1990 WL 67839, at *13
(Del. Ch. May 16, 1990) (“[C]laims . . . that are derivative in nature and precluded for lack
of standing, may be considered in the appraisal phase of this litigation.”).
23
See Morris v. Spectra Energy P’rs (DE) GP, LP, 246 A.3d 121, 136–38 (Del.
2021) (reversing trial court’s dismissal of claim for breach of fiduciary duty based on
merger price failure to include value for viable derivative claim); In re Happy Child World,
Inc., 2020 WL 5793156, at *10–22 (Del. Ch. Sept. 29, 2020) (adjudicating derivative
claims in plenary action for appraisal and breach of fiduciary duty and including net value
of derivative claims in appraisal award); Zutrau v. Jansing, 2014 WL 3772859, at *2 (Del.
Ch. July 31, 2014) (“[T]he monetary value of the meritorious derivative claims that the
company had against the defendant at the time of the reverse stock split should be treated
as a non-operating corporate asset and added to the value of the company.”), aff’d, 123
A.3d 938 (Del. 2015) (TABLE); Oliver v. Bos. Univ., 2006 WL 1064169, at *19–21 (Del.
Ch. Apr. 14, 2006) (including value of potential derivative claims in entire fairness award);
Nagy v. Bistricer, 770 A.2d 43, 55 n.23 (Del. Ch. 2000) (“To the extent that the entity
possessed valuable legal claims, the value of those claims is part of the overall value of the
entity . . . .”); Merritt v. Colonial Foods, Inc., 505 A.2d 757, 765–66 (Del. Ch. 1986) (Allen,
62
Premium Provision, giving rise to a derivative claim that belonged to the Partnership.
Salem Contract Decision, 2021 WL 4438046, at *77–80 & n.61. The Freeze-Out price did
not afford any value to that litigation asset.
Taylor’s analyses did not fix these problems. At trial, Taylor gave not-credible
testimony to the effect that she believed that AT&T’s financial and accounting procedures
complied with the Management Agreement. Taylor Tr. 707. That was an improper legal
conclusion, and she lacked any foundation for her opinion.
C.) (holding defendants failed to prove “that the merger price was fair considering the value
of the then pending derivative claims to the corporation” and that “[i]n the relief phase of
this class-action litigation, plaintiffs will be free to introduce evidence relating to . . . . the
value . . . of the claims previously asserted in the derivative litigation”); id. at 763 n.3
(extinguishment of derivative claim in merger transaction does not “permit self-dealing
fiduciaries inappropriately to avoid their duty to account to minority shareholders” in
subsequent entire fairness action); see also Brinckerhoff v. Tex. E. Prods. Pipeline Co.,
LLC, 986 A.2d 370, 393–94 (Del. Ch. 2010) (failure by special committee’s financial
advisor to value derivative action “as a non-operating asset” weighed against finding that
merger price was fair); In re TD Banknorth S’holders Litig., 938 A.2d 654, 665 (Del. Ch.
2007) (declining to approve settlement that would extinguish derivative claims because
“[a] reasonable class representative in the plaintiff’s position certainly would have tried to
extract substantial consideration for the settlement of these claims”); Kohls v. Duthie, 765
A.2d 1274, 1277–78, 1284 (Del. Ch. 2000) (recognizing that challenged management
buyout of corporation would extinguish derivative claim whose value would “positively
affect” the value of stockholders’ shares but declining to enjoin transaction on basis of
director defendant’s status as a defendant in derivative action because derivative claim was
weak, chance of recovery against director defendant in derivative action appeared
“exceedingly remote,” and challenged transaction included “a substantial premium” that
would compensate for value of derivative claim). See generally David A. Drexler et al.,
Delaware Corporation Law and Practice § 36.08(1)(E), at 36-25 (1999 & Supp.) (“Where
the corporate cause of action involves a reasonably readily ascertainable quantum of
damages . . . and the cause of action is clearly established in the record, the Delaware courts
appear willing to include the value of such a cause of action in determining the overall
asset value.” (collecting cases)).
63
In post-trial briefing, AT&T cited testimony from Taylor, Barrick, and Musey in
which each stated that they had sufficient information to reach their valuation opinions.
See Musey Dep. 24–25; Barrick Dep. 51–52. AT&T seems to think that this testimony
addressed its failure to comply with the Management Agreement, but that testimony
addressed a different issue. The experts were speaking to whether they had sufficient
information to prepare valuations, assuming the numbers were accurate. The fact that the
parties’ experts performed their roles and valued the entities based on the information they
had does not excuse AT&T’s pervasive breaches of the Management Agreement.
The Premium Provision implies that the value of the Partnership would be at least
25% higher than AT&T’s valuation. In addition, the Partnership would have a claim for
revenue to which it was entitled and expenses for which it was overcharged during prior
years. The Freeze-Out price did not account for these sources of value.
2. AT&T’s Contemporaneous Beliefs About Value
AT&T’s attempts to establish the fairness of the Freeze-Out price likewise founder
because contemporaneous documents show that AT&T placed a significantly higher value
on the Partnership and its sister entities than it paid. Those internal analyses provide
persuasive valuation evidence. Cf. Verition P’rs Master Fund Ltd. v. Aruba Networks, Inc.,
210 A.3d 128, 137 (Del. 2019) (explaining that a buyer who possesses “material nonpublic
information about the seller is in a strong position (and is uniquely incentivized) to properly
value the seller”).
In January 2008, AT&T’s Corporate Development Team prepared a “Minority
Partnership Acquisition Summary” which projected that a “buyout of the 18 non-wireline
64
partnerships . . . at $100M” would create “[t]otal savings of $243M in perpetuity.” JX 344
at 7. In other words, AT&T believed that the value of the avoided distributions would be
more than double the transaction price. AT&T ultimately paid a higher price of $130
million. See JX 1123 at ’236. Even at that price, AT&T’s January 2008 analysis indicated
that the partnerships were worth 87% more.
After initially prioritizing other opportunities, such as outright purchases of
spectrum, AT&T returned to the buyouts. This time, AT&T brought in outside counsel to
shape the record for litigation. With outside counsel involved, AT&T’s internal documents
did not openly reveal AT&T’s valuations. See, e.g., JX 3514 at ’578. But in a supporting
spreadsheet for a presentation used to obtain CEO approval for the buyouts, AT&T’s
Corporate Development team calculated a sanitized valuation of the avoided distributions
by capitalizing the distributions it paid minority partners in 2009. The calculation used an
annual growth rate of 3%, which was strikingly pessimistic relative to the growth rates in
the Corporate Development team’s original model. JX 3516, “actualDist” tab; see JX 347
App’x (projecting growth of 8% in 2010, 6.5% in 2011, and 3.5% thereafter). Teske had
questioned why that earlier analysis included “a quick decline to 3.5%,” implying that even
that growth rate was unreasonably low. JX 345 at ’254.
Digging into the spreadsheet reveals that AT&T had much higher valuation
expectations. In a separate worksheet, the spreadsheet estimated the “cash flow per share
impact” of the Freeze-Out by projecting the Partnership’s free cash flow from 2010 through
2014. JX 3516, “CFPS” tab. Constructing a simple DCF analysis based on those
projections and using the Corporate Development team’s discount rate and long-term
65
growth rate supports a value for the Partnership of $334.8 million, 53% more than the
Freeze-Out price:
2011 2012 2013 2014 Terminal
CF $ 23,742 $ 23,688 $ 23,817 $ 28,080 $ 28,923
PV $ 21,584 $ 19,577 $ 17,894 $ 19,179
WACC 10%
G 3%
Terminal CF $ 413,179
PV Terminal CF $ 256,551
PV Discrete Period $ 78,234
Concluded Value $ 334,785
AT&T’s internal documents, used by AT&T when making actual decisions
regarding the Partnership, provide strong evidence that the price was unfair.
3. Problems With AT&T’s Analyses
Finally, AT&T relied on unsound valuations. As noted, AT&T did not present
Gilcreast or rely on PwC’s valuation work at trial. Instead, AT&T sought to prove the
fairness of the Freeze-Out price by having Taylor create different valuation analyses and
vouch for PwC’s conclusion. That tactic implies that PwC’s work could not stand on its
own, undermining the credibility of the basis for the Freeze-Out price.
As noted, PwC prepared a comparable companies analysis and a DCF analysis and
gave 50% weight to each. Taylor prepared a comparable companies analysis, a comparable
transactions analysis, and a DCF analysis, then gave 50% weight to the DCF analysis and
25% to each of the other analyses. She derived a range of value that bracketed PwC’s
valuation conclusion, then pronounced PwC’s valuation fair. No one disputes that the
66
comparable companies, comparable transactions, and DCF methodologies are viable
valuation techniques. The problem is that AT&T failed to prove that those methodologies
as applied in this case provided reliable evidence of the fair value of the Partnership. Setting
aside the failure of PwC—and subsequently Taylor—to take into account the Management
Agreement and the Premium Provision, AT&T still failed to prove that their valuations
were sufficiently reliable on the facts of this case.
a. The Comparable Companies Model
PwC and Taylor each valued the Partnership using a comparable companies
methodology. As the proponent of using a comparable company method, AT&T bore the
burden of proving the reliability of that model as applied in this case.
The comparable companies methodology “endeavors to draw inferences about a
company’s future expected cash flows from the market’s expectations about comparable
companies.” Merion Cap., L.P. v. 3M Cogent, Inc., 2013 WL 3793896, at *5 (Del. Ch. July
8, 2013). “The idea is that if the market expects comparable companies to grow at a certain
rate, then one can infer the growth of the subject company by applying a multiple drawn
from the comparables to a relevant metric, such as EBITDA or revenues.” In re Appraisal
of Orchard Enters., Inc., 2012 WL 2923305, at *9 (Del. Ch. July 18, 2012).
The comparable companies method involves
(1) finding comparable, publicly traded companies;
(2) deriving valuation ratios based on the trading prices of the shares of
the comparable companies and some recognized financial attribute,
such as revenue, EBIT, or EBITDA;
67
(3) correcting the ratios to account for differences in capital structure or
other factors; and
(4) applying a multiple derived from the valuation ratios to the financial
attribute of the subject company.
See Merion, 2013 WL 3793896, at *6. The reliability of the comparable companies method
thus depends in the first instance on having companies that are sufficiently comparable that
their valuation ratios provide insight into the value of the subject company. See Orchard,
2012 WL 2923305, at *9.
Taylor and PwC started with the same set of eight comparable companies. PwC used
all eight. Taylor excluded Sprint Nextel Corp. and Leap Wireless International Inc. because
they had low or negative EBITA that produced outlier results. See Taylor Report at 105,
110; JX 1209 at ’083. By excluding two companies from the list that PwC used, Taylor
undercut the reliability of PwC’s valuation.
Taylor ended up using the following six companies:
• AT&T
• Verizon
• Shenandoah Telecommunications Co.
• NTELOS Holdings Corp.
• MetroPCS Communications Inc.
• United States Cellular Corp.
AT&T failed to prove that the comparable companies that Taylor selected were sufficiently
comparable to generate a reliable indicator.
68
Determining whether a company is sufficiently comparable requires considering
firm-specific characteristics such as the size of the subject company, its line of business,
and its stage in its life cycle. Shannon P. Pratt, Valuing a Business: The Analysis and
Appraisal of Closely Held Companies 276–77 (5th ed. 2008). If the firm-specific
characteristics diverge from those of the subject company, then the valuator must discard
the comparable or make adjustments to account for the divergence.24 “At some point, the
differences become so large that the use of the comparable company method becomes
meaningless for valuation purposes.” In re Radiology Assocs., Inc. Litig., 611 A.2d 485,
490 (Del. Ch. 1990). When there are significant differences “as to product mix, revenues,
profit margins, revenue and earnings growth rates, assets and geographic markets,” then
those differences can “combine to make any comparison with [the subject company]
meaningless.” Id.AT&T faced an uphill battle in arguing for the use of a comparable companies
analysis in this case. The partnerships were unique because their primary asset was their
spectrum licenses, which were the “crown jewel” of AT&T’s wireless business. Barrick
24
See Merion, 2013 WL 3793896, at *7 (rejecting use of comparable companies
analysis due to difference in size between “comparable” companies and subject company);
Orchard, 2012 WL 2923305, at *9 (“When . . . it is difficult to find companies that actually
do the same thing as the subject company, the comparables method is less reliable.”); id.
at *10 (rejecting use of comparable companies analysis when the “comparable” companies
“often make money in ways that do not resemble [the subject company’s] business at all”);
Est. of Gallo v. Comm’r, 50 T.C.M. (CCH) 470 (T.C. 1985) (acknowledging that a “mature
company with [a] vast national distribution system in place” was not “remotely
comparable” to a much smaller, private company in the same industry, necessitating the
use of comparable companies from other industries); Pratt, supra, at 276–77.
69
Tr. 1326; see JX 3551 at 3 (AT&T “considers its spectrum holdings to be strategic in
nature”). Because each partnership was a single patch in the quilt of AT&T’s nationwide
wireless network, it is difficult to find public companies with comparable assets,
operations, and business models. The partnerships also were unique because they were
organized as pass-through entities for tax purposes and remitted the overwhelming majority
of their earnings as distributions to the partners. For similar reasons, this court has rejected
the use of the comparable companies method in prior cases involving similar entities. See
B&L Cellular, 2014 WL 6882207, at *3–4; AMWOH, 2013 WL 3865099, at *2–3.
The first four companies Taylor selected were integrated telecommunications
companies. Those companies were much larger than the Partnership and engaged in lines
of business the Partnership did not pursue. The remaining two companies—MetroPCS and
United States Cellular—were “pure-play” wireless companies, but they were much larger
than the Partnership.
AT&T failed to prove that the integrated telecommunications companies were
comparable to the Partnership. Taylor’s use of AT&T as a purportedly comparable
company is the most straightforward example of an inappropriate comparison. The
companies obviously differed in size, with AT&T dwarfing the Partnership. The nature of
the companies’ respective businesses also differed. The Partnership was a market-level
entity owned by Mobility, an operating company. AT&T was a holding company that
owned multiple operating companies, including Mobility. Contractually, Mobility and the
Partnership were supposed to operate in the same lines of business, but AT&T did not
permit the Partnership to engage in the full spectrum of business in which Mobility
70
engaged. See Salem Contract Decision, 2021 WL 4438046, at *19–22, *51. Instead, AT&T
enforced an unwritten rule that limited the Partnership’s business to “things . . . we do with
our phones.” Wages Tr. 395. At the time of the Freeze-Out, AT&T’s various business units
were in different stages of their respective life cycles. The Partnership was part of Mobility
and poised for significant growth. AT&T’s other businesses were mature or declining.
Finally, as noted, the Partnership represented a concentrated investment in a particular
chunk of spectrum, a “[s]carce, [c]ontrolled [a]sset,”25 making the Partnership
comparatively more valuable than the diversified portfolio that AT&T held. Barrick Tr.
1326.
AT&T also failed to prove that the other integrated telecommunications companies
were comparable to the Partnership. Like AT&T, Verizon, Shenandoah, and NTELOS
were considerably larger than the Partnership. Compare JX 3530, “Salem Revenue Build”
tab, with Taylor Report Ex. 35. Like AT&T, the other integrated telecommunications
25
JX 1239 at 12; see id. at 6 (observing AT&T had a “[s]olid [p]latform[]” due to
its “[s]trong spectrum position,” while its competitors’ “biggest challenge” was their “lack
of scale to compete with the likes of Verizon and AT&T”); JX 1209 at ’046 (“Wireless
carriers license spectrum from the FCC. Complete spectrum coverage of a geographic area
served by a carrier is critical to success. Just as critical is the ability to broadcast more data
over the same spectrum (i.e., increasing bandwidth).”); JX 3551 (Duff & Phelps license
impairment test explaining that AT&T’s “marketing and branding strategy is centered on
being able to provide subscribers with a national wireless service footprint,” which
depended on AT&T’s ownership of spectrum licenses); see also Taylor Report at 10
(“Going forward, consolidation activity in the sector would be heavily influenced by the
need of carriers to access additional spectrum, because fulfilling data demand for more
advanced 4G services would require significantly greater bandwidth.”).
71
companies engaged in other lines of business that were in different stages of their life
cycles.
AT&T also failed to prove that the two pure-play wireless companies were
comparable to the Partnership. Both were orders of magnitude larger than the Partnership.
While the Partnership had approximately 100,000 subscribers, MetroPCS and United
States Cellular had 6.6 million and 6 million subscribers, respectively. Taylor Report at
103–04. While the Partnership generated total revenue in 2009 of $70.6 million, MetroPCS
and United States Cellular generated total revenue of $3.5 billion and $4.2 billion,
respectively. Taylor Report at 106, Ex. 35. MetroPCS also operated under a different
business model by offering only prepaid, flat-rate contracts. See Taylor Report at 103.
Further, MetroPCS operated only in “select major metropolitan areas,” all of which were
much larger than the Salem Market. Id.
Among the companies Taylor selected, United States Cellular was the closest
comparable. Although much bigger, it was a pure-play wireless company and used a similar
business model. But having one comparable company does not save Taylor’s analysis,
because relying exclusively on United States Cellular produces an outlier result. See
AMWOH, 2013 WL 3865099, at *2. In this case, Taylor acknowledged that there were
“valid arguments for excluding” United States Cellular due to its “marginal profitability.”
Taylor Report at 111.
AT&T failed to carry its burden of proving that the companies in Taylor’s model
were comparable to the Partnership. PwC used those same companies, plus two that Taylor
72
discarded. AT&T thus failed to prove that either PwC or Taylor’s comparable companies
analysis provided reliable evidence of fair price.
b. The Comparable Transactions Model
Taylor also valued the Partnership using a comparable transactions model. PwC did
not use a comparable transactions model, which calls into question Taylor’s use of the
method. As the proponent of using a comparable transactions method to justify the fairness
of the Freeze-Out, AT&T bore the burden of proving the model’s reliability and
persuasiveness. In this case, the evidence demonstrates that Taylor’s comparable
transactions model did not produce reliable valuation evidence.
The comparable transactions method involves identifying transactions involving
similar companies, using financial metrics to derive ratios that quantify the value placed
on the companies, and then applying the resulting ratios to the company at issue. See
Highfields Cap., Ltd. v. AXA Fin., Inc., 939 A.2d 34, 54 (Del. Ch. 2007). The comparable
transactions method thus operates similarly to the comparable companies method, but it
uses transaction prices rather than trading prices to derive valuation ratios.
As with the comparable companies methodology, the reliability of a comparable
transactions methodology depends on the degree of similarity between the company in the
transaction being used for comparison and the company being valued. See LongPath Cap.,
LLC v. Ramtron Int’l Corp., 2015 WL 4540443, at *18 (Del. Ch. June 30, 2015). The
reliability of the method also depends on the nature of the transaction being used for
comparison and its ability to provide clear valuation metrics. See id.; Cede & Co. v. JRC
Acq. Corp., 2004 WL 286963, at *3–4 (Del. Ch. Feb. 10, 2004).
73
As with the comparable companies method, AT&T faced an uphill battle in arguing
for the use of a comparable transactions analysis in this case. The partnerships’ spectrum
licenses were unique. AT&T did not have any “significant past practice of voluntarily
selling licenses.” JX 3551 at 3. It was therefore unlikely that other transactions would be
sufficiently similar to the Freeze-Out.
Taylor relied on four transactions. Two involved public company targets: Sprint
Nextel Corp.’s acquisition of iPCS, Inc., in October 2009, and AT&T’s acquisition of
Centennial Communications Corp. in November 2008. Taylor Report at 116–17. The other
two involved private company targets: Verizon’s acquisition of Alltel Corp. in June 2008,
and AT&T’s effort to purchase T-Mobile in March 2011. Taylor Report at 117–18. The
last of the four did not close. AT&T abandoned its pursuit of T-Mobile after the FCC and
the Department of Justice opposed the transaction on antitrust grounds. See JX 2087 at 4–
5.
AT&T failed to prove that the transactions Taylor selected were sufficiently
comparable to the Freeze-Out. None of the transactions resembled a sale of a discrete,
market-level entity whose primary asset was spectrum. All four involved much larger
companies, and all but the Sprint Nextel-iPCS acquisition involved purchase prices that
were orders of magnitude greater. The AT&T-T-Mobile transaction was valued at $39
billion. Taylor Report Ex. 43. The Verizon-Alltel transaction was valued at $28.5 billion.
Id. The AT&T-Centennial transaction was valued at $2.8 billion. Id. The Sprint Next-iPCS
transaction was valued at $833 million. Id.74
For the public company transactions, Taylor rendered her analysis suspect by failing
to use the actual transaction price. Instead, she calculated the market value of 100% of the
target’s equity based on its stock price immediately before the announcement date. She
then added a control premium of 15% and subtracted the company’s net debt. Id. at 119.
Taylor claimed that she made these adjustments to remove the effects of merger synergies.
Id. at 115–16, 119. Taylor did not provide a persuasive explanation for her methodology,
and AT&T did not demonstrate that Taylor’s adjusted-transaction-price approach was
generally accepted in the financial community.
AT&T likewise failed to justify Taylor’s use of a 15% across-the-board control
premium. Taylor simply claimed that she had been using that figure for years. See Taylor
Tr. 977. Her blanket use of a 15% control premium distorted her analysis of the AT&T-
Centennial transaction, making it unusable in a comparable transactions analysis. AT&T
paid a 121.4% premium to acquire Centennial. Taylor Report at 119. By assuming that all
of the premium beyond her blanket 15% control premium derived from synergies, Taylor
artificially depressed the resulting EBITA multiple, resulting in a lower concluded
valuation for the Partnership.
The Sprint Nextel-iPCS acquisition was not comparable for the additional reason
that the buyer was “motivated by the desire to resolve a longstanding litigation, which
would have otherwise forced Sprint Nextel to make certain divestitures.” Taylor Report at
116 n.229. Taylor rejected another comparable transaction because the forced nature of the
sale meant that it “occurred at suppressed EBITDA multiples, which make it incompatible
75
for use in [her] valuation.” Taylor Report at 118. The same rationale undercuts the
comparability of the Sprint Nextel-iPCS acquisition.
For the private company transactions, Taylor took a different approach. To estimate
an EBITA multiple, she netted out synergistic value by taking the target’s last-twelve-
months EBITA “inclusive of the announced synergies” and dividing it by the target’s last-
twelve-months EBITA without synergies. Id. at 119. Then she divided the transaction price
by the ratio of EBITA including synergies to EBITA excluding synergies. Finally, she
calculated “adjusted Revenue and EBITA multiples by dividing the adjusted transaction
Enterprise Value by the reported revenue and EBITA.” Id. at 120. By combining historical
data with purely speculative “announced synergies,” Taylor made the private companies
seem more profitable than they actually were, resulting in lower EBITA multiples and
depressing the concluded value of her comparable transactions analysis. For example, by
simply assuming that the Alltel transaction would generate $1 billion in annual synergies,
Taylor depicted Alltel as 60% more profitable than it actually was. Id.; see Taylor Tr. 983–
84. That calculation reduced the derived EBITA multiple from 15.9x to 10.2x, depressing
the concluded value of Taylor’s comparable companies analysis. Taylor Tr. 984.
For the T-Mobile transaction, Taylor made additional adjustments that rested on a
mistaken reading of internal AT&T documents and rendered her analysis unusable. To
calculate an EBITA multiple for T-Mobile, Taylor added $1.8 billion in “Announced
Synergies” to T-Mobile’s Last Twelve Months EBITA. Taylor Report at 120, Ex. 46;
Taylor Tr. 984. That calculation reduced the derived EBITA multiple from 17.1x to 9.6x,
depressing the concluded value. See Taylor Report at 120. The plaintiffs proved that the
76
synergy value Taylor added actually consisted of “[a]voided spectrum acquisitions” that
AT&T would achieve in 2012 by acquiring T-Mobile’s spectrum in the transaction. JX
1893 at 6. In reality, AT&T already was planning to spend $1.8 billion to purchase
additional spectrum in 2010. See id. at 7; Taylor Tr. 988. Without the value of the avoided
spectrum acquisitions, the T-Mobile transaction would have had a negative $1.8 billion
EBITA impact in 2012. See Taylor Report at 120, Ex. 46 ($1.8B – $3.6B = $-1.8B).
Taylor’s erroneous synergies adjustment misleadingly depicted T-Mobile as being 80%
more profitable than it actually was. Taylor Tr. 984. Put differently, without Taylor’s
erroneous synergies adjustment, her analysis of the T-Mobile transaction translates to an
81.7x adjusted EBITA multiple, nearly eight times higher than any of the adjusted EBITA
multiples she calculated. See Taylor Report at 119–20, Exs. 45–46 (($21.793B * 1.8x) /
($2.28B – $1.8B) = 81.72x). The absurdity of the multiple that results from correcting
Taylor’s erroneous synergy adjustment warrants discarding her analysis of the T-Mobile
transaction.
The four transactions also are not comparable to the Freeze-Out because all but one
occurred during or immediately after the Great Recession of 2008. In an earlier case
involving similar freeze-outs by AT&T, this court noted that PwC did not use a comparable
transactions analysis and that “[t]he timing of each of Taylor’s selected comparable
transactions renders the method unreliable in this case, consistent with PwC’s view.”
AMWOH, 2013 WL 3865099, at *3. The only “transaction” that did not occur during this
period was AT&T’s failed bid for T-Mobile. This decision already has rejected Taylor’s
77
reliance on the T-Mobile “transaction” due to her inaccurate EBITA calculations. It bears
adding that the “transaction” (i) did not close and (ii) arose after the Freeze-Out.
Taylor’s own expert report in B&L Cellular provides additional persuasive reasons
to discard her comparable transactions analysis. There, she opined that “minimal weight
should put [sic] on the transactions approach considering the lack of transactions that
occurred in the recent period and the material adjustments required to strip out the
significant strategic premium [sic] that were included in these deals.” JX 3552 at 54. That
reasoning applies in this case.
AT&T failed to carry its burden of proving that the transactions in Taylor’s model
were comparable to the Freeze-Out. This decision therefore declines to adopt Taylor’s
comparable transactions analysis as reliable evidence of fairness.
4. The DCF Model
AT&T finally sought to prove the fairness of the Freeze-Out price by relying on
Taylor’s DCF model, which produced a valuation range of $193 to $279.27 million. AT&T
failed to prove that Taylor’s DCF valuation model established that the Freeze-Out price
was fair.
The DCF methodology is a “standard” technique that “gives life to the finance
principle that firms should be valued based on the expected value of their future cash flows,
discounted to present value in a manner that accounts for risk.” Andaloro v. PFPC
Worldwide, Inc., 2005 WL 2045640, at *9 (Del. Ch. Aug. 19, 2005).
The DCF model entails three basic components: an estimation of net cash
flows that the firm will generate and when, over some period; a terminal or
residual value equal to the future value, as of the end of the projection period,
78
of the firm’s cash flows beyond the projection period; and finally a cost of
capital with which to discount to a present value both the projected net cash
flows and the estimated terminal or residual value.
Radiology Assocs., 611 A.2d at 490 (internal quotation marks omitted). A DCF model
incorporates numerous assumptions about the future performance of the subject company.
Those assumptions “may always be challenged in any particular case.” Pinson v.
Campbell-Taggart, Inc., 1989 WL 17438, at *8 n.11 (Del. Ch. Feb. 28, 1989). AT&T bore
the burden of proving that Taylor’s assumptions were reasonable such that her DCF model
provided reliable evidence of the value of the Partnership as of the Freeze-Out.
a. Problems With The Flowshare Model
Taylor forecasted cash flows for a discrete projection period of 9.25 years using a
“flowshare model” prepared by PwC. See Taylor Tr. 739–40 A flowshare model is a
standard methodology used in the wireless industry to forecast market share. Taylor Tr.
738; see Musey Report at 38. Taylor tweaked the flowshare model slightly to reflect the
fact that PwC used a valuation date of March 31, 2010, over six months before the Freeze-
Out, while Taylor used a valuation date of October 12, 2010, when the Freeze-Out actually
occurred. See JX 1209 at ’018; Taylor Report at 29–31.
A flowshare model estimates the number of subscribers that a wireless firm will
have during a given period. To accomplish this, the model starts with the number of
existing subscribers at the beginning of the period. The model then forecasts the number
of potential subscribers that competing businesses in that market can capture during the
period based on the size of the market population. Some of the potential subscribers are
new wireless subscribers; others are existing subscribers who may be induced to switch
79
providers. Using these inputs, the model estimates the expected number of new subscribers,
or “gross adds,” that the firm captures. The model also estimates the number of subscribers
that the firm loses—a concept called “churn.” For each subsequent period, the flowshare
model performs the same calculations using the figures from the prior period as a starting
point. See Taylor Tr. 738–40; Musey Report at 39. Because it builds on itself, a flowshare
calculation can “roll . . . forward” indefinitely. Taylor Tr. 739. To convert the model’s
outputs into a revenue projection, the flowshare model multiplies the number of each type
of subscriber by the ARPU for that type of subscriber.
Two of the starting inputs in the Taylor/PwC flowshare model were unreliable: (i)
the number of existing subscribers at the beginning of the initial period, and (ii) the
population of the Salem Market at the beginning of the initial period. The starting figures
are critical, because each layer of estimation in a flowshare model “is dependent on the
accuracy of the estimates preceding it.” Musey Report at 41. In other words, errors in the
starting figures compound throughout the model.
AT&T failed to prove that the flowshare model used a reliable estimate of the
number of subscribers at the beginning of the initial period. Taylor and PwC used AT&T’s
subscriber count for the Partnership, which was equal to the number of AT&T subscribers
who had NPA-NXX numbers associated with the Partnership’s area. See Taylor Report at
67–68. In the Salem Contract Decision, this court found that by the time of the Freeze-Out,
NPA-NXX no longer served as a reliable proxy for a subscriber’s primary place of use.
2021 WL 4438046, at *10. In this litigation, AT&T could not answer basic questions about
the number of AT&T subscribers who resided in the Partnership’s service area but used a
80
non-Partnership NPA-NXX number. Id. AT&T could not say whether the subscriber
counts were off by 25%, 50%, or even 75%. Id. at *79. Assuming the subscriber count was
75% lower than it should have been, then a traffic-based allocation would have resulted in
four times as much revenue going to the Partnership. Id. Taylor acknowledged the lack of
reliability in the NPA-NXX counts by conceding that “[t]he financial impact of using NPA-
NXX to attribute subscribers to a Partnership cannot be readily determined.” Taylor Report
at 68. The lack of reliability in the NPA-NXX system renders the flowshare model
unreliable.26
26
Taylor tried to justify using NPA-NXX by claiming that any inaccuracy would be
partially offset by the outcollect roaming revenue the Partnership would receive when
would-be subscribers of the Partnership used their wireless devices in the Salem Market.
Taylor Report at 68. In the Salem Contract Decision, this court found that
the outcollect-incollect system did not fairly account for the revenue and
expense of the Partnership. Under AT&T’s system, a subscriber who moved
out of the Partnership’s area continued to generate incollect expense for the
Partnership, but the subscriber no longer generated any offsetting network
usage revenue for the Partnership. AT&T’s methodology thus generated
expense that had nothing to do with the operation of the Partnership’s
business.
Salem Contract Decision, 2021 WL 4438046, at *54. AT&T also suggested that the use of
the NPA-NXX benefited the Partnership because a subscriber who relocated to another
market still would generate monthly access fees that would be allocated to the Partnership.
Dkt. 614 at 12–13; see Wages Tr. 344–45; Hall Tr. 1040–41. In the Salem Contract
Decision, this court held that “the evidence does not establish that the monthly recurring
access fees exceeded the incollect expense, and there is no evidence that anyone at AT&T
ever conducted the cost-benefit calculation in real time. Instead, the record evidence makes
it unlikely that the Partnership came out ahead.” 2021 WL 4438046, at *54. Taylor
recognized that the partnerships likely incurred net losses from subscribers who relocated.
Taylor Tr. 711–12. And, as noted in the Salem Contract Decision, AT&T continued to
allocate equipment expenses to the Partnership each time a relocated subscriber purchased
a new cell phone, even though the Partnership no longer received revenue from that
81
AT&T also failed to prove that the flowshare model started with a reliable estimate
of the population of the Salem Market. Taylor adopted PwC’s figure of 387,965 people.
Taylor then grew the population of the Salem Market using PwC’s annual population
growth rate of 1.5%. JX 3530, “Salem Subs Build” tab.
PwC obtained its population estimate from AT&T management. See JX 1209 at
’031. PwC had data from the United States Census Bureau (the “Census Bureau”)
indicating that the population of the Salem Market was 396,103 people, not 387,965. See
id. at ’206. PwC consulted with AT&T about the divergence, and AT&T claimed that the
Census Bureau’s measuring area did “not precisely match [the Partnership’s] coverage
area.” Id. at ’031 & n.1. PwC turned to Census Bureau survey data from 2006 to 2008 to
corroborate AT&T’s estimate, which PwC ultimately used. See id. at ’112.
The evidence shows that AT&T provided PwC with data for the wrong area. The
Partnership was licensed to provide wireless services in the Salem, Oregon metropolitan
statistical area, consisting of Marion County and Polk County. See PA at ’415; MNSA at
’748–49; Common Carrier Public Mobile Services Information, Rep. No. CL-92-40, 7
FCC Rcd. 742, 752 (Jan. 24, 1992). AT&T provided PwC with the population of a different
subscriber’s use of AT&T’s network. 2021 WL 4438046, at *54. Worse still for the
Partnership, AT&T allocated three categories of General and Administrative expenses to
the Partnership for each subscriber whose NPA-NXX number was associated with the
Partnership, regardless of whether that subscriber long since had relocated outside of the
Salem Market. Wages 2019 Dep. 174–75. The Partnership thus incurred those categories
of General and Administrative expense without offsetting revenue. By the time of the
Freeze-Out, the NPA-NXX system was unreliable, and its continued use penalized the
Partnership.
82
FCC designation—the Cellular Geographical Service Area (“CGSA”). Musey Report at
49. By 2009, the CGSA was outdated. See id. at 50. The proper metric was the Salem
metropolitan statistical area. See id. Taylor thus started with an erroneous population
estimate, and that error persisted and compounded throughout her calculations.
AT&T also failed to prove that Taylor used a reliable churn estimate. Taylor
adopted PwC’s figure. PwC lacked confidence in its estimate, noting in its report that
“[l]imitations on available data (specifically, market churn by type for wireless providers
in the region) did not permit, in our view, a reliable analysis at this level.” JX 1209 at ’052.
PwC claimed that it used data from the Partnership about its historical levels of churn
combined with data from AT&T about historical levels of churn for AT&T and its
competitors in the Portland-Salem market area. See id. at ’028, ’052–53; Gilcreast Dep.
169. PwC also took into account national churn estimates for AT&T’s competitors. JX
1209 at ’052. Based on its analysis, PwC used a monthly churn rate for the Partnership of
1.79%, significantly higher than both AT&T’s national average of 1.48% and AT&T’s
churn rate of 0.98% in the Portland-Salem market area. See id. at ’033–34, ’052–53; JX
1215, “Inputs” tab, Cell G105. PwC suggested that it projected future churn rates based on
AT&T’s national churn rates, with the only logical source being the Ten-Year Plan. See
JX 1209 at ’053. Compare JX 725, “Subscriber Funnel” tab, Row 126 (Ten-Year Plan),
with JX 1215, “Inputs” tab, Row 101 (identical numbers in PwC model). But the forecasted
churn rates for the Partnership do not correspond to the forecasted churn rates in the Ten-
Year Plan. The Ten-Year Plan projected that national churn would decrease from 1.39% to
1.26%, for a compound annual growth rate (“CAGR”) of -1.09%. During the same period,
83
PwC forecasted that the Partnership’s churn would decrease much more slowly—from
1.79% to 1.7%, for a CAGR of only -0.57%. The Partnership inputs do not result from a
formula; they appear to have been hard coded. See JX 1215, “Inputs” tab, row 107. No one
provided a convincing explanation for the Partnership’s significantly higher churn rates.
To backfill for PwC, Taylor had one of her associates interview an AT&T employee
named Vinay Gaddamanugu, who claimed that AT&T used a system called
“SMARTFlow” to track churn rates. Taylor Rebuttal Report at 36. Taylor relied on this
unsworn statement to argue that PwC’s estimate was appropriate. Id. at 36–37. As an
expert, Taylor could rely on her associate’s conversation with Gaddamanugu. See D.R.E.
703. The court, however, will give this aspect of Taylor’s opinion little weight. AT&T
previously designated Wages, not Gaddamanugu, as its Rule 30(b)(6) witness on the
subjects of how AT&T tracked subscribers. Wages 2019 Dep. 54–55, 335–44. AT&T
could and should have produced Gaddamanugu as its Rule 30(b)(6) witness or called him
at trial, if he was indeed the appropriate person to address this topic. Taylor’s understanding
of her associate’s conversation with Gaddamanugu does not save PwC’s otherwise
unreliable churn estimate.
The errors discussed up to this point affect the number of Partnership subscribers
that the flowshare model forecasts for each year. Converting those numbers into revenue
projections requires an estimate of revenue that each subscriber generates. To reiterate, the
industry metric for this input is ARPU, an acronym for average revenue per user. Different
types of subscribers generate different levels of ARPU. Postpaid subscribers, who pay at
the end of a billing cycle based on actual usage, generate the highest ARPU. Reseller
84
subscribers, who are customers of third party resellers who purchase network usage from
AT&T in bulk, generate the lowest ARPU. Prepaid subscribers, who pay in advance for a
preset level of usage, generate lower ARPU than postpaid subscribers and higher ARPU
than reseller subscribers.
The flowshare model projected subscriber counts for a 9.25 year period. To
calculate ARPUs, the model started by dividing the Partnership’s 2010 subscriber revenue
by its average number of subscribers in 2010, generating an ARPU for the Partnership of
$45.68 (the “Base Partnership ARPU”). Taylor Report at 77; JX 3530, “Salem Revenue
Build” tab, Cell M29. The model used the Base Partnership ARPU as the starting point for
its ARPU projections.
To estimate how ARPU would grow over time, the model performed a complicated
calculation involving AT&T’s national ARPUs, AT&T’s projected national growth rates,
and the Partnership’s subscriber mix for each year. The model started with AT&T’s
national ARPU figures in 2010 for each type of subscriber, then used the Partnership’s mix
of subscribers in 2010 to calculate a weighted-average ARPU that the Partnership would
have had if the Partnership generated ARPU at AT&T’s national rates (the “Implied
National ARPU”). For 2010, the Partnership’s Implied National ARPU was $49.92. The
model then used AT&T’s projections for growth in ARPU of each subscriber type to
calculate new ARPUs for the subsequent year. Using the projected subscriber mix from
2010, the model then calculated a new Implied National ARPU for the Partnership. For
2011, the new Implied National ARPU was $50.52. To derive a growth rate, the model
compared the two Implied National ARPUs, resulting in a growth rate of 1.2% from 2010
85
to 2011. The model then applied that growth rate to the Base Partnership ARPU, resulting
in a Partnership ARPU for 2011 of $46.23.
These calculations effectively meant that the Partnership’s ARPUs for its
subscribers would always remain below AT&T’s National ARPUs. Because the model
used national ARPU growth rates to adjust the Partnership’s ARPU, the Partnership could
never increase its ARPU by more than AT&T.
Taylor could not provide a persuasive explanation for why the Partnership’s
weighted average ARPU would lag perpetually behind AT&T’s. See Taylor Tr. 931–33.
AT&T set its rate and pricing plans on a nationwide basis to serve all of AT&T’s
subscribers. Wages Tr. 331–32; JX 3597 at ’116. At a minimum, the Partnership’s ARPUs
should have converged towards AT&T’s.
The source of the critical assumptions about the growth of weighted-average
ARPUs appears to be a conversation that PwC had with Gopalan, who suggested that
growth in ARPUs would generally follow AT&T’s overall trend. See JX 1209 at ’053–54,
’213. PwC then operationalized this general comment using the complex calculations in
the flowshare model, which fixed the Partnership’s subscriber mix as of 2010 and locked
the Partnership into a lagging position relative to AT&T. PwC’s valuation report does not
suggest that PwC conducted any independent analysis of the reasonableness of the ARPU
assumptions. PwC noted that Leap Wireless and MetroPCS competed in the Salem Market,
which in theory might have led the Partnership to charge lower rates and depress ARPU.
JX 1209 at ’211. But Leap Wireless and MetroPCS competed in many other AT&T
markets, see id. at ’033, and as noted, AT&T priced its plans nationally.
86
AT&T failed to prove that the ARPU estimates for the Partnership were reliable.
AT&T long ago had shifted toward nationwide plans, making it likely that ARPUs would
converge and become relatively uniform across markets. Even if the Partnership started out
at a lower level of ARPU, it was not reasonable to assume that the Partnership would lag
AT&T by a fixed margin indefinitely.
AT&T also failed to prove that PwC’s and Taylor’s projections for changes in
subscriber mix were reliable. PwC and Taylor adopted two different approaches to project
subscriber mix. Neither was reasonable.
PwC’s model projected that the Partnership’s subscriber mix would remain the same
as it stood in 2010. In 2010, the Partnership had a less lucrative subscriber mix than AT&T
as a whole, resulting in a lower weighted average ARPU. PwC’s approach effectively
locked in the Partnership to an inferior position relative to AT&T. That choice compounded
the negative effect of using the Partnership’s lower 2010 ARPUs to create a weighted
average ARPU, then grow it at the rate AT&T projected for its national business. In reality,
AT&T projected that its overall subscriber mix would become more favorable over time,
with an increasing share of higher-valued postpaid subscribers and a decreasing share of
lower-valued reseller subscribers. See JX 725, “Subscriber Report” tab; JX 1132,
“Subscribers” tab.
Taylor tweaked PwC’s model by making changes that were modestly more
favorable to the Partnership. She projected that the share of the reseller subscribers, which
had the lowest ARPU, would decrease from 18.4% to 14.5% during the discrete projection
period. And she projected that the share of prepaid subscribers, which had higher ARPU,
87
would increase from 4.4% to 7.6%. Finally, she projected that the share of postpaid
subscribers, which had the highest ARPU, would increase from 77.3% to 77.9%. JX 3530,
“Salem Subs Build” tab, Rows 32:34.
Taylor’s tweaks blunted some of the unfavorable inputs that PwC used, but they
only went part of the way. Even after Taylor’s tweaks, the Partnership’s subscriber mix
and ARPUs remained meaningfully less favorable than what AT&T was forecasting. For
example, AT&T projected that the share of postpaid subscribers in its national business
would increase at a CAGR of 0.29% between 2010 and 2019, while Taylor projected that
same metric for the Partnership would increase at a CAGR of 0.09%. Compare JX 1132,
“Subscribers” tab, Row 22, with JX 3530, “Salem Subs Build” tab, Row 32. AT&T also
projected that the share of both prepaid and reseller subscribers would decrease. In Taylor’s
model, the two categories offset each other. Taylor did not provide a convincing
explanation for her changes. They appeared designed to make her version of the flowshare
model a little more favorable to the Partnership than PwC’s, while not resulting in any big
changes that would meaningfully affect valuation. Taylor claimed that she derived the
trends in subscriber mix by “trending the last four years of data” for the Partnership. See
Taylor Tr. 935–37. But they do not actually reflect the trends in the Partnership’s subscriber
mix during that period.27
27
See JX 3530, “Salem Subs Build” tab, Rows 32:34. For example, the three-year
CAGR for the share of postpaid subscribers in Taylor’s model was 0.5%. See id., Cells
D32:F32, I32. Following that trend would have been even more favorable to the
Partnership than following AT&T’s projections, which projected that the share of postpaid
subscribers would increase at a 0.29% CAGR. Instead of growing postpaid subscribers at
88
For all of these reasons, AT&T failed to show that the flowshare operated reliably.
The flowshare model drove the cash flow projections that PwC and Taylor used in their
DCF analysis. Those do not provide persuasive evidence that the Freeze-Out price was fair.
b. Other Problems With The Cash Flow Projections
In addition to the defects resulting from the flowshare model, the cash flow
projections that PwC and Taylor used failed to account for Connected Devices. Although
AT&T had not yet started to generate significant revenue from Connected Devices, AT&T
was projecting major growth in that area. By failing to account for Connected Devices,
PwC and Taylor undervalued the Partnership.
PwC did not include any meaningful value for Connected Devices. PwC asked
AT&T about its projections for Connected Devices. See JX 850; JX 867 at ’251. Teske
agreed that the information was worth considering, but he declined to provide it because
AT&T did not want the Freeze-Out to be delayed. JX 867 at ’251.
Taylor also did not include meaningful cash flow projections for Connected
Devices. Like PwC, she used the 2009 version of the Ten-Year Plan, which blended
Connected Devices with other types of subscribers. Taylor Report at 78 n.201. She claimed
that to break out Connected Devices separately, she would have had to make difficult
that rate, Taylor grew postpaid subscribers at a 0.09% CAGR over the ten-year discrete
projection period. More strikingly, prepaid subscribers declined during the four years
before the Freeze-Out, but Taylor forecasted that they would increase as a share of total
subscribers. Id., Row 33.
89
adjustments to the “subscriber ARPU, penetration, gross adds, and average subscribers.”
Taylor Report at 45.
Taylor’s approach was not persuasive. At the time of the Freeze-Out, there were
two principal types of Connected Devices: “those with NPA-NXX numbers (usually for
consumers) and those without (usually for enterprise users).” Musey Report at 61. The
2009 version of the Ten-Year Plan accounted for devices with NPA-NXX numbers, but
not for Connected Devices without NPA-NXX numbers. See JX 838 at ’555. And AT&T
did not attribute revenue from Connected Devices without NPA-NXX numbers to market-
level entities like the Partnership. Wages Tr. 402. This court found in the Salem Contract
Decision that before AT&T adopted the kilobyte-fee method in July 2010, AT&T “had the
capability to track data usage” for Connected Devices without NPA-NXX numbers, but
“simply chose not to do so.” 2021 WL 4438046, at *56–57. Because her analysis was based
on the 2009 version of the Ten-Year Plan, Taylor’s cash flow projections did not capture
the value of those subscribers.
Taylor thus failed to include a segment of Connected Device subscribers that AT&T
expected to grow rapidly. See Taylor Report at 8. The Ten-Year Plan projected a total of
53 million Connected Devices by 2019, encompassing both Connected Devices with NPA-
NXX numbers and those covered by Commercial Network Agreements. Id. at 61. That
figure reflected the addition of approximately 23 million Connected Devices.
A valuation of the Partnership needed to incorporate the projected benefits of
Connected Devices. Neither Taylor nor PwC took adequate steps to do that. For this
90
additional reason, AT&T failed to prove that PwC and Taylor’s cash flow estimates were
sufficiently reliable to support a meaningful valuation estimate.
c. Problems With The Perpetuity Growth Rate
For the reasons discussed up to this point, the projections that PwC and Taylor
created for the discrete projection period covered by their DCF analyses were unreliable.
A discounted cash flow model also includes a terminal period, which extends the cash flow
projections perpetually into the future. When calculating cash flows for the terminal period,
both PwC and Taylor used a perpetuity growth rate that was unreasonably low.
Conventional valuation wisdom holds that the perpetuity growth rate generally
should fall somewhere between the rate of inflation and the projected growth rate of the
nominal gross domestic product (“GDP”). Merion, 2013 WL 3793896, at *21. “A viable
company should grow at least at the rate of inflation and . . . the rate of inflation is the floor
for a terminal value estimate for a solidly profitable company that does not have an
identifiable risk of insolvency.”28 “But, a terminal growth rate should not be greater than
28
Glob. GT LP v. Golden Telecom, Inc., 993 A.2d 497, 511 (Del. Ch. 2010), aff’d,
11 A.3d 214 (Del. 2010); see Pratt, supra, at 247 (explaining that when a capital asset
pricing model (“CAPM”) “is used to develop the present value discount rate from which
the growth rate is to be subtracted, . . . that discount rate incorporates the expected rate of
inflation as part of the required rate of return,” so “the selected long-term growth rate
should also reflect the impact of expected inflation”); see also Maclane Gas Co. Ltd. P’ship
v. Enserch Corp., 1992 WL 368614, at *2 (Del. Ch. Dec. 9, 1992) (adopting inflation
adjustment to terminal year cash flow to “achieve a model that fairly depicts a firm whose
capital base is maintained in value”), aff’d, 633 A.2d 369 (Del. 1993) (TABLE); Cede &
Co. v. Technicolor, 1990 WL 161084, at *26–27 (Del. Ch. Oct. 19, 1990) (describing
upward adjustment to terminal period cash flow to account for inflation as not “wrong as a
91
the nominal growth rate for the United States economy, because if a company is assumed
to grow at a higher rate indefinitely, its cash flow would eventually exceed America’s
[GDP].” Merion, 2013 WL 3793896, at *21 (cleaned up).
At the time of the Freeze-Out, credible projections placed the rate of long-term
inflation at between 2.4% and 2.6%. See Barrick Report at 26; JX 628 at 3; JX 1209 at
’039. At the time of the Freeze-Out, credible projections placed the rate of nominal GDP
growth at 5.1%, suggesting real GDP growth of 2.5% to 2.7%. See Barrick Report App’x
B at 1; JX 628 at 3. These figures suggested a perpetuity growth rate of between 2.4% and
2.7%.
PwC used a perpetuity growth rate of 1.5%. JX 1215, “DCF” tab. Taylor used
PwC’s perpetuity growth rate. JX 3530, “Salem DCF” tab, Cell D46.
Those perpetuity growth rates were lower than expected inflation, meaning that
PwC and Taylor treated the Partnership as a wasting asset. See Golden Telecom, Inc., 993
A.2d at 511. AT&T did not believe that the Partnership was a wasting asset. The
Partnership was part of AT&T’s nationwide wireless network. AT&T consistently touted
its wireless business as the future of the company.
Taylor tried to justify the low growth rates by pointing to AT&T’s projections in
the 2010 Ten-Year Plan for its overall revenue and wireless revenue. Both declined to 1.7%
matter of logic” and “present[ing] a reasonable approach recognized in the world of
financial analysis”).
92
and 1.6%, respectively, by the end of the discrete projection period.29 That comparison was
not persuasive, because as a general matter, revenue growth does not necessarily
correspond to the growth of free cash flows. Depending on its cost structure and return on
invested capital, a company could grow free cash flows at a rate higher than revenue rate.
See Aswath Damodaran, Investment Valuation: Tools and Techniques for Determining the
Value of Any Asset 312–14 (3d ed. 2012). Duff & Phelps recognized this fact when it valued
AT&T’s spectrum licenses on the eve of the Freeze-Out, projecting that revenue growth
would decline to 1.4%, while free cash flows would grow at 2%. JX 3551 at 4.
Taylor’s perpetuity growth rate also resulted in a sharp drop-off in free cash flow
growth at the end of her discrete projection period. During the last four years of the discrete
period, Taylor projected that the Partnership’s free cash flows would grow at a CAGR of
2.46%. Taylor then set the perpetuity growth rate equal to 1.5% for the perpetuity period,
resulting in a 39% drop-off.
Taylor tried to justify the low growth rates with data from the United States Bureau
of Labor Statistics showing that “wireless inflation” was consistently negative from 2002
to 2010. Taylor Rebuttal Report at 47, Ex. 27; see Taylor Tr. 790–91. Put differently, the
29
See JX 1132, “Revenue Excl Overlays” tab, Cells BO56, BO64. For the reasons
stated above the line, the comparison is not valid. In addition, the revenue growth forecasts
in the Ten-Year Plan did not include Connected Device revenue, meaning that they omitted
the effects of the high-growth segment that AT&T saw as the future of its wireless business.
See JX 1132, “Revenue Excl Overlays” tab; see also id., “Conn_Devices” tab (Connected
Devices accounted for separately and not included in “Revenue Excluding Overlays” tab
Taylor used to support her growth rates).
93
prices per unit for wireless services came down during that period. But the prices that a
company charges are only one component of a cash flow projection. An equally important
component is the cost of producing each unit of wireless service. The difference between
the price charged and the cost of production is the margin, which determines how much
free cash flow becomes available. AT&T projected that network efficiencies would result
in “stable margins” for the foreseeable future. JX 367 at ’322; accord JX 1330 at ’077
(“industry margins remain high at 34%”). When Duff & Phelps valued AT&T’s licenses,
it predicted expanding margins for the indefinite future. JX 3551 at 4; accord JX 662 at 47
(projecting expanding margins from 38% to 40%). Those expectations only could hold if
AT&T anticipated that the cost per unit would come down dramatically as well and that its
cash flows would continue to experience healthy growth.
More generally, Taylor and AT&T tried to argue that a low perpetuity growth rate
was justified because the wireless business was maturing, citing a page in a presentation to
the AT&T board of directors. That page depicted the legacy wireless voice business as
maturing, but not yet fully mature. See JX 1330 at ’079. The same chart cited the
“Applications and Services” business as being in the growth phase and the “Interactive
Advertising” business as just starting its growth phase. Id. The same presentation projected
a five-year CAGR for applications and services of 10%. Id. at ’077. There is no reason to
believe that AT&T’s wireless business would grow at levels below inflation.
The low perpetuity growth rate that Taylor and PwC used conflicted with cash flow
estimates during the discrete projection period, which included significant capital
expenditures that would be expected to drive future growth. See JX 3530, “Salem DCF”
94
tab, Row 29. When Kobos provided input on the PwC valuation, he advised against
including high levels of capital expenditures in the perpetuity calculation. JX 3577 at ’181.
Taylor and PwC kept them in. At best for Taylor and PwC, their DCF model presented a
picture of a company in a competitive industry, with little prospect for real growth. But
that would have counseled in favor of setting perpetual growth at the rate of long-term
inflation. See Pratt, supra, at 248. Nothing warranted treating the Partnership as a wasting
asset by projecting negative real growth into perpetuity. See id.
PwC and Taylor used artificially low perpetuity growth rates that treated the
Partnership as a wasting asset, rather than as an integral part of AT&T’s nationwide
wireless network and the future of the company. AT&T therefore failed to prove that PwC
and Taylor used a reliable perpetuity growth rate. Their use of such a low perpetuity growth
rate provides evidence that the Freeze-Out price was unfair.
d. The Tax Rate
So far, the problems with PwC’s and Taylor’s DCF valuations have related to the
cash flow projections. PwC and Taylor also took an unpersuasive approach when tax-
effecting the cash flows. Both applied AT&T’s blended corporate tax rate of 38.5%, even
though the Partnership is a pass-through entity that does not pay tax at the entity level.
AT&T failed to prove that PwC and Taylor used a tax rate that was fair.
95
For a tax-paying entity like a corporation, financial analysts and this court typically
tax-effect cash flow projections using the corporation’s marginal tax rate. 30 For a pass-
through entity that does not pay entity-level tax, the analysis is more difficult. This court
has recognized that pass-through entity status confers additional value on the entity’s
investors by avoiding “double taxation” at both the entity and investor levels. Del. Open
MRI, 898 A.2d at 326; see Radiology Assocs., 611 A.2d at 494–95. By contrast, the use of
a corporation’s marginal tax rate in a DCF model simulates the effects of double taxation,
understating the value of the entity’s cash flows.
PwC and Taylor used AT&T’s tax rate of 38.5%, which was a blended tax rate that
took into account AT&T’s marginal federal tax rate of 35%, added an estimate of the state
taxes AT&T paid, and accounted for the fact that paying state taxes would result in a federal
tax deduction for AT&T. See Barrick Report at 33–34. PwC and Taylor thus valued the
Partnership as if it was a tax-paying C-corporation.
AT&T dictated that PwC use the 38.5% rate. See JX 893. PwC inquired about using
the Delaware Open MRI methodology, but AT&T instructed PwC to stick with AT&T’s
corporate rate of 38.5%, without providing any explanation. See Teske Tr. 668–70;
Gilcreast Dep. 243–45; JX 893.
30
See Damodaran, supra, at 252 (“If the same tax rate has to be applied to earnings
every period, the safer choice is the marginal tax rate . . . .”); accord In re Appraisal of
Ancestry.com, Inc., 2015 WL 399726, at *20 (Del. Ch. Jan. 30, 2015) (holding that it would
be “overly speculative to apply the current tax rate in perpetuity” because use of effective
tax rate failed to account for “the transitory nature of tax deductions and credits” (cleaned
up)).
96
Taylor’s adoption of AT&T’s corporate tax rate was not fair to the minority
partners. A legitimate debate exists about how to tax effect a pass-through entity’s cash
flows. See Part III.A.4, infra. What is not up for debate is the impropriety of using a
corporate tax rate to tax effect a pass-through entity’s cash flows.
AT&T failed to provide any credible justification for PwC and Taylor using
AT&T’s tax rate. That failure is an additional reason why AT&T failed to prove that
Taylor’s DCF model establishes the fairness of the Freeze-Out price.
5. The Finding Regarding Fair Price
AT&T failed to prove that the Freeze-Out price was fair. The Freeze-Out therefore
failed to pass muster under the entire fairness standard.
III. THE REMEDY
By effectuating the Freeze-Out in a manner that was not entirely fair, AT&T
breached its duty of loyalty. Once a transaction fails to satisfy the entire fairness test, then
the Court of Chancery has broad power in crafting a remedy. The court’s charge is to
fashion an award of equitable and monetary relief that is appropriate to address the wrong.
Bomarko, 766 A.2d at 440.
“Delaware law dictates that the scope of recovery for a breach of the duty of loyalty
is not to be determined narrowly.” Thorpe v. CERBCO, Inc., 676 A.2d 436, 445 (Del.
1996). “Once disloyalty has been established,” the court’s remedy should seek to ensure
“that a fiduciary not profit” from its actions “and that the beneficiary not be harmed.” Id.The court need not stop at an award of compensatory damages measured at the time of
breach. See Cede & Co. v. Technicolor, Inc., 634 A.2d 345, 371 (Del. 1993). A remedy
97
instead may take into account principles of disgorgement and award damages “designed to
eliminate the possibility of profit flowing to defendants from the breach of the fiduciary
relationship.” Gesoff, 902 A.2d at 1154. Although the plaintiffs do not seek them here, a
court also may award rescissory damages. Reis, 28 A.3d at 466.
The plaintiffs advanced a theory of damages rooted in the present value of the
distributions they would have received as minority partners but for the Freeze-Out. The
plaintiffs thus seek an award of damages based on the value of their interests in the
Partnership at the time of the Freeze-Out, with the value of their interests measured by the
present value of the cash flows those interests would generate. See Barrick Report at 13.
The court adopts the plaintiffs’ basic approach to damages, but the court declines to use
the present value of distributions to quantify the award.
Instead, for purposes of crafting a remedy, the court will seek to determine the value
of what the plaintiffs had before the Freeze-Out based on the operative reality of the
Partnership at that time of the transaction. The asset that the plaintiffs held took the form
of a partnership interest in an entity whose business was an essential part of AT&T’s
nationwide wireless network, which AT&T operated on an integrated basis, and which was
expected to be entering a prolonged period of growth as a result of the data revolution. The
value of AT&T’s network lay in the promise of ubiquity, and the Partnership’s market area
was critical to that offering. Because of that fact, AT&T subsidized the Partnership by
providing capital for investment at AT&T’s weighted average cost, rather than at the higher
cost that the Partnership would have had to pay if it were a stand-alone entity operating an
isolated cellular network. The Partnership also benefitted from other relationships with
98
AT&T. And because of its pass-through status, the Partnership could make distributions to
its investors that were not reduced by entity-level taxes.
This decision uses a DCF model to determine the value of what the plaintiffs held.
In doing so, the court selects reasonable inputs that result in a responsible estimate, but
gives the plaintiffs the benefit of the doubt. That approach follows the general principle
that “[t]he law does not require certainty in the award of damages where a wrong has been
proven and injury established. Responsible estimates that lack m[a]thematical certainty are
permissible so long as the court has a basis to make a responsible estimate of damages.”
Red Sail Easter Ltd. P’rs v. Radio City Music Hall Prods., Inc., 1992 WL 251380, at *7
(Del. Ch. Sept. 29, 1992) (Allen, C.). It also follows the corollary principle that
“uncertainties in awarding damages are generally resolved against the wrongdoer.” Thorpe
v. CERBCO, Inc., 1993 WL 443406, at *12 (Del. Ch. Oct. 29, 1993).
Giving the plaintiffs the benefit of the doubt also is warranted by case-specific
factors. As the court found in the Salem Contract Decision, AT&T failed to adopt accurate
methods for tracking its subscribers, relying instead on the NPA-NXX convention even
after AT&T knew the methodology was no longer a viable proxy for principal place of use.
2021 WL 4438046, at *50–54. AT&T’s use of NPA-NXX created a situation in which
AT&T did not know which subscribers should have been attributed to the Partnership and
which should not. AT&T did not even know how to estimate the magnitude of the error.
Id. at *79.
The court also found in the Salem Contract Decision that AT&T pervasively
disregarded the contractual obligations it undertook to the Partnership. “AT&T committed
99
to treat the Partnership better than its own business units and indisputably better than an
arm’s-length third party.” Id. at *14. AT&T agreed to operate the Partnership as an integral
part of its overall nationwide network, and for purposes of running its own business, that
is how AT&T managed the Partnership’s assets. Id. at *20. But when it came time to
allocate revenue and expense to the Partnership, “AT&T treated the Partnership as an
independent, stand-alone entity whose business had not progressed beyond providing voice
and basic data services to a limited group of subscribers in a specific geographic area.” Id.
AT&T did not follow the methods for allocating revenue and expense that the Management
Agreement specified. Most notably, AT&T did not follow the Shared Revenues Formula,
and it did not comply with the Premium Provision.
It is impossible to know how the Partnership’s financial performance would have
changed if AT&T had complied with its contractual commitments to the Partnership. Id. at
*51. This court has already held that “[i]t seems more likely than not that the Partnership
would have benefited materially from a traffic-based allocation.” Id. This court also has
noted that AT&T failed to comply with the Premium Provision, which called for adding a
25% premium to any revenue allocated to AT&T.
As the basic framework for the DCF model, this decision uses the model that PwC
prepared and which Taylor modified. By embracing this model, the court admittedly
creates some tension with its earlier finding that the outputs of PwC’s and Taylor’s version
of the model did not provide persuasive evidence on the question of fair price. But that was
due to errors in application. The validity of the DCF model as a conceptual approach is
beyond question.
100
The problem with PwC’s and Taylor’s versions of the DCF model fell into two
broad categories. First, PwC and Taylor used financial information that was based on
unreliable subscriber counts and which did not take into account AT&T’s contractual
obligations to the Partnership. Unfortunately, those problems cannot be fixed. The financial
records that AT&T maintained are all that we have.
Second, PwC and Taylor used erroneous and unreliable assumptions. Those aspects
can be fixed. By doing so, the court can arrive at a responsible estimate of damages.
A. The Estimate Of Cash Flows
The first step in the DCF analysis is to determine a reliable estimate of the
Partnership’s future cash flows. For remedial purposes, the court uses a ten-year projection
period, draws information from both the Three-Year Plan and the Ten-Year Plan, employs
a flowshare model with corrected inputs, and adjusts the tax rate.31
1. The Length Of The Explicit Forecast Period
When constructing a DCF model, the goal is to create an explicit forecast period
that is “long enough that the business will have reached a steady state by the end of the
period.” Tim Koller et al., Valuation 221 (5th ed. 2010). Whether a company has reached
steady state turns on the company’s “current stage within its lifecycle [and] the length of
31
The parties focused their disputes on the determinants of revenue. The parties
generally agreed on how to calculate expense. Most of the Partnership’s expenses are
calculated on a per-subscriber basis, so adjustments to the subscriber counts in the
flowshare model result in adjustments to expenses.
101
time it will remain in that stage.” In re ISN Software Corp. Appraisal Litig., 2016 WL
4275388, at *5 (Del. Ch. Aug. 11, 2016), aff’d, 173 A.3d 1047 (Del. 2017) (TABLE).
At the time of the Freeze-Out, different segments of AT&T’s wireless businesses
were at different stages of maturity. The legacy wireless businesses involving voice
communications and text messages was still growing, but were relatively more mature than
the newer services business and application business. The Connected Devices business was
in an early stage. Other data-driven ventures were newer still. See JX 1330 at ’079–80;
Musey Report at 17, 49, 69. The early stages of many of these businesses counsel in favor
of a longer projection period.
Another factor favoring a longer projection period was AT&T’s plans to invest in
network infrastructure, including a rollout of new 4G technology. See Musey Report at 10–
11. A shorter projection period could exclude significant capital expenditures. Doing so
might benefit the plaintiffs by reducing expenses in the discrete period, or it might harm
them by omitting the returns on those investments. Either would be inaccurate.
The Partnership was part of AT&T’s integrated wireless business. A ten-year
projection period better reflects the record evidence about where that business was in its
life cycle. This decision therefore uses a ten-year projection period to calculate damages.
2. The Three-Year Plan Versus The Ten-Year Plan
The next issue is which set of projections to use to forecast the Partnership’s cash
flows. The plaintiffs prefer the Three-Year Plan. AT&T prefers the Ten-Year Plan. As
discussed in the Factual Background, both plans are sufficiently reliable to use as a source
of evidence. See In re Appraisal of Regal Ent. Gp., 2021 WL 1916364, at *21 (Del. Ch.
102
May 13, 2021) (collecting cases). For the reasons provided in the Factual Background, the
Three-Year Plan generally is more credible and reliable than the Ten-Year Plan.
Because the Three-Year Plan is more reliable, this decision generally uses the
Three-Year Plan for periods when it applies. For periods post-dating the Three-Year Plan,
this decision uses the Ten-Year Plan as the best source of information available. When
drawing on the Ten-Year Plan, however, this decision takes into account evidence
indicating that AT&T included assumptions in the 2010 Ten-Year Plan that AT&T knew
were excessively conservative. See JX 1598 at ’355; JX 3577 at ’181; Kobos Dep. 165–66.
Although the Three-Year Plan wins out over the Ten-Year Plan, the choice between
them has a minimal effect on the damages calculation. Neither plan contains projections
for the Partnership’s business. Both plans contain projections for AT&T’s wireless
business as whole. The two plans are relevant primarily because they inform assumptions
about inputs such as revenue growth rates and the net effect of roaming.
3. The Forecasting Methodology
Because AT&T did not prepare a specific forecast for the Partnership, it is necessary
to create one. This decision uses the flowshare model that PwC created and Taylor
modified, but corrects for the value-depressing inputs that PwC and Taylor used.
For the number of subscribers at the start of the projection period, this decision uses
AT&T’s subscriber count. Although this decision has found those counts to be unreliable,
they are “the only starting point we have for the purpose of estimating subscriber growth.”
Musey Rebuttal Report at 6. AT&T also cannot complain, because AT&T itself used the
NPA-NXX counts.
103
For purposes of calculating damages, this decision uses the following inputs:
• The starting population of the Salem Market was 396,103 residents as of the
beginning of 2010. See Part II.C.4.a, supra.
• The rate of population growth in the Salem Market was 1.5%, based on data from
the Census Bureau. See JX 3530, “Salem Subs Build” tab, Row 9; see JX 1209 at
’029, ’212.
• The wireless penetration rate for the Salem Market at the start of the model is 88%.32
• The annual rates of increase in wireless penetration in the Salem Market are 3% for
2010, 2.5% for 2011–2013, 2.4% for 2014–2016, and 2.3% for 2017–2019. See JX
1209 at ’079.
Compared to PwC’s and Taylor’s model, these more realistic inputs increase the total
number of subscribers in the Salem Market, raise the number of subscribers available to
the Partnership, and result in the Partnership capturing more subscribers over time.
a. Partnership Churn And Market Churn
As discussed in the ruling on fairness, AT&T failed to prove that Taylor’s estimate
of Partnership churn was reliable. Musey argued for using AT&T’s level of churn in the
32
Musey Report at 57; see Taylor Report at 42, Ex. 2. PwC used 78.4% as its initial
estimate based on Gopalan’s statement that “a market penetration in the high 70s for 2009
is reasonable.” JX 1209 at ’033, ’212. Neither Gopalan nor PwC possessed any data to
support this figure. Musey explained persuasively that the estimate was too low. See Musey
Report at 56–57. Five months before the Freeze-Out, the FCC published the national and
local penetration rates for the United States in 2008, reporting a national penetration rate
of 90% and a penetration rate of 86% in the economic area that included the Salem Market.
JX 1142 at 9, 221. Musey projected that the penetration rate for the Salem Market would
grow to 88% in 2009, consistent with the increase Taylor used. That figure is likely to be
conservative. For example, between 2007 and 2008, the national wireless penetration rate
increased by 5%. JX 1142 at 9; see also JX 1994 at 95 (2011 FCC report stating that
national wireless penetration rate for year-end 2009 was 93.5%).
104
Portland-Salem market area, which was 0.98%. Musey Report at 45, Fig. 3-2. That figure
was likely somewhat low, because churn in smaller markets usually exceeds churn in larger
markets. Taylor Report at 74. The principal alternative, however, is AT&T’s national rate
of 1.48%, which is even higher than the Portland-Salem rate. This decision uses the figure
of 0.98%, which is a responsible estimate.
The estimate of market churn is an easier issue. Taylor used PwC’s estimates, but
she rounded the percentages to the nearest tenth. Compare JX 3530, “Salem Subs Build”
tab, Row 22, with JX 1215, “Sub build” tab, row 19. That choice caused her estimates to
fluctuate within a tight range above and below PwC’s estimates. Compare JX 3530, “Salem
Subs Build” tab, Row 25, with JX 1215, “Sub build” tab, row 24. Although the impact was
small, it negatively affected the value of the Partnership. Taylor did not offer a convincing
reason for the change, which exemplifies the type of small adjustment that a valuator can
make to influence the output. This decision reverts to PwC’s estimates for market churn.
b. Gross Adds
Taylor adopted PwC’s forecast for the Partnership’s share of gross adds. Compare
JX 3530, “Salem Subs Build” tab, Row 31, with JX 1215, “Sub Build” tab, Row 43. PwC
estimated the Partnership’s share of gross adds to be 22.8% for 2010. PwC then gradually
decreased the Partnership’s share of gross adds to a floor of 21.1% in 2014 and maintained
that estimate for the remainder of the discrete projection period.
Musey criticized PwC’s estimate because it appeared to be based solely on the
Partnership’s gross adds in 2009. Musey noted that “[i]nternal PwC emails indicate
flowshare should equal market share,” but that PwC’s model set flowshare significantly
105
lower than PwC’s estimate of the Partnership’s market share of 32.4%. Musey Report at
41; see JX 1209 at ’033.
The explanation lies in the fact that PwC and Taylor underestimated the population
in the Salem Market. See Part II. As a result, they underestimated the number of subscribers
and the Partnership’s market share. See generally Taylor Tr. 941–42. This decision corrects
those errors and therefore does not make any further adjustment to gross adds.
c. ARPUs And Subscriber Mix
The final step in correcting the flowshare model involves adjusting the Partnership’s
subscriber mix and ARPUs. As discussed in the preceding section, Taylor failed to justify
her inputs.
Taylor’s ARPUs assumed without sufficient basis that the Partnership would
underperform AT&T’s national ARPUs indefinitely. It seems reasonable to assume that
the Partnership’s weighted average ARPUs would converge toward AT&T’s, but neither
party presented a calculation that operationalizes that intuition. Changing the Partnership’s
weighted average ARPUs to equal AT&T’s would increase the ARPUs abruptly at the
beginning of the discrete projection period, when in reality the change likely would have
been more gradual.
Taylor also failed to justify her changes in the Partnership’s subscriber mix. She
claimed to have followed the trends in the Partnership’s subscriber mix in the years leading
up to the Freeze-Out, but her projections did not match any recognizable trends in the data.
Her projections also were less favorable than the changes AT&T projected for its overall
wireless business.
106
These errors are difficult to correct. To approximate the value Taylor excluded, this
decision leaves the subscriber mix projections unchanged but adopts the ARPUs implied
by AT&T’s national averages. Using AT&T’s national averages recognizes that by the
time of the Freeze-Out, AT&T had adopted uniform rate plans on a nationwide basis,
making it likely that Partnership ARPUs would converge to AT&T’s nationwide ARPUs.
This choice results in an abrupt increase in Partnership ARPUs at the beginning of the
projection period. While that choice favors the plaintiffs, it is offset in part by leaving in
place Taylor’s unfavorable projections for subscriber mix. The resulting approach
generates a responsible estimate on an issue where further precision is not possible.
d. Other Revenue Sources
The adjusted flowshare model generates responsible estimates of revenue for voice
and data services. It does not address three additional sources of revenue: Connected
Devices, voice roaming, and data roaming.
i. Connected Devices
As discussed in the preceding section, AT&T failed to prove that Taylor’s DCF
model adequately accounted for revenue from Connected Devices. It is therefore necessary
to estimate the amount of omitted revenue.
Musey provided his own forecast for Connected Devices, but Musey did not explain
how he derived his calculations. The court therefore will not use his figures.
It is possible to approximate the number of Connected Devices not captured by
Taylor’s projections. The 2009 Ten-Year Plan projects a total of 118.33 million AT&T
subscribers by 2019. JX 725, “Subscriber Report” tab, Cell CL16. The 2010 Ten-Year Plan
107
projects 95.76 million traditional subscribers by 2019.33 The additional 22.57 million
subscribers represent the sum of (i) the number of subscribers who did not have NPA-NXX
numbers in 2009 that AT&T gained the ability to track in 2010 and (ii) an upward revision
that AT&T made for additional growth in Connected Devices.
To include the value of the additional Connected Devices, this decision adds the
resulting revenue as an overlay to the Partnership’s services revenue. That approach avoids
the difficulties Taylor identified with disaggregating Connected Devices and reconfiguring
subscriber ARPUs. See Taylor Report at 72–73.
For simplicity, this decision calculates the additional Connected Devices on a
national level as a percentage of overall AT&T subscribers. It then multiplies the resulting
percentage by the total Partnership subscribers on the date of the Freeze-Out to estimate
the additional Connected Devices that should have been attributed to the Partnership. It
then multiplies the number of subscribers by the ARPU for Connected Devices from the
2010 Ten-Year Plan to estimate incremental Connected Device revenue.34
33
Compare JX 1132, “Subscribers” tab, Cell BN12 (148.7 million total subscribers
in 2019), with JX 1132, “Subscribers” tab, Cell BN72 (52.94 Connected Device subscribers
in 2019).
34
This decision uses the 2010 Ten-Year Plan’s ARPU projections for “Reseller
Connected Device Subscribers” instead of the ARPU projections for “Retail Connected
Devices Subscribers.” The ARPU for Retail Connected Device Subscribers was over four
times higher than the ARPU for Reseller Connected Device Subscribers. Compare JX
1132, “ARPU” tab, Row 373, with JX 1132, “ARPU” tab, Row 328. Two reasons support
this selection. First, prepaid Connected Device subscribers included iPad users, and it
appears likely that those subscribers had NPA-NXX numbers, making it more likely that
they already were included within Taylor’s estimates of traditional prepaid subscribers. See
Taylor Tr. 958–59; Taylor Report at 16–17. Second, the overwhelming majority of the
108
The following table depicts these calculations:
This estimate is conservative and has minimal effect on the Partnership’s cash flows. It
also uses a lower initial estimate of low-ARPU Connected Devices as a percentage of all
subscribers (2.1%) than Taylor used (8%). See Taylor Report at 45.
ii. Voice Roaming
growth in Connected Devices that AT&T projected in the 2010 Ten-Year Plan was from
reseller Connected Devices. Compare JX 1132, “Subscribers” tab, Row 82, with JX 1132,
“Subscribers” tab, row 86. That dynamic reflects a majority of the growth in Connected
Devices coming from Commercial Network Agreements, which used wholesale prices and
thus had lower ARPUs. See Taylor Report at 16–17. Using the ARPU estimate for reseller
Connected Devices results in a more conservative estimate. See Musey Tr. 1260–61.
109
The cash flow projections also must account for net revenue from voice roaming.
By the time of the Freeze-Out, AT&T had implemented an arrangement under which
market entities like the Partnership earned and were charged a flat rate per minute for both
intra-carrier and inter-carrier voice roaming. See JX 1209 at ’032.35
Taylor incorporated AT&T’s system into her DCF model. The plaintiffs and their
experts generally did not contest Taylor’s calculations for voice roaming. Musey used a
different approach, but Taylor persuasively critiqued it as depending on a non-existent
relationship between outcollect roaming revenues and wireless service revenues. Taylor
Rebuttal Report at 33. For purposes of the damages calculation, this decision adopts
Taylor’s projections for voice roaming.
iii. Data Roaming
35
In the Salem Contract Decision, the court found that the roaming methodology
violated AT&T’s contractual commitments. AT&T agreed to eliminate any direct charges
associated with intra-company roaming, and the wholesale fee arrangement violated that
commitment. See Salem Contract Decision, 2021 WL 4438046, at *19. AT&T committed
to include any roaming-related revenue, whether from intra-carrier roaming or inter-carrier
roaming, within the Shared Revenues Formula as revenue generated by subscribers using
the Entire Network. AT&T then committed to allocate a proportionate share of the revenue
to the Partnership based on the Traffic Ratio and add a premium to comply with the
Premium Provision. AT&T did not do that. See id. at *53–55, *57–58. Instead, AT&T used
its wholesale system for voice roaming. AT&T’s noncompliant system constituted the
operative reality of the Partnership at the time of the Freeze-Out.
110
The final source for revenue that the cash flow projections must address is data
roaming.36 Taylor annualized the Partnership’s results from the first nine months of 2010,
then applied growth rates from the 2010 Ten-Year Plan. See Taylor Report at 81; JX 3530,
“Salem Revenue Build” tab, Row 41; JX 1598a, “P&L” tab, Row 5. That approach
underestimated intra-carrier data roaming revenue because it used the excessively
conservative growth rates implied by the 2010 Ten-Year Plan. This decision has found that
the Three-Year Plan provides a more realistic picture of AT&T’s expectations for data
services revenue, and it therefore uses the growth rates from the Three-Year Plan. That
modification remains conservative, because this decision does not extend those higher
growth rates beyond 2013.
The Three-Year Plan’s higher estimates for data roaming applied to AT&T’s system
as a whole. Some of that increased data roaming would manifest as increased in-collect
revenue for the Partnership. But some of that increased data roaming would result from the
Partnership’s subscribers using the network elsewhere and manifest as increased out-
collect expense for the Partnership. This decision therefore also increases the anticipated
growth rates in intra-carrier data roaming expense for the years covered by the Three-Year
Plan by an amount equal to the difference between Taylor’s growth estimates and the
Three-Year Plan’s growth estimates. The following table shows the calculations:
2011 2012 2013
36
As with voice roaming, AT&T’s approach to data roaming violated AT&T’s
contractual commitments in the Management Agreement. It nevertheless constituted the
Company’s operative reality at the time of the Freeze-Out.
111
Taylor Revenue Growth Rate 20.66% 17.29% 15.91%
3YP Revenue Growth Rate 9.74% 11.13% 11.41%
Difference 10.92% 6.16% 4.50%
Taylor Expense Growth Rate 6.39% 8.43% 8.14%
Modified Expense Growth Rate 17.31% 14.59% 12.64%
The Partnership was a net loser on roaming, so the effect of this change is to increase the
Partnership’s loss position on roaming and decrease the value of the Partnership.
The record does not provide any basis to incorporate inter-carrier data roaming.
AT&T did not track inter-carrier data roaming in the period leading up to the Freeze-Out.
Taylor did not model inter-carrier data roaming at all, and Musey and Barrick modeled all
roaming together using an unpersuasive methodology. The damages award therefore omits
any amount for inter-carrier data roaming.
4. The Tax Rate
In a dispute with significant implications for the damages award, the parties
disagreed over the tax rate to apply to the Partnership’s projected cash flows. As discussed
in the previous section, AT&T failed to prove that using AT&T’s corporate tax rate was
fair given that the Partnership was a pass-through entity for tax purposes. That finding does
not address how the pass-through status of the Partnership should be handled.
As noted, this court has recognized the need to account for the value of pass-through
entity status. In Radiology Associates, this court declined to deduct any taxes from a pass-
through entity’s cash flows. 611 A.2d at 494–95. The court valued the entity based on its
operative reality as a stand-alone entity, in which the entity did not pay taxes before making
distributions to investors. Id. The court concluded that in light of that operative reality, the
112
plaintiff “failed to prove that the adjustment is ‘generally considered acceptable in the
financial community.’” Id. (quoting Weinberger, 457 A.2d at 713).
Fifteen years later, the court took a different approach. This time, the court valued
the entity based on what an acquirer would pay for the business. Del. Open MRI, 898 A.2d
at 329. The court concluded that when applying that metric, it needed to consider the
personal taxes that the investors paid, because “[t]o ignore personal taxes would
overestimate the value . . . and would lead to a value that no rational investor would be
willing to pay to acquire control.” Id. The court therefore derived an implied tax rate by
starting with the amount of cash that the investors in the pass-through entity retained in
their personal capacities after paying personal taxes. Id. at 327–30. Using that amount of
cash, the court then calculated a synthetic “effective tax rate” that a corporate entity would
pay to generate the same after-tax cash flows. The following chart depicts the analysis:
C Corp. Pass- Equivalent C-
Through Corp. Tax Rate
Income Before Tax $100 $100 $100
Corporate Tax Rate 40% – 29.4%
Available Earnings $60 $100 $70.60
Dividend or Personal 15% 40% 15%
Income Tax Rate
Available After Dividends $51 $60 $60
Id. at 330 (formatting added). As the chart shows, the investors in the pass-through entity
netted $60 after paying no tax at the entity level but 40% tax at the personal level. To
generate the same amount of after-tax proceeds, the taxable corporation would have to pay
tax at 29.4%, rather than at its actual marginal rate of 40%. The court therefore used the
synthetic tax rate of 29.4% in its DCF model to value the firm. Id.113
The Delaware Open MRI method is one way of approximating the benefits of pass-
through entity status in a judicial valuation. This court adopted the Delaware Open MRI
method in B&L Cellular, which was an appraisal proceeding involving a similar market-
level entity. See 2014 WL 5342715, at *3. But the fact that the court applied a valuation
principle previously does not create a rule of law. Valuing an entity is “at bottom, a fact
finding exercise.” In re Appraisal of Jarden Corp., 2019 WL 3244085, at *1 (Del. Ch. July
19, 2019), aff’d sub nom. Fir Tree Value Master Fund, LP v. Jarden Corp., 236 A.3d 313,
322 (Del. 2020). “[B]y functional imperative, the evidence, including expert evidence, in
one [valuation] case will be different from the evidence presented in any other [valuation]
case. Different evidence, of course, can lead to different decision paths and different
outcomes.” Id.; accord Merion Cap. L.P. v. Lender Processing Servs., Inc., 2016 WL
7324170, at *16 (Del. Ch. Dec. 16, 2016) (noting that an argument which succeeds in one
case “may not prevail in another case if the proponents fail to generate a similarly
persuasive level of probative evidence or if the opponents respond effectively”).
In this case, both sides have advanced persuasive reasons for not following
Delaware Open MRI. AT&T and Taylor explained that the method is fact-specific and
subjective, because the synthetic tax rate will vary based on the marginal tax rates of the
investors. Dkt. 614 at 52. In Delaware Open MRI, the court was valuing a small and closely
held entity, operating exclusively in Delaware, owned by similarly situated physicians who
all paid federal and state taxes at the maximum rate. With that homogenous investor base,
the court could construct a responsible model of the after-tax cash flows that the investors
would receive, then use those cash flows to calculate what a buyer paying corporate-level
114
tax would pay to acquire those tax flows in an arm’s-length sale. AT&T points out that the
plaintiffs in this case “included individuals, trusts, corporations, and other pass-through
entities located in numerous states and with vastly different tax circumstances.” Dkt. 614
at 53. AT&T concludes that Delaware Open MRI should not apply as a general principle,
particularly to an entity like the Partnership where the various investors pay taxes at
different and unknown rates.
AT&T’s criticisms are valid, but while AT&T advanced them in an effort to justify
using AT&T’s corporate tax rate, they instead point in favor of recognizing that the
investors in the Partnership, including AT&T, receive distributions from the Partnership
free of entity-level tax. Whether they paid any investor level tax depended on their
individual tax situations, which could vary widely. The Partnership as an entity, however,
did not pay taxes. For purposes of a remedial calculation, those criticisms point in favor of
the approach taken in Radiology Associates, where the court valued the entity in its
operative reality, not based on what a hypothetical buyer would pay, and did not tax-effect
the cash flows of the entity.
For their part, the plaintiffs also object to the use of Delaware Open MRI. They
contend that the court should apply no tax rate at all because the court is calculating
damages, not conducting an appraisal, and “damage awards are given pre-tax.” Dkt. 626 at
63–64 (quoting Barrick Tr. 1286). The plaintiffs then pay tax on the damages award based
on their individual tax situations. Not tax effecting the cash flows thus does not mean that
the plaintiffs will not pay tax on what they receive; it only means they will not pay
115
effectively two levels of tax on the damages award when they only paid one level of tax
before the Freeze-Out. That argument has considerable force.
The plaintiffs also advance a persuasive policy criticism of Delaware Open MRI,
which applies all the more strongly to PwC and Taylor’s decision to use AT&T’s corporate
tax rate. Under the Partnership’s operative reality at the time of the Freeze-Out, the
investors in the Partnership could expect to receive distributions into perpetuity from an
entity that did not pay entity-level tax. An approach that permits a controller to apply a
higher tax rate for purposes of a freeze-out results in a lower value for the interests held by
the eliminated investors relative to the entity’s operative reality. But the value of the
entity’s pass through status does not disappear. Through the Freeze-Out, the controller
retains that value and enjoys distributions from the entity without paying entity-level tax.
Applying a tax rate to the pass-through entity for purposes of the valuation in the Freeze-
Out thus transfers value from the eliminated minority to the controller. It enables the
controller to profit from its wrongdoing, contrary to settled remedial principles.
The Delaware Open MRI method also fails to account for the fact that the acquirer
obtains tax benefits from acquiring the minority interest. When evaluating the Freeze-Out,
AT&T considered the value it would receive from the step-up in the tax basis of the
Partnership’s assets. Dkt. 626 at 31–32 (citing JX 3514 at ’580). Using a range of discount
rates, AT&T valued the tax benefits of the freeze-outs at between $16 million and $18.9
million, translating to a tax benefit in the range of 21% to 24% of the total purchase price.
JX 3514 at ’580.
116
In an effort to respond to these criticisms, Taylor and PwC argue that “in the
hypothetical willing buyer/willing seller valuation context, the only realistic willing buyers
for the Partnerships would be other corporate entities” that would have similar tax
characteristics as AT&T. Taylor Report at 91–92. But the hypothetical sale standard is not
the test. The court is crafting a damages award for a breach of the duty of loyalty, resulting
in an inquiry that looks to whether the eliminated investors received the “substantial
equivalent in value of what he had before.” Sterling, 93 A.2d at 114. The test does not look
to fair market value, but rather the value of what was taken from the beneficiary.
It is true that this court has looked to what a third party would pay when relevant to
determining whether the controller paid a price that was too low. See Reis, 28 A.3d at 466;
Kahn v. Tremont Corp., 1996 WL 145452, at *1 (Del. Ch. Mar. 21, 1996) (Allen, C.), rev’d
on other grounds, 694 A.2d 422 (Del. 1997); Cinerama, Inc. v. Technicolor, Inc., 663 A.2d
1134, 1143 (Del. Ch. 1994) (Allen, C.), aff’d, 663 A.2d 1156 (Del. 1995). That reference
point is not a license for controllers to impose prices that might be justifiable by a third-
party metric but which ignore important attributes of the operative reality of the entity in
question. The court’s ultimate task is not to value the plaintiffs’ interest based on a
hypothetical arm’s-length sale of the Partnership. Such an approach would treat the
Partnership as if it were an isolated standalone entity, operating a tiny legacy wireless
business in the Salem Market. The court’s task is to value a hypothetical alternative in
which the investors could remain holders of Partnership interests based on the operative
reality of the Partnership at the time of the Freeze-Out. In that operative reality, the
117
Partnership was an integral part of AT&T’s nationwide network, and the plaintiffs received
distributions from an entity that did not pay entity-level tax.
Both sides thus advanced compelling arguments against applying Delaware Open
MRI on the facts of this case. The principal alternative method is to follow Radiology
Associates. Declining to introduce entity-level tax for purposes of the damages calculation
is consistent with the Partnership’s status as a tax-free entity. The resulting damages award
risks overstating the value of the Partnership, but it generates a responsible estimate, and
“once a breach of duty is established, uncertainties in awarding damages are generally
resolved against the wrongdoer.” Thorpe, 1993 WL 443406, at *12.
Using the Radiology Associates approach also is consistent with the maxim that
“[o]nce disloyalty has been established,” the court’s remedy should seek to ensure “that a
fiduciary not profit” from its actions “and that the beneficiary not be harmed.” Thorpe, 676
A.2d at 445. In this case, AT&T reaped significant benefits through the Freeze-Out,
including from the step-up in basis. By imposing a tax-free damages award on AT&T, this
decision assures that AT&T will not enjoy the benefit that the Freeze-Out created.
This decision therefore follows Radiology Associates when awarding damages.
Operationally for purposes of the DCF model, that means a tax rate of 0%.
B. The Discount Rate
The next step in the damages calculation is to calculate a discount rate. The experts
for both sides used the Partnership’s after-tax weighted average cost of capital (the
118
“WACC”).37 Taylor derived a WACC of 8.5%. Barrick derived a WACC of 6.3%. The
parties disagreed about five components: (i) the pre-tax cost of debt, (ii) the equity risk
premium, (iii) whether to include a size premium, (iv) the Partnership’s beta, and (v) the
after-tax cost of debt.38
a. The Pre-Tax Cost Of Debt
When calculating the Partnership’s pre-tax cost of debt. Barrick followed PwC and
used the Moody’s average bond yield on A-rated debt securities, which Barrick estimated
37
The formula is: WACC = Kd * (1 – T) (D/V) + Ke * (E/V), where Kd is the
company’s pre-tax cost of debt capital, T is the company’s tax rate, D/V is the proportion
of the company’s operations that are funded by debt capital, Ke is the company’s cost of
equity capital, and E/V is the proportion of the company’s operations that are funded by
equity capital. The relative proportions of D/V and E/V comprise the company’s capital
structure. See Richard A. Brealey & Stewart C. Myers, Capital Investment and Valuation
365 (2003).
The cost of equity represents the market rate of return on an equity investment.
Valuation professionals typically use a CAPM to calculate the cost of equity. CAPM starts
with the “risk-free rate,” meaning the market rate of return on a debt security with no
default risk. CAPM then adds the risk-free rate to the product of the “equity risk premium,”
meaning the required rate of return on equity securities above the risk-free rate, and the
company’s “beta,” a measure of the company’s sensitivity to systemic market risk. The
result of that calculation is the cost of equity: Ke = Rf + (ß * Rm), where Rf is the risk-free
rate, ß is the company’s beta, and Rm is the equity risk premium.
38
The experts agreed on a risk free rate of 3.4%, which this decision adopts. The
experts also agreed on using AT&T’s capital structure when calculating the Partnership’s
WACC. Barrick used AT&T’s five-year average capital structure. See Barrick Report
Sched. 15. Taylor used AT&T’s reported capital structure on the day of the Freeze-Out.
See Taylor Report at 98–100, Ex. 33. Because the value of the Partnership must be
determined on the day of the Freeze-Out, the decision adopts Taylor’s approach. On the
day of the Freeze-Out, AT&T’s capital structure consisted of 29.1% debt and 70.9% equity.
Id. Ex. 33.
119
to be 5.7%. Barrick Report at 33. AT&T had an A credit rating, so AT&T could borrow at
that rate. Taylor used the Moody’s average bond yield on Ba1-rated corporate borrowers,
which was 5.4% at the time of the Freeze-Out. Taylor Report at 97, 98 n.222, 100.
Normally, the yield on a Ba1-rated bond would be higher than the rating on an A-rated
bond, but Barrick calculated the cost of debt using AT&T’s anticipated average cost of
debt over the next five years per AT&T’s 2009 Form 10K, while Taylor simply used the
yield on a Ba1-rated bond on the date of the Freeze-Out.
Damages should be measured at the time of the Freeze-Out, so this decision does
not use Barrick’s pre-tax cost of debt. This decision also rejects Taylor’s number, which is
based on a fictional credit rating for AT&T. This decision uses a pre-tax cost of debt for
AT&T of 5.2%. That figure appears in PwC’s valuation report for a different partnership
that was prepared on June 30, 2010, approximately ten weeks before the Freeze-Out. See
JX 1752 at ’673.
b. Equity Risk Premium
For purposes of an equity risk premium, each expert ironically used a figure that
was more favorable to the other side. Taylor adopted a risk premium of 5.2% from the 2010
Ibbotson SBBI Yearbook. Taylor Report at 95 & n.221; see JX 668 at 3. Barrick adopted
a risk premium of 6% from the 2011 Ibbotson SBBI Yearbook. Barrick Report at 29–30;
see JX 1793 at 3. Damages must be calculated as of the date of the Freeze-Out. The 2011
Ibbotson SBBI Yearbook reported 2010 data, so the equity risk premium it reported was in
effect when the Freeze-Out took place. This decision therefore uses Barrick’s figure.
120
c. Beta
To calculate beta, Taylor averaged the five-year weekly unlevered beta of a set of
comparable companies, then relevered the result using AT&T’s capital structure for a beta
of 0.78. Taylor Report at 96, Ex. 31. In her opening report, Barrick argued for a beta of
0.66, which she claimed to have calculated using AT&T’s five-year weekly levered beta.
Barrick Report at 30–31. In her rebuttal report, Barrick agreed on using a five-year weekly
beta but argued that Taylor’s methodology was incorrect because “[t]he operative reality
of the Partnerships is that they are a part of AT&T as a whole and it is AT&T’s cost of
capital that is relevant.” Barrick Rebuttal Report at 24–25. Barrick noted that Taylor had
calculated AT&T’s five-year weekly beta to be 0.75 and appeared to suggest that 0.75 was
correct, even though Barrick had calculated a figure of 0.66.
The court agrees that the Partnership was so intertwined with AT&T as to warrant
using AT&T’s five-year weekly beta. In the AMWOH decision, this court used AT&T’s
beta in a valuation of market-level entities to “reflect[] the reality that the Companies were
operated as part of Mobility, which in turn is part of AT&T.” 2013 WL 3865099, at *4.
The same reasoning applies here. The Partnership was a market-level entity that operated
as part of AT&T’s nationwide business.
The experts generated different figures for the five-year weekly beta. The court
confirmed the accuracy of Taylor’s figure. This decision therefore uses a beta of 0.75.
d. Size Premium
In a debate with significant implications for the damages award, the parties
disagreed over whether to apply a size premium. Taylor applied a size premium of 2.99%.
121
PwC did not apply a size premium. Neither did Barrick. This decision adopts Taylor’s size
premium.
A size premium increases the subject company’s cost of equity, increasing the
discount rate. That in turn lowers the present value of cash flows and results in a lower
valuation estimate.39 The rationale is that a small business faces greater overall risk than a
larger, more diversified one. Pratt, supra, at 193–94. The proponent of a size premium must
“prove the amount” of the premium. ONTI, Inc. v. Integra Bank, 751 A.2d 904, 921 (Del.
Ch. 1999); see Orchard, 2012 WL 2923305, at *21.
Whether to apply a size premium in this case presents a close question. It is
undeniable that the Partnership and its sister entities were small entities that operated in
geographically discrete areas. At the same time, AT&T managed the Partnership and its
sister entities as integral parts of its nationwide wireless business. A strong argument can
be made that each of the determinants of the Partnership’s cost of capital in a DCF model
should reflect the reality that AT&T was its capital provider. For example, the experts
agreed on using AT&T’s capital structure when calculating the WACC, and both used
AT&T’s cost of debt. Due to the imperative of controlling spectrum to support a nationwide
network, AT&T was highly unlikely to sell a market-level entity voluntarily. Using a size
premium thus fails to reflect their operative reality as permanent parts of AT&T’s national
wireless business.
39
The inclusion of a size premium changes the conventional CAPM equation for a
company’s cost of equity from Ke = Rf + (B * Rm), to Ke = Rf + (B * Rm) + Size Premium.
122
In two earlier appraisal cases involving similar market-level entities, the court made
different findings. In AMWOH, Taylor and AT&T failed to convince the court to apply a
size premium of 3.99%. The court acknowledged that a size premium “may be appropriate
to capture the additional risk associated with the unique risk factors of small entities,” but
the court found that “PwC’s determination not to use a small company risk premium is
more persuasive and recognizes that the Companies operated as part of a larger entity.”
2013 WL 3865099, at *4. In B&L Cellular, by contrast, the local partnership was operated
as part of United States Cellular. Taylor provided a persuasive explanation for a size
premium of 1.85%, and the court adopted it. 2014 WL 6882207, at *2.
Here, Taylor and AT&T contend that a size premium of 2.99% is appropriate.
Taylor explained that the small size of the Partnership and its sister entities put them at
“greater risk of significant variability in financial results.” Taylor Report at 96–97.
Speaking of the partnerships as a group, she cited possibilities such as “a Partnership
experiencing unique competition differences such as Metro or Leap launching in their
territory with much lower prices, or a severe weather event impacting the city, or a major
employer in the city leaving.” Id. at 97. At the same time, Taylor opined that it was
“important to recognize that the Partnerships have benefitted from being part of the larger
AT&T, which suggests that the Partnerships would be viewed as less risky than
independent companies of similar size given a stabilizing factor from their relationship as
part of AT&T.” Id. To account for these competing considerations, Taylor started with the
3.99% premium indicated by the “micro-cap” decile from the 2010 Ibbotson SBBI
123
Yearbook, then subtracted 1% to reflect AT&T’s involvement for a total size premium of
2.99%. Id.
Taylor’s approach accounted for the operative reality of the Partnership and its sister
entities. They were small companies operating in non-diversified local markets. Their
revenues in particular were vulnerable to competitive and economic conditions within their
geographic footprints. Compared to an integrated telecommunications business, the
partnerships lacked diversification. They therefore carried higher risk than larger entities.
At the same time, the entities were part of AT&T, which managed them as part of
its nationwide wireless network. This decision has accounted for that fact by using AT&T’s
financial metrics for other components of the WACC, such as AT&T’s capital structure,
cost of debt, and beta. AT&T’s support did not eliminate the Partnership’s dependence on
its local market for revenue, making a size premium appropriate. Taylor also made a
reasoned judgment to reduce her size premium by 1% to reflect AT&T’s support.
The plaintiffs note that PwC declined to use a size premium. That is true, but PwC
achieved a similar result through a higher beta. See JX 1209 at ’060. A higher beta indicates
that a company is relatively more sensitive to systemic risk than a diversified portfolio of
securities. Shannon P. Pratt & Roger Grabowski, Cost of Capital: Applications and
Examples 187–88 (4th ed. 2010). A size premium attempts to account for greater sensitivity
to risk due to firm size. Although not equivalent, both increase the subject company’s cost
of capital to account for attributes that make the company riskier than a diversified
portfolio. As Taylor explained, PwC first calculated a beta of 0.80, then “subjectively
increased the AT&T Inc. beta by 0.2 to account for the fact that pure-play wireless
124
companies have historically demonstrated higher systematic risk than diversified telecoms
(such as AT&T Inc.).” Taylor Report at 95; see JX 1209 at ’060 (“Given our observation
of higher betas for pure wireless companies, we selected 1.0 as the unlevered beta for the
WACC calculation.”). That adjustment was an alternative approach to applying a size
premium.
The plaintiffs have pointed out that the specific reasons Taylor cited in favor of a
size premium did not necessarily apply to the Partnership. Taylor provided those reasons
as examples. Small companies face greater risks, including risks that are unknown and
unanticipated. The size premium accounts for that increased risk.
This decision therefore applies Taylor’s size premium when quantifying the
damages award.
e. After-Tax Cost Of Debt
The parties’ experts used different tax rates when calculating the after-tax cost of
debt. Taylor used AT&T’s effective tax rate of 32.9% over a five-year period. Barrick used
AT&T’s marginal tax rate of 38.5%. Valuation treatises generally support using the
marginal tax rate of the borrower.40 Here, the borrower was AT&T, which supplied capital
to the Partnership. This decision therefore uses AT&T’s marginal tax rate of 38.5%.
40
See Robert F. Bruner, Applied Mergers & Acquisitions 266 (2004) (presenting
beta levering and unlevering formulas using “the target’s marginal tax rate”); accord
Richard A. Brealey & Stewart C. Myers, Principles of Corporate Finance 366, 525 (7th
ed. 2003); see also Robert W. Holthausen & Mark E. Zmijewski, Corporate Valuation:
Theory, Evidence & Practice 174 (1st ed. 2014) (recommending use of “the corporate tax
rate that is applicable to the interest deduction”); Shannon P. Pratt & Roger J. Grabowski,
Cost of Capital in Litigation: Applications and Examples 168 (2011) (noting that the Court
125
f. The Finding Regarding The WACC
The ingredients of the WACC formula and the CAPM calculation are as follows:
• The Partnership’s pre-tax cost of debt equals 5.2%.
• The Partnership’s capital structure for purposes of the WACC calculation is 29.1%
debt, 70.9% equity.
• The cost of equity is 10.89%, comprising a risk-free rate of 3.4%, an equity risk
premium of 6.00%, a beta of 0.75, and a size premium of 2.99%.
• The tax rate for valuing the after-tax cost of debt equals 38.5%.
Putting it all together, those calculations result in a WACC of 8.65%:
Ke = Rf + (ß * Rm) + Size Premium
Rf 3.40%
ß 0.75
Rm 6.00%
Size Premium 2.99%
Ke 10.89%
WACC = Kd * (1 – T) (D/V) + Ke * (E/V)
Kd 5.20%
T 38.50%
(D/V) 29.10%
Ke 10.89%
(E/V) 70.90%
WACC 8.65%
of Chancery “normally tax-effects the cost of debt based on the company’s marginal tax
rate”) (citing Cede & Co. v. Technicolor, Inc., 2003 WL 23700218, at *43 (Del. Ch. Dec.
31, 2003) (“Using the 46% tax rate agreed upon by both experts, the resulting after-tax cost
of debt is 7.54% . . . .”), aff’d in part, rev’d in part on other grounds, 884 A.2d 26 (Del.
2005)). But see Pratt, supra, at 217.
126
2. Perpetuity Growth Rate
The final step in projecting cash flows is to calculate a terminal value by extending
the cash flows into perpetuity then discounting the resulting value to the present. This
decision already has found that Taylor used an unreasonably low perpetuity growth rate
and that her justifications for adopting it were unpersuasive.41
The plaintiffs maintain that the court should adopt Barrick’s perpetuity growth rate
of 3%. They emphasize that when AT&T modeled the net present value of avoided
distributions in its buyout analysis, AT&T consistently forecasted an annual growth rate of
3.5% for the distributions. See, e.g., JX 547 at ’326. In response, Taylor pointed out that
AT&T’s valuation also used a higher discount rate, resulting in a lower present value than
Taylor derived. Taylor Tr. 799. So be it, but a discount rate is not interchangeable with a
perpetuity growth rate. AT&T’s valuation was made by finance executives to support a
request for a multibillion dollar investment that AT&T’s CEO ultimately approved, so that
growth rate is highly informative.
As discussed previously, conventional wisdom holds that a perpetuity growth rate
should not exceed real growth in GDP. At the time of the Freeze-Out, credible projections
41
AT&T notes that in the AMWOH decision, this court adopted Taylor’s perpetuity
growth rates of 1.5% to 1.7%. 2013 WL 3865099, at *5. In that case, the petitioners had
“largely abandoned” their expert’s calculations and had failed to support his growth rates
with “any underlying facts.” Id. This is a different case with a different record. Here, the
plaintiffs and their experts have persuasively established that AT&T expected significant
growth in its wireless business.
127
placed the rate of nominal GDP growth at 5.1%, suggesting real GDP growth of 2.5% to
2.7%. See Barrick Report App’x B; JX 628 at 3.
For purposes of its damages calculation, the court will use a perpetuity growth rate
of 2.7%. That growth rate adopts the high end of the range for real growth in GDP, thereby
recognizing that AT&T anticipated significant growth in its wireless business and that
although the legacy voice wireless was maturing, AT&T expected growth from a host of
other wireless-related businesses.
3. The Remedial Calculation
For purposes of awarding a damages remedy, the foregoing inputs result in an
implied valuation of the Partnership at the time of the Freeze-Out of $714,039,606.48. For
comparison, PwC valued the Partnership at $219 million. Taylor valued the Partnership at
between $171 million and $224 million. Barrick valued the Partnership at $946 million.
AT&T tried to cast doubt on any approach that would generate a valuation
significantly higher than the Freeze-Out price by drawing comparisons to AT&T’s failed
bid for T-Mobile. AT&T argued at length that its “valuation of T-Mobile corroborates
PwC’s and Taylor’s valuations,” even going so far as to argue that because the T-Mobile
transaction “bore objective indicia of reliability,” the court should defer to the prices that
AT&T paid. Dkt. 631 at 16–17. AT&T thus attempted to obtain the benefit of deference to
the deal price that is warranted in an arm’s-length transaction, despite relying on (i) a
different deal, that was (ii) signed after the Freeze-Out, (iii) ultimately failed to close, and
(iv) incorporated a range of value-confounding factors in the price.
128
This decision already has held that Taylor’s comparable transactions, including the
T-Mobile transaction, were not sufficiently comparable for valuation purposes. For the T-
Mobile transaction, the divergence is stark. The transaction multiples implied by AT&T’s
failed bid for T-Mobile reflected risks and costs that were not present in the Freeze-Out.
First, the T-Mobile deal included a $4.2 billion termination fee that AT&T would have to
pay if the deal did not close. See JX 2076 at 3, 17; see also Taylor Tr. 846–48. Due to the
relative size of the two companies, the antitrust risk was real. It also came to pass: The
antitrust authorities killed the deal, and AT&T took a $4.2 billion charge in the fourth
quarter of 2011. JX 2076 at 3, 17. A buyer like AT&T that faced antitrust risk would
discount its bid for that risk, resulting in lower revenue and EBITDA multiples. The
Freeze-Out carried zero antitrust risk.
Second, the T-Mobile deal carried integration risk. AT&T planned to spend $1.9
billion in 2012 and $1.8 billion in 2013 to integrate T-Mobile’s subscribers into AT&T’s
business. JX 1893 at ’938. A buyer facing significant integration risk would account for it
by reducing the purchase price, again resulting in lower multiples. The Freeze-Out carried
zero integration risk.
The attempt to compare the Freeze-Out to the T-Mobile transaction also fails
because it seeks to use a fair market value standard to evaluate the fairness of a controller
freeze-out. AT&T justifies the Freeze-Out by contending that “Plaintiffs could not have
negotiated a materially higher price from a third party.” Dkt. 631 at 2. But as discussed
previously, the question in a controller freeze-out is whether the minority investor has
received “the substantial equivalent in value of what he had before.” Sterling, 93 A.2d at
129
114. If the controlled entity benefitted from relationships with its affiliates or was part of a
larger business that operated synergistically, then the “what” that the minority investor
“had before” must include those benefits as part of its operative reality. Cf. Dobler v.
Montgomery Cellular Hldg. Co., 2004 WL 2271592, at *9 (Del. Ch. Oct. 4, 2004)
(rejecting argument that “stand-alone” value required exclusion of relationships with
affiliates), aff’d in part, rev’d in part, 880 A.2d 206 (2005); see also Boyer v. Wilm. Mat’ls,
Inc., 754 A.2d 881, 901 (Del. Ch. 1999) (including value of long-term contract in fair value
award). The controller does not get to create a hypothetical company devoid of those
benefits by imagining what a third-party buyer might pay for the entity standing alone.
Such an approach would result in the valuation of a hypothetical standalone entity, not the
actual entity that was the subject of the Freeze-Out.
That distinction applies in this case. AT&T operated the Partnership and its sister
entities as integrated parts of its nationwide wireless network. AT&T was not going to sell
the Partnership or any of its sister entities, because doing so would create holes in its
network and negate the promise of ubiquity that AT&T offered its customers. See JX 3551
at 3 (AT&T did not have any “significant past practice of voluntarily selling licenses”).
The remedial award must take into account the operative reality of the Partnership as it
existed at the time of the Freeze-Out, namely that it was an integrated part of AT&T’s
network. The analysis cannot treat the Partnership as if it were a fungible asset.
Barrick provided a more realistic reality check. Using AT&T’s corporate hurdle rate
of 6.2% and a 3% growth rate, Barrick capitalized the Partnership’s 2009 distributions into
perpetuity. That calculation produced a fair value estimate for the Partnership of $571.1
130
million. JX 3604 at 3. And simply capitalizing the 2009 distributions does not account for
the higher distributions that AT&T paid in 2010 or the growth that AT&T anticipated in
the Three-Year Plan.42
The remedial award also does not incorporate value for AT&T’s pervasive breach
of the Management Agreements. If the cash flow projections were adjusted to comply with
the Premium Provision and include a 25% premium for Shared Revenues and the 10%
discount for Sales and Marketing Expenses, then the concluded value of the DCF model
would have been $920,830,852.18. That valuation would not have accorded any value to
the derivative claim for AT&T’s past breaches.
For purposes of a remedial damages calculation following a proven breach of
fiduciary duty, valuing the Partnership at $714,039,606.48 is a responsible estimate. The
plaintiffs are entitled to 1.881% of that amount ($13,431,085), less what they already
received in the Freeze-Out ($4,119,390). The plaintiffs’ damages therefore equal
$9,311,695.
42
AT&T attempts to avoid the comparison by claiming that the Partnership’s
“distributions in the years leading up to the transactions exceeded both EBITDA and free
cash flow,” and that “Barrick extrapolates these excessive distributions in perpetuity, which
is impossible.” Dkt. 614 at 69; see JX 1209 at ’124; JX 3530, “Salem DCF” tab, Cell D31.
After removing the effect of taxes, applying the same analysis to Taylor’s projections for
the Partnership’s free cash flows in 2010 translates to a fair value estimate of $487.6
million.
131
C. An Award Of Attorneys’ Fees
The plaintiffs seek an award of attorneys’ fees. They contend that an award is
warranted because AT&T hired PwC, “a conflicted advisor who [AT&T] thought would
be ‘effective[] as a potential witness’ to put its name on valuations [AT&T] essentially
prepared itself.” Dkt. 626 at 87 (footnote omitted). The plaintiffs argue that failure to grant
their request would “encourage such outright disloyalty,” and that an award of attorneys’
fees is warranted for the additional reason that AT&T exits “this litigation with the
distinction of being the party most obstructive of the discovery process in this Court’s
experience.” Id. This decision denies the plaintiffs’ request.
“Delaware follows the ‘American Rule,’ which provides that each party is generally
expected to pay its own attorneys’ fees regardless of the outcome of the litigation.” Shawe
v. Elting, 157 A.3d 142, 149 (Del. 2017). There are exceptions to the American Rule. One
of them permits the court “to award attorney’s fees when the ‘losing party has acted in bad
faith, vexatiously, wantonly, or for oppressive reasons.’” Brice v. Dep’t of Corr., 704 A.2d
1176, 1179 (Del. 1998) (quoting Alyeska Pipeline Serv. Co. v. Wilderness Soc’y, 421 U.S.
240, 258–59 (1975)). “Although there is no single definition of bad faith conduct, courts
have found bad faith where parties have unnecessarily prolonged or delayed litigation,
falsified records, or knowingly asserted frivolous claims.” Shawe, 157 A.3d at 149 (cleaned
up).
“The bad faith exception is applied in extraordinary circumstances as a tool to deter
abusive litigation and protect the integrity of the judicial process.” Id. at 149–50 (internal
quotation marks omitted). A party who invokes the bad faith exception “must demonstrate
132
by clear evidence that the party from whom fees are sought acted in subjective bad faith.”
Lawson v. State, 91 A.3d 544, 552 (Del. 2014) (cleaned up).
In this case, AT&T has not engaged in the sort of conduct that warrants fee shifting.
As an initial matter, the fact that AT&T wanted to assess Gilcreast’s effectiveness as a
potential witness does not constitute an abuse of process. The record shows that AT&T
anticipated litigation over the Freeze-Out, and AT&T tried to shape the record to its
advantage. The court considered that evidence in the course of making its finding that
AT&T knew it was underpaying for the minority partners’ interests in the Partnership. That
evidence in turn supported the court’s finding that AT&T breached its duty of loyalty. It
does not follow that AT&T’s conduct constitutes an abuse of the litigation process itself.
As to AT&T’s conduct during the litigation, the plaintiffs failed to make the
showing required for the bad faith exception to apply. The plaintiffs did not point to any
evidence. They only paraphrased the court’s comment about AT&T’s conduct during
discovery. The bad faith exception requires more. The plaintiffs’ request therefore is
denied.
IV. CONCLUSION
The plaintiffs proved that AT&T breached its fiduciary duty of loyalty by effecting
the Freeze-Out in a manner that was not entirely fair to the plaintiffs. The plaintiffs are
entitled to damages of $9,311,695. Pre- and post-judgment interest will accrue on this
amount, compounded monthly, until the date of payment. Interest will accrue at the legal
rate, with the legal rate changing with fluctuations in the underlying reference rate.
133
Within thirty days, the parties will prepare a form of final judgment that has been
agreed upon as to form. If there are issues that need to be resolved before a final judgment,
then the parties instead will submit a joint letter that identifies those issues and proposes a
schedule for resolving them.
134