United States District Court, E.D. Virginia, Alexandria Division
TIM P. BRUNDLE, on behalf of the Constellis Employee Stock Ownership Plan and a class of other individuals similarly situated, Plaintiff
WILMINGTON TRUST N. A., as successor to Wilmington Trust Retirement and Institutional Services Company Defendant. Item Court's Findings of Impact on Price
Tim P. Brundle ("plaintiff" or "Brundle")
is a former employee of Constellis Group, Inc.
("Constellis"), and a former participant in an
Employee Stock Ownership Plan ("ESOP") sponsored by
Constellis. Defendant Wilmington Trust N.A.
("Wilmington") was the trustee for the ESOP in
connection with Constellis' creation of the ESOP.
Creating the ESOP involved purchasing 100% of Constellis'
voting stock in December 2013 (the "2013
Purchase"). Less than a year after the ESOP was created,
all its stock was sold (the "2014 Sale"). Plaintiff
alleges that the 2013 Purchase involved transactions and
payments prohibited by the Employee Retirement Income
Security Act of 1974 ("ERISA"), 29 U.S.C. §
1106, resulting in the ESOP paying an inflated price for the
Constellis stock. Specifically, plaintiff alleges that the
$4, 235 per share paid in 2013 was not the fair market value
of such stock, resulting in the ESOP overpaying for the stock
by $103, 862, 000, see PTX 512, which plaintiff seeks to
recover for the ESOP.
a bench trial, and for the reasons that follow, the Court
holds that Wilmington is liable for violating §
1106(a)(1)(A), causing $29, 773, 250 in damage to the ESOP,
but not liable for violating § 1106(a)(1)(B) or §
FINDINGS OF FACT
is a form of employee retirement benefit plan "designed
to invest primarily in securities issued by its sponsoring
company." Donovan v. Cunningham. 716 F.3d 1455,
1458 (5th Cir. 1983). Under the terms of the Plan Document
("Plan") governing Constellis' ESOP, the
participant employees would not immediately own the stock
purchased by the ESOP. Rather, the stock would be held by a
trust and allocated to employees over time, as an incentive
to remain with Constellis. The Plan defines a year of service
as 1, 000 hours. DTX 120 at 28, 13. An employee begins to
become vested in his shares at the completion of the second
year of service, an amount that increases 20% per year until
reaching 100% in the sixth year. Id. Alternatively,
the employee would become fully vested upon reaching the
designated retirement age while employed by Constellis.
Id. A. Constellis Background Thomas Katis
("Katis") and Matthew Mann ("Mann")
founded Constellis' predecessor, Triple Canopy, Inc.
("Triple Canopy"), as a private security firm in
the fall of 2003. Tr. at 1091:5-8. Troubled by the unsavory
reputation of the private security industry at that time,
they hoped to build Triple Canopy into a company that would
be a source of pride for its employees, who were primarily
retired members of the U.S. Armed Forces. Id. at
1095:20-22. Although the founders had a significant
operational role at the outset, they began to step back in
the spring of 2004, naming Ignacio Balderas
("Balderas") as Chief Executive Officer
("CEO"). Id. at 1092:8-14. Triple
Canopy's primary clients were the U.S. Department of
State ("DoS") and U.S. Department of Defense
("DoD"). Id. at 1094:23-24. Over the next
several years, the company experienced steady growth,
Id. at 1095:13-14, landing a series of large
contracts to provide security both for physical locations,
known as "fixed site security, " and for key
personnel, known as "personal security, " Tr. at
19:14-19. Triple Canopy also began to acquire other security
firms. See Id. at 1093:17-1094:3. These acquisitions
led the company to reorganize itself into Constellis, which
became the parent company, with Triple Canopy being its
largest subsidiary. See id.
first form of deferred compensation Constellis offered to its
employees was an employee stock option program. See Tr. at
1099:5-13. Driven in part by a desire to monetize that
program for the employees, in the fall of 2007 Constellis
began considering an offer from Cerberus Capital Management
to purchase the company. Id. Although the parties
engaged in extensive negotiations, the deal fell through. In
2010, with no prospective sale on the horizon, Constellis
cashed out its employee stock option plan as part of a
restructuring and replaced it with a profit sharing plan that
used Earnings Before Interest, Taxes, Depreciation, and
Amortization ("EBITDA") as its benchmark. Tr. at
next serious offer to buy Constellis came in 2012, when the
investment banking firm Teneo Holdings, Inc., attempted to
broker a deal in which a private equity firm, Vestar Capital
Partners ("Vestar"), would be the buyer. Although
Vestar initially offered a price in the range of $340 to $350
million, it reduced its offer to $275 million as the closing
date approached. Tr. at 1136:8-9. The deal fell through
because of this last minute reduction, which Katis
characterized as a "classic private equity move."
Id. at 1100:6-12 By the end of 2013, Katis owned
31.1% of Constellis' outstanding stock, Mann owned 29.2%,
Howard Acheson ("Acheson") owned 11.1%, John Peters
("Peters") owned 5.6%, and "other minority
shareholders" held the remaining shares. PTX 2 at 14.
Collectively, Katis, Mann, Acheson, and Peters are known as
Considering an ESOP
of 2013, Constellis board member Simon Crane
("Crane") approached its general counsel Juliet
Protas ("Protas") about the possibility of forming
an ESOP. Tr. at 1417:9-14. Several factors motivated Crane
and the Sellers to consider this option. Katis believed that
an ESOP provided him an "exit strategy" while being
consistent with his vision of Constellis as a company
focusing on taking care of its employees. Id. at
1418:16-22. Creating an ESOP is often driven by the
shareholders' desire for liquidity, Id. at
945:10-11, although Katis testified that he had no urgent
need for liquidity when the Constellis ESOP was formed,
Id. at 1184:3.
was receptive to the idea of an ESOP, in part because she had
participated in a panel discussion with ESOP specialist Scott
Meza of Greenburg Traurig LLP ("GT") a month
earlier. Tr. at 1417:18-19. She took Crane's suggestion
to Balderas, who authorized her to explore the idea.
Id. at 1418:3-7. At GT's recommendation, Protas
retained investment bankers CSG International
("CSG"), including managing partner George Thacker
("Thacker"), to advise Constellis on the prospect
of an ESOP. Id. at 1421:9-11.
prepared a series of PowerPoint presentations for the Sellers
and Constellis management about what an ESOP is and how to go
about forming one. See PTX 626, 628. The first, titled
"Preliminary ESOP Structural Analytics" and dated
September 25, 2013, explained that the "Status Quo"
left the shareholders with "no exit strategy" and
that they were "limited to dividends for
liquidity." PTX 626 at 6. Under those circumstances, CSG
reported that an ESOP "optimizes results for all
parties." Id. at 9 (emphasis in original).
For the Sellers, this meant creating an "attractive
liquidity structure, " optimizing "after-tax cash
flow, " receiving "$307.8MM net of taxes
over 6 years, " retaining "control until fully
paid, " and participating "in future
growth/[mergers and acquisitions] opportunities."
Id. (emphasis in original). CSG also pointed out
that by selling their shares to an ESOP, the Sellers would be
able to roll over capital gains taxes, saving 23.8% on
federal taxes. Id. at 18. Employees would benefit
through a management incentive program and by obtaining
equity in the ESOP. Id. at 9. Creating an ESOP in
2013 would result in Constellis becoming a "tax-free
business" beginning in 2014 and receiving approximately
$30 million through a tax refund for taxes paid in 2013;
however, to fully realize these tax benefits, an ESOP had to
own 100% of the company's stock. Id.
not recommend a "traditional 100% ESOP." PTX 626 at
17. Instead, it proposed "a more advanced 100% structure
that CSG developed." Id. In this "more
advanced" ESOP structure, the Sellers would sell the
ESOP 90% of their shares, and exchange the remaining 10% for
warrants, id, which Katis described as "equity-like,
" Tr. at 1173:15. The warrants would be financial
instruments entitling the Sellers to buy back equity in
Constellis at a designated price, known as the "strike
price, " during a certain period of time. See
Id. at 1173:19-21. Under CSG's proposed
structure, the Sellers as warrant holders would retain
significant elements of control over the company, most
notably the ability to appoint a majority of the board of
directors. Id. at 1603:21-22. Although the warrants
would not create immediate value for the Sellers, eventually
the sellers would be able to capture the warrants' value
"by (1) selling the warrants back... in the future, or
(2) selling the company." PTX 626 at 17. Under CSG's
proposal, the Sellers would "control the company post
transaction, " giving them the "flexibility"
to follow through on these options. Id.
addition, because an ESOP rarely, if ever, begins its life
with substantial assets, it is created by a heavily leveraged
transaction in which the ESOP borrows the funds with which it
buys the stock. Under the plan proposed by CSG, the
Constellis ESOP was to borrow from the Sellers to buy their
stock, meaning that the Sellers would also become the
company's creditors. CSG pointed out that because the
Sellers would "control the [b]oard, " they would
also "control the timing of payment" on that debt.
Id. at 19. CSG even suggested that Constellis might
take on new debt in order to pay down the seller notes more
presentation stressed the after-tax benefits to the
shareholders of selling Constellis to the ESOP, PTX 626 at
26, and the Court finds that these after-tax benefits for the
Sellers were a major incentive for creating the ESOP. To
highlight the advantages of an ESOP sale, CSG pointed out
that to "match the $307.8MM in after-tax proceeds [to
the Sellers], a traditional [mergers and acquisitions]
transaction would have to" value Constellis at $421
million, which exceeded CSG's estimated valuation of
Constellis by $100 million. Id. CSG continued to
stress the benefit to the Sellers, reporting on September 26,
2013, one day after its preliminary presentation, that CSG
had "looked at every possible scenario" and had not
"seen any possible alternative that produces more after
tax cash than the ESOP." PTX 625. Similarly, in a
presentation to the Sellers and Constellis' management
team on October 4, 2013, CSG focused on demonstrating the
benefits to the shareholders of the proposed
"advanced" ESOP. PTX 628.
after the September 26, 2013, presentation, Constellis'
management and the Sellers decided to move forward with the
proposed ESOP. CSG had recommended that Constellis hire
Wilmington to act as the trustee for the ESOP. Protas
approved that choice after speaking with Greg Golden
("Golden"), a vice president of Wilmington,
discussing the choice with Constellis' outside counsel,
and conducting Internet research on Wilmington. Tr. at
1422:24-1423:4. Wilmington was chosen because Protas was
impressed by her initial telephone conversation with Golden
and considered Wilmington to be a "household name."
Id. Within Wilmington, the Constellis transaction
was overseen by the Fiduciary Services Sub-Committee
("FSSC"), which had four voting members for
purposes of the Constellis transactions: Golden; Jennifer
Matz ("Matz"), who was a Wilmington vice president
and chair of the FSSC; John Lindak ("Lindak"); and
Boyd Minnix ("Minnix"). The fifth FSSC member,
Karen Bonn ("Bonn"), served as Wilmington's
relationship liaison for Constellis and was therefore recused
from voting on matters related to the Constellis
Id. at 84:5-9. Wilmington, in turn, retained Taylor
English Duma, LLP ("Taylor English"), to act as its
legal counsel because that firm was already on a list of
approved law firms maintained by the legal team of
Wilmington's parent corporation, M&T Bank.
Id. at 326:7-15. The Taylor English attorneys
working with Wilmington included corporate partners Denny
Summers ("Summers") and Emily Horn
("Horn"), both of whom had worked on ESOP
transactions for many years. DTX 251, 252.
also needed a financial advisor, for which it turned to the
firm of Stout Risius and Ross ("SRR"), which had
extensive ESOP experience and had been on Wilmington's
list of approved financial advisors since 2009. DTX 5 at 1.
As of 2013, it had provided "valuation and advisory
services to every major institutional ESOP trustee in the
United States" and its ESOP practice included 15 full
time professionals, including Scott Levine
("Levine") and Aziz El-Tahch
("El-Tahch"), who were the primary analysts tasked
with the Constellis valuations. See Id. at 1-2.
Levine was a Certified Public Accountant ("CPA"),
Accredited in Business Valuation ("ABV"), a
Chartered Financial Analyst ("CFA"), and an
Accredited Senior Appraiser ("ASA") with 18 years
of experience who had performed roughly 1, 500 ESOP
valuations. Id. at 2. El-Tahch was a CFA with 12
years of experience who had conducted roughly 750 ESOP
Constellis had no previous relationship with CSG or
Wilmington, SRR, TED, CSG, GT, and Wilmington had worked with
one another regularly. Records show that between 2010 and
2015, CSG was listed as the referral for more than 50% of
Wilmington's ESOP projects. See PTX 159. SRR acted as
Wilmington's financial advisor for 14 of those ESOPs.
Id. As of the date Constellis engaged Wilmington,
CSG had already referred 23 ESOPs to Wilmington. Id.
GT had referred business to Wilmington on at least four
occasions, and actually served as Wilmington's counsel in
one instance. Id. SRR and Wilmington have worked
together at least 25 times, id, and Matz could not recall
Wilmington ever asking SRR to revise any of its valuation
reports, Tr. at 1405:4-13.
services in connection with the 2013 Purchase, Wilmington
charged Constellis a flat fee of $150, 000, to be paid
regardless of whether or not the ESOP transaction closed. Tr.
at 335:19-23. If the transaction closed, Wilmington would
also receive a minimum payment of approximately $80, 000 per
year in fees for serving as the ESOP's ongoing
Id. at 336:16. Protas approved the fee arrangement,
after consulting those familiar with the industry, finding it
to be reasonable in light of the overall amount at stake in
the deal. Id. at 1423:18-21. There is no evidence in
the record suggesting that these fees were unreasonable and
the Court finds that they were reasonable.
Constellis' Value in 2013
crucial step in forming an ESOP is to arrive at the fair
market value of the stock that the ESOP is going to purchase.
The "fair market value" is "the price at which
the property would change hands between a willing buyer and a
willing seller, neither being under any compulsion to buy or
sell and both having reasonable knowledge of relevant
facts." Estate of Godley v. Comm'r of Internal
Revenue. 286 F.3d 210, 214 (4th Cir. 2002) (internal
citations and quotation marks omitted). It is a widely shared
view, and the Court finds, that valuing closely held stock is
an "inexact science, " see, e.g., Donovan.
716 F.2d at 1473, in which valuation professionals commonly
employ two basic methodologies, each of which incorporates a
substantial degree of latitude for "judgment calls"
by the analyst, see, e.g., Tr. at 672:19-20;
first method, which all of the experts considered more useful
in trying to value a closely held corporation like
Constellis, is the Discounted Cash Flow ("DCF")
method. See PTX 1 at 53. This method begins by estimating the
company's projected cash flow far enough into the future
to arrive at the company's "terminal growth rate,
" or the rate at which it can plausibly be expected to
grow indefinitely into the future. Id. at 54. That
projection is then discounted by a number of factors designed
to take into account current cash reserves and liabilities.
Id. at 55. Because the DCF is calculated by default
on a marketable, controlling-interest basis, additional
discounts may be necessary to reflect the nature of the
shareholder's stake. Id.
second method, the Guideline Company Method
("GCM") begins by identifying various publicly
traded companies (the "guideline companies") that
experience similar market forces as the company that is the
subject of the valuation (the "subject company").
PTX 1 at 45. When conducting a GCM analysis, the goal is to
identify a factor by which the subject company's EBITDA
or revenue must be multiplied to estimate the company's
total enterprise value (a "multiple"). Id.
The first step in determining the subject company's
multiples is to identify the guideline companies'
multiples. Id. To calculate the guideline
companies' multiples, each guideline company's stock
price is used to derive an enterprise value for that company.
See Id. at 46. That enterprise value is then divided
by the guideline company's EBITDA (or revenue) to arrive
at the multiple. Id. at 47. By default, these
calculations are made on a single share basis, and may be
adjusted based on market forces. PTX 2 at 46. Once the
multiples for the guideline companies have been identified,
the risk that the guideline companies face is compared to any
risks the subject company may have. Id. at 50. If
the subject company is riskier than the guideline companies,
lower multiples should be applied to the subject
company's EBITDA and revenue to estimate an enterprise
value for the subject company. Tr. at 1216:13-20. If the
subject company is less risky, higher multiples are used. See
analyst must next determine how to reconcile the results of
the DCF and GCM methods. All analysts agreed, and the Court
finds, that it is preferable for these numbers to be close to
each other. Tr. at 1534:19-22. If they are not, some process
must be used to merge the two results into a single
enterprise value for the company. Once the enterprise value
is determined, the company's liabilities must be
subtracted and any cash on hand added to arrive at an equity
value. PTX 2 at 59. The equity value may be discounted
further to account for certain business realities, such as a
lack of control or a lack of marketability. See Id.
at 60. When the final equity value has been reached, that
number is divided by the number of outstanding shares to
arrive at a per share value for the stock in question.
Id. at 61.
the second valuator to calculate a fair market value for
Constellis' stock in 2013. The first was Andy Smith
("Smith") of the McLean Group ("McLean"),
who had been performing an annual valuation of a single share
of Constellis stock for at least the three preceding years.
See PTX 1, 4, 5. Constellis used the McLean reports to price
employee stock options and for financial reporting purposes.
Tr. at 536:17-19. The 2013 McClean report provided a
valuation of Constellis as of January 31, 2013, incorporating
only information that was "known and knowable" as
of January 31, even though the report was not completed until
June 19, 2013. Tr. at 652:16-19. Smith performed both a DCF
and GCM evaluation, although he found the guideline companies
to be so different from Constellis that he did not ultimately
give the GCM result any weight in his final valuation. See
Tr. at 655:1-23. This approach was consistent with his habit
of performing both evaluations and then choosing the more
reliable, except where the two results were relatively close.
Id. at 655:14-19.
performing his DCF analysis, Smith relied on
"management's prepared forecasts" which only
covered "the remaining 11 months of 2013 to 2014."
PTX 1 at 54. In using those projections to arrive at an
enterprise value, Smith had to calculate a "weighted
average cost of capital" ("WACC"), which is
used to discount projected cash flows. Id. at 55.
Smith calculated a WACC of 15%, in part by factoring in a
"Specific Company and Industry Risk Factor" of 7%.
Id. Smith acknowledged that a 7% risk is relatively
high, but he felt it was necessary because "Constellis
represents more of a risk to potential investors than that of
the sector as a whole, " in large part because of its
"heavily concentrated contract backlog."
Id. at 57. Given that a high percentage of
Constellis' revenue was tied up in a small number of
contracts primarily involving two government agencies, the
Court finds that it was reasonable and prudent for Smith to
adjust his valuation downward due to the company's
performance being so highly dependent on the continued
performance on those few contracts. Factoring in the 15%
WACC, Smith estimated Constellis' enterprise value to be
$213, 397, 018. Id. at 10. After adding in cash and
subtracting debt, he calculated Constellis' equity value
at $165, 015, 140. Id. This resulted in an estimated
fair market value of $1, 838.11 per share for the voting
stock and $1, 746.20 for the non-voting stock after further
discounting for lack of control and lack of marketability.
than one year later, SRR produced a draft evaluation, dated
November 14, 2013. PTX 170. Using the same basic approach
used by McLean but differing in the details, that evaluation
concluded that the fair market value range for a single share
of stock on a controlling interest basis was between $3, 865
and $4, 600. Id. at 35. One of the most significant
differences between the two valuations was that SRR factored
both the GCM and DCF methods into its final conclusion, and
added a 10% control premium to the comparable company
multiples in its GCM analysis. According to SRR, adding a
control premium was appropriate because the ESOP would hold
100% of the stock and have some rights typically associated
with being a majority shareholder. See PTX 2 at 116. SRR chose
multiples for Constellis that were at or below the low end of
the range produced by the guideline companies' multiples
for EBITDA, but mostly above the median for revenue. PTX 170
at 24. This produced a range of enterprise value from $290
million to $345 million. Id. Like McLean, SRR's
DCF analysis began by adopting management projections. See
PTX 170 at 26. According to SRR and Wilmington, Constellis
had a relatively "robust" process for developing
projections. Tr. at 1378:19-25. Both lauded the "bottoms
up" approach that Constellis used, Id. at
1257:17, meaning that the projections were developed by
looking at each contract or prospective contract, and giving
the contract one of three designations: "backlog, "
for an awarded contract upon which more revenue was due;
"option, " for contracts that Constellis had the
option to obtain; or "pipeline, " for contracts for
which Constellis was planning to submit a proposal and
anticipated a non-trivial chance of winning. Id. at
1354:22-1355:5. The process was overseen by Raymond Randall
("Randall"), who was then Constellis' Senior
Vice President for Strategic Initiatives, and who reported to
Magnani. Id. at 1353:16-17; 1355:11.
gave SRR projections that extended to the year 2018, as
opposed to the one year of projections that McLean received.
In these projections, Constellis advised SRR that it expected
EBITDA of $67.5 million for 2014, compared to the $52.2
million figure it had given McLean. Compare PTX 170
at 26, with PTX 1 at 59. SRR considered those projections to
be "conservative, " both because of the method
Constellis used and because historical data showed that
Constellis had beaten its projections almost every year. Tr.
at 1204:9-22; 1517:2-6.
also differed somewhat from McLean in how it calculated the
WACC discount rate. Of particular significance, SRR used a
metric called "beta" to assess the risk of
Constellis relative to that of the industry overall. PTX 170
at 27. A beta of 1.0 would mean that Constellis faced the
same risk as the industry as a whole. See Tr. at 388:9-16. A
lower number meant Constellis was less risky, and a greater
number meant Constellis was more risky. See id SRR chose a
beta of 0.7, reflecting its belief that Constellis was a less
risky investment than other companies in the industry. PTX
170 at 27. SRR's overall WACC was 11.5%. Id.
Using these numbers, SRR's enterprise value range under
the DCF method was $290 million to $322 million. Id.
all the experts in this case, SRR felt that the DCF method
was better suited to valuing a closely held corporation like
Constellis. Accordingly, it gave the DCF value twice the
weight of the GCM result. See PTX 170 at 35. As a result, SRR
concluded that Constellis' enterprise value ranged from
$275 million to $330 million. Id. After adding cash
and removing debt, Constellis' equity value was found to
range between from $283 million and $338 million.
Id. SRR then applied a 5% discount for lack of
marketability. Id. Although there was no
readily accessible market for Constellis' stock, ESOP
participants would be able to exercise a "put"
option to sell the stock back to the company, which SRR felt
justified a small discount.
concluded that the range of fair market value of equity was
$256 million to $305 million, with the median being $281
million. PTX 170 at 35. Divided equally among the number of
shares outstanding, SRR found the fair market value per share
to range from $3, 865 to $4, 600, with the median fair market
value of equity per share being $4, 232.50, which SRR rounded
up to $4.235. See id.
party called a financial expert to testify concerning
SRR's work. Dana Messina ("Messina"),
plaintiffs expert, is the Chief Executive Officer of Kirkland
Messina, a financial services firm. Tr. at 693:13-21. He
holds a Master's in Business Administration from the
Harvard Business School, PTX 152 at 33, has 25 years of
experience providing business valuations in ESOP
transactions, Tr. at 694:2-13, and has consulted for the
Department of Labor in over 100 investigations, Id.
at 696:7-12. Jeffrey Tarbell ("Tarbeir), defendant's
expert, is a director of the financial services firm Houlihan
Lokey and a Co-Head of their ESOP Valuation practice. DTX 232
at 5. He holds a Master's in Business Administration from
the University of Chicago and the ASA and CFA accreditations.
Id. He is currently vice chair of the Business
Valuation Committee of the American Society of Appraisers and
is the immediate past chair of the Valuation Advisory
Committee of the ESOP Association. Id.
Wilmington's Evaluation of the Constellis ESOP
Court finds that Wilmington rushed its evaluation of the
Constellis ESOP, failed to follow its own policies, and
failed to adequately vet SRR's conclusions.
had scheduled a meeting to discuss the November 2013 SRR
valuation report for Thursday, November 14, 2013 at 1:00 p.m.
Although Wilmington's policy called for receiving any
valuation report at least 48 hours in advance of any meeting
scheduled to discuss the report, Tr. at 72:10-18, SRR's
November 2013 Report was not forwarded to the FSSC until 4:58
p.m. on Tuesday, November 12, and not all members of the FSSC
had reviewed the entire draft before the meeting with SRR.
Golden testified that he did not need to review every part of
the report because he had already seen Constellis'
audited financial statements as well as a Confidential
Information Memorandum ("CIM"), which included much
of the raw data underlying SRR's analysis. Tr. at
98:16-99:3. Mate's testimony about whether she reviewed
the entire report before the meeting was inconsistent, and
the Court finds her original answer at her deposition when
she testified that she could not specifically recall reading
the report more credible than her contradictory trial
Id. at 1400:15-1401:17.
Golden and Matz took handwritten notes at the November 14
meeting, which was conducted by conference call. See PTX 173;
174. Each relied heavily on those notes when asked to discuss
what occurred at the meeting, and neither had any specific
recollection beyond what was preserved in the notes. Those
notes showed that SRR discussed several features of its
report, including Constellis' growth and diversification
efforts; SRR's conclusion that the projections provided
by Constellis' management were "conservative, "
PTX 174 at 2; the management role of the founders;
Constellis' proposed acquisition of CGI Group and
Strategic Social, LLC; the 2012 Vestar offer to buy
Constellis, which SRR reported as having involved a $338
million enterprise value; an audit by the Defense Contracting
Auditing Agency ("DCAA") for overtoiling by $62
million; and the valuation conclusions that SRR reached in
its draft report. Id. at 2-4.
El-Tahch finished presenting SRR's report, members of the
FSSC asked a number of questions about SRR's analysis. A
non-voting FSSC member asked about the financing terms for
the ESOP transaction. PTX 174 at 5. Lindak inquired about how
Constellis' projections incorporated risk. Id.
Tr. at 1487:10-15. He also asked whether the different types
of outstanding stock were treated differently for purposes of
the control premium (they were not). PTX 174 at 5. Summers
asked about the DCAA audit, and was assured that Constellis
was confident that it would not be liable for having to repay
$62 million, in part because Constellis believed that the
statute of limitations had run on some claims. Id.
at 6. Matz asked whether there was a chance that
Constellis' planned acquisitions of CGI and Strategic
Social could fall through, and El-Tahch advised that they
were "advanced but could always fall through."
Id. He assured her that if that happened, the
transaction price would not "fall out of [SRR's]
range." Id. A question was asked about who
formulated the projections, and El-Tahch responded that it
was independent management rather than the Sellers.
November 14 meeting, Golden admits that he knew that another
firm had prepared a valuation of Constellis in 2013. SeeTr.
at 120:23-121:6. Even so, no one from Wilmington asked to see
the McLean report. See Id. at 121:14. In fact,
Golden could not even say that anyone from Wilmington had
asked basic questions about the McLean report, including who
performed it, when, and what methodology it employed. See
Id. at 121:7-14.
meeting, Wilmington also held a discussion about how to
proceed with the deal. The FSSC decided that it would seek a
special indemnity from the Sellers for the full amount of the
potential DCAA audit liability, notwithstanding
Constellis' assurances that it was unlikely to have to
pay anything. Tr. at 1013:8-14. With that proviso, the FSSC
authorized Golden, Bonn, SRR, and Taylor English to carry out
negotiations on a price per share within a range of $3, 900
to $4, 235. PTX 174 at 7. There is nothing in any of the
Wilmington officials' notes reflecting any questioning
about the range of share prices SRR presented or about how to
negotiate an appropriate price. In terms of timing, SRR noted
that minority shareholders had to be given a tender offer at
least 20 days before closing. Between that requirement and
the December 31 deadline for being eligible to obtain the tax
advantages for 2013, there was a relatively short window in
which to agree on a price with Constellis and extend the
tender offer. PTX 174 at 1; DTX51 at 1.
and CSG sent their opening position of $4, 525 per share, a
figure just $75 below the maximum of the range calculated by
SRR, on Wednesday, November 13, 2013, a day before
Wilmington's meeting. Negotiations over the price and
term sheet began in earnest on Friday, November 15, 2013, the
day after Wilmington's meeting with SRR. Throughout those
negotiations, Constellis was represented by CSG's
Thacker, and Wilmington communicated with Thacker only
through representatives from Taylor English, although the
lawyers spoke with Golden and Bonn to develop a negotiation
strategy. Wilmington's opening bid was $3, 900, the
bottom of the range it was willing to pay. DTX 56 at
TED016013. Thacker replied at 3:02 p.m., proposing that the
parties "come to an agreement on the per share
value" while Constellis and CSG worked on their response
to the term sheet "because the tender offer documents
are scheduled to go out next Monday." PTX 50 at 8. In
the same email, he lowered Constellis' asking price to
$4, 350 per share. Id. Sometime shortly after this
email, someone from Wilmington's team proposed a price of
$4, 100 per share, to which Thacker responded with $4, 250 as
the per share price, remarking that "in the interest of
time, " Wilmington should respond to the share price
"while the term sheet is being finalized."
Id. at 7. At 4:50 p.m., Wilmington rejected $4, 250,
informing Thacker that its "best and final offer"
was $4, 235 per share and requiring that "the period
during which the Warrant Holders can control the Board of
Directors... be limited to 12 years." Id. at 6.
Thacker accepted the share price on behalf of Constellis at
4:59 p.m. Id. at 5-6.
parties did not reach a final agreement on the complete term
sheet until Friday, November 22, 2013. In the final version,
the warrants were divided into two series, with "Series
A" expiring at the end often years and "Series
B" expiring at the end of 15 years. PTX 50 at ¶
709. Only the holders of the Series A warrants were entitled
to designate members of the Constellis board of directors.
Id. at ¶ 710. According to the term sheet, the
strike price of the warrants was to be set at "no less
than 90 percent of the post-transaction fair market value per
share." Id. at ¶ 709. By email, Thacker
assured Wilmington's team that Constellis would be using
100% of the fair market value as the strike price, and that
"a higher strike price is more beneficial for the
Trustee." Id. at 2.
the terms of the transaction, Constellis agreed to adopt a
"management retention plan." PTX 50 at ¶ 710.
Essentially, this meant that it would pay a cash bonus to
certain key members of the management team if they agreed to
stay on for a set period of time following the transaction to
ensure continuity of management, which would be a benefit to
the ESOP as the new owner of Constellis. The final term sheet
did not specify how long management would need to stay to
earn the bonus, a change from an earlier version that would
have required staying on for 36 months. Compare PTX
50 at ¶ 710, with DTX 56 at TED016019. It was
"anticipated" in the term sheet that the management
incentive plan would be paid in cash and would be worth 5% of
the transaction price in total, meaning that the amount paid
to management was linked to the overall sale price. See PTX
50 at ¶ 710.
of the cash bonus, the term sheet authorized Constellis to
issue stock appreciation rights ("SARs") to
management. PTX50atWT710. Although the exact group of people
eligible for SARs was to be determined later, the Sellers
were prohibited from participating in the SARs program. PTX
50 at ¶ 710. When the transaction closed, there was no
guarantee that the SARs would be issued or exercised.
Nevertheless, the SARs were to be an incentive to "help
management get to" Constellis' projected growth
levels, Tr. at 156:25, and the Court finds that all parties
expected the SARs eventually to be issued.
the Sellers and Constellis agreed to include "reasonable
commercially acceptable and customary representations and
warranties[.]" PTX 50 at ¶ 711-12. For most
potential breaches, the aggregate amount of the
indemnification would be capped at "30% of the
Transaction Price for actual loss due to breach of
representation and warranties through setoff of the Seller
notes." Id. at ¶ 712. In other words, the ESOP would
only be reimbursed for a breached warranty for an amount up
to 30% of the total purchase price. Id. Moreover,
the ESOP would only receive cash as compensation for a
breached warranty if the value of any resulting injury
exceeded the balance of the debt owed to the Sellers;
otherwise, it was to be compensated by reducing the amount
owed to the Sellers. Id. Notwithstanding the general
30% cap, the Sellers agreed to indemnify the ESOP trust for
the full amount of any liability arising from any
"adverse outcome in the DCAA matter." Id.
tender offer issued on Monday, November 18, 2013, four days
after Wilmington's meeting with SRR, with December 20 set
for closing. SRR issued a complete opinion on the fairness of
the transaction, including a revised valuation of the stock,
in the days leading up to the closing. That December 2013 report
contained several revisions from the November draft. By
December, Constellis' proposed acquisition of CGI had
fallen through. As a result, projected revenues had declined.
Tr. at 472:11-13. The second major change related to
outstanding stock options. In November, SRR had assumed that
nearly all people holding stock options would exercise them
before the transaction closed, see Tr. at 415:7-12, diluting
the value of the company by $4.5 to $5.9 million, PTX 170 at
35. By December, it was clear that far fewer people had
actually done so, Tr. at 1262:8-9, reducing the range of
dilution to $300, 000 to $400, 000, PTX 2 at 61. These
changes largely balanced one another out in the final
valuation reconciliation. In the end, the GCM enterprise
value range was unchanged, and the DCF range shifted downward
by $2 million at the low end and $8 million at the high end.
Compare PTX 2 at 61, with PTX 170 at 35.
This brought the concluded enterprise value range down to
$275 million to $325 million. Compare PTX 2 at 61,
with PTX 170 at 35. Instead of providing a revised median
market value on a per share basis, the December report simply
repeated the agreed upon purchase price of $4, 235. PTX 2 at
numbers from the 2013 McLean and SRR reports discussed above
are summarized in the following table:
McLean January 2013
SRR November 2013
SRR December 2013
GCM Enterprise Value
$290 to $345 million
$290 to $345 million
DCF Enterprise Value
$213, 397, 018
$267 to $322 million
$265 to $314 million
Concluded Enterprise Value
$213, 397, 018
$275 to $330 million
$275 to $325 million
Marketable, Controlling Interest Value of Equity
$165, 015, 140
$283 to $338 million
$279 to $329 million
Fair Market Value on a Per Share Basis
$3, 865 to $4, 600
$3, 865 to $4, 555
those changes, SRR opined that
the consideration to be paid by the ESOP for shares of Class
A Common stock pursuant to the terms of the ESOP Purchase is
not greater than Fair Market Value of such shares; the
interest rate[s][on the debt the ESOP was to take on]... are
not in excess of a reasonable rate; the financial terms ...
are at least as favorable as would be the financial terms of
a comparable loan resulting from arm's-length
negotiations between independent parties; and the exercise
price of the Warrants, on a per share basis, is at least
equal to 90% of the Fair Market Value of the underlying
common stock on a per share basis; and the terms and
conditions of the Transaction, taken as a whole, are fair to
the ESOP from a financial point of view.
DTX 111 at 4.
sent the draft of the report to the FSSC at 4:06 p.m. on
December 18. DTX 97. Again, less than 48 hours after
receiving that 125-page report, on December 19, 2013, the
trustee team held a final meeting with SRR to discuss the
fairness opinion. The meeting lasted only half an hour, and
FSSC members asked very few questions. DTX 106; 107; see also
Tr. at 185:4-5. Indeed, Lindak did not even attend the
meeting. Tr. at 1491:23. At the conclusion of this meeting,
the FSSC approved the 2013 Purchase. Tr. at 1381:20-21.
December 20, 2013, the 2013 Purchase closed. The parties have
stipulated that the ESOP purchased purchased 47, 586.55
shares of Class A Common Stock at $4, 235 per share, DTX 112
at 1, 8, making the total purchase price $201, 529, 032.77.
[Dkt. 139] at 3. The Stock Purchase Agreement
("SPA") provided that the Sellers would exchange
the remainder of their stock for warrants, making the ESOP
the owner of 100% of Constellis' outstanding stock. DTX
112 at 1, 28-29. The funding for the 2013 Purchase came from
three sources: approximately 24% from a cash contribution
from Constellis, approximately 7% from a loan from
Constellis, and approximately 69% from a loan from the
Sellers. PTX 2 at 7. On December 27, the parties refinanced
the transaction, DTX 118, with Constellis assuming the
ESOP's debt to the Sellers, and the ESOP executing a note
in favor of Constellis for that amount. PTX 2 at 8.
with the term sheet, the SPA contained a series of
representations and warranties, which were divided between
those made by the Sellers and those made by Constellis.
See DTX 112 at 8-25. Significantly, it was
Constellis, not the Sellers, that represented to the ESOP
that "there has not been any event or condition of any
character that has or would be reasonably expected to have a
Material Adverse Effect" on the company. Id. at 11.
By contrast, the Sellers made no warranty about the accuracy
of Constellis' financial projections.
Investor Rights Agreement ("IRA"), executed at the
same time as the SPA, essentially enabled the Sellers to
maintain control of Constellis by providing that, out of a
five-member board of directors:
[F]or so long as all of the Series A warrants have not been
exercised or terminated, the holders of at least 51% of the
outstanding Series A Warrants shall have the right to
designate three of the members of the Board, and to select
the replacement Board member in the event of the resignation,
death or incapacity of any member of the Board designated by
DTX 115 at 6. A fourth director would "be independent
from management of the Company, who shall be selected by the
remaining members of the Board[.]" Id. In
short, under the IRA, the ESOP had the power to appoint only
one member of the board. The IRA provided that the warrant
holders could not exercise their power in a way that caused
"the Trustee to violate its fiduciary obligations under
ERISA, " but did not identify any specific mechanism by
which the trustee could stop the designation of a board
member if it felt that its obligations under ERISA were in
danger of being violated. See id.
extremely unusual for the warrant holders to be able to
appoint a majority of the board members when the ESOP held
100% of the company's shares. Smith testified that he
knew of only one other ESOP that owned 100% of the shares but
did not have a majority of the board of directors, and that
the Department of Labor had opened an investigation into that
619:25-620:2. Indeed, Constellis' own counsel,
Marc Baluda ("Baluda") of GT, sent an email to
Protas and Thacker, among others, in November asking how many
board members they envisioned after the transaction. PTX 242
at CSG027460. He observed that "warrant holders
nor[m]ally have the right to appoint a minority, so two of
five, three of seven, etc." Id. Thacker replied
that in CSG's deals, warrant holders have "been
given the right to appoint a majority." Id., He
explained that this was "how they retain operational
control over the company until they've been paid."
Id. The ESOP's lack of control was further
reflected in the plan document ("Plan"), which
directed Wilmington to "vote all Employer Securities
held by it at such time and in such manner as the
Administrator [i.e., Constellis] decides[.]" DTX 120 at
46. The Plan identified several exceptions to that
requirement, with the most relevant being for a "merger
or consolidation, recapitalization, reclassification,
liquidation, dissolution, sale of substantially all assets of
a trade or business, or such similar transaction as may be
prescribed in Treasury regulations." Id. Golden
and Summers agreed that a sale of the company, as opposed to
a sale of its assets, was not covered by this exception. Tr.
at 249:24-250:1; 1041:14-16. Therefore, under the Plan
Wilmington had no authority to break from Constellis'
instructions when voting on a sale of the company.
companion document, executed to establish the trust to hold
the ESOP's assets (the "Trust Agreement"),
further outlined the trustee's duties. According to
§ 3.03 of the Trust Agreement, the trustee had to vote,
tender, and exchange the stock it held on behalf of the ESOP
"in the manner set forth in the Plan... and consistent
with its duties described in Section 2.03 herein." DTX
121 at 9. Section 2.03, in turn, provided that the Trustee
must perform its duties "in accordance with the terms of
the Trust Agreement to the extent, in the good faith judgment
of the Trustee, that the Trust Agreement is consistent with
the provision of the [Internal Revenue] Code and ERISA."
Id. at 3.
together, the Plan, the Trust Agreement, Triple Canopy's
by-laws, and the IRAdid not provide a clear avenue for
relief if the trustee felt that directions from Constellis
with respect to a sale of the company were in tension with
its ERISA obligations. When asked how the ESOP might go about
stopping an action by the board of directors given its
minority position, if Wilmington felt the action was
inconsistent with ERISA, Protas admitted that the only
recourse would be to "file a lawsuit to fight about
it." Tr. at 1448:6-8. As a practical matter, she
continued, "I guess [the board] could do anything."
Developments in Early 2014
experienced a number of setbacks at the beginning of 2014,
the most significant of which related to a contract with DoS
to provide fixed-site security in Iraq. The umbrella
contract, known as the Worldwide Protective Services
("WPS") contract, was divided into task orders.
Constellis had been awarded a $300 million task order to
provide security in Basra, Iraq. DTX 2 at 17. The incumbent
protested that award, and in January 2014 the DOS rescinded
the award and requested new, best and final offers from
Triple Canopy and its competitor. Id. This change
both reduced Constellis' projected revenue from the
contract by over $100 million and pushed that revenue further
into the future. Id. Another major development in
2014 involved Constellis' sub-contract on the Kuwait Base
Operations Security Support Services ("KBOSSS")
contract, which involved providing fixed-site security at two
military bases in Kuwait. DTX 2 at 17. In February 2014, the
prime contractor, Exelis Inc., informed Constellis that it
planned to hold a competition to fill the subcontractor role
that Constellis had been expecting to fill. Id. This
development also reduced Constellis' projected revenues
by tens of millions of dollars. Id.
suffered a number of smaller problems, including the DoS
making a formal demand for payment of $62.2 million in
connection with the DCAA audit. DTX 2 at 18. Although
Constellis planned to defend against that demand, eventual
liability seemed more likely. See id Even though the ESOP had
successfully negotiated an indemnification for the full
amount of any ensuing liability, that amount would have to be
set off against the Seller notes rather than being paid to
the company in cash. See Id. Consequently, as SRR
observed in the summer of 2014, "repayment of the claim
would negatively affect [Constellis'] operating
other developments contributed to the inauspicious beginning
of 2014 for Constellis. First, Constellis lost its bid for a
$338 million contract to provide protective services in
Germany. DTX 2 at 18. The company persuaded the DoD to
rescind its award to a competitor and reopen discussions with
Constellis, but management put the odds of winning this
contract no higher than 33%, and probably lower than that.
Id. Second, the DoD informed Constellis that it
would be terminating a contract to provide security to a U.S.
Marine Corps base in Afghanistan, Camp Leatherneck, four
months earlier than anticipated. Id. This decision
reduced Constellis' backlog for 2014 by several million
bad news was breaking, in February 2014, Katis was contacted
by Jason DeYonker ("DeYonker"), the founder and
managing partner of Forte Capital Advisors
("Forte"), who was interested in buying Constellis.
Tr. at 1118:18-24. Forte owned ACADEMI, which was formerly
known as Blackwater, a major competitor of Constellis in the
private security market. Id. at 1221:10-14;
1119:16-21. DeYonker's interest in Constellis had been
piqued by the press release announcing Constellis' sale
to the ESOP. Id. at 1118:18-24.
initially balked at the idea of selling Constellis to
ACADEMI, in part because there was bad blood between the two
companies. Not only had Katis and Mann founded Constellis in
an effort to combat the bad reputation that Blackwater had
given the industry, the two companies openly feuded for
years. Tr. at 1119:16-21. Katis believed that Blackwater had
used unfair tactics in competing against Triple Canopy and
Constellis. Id. at 1119:22-25. DeYonker managed to
convince Katis that ACADEMI had changed under new leadership.
Id. at 1120:1-15. In Katis' words, DeYonker and
his partners had come in "and bought the ...