Manichaean Capital, LLC v. SourceHOV Holdings, Inc.

• Court: Court of Chancery of Delaware
• Decided: January 30, 2020
• Docket: C.A. No. 2017-0673-JRS
• Status: Published

Opinion

IN THE COURT OF CHANCERY OF THE STATE OF DELAWARE

MANICHAEAN CAPITAL, LLC, ) CHARLES CASCARILLA, EMIL KHAN ) WOODS, LGC FOUNDATION, INC., and ) IMAGO DEI FOUNDATION, INC. ) ) Petitioners, ) ) v. ) C.A. No. 2017-0673-JRS ) SOURCEHOV HOLDINGS, INC., ) ) Respondent. )

MEMORANDUM OPINION

Date Submitted: October 31, 2019 Date Decided: January 30, 2020

Rudolf Koch, Esquire and Matthew W. Murphy, Esquire of Richards, Layton & Finger, P.A., Wilmington, Delaware and Samuel J. Lieberman, Esquire and Michelle Malone, Esquire of Sadis & Goldberg LLP, New York, New York, Attorneys for Petitioners.

T. Brad Davey, Esquire, Matthew F. Davis, Esquire, Andrew H. Sauder, Esquire and Caneel Radinson-Blasucci, Esquire of Potter Anderson & Corroon LLP, Wilmington, Delaware, Attorneys for Respondent.

SLIGHTS, Vice Chancellor SourceHOV Holdings, Inc. (“SourceHOV” or the “Company”) executed a

series of transactions in 2017 that converted certain of its minority stockholders into

unitholders of a limited liability company. These transactions facilitated a three-

party business combination between SourceHOV, Novitex Holding Inc. (“Novitex”)

and Quinpario Acquisition Corp. 2 (“Quinpario”) wherein SourceHOV merged into

Quinpario and became a publicly traded company (the “Business Combination”).

Petitioners were SourceHOV minority stockholders. The Business Combination

triggered their appraisal rights under 8 Del. C. § 262, which they now seek to

exercise.

In the wake of recent guidance from our Supreme Court, this Court typically

begins its statutory appraisal function by focusing on market-based evidence of fair

value.1 In this case, however, the parties agree that market evidence is not useful

because SourceHOV was privately held and its managers made no real effort to run

a “sale process” in advance of the Business Combination. Accordingly, the parties

rely on traditional valuation methodologies, as presented by their experts, to advance

their divergent views of SourceHOV’s fair value. After completing their valuation

analyses based on several approaches, the experts agree that a discounted cash flow

1 See DFC Global Corp. v. Muirfield Value P’rs, L.P., 172 A.3d 346 (Del. 2017); Dell, Inc. v. Magnetar Global Event Driven Master Fund Ltd., 177 A.3d 1 (Del. 2017); Verition P’rs Master Fund Ltd. v. Aruba Networks, Inc., 210 A.3d 128 (Del. 2019).

1 analysis (“DCF”) is the most reliable tool to determine SourceHOV’s fair value. Of

course, they disagree on multiple crucial inputs in their DCF analyses, and these

disagreements have placed the Court in the now familiar position of grappling with

expert-generated valuation conclusions that are solar systems apart. Good times. . . .

Petitioners’ expert calculates SourceHOV’s fair value at $5,079 per share;

Respondent’s expert sets the fair value mark at $2,817 per share. While frustrating,

the fact that appraisal experts so profoundly disagree on what is, in essence, a fixed

point is no longer surprising.2 If that were as far as the disagreements went, this

appraisal dispute would not be particularly remarkable. But this case comes with a

twist. Not only does Respondent disagree with Petitioners’ expert, it disagrees with

its own expert—it has rested on a fair value for SourceHOV ($1,633 per share) that

comes in well below even its own expert’s appraisal.

The evidentiary framework for appraisal litigation, while strange, is well

settled. Both sides have the burden of proving their respective valuation positions

by a preponderance of evidence. If the parties fall short in meeting their respective

2 See, e.g., Golden Telecom, Inc. v. Global GT LP, 11 A.3d 214, 218 (Del. 2010) (“[I]t is difficult for . . . Vice Chancellors to assess wildly divergent expert opinions regarding value.”); In re Appraisal of Jarden Corp., 2019 WL 3244085, at *1 (Del. Ch. July 19, 2019) (observing that well credentialed experts were “miles apart”); Gonsalves v. Straight Arrow Publ’rs, Inc., 1996 WL 696936, at *1 (Del. Ch. Nov. 27, 1996) (“Gonsalves I”), rev’d, 701 A.2d 357 (Del. 1997) (“Gonsalves II”) (stating it is “rather a typical appraisal trial” when experts advance “absurdly differing values”).

2 burdens, then the court must sift through the evidence to perform its own appraisal.3

After carefully considering the evidence, I am satisfied that I need not undertake my

own appraisal of SourceHOV. Petitioners’ expert, with one minor exception, has

presented a credible valuation analysis from which I see no legal or evidentiary basis

to depart. In other words, I have more confidence in Petitioners’ presentation than

I have in my own ability to translate any doubts I may have about it into a more

accurate DCF valuation.

After reviewing Respondent’s fair value presentation, I am struck by the fact

that it has disagreed with its own valuation expert, relied on witnesses whose

credibility was impeached and employed a novel approach to calculate

SourceHOV’s equity beta that is not supported by the record evidence. In a word,

Respondent’s proffer of fair value is incredible.

With these factual conclusions in hand, I have determined the fair value of

SourceHOV’s stock at the time of the Business Combination was $4,591 per share.

I explain my reasons below.

3 M.G. Bancorporation, Inc. v. LeBeau, 737 A.2d 513 (Del. 1999).

3 I. FACTUAL BACKGROUND

The following facts were proven by a preponderance of the credible evidence

after a three-day trial. 4

A. Parties and Relevant Non-Parties Respondent, SourceHOV, was a Delaware corporation with its principal place

of business in Irving, Texas.5 It provided process outsourcing and financial

technology services within several industries.6

Petitioners, Manichaean Capital, LLC, Charles Cascarilla, Emil Khan Woods,

LGC Foundation, Inc. and Imago Dei Foundation, Inc. (collectively “Manichaean”)

owned 3,574, 4,418, 2,024, 205 and 83 shares of SourceHOV common stock,

respectively, as of the Business Combination. 7 Manichaean received its 10,304

shares—about 6.5% of SourceHOV’s common stock—in February 2014 when

4 The trial record consists of testimony from 15 fact witnesses, 2 expert witnesses and more than 450 trial exhibits. See Stipulated Joint Pre-Trial Order at Ex. A (“PTO”) (D.I. 81); Pet’rs’ Notice of Lodging (D.I. 86). Citations will appear as follows: “PTO __” will refer to stipulated facts in the Pretrial Order; “Tr. __ ([Name])” will refer to witness testimony from the trial transcript; “JX __” will refer to the trial exhibits; and “([Name]) Dep. (JX __ or D.I. __)” will refer to witness testimony from a deposition transcript lodged with the Court for trial. 5 PTO ¶ 7. 6 Id.; Tr. 305:23–306:2 (Chadha). 7 PTO ¶¶ 1–5.

4 SourceHOV acquired BancTec Group (“BancTec”). 8 In total, Petitioners invested

about $32 million in SourceHOV. 9

Non-party, HandsOn Global Management, LLC (“HGM”), is a family

investment business operated by non-party, Parvinder Chadha (“Chadha”), who

served as HGM’s CEO and Chief Investment Officer.10 Chadha founded HGM in

2001 “to invest [his] . . . personal capital in [order to] build [a] business process

services company that acquired technology.” 11 HGM and its affiliates held about

80% of SourceHOV’s common stock. 12 Its investment in SourceHOV traces back

to 2007, when HGM acquired SourceHOV’s predecessor entity,

HOV Services LLC. 13

8 Id.; JX 265 at 270 (providing the BancTec acquisition background); JX 274 at 35–38 (showing SourceHOV’s ownership as of June 30, 2017). 9 Tr. 450:7–11 (Cascarilla). 10 PTO ¶ 9. 11 Tr. 304:11–14 (Chadha). 12 PTO ¶¶ 10–11; Tr. 368:20–23 (Chadha). 13 Tr. 304:18–20 (Chadha); JX 265 at 270.

5 Chadha works at HGM with non-party, Jim Reynolds, a CPA who serves as

HGM’s COO. 14 Together, Chadha and Reynolds functionally comprised the

SourceHOV board of directors (the “SourceHOV Board”).15

Non-party, Apollo Global Management, LLC (“Apollo”), is a global

investment firm. 16 Apollo was a majority stockholder of Novitex, a participant in

the Business Combination. 17

Non-party, Delos Capital Management, LP (“Delos”), is a private equity and

venture capital firm. 18 Delos was an investor in SourceHOV. 19

B. SourceHOV’s Origins, Growth and Governance

SourceHOV was formed in April 2011 through the combination of two

process outsourcing service providers.20 In November 2014, SourceHOV acquired

BancTec, which expanded SourceHOV’s business automation services into the

14 Tr. 519:8–10 (Reynolds). 15 Chadha was Chairman and Reynolds Co-Chair of the SourceHOV Board. See PTO ¶¶ 10–11. 16 Id. ¶ 17. 17 Id. 18 Id. ¶ 22. 19 Id. 20 JX 265 at 246.

6 banking and payments industry. 21 Upon consummation of the BancTec transaction,

Manichaean’s BancTec investment rolled into SourceHOV and Manichaean became

SourceHOV’s second largest, non-HGM affiliated investor. 22

SourceHOV continued to grow through acquisitions. In 2016, it acquired

TransCentra, Inc. (“TransCentra”), a provider of remittance transaction processing

services.23 That transaction, like the BancTec transaction before it, left SourceHOV

in a highly leveraged state.24

By the time of the Business Combination, SourceHOV had grown into a

global business process outsourcing and financial technology company with a work

force of 16,000 employees. 25 It provided information and transaction processing

solutions to clients in three major industry segments: (i) Information and Transaction

Processing Solutions (“ITPS”); (ii) Healthcare Solutions (“HS”) and (iii) Legal and

Loss Prevention Services Solutions. 26 In 2015 and 2016, SourceHOV generated

21 Id. at 270; Tr. 305:10–13 (Chadha). 22 JX 274 at 36–38. 23 JX 265 at 270–71, 397; Tr. 115:3–5 (Verma). 24 Tr. 306 (Chadha). 25 JX 265 at 32, 260, 396. 26 Id. at 396.

7 roughly $800 million in “recurring” revenue, 27 meaning that approximately 90% of

its revenues flowed from long-term client contracts. 28

SourceHOV’s governance structure was not a model for best practices.

Its Board appears to have comprised two members—Chadha and Reynolds—both

of whom were nominated by HGM. 29 While minority investors (including

Manichaean) had information rights, SourceHOV went years without holding a

Board meeting, and Manichaean seldom received SourceHOV’s financial

statements.30

C. The 2014 BancTec Acquisition and Future Debt/EBITDA Ratio Stepdowns

In connection with the 2014 BancTec acquisition, SourceHOV raised debt in

two separate agreements: the First Lien Credit Agreement (the “First Lien”) and the

Second Lien Credit Agreement (the “Second Lien”).31 The First Lien included both

27 Id. at 392. 28 Id. at 270. 29 Tr. 14–15 (Verma) (stating that Chadha and Reynolds were the SourceHOV Board members who reviewed management’s projections); Tr. 265–66 (Chadha) (expressing confusion over whether Delos had a Board seat); PTO ¶¶ 10 (stipulating that Chadha was Chairman of the SourceHOV Board), 11 (stipulating that Jim Reynolds was Co-Chairman of the SourceHOV Board). At trial, Chadha, the nominal Chairman of the SourceHOV Board, was unable to say for sure who the other Board members were. Tr. 266 (Chadha). 30 JX 42; JX 414 at 72; Tr. 450, 452–55 (Cascarilla). 31 JX 6 (the “First Lien”); JX 7 (the “Second Lien”).

8 a $780 million term loan and $75 million in revolving credit. 32 It carried a coupon

rate of 7.75% and was due in October 2019, with required quarterly payments on

principal at .625% for the first year and 1.25% thereafter. 33 The Second Lien

included a $250 million term loan with an 11.5% coupon rate due in April 2020.34

Because they were issued at a discount to par, the weighted average yield to maturity

of the First and Second Liens at issuance was 9.5% 35 The First and Second Liens

remained outstanding until the Business Combination and constituted the “vast

majority” of SourceHOV’s outstanding debt.36

The First and Second Liens contained covenants requiring SourceHOV to

maintain a defined leverage ratio, with a numerator of total net debt and a

denominator of a specified measure of EBITDA (the “Leverage Ratio”). 37 The

Leverage Ratio was initially 6.375x with scheduled “stepdowns” every six months.38

32 First Lien at 6 (Recitals). 33 Id. § 2.11(a)(i); JX 357. 34 Second Lien at 7 (Recitals). 35 Meinhart Dep. (D.I. 87) 72:16–73:12. 36 PTO ¶ 29. 37 See Tr. 82:12–83:21 (Verma). The Leverage Ratio uses a measure of EBITDA (“adjusted EBITDA”), defined on a “Pro Forma Basis,” and includes certain adjustments. See First Lien §§ 1.01, 6.10; Second Lien §§ 1.01, 6.10. 38 First Lien § 6.10; Second Lien § 6.10.

9 Under the designated schedule, the Leverage Ratio had to be reduced to 4.75x and

then 4.25x by the end of December 2016 and June 2017, respectively. 39 Because the

Leverage Ratio depends on the relationship between debt and EBITDA, SourceHOV

would either have to increase its EBITDA, reduce its total debt or both to satisfy the

Leverage Ratio covenant.40

Under both the First and Second Liens, SourceHOV was required to certify

its compliance with the Leverage Ratio at regular intervals.41 If SourceHOV failed

to satisfy the Leverage Ratio, it would be in default, and its repayment obligations

would be accelerated, unless cured.42 The cure right allowed SourceHOV “a period

in which it [could] bring in contributions from shareholders’ equity, . . . and add that

to the EBITDA.” 43

D. The 2016 TransCentra Acquisition and the Novitex/Quinpario Letter of Intent

By September 2016, SourceHOV had acquired TransCentra. 44 Two months

later, it turned its sights to Novitex, a provider of document management services.

39 First Lien § 6.10; Second Lien § 6.10. 40 Tr. 86:9–12 (Verma). 41 First Lien § 5.04(c); Second Lien § 5.04(c); Tr. 90:9–20 (Verma). 42 First Lien §§ 7.01–7.02; Second Lien §§ 7.01–7.02. 43 Tr. 88:11–15 (Verma). 44 JX 265 at 397.

10 At first, SourceHOV considered an all-cash acquisition of Novitex and engaged

Millco Advisors, LP (“Millstein”) to assist with securing financing.45 By January 3,

2017, however, SourceHOV had decided to pursue a different structure, a merger,

with new financial advisors, Rothschild, Inc. (“Rothschild”) and Morgan Stanley &

Co., LLC (“Morgan Stanley”).46 There were two main reasons for the change in the

acquisition strategy. First, SourceHOV did not want to pursue the “significant

additional equity infusion” that would be required for an all-cash acquisition, nor did

it want to pursue a change of control transaction. 47 Second, a merger with a publicly

traded company would provide SourceHOV with greater liquidity and access to

public markets for future financing.48

Given these factors, Rothschild suggested that SourceHOV pursue the

Business Combination among SourceHOV, Novitex and Quinpario rather than an

all-cash acquisition.49 Quinpario was a NASDAQ-listed, blank-check, special

45 PTO ¶¶ 35, 36; JX 63 (Millstein engagement letter dated November 1, 2016). 46 PTO ¶¶ 25, 27, 41. 47 JX 265 at 172–73; Tr. 215 (Chadha) (confirming that Rothschild proposed a transaction through which HGM retained majority control of any SourceHOV/Novitex combined entity). 48 PTO ¶ 41. 49 Id.

11 purpose acquisition company (“SPAC”) formed to find a merger opportunity.50

The initial plan was for SourceHOV and Novitex to roll all of their equity interests,

along with at least $225 million of Quinpario’s cash, into Quinpario, which would

then be renamed Exela. 51

Investors in SPACs like Quinpario have the right to require the SPAC to

redeem their shares rather than roll their shares into a post-acquisition company. 52

This dynamic can result in last-minute re-negotiations of SPAC deals when there are

more redemptions than anticipated.53 Quinpario had $350 million in cash to deploy

for a potential transaction, but the fund was set to expire on January 22, 2017,

at which point, absent an extension approved by its investors, it would return that

cash to its investors and wind up.54 With its sunset looming, Quinpario needed to

move quickly.

On January 13, 2017, SourceHOV, Novitex and Quinpario signed a letter of

intent (the “LOI”) memorializing their initial plan for the Business Combination.55

50 Id. ¶ 20. 51 JX 117 at 3–4; PTO ¶ 19. 52 See JX 265 at 143. 53 JX 11 at 8. 54 JX 104 at 6. 55 JX 115; JX 117.

12 According to the LOI, SourceHOV, Novitex and Quinpario would all combine, and

SourceHOV stockholders would receive between 53% and 58% of Exela’s stock.56

Together, SourceHOV and Novitex stockholders would own approximately 77% of

Exela.57 Quinpario would contribute between $225 and $350 million in exchange

for about 17% of the remaining stock. 58 The LOI contemplated a total enterprise

value for Exela of approximately $3.1 billion.59

E. The First and Second Liens Cause Liquidity Pressure

During 2016 and early 2017, before SourceHOV signed the LOI, SourceHOV

faced stepdowns for the Leverage Ratio under the First and Second Liens.60 The

stepdowns posed particular challenges because SourceHOV had experienced flat

top-line revenue for at least the past two years.61 These trends required SourceHOV

to juggle liquidity issues at the same time it was negotiating the Business

Combination. 62

56 JX 117 at 5. 57 Id. 58 Id. 59 Id. at 3. 60 First Lien § 6.10. 61 JX 395 at 13, 45; JX 265 at 68; JX 292. 62 JX 10 at 1–2.

13 Historically, SourceHOV had a strong record of meeting or exceeding its

revenue projections. On April 1, 2013, SourceHOV presented a set of projections

to rating agencies predicting its income each year until 2015. 63 In 2013, 2014 and

2015, SourceHOV exceeded those projections—showing annual growth rates of

10%, 12.8% and 23.7%, respectively. 64 A July 2017 “Roadshow Presentation,”

circulated just four days before the Business Combination, disclosed a cumulative

annual growth rate of 10.1% between 2014 and March 31, 2017. 65 But that same

presentation disclosed that SourceHOV’s “core business” had been mostly

“steady.” 66 SourceHOV’s “key drivers” for the future depended on acquisitions that

would enable the Company to grow through cross-selling, synergy realization, a shift

to higher-margin products and a reduction in operating leverage.67 As the Roadshow

Presentation hinted, SourceHOV did not grow as fast as expected in 2015 and 2016.

63 JX 393 at 63. 64 Id. at 13, 45 (showing SourceHOV’s actual results for 2013); JX 265 at 68 (showing SourceHOV’s actual results for 2014 and 2015); JX 393 at 63 (showing SourceHOV’s 4/1/14 Rating Agency Presentation multi-year projections). 65 JX 275 at 35. 66 Id. 67 Id. at 24.

14 Indeed, the Roadshow Presentation acknowledged that the ITPS segment’s top-line

revenue had “remained stable” since 2014. 68

Debt rating agencies noticed these trends. Shortly after the BancTec

acquisition in 2014, Moody’s gave SourceHOV a corporate family rating of B2 and

rated the First and Second Liens at B1 and Caa1, respectively. 69 A year later,

Moody’s downgraded SourceHOV to B3 and the First and Second Liens to B2 and

Caa2, respectively, based on revenue and earnings shortfalls.70

Moody’s ratings changes reflected liquidity pressure created by the Leverage

Ratio stepdowns as well as SourceHOV’s recent revenue stagnation.71

To compound the liquidity pressure, the First and Second Liens stepped up their

required amortization payments in 2015. 72 With these trends unfolding, cash

generation became a key priority for SourceHOV. 73

68 Id. at 31 (noting “Topline Trends” for the ITPS segment: “[c]ore revenue, which has remained stable during 2014–16, is poised for growth.”). 69 JX 4. 70 JX 10. 71 Id. 72 First Lien § 2.11(a)(i); JX 205 at 10 (statement of cash flows showing SourceHOV’s principal payment obligations from 2014 through 2016). 73 Tr. 219 (Chadha).

15 On May 3, 2016, Moody’s again downgraded the First and Second Liens to

B3 and Caa3 respectively—noting SourceHOV’s weak liquidity position.74 Three

months later, Standard & Poor’s followed suit—downgrading the First Lien to

CCC+ and the Second Lien to CCC-.75

Consistent with past practice, SourceHOV looked to increase its revenue

through acquisitions. In September 2016, SourceHOV alleviated some pressure

when it acquired TransCentra. The TransCentra acquisition was a “de-leveraging”

transaction because TransCentra had a much lower leverage ratio than

SourceHOV. 76 Following the TransCentra deal, SourceHOV’s total Leverage Ratio

decreased from 5.341x to 5.235x at the end of Q3 2016. 77 But SourceHOV still had

more work to do because it needed to bring its Leverage Ratio down to 4.75x by the

end of 2016, around the same time SourceHOV negotiated the LOI. 78

Management projections showed SourceHOV would meet its Leverage Ratio

goal if it “execut[ed] exactly according to the business plan,” but there was little

74 JX 12. 75 JX 30 (noting SourceHOV’s “tight covenant cushion, upcoming maximum leverage stepdowns, and less-than-adequate liquidity[]”). 76 Tr. 93 (Verma); JX 66 at 2; JX 30 (Standard & Poor’s noted that SourceHOV could improve its covenant cushion percentage by “closing on the TransCentra acquisition.”). 77 Tr. 93 (Verma); JX 66 at 2. 78 First Lien § 6.10.

16 room for error.79 As year-end 2016 approached, SourceHOV determined that it

needed a small equity infusion to meet the required Leverage Ratio.80 In September,

four months before signing the LOI, SourceHOV sought a $23 million investment

from existing equity stockholders including HGM, Delos and Manichaean.81

The offer was for $1,600 per share, valuing SourceHOV’s equity at $231 million. 82

SourceHOV’s offer triggered matching rights for Manichaean that it could

have exercised to invest at the $1,600 per share price. 83 It was reluctant to make the

investment, however, because it had limited information about SourceHOV and no

expectation that SourceHOV management would suddenly open the information

pipeline. Manichaean acknowledged that SourceHOV had “significant upside on

the [] equity” but was concerned by the “lack of reliable transparency in terms of

general business prospects (management updates, dissemination of financials) and

governance (e.g., lack of any board meetings).”84 Indeed, Manichaean’s managing

partner, Charles Cascarilla, had been complaining since September 2016 that

79 Tr. 94 (Verma). 80 Tr. 94–96 (Verma). 81 JX 35. 82 Id. at 27. 83 Id. at 1; JX 82. 84 JX 82 at 1; Tr. 452 (Cascarilla) (Manichaean never received unaudited financial statements.).

17 SourceHOV had not held a board meeting in the two years since Manichaean became

a minority owner. 85 Cascarilla was so frustrated, in fact, that he considered selling

his interest in SourceHOV to Delos without having any real sense of what that

interest was worth.86

Ultimately, Delos and HGM invested $9 million around January 2017.87

SourceHOV used $6 million of the new equity to “cure” its shortfall and meet the

required Leverage Ratio stepdown under the First and Second Liens.88

Manichaean initially declined to participate in the equity raise. But after

Manichaean had the chance to “see presentations . . . about how [SourceHOV] was

performing” and “assess whether or not [to] participat[e] on a pro rata basis, so

[Manichaean] [would not be] diluted down,” Manichaean agreed to invest an

additional $1.5 million on February 8, 2017. 89

85 JX 42 (Cascarilla expressing his frustration that “it’s hard to think of a company of this size ($1.3bn EV) with such poor governance and communication” and “[w]e keep waiting to be treated as partners, but that is not happening”); Tr. 447 (Cascarilla). 86 JX 79; JX 82; JX 83. 87 JX 155; JX 120; Tr. 522–24 (Reynolds). Delos invested $5 million on January 20, 2017 and HGM made a separate $4 million investment in late December 2016. JX 155; JX 120; Reynolds Dep. (D.I. 87) 36:7–15; Chadha Dep. (D.I. 86) 285 (tab 2). 88 JX 205 at 59. 89 Cascarilla Dep. (D.I. 86) 82, 84–85; JX 155; Tr. 497–98 (Cascarilla). Testimony from Cascarilla credibly explains that Manichaean’s frustration with the lack of information made Manichaean reluctant to invest more into SourceHOV. Cascarilla Dep. (D.I. 86) 82– 85; Tr. 500 (Cascarilla). That changed when “other firms that had access to information” 18 The combination of the TransCentra acquisition and the equity investment

alleviated enough liquidity pressure to allow SourceHOV to focus on the Business

Combination. SourceHOV certified in a March 14, 2017 Going Concern

Memorandum that it did not “anticipate any defaults” in 2017.90 The debt rating

agencies apparently agreed that a default was unlikely, as they did not downgrade

SourceHOV’s debt between September 2016 and the Business Combination. 91

F. The Business Combination Agreement As noted, Quinpario stockholders were entitled to exercise redemption rights

and withdraw their investments before the Business Combination closed, and the

fund was set to wind up on January 22, 2017.92 With negotiations relating to the

Business Combination in full swing, Quinpario’s stockholders approved an

extension of the wind up to July 24, 2017. 93 While the extension allowed more time

for negotiations, it also allowed more time for redemptions. By the time the parties

were prepared to consummate the Business Combination, more than 14 million

shares of Quinpario stock had been redeemed, leaving just $200 million in

(i.e., Quinpario) announced that SourceHOV’s equity was worth $806 million. Tr. 500 (Cascarilla). 90 JX 191 at 15. 91 Tr. 223–24 (Chadha). 92 JX 265 at 143; JX 104 at 6. 93 JX 265 at 176.

19 Quinpario’s trust account. 94 And Quinpario stockholders would have another

opportunity to demand redemption in connection with the ultimate vote to approve

the Business Combination. 95 Thus, it became clear to all that Quinpario would bring

less cash to the table than originally anticipated in the LOI.

On February 21, 2017, SourceHOV, Novitex, Quinpario and other entities

executed the Business Combination Agreement (the “BCA”). 96 Under the BCA, and

as contemplated by the LOI, both SourceHOV and Novitex would merge into

separate wholly-owned Quinpario subsidiaries. 97 Quinpario would then assume the

name Exela. 98 SourceHOV and Novitex stockholders would roll over all of their

equity into Exela. 99 Quinpario and other private (“PIPE”) 100 investors would

contribute $200 million and ˜$75 million, respectively, for their shares. 101 Crucially,

94 Id. 95 Id. at 143. 96 JX 169. 97 JX 265 at 2–3, 133; JX 173. 98 JX 173; PTO ¶ 19. 99 JX 265 at 2–3, 133. 100 PIPE is an acronym that stands for “private investment in public equity.” Tr. 357 (Chadha). 101 JX 215 at 61. Under the original BCA, closing was conditioned on Quinpario providing at least $275 million in cash, which could consist of funds remaining in its trust account following redemptions coupled with proceeds from the PIPE investment. See JX 169 §§ 6.15, 8.1(g), 8.3(c); JX 265 at 2–3.

20 the BCA’s closing was conditioned on Quinpario and the PIPE investors

contributing at least $275 million in total (the “Cash Condition”). 102 On top of these

equity investments, the BCA contemplated raising $1.35 billion in new debt.103

G. The Revised BCA

SourceHOV began working on the PIPE financing before signing the BCA.104

To help satisfy the Cash Condition, (i) financial advisors working on the Business

Combination agreed to invest their fees in the PIPE investment and (ii) SourceHOV

obtained additional debt financing (the “Margin Loan”) to generate $57.5 million of

proceeds, which were also put towards the PIPE investment.105

To secure the Margin Loan, it was agreed that the Company’s former-

stockholders’ merger consideration (i.e., Exela stock) would be held by a new entity,

Ex-Sigma LLC (“Ex-Sigma”). 106 The Margin Loan required SourceHOV to merge

into a wholly-owned Ex-Sigma subsidiary immediately before the Business

102 JX 265 at 2–3. 103 Id. at 26, 146. 104 JX 132. 105 JX 265 at 3. 106 Id. Ex-Sigma agreed to purchase up to $57.5 million of the total $275 million private placement of Exela’s common and Series A Perpetual Convertible Preferred Stock sold in the PIPE investment. Id. at 188.

21 Combination. 107 Each share of SourceHOV stock would then convert into Ex-Sigma

membership units (the “Ex-Sigma Merger”). 108 During the negotiation of the Ex-

Sigma Merger, Chadha and Reynolds acted on SourceHOV’s behalf without an

independent committee of SourceHOV directors.109

After the Ex-Sigma Merger, Reynolds and Chadha became Ex-Sigma’s sole

managers.110 This dynamic put SourceHOV’s former-minority stockholders in a

particularly illiquid position. The terms of the Margin Loan require Ex-Sigma to

hold its Exela stock as security until the Margin Loan is repaid.111 And Ex-Sigma’s

LLC agreement gives Reynolds and Chadha full discretion to decide when, and

whether, to repay the Margin Loan.112 Taken together, Chadha and Reynolds

maintained exclusive voting control over all the Exela stock SourceHOV’s former

107 Id. at 3. 108 Id. 109 Tr. 182–84 (Chadha). SourceHOV made no real effort to run a sale process. Its Board did not hold a single meeting to consider either the Ex-Sigma Merger or the Business Combination more generally. JX 316, Resp. No. 11. One of SourceHOV’s financial advisors, Morgan Stanley, operated under a conflict of interest because it had served as financial advisor on prior transactions at HGM’s behest, receiving $40 million in fees. JX 316, Resp. No. 13. Morgan Stanley also invested in the Business Combination. JX 236 at 13 (noting “MS” had acquired 642,860 Exela shares with two other banks). 110 Tr. 205–06 (Chadha). 111 Tr. 208 (Chadha). 112 JX 99 at 34.

22 stockholders received in the Business Combination. In light of the new structure

required to facilitate the Margin Loan, the parties to the Business Combination

revised the BCA (the “RBCA”) to (i) make Ex-Sigma the recipient of the former-

SourceHOV stockholders’ consideration, (ii) have Quinpario provide less equity and

(iii) increase the PIPE financing.113

The Ex-Sigma Merger and the Business Combination closed in July 2017.114

The stockholders, directors and managing members of Novitex, SourceHOV, Ex-

Sigma and Exela passed written consents approving the RBCA and various

financing transactions for the Business Combination. 115 Ultimately, Exela stock was

distributed as follows: (i) Ex-Sigma 54.9%, (ii) Apollo 20.9%, (iii) Quinpario’s

stockholders 8.3%, (iv) PIPE investors 14.2% and (v) financial advisors 1.7%.116

Based on Exela’s closing stock price of $8.61 per share on July 12, 2017, the market

value of the consideration provided to Ex-Sigma implies an aggregate equity value

for SourceHOV of $694 million, or $4,177.10 per share. 117

113 JX 236 at 8; JX 265 at 2–4, 61; see id. at 58 (containing a helpful illustration of the Business Combination’s structure). 114 JX 265 at 1–4; JX 287. 115 Id. 116 JX 236 at 6. 117 80,600,000 Exela shares x $8.61 per share. This implies a $4,177.10 per share price for SourceHOV’s stock ($694 million ÷ 166,136 SourceHOV shares).

23 H. SourceHOV Revenue Projections

Because SourceHOV’s growth strategy depended on buying companies, its

management regularly made financial projections to facilitate acquisitions.118

As discussed below, three sets of projections are particularly relevant to this dispute:

the Equity Case, the Lender Model and the Bank Case.119 Each set is depicted in the

chart below:

Transaction Projection Date [1] 2013 2014 2015 2016 2017 2018 2019 2020 2021 2013 Rating Agency N/A Presentation [3] $ 555 $ 575 $ 596 $ 618 $ 693 $ 661 BancTec Sep. 26, 2014 [4] $ 872 $ 976 $ 1,012 $ 1,052 $ 1,087 TransCentra Aug. 20, 2016 [5] $ 930 $ 994 $ 1,062 $ 1,135 $ 1,213 $ 1,296 Nov. 2016 [6] $ 913 $ 960 $ 1,007 $ 1,057 $ 1,110 $ 1,166 Jan. 23, 2016 "Equity Case" [7] $ 927 $ 974 $ 1,022 $ 1,074 $ 1,127 Dec. 5, 2016 Novitex "Bank Case" [8] $ 918 $ 937 $ 957 $ 978 $ 999 Mar. 20, 2017 [9] $ 917 $ 960 $ 1,008 $ 1,059 $ 1,112 Mar. 21 - June 2017 "Lender Model" [10] $ 911 $ 960 $ 1,008 $ 1,059 $ 1,112 $ 427 SourceHOV Actual Results [2] $ 577 $ 651 $ 805 $ 789 (.5 year) [1] Values shown in millions. [2] Actual Results taken from JX 395 at 13, 45 for 2013; JX 265 at 68 for 2014-2016; JX 292 for the first half of 2017. [3] JX 393 at 63. [4] JX 3 at 13. [5] JX 32E at "Consolidated SourceHOV" tab - Total Revenue line. [6] JX 62 (Morgan Stanley slide deck prepared as "Discussion Materials" from the "Base Case"). [7] JX 136E (at "Bank" tab); JX 158E (with Equity Case selected at "Case Selection" tab). [8] JX 158E ("Bank" tab with Bank Case selected at "Case Selection" tab). [9] JX 192 at 3. [10] JX 227 at 3.

118 Tr. 107 (Verma); see, e.g., JX 393 (2013 Ratings Agency Presentation); JX 158 at 17– 18 (Feb. 2017 “Equity” Case); JX 227 at 3 (the “Lender Model”); JX 3 at 13 (BancTec acquisition projections). 119 JX 158E (the Equity Case and the Bank Case are found on the “Bank” tab by toggling between Assumptions 1 and 2 on the “Case Selection” tab); JX 227 at 3 (the Lender Model).

24 SourceHOV primarily used the Equity Case and its derivative, the Lender

Model, in its financial analyses and reporting. 120 Both models assumed revenue

growth for SourceHOV at 5% per year. 121 SourceHOV management developed and

“stood behind” the Equity Case, which it created along with SourceHOV’s Board,

sales team, operations team, investors and financial advisors through an “iterative

process.”122 The Lender Model reflected a minor “haircut” to improve the accuracy

of the Equity Case based on iterative feedback from SourceHOV’s bankers.123

SourceHOV used either the 5% Equity Case or the 5% Lender Model for

making investor presentations, interacting with its financial advisors, making lender

pitches, reporting to credit rating agencies, making public filings and working with

its accountants. 124 Indeed, at least 10 SourceHOV presentations relied exclusively

on 5% growth projections.125 In accounting memoranda, SourceHOV described

120 Tr. 63, 66, 72–74, 126 (Verma). 121 (Year 2 – Year 1) / Year 1. 122 Tr. 14–16 (Verma); Verma Dep. (JX 338) 34. 123 Verma Dep. (JX 338) 34–35. 124 JX 102 at 17 (SourceHOV “Management Presentation” from January 2017); JX 100 at 4; JX 101 at 4 (presentations by Morgan Stanley and Rothschild); JX 136 at 8, 30 (presentations for lenders and ratings agencies); JX 394 (same); Tr. 58–65 (Verma) (5% models were used for presentations to potential lenders and auditors); JX 377E (same); JX 302 at 3 (discussing Rothschild’s analysis); JX 173 at 37 (SEC filings incorporating the Equity Case). 125 See Pet’rs’ Post-Trial Answering Br. (“PPTAB”) (D.I. 101) at 12–13 (citing JX 102 at 17; JX 100 at 4; JX 101 at 4; JX 136 (data room for lenders); JX 173 at 37 (public S.E.C. 25 these 5% models as “conservative” and sometimes labeled them as the “base

model.” 126

Unlike the Equity Case and the Lender Model, the Bank Case projected

approximately 2% growth for SourceHOV after 2018. 127 SourceHOV seldom used

the Bank Case and did not update the model after it was created. 128

I. “Contemporaneous” SourceHOV Valuations

As a part of its assignment, Rothschild was asked in February 2017 to value

SourceHOV’s equity in a “fairness or unfair opinion.”129 Not surprisingly,

Rothschild selected the oft-used 5% Equity Case revenue projections as the

foundation for its work and calculated a 12% cost of capital “based on comparable

companies.”130 It then incorporated these assumptions into a DCF analysis that

filings); JX 302 at 11 (Rothschild analysis); JX 191E; JX 191 at 11; JX 394E (credit rating agency presentation); JX 229 at 3; JX 234 at 3). 126 Tr. 69–70, 126 (Verma); JX 377 at 5; JX 191 at 10, 13–14, 18–19, 28, 31. 127 (Year 2 – Year 1) / Year 1. 128 Cf. JX 192 (discussing Lender Model update on March 20, 2017); Verma Dep. (JX 338) 34 (same). 129 JX 302 at 3. Rothschild’s February Valuation was not a formal fairness opinion, although SourceHOV management relied on it when assessing the fairness of the Business Combination. JX 265 at 5 (no formal fairness opinion); JX 302 at 3 (Rothschild’s February Valuation was used to help SourceHOV management “decide[] on fairness or unfair[ness].”). 130 JX 302 at 17.

26 yielded a “Standalone SourceHOV” enterprise valuation of “$2.035” billion, and an

“equity value” of “$931” million (the “February Valuation”).131 The February

Valuation was the “last” valuation that Rothschild presented to the SourceHOV

Board before the Business Combination. 132 But that is not what Chadha wanted the

outside world to believe.

Almost four months after this litigation began, Chadha asked his son-in-law,

Andrej Jonovic (who also works at HGM), to request a “revised” valuation from

Rothschild.133 In January 2018, Rothschild responded with a so-called

“retrospective valuation update as of July 2017 . . . reflecting the final transaction

structure and updated assumptions at that time” (the “Backdated Valuation”).134

The Backdated Valuation used lower revenue growth projections (i.e., 2.4%–3.5%

per year) and calculated SourceHOV’s equity value at $675 million. 135 A few days

after reviewing the Backdated Valuation, Jonovic responded to Rothschild, “the

131 Id. 132 Rothschild Dep. (JX 322) 174. 133 JX 301 at 1. 134 JX 309 at 2. 135 See id. at 10 (showing a “Discounted Cash Flow” “Total equity value” of between $451 and $994 million); 15 (showing an “Implied equity value” of $675 million); but see JX 302 at 12 (the February Valuation calculated a “Total equity value” based on a “Discounted Cash Flow” analysis of between $680 million and $1.29 billion), 17 (showing an “Implied equity value” of $931 million).

27 cover page says Jan 2018 . . . Happy for it to simply say July 2017.” 136 After

Rothschild agreed to change the date, Jonovic forwarded the Backdated Valuation

to Chadha by email. The transmittal contained one word: “Done.” 137

Ultimately, the Proxy Statement for the Business Combination disclosed

SourceHOV’s existing equity value was $645 million based on a $644,800,000 value

for the Exela shares paid to Ex-Sigma. 138 The Proxy Statement arrived at this

valuation after applying a 25% “IPO discount.”139 The term “IPO discount”

apparently was meant to convey that the parties valued SourceHOV’s stock

differently depending on whether one considered the Exela transaction on a “fully

distributed” or a “pre-listing” basis.140 The “fully distributed” value represented “the

valuation [] [at which Exela] would trade [] at some point, once it’s fully distributed

into the market.”141 In contrast, the “pre-listing” value was “the price that investors

136 JX 309 at 1. 137 Id. 138 JX 265 at 78, 101, 126. 139 Surjadinata Dep. (JX 322) 126. 140 Id. 141 Id.

28 would receive in a regular IPO.” 142 The Proxy Statement valued SourceHOV’s

equity at $645 million based on the lower, pre-listing value of Exela’s stock.143

J. Procedural Posture

Manichaean filed its petition for appraisal under Section 262 of the Delaware

General Corporation Law on September 21, 2017. 144 It seeks appraisal for

10,304 shares of SourceHOV’s common stock that were converted into Ex-Sigma

membership units in the Ex-Sigma Merger. 145 Manichaean and SourceHOV both

presented expert witnesses at trial in support of their proffered views of the fair value

of SourceHOV stock. I summarize these opinions below. Before doing so, however,

I discuss certain discovery-related events that have influenced my assessment of

witness credibility.

142 Id.; JX 297 at 8 (showing a “Final Transaction Consideration” for “Standalone SourceHOV” on a “Long-Term FD” basis of $806 million but a value of $645 million on a “pre-listing” basis); JX 309 (same). 143 JX 265 at 78. 144 Verified Pet. for Appraisal of Stock (D.I. 1). 145 Id.

29 1. Manichaean Discovers the Backdated Valuation

Respondent produced the Backdated Valuation with a date stating it was

created during “July 2017,” instead of 2018 when it was actually created.146 Later,

Manichaean e-mailed Respondent’s counsel requesting “other information

underlying the analysis of the [Backdated Valuation].” 147 While Respondent

produced some responsive information, it did not produce the e-mails between

Jonovic and Rothschild discussing the Backdated Valuation. Instead, Respondent

claimed in a sworn Interrogatory Response that “Rothschild made [a] presentation[]

concerning the Merger . . . in . . . July of 2017 [during] meetings” with

SourceHOV. 148

As discovery wore on, Manichaean learned that the Backdated Valuation had

actually been created in January 2018.149 Manichaean demanded production of the

January 2018 e-mails surrounding the Backdated Valuation, 150 and Respondent

finally produced the e-mails in November 2018.151 Yet, when Manichaean deposed

146 JX 297 at 1, 3. Current counsel for Respondent was not yet involved in the case at this stage of the discovery. 147 JX 313 at 1. 148 JX 316, Resp. No. 5 at 6–7. 149 Rothschild Dep. (JX 322) 147–49. 150 Id. at 152:8–19. 151 JX 331.

30 Jonovic and Chadha to address these developments, they still maintained that the

Backdated Valuation was presented to SourceHOV before the Business

Combination closed in 2017 and that Respondent’s interrogatory responses stating

as much were correct. 152 It was not until the eve of trial when Respondent finally

amended its Interrogatory Response to admit there was no Rothschild “July 2017”

presentation. 153

2. Manichaean’s Expert

Before trial, Manichaean retained Timothy J. Meinhart to appraise the fair

value of SourceHOV as a standalone entity immediately before the Business

Combination. 154 He ultimately concluded SourceHOV’s equity was worth

$798.711 million or $5,079 per share.155

152 Chadha testified that Rothschild made a presentation in July 2017 related to “the proxy and the road show with the bankers” and that SourceHOV’s Interrogatory Response “sounds accurate.” Chadha Dep. (JX 359) 119:14–121:11. Jonovic testified that SourceHOV’s Interrogatory Response “looks correct, yes” and claimed that “Rothschild could have made a presentation over, you know, video conference” and “provided it to us in July.” Jonovic Dep. (JX 337) 143:2–145:17. 153 JX 361, Resp. No. 5 at 5. 154 JX 350a (the “Meinhart Op.”) at 4. Meinhart holds a BS in finance from Northern Illinois University, an MBA degree from the DePaul University Graduate School of Business and is an accredited senior appraiser of the American Society of Appraisers, accredited specifically in business valuation. Meinhart Op. at 5. 155 Id. at 46.

31 In preparing his valuation, Meinhart employed three valuation methodologies:

(i) DCF, (ii) Capital Cash Flow (“CCF”) and (iii) Guideline Publicly Traded

Company (“GPTC”). 156 These approaches yielded enterprise values for SourceHOV

of $1.788 billion, $1.831 billion and $2.074 billion respectively. 157 While Meinhart

considered all three, he ultimately based his conclusions only on the DCF and CCF

methods—both of which are “income approaches.” 158

DCF posits that the value of a business is the present value of the future

income that will be received by the owners of a business.159 It uses a weighted

average cost of capital (“WACC”) to discount the future cash flows a company’s

owners expect to receive. 160 A CCF is a variation of DCF that is better suited to

156 Id. at 17–18. 157 Id. at 36–40, 46. 158 Id. at 18. Income approaches seek to convert a company’s expected future cash flows into a single “present value.” Id. Meinhart rejected the guideline merged and acquired company method because he could not identify any transactions involving companies sufficiently similar to SourceHOV where the transaction closed within a reasonable period before the Business Combination. Id. He rejected the asset accumulation method because a discrete valuation of each of the SourceHOV assets was beyond the scope of his engagement. Id. 159 Id. at 17. 160 Id.

32 value future cash flows where a company’s capital structure is expected to change.161

Ultimately, a traditional DCF and CCF are “algebraically equivalent.”162

According to Meinhart, the first step in applying either the DCF or the CCF

models is to project SourceHOV’s future cash flows. In this regard, he placed

“primary reliance” on the Lender Model because it was (i) frequently “updated” and

“circulated” before the Business Combination, (ii) “vetted” by the participants of the

Business Combination and their advisors and (iii) the “most conservative” of the

updated projections.163

Meinhart then made two adjustments to the Lender Model. First, he adjusted

projected cash flows to include the continued amortization of goodwill. 164 Second,

he accounted for management fees, board fees and expenses, and non-cash equity

compensation expenses.165 Both adjustments led to lower cash flows than

SourceHOV management originally projected. 166

161 Id. (citing JX 422). 162 JX 422 at 1–2. 163 Meinhart Op. at 20. 164 Id. at 20–21 (explaining that SourceHOV was amortizing its goodwill over a period of 10 years beginning in 2014, but changed its treatment of goodwill in 2016 in anticipation of the Business Combination, and further explaining that a reversion to a private company would probably cause a reversion to amortization of goodwill). 165 Id. at 21. 166 Id.

33 For the DCF model, Meinhart (i) projected the future cash flows to holders of

SourceHOV’s debt and equity using the Lender Model and (ii) applied a present

value discount rate (or WACC) to those cash flows.167 He began by converting the

Lender Model into a cash flow projection that incorporated taxes, depreciation and

amortization expenses, capital expenditures and changes to net working capital.168

Next, he applied a WACC discount rate of 11.2% to SourceHOV’s future cash flows

to arrive at a net present value based on SourceHOV’s cost of debt and equity

capital.169

To calculate industry beta, which is one of the key inputs in a WACC

calculation, Meinhart selected 19 publicly traded guideline companies and

167 Id. at 22, 27. 168 Id. at 22. 169 Id. at 23, 27. To arrive at a single WACC, Meinhart calculated a cost of equity capital and a cost of debt capital separately. Id. at 26–27. He calculated the WACC discount rate based on the capital asset pricing model (“CAPM”)—the general formula for which is: Ke = Rf + [β * ERP] + SRP. Ke represents the cost of equity capital. Rf represents the risk- free rate of return (which is based on U.S. Government debt). Id. β represents the industry beta. Industry beta is a measure of the systematic risk of a stock. JX 420 at 8–9. It shows the tendency of a specific stock’s price to correlate with changes in the broader market. Meinhart Op. at 22. ERP represents the equity risk premium. This is the extra return that investors demand to compensate them for investing in common stocks rather than investing in risk-free securities. SRP is the size-related equity risk premium. The size premium represents the empirical observation that companies of smaller size are linked to greater risk and thus have greater cost of capital. JX 426. It is needed to adjust for the size differential between a specific company and the empirical data from which the equity risk premium is derived. Meinhart Op. at 23. Meinhart applied an equity risk premium of 5.97% and a size premium of 2.08% based on the size premium table from the Duff & Phelps 2017 Valuation Handbook. Id. (citing JX 426).

34 calculated their “raw, levered betas.”170 He based his selection of comparable

companies on (i) SourceHOV’s public filings, (ii) those identified by Rothschild in

the February Valuation and (iii) his own independent research.171 Next, he

“unlever[ed]” each specific beta to focus only on industry risk instead of the risk

created “by the [guideline] company’s particular capital structure.”172 To be

conservative, and to account for SourceHOV’s high debt load, Meinhart “selected

the highest unlevered equity betas of the guideline company group of 1.203 and

1.210.” 173 He then further increased the beta in a re-levering process to account for

SourceHOV’s projected capital structure.174

Another key input for a WACC calculation is a company’s size premium. The

size premium accounts for the additional risk of investing in a smaller company

compared with the broader index of companies represented in a market index.175

170 Meinhart Op. at 25. 171 Id. at 36–37. 172 Id. at 25. For the de-levering calculation, Meinhart used the “Hamada formula, the Harris-Pringle formula, and the Fernandez formula.” Id. He ultimately chose the Hamada formula, even though the other formulas tended to produce lower betas, because Hamada is “widely accepted,” “used by many analysts” and more commonly used “than any other re-levering model.” Tr. at 633–34 (Meinhart). 173 Meinhart Op. at 26. 174 Id. 175 Id. at 23; JX 340 at 67.

35 As noted, Meinhart applied a size premium of 2.08% based on statistical analysis

from the 2017 Duff & Phelps Valuation Handbook for companies with market

capitalization between $569.279 million to $1.030 billion.176

For his alternative CCF model,177 Meinhart (i) adjusted the Lender Model

into a cash flow projection and (ii) considered a present value discount rate that is

easier to calculate than WACC—the unlevered cost of equity capital (“UCEC”).178

To calculate the appropriate UCEC, Meinhart used the same unlevered guideline

company beta of 1.21 that he used in the WACC calculation for the traditional DCF

analysis. His calculations led to a 12.4% discount rate for the CCF analysis.179

One component of Meinhart’s CCF analysis was to project SourceHOV’s

future interest expenses and their impact on future cash flows. He estimated the

present value of SourceHOV’s income tax shield based on an assumption that the

176 Meinhart Op. at 23 (citing JX 426). 177 Meinhart explained he applied the CCF model to address concerns about whether SourceHOV would be able to (i) exploit its interest expense deductions in the years in which expenses were incurred and (ii) maintain a constant capital structure where the percentages of debt and equity capital are essentially unchanged over time. Id. at 29. 178 Id. at 31. 179 Id.

36 Company would pay down debt at a weighted average interest rate of 9%. 180 This

is in accord with the management projections SourceHOV provided to its auditors.181

Meinhart ultimately weighted his DCF and CCF valuations equally—yielding

a SourceHOV enterprise value of $1.810 billion.182 In reaching this conclusion,

Meinhart chose not to rely on the Equity Case or his GPTC valuation.183

He preferred the Lender Model because it was “more updated”; and he rejected the

GPTC valuation because there was “not a perfect guideline company for

SourceHOV.” 184

As a final step, Meinhart tweaked his valuation to account for

(i) SourceHOV’s cash, (ii) net operating loss carryforwards (“NOLs”) and

180 Id. at 72; Tr. 650–52, 571–77 (Meinhart). 181 See JX 191E (“Debt Schedule” Tab, line 58, columns G–Z). 182 Meinhart Op. at 41, 46. 183 See id. at 41 (noting that reliance on either the Equity Case or the GPTC valuation would have yielded higher values for SourceHOV’s equity). 184 Id. Meinhart also considered other “indications of value.” He noted that his valuation of SourceHOV’s equity at $798,711 was within the range Rothschild found in the February Valuation of $527 million to $993 million. Id. at 42–43. He also considered the transaction price at which Delos, HGM and Manichaean purchased additional SourceHOV stock in early 2017 (implying an equity value of $231 million). See JX 35 at 27. Meinhart chose not to rely on these transactions because of the “conflicted nature” of the parties to the transactions and “the fact that the transaction price was well below the range of value established by SourceHOV’s own financial advisor in the subsequent work it performed.” Meinhart Op. at 43.

37 (iii) various sources of debt.185 When all was said and done, Meinhart concluded

that the fair value of SourceHOV as of the Business Combination was $798 million,

or $5,079 per share.186

3. Respondent’s Expert

Respondent’s valuation expert was Gregg Jarrell. 187 In his report, Jarrell

opined that SourceHOV’s equity value was $286.4 million (or $1,723 per share).188

He amended that view during his deposition after making certain changes that drove

his valuation 63% higher than his original assessment, to $468.1 million

(or $2,817 per share). 189

Like Meinhart, Jarrell relied on a type of DCF analysis to value

SourceHOV. 190 Specifically, he relied on an adjusted present value (“APV”)-based

185 Id. at 41–42. 186 Id. at 42. 187 JX 340 (“Jarrell Op.”) at 4. Jarrell holds both a Ph.D. in Business Economics and a MBA from the University of Chicago. Jarrell Op. at 5. He has experience as an economics professor, an economics consultant and the Chief Economist for the Securities and Exchange Commission. Jarrell Op. at 5. 188 Id. at 4. 189 Jarrell Dep. (JX 356) 15; JX 353 Ex. 8 (updated). Jarrell’s adjustments were prompted by Meinhart’s suggestions regarding SourceHOV’s projected depreciation and amortization as well as its NOL projections. Jarrell Dep. (JX 356) 19, 21–22. 190 Jarrell Op. at 39.

38 DCF model to value the projected cash flows associated with SourceHOV equity.191

Jarrell’s APV model is functionally the same as Meinhart’s CCF model. 192 Both the

APV and CCF models seek to simplify the valuation exercise for a company with a

changing capital structure. 193

Jarrell began by selecting the Equity Case as the foundation for his APV

valuation. 194 He made this selection because he wanted to be as “conservative as the

expert for Respondent.” 195 Jarrell noted his “serious reservations” about the Equity

Case’s reasonableness based, in part, on SourceHOV management’s “aggressive”

accounting practices. 196 On the other hand, he observed that Rothschild used the

Equity Case projections when performing its February Valuation, which, in his

mind, increases their reliability. 197

191 See id. at 37–38 (explaining that a traditional WACC-based DCF handles income tax savings from tax deductible interest payments by incorporating those savings directly into the WACC, while an APV model uses an unlevered cost of equity and separately values the present value of interest tax shields). 192 Tr. 802 (Jarrell) (explaining that an APV model is “mathematically virtually identical” to a CCF model); Tr. 571 (Meinhart). 193 Tr. 577 (Meinhart); Jarrell Op. at 41–42. 194 Jarrell Op. at 53. 195 Id. 196 Id. at 50, 60. 197 Id. at 53.

39 After applying certain adjustments to the Equity Case, he then turned to the

Modigliani and Miller theorem (the “M&M Theorem”) as the foundation for his

discount rate. 198 That theorem posits “that the risk (beta) of the firm’s debt must

always be less than the risk (beta) of the firm’s equity.” 199 He adopted this concept

as the “methodological basis for how [he] estimate[d] the unlevered cost of equity

for SourceHOV.” 200 Jarrell explained:

I use the available evidence to determine the minimum reasonable cost of debt of a standalone SourceHOV as of the valuation date, which then yields an implied minimum reasonable debt beta based on this minimum reasonable cost of debt. I then conservatively use this implied debt beta as a minimum possible estimate of the overall beta of SourceHOV’s assets (also called the unlevered equity beta). Because I use the APV approach, instead of the WACC approach, all I need to calculate the appropriate unlevered equity discount rate is the unlevered equity beta, which in theory cannot be less than the beta of the firm’s debt as explained above. 201

Jarrell took this approach for the same reasons that Meinhart rejected the

GPTC approach.202 SourceHOV was a private company, so the appraiser cannot

198 Id. at 64–65; Tr. 755–57 (Jarrell) (discussing his adjustments to the Equity Case for stock-based compensation, depreciation and amortization, taxes and capital expenditures). 199 Jarrell Op. at 65. 200 Id. at 66. 201 Id. 202 Id.

40 measure equity beta directly. 203 Unlike Meinhart, however, Jarrell took the dearth

of comparable companies a step further and concluded he could not use “indirect or

regression-based betas . . . to estimate SourceHOV’s unlevered equity beta.”204

Instead, Jarrell estimated that SourceHOV’s unlevered cost of equity was 13.69%

by starting with the “market-based yields for [SourceHOV’s] traded debt” and

plugging that value into the capital asset pricing model.205

Beta is one of the key inputs for a CAPM analysis. In this regard, Jarrell

noted, “industry or peer group averages are commonly used when the beta of a

company . . . cannot be determined.”206 Even so, based on the M&M Theorem,

Jarrell chose to estimate SourceHOV’s equity beta directly by calculating its debt

beta. 207

203 Id. 204 Id. 205 Id. at 67–68 (explaining CAPM’s basic formula: the cost of equity = (i) the risk-free rate plus (ii) a firm’s beta multiplied by the equity risk premium plus (iii) the equity size premium). This is the same formula Meinhart used. See Meinhart Op. at 23. 206 Id. at 69 (internal citation and quotation omitted). 207 Id. at 66–70 (noting that the comparable companies Rothschild identified for the February Valuation were “of considerably greater size than SourceHOV” and that SourceHOV had “high financial leverage”). With this in mind, Jarrell looked at SourceHOV’s actual debt (primarily, the First and Second Liens) to calculate debt beta. Id. at 71. In this regard, Jarrell reviewed all of SourceHOV’s debt and determined that its average cost of debt was 11%. Id. He then determined the debt beta implied by an 11% cost of debt. Id. at 71–72.

41 Two other key components of Jarrell’s analysis relate to SourceHOV’s size

premium and future interest expenses. To calculate SourceHOV’s size premium,

Jarrell used the same sources as Meinhart.208 But Jarrell concluded that SourceHOV

had a smaller market capitalization. Therefore, he increased SourceHOV’s size

premium from 2.08% (corresponding to a $569 million–$1.03 billion market

capitalization) to 2.68% (corresponding to a $264 million–$568 million market

capitalization). 209 He based this decision on Exela’s trading prices after the Business

Combination closed.210

To calculate the present value associated with SourceHOV’s future interest

expense tax deductions, Jarrell concluded that SourceHOV would pay off all of its

sizable debt before 2020.211 This assumption decreases the present value of the tax

deductions.212 Jarrell based this decision on his opinion that “the repayment of debts

would likely be financed through new equity investments.” 213

208 Id. at 74 (citing JX 426). 209 Id. at 75–76. 210 Id. 211 Id. at Ex. 7. 212 Id. 213 Id. at 79.

42 After calculating the present value of SourceHOV’s future cash flows, Jarrell

made adjustments for NOLs, amortization, interest tax shields and debt.214

Ultimately, he determined SourceHOV’s total equity value was $468.1 million or

$2,817 per share.215 He reached this final determination, laudably, after

incorporating input from Meinhart and adopting portions of Meinhart’s expert

opinion that he found persuasive. 216 Even with the adjustments, however, the

experts’ fair value determinations miss each other by ˜44%.

II. ANALYSIS

After considering all relevant factors, I have determined the fair value of

SourceHOV as of the Business Combination was $4,591 per share. I reach this

conclusion in four steps. First, I review the legal standards and burdens of proof

applicable in a statutory appraisal proceeding. Second, I summarize the unique

factors that have led me to conclude that DCF is the only reliable method to reach

SourceHOV’s fair value. Third, given the wide divergence between the parties’

litigation positions, I assess the credibility of each expert’s analysis at the macro

level to determine the extent to which their opinion is dispositive, or informative,

214 Id. at 97. 215 Jarrell Dep. (JX 356) 15; JX 353 Ex. 8 (updated). 216 Jarrell Dep. (JX 356) 15.

43 of fair value. 217 Fourth, upon determining that Meinhart, on behalf Manichaean, has

presented a credible valuation opinion, I discharge my independent obligation to

determine SourceHOV’s fair value by reviewing the record to assess whether there

are opportunities for the Court to improve upon what Meinhart has done.218 With

one minor exception, I see no basis in the evidence to depart from Meinhart’s

calculations.

A. The Statutory Appraisal Remedy The Delaware appraisal statute “provide[s] equitable relief for shareholders

dissenting from a merger on grounds of inadequacy of the offering price.”219

The statute directs the court to:

determine the fair value of the shares exclusive of any element of value arising from the accomplishment or expectation of the merger or consolidation, together with interest, if any, to be paid upon the amount

217 Cede & Co. v. Technicolor, Inc., 884 A.2d 26, 35–36 (Del. 2005) (“Cede III”) (observing that “[i]t is often the case in statutory appraisal proceedings that a valuation dispute becomes a battle of experts . . . present[ing] [] conflicting expert testimony” and “[t]he Court of Chancery, as the finder of fact in an appraisal case, enjoys the unique opportunity to examine the record and assess the demeanor and credibility of witnesses” as “the sole judge of the credibility of live witness testimony”) (internal quotations omitted). 218 See, e.g., In re Appraisal of Jarden, 2019 WL 3244085 (finding that neither valuation expert had presented an entirely credible DCF valuation and, therefore, undertaking a separate analysis). 219 Cede & Co. v. Technicolor, Inc., 542 A.2d 1182, 1186 (Del. 1988) (“Cede I”).

44 determined to be the fair value. In determining such fair value, the Court shall take into account all relevant factors.220

“Fair value,” in the statutory appraisal context, “is not equivalent to the

economic concept of fair market value.”221 Rather, it is “a jurisprudential concept”

that seeks to calculate “the value of the company as a going concern, rather than its

value to a third party as an acquisition.” 222 When assessing fair value, Delaware

courts have understood that the statutory direction to consider “all relevant factors”

mandates consideration, at least, of “all generally accepted techniques of valuation

used in the financial community.” 223 Even so, our Supreme Court has made clear

that statutory appraisal is a “flexible process” that vests the Court of Chancery with

“significant discretion” to determine fair value. 224 In exercising this discretion, the

court may “select one of the parties’ valuation models as a general framework, or

fashion its own.”225

220 8 Del. C. § 262(h) (emphasis supplied). 221 Merion Capital L.P. v. Lender Processing Servs., Inc., 2016 WL 7324170, at *13 (Del. Ch. Dec. 16, 2016) (quotation and citation omitted). 222 Del. Open MRI Radiology Assoc., P.A. v. Kessler, 898 A.2d 290, 310 (Del. Ch. 2006); M.P.M. Enters., Inc. v. Gilbert, 731 A.2d 790, 795 (Del. 1991). 223 Cede I, 542 A.2d at 1186–87 (citing Weinberger v. UOP, Inc., 457 A.2d 701, 712–13 (Del. 1983)). 224 Golden Telecom, 11 A.3d at 218. 225 Cede & Co. v. Technicolor, Inc., 684 A.2d 289, 299 (Del. 1996) (“Cede II”).

45 In the unique creature that is a Delaware appraisal trial, both parties “bear the

burden of establishing fair value by a preponderance of the evidence, which

effectively means that neither party has the burden, and the burden instead falls on

this court.”226 Over the years, the court has not been shy about expressing its

exasperation with the task of sifting through complex financial data to reach a fair

value determination, particularly when the parties’ valuation experts, who ostensibly

are meant to “help the trier of fact,”227 view their roles, instead, as advocates. 228 This

frustration was on full display in Gonsalves v. Straight Arrow Publishers, Inc., where

Chancellor Allen was tasked with determining the fair value of Straight Arrow

226 In re Appraisal of Ancestry.com, Inc., 2015 WL 399726, at *16 (Del. Ch. Jan. 30, 2015) (citing Huff Fund Inv. P’ship v. CKX, Inc., 2013 WL 5878807, at *9 (Del. Ch. Nov. 1, 2013); Highfields Capital, Ltd. v. AXA Fin., Inc., 939 A.2d 34, 42–43 (Del. Ch. 2007) (“[I]f neither party adduces evidence sufficient to satisfy this burden, the court must then use its own independent judgment to determine fair value.”)). Each party’s burden includes the burden of proving the propriety of their proffered valuation method. Merion Capital, 2016 WL 7324170, at *12 (internal citations omitted). 227 Del. R. Evid. 702. 228 See, e.g., In re Orchard Enters., Inc., 2012 WL 2923305, at *18 (Del. Ch. July 18, 2012) (“As a law-trained judge who has to come up with a valuation deploying the learning of the field of corporate finance, I choose to deploy one accepted method as well as I am able, given the record before me [of competing expert testimony] and my own abilities.”); In re Emerging Commc’ns Inc. S’holders Litig., 2004 WL 1305745, at *11 (Del. Ch. May 3, 2004) (comparing petitioners’ valuation at $41 per share with respondent’s valuation of $10.38 per share and noting that “[t]hese wildly differing valuations of the same company result from quite different financial assumptions that each sponsoring [expert] exhorts this Court to accept”).

46 Publishers, Inc.229 Having seen in pretrial submissions that the experts were light

years apart in their appraisals, Chancellor Allen quipped that he was inclined to take

a “temperamental approach . . . [by] accept[ing] one expert or the other hook, line

and sinker.”230 While the trial court’s comment clearly was intended to express a

fact-finder’s frustration, our Supreme Court took the comment literally when

Chancellor Allen ultimately decided that one of the experts, in fact, had presented a

credible analysis of fair value and accepted that opinion as his own.231

On appeal, the Supreme Court reversed and remanded upon concluding that

the trial court erred when it made a “pretrial decision to adhere to, and rely upon, the

methodology and valuation factors of one expert to the exclusion of other relevant

evidence.”232 The Supreme Court was particularly concerned that the trial court

stated before the trial (likely in jest) that it intended to listen to the evidence, pick an

229 Gonsalves I, 1996 WL 696936, at *1. 230 Gonsalves II, 701 A.2d at 358 (emphasis supplied). 231 Id. 232 Id. (emphasis supplied).

47 expert and call it a day. 233 According to the Supreme Court, this approach was

“at odds” with Section 262’s command “that the Court ‘shall appraise’ fair value.”234

While the Supreme Court took issue with the trial court’s pretrial comments,

it was careful to explain, “the selection of one expert to the total exclusion of another

is [not], in itself, an arbitrary act.”235 The Court acknowledged that, even in appraisal

cases, the Court of Chancery should assess expert testimony under the “usual

standards which govern the receipt of such evidence.”236 Since Gonsalves, the

Supreme Court has acknowledged, “the Court of Chancery’s role as an independent

appraiser does not necessitate a judicial determination that is completely separate

and apart from the valuations performed by the parties’ expert witnesses.” 237 Indeed,

as long as the trial court “carefully consider[s] whether the evidence supports the

valuation conclusions advanced by the parties’ respective experts,” “it is entirely

proper for the [court] to adopt any one expert’s model, methodology, and

233 Id. at 361 (“[T]he nub of the present appeal is not merely that the Chancellor made an uncritical acceptance of the evidence of [one expert] but that he announced in advance that he intended to choose between absolutes.”); id. (holding that such a pre-determination established an impermissible “evidentiary construct”). 234 Id. See also id. at 358, 360 (emphasizing that that the Court of Chancery must “employ its own acknowledged expertise” and not “exclu[de]” any “relevant evidence”). 235 Id. at 361. 236 Id.; Cede III, 884 A.2d at 35 (“[T]he Court of Chancery is the sole judge of the credibility of live witness testimony.”) (internal quotation omitted). 237 M.G. Bancorporation, 737 A.2d at 525–26.

48 mathematical calculations, in toto, if that valuation is supported by credible evidence

and withstands a critical judicial analysis on the record.”238

B. DCF Is the Only Reliable Means By Which to Appraise SourceHOV

In fulfilling the statutory mandate to account for “all relevant factors” bearing

on “fair value,” Delaware courts consider a range of evidence that often includes

(i) “market evidence,” such as a company’s unaffected trading price or the

“deal price” following an appropriate “market check”239 and (ii) “traditional

valuation techniques,” 240 such as a comparable company, comparable transaction or

DCF analysis. 241 In this case, however, the parties and their experts agree that the

238 Id. at 526 (emphasis in original) (citing Cede II, 684 A.2d at 299; Gonsalves II, 701 A.2d at 361–62). The parties here both acknowledge, as a general matter, that it is proper for the trial court to accept the opinion of a valuation expert in its entirety, to the exclusion of other evidence of fair value, even in the statutory appraisal paradigm. Importantly, they also agree that this is a proper case for the Court to take that approach. Of course, they disagree on which of the experts has offered the definitive fair value appraisal of SourceHOV. Post-Trial Oral Arg. (D.I. 109) at 30–37, 65–68. 239 Dell, Inc., 177 A.3d at 27–30. 240 Cede I, 542 A.2d at 1186–87 (noting that Weinberger directs that the trial court consider traditional valuation techniques if relevant to fair value); Highfields Capital, 939 A.2d at 37 (describing DCF as a “traditional valuation methodology”). 241 Cede II, 684 A.2d at 297; see, e.g., Merion Capital, 2016 WL 7324170, at *14 (considering deal price); Dell, Inc., 177 A.3d at 5 (considering, among other factors, unaffected stock price and a DCF analysis); In re Appraisal of Dole Food Co., Inc., 114 A.3d 541, 550 (Del. Ch. 2014) (listing factors the court often considers such as market price, merger price, other offers for the company, prices at which knowledgeable insiders sold their shares, internal corporate documents and valuation work prepared for non- litigation purposes).

49 circumstances surrounding the Business Combination disqualify market evidence as

reliable inputs for a fair value analysis.242 Accordingly, the valuation presentation

from both sides focused on DCF. In my view, that focus was well placed.

SourceHOV’s deal process (or lack thereof) undermines any reliance on deal

price as an indicator of fair value.243 Moreover, as a private company, SourceHOV’s

equity was not traded in an efficient market, so its unaffected market price is also an

unreliable indicator of fair value.244 Without reliable market evidence of fair value,

242 JX 346 at 4 (summarizing points of agreement). See also Post-Trial Oral Arg. (D.I. 109) at 55, 103; Meinhart Op. at 43 (rejecting certain pre-Business Combination transactions because they were “conflicted”); Jarrell Op. at 1 (basing his estimation of fair value “primarily” on a DCF analysis). See Dell, Inc., 177 A.3d at 22 (“In some cases, it may be that a single valuation metric is the most reliable evidence of fair value and that giving weight to another factor will do nothing but distort that best estimate.”). 243 SourceHOV did not hold a single Board meeting to consider the Business Combination, nor did it solicit offers from third parties after Quinpario made its overture in January 2017. Tr. 210 (Chadha); see, e.g., Dell, Inc., 177 A.3d at 5–13 (reviewing a deal process to assess whether deal price was a persuasive indicator of fair value); In re Appraisal of AOL Inc., 2018 WL 1037450, at *8 (Del. Ch. Feb. 23, 2018) (inquiring whether a sale process was “Dell Compliant”); Merion Capital, 2016 WL 7324170, at *16–18 (reviewing a board’s sale process when considering the “persuasiveness of the initial merger consideration” as an indicator of fair value). 244 Tr. 870 (Jarrell) (SourceHOV had “no publicly traded stock prices”); JX 419 (debt pricing services observed incorrect and incomplete information regarding SourceHOV’s debt); Tr. 496–97 (Cascarilla) (SourceHOV’s stock “isn’t traded on any exchange.”); see, e.g., Merion Capital, 2016 WL 7324170, at *14 (noting that trading prices can be persuasive indicators of fair value when pricing data is available from a “thick and efficient market”) (internal citation omitted). Respondent does point to SourceHOV’s conversations with Madison Dearborn Partners (“MDP”) regarding an October 18, 2016, term sheet contemplating a $100 million investment into SourceHOV (valuing SourceHOV at $275–355 million) as some “market evidence” of fair value. RPTOB at 5, 16–17 (citing JX 45). I disagree. Contemporaneous documents reveal that the MDP discussions 50 the parties were left to focus on “traditional valuation methods” to appraise

SourceHOV. 245 This, of course, places the spotlight squarely on their competing

valuation experts. In other words, as I see it, this case has played out as the

quintessential “battle of the experts.”

Both experts agree there are no sufficiently comparable companies or

transactions with which to perform either a trading multiples or a transaction

multiples analysis. 246 Given that other valuation techniques do not fit here, both

experts also agree that a DCF analysis is the only reliable method to calculate

SourceHOV’s fair value. 247 In light of the experts’ agreement, and seeing no reason

to disagree, I am satisfied that a DCF analysis is the only reliable indicator of

SourceHOV’s fair value.248

were, at best, preliminary and did not proceed into anything meaningful because MDP simply “couldn’t move as quickly” as Exela. See JX 431 at 1. 245 Jarrell Op. at 4–5 (concluding that a DCF is the only “reliable indicator of value” for SourceHOV); Meinhart Op. at 18 (same). 246 Jarrell Op. at 4–5, 101–08; Meinhart Op. at 18; JX 346 at 4 (“[B]oth [experts] reject trading and transaction multiples as an indication of SourceHOV’s value.”); Tr. 677–78 (Meinhart). 247 JX 346 at 4 (sub-point C). 248 See Owen v. Cannon, 2015 WL 3819204, at *16 (Del. Ch. June 17, 2015) (exclusively relying on a DCF analysis when “[t]he parties’ post-trial briefing focused exclusively on the use of a . . . DCF analysis”).

51 C. Respondent’s Fair Value Presentation Is Not Credible

While the experts agree that DCF is the most reliable means to appraise

SourceHOV, they, and others who have undertaken a DCF analysis with respect to

SourceHOV, all reach remarkably divergent fair value conclusions. A summary of

the DCF values in the record appears in the chart below:

Valuation Revenue Total Equity Per Share Projection Used Value Value Respondent’s Litigation Position249 Bank Case 250 $271.4 million $1,633 Jarrell’s Opinion Equity Case 252 $468.1 million $2,817 (Respondent’s expert) 251 Rothschild’s Backdated Valuation 253 3.4% model 254 $675 million N/A The Court’s Determination of Fair Lender Model $722 million $4,591 Value Meinhart’s Opinion Lender Model 256 $798 million $5,079 (Manichaean’s expert) 255 Rothschild’s February Valuation 257 Equity Case 258 $931 million N/A

249 RPTOB at 1. 250 Id. at 46. 251 Jarrell Op. at 4. 252 Id. at 50–51. Jarrell’s ultimate conclusion is not in his opinion because he adjusted his opinion after incorporating feedback from Meinhart. See Jarrell Dep. (JX 356) 15; JX 353 at Ex. 8 (updated). 253 JX 309 at 15. 254 The Backdated Valuation applied a separate 3.4% growth model that falls in between the Bank Case and the Equity Case / Lender Model. See id. 255 Meinhart Op. at 4–5. 256 Id. at 20. 257 JX 302 at 17. 258 Id.

52 Before drilling down on the elements of the experts’ competing analyses, it is

appropriate first to dilate on what is an important consideration in any adversarial

proceeding—even appraisal litigation—credibility. 259 Who presented the more

credible witnesses; who presented the more credible case? After carefully

considering the evidence, I find Respondent’s presentation lacked credibility for

three main reasons: (i) Respondent disagreed with its own expert over which revenue

projections to use in the DCF analysis and ultimately separated from its expert with

respect to SourceHOV’s fair value; (ii) Chadha, one of Respondent’s key witnesses,

was not at all forthright in explaining the circumstances surrounding the creation of

the Backdated Valuation; and (iii) Jarrell’s bespoke approach to calculating

SourceHOV’s beta lacks foundation, both within the expert valuation community

and the facts of record.

1. Respondent Disagrees With Its Own Expert

Both experts, Meinhart and Jarrell, agree that either the Lender Model or the

Equity Case are the best revenue projections to use in a SourceHOV DCF

259 Post-Trial Oral Arg. (D.I. 109) at 30–37, 65–68; Cede III, 884 A.2d at 35–36 (acknowledging the importance of the trial court’s assessment of “the demeanor and credibility of witnesses” in an appraisal proceeding); M.G. Bancorporation, 737 A.2d at 525–26 (holding that even though the court has a role as an “independent appraiser,” it may “adopt any one expert’s model, methodology, and mathematical calculations” if they are “supported by credible evidence and withstand[] a critical judicial analysis”).

53 valuation. 260 They used these forecasts because SourceHOV itself relied on them

when working with its auditor, financial advisors and debt rating agencies in the

period before the Business Combination.261 Indeed, the Lender Model was the most

up-to-date set of projections created before the Business Combination.262

Notwithstanding this persuasive evidence of reliability, Respondent disagrees

with its own expert and insists that the Bank Case is the best projection of

SourceHOV’s future cash flows. 263 Thus, in its zeal to reach a desired litigation

outcome, Respondent finds itself in the awkward position of advancing a position at

odds with its own expert and advisor, Jarrell and Rothschild.264

SourceHOV’s relatively poor performance in 2016 is not a sufficient reason

to ignore multiple experts’ opinions regarding likely future performance in favor of

260 Jarrell Op. at 50, 53; Meinhart Op. at 19–20. 261 Tr. 71–74 (Verma); JX 191 (providing Ernst & Young projections for a going concern memo); JX 191E (FCF-Base Model); Tr. 48–50 (Verma) (confirming SourceHOV provided Ernst & Young 5% revenue growth projections); JX 302 at 11 (the February Valuation using the Equity Case); JX 394 at 1 (sending the Lender Model to rating agencies); Jarrell Op. at 53 (observing that Rothschild used the Equity Case in its analysis); JX 136E (“Working Cap” tab, “SourceHOV Standalone”); Tr. 32–35 (Verma) (explaining that the Equity Case projections were provided to certain lenders before the Business Combination). 262 Tr. 17 (Verma) (the Lender Model was a “haircut” on the Equity Case.); Tr. 567, 686 (Meinhart). 263 See RPTOB at 46; Tr. 752–53 (Jarrell). 264 See Jarrell Op. at 50; JX 302 at 11 (the February Valuation assuming 5% growth in 2018–2021).

54 the seldom-used Bank Case (which projects only 2.2% revenue growth per year).265

SourceHOV’s compound annual revenue growth was 10.1% from 2014 until just

before the Business Combination. 266 Unlike Respondent’s recently minted litigation

position, the Equity Case and the Lender Model were not created as convenient

afterthoughts. SourceHOV’s management created both models after engaging in a

robust “iterative process” that ultimately allowed them to “st[and] behind” the work

they did to create them. 267

In any event, Respondent engaged an expert to opine on the most accurate

revenue projections for SourceHOV. 268 For his own calculations, he selected the

Equity Case. 269 Respondent’s willingness to continue to argue for the Bank Case—

even when its own expert rejected those projections—renders its overall presentation

substantially less credible. 270

265 RPTOB at 46–49; Resp’t’s Post-Trial Answering Br. (“RPTAB”) (D.I. 100) at 18; Jarrell Op. at 47. 266 JX 275 at 35. 267 Tr. 14–16 (Verma). 268 Jarrell Op. at 50, 53. 269 Id. 270 On this topic, I note that Meinhart enhances his credibility by decreasing management’s projections for Board-related expenses, stock-based compensation expenses and aggressive depreciation projections (thus lowering the level of SourceHOV’s projected income). Tr. 582–83 (Meinhart); Tr. 777, 787 (Jarrell) (Meinhart cured “an important defect [in management’s] projections” for depreciation and capital expenditures). See also Tr. 760 (Jarrell) (acknowledging the Lender Model had “advantages” over the other 55 2. Chadha’s Trial Testimony Was Not Credible

Chadha was the centerpiece of Respondent’s effort to paint the picture of a

company in trouble in order to lay foundation for its argument that SourceHOV’s

fair value was substantially south of Manichaean’s fair value mark. Indeed,

according to Chadha: (i) SourceHOV’s equity was worthless; 271 (ii) MDP turned

down an investment in SourceHOV because it did not think the Company’s equity

was worth $257 million; 272 (iii) SourceHOV was totally shut out from the debt

markets; 273 and (iv) all strategies to keep SourceHOV afloat, other than the Business

Combination, were hopeless.274

But Chadha simply was not believable. His litigation-driven effort to

persuade Rothschild to create the Backdated Valuation to appear as if it had been

revenue forecasts because they “were more current” and “reflected feedback from [] lenders,” “the parties” and “Apollo”). Jarrell testified he was “not qualified to second- guess [management]” on the Equity Case, and he did not think that management’s aggressive accounting tactics “ruin[] the projections” or require him to “go in and undo” management’s work. Tr. 769 (Jarrell). See also Tr. 775, 786 (Jarrell) (testifying there “should be a high standard for concluding that the projections are . . . not reasonable enough to use for a DCF”). 271 Tr. 216 (Chadha). 272 RPTOB at 16–17 (citing Tr. 339–40 (Chadha)). 273 Id. at 65 (citing Tr. 384–85 (Chadha)). 274 Id. at 15–19 (citing Tr. 322, 341, 335–36, 339–43 (Chadha)). Chadha also described the Business Combination as a “miracle.” Tr. 368 (Chadha).

56 prepared before the Business Combination was bad enough. 275 His failure even to

acknowledge that scheme, when it was finally exposed in discovery, taints all of his

testimony. 276

3. Jarrell’s Novel Approach To Determine SourceHOV’s Beta Undermines His Credibility

Perhaps the most consequential point of disagreement between the experts is

the appropriate method for calculating SourceHOV’s equity beta. 277 Meinhart

calculated SourceHOV’s beta indirectly based on 19 publicly traded comparable

companies; 278 Jarrell estimated SourceHOV’s beta directly using the yields and

interest rates on the First and Second Liens.279 Meinhart’s methodology is generally

accepted among valuation experts and finds direct support in academic literature,

275 Tr. 279–82 (Chadha); JX 301; JX 309 (Chadha receiving an email where Jonovic told Rothschild “the cover page says Jan 2018 . . . Happy for it to simply say July 2017”); JX 309 (Jonovic reporting back to Chadha with a single word when Rothschild finally agreed to remove all references to 2018—“Done”). 276 JX 316, Resp. No. 5 at 6–7. It is also worth noting that Respondent’s litigation position (i.e., that SourceHOV’s equity was worth $271 million) is a far cry from Rothschild’s Backdated Valuation (which valued SourceHOV’s equity at $675 million). See JX 309 at 15. 277 Tr. 809–10, 814 (Jarrell); JX 346 at 21–22; JX 343 at 13–19. 278 Meinhart Op. at 24–26, 36–37; Tr. 597–98 (Meinhart). 279 Jarrell Op. at 71 (Jarrell also considered, among other factors, the stated interest rates on Exela’s acquisition financing); Tr. 868–69 (Jarrell).

57 while Jarrell’s alternative method, by his own admission, does not.280 Jarrell

employed his methodology because, in his view, it fit the facts. In other words,

nothing “connects” this expert’s “opinion evidence . . . to existing data” except

“the ipse dixit of the expert.” 281 That Jarrell was so willing to go out on a limb to

support a forensic valuation opinion, of course, raises serious admissibility issues

under Daubert. 282 It also raised serious questions about the credibility of his entire

valuation analysis.

Not only is Jarrell’s approach to estimating Beta methodologically novel,

it also starves for want of support in the record. Beta is a measure of the systematic

risk of a stock—that is the tendency of a stock’s price to correlate with changes in

the market. 283 Valuation experts calculate beta in two ways: (i) directly through a

regression analysis of a public company’s stock prices and the market or

280 Tr. 826–30 (Jarrell); Tr. 597–98, 659 (Meinhart). 281 Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997). 282 See M.G. Bancorporation, 737 A.2d at 521 (adopting as Delaware law the United States Supreme Court’s seminal opinion in Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993), where the Court addressed the trial court’s discretion to exclude unreliable expert testimony under the federal counterpart to DRE 702). 283 JX 420 at 8–9. The “market” is typically represented by a broad-based equity index that includes a wide range of industries. Id. at 3. The market’s beta is 1.0 by definition. A company with a beta equal to 1 has the same risk as the market (it theoretically moves up and down with the market in tandem). A company with a beta greater than 1 is riskier than the market (i.e., it theoretically moves up and down to a greater degree than the market). Id.

58 (ii) indirectly by proxy. 284 But direct calculation is impossible when the target

company is privately held.285 Indeed, Jarrell admitted that “[a] stack of books”

supports the proposition that one “need[s] to use a proxy beta when the subject

business is . . . closely held.” 286 Even so, Jarrell was undeterred.

Using the courtroom as incubator for his experiment, Jarrell proceeded to

calculate SourceHOV’s equity beta directly by looking to market evidence of

SourceHOV’s debt. 287 In doing so, Jarrell abandoned the traditional, indirect, beta

approximation method because he did not believe there were public companies

sufficiently comparable to SourceHOV. 288 Instead, he employed his novel approach,

284 JX 420 at 2–4. 285 Tr. 809, 867–68 (Jarrell); JX 420 at 5–6 (“You need to use a proxy beta when the subject business is a division, reporting unit, or closely held business.”). 286 Tr. 829 (Jarrell) (emphasis supplied); see JX 343 at 16–19 (Meinhart’s rebuttal report discussing Jarrell’s methodology); JX 420; JX 423 at 3–8; Tr. 634 (Meinhart); Rothschild Dep. (JX 322) 240 (“[T]his is a private firm, generally, so you have to infer what investors would require as a cost of equity from publicly-traded, quote, unquote, ‘peers.’”). 287 Jarrell Op. at 68. See Steven J. Breyer, Introduction to Reference Manual on Scientific Evidence, FED. JUD. CTR., at 4 (3d Ed. 2011) (observing, in the context of Daubert, that “the courtroom is not a laboratory”); Minner v. Amer. Mort. & Guarantee Co., 791 A.2d 826 (Del. Super. Ct. 2000) (Quillen, J.) (“[T]here are important differences between the quest for truth in the courtroom and the quest for truth in the laboratory.”) (citation omitted). 288 Tr. 809–10 (Jarrell).

59 which he admitted he had “not seen” or “done” before. 289 Jarrell “thought of [his

method] for this case” and hoped that it would “catch on” in the future.290

I do not foreclose the possibility that Jarrell’s method will “catch on” after

proper vetting by his peers. If it does, perhaps then this court might be persuaded.

For now, however, I am not inclined to ignore the “stack of books” to chase

Respondent’s proffer of a shiny new penny. 291

Another, more fundamental, problem with Jarrell’s approach is his reliance

on SourceHOV’s debt yields as market evidence of SourceHOV’s actual cost of

debt.292 Jarrell looked to the weighted average yield to maturities of the First and

Second Liens, both at their issuance and their subsequent trading prices, to determine

SourceHOV’s cost of debt.293 The problem with this approach is that SourceHOV’s

289 Tr. 828 (Jarrell); Tr. 624 (Meinhart) (The comparable company method is the “preferred method for estimating a beta when you’re valuing a privately held company.”). 290 Tr. 828 (Jarrell). 291 The Court is ill-equipped to assess the merits of the theoretical debate in which Jarrell and Meinhart engaged regarding novel implications of the M&M Theorem for beta approximation, much less who will ultimately prevail should the debate continue in the academy where it belongs. Jarrell admits his theory is new and that Meinhart’s approach is tried and tested. Tr. 828–29 (Jarrell); Tr. 667–71 (Meinhart). As lay fact finder, I place my trust in the generally accepted methodology. 292 Jarrell Op. at 31, 71 (considering the current yield to maturity on the First and Second Liens ranging from 8.48% to 17.96% based on Bloomberg data). 293 Id.

60 debt was not publicly traded.294 The First and Second Liens were private loans that

traded only by appointment.295 And, at least once, Bloomberg reported prices on

SourceHOV bonds that no longer existed. 296 Thus, even if Jarrell’s approach were

sound in theory, his flawed execution does not engender confidence in the results.297

This leaves the traditional, indirect method Meinhart employed to

approximate SourceHOV’s equity beta. Respondent takes issue with Meinhart’s

beta calculation on two grounds. First, it argues Meinhart’s calculation is unreliable

because he derived SourceHOV’s beta from public companies that are not

comparable to SourceHOV in terms of industry or market focus.298 Second, and

294 Jarrell Dep. (JX 356) 47–53. 295 Id.; Tr. 472–76 (Cascarilla). 296 Jarrell Dep. (JX 356) 47–53; JX 419 (debt pricing services listed incorrect and incomplete information for SourceHOV’s debt). 297 In its Answering Brief, Respondent stresses that Jarrell only considered the flawed Bloomberg data as “confirmatory” evidence for his assessment of SourceHOV’s cost of debt. See RPTAB at 37. Respondent argues Jarrell primarily considered the yield to maturity on the First and Second Liens. Id. Ultimately, I find Respondent has failed to present enough credible evidence to support the conclusion that the yield to maturity on the First and Second Liens accurately represents SourceHOV’s cost of debt. Respondent cites Meinhart’s testimony that considering the yield to maturity on a company’s debt is “one of the ways to look at [cost of debt].” Id. at 33–34 (citing Tr. 717–18 (Meinhart)). But this is a far cry from the firm evidentiary foundation that would be required to conclude that the yield to maturity on the First and Second Liens reflected the Company’s “actual cost of debt, in an efficient market, full stop.” See JX 427 at 3 (warning against “using the debt yield as [a company’s] cost of capital” because “[w]hen the firm’s debt is risky, [] the debt yield will overestimate the debt cost of capital”). 298 RPTOB at 52–55.

61 relatedly, Respondent says Meinhart’s portfolio contained companies that were less

levered and much larger than SourceHOV. 299 These criticisms do not shake my

confidence in Meinhart’s valuation methodology.

As for comparability, Meinhart used many of the same comparable companies

that SourceHOV, its accountants and Rothschild used in their own beta calculations

before the Business Combination.300 While there may be some imprecision

associated with indirect beta estimates, it is generally accepted that when a company

is privately held, a comparable companies analysis is the best tool available to derive

beta, even if the comparable companies are larger or less levered.301 Jarrell does not

dispute this fact.302

Betas for any specific stock incorporate two risk factors: business

(or operating) risk and financial (or capital structure) risk. 303 By starting with a

comparable company’s beta and removing the effect of financial leverage

(i.e., unlevering the beta), the appraiser is left only with the effect of business risk,

299 Id. at 52; RPTAB at 38; JX 346 at 23–24. 300 Tr. 600–08 (Meinhart); Meinhart Op. at 64. 301 Tr. 629–39 (Meinhart); JX 420 at 4–6; JX 423 at 8–13. 302 Tr. 828–29 (Jarrell). 303 JX 420 at 8.

62 which can then be used to estimate the business risk of the target company. 304 While

this process necessarily invites some measurement error, appraisers decrease the risk

of error by employing a large pool of comparable companies, as Meinhart did in his

analysis.305 Indeed, the valuation authorities relied upon by the parties, and in the

record, recommend the exact de-levering process Meinhart employed as the best way

to adjust for differences in leverage between the target company and the comparable

companies.306

Both experts considered the Pratt and Grabowski text’s discussion of de-

levering betas. 307 It provides:

If the leverage of . . . [a] closely held company subject to valuation differs significantly from the leverage of the guideline public companies selected for analysis . . . it typically is desirable to remove the effect that leverage has on the betas before using them as a proxy to estimate the beta of the subject company. 308

304 Id. at 8–9. 305 Meinhart Op. at 36–37; Tr. 637 (Meinhart); Tr. 704–07 (Meinhart) (discussing his statistical analysis of SourceHOV’s comparable companies). 306 JX 420 at 8–9. 307 Tr. 634, 638 (Meinhart); Jarrell Op. at 68 n.201. 308 JX 420 at 9. Meinhart acknowledged that SourceHOV was smaller than his selected comparable companies, but he also included a size premium in his analysis. See Meinhart Op. at 23 (discussing his use of a size premium to “adjust for the size differential between SourceHOV and the empirical data from which the equity risk premium is derived”).

63 This is exactly what Meinhart did.309 After applying widely accepted de-levering

formulas, Meinhart chose the highest beta in his analysis to adjust for the possibility

that the de-levering and re-levering process underestimated SourceHOV’s beta.310

Given Meinhart’s conservative execution of widely accepted beta approximation

methods, his beta value is both reasonable and credible, while Jarrell’s admittedly

novel process does not survive judicial scrutiny—at least not on this record.

*****

Jarrell’s presentation stood out as untethered to accepted methods and

generally not credible. Since Respondent’s fair value position rested on Jarrell’s

opinion,311 its fair value proffer suffers the same flaws. In other words, Respondent

has failed to prove its valuation position by a preponderance of the evidence.312

Respondent maintains that Meinhart’s appraisal substantially overvalues

SourceHOV. I address those criticisms below.

309 Meinhart Op. at 24–25. 310 Tr. 638–39 (Meinhart); Meinhart Op. at 26. 311 As noted, at trial, Respondent started with its expert’s conclusions and then endeavored to adjust them downward to reach a lower fair value for SourceHOV. 312 Highfields Capital, 939 A.2d at 42.

64 D. Meinhart’s Fair Value Opinion Is Both Credible and Reasonable

Respondent has identified five areas where Meinhart’s DCF inputs are flawed:

(1) debt load projections, (2) depreciation and amortization projections, (3) the

appropriate set of SourceHOV financial statements to use in a DCF analysis, (4) the

total shares of SourceHOV stock outstanding before the Business Combination and

(5) the appropriate size premium to apply in a CAPM analysis. 313 I address each

input below.

1. Debt Load Projections

When employing either the CCF or APV model, the appraiser must calculate

the net present value of a company’s income tax shields using reliable projections of

the company’s future debt load. 314 Without an accurate projection of future debt, it

is impossible accurately to predict tax savings. 315 Meinhart assumed SourceHOV

would carry significant debt past the year 2020.316 Jarrell, on the other hand,

313 Meinhart and Jarrell use the risk-free rate of return and the same equity risk premium. Tr. 809 (Jarrell). 314 See Tr. 571–74 (Meinhart). 315 Id. 316 See, e.g., Meinhart Op. at 59.

65 predicted that SourceHOV would have to refinance all of its debt—leading to lower

tax savings from interest deductions. 317

Rather than create his own forecasts, Meinhart based his projection on

SourceHOV management’s own forecasts.318 These forecasts predicted SourceHOV

would continue to carry debt even after SourceHOV repaid the First and Second

Liens. 319 Contrary to Respondent’s criticism, Meinhart did not ignore SourceHOV’s

high leverage ratios. He recognized SourceHOV’s high debt loads were stressing

the Company. Accordingly, he assumed SourceHOV would “try[] to reduce debt as

rapidly as it could.”320

For his part, Jarrell assumed SourceHOV would retire 100% of its debt in

2020 when SourceHOV repaid the First and Second Liens.321 When considered in

context with the entire record, Jarrell’s assumption is not reasonable. Given

SourceHOV’s acquisitive history, and its past tolerance for high debt loads, it is

317 Tr. 654 (Meinhart); Tr. 837 (Jarrell). 318 Tr. 645 (Meinhart); Meinhart Op. at 59 (incorporating debt projections from the Lender Model). 319 Tr. 654, 571–74 (Meinhart); JX 75E (Nov. 2016 “Debt Schedule” tab, line 58, columns G–Z); JX 191E (March 2017 “Debt Sheet” Tab, line 81 columns H–AI); JX 211E. 320 Tr. 575 (Meinhart). Respondent argues that SourceHOV would have been cut off from the debt markets when the First and Second Liens matured. See RPTOB at 65. Respondent cites testimony from Chadha for this proposition. RPTOB at 65 (citing Tr. 384 (Chadha)). As discussed above, Chadha was not credible. 321 Tr. 571–74 (Meinhart); Tr. 837 (Jarrell).

66 unlikely SourceHOV would have abruptly abandoned its strategy of using debt to

fuel future acquisitions.322 Management’s projections realistically forecast that

SourceHOV would continue to carry debt after the First and Second Liens

matured. 323 Meinhart’s reliance upon these projections was reasonable and

supported by credible evidence.

2. Depreciation and Amortization Projections

In his analysis, Meinhart recognized that SourceHOV management had

forecast “very high depreciation and amortization and relatively low capital

expenditures.”324 This forecast led to “depreciating and amortizing more asset value

than [SourceHOV] even ha[d] on the books.”325 If Meinhart had accepted this high

level of depreciation and amortization (as Jarrell did), the result would have been to

increase SourceHOV’s value in a DCF analysis. 326 Instead, to account for his

concern that depreciation and amortization forecasts were too high, Meinhart made

322 Tr. 306 (Chadha) (“[A]lmost all” of SourceHOV’s later acquisitions were funded with 100% debt.). 323 JX 75E (Nov. 2016 “Debt Schedule” tab, line 58, columns G–Z); JX 191E (March 2017 “Debt Sheet” Tab, line 81 columns H–AI); JX 211E. 324 Tr. 585 (Meinhart). 325 Id. 326 See Jarrell Op. at 60 (accepting management’s “very aggressive” reinvestment rates).

67 a Respondent-friendly adjustment to provide a more accurate calculation.327 Once

he made this adjustment, Jarrell, in large measure, followed suit.328

Even though Respondent and its expert abandoned their own depreciation and

amortization calculations in favor of Meinhart’s, they criticize his approach for

treating certain asset depreciation values as tax deductible when the tax code would

treat them as non-deductible. 329 Meinhart responds by arguing that SourceHOV did

not produce a “tax basis runout” of its assets before he prepared his expert report.330

Thus, while Meinhart would have preferred to begin with tax basis instead of book

basis when preparing his depreciation and amortization schedules, that option was

not available to him. 331 Accordingly, he used the same available book values and

corresponding projections that SourceHOV, itself, had created and used for its own

forecasts.332 Again, this was the only data made available to him.

327 Meinhart Op. at 20–21. 328 See Jarrell Dep. (JX 356) 15; JX 353 at Ex. 8 (updated). 329 See RPTOB at 49–51. 330 Tr. 588 (Meinhart); Jarrell Dep. (JX 356) 36 (confirming SourceHOV did not produce a full set of contemporaneous documents showing the full tax basis of its goodwill and other assets). 331 Tr. 588 (Meinhart). 332 Meinhart Op. at 20–21; Tr. 589–90, 593 (Meinhart).

68 In his rebuttal report, Jarrell “recalculated [Meinhart’s] D&A Projections, but

replace[d] [Meinhart’s] forecasted total goodwill with just the portion of goodwill

that is tax deductible.” 333 Meinhart objects to Jarrell’s recalculation, and for good

reason. Tax basis accounting and book basis accounting involve fundamentally

different rules.334 The appraiser should analyze either book depreciation or tax

depreciation since the two numbers can be vastly different. 335 I reject Jarrell’s

argument that the “default rule” should be that goodwill is not tax deductible.336

Allowing Respondent to modify Meinhart’s book basis depreciation runouts would

reward the lack of information flow between the parties and give an unreasonable

inference to SourceHOV. 337

Jarrell’s effort to do a tax analysis on book values, in my view, is not

reasonable. I am persuaded Meinhart’s depreciation and amortization projections

are the best-available forecasts. Indeed, Petitioners’ point that Meinhart’s

333 JX 346 at 14. 334 Tr. 587–88 (Meinhart); JX 373; Jarrell Dep. (JX 356) 37–38. Jarrell admits he did not consider or assess whether other intangible assets subject to depreciation and amortization (such as tradenames) were tax deductible and that he is “way out of [his] league with accounting questions.” Jarrell Dep. (JX 356) 38. 335 Tr. 589–93 (Meinhart) (explaining the nuances of a tax depreciation runoff schedule that made him “uncomfortable with the mixing” and why he “decided to stick with [his] schedule”). 336 Tr. 874–75 (Jarrell). 337 Jarrell Dep. (JX 356) 36–37 (SourceHOV did not produce tax documents).

69 calculation “is the only full book-basis or tax-basis calculation provided by either

party” is well taken.338 I find Meinhart’s approach to be both reasonable and

3. The Selection of Appropriate Financial Statements and Forecasts

The parties dispute which SourceHOV financial statements and forecasts most

accurately project the Company’s future cash flows.339 For his calculations,

Meinhart used SourceHOV’s balance sheet, cash flow and net debt financial

information as of March 31, 2017, because, as a practical matter, these results were

the last SourceHOV numbers available before the Business Combination.340

Multiple sources corroborate the reasonableness of Meinhart’s choice.

First, SourceHOV’s management represented that, as of July 12, 2017, there

were no more updated financial statements than those Meinhart used in his

analysis.341 Second, on July 11, 2017, SourceHOV also told its auditor that it only

had “best estimates” for May and June income statements. 342 Third, when

338 PPTAB at 63 (emphasis supplied). 339 Tr. 656, 681–83 (Meinhart); JX 346 at 15–16. 340 Meinhart Op. at 21, 47–51; Tr. 654–56, 686 (Meinhart). 341 JX 411 (management confirming that “n[o] consolidated financial statements are available as of any date or for any period subsequent to March 31, 2017.”). 342 JX 378 at 2.

70 Rothschild performed its Backdated Valuation for litigation purposes, it used the

same financial statements as Meinhart. 343

On the other hand, Respondent asks the Court to rely on second-quarter

information that was not realistically available until about a month after the Business

Combination closed.344 While second-quarter data may have existed before July 12,

on this record, I find Meinhart’s decision to use the March 31 financial statements

both reasonable and supported by credible evidence.345

4. The Correct Calculation of Total Outstanding Shares

It is undisputed SourceHOV’s fully “diluted” share count at the time of the

Business Combination was 157,249. 346 But the parties disagree over whether

SourceHOV’s Restricted Stock Units (“RSUs”) should be included in the count of

total outstanding shares. This disagreement is important because if the RSUs are

included in the count, then the effect is to dilute the holdings of existing

stockholders, including Petitioners. 347 In his analysis, Meinhart did not count any of

343 JX 308 at 4 (considering “net debt figures as of March 31, 2017”). 344 JX 292 (Exela 8-K releasing second quarter financial statements on August 9, 2017); RPTOB at 66–67. 345 Tr. 539–40 (Reynolds) (admitting that “it usually takes time” to prepare financial statements after a quarter ends and that financial statements are not “instantaneously” available). 346 JX 292 at 5; JX 265 at 68. 347 JX 346 at 61.

71 SourceHOV’s 8,887 RSUs granted under the Company’s Long-Term Incentive Plan

(the “Plan”) because, in his view, whether vel non those RSUs would vest was, at

best, speculative. 348

According to the Plan, the holder must be alive, not disabled and employed

with the Company in order to convert her RSUs. 349 It is undisputed that at least

1,192 of the unvested RSUs were forfeited within 5 months of the Business

Combination and over 10,000 were forfeited from 2014–16.350 Given this history,

Meinhart’s reluctance to count the RSUs in the share count was justified.

5. The Applicable Size Premium

The parties agree that applying a size premium is appropriate and that it should

be determined using Duff & Phelps’ 2017 Valuation Handbook, which provides size

premiums based on market capitalization. 351 But, of course, the parties dispute

SourceHOV’s market capitalization at the time of the Business Combination and,

therefore, the experts disagree on the appropriate size premium. 352 Meinhart

348 Tr. 734–35 (Meinhart); JX 343 at 24. 349 JX 265 at 428–29; JX 383 §§ 11(a) at 11–12, 12(a)(ii) at 12, 13(a) at 14. 350 Tr. 543–44 (Reynolds); JX 385 at 105; JX 265 at 429–30. 351 RPTOB at 58 (citing Tr. 690 (Meinhart)). As noted, a size premium accounts for the additional risk of investing in a smaller company compared with the broader index of companies represented in a market index. Meinhart Op. at 23; Jarrell Op. at 67. 352 Compare Jarrell Op. at 74–78 (using decile 9 and a size premium of 2.68%), with Meinhart Op. at 23 (using decile 8 and a size premium of 2.08%).

72 determined SourceHOV’s market capitalization using Exela’s stock price just after

the Business Combination and Rothschild’s analyses of SourceHOV’s share of the

consideration in the Business Combination. 353 Based on Exela’s $8.61 per share

stock price on July 12, 2017, and Rothschild’s calculations in the February

Valuation, Meinhart concluded SourceHOV’s market capitalization was greater than

$569.279 million. 354 This puts SourceHOV in Duff & Phelps’ “8th decile”—

yielding a size premium of 2.08%.355

On the other hand, Jarrell considered a post-closing decrease in Exela’s stock

price to determine the applicable size premium. 356 Specifically, one week after the

Business Combination, Exela released an 8-K disclosing that many Quinpario

stockholders had elected to redeem their shares rather than participate in the

Business Combination. 357 In response, Exela’s stock price decreased to $6.98 per

353 Meinhart Op. at 23–24; JX 302 at 12 (showing Rothschild’s February Valuation calculating SourceHOV’s merger consideration between $806 million and $1.003 billion); JX 309 at 10 (showing Rothschild’s Backdated Valuation valuing SourceHOV’s merger consideration between $645 million and $806 million). 354 Meinhart Op. at 23–24 (citing JX 302 at 12). 355 Id. 356 Tr. 835–36 (Jarrell). 357 JX 288.

73 share on July 19, implying a $563 million value for SourceHOV—below the

$569 million market capitalization threshold for decile 9. 358

Ultimately, Jarrell chose the 2.68%, decile 9, size premium for two reasons.

First, the trading activity on July 19 reflected the market’s informed reaction to

Quinpario’s redemptions, an outcome that was knowable before the Business

Combination. 359 Second, the market price of the Exela stock SourceHOV received

in the Business Combination necessarily overstates SourceHOV’s value because it

includes synergies arising from the Business Combination.360

After reviewing both parties’ arguments related to the applicable size

premium, I find that both sides have presented reasonable arguments for either the

2.08% or the 2.68% size premiums. But I am persuaded the 2.68% size premium is

more accurate on this record. In reaching this conclusion, I am cognizant that

selecting the applicable size premium requires some circularity since its main input

(market capitalization) is usually a strong indicator of a company’s fair value.361

358 See Jarrell Op. at 26–27, 75 (considering Exela’s stock price on July 19, 2017 and August 10, 2017). 359 Jarrell Op. at 75–77; In re Appraisal of AOL, 2018 WL 1037450, at *10 (citations omitted). 360 Tr. 836 (Jarrell); Jarrell Op. at 76. 361 See Market Capitalization, Merriam-Webster (Dec. 31, 2019, 10:59 AM), https://www.merriam-webster.com/dictionary/market%20capitalization (defining market 74 I am also aware that I am selecting a market capitalization for size premium purposes

that contradicts my ultimate determination of SourceHOV’s fair value. But, on this

record, both experts applied a size premium based on Exela’s post-Business

Combination stock price.362 The question becomes which expert’s assumptions

were more reliable and a better reflection of SourceHOV’s operative reality.

Between the two experts’ approaches, I am persuaded a 2.68% size premium is more

accurate given that it incorporates information that was knowable as of the Business

Combination.

E. SourceHOV’s Fair Value and The Court’s Independent Burden

After reviewing the parties’ evidentiary presentations and arguments in the

context of the entire record, I determine SourceHOV’s fair value immediately before

the Business Combination was $4,591 per share. 363 This valuation incorporates my

judgment that Meinhart’s DCF model accurately reflects SourceHOV’s fair value.

After carefully reviewing the analysis, I adopt it in toto, except for my adjustment to

the applicable size premium. I reach this conclusion after considering whether there

are any additional adjustments to Meinhart’s DCF valuation that are justified in the

capitalization as a company’s “current stock price” multiplied by its total “shares outstanding”); Jarrell Op. at 72 n.244 (same). 362 Jarrell Op. at 75; Meinhart Op. at 23. 363 I arrive at this number by modifying the size premium Meinhart applied in his calculations. See JX 351E; Appendix 1–3 (attached to this Opinion).

75 record.364 After applying my own “critical judicial analysis,” I see no basis to tinker

with the careful analysis of a valuation expert whose testimony I have found to be

credible and whose conclusions are well supported by the evidence.365

III. CONCLUSION

For the foregoing reasons, I have found the fair value of SourceHOV as of the

Business Combination was $4,591 per share. The legal rate of interest, compounded

quarterly, shall accrue from the date of the Business Combination’s closing to the

date of payment. The parties shall confer and submit an implementing order and

final judgment within ten days.

364 See M.G. Bancorporation, 737 A.2d at 526–27. 365 Id. at 526.

76 APPENDIX 1

EXHIBIT 11a SOURCEHOV HOLDINGS, INC. DISCOUNTED CASH FLOW METHOD WEIGHTED AVERAGE COST OF CAPITAL AS OF JULY 12, 2017

Cost of Equity Capital:

Modifie d Capital Asse t Pricing Mode l:

Risk-Free Rate of Return [a] 2.65% 2.65% 2.65%

General Equity Risk Premium [b] 5.97% 5.97% 5.97% Multiplied by: Industry Beta (rounded) [c] 1.37 2.46 2.02 Industry-Adjusted General Equity Risk Premium 8.18% 14.69% 12.06%

Size Equity Risk Premium [d] 2.68% 2.68% 2.68% Indicated Cost of Equity Capital 13.51% 20.02% 17.39%

Se le cte d Cost of Equity Capital (rounde d) 13.5% 20.0% 17.4%

Cost of De bt Capital:

Before-T ax Cost of Debt Capital 4.42% [e] 9.73% [f] 9.73% [f] Income T ax Rate [g] 37.0% 37.0% 37.0%

Se le cte d Cost of De bt Capital (rounde d) 2.8% 6.1% 6.1%

We ighte d Ave rage Cost of Capital Calculation:

Selected Cost of Equity Capital 13.5% 20.0% 17.4% Multiplied by: Equity/Invested Capital (rounded) 83.0% [h] 37.9% [i] 48.5% [j] Equals: Weighted Cost of Equity Capital 11.2% 7.6% 8.4%

Selected Cost of Debt Capital 2.8% 6.1% 6.1% Multiplied by: Debt/Invested Capital (rounded) 17.0% [h] 62.1% [i] 51.5% [j] Equals: Weighted Cost of Debt Capital 0.5% 3.8% 3.1%

We ighte d Ave rage Cost of Capital (rounde d) 11.7% 11.4% 11.6%

Se le cte d We ighte d Ave rage Cost of Capital 11.6%

Definitions are presented in Schedule A. [a] 20-year U.S. T reasury bond, Federal Reserve Statistical Release , as of July 12, 2017. [b] Duff & Phelps, 2017 Valuation Handbook: U.S. Guide to Cost of Capital . [c] As presented in Exhibit 13, Hamada relevered beta. [d] Duff & Phelps, 2017 Valuation Handbook: U.S. Guide to Cost of Capital , 8th size decile. [e] Moody's Baa corporate bond yield as of July 12, 2017. [f] S&P high yield CCC corporate bond yield as of July 12, 2017.

[g] Based on the Company-provided expected income tax rate of 37 percent. See ROT HSCHILD00107, Excel file and SHOV-QP-00036116. [h] Based on the median capital structure of the guideline publicly traded companies. See Exhibit 12. [i] Based on SourceHOV estimated pre-listing equity value of $645 million and debt as of March 31, 2017, of $1,055 million. [j] Based on SourceHOV estimated fully distributed equity value of $1,003 million and debt as of December 31, 2016, of $1,064 million. Sources: As indicated above. APPENDIX 2

EXHIBIT 11b SOURCEHOV HOLDINGS, INC. CAPITAL CASH FLOW METHOD PRETAX WEIGHTED AVERAGE COST OF CAPITAL AS OF JULY 12, 2017

General Equity Risk Premium [b] 5.97% 5.97% 5.97% Multiplied by: Industry Beta (rounded) [c] 1.37 2.46 2.02 Industry-Adjusted General Equity Risk Premium 8.18% 14.69% 12.06%

Size Equity Risk Premium [d] 2.68% 2.68% 2.68% Indicated Cost of Equity Capital 13.51% 20.02% 17.39%

Se le cte d Cost of Equity Capital (rounde d) 13.5% 20.0% 17.4%

Before-T ax Cost of Debt Capital 4.42% [e] 9.73% [f] 9.73% [f] Income T ax Rate [g] - - -

Se le cte d Cost of De bt Capital (rounde d) 4.4% 9.7% 9.7%

Selected Cost of Equity Capital 13.5% 20.0% 17.4% Multiplied by: Equity/Invested Capital (rounded) 83.0% [h] 37.9% [i] 48.5% [j] Equals: Weighted Cost of Equity Capital 11.2% 7.6% 8.4%

Selected Cost of Debt Capital 4.4% 9.7% 9.7% Multiplied by: Debt/Invested Capital (rounded) 17.0% [h] 62.1% [i] 51.5% [j] Equals: Weighted Cost of Debt Capital 0.7% 6.0% 5.0%

Pre tax We ighte d Ave rage Cost of Capital (rounde d) [l] 12.0% 13.6% 13.4%

Unle ve re d Cost of Equity Capital:

Risk-Free Rate of Return [a] 2.65%

General Equity Risk Premium [b] 5.97% Multiplied by: Industry Beta (rounded) [k] 1.21 Industry-Adjusted General Equity Risk Premium 7.2%

Size Equity Risk Premium [d] 2.68% Indicated Cost of Equity Capital 12.55%

Conclude d Unle ve re d Cost of Equity Capital (rounde d) 12.6%

Se le cte d Pre tax We ighte d Ave rage Cost of Capital 12.9%

Definitions are presented in Schedule A. [a] 20-year U.S. T reasury bond, Federal Reserve Statistical Release , as of July 12, 2017. [b] Duff & Phelps, 2017 Valuation Handbook: U.S. Guide to Cost of Capital . [c] As presented in Exhibit 13, Hamada relevered beta. [d] Duff & Phelps, 2017 Valuation Handbook: U.S. Guide to Cost of Capital , 8th size decile. [e] Moody's Baa corporate bond yield as of July 12, 2017. [f] S&P high yield CCC corporate bond yield as of July 12, 2017. [g] Income tax rate is eliminated to arrive at a pretax weighted average cost of capital. [h] Based on the median capital structure of the guideline publicly traded companies. See Exhibit 12. [i] Based on SourceHOV estimated pre-listing equity value of $645 million and debt as of March 31, 2017, of $1,055 million. [j] Based on SourceHOV estimated fully distributed equity value of $1,003 million and debt as of December 31, 2016, of $1,064 million. [k] As presented in Exhibit 13, highest unlevered beta of the guideline companies. [l] T he unrounded pretax weighted average cost of capital ranged from 11.455 percent to 13.376 percent. Sources: As indicated above. APPENDIX 3

EXHIBIT 1 SOURCEHOV HOLDINGS, INC. VALUATION SUMMARY AS OF JULY 12, 2017

Indicated Business Enterprise Weighted Exhibit Value Relative Value Valuation Method Reference $000 Emphasis $000

Management Projections—Discounted Cash Flow Method 8a 1,911,000 0% -

Management Projections—Capital Cash Flow Method 8b 1,841,000 0% -

Lender Model Projections—Discounted Cash Flow Method 10a 1,715,000 50% 857,500

Lender Model Projections—Capital Cash Flow Method 10b 1,751,000 50% 875,500

Market Approach—Guideline Publicly T raded Company Method 15 2,074,000 0% - 100% Business Enterprise Value before Adjustments 1,733,000

Plus: Cash and Cash Equivalents [a] 2 15,916 Plus: Value of Net Operating Loss Carryforwards 16b 50,381 Less: Interest-Bearing Debt 2 (1,055,115) Less: After-T ax Pension Liability [b] 2 (18,026) Less: Long-T erm T ax Liability 2 (3,063) Less: Other Long-T erm Liabilities less Other Assets 2 (1,239) (1,011,146)

Fair Value of Equity 721,854

Fair Value of Equity 721,854 T otal Common Shares Outstanding (000) [c] 157.249 Per-Share Fair Value of Equity ($, rounded) 4,591

Definitions are presented in Schedule A. [a] Equates to all unrestricted cash and cash equivalents. [b] Estimated as the pension liability of $28.612 million less income taxes at a rate of 37 percent. [c] T otal shares outstanding as of June 30, 2017, per Exela T echnologies, Inc., SEC Form 8-K/A filed August 9, 2017. SourceHOV had 157,243 shares outstanding as of March 31, 2017, per the Notice, F-76. Sources: As indicated above and Willamette Management Associates estimates and calculations.