Tribune Company v.

PRECEDENTIAL

UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT

No. 18-2909

In re: TRIBUNE COMPANY, et al., Debtors

DELAWARE TRUST COMPANY, as successor indenture trustee for certain series of Senior Notes and DEUTSCHE BANK TRUST COMPANY AMERICAS, solely in its capacity as successor Indenture Trustee for certain series of Senior Notes, Appellants

Appeal from the United States District Court for the District of Delaware (D.C. Civil Action Nos. 1-12-mc-00108, 1-12-cv- 00128/01072/1073/1100/1106) District Judge: Honorable Gregory M. Sleet

Argued November 12, 2019 Before: AMBRO, KRAUSE, and BIBAS, Circuit Judges

(Opinion filed: August 26, 2020)

Roy T. Englert, Jr. (Argued) Matthew M. Madden Mark T. Stancil Robbins Russell Englert Orseck Untereiner & Sauber 2000 K Street, N.W., 4th Floor Washington, DC 20006

Counsel for Appellant Delaware Trust Company

David J. Adler McCarter & English 825 Eighth Avenue Worldwide Plaza, 31st Floor New York, NY 10019

Counsel for Appellant Deutsche Bank Trust Co Americas, as Successor Indenture Trustee

Kenneth P. Kansa Sidley Austin One South Dearborn Street Chicago, IL 60603

James O. Johnston (Argued) Jones Day 555 South Flower Street, 50th Floor Los Angeles, CA 90071

J. Kate Stickles Cole Schotz 500 Delaware Avenue, Suite 1410

Wilmington, DE 19801

Counsel for Appellee Tribune Co.

Adam Hiller Hiller & Arban 1500 North French Street 2nd Floor Wilmington, DE 19801

Jay Teitelbaum (Argued) Teitelbaum Law Group 1 Barker Avenue, Third Floor White Plains, NY 10601

Counsel for Appellee TM Retirees

OPINION OF THE COURT

AMBRO, Circuit Judge

Many of the contentious battles in bankruptcy involve the allocation of distributions among similarly situated creditors. We have such a battle here, where certain creditors of the Tribune Company, called the “Senior Noteholders,” claim Tribune’s plan of reorganization (the “Plan”) misapplies their rights under the Bankruptcy Code by not according them the full benefit of their subordination agreements with other Tribune creditors. The Bankruptcy Court confirmed the Plan over the Senior Noteholders’ dissenting votes. In bankruptcy parlance, they were “crammed down.”

The provision in play was 11 U.S.C. § 1129(b)(1), which provides (with explanatory annotations) as follows:

Notwithstanding section 510(a) of this title [making subordination agreements enforceable in bankruptcy to the extent they would be in non- bankruptcy law], if all of the applicable requirements of subsection (a) of this section [1129] other than paragraph (8) [for our purposes, this paragraph requires that each class of claims has accepted the plan] are met with respect to a plan, the court, on request of the proponent of the plan, shall confirm the plan notwithstanding the requirements of such paragraph [8] if the plan does not discriminate unfairly, and is fair and equitable, with respect to each class of claims or interests that is impaired under, and has not accepted, the plan.

To unpack terms of art, “discriminate unfairly” is a horizontal comparative assessment applied to similarly situated creditors (here unsecured creditors) where a subset of those creditors is classified separately, does not accept the plan, and claims inequitable treatment under it. Bruce A. Markell, A New Perspective on Unfair Discrimination in Chapter 11, 72 Am. Bankr. L.J. 227, 227–28 (1998). “[F]air and equitable” (a redundant term) should be pictured vertically, as it “regulates priority among classes of creditors having higher and lower priorities,” id. at 228. For example, secured creditors are a higher priority for payment than unsecured creditors. For the sake of completeness, “impaired” means a creditor whose rights under a plan are altered (obviously adversely). 11 U.S.C. § 1124(1). In our case, the Senior Noteholders were assigned their own class (1E) of unsecured creditors in Tribune’s Plan. When

they did not accept the Plan but other classes did, the Bankruptcy Court confirmed it under the cramdown provision, and they became bound by it. They appeal to us, contending that “[n]otwithstanding” in § 1129(b)(1) entitles them to their full recovery from the strict enforcement of the subordination agreements, and, in any event, the Plan’s proposed distributions were unfairly discriminatory in favor of another unsecured class (1F) that shared in the subordinated sums.

We agree with the Bankruptcy and District Courts that the text of § 1129(b)(1) supplants strict enforcement of subordination agreements. Instead, when cramdown plans play with subordinated sums, the comparison of similarly situated creditors is tested through a more flexible unfair- discrimination standard. Applying that standard here, we affirm the result determined by those Courts. The facts that follow, as typical in the transactional world, are complicated, and so at times is the legal analysis. Frame them, however, in the context set out above.

I. FACTS AND PROCEDURAL HISTORY Prior to its bankruptcy, Tribune was the largest media conglomerate in the country, reaching 80% of American households each year. It owned the Chicago Tribune and the Los Angeles Times, as well as many regional newspapers, television and radio stations.

The Company’s 2008 bankruptcy followed on the heels of its failed leveraged buyout (“LBO”), 1 which left it with

1 An LBO is typically a transaction where the purchaser acquires an entity with borrowings secured by the assets of that entity. The LBO here left Tribune saddled with debt, while the

almost $13 billion of debt and a complex capital structure. In 2012, after years of contentious proceedings, the Bankruptcy Court confirmed the Plan over the dissenting votes of the Senior Noteholders. 2 Initial distributions under the Plan were made at the end of that year, as the Bankruptcy Court rejected the Senior Noteholders’ request for a stay. This is the second time Tribune’s dissenting creditors are before us. In In re Tribune Media Co., 799 F.3d 272 (3d Cir. 2015) (“Tribune I”), we reversed in part the District Court’s determination that the Senior Noteholders’ claims were, because the Plan’s distributions had already occurred, equitably moot and sent them back for further proceedings on the merits. On remand, that Court affirmed the Bankruptcy Court’s confirmation order over the Senior Noteholders’ objections. In re Tribune Media Co., 587 B.R. 606 (D. Del. 2018). We review their appeal here. 1. Overview of Tribune’s creditors

Prior to the 2007 LBO, Tribune had a market capitalization of $8 billion and $5 billion in debt, which had been amassed over decades. The Senior Noteholders loaned to Tribune unsecured debt between 1992 and 2005 (the “Senior

purchaser put none of its own money at risk. In re Tribune Media Co., 799 F.3d 272, 275 (3d Cir. 2015). 2 Appellants technically are the Delaware Trust Company and Deutsche Bank Trust Company Americas (the “Trustees”), which, in their capacities as successor indenture trustees, represent the interest of the Senior Noteholders. For simplicity, we refer throughout this opinion to the Senior Noteholders and the Trustees collectively as the Senior Noteholders.

Notes”). Covenants in the Senior Notes’ indentures require that they are paid before any other debt incurred by the company. When Tribune filed for reorganization, the outstanding amount due on those Notes was $1.283 billion.3 In 1999, Tribune also issued $1.256 billion of unsecured exchangeable subordinated debentures (the “PHONES Notes”). Their indenture provided that they are subordinate in payment to all “‘Senior Indebtedness’ of Tribune,” which included the Senior Notes. In re Tribune Co., 464 B.R. 126, 138 (Bankr. D. Del. 2011) (the “2011 Opinion”). At the time of its bankruptcy, the outstanding principal on the PHONES Notes was $759 million.

The LBO added approximately $8 billion of debt to Tribune’s capital structure. As part of the merger financing, Tribune issued $225 million of unsecured debt (the “EGI Notes”). That indenture also subordinated repayment to “Senior Obligations.” 4

Also among the billions of dollars of Tribune’s debt are an unsecured $150.9 million “Swap Claim” (which is tied to the termination of an interest rate swap agreement to offset the interest rate exposure from the LBO); $105 million of

3 For consistency, we use the claim amounts from the Stipulated Recovery Percentage Table set out below. 4 The Bankruptcy Court determined that the definition of “Senior Obligations” in the EGI Notes was broader than the definition of “Senior Indebtedness” in the PHONES Notes. As any creditor determined to be Senior Indebtedness would also qualify as a Senior Obligation, for simplicity we refer to both provisions collectively as the Senior Obligations.

unsecured claims by Tribune Media Retirees (the “Retirees”); and $8.8 million of unsecured claims by trade and miscellaneous creditors (the “Trade Creditors”). From Tribune’s perspective, these unsecured creditors—the holders of Senior Notes, the PHONES and EGI Notes, and Swap Claim, plus the Retirees and the Trade Creditors—are of equal priority. But the subordination provisions in the PHONES and EGI Notes’ indentures—which were entered outside Tribune’s bankruptcy—limit their repayment until all Senior Obligations are paid in full. Thus, whether a creditor should benefit from these subordination provisions depends on its claim qualifying as a Senior Obligation.

2. The Bankruptcy Court Proceedings

During the bankruptcy proceedings, several groups of stakeholders proposed plans to reorganize Tribune’s debt. The Bankruptcy Court’s 2011 Opinion on competing plans coalesced support around the Plan sponsored by Tribune, its Official Committee of Unsecured Creditors, and certain lenders. It settled many of the Creditors’ Committee’s claims against the LBO lenders, directors and officers of the old Tribune, real estate investor Samuel Zell (who orchestrated the LBO 5), and others, for $369 million paid to Tribune’s estate.

The Plan organized Tribune’s unsecured creditors into distinct classes. The Senior Noteholders, which comprise

5 Mr. Zell called it “the deal from hell.” Michael Arndt and Emily Thornton, Sam Zell Speaks His Mind, Bloomberg Bus. News (July 30, 2008, 12:00 AM), https://www.bloomberg.com/news/articles/2008-07-29/sam- zell-speaks-his-mind.

Class 1E, argue that the Plan favored Class 1F, which is made up of the Swap Claim, the Retirees, and the Trade Creditors (collectively this class includes over 700 unsecured creditors). It paid both Class 1E and Class 1F creditors 33.6% of their outstanding claims from the initial distributions under the Plan. In re Tribune Co., 472 B.R. 223, 237 & n.17 (Bankr. D. Del. 2012) (the “Allocation Opinion”). These payments included monies from the subordination of the PHONES and EGI Notes.

The Senior Noteholders objected to the Plan. They argued that it allocated more than $30 million of their recovery from the subordinated PHONES and EGI Notes to Class 1F when only the Senior Noteholders in Class 1E qualified as Senior Obligations, and thus they alone should benefit from those subordination agreements. Specifically, they asserted that the Plan violated the Bankruptcy Code’s standards for confirmation because it did not fully enforce the subordination provisions per § 510(a). In the alternative, they claimed that the allocation of subordination payments to Class 1F unfairly discriminated against their Class 1E.

To resolve these and other intercreditor claims, the Bankruptcy Court established an allocation dispute process, which called for extensive briefing followed by a hearing on the Senior Noteholders’ objections. As a starting point, the creditors stipulated to their initial recovery percentages under the Plan depending on which creditors ultimately qualified as Senior Obligations under the PHONES and EGI indentures. It was undisputed that the Class 1E Senior Notes were Senior Obligations and thus entitled to payments from the subordinated creditors. A principal dispute concerned whether other creditors also qualified; if so, they would recover additional payment from the subordination provisions in the PHONES and EGI Notes. Thus the Senior Noteholders stood to increase their recovery by limiting which, if any, other creditors qualified as Senior Obligations. The chart below

Allocation Opinion, 472 B.R. at 238 (describing recovery under the Plan); J.A. 327.6

In its Allocation Opinion, the Bankruptcy Court determined that § 1129(b)(1) does not require that the subordination agreements be strictly enforced for a plan to be confirmed. It also rejected the Senior Noteholders’ unfair- discrimination claim, explaining that it failed even if the Court “assume[d] (without deciding) that[ ] none of the [Class 1F creditors] are Senior [Obligations] [ ] and are not entitled to the benefit of either subordination agreement.” 472 B.R at 238. However, it resolved in a footnote that the Swap Claim, which comprised 57% of Class 1F’s aggregate claims, was a Senior Obligation and thus should benefit from the partial enforcement of the subordination agreements.7 Id. at 238–39

6 The parties stipulated that the Swap Claim would recover under the Plan 36.0% of its claim on J.A. 327. However, given descriptions of the Plan in the Bankruptcy Court’s Allocation Opinion and the parties’ briefs, this looks to be incorrect, see 472 B.R. at 238; Trustees’ Br. 39; Tribune’s Br. 3. All Class 1E and 1F creditors, including the Swap Claim, appear to have recovered under the Plan 33.6% of their respective claims. 7 The Senior Noteholders appealed this categorization of the Swap Claim to the District Court, which affirmed the Bankruptcy Court’s determination. 587 B.R. at 618–19. They do not appeal that ruling to us.

n.19. Notably, the Court did not decide whether the Retirees’ claim qualified as a Senior Obligation.8 Id.

The Court’s footnote stating that the Swap Claim qualified as a Senior Obligation reduced the Senior Noteholders’ unfair-discrimination claim of approximately $30 million by over $17 million, thereby leaving roughly $13 million in dispute (a small sum relative to their overall $1.283 billion claim). 9 To put this in a picture, they ask us to reallocate payments to reflect the fourth row of the Stipulated Recovery Percentage Table (“If Class 1E and the Swap Claim benefit from subordination”) rather than the first row, increasing initial distributions toward their claim recovery from 33.6% to 34.5%. 10

8 In their Brief to us, the Retirees contend that, should we not affirm and remand the case, this issue remains for determination. Retirees’ Br. 14–15. Yet, they state two pages later that they “are not subject to the subordination agreements.” Id. at 17. Though this is ambiguous, we affirm the Bankruptcy and District Courts here, and thus we do not need to resolve this issue or remand for its resolution. 9 Aligning the Swap Claim with the same benefits accorded the Senior Noteholders begs us to ask whether it was misclassified by being assigned to Class 1F. That is a good question, but no one teed it up for resolution on appeal. Thus it is not before us. 10 We, like the parties, stick with initial Plan distributions that do not include hypothetical future recoveries that may be gained by a litigation trust under the Plan.

3. The Senior Noteholders’ appeal

The Senior Noteholders appealed the Plan’s confirmation to the District Court, renewing their argument that it violated § 1129(b)(1) by not exactly enforcing the subordination agreements. In the alternative, they challenged the particulars of the Bankruptcy Court’s unfair-discrimination analysis, arguing that it erred, first, by including the recoveries due them from the subordination agreements in their Plan distributions, and second, by failing to compare their recovery under the Plan to that of Class 1F.

While the appeal was pending, Tribune consummated the Plan by making the distributions called for in it. We nonetheless held that the Senior Noteholders’ arguments before us now could proceed. 11 Tribune 1, 799 F.3d at 283– 84. They still came up short on remand, and appeal to us again.

II. JURISDICTION AND STANDARD OF REVIEW

We have jurisdiction over this appeal under 28 U.S.C. §§ 158(d) and 1291. We exercise plenary review of the District Court’s conclusions of law, including its interpretation of the Bankruptcy Code. In re Goody’s Family Clothing Inc., 610 F.3d 812, 816 (3d Cir. 2010). “Because the District Court sat as an appellate court to review the Bankruptcy Court, we review the [latter’s] legal determinations de novo, its factual

11 In Tribune I we explained that, as payments under the Plan have already been distributed, if the Senior Noteholders (referred to there as the Trustees) are successful, their relevant relief, inter alia, is an increase in recovery payments through disgorgement “ordered against those Class 1F holders who have received more than their fair share.” 799 F.3d at 282–83.

findings for clear error, and its exercises of discretion for abuse thereof.” Id.

III. DISCUSSION Cramdown plans are an antidote to one or more classes of claims holding up confirmation of an otherwise consensual plan. See generally Kenneth N. Klee, All You Ever Wanted to Know About Cram Down Under the New Bankruptcy Code, 53 Am. Bankr. L.J. 133 (1979). The cramdown provision in 11 U.S.C. § 1129(b)(1) waives § 1129(a)(8)’s mandate that all classes either vote to accept the plan or recover their debt in full under it. Yet the provision also affords unique safeguards: the fair-and-equitable and unfair-discrimination standards. While we have spent considerable time outlining the boundaries of what is fair and equitable (as noted below, not at issue here), see, e.g., In re Armstrong World Indus., Inc., 432 F.3d 507, 512–13 (3d Cir. 2005); In re PWS Holding Corp., 228 F.3d 224, 237–42 (3d Cir. 2000); see also Bank of Am. Nat’l Tr. and Sav. Ass’n v. 203 N. LaSalle St. P’ship, 526 U.S. 434, 444–49 (1999), the unfair-discrimination standard has received little analysis.

A. Subsection 1129(b)(1) does not require subordination agreements to be enforced strictly.

The Senior Noteholders first contend that the Bankruptcy Court should not have confirmed the Plan because it does not enforce strictly the PHONES and EGI Notes’ subordination agreements under Code § 510(a). Thus we must determine the effect of 11 U.S.C. § 1129(b)(1)’s explicit reference to the earlier provision. Put another way, how does the cramdown provision’s authority interact with intercreditor subordination agreements?

The Senior Noteholders argue § 1129(b)(1) “explain[s] that the cramdown safeguards must be applied ‘[n]otwithstanding section 510(a),’” Trustees’ Br. 18 (quoting 11 U.S.C. § 1129(b)(1)), meaning that subordination agreements cannot be interfered with in cramdown cases. Both the Bankruptcy and District Courts rejected this argument, explaining that it is at odds with the plain meaning of § 1129(b)(1). We agree.

1. The text of § 1129(b)(1)

Section 510(a) provides that “[a] subordination agreement is enforceable in [bankruptcy] to the same extent that such agreement is enforceable under applicable nonbankruptcy law.” But § 1129(b)(1) states that a nonconsensual plan may be confirmed “[n]otwithstanding section 510(a).” We have previously defined the phrase “notwithstanding” in the bankruptcy context to mean “‘in spite of’ or ‘without prevention or obstruction from or by.’” Goody’s, 610 F.3d at 817 (quoting Webster’s Third Int’l Dictionary 1545 (1971)); see also In re Federal-Mogul Global Inc., 684 F.3d 355, 369 (3d Cir. 2012) (reading the lead-in to Bankruptcy Code § 1123(a)—“notwithstanding any otherwise applicable non-bankruptcy law”—to mean that what follows in subsection (a) displaces conflicting state nonbankruptcy law). Although these cases interpret different sections of the Code, their analysis applies equally to § 1129(b)(1) because, “[p]resumptively, identical words used in different parts of the same act are intended to have the same meaning.” U.S. Nat’l Bank of Or. v. Indep. Ins. Agents of Am., Inc., 508 U.S. 439, 460 (1993) (internal quotation marks and citation omitted). Further, as we explained in Federal-Mogul, “[w]hen a federal law contains an express preemption clause, we focus on the plain wording of the clause, which necessarily contains the best

evidence of Congress’ preemptive intent.” 684 F.3d at 369 (internal quotation marks omitted).

Applying the lessons of Goody’s and Federal-Mogul here, § 1129(b)(1) overrides § 510(a) because that is the plain meaning of “[n]otwithstanding.” Thus our holding becomes simple: Despite the rights conferred by § 510(a), “if all of the applicable requirements of subsection (a) of this section [1129] . . . are met with respect to a plan, the court . . . shall confirm the plan . . . if [it] does not discriminate unfairly, and is fair and equitable,” for each impaired class that does not accept the plan.

2. The purpose of § 1129(b)(1)

Section 1129(b)(1)’s purpose affirms this analysis. The provision allows a court to confirm a plan if it protects the interests of a dissenting class, here the Class 1E Senior Noteholders. Those interests are primarily preserved by the fair-and-equitable test (not in play here, as our dispute involves only unsecured creditors) and the unfair-discrimination test, the latter protecting the relative payments to same-rank creditor classes whose recovery has been affected, inter alia, by intercreditor subordination agreements. 11 U.S.C. § 1129(b)(1); Armstrong World Indus., 432 F.3d at 512; In re Aztec Co., 107 B.R. 585, 589 (Bankr. M.D. Tenn. 1989).

Both § 510(a) and the cramdown provision’s unfair- discrimination test are concerned with distributions among creditors. The first is by agreement, while the second tests, among other things, whether involuntary reallocations of subordinated sums under a plan unfairly discriminate against the dissenting class. Only one can supersede, and that is the cramdown provision. It provides the flexibility to negotiate a confirmable plan even when decades of accumulated debt and private ordering of payment priority have led to a complex web

of intercreditor rights. It also attempts to ensure that debtors and courts do not have carte blanche to disregard pre- bankruptcy contractual arrangements, while leaving play in the joints. 12 To date, we are aware of only one court that has spoken in a published opinion to the effect of § 1129(b)(1)’s notwithstanding proviso. See In re TCI 2 Holdings, 428 B.R. 117, 141 (Bankr. D.N.J. 2010) (“The phrase ‘[n]otwithstanding section 510(a) of this title’ removes section 510(a) from the scope of 1129([b])(1)[.]”). That decision aligns with ours here.

The House Report for the Bankruptcy Code also supports this interpretation. Though the Report’s discussion of unfair discrimination is quite brief, its examples exclusively involve the relative treatment of like-kind creditors affected by subordination agreements. See H.R. Rep. No. 95-595, at 416– 17 (1977), reprinted in 1978 U.S.C.C.A.N. 5963, 6372–73 (the

12 Further, as illustrated by the facts of this case, diverse groups—including former employees, trade creditors, and bondholders—typically make up a debtor’s unsecured creditors. Agreements reached regarding repayment outside of the Chapter 11 context may no longer serve the creditors or the debtor “when each [creditor] does not contribute proportionately to [the reorganized debtor’s] creation and maintenance.” 7 Collier on Bankruptcy ¶ 1129.03 (16th ed. 2020). Thus the flexibility provided by the unfair- discrimination test may be a welcome relief.

“House Report”). They rely on that discrimination principle, and not on § 510(a), to enforce subordination agreements. 13 Id.

To save their reading of § 1129(b)(1), the Senior Noteholders cite the 1995 article by Professor Kenneth Klee (one of the principal drafters of the Bankruptcy Code and, coincidentally, an examiner appointed by the Bankruptcy Court in this case) recommending that Congress delete the reference to § 510(a) in § 1129(b) to prevent the “anomalous result of overriding § 510(a) and eliminating the enforcement of subordination agreements in cases in which the class rejects the plan.” Kenneth N. Klee, Adjusting Chapter 11: Fine Tuning the Plan Process, 69 Am. Bankr. L.J. 551, 561 (1995). They urge that we adopt this recommendation to “avoid that bizarre result.” Trustees’ Br. 29. Their problem is that Professor Klee’s recommendation to Congress is not evidence of the legislature’s intent to favor § 510(a) (indeed, by inference he acknowledges that the cramdown provision prevails). As Congress has not changed the law to reflect

13 The Report’s discussion of unfair discrimination has received criticism over the years. See In re Armstrong World Indus., Inc., 348 B.R. 111, 121 (D. Del. 2006) (“Unfair discrimination is not defined in the Bankruptcy Code, nor does the statute’s legislative history provide guidance as to its interpretation.”); Markell, A New Perspective, supra, at 237– 38 (describing the examples of unfair discrimination in the legislative history as “roundabout, almost otiose”). Nevertheless, Professor (and former Judge) Markell relies on this legislative history to guide in part his analysis of the provision, ultimately arriving at the test used by the Bankruptcy Court here. Id. at 236–38, 249–50. We too find some guidance in the Report, even as we acknowledge that it is no model of clarity on our issue.

Professor Klee’s proposal, we rely on the plain language of § 1129(b)(1).

Hence we affirm the holding of the Bankruptcy and District Courts that subordination agreements need not be strictly enforced for a court to confirm a cramdown plan. They correctly evaluated the Senior Noteholders’ claim under the unfair-discrimination test rather than a rigid application of § 510(a).

B. The Plan’s allocation of a small portion of subordinated sums to Class 1F creditors does not unfairly discriminate against the Senior Noteholders. As we resolved the first issue in favor of Tribune, we turn to the Senior Noteholders’ alternative argument: The Plan unfairly discriminates against them. The Bankruptcy Court’s unfair-discrimination analysis compared Class 1E’s initial distribution recovery percentage under the Plan—33.6%—to its recovery were there strict enforcement of the subordination agreements—34.5%—and determined that 0.9% was not a material difference in recovery. Allocation Opinion, 472 B.R. at 243 n.21. Thus it held there was no unfair discrimination to bar Plan confirmation.

The Senior Noteholders allege two flaws in that analysis. First, they claim the Court failed to compare only recoveries from the Tribune estate. That is, it should have compared Class 1E’s and Class 1F’s Plan recoveries as if no subordination agreements were in effect. As the parties stipulated in the Stipulated Percentage Recovery Table that the Senior Noteholders in Class 1E recovered only 21.9% of their claims absent subordination payments, this meant, they argue, that the rest of their 33.6% recovery under the Plan came from

the subordinated creditors and should be excluded from the unfair-discrimination analysis.

Second, they claim that the Court’s refusal to compare their Class IE percentage recovery with Class 1F’s percentage recovery, and its decision to compare instead only Class 1E’s recovery under the Plan with its recovery had the subordination agreements been fully enforced, were incorrect. Once these errors are addressed, they assert the difference between Class 1E’s recovery under the Plan absent subordination (21.9%), and the Plan recovery of Class 1F’s non-Swap Claim creditors including subordination benefits (33.6%), is material, evidencing unfair discrimination that should have prevented the Plan’s confirmation.

The Bankruptcy Code does not define unfair discrimination. It “is something of an orphan in Chapter 11 reorganization practice. . . . [J]ust what suffices to avoid unfair discrimination is uncertain.” Markell, A New Perspective, supra, at 227. “Generally speaking, this standard ensures that a dissenting class will receive relative value equal to the value given to all other similarly situated classes.” In re Armstrong World Indus., Inc., 348 B.R. 111, 121 (D. Del. 2006) (quoting In re Johns-Manville Corp., 68 B.R. 618, 636 (Bankr. S.D.N.Y. 1986)). Since unfair discrimination’s inclusion in the Bankruptcy Code (it appeared for a short time in the 1930s in revisions to the Bankruptcy Act of 1898), courts have relied primarily on one of four tests to determine what unfairness means and, in some of those tests, whether, if a presumption of unfairness exists, it can be rebutted. See generally Denise R. Polivy, Unfair Discrimination in Chapter 11: A Comprehensive Compilation of Current Case Law, 72 Am. Bankr. L.J. 191, 196–208 (1998) (collecting and discussing cases applying the various tests).

The “mechanical” test prohibits all discrimination, that is, it requires that similarly situated creditors’ recoveries be 100% pro rata. See In re Greystone III Joint Venture, 102 B.R. 560, 571–72 (Bankr. W.D. Tex. 1989) (reasoning that paying the trade creditors a higher percentage of their claims than other unsecured creditors would constitute unfair discrimination), rev’d on other grounds, 995 F.2d 1274 (5th Cir. 1992). The “restrictive” approach narrowly defines unfair discrimination such that, “[i]n the absence of subordination, . . . no disparate treatment of similarly situated creators would qualify,” Polivy, supra, at 200. Both tests have support in the House Report: it noted “there is no unfair discrimination as long as the total consideration given all other classes of equal rank does not exceed the amount that would result from an exact aliquot distribution,” House Report, supra, at 416, and the examples given involved subordinated creditors, id. at 416– 17; see also In re Acequia, Inc., 787 F.2d 1352, 1364 & n.18 (9th Cir. 1986). Neither of these tests appears to be widely adopted, however. See Polivy, supra, at 200–201 (collecting cases); cf. Aztec, 107 B.R. at 588–89 (rejecting both the restrictive and mechanical tests).

The “broad” approach is generally applied as a four- factor test that originated in the Chapter 13 case In re Kovich, 4 B.R. 403, 407 (Bankr. W.D. Mich. 1980). To determine whether the plan unfairly discriminates, the test considers whether: (1) a reasonable basis for discrimination exists; (2) the debtor cannot consummate its plan without discrimination; (3) the discrimination is imposed in good faith; and (4) the degree of discrimination is directly proportional to its rationale. 14 See, e.g., Aztec, 107 B.R. at 590 (laying out and

14 Some courts, finding the factors redundant, pared down the test to ask whether there is “a rational or legitimate

applying the test). Although this analysis has received criticism, see, e.g., In re Dow Corning Corp., 244 B.R. 696, 702 (Bankr. E.D. Mich. 1999); In re Brown, 152 B.R. 232, 235 (Bankr. N.D. Ill. 1993), rev’d on other grounds sub nom. McCullough v. Brown, 162 B.R. 506 (N.D. Ill. 1993); Markell, A New Perspective, supra, at 242–48, 254–55, it has been applied often, see Polivy, supra, at 203 n.102 (collecting cases applying the broad test). 15

In response to criticisms of these tests, Professor Bruce Markell proposed the “rebuttable presumption” test, which was applied by the Bankruptcy Court in this case, 472 B.R. at 242.16

basis for discrimination” and if it is “necessary for the reorganization.” In re Dow Corning Corp., 244 B.R. 696, 701 (Bankr. E.D. Mich. 1999) (citation omitted). 15 These critics reject the application of the broad test because, among other things, it was developed for Chapter 13 cases, which require the protection of all creditors, as they do not have voting rights, and provides amorphous limits on discrimination. Dow Corning, 244 B.R. at 702. The rebuttable presumption test, dealt with below and developed for the Chapter 11 context, is tailored to the specific circumstances of cramdown, where only the interest of the dissenting class is at issue. Id. It also eschews concepts such as reasonableness, whether a plan can be confirmed without discrimination, and good faith. 16 This holding is not before us, as the Senior Noteholders endorse it, Trustees’ Br. 37. Tribune accepts it, Tribune’s Br. 17, as do we. See In re Tribune Media Co., 587 B.R. at 617–18; see also In re Nuverra Envtl. Sols., Inc., 590 B.R. 75, 90 (D. Del. 2018).

A rebuttable presumption of unfair discrimination exists when there is

(1) a dissenting class; (2) another class of the same priority; and (3) a difference in the plan’s treatment of the two classes that results in either (a) a materially lower percentage recovery for the dissenting class (measured in terms of the net present value of all payments), or (b) regardless of percentage recovery, an allocation under the plan of materially greater risk to the dissenting class in connection with its proposed distribution.

Markell, A New Perspective, supra, at 228, 249; see also Dow Corning, 244 B.R. at 702.

Under this test, a presumption of unfair discrimination may be overcome if the court finds that

a lower recovery for the dissenting class is consistent with the results that would obtain outside of bankruptcy, or that a greater recovery for the other class is offset by contributions from that class to the reorganization. The presumption of unfairness based on differing risks may be overcome by a showing that the risks are allocated in a manner consistent with the prebankruptcy expectations of the parties. Markell, A New Perspective, supra, at 228; cf. Comm. on Bankr. and Corp. Reorg. of the Ass’n of the Bar of the City of New York, Making the Test for Unfair Discrimination More “Fair”: A Proposal, 58 Bus. Law. 83, 106–07 (2002) (proposing amendments to this test that narrow the

circumstances where it is appropriate to overcome the presumption of unfair discrimination).

The Senior Noteholders’ unfair-discrimination claim involves mixed questions of law and facts. A bankruptcy court’s initial determination of which test to use is reviewed as “a legal conclusion without the slightest deference.” U.S. Bank Nat’l Ass’n ex rel. CWCapital Asset Mgmt. LLC v. Vill. at Lakeridge LLC, 138 S. Ct. 960, 965 (2018). Reviewing the Bankruptcy Court’s choice of legal test de novo, we agree that it was appropriate in these circumstances to take a pragmatic approach to measure the Plan’s discrimination. Thereafter, Village at Lakeridge asks whether applying the law to the facts “entails primarily legal or factual work.” Id. at 967. This inquiry sounds simple, but often it will depend on how a court approaches its analysis. Our approach here sets up a fact- specific question: How does discrimination affect Class 1E’s actual recovery? Thus we review the application of legal precepts to the facts in this instance for clear error. (Even were we instead to apply de novo review, the result would not change).

1. Principles framing the “unfair-discrimination” standard We distill the following principles from the various unfair-discrimination analyses. First, though § 1129(b)(1)’s legislative history speaks of discrimination as unfair once there is breached a pure pro rata division of plan distributions among like-priority creditors, that runs counter to the text. See Aztec, 107 B.R. at 588–89. “Discriminate unfairly” is simple and direct: you can treat differently (discriminate) but not so much as to be unfair. There is, as is typical in reorganizations, a need for flexibility

over precision. The test becomes one of reason circumscribed so as not to run rampant over creditors’ rights.

This reading is also consistent with our holding in Section A above. A subordination agreement does not need to be enforced to the letter in the case of a cramdown, and subordinated amounts may be allocated to other classes not entitled outside bankruptcy to benefit from subordination agreements as long as that allocation is not presumptively unfair (and, if so, the presumption is not rebutted). Second, the cramdown provision’s text also makes plain that unfair discrimination applies only to classes of creditors (not the individual creditors that comprise them), and then only to classes that dissent. Thus a disapproving creditor within a class that approves a plan cannot claim unfair discrimination, and the standard does not “apply directly with respect to other classes unless they too have dissented.” Klee, Cram Down, supra, at 141 n.67.

Third, unfair discrimination is determined from the perspective of the dissenting class. House Report, supra, at 416–17. What this means, however, is subject to interpretation. Courts and commentators nearly always consider this a comparison between the allegedly preferred class and the dissenting class. See, e.g., In re Greate Bay Hotel & Casino, Inc., 251 B.R. 213, 231 (Bankr. D.N.J. 2000) (collecting cases comparing the recovery of the dissenting class to that of the preferred class or classes); Klee, Cram Down, supra, at 142 (“[I]f the plan protects the legal rights of a dissenting class in a manner consistent with the treatment of other classes . . . , then the plan does not discriminate unfairly with respect to the dissenting class.”). However, as was done in this case, a court may in certain circumstances consider the difference between what the dissenting class argues it is entitled to recover and what it actually received under the plan.

In other words, a comparison between the recovery of the preferred class and the dissenting class is by far the preferred but not always the only acceptable approach. Other measures that allow courts to assess the magnitude of harm to the dissenting class may also be appropriate in some cases.

Fourth, the need for classes to be aligned correctly is a precursor to an effective assessment. A typical refrain in bankruptcy is that many plan disputes in § 1129 begin as misclassifications under § 1122. 17 See, e.g., In re Woodbridge Assocs., 19 F.3d 312, 317–321 (7th Cir. 1994) (considering both the dissenting creditors’ misclassification and unfair- discrimination claims); In re Unbreakable Nation Co., 437 B.R. 189, 200–202 (Bankr. E.D. Pa. 2010) (same); Greate Bay Hotel, 251 B.R. at 223–32 (same); see generally G. Eric Brunstad, Jr. and Mike Sigal, Competitive Choice Theory and the Unresolved Doctrines of Classification and Unfair Discrimination in Business Reorganizations Under the Bankruptcy Code, 55 Bus. Law. 1, 72–73 & n.289, 78 (1999) (discussing the need to enforce subordination principles at classification to avoid “perverse incentives” and unfair- discrimination claims).

17 Technically a plan objection, if made, would be under § 1129(a) by claiming that the plan did not comply with the classification requirements of § 1122(a), which requires that “substantially similar” claims be placed in the same class. Markell, A New Perspective, supra, at 238 n.56. Where subordination agreements are in play, a gateway to unfair- discrimination determinations is to separate those whose claims benefit from the agreements from those who do not. Placing a subordination beneficiary with a non-beneficiary in a single class bleeds over clear analysis.

Fifth, courts should resolve how a plan proposes to pay each creditor’s recovery “measured in terms of the net present value of all payments” or the “allocation . . . of materially greater risk . . . in connection with its proposed distribution.” Markell, A New Perspective, supra, at 228. This allows future distributions to be made reasonably equivalent to the actual value distributed at the time of the unfair-discrimination comparison.

Sixth, in making an unfair-discrimination determination, start by adding up all proposed plan distributions from the debtor’s estate and divide by the number of creditors sharing the same priority. This provides a pro rata baseline. Then look at what actually happens if the plan is implemented. Where there are no subordination agreements involved, the analysis is simple: look at the difference between the recovery percentage under the plan of a preferred class and that of a dissenting class. Where subordination agreements are involved, courts should resolve in the first instance which creditors are entitled to benefit from those agreements. They should make their comparisons after including subordinated sums in the plan distributions, for what may be in dispute often is the amount the dissenting class would be entitled under full enforcement of § 510(a) but did not get under the plan.

Seventh, to presume unfair discrimination, there must be “a ‘materially lower’ percentage recovery for the dissenting class or a ‘materially greater risk to the dissenting class in connection with its proposed distribution.’” Greate Bay Hotel, 251 B. R. at 229 (quoting Dow Corning, 244 B.R. at 702). The rebuttable presumption test intentionally leaves opaque what is, under the circumstances, “material.” Such line drawing has been left primarily to bankruptcy courts. See Bruce A. Markell, Slouching Toward Fairness: A Reply to the ABCNY’s Proposal on Unfair Discrimination, 58 Bus. Law. 109, 116 (2002) (“Congress has left the important area of nonconsensual

confirmation to the common law method of incremental decision-making.”). We too leave this for judicial development. Eighth, if courts find plans materially discriminate against the dissenting class and follow the rebuttable presumption test or some variation, that finding is by definition presumptive and can be rebutted. Though we could make general suggestions for what qualifies as an adequate rebuttal (e.g., contributions to the reorganization by a preferred class may rebut unfair discrimination), those determinations are for bankruptcy courts to decide initially. Id.

2. Application of the principles

To review, the Bankruptcy Court compared Class 1E’s recovery under the Plan (33.6%) to its recovery if it and the Swap Claim were the only creditors to benefit from the subordination agreements (34.5%). 472 B.R. at 243 n.21. The Senior Noteholders point out that typically a court will compare the recovery percentages of the dissenting and preferred classes and ask whether the difference in recovery, if any, is material. Trustees’ Br. 41. If that analysis had been applied here, the Court would have needed to resolve the relative priority of all the creditors in Classes 1E and 1F to determine which creditors qualified as Senior Obligations under the PHONE and EGI Notes before comparing the treatment of the two classes under the Plan.

Yet neither the text of 11 U.S.C. § 1129(b)(1) nor the rebuttable presumption test explicitly limits the unfair- discrimination analysis to only a class-to-class comparison. As the Bankruptcy Court noted, unfair discrimination requires that a court evaluate whether “there was a materially lower percentage recovery for the dissenting class.” 472 B.R. at 244 (internal quotation marks omitted) (quoting Greate Bay Hotel,

251 B.R. at 231). In cases where a class-to-class comparison is difficult—for instance, here 57% of Class 1F (the Swap Claim) is entitled to benefit from the subordination of the PHONES and EGI Notes, while the Trade Creditors (and perhaps the Retirees) are not—a court may opt to be pragmatic and look to the discrepancy between the dissenting class’s desired and actual recovery to gauge the degree of its different treatment. Either way the perspective remains that of the dissenting class.

The Senior Noteholders argue that the Court should have compared their recovery from the estate absent subordination (21.9%) to the Trade Creditors’ recovery under the Plan with the reallocated subordination payments (33.6%). To measure discrimination this way is to ignore that the Plan brought into the Tribune estate not only the subordinated sums distributed to non-beneficiaries of that subordination, but all payments from the subordinated creditors (and indeed it allocated the overwhelming majority of those sums to the Senior Noteholders and the Swap Claim).

In this context, the Bankruptcy Court did not necessarily err when it compared the Senior Noteholders’ desired recovery under the fourth row of the Stipulated Recovery Percentage Table (34.5%) to their actual recovery under the Plan (33.6%). To repeat, this is not the preferred way to test whether the allocation of subordinated amounts under a plan to initially non-benefitted creditors unfairly discriminates. It may, however, be an appropriate metric (or cross-check) given the circumstances of a case. This is such a case. Because the claims of the Retirees ($105M) and Trade Creditors ($8.8M) are so substantially smaller than the Senior Noteholders’ claims ($1.283B), the increases in the recovery percentage for the Retirees’ and Trade Creditors’ claims from reallocated subordinated

amounts result in only a minimal reduction of the recovery percentage for the Senior Noteholders. Specifically, the subordinated sums allocated to the Retirees and Trade Creditors comprised 11.7 percentage points toward their 33.6 percentage recovery, but only reduced the Senior Noteholders’ recovery by nine-tenths of a percent. Thus we agree with the Bankruptcy and District Courts that this difference in the Senior Noteholders’ recovery is not material. Although the Plan discriminates, it is not presumptively unfair when understood, as ruled above, that a cramdown plan may reallocate some of the subordinated sums.

As an aside, we note that the Bankruptcy Court looked to cases comparing the differences in the dissenting class and the preferred class recoveries as a baseline for its materiality determination. See Allocation Opinion, 472 B.R. at 243 (collecting cases). Because it adopted a different framework for its analysis than the courts it cited, id. at 242–43, it did not need to apply their metric for materiality. What constitutes a material difference in recovery when analyzing the effect of a plan on the dissenting class is a distinct and context-specific inquiry. We do not address the outer boundary of that inquiry here. Wherever it may lie, the nine-tenths of a percentage point difference in the Senior Noteholders’ recovery is, without a doubt, not material. * * * * * Unfair discrimination is rough justice. It exemplifies the Code’s tendency to replace stringent requirements with more flexible tests that increase the likelihood that a plan can be negotiated and confirmed. This flexibility is balanced by the Code’s inherent concern with equality of treatment. We seek to maintain this balance in our interpretation of § 1129(b)(1) here.

The Code does not compel courts reviewing cramdown plans to enforce subordination agreements strictly, though not to do so must conform with the constraints set out in the cramdown provision. The pragmatic approach taken by the Bankruptcy Court, affirmed by the District Court, reached the right result. Thus we also affirm.