Opinion
Case No.: 09–50026 REG Jointly Administered
10-14-2015
In re Motors Liquidation Company, et al., f/k/a General Motors Corp., et al. Debtors.
Brown Rudnick, Designated Counsel for Ignition Switch and Non–Ignition Switch Plaintiffs, Seven Times Square, New York, New York 10036, by: Edward S. Weisfelner, Esq. (argued) Gibson, Dunn & Crutcher, LLP, Counsel for Wilmington Trust Company as GUC Trust Administrator, 200 Park Avenue, New York, New York 10166, by: Lisa H. Rubin, Esq. (argued) Akin, Gump, Strauss, Hauer & Feld, LLP, Counsel for Participating GUC Trust Unit Trust Holders, One Bryant Park, New York, New York 10036, by: Deborah Newman, Esq. (argued)
Brown Rudnick, Designated Counsel for Ignition Switch and Non–Ignition Switch Plaintiffs, Seven Times Square, New York, New York 10036, by: Edward S. Weisfelner, Esq. (argued)
Gibson, Dunn & Crutcher, LLP, Counsel for Wilmington Trust Company as GUC Trust Administrator, 200 Park Avenue, New York, New York 10166, by: Lisa H. Rubin, Esq. (argued)
Akin, Gump, Strauss, Hauer & Feld, LLP, Counsel for Participating GUC Trust Unit Trust Holders, One Bryant Park, New York, New York 10036, by: Deborah Newman, Esq. (argued)
BENCH DECISION ON REQUEST FOR STAY
I use bench decisions to lay out in writing decisions that are too long, or too important, to dictate in open court, but where the circumstances don't permit more leisurely drafting or more extensive or polished discussion. Because they often start as scripts for decisions to be dictated in open court, they tend to have a more conversational tone.
Robert E. Gerber, United States Bankruptcy Judge
In this contested matter in the chapter 11 case of debtor Motors Liquidation Company, the Ignition Switch Plaintiffs move, under Fed.R.Bankr.P. 8007(a), for a stay pending their appeal from the June 1, 2015 Judgment of a contemplated $135 million distribution from the GUC Trust to its unitholders (the “Unitholders ”). I'm granting the request for a stay, subject to the posting of a bond in the amount of $10.6 million. My Findings of Fact, Conclusions of Law, and bases for the exercise of my discretion in connection with this determination follow.
Unless otherwise noted, terms here are as defined in the Stipulations of Fact Regarding Request for Stay (ECF # 13441).
The Ignition Switch Plaintiffs don't specify the exact subsection of Bankruptcy Rule 8007(a) under which they rely. Since they aren't seeking a stay of the Judgment itself, it must be Rule 8007(a) ( (1)(C): “an order ... granting an injunction while an appeal is pending....”
ECF # 13177 (the “Judgment”).
Findings of Fact
Background facts are set forth in the Motion to Enforce Decision, familiarity with which is assumed. As additional Findings of Fact, I find as facts each of the facts stipulated to between the two sides. I also accept as true (a) the underlying factual information in the declaration and subsequent testimony of the GUC Trust's expert, Andrew Scruton, and (b) the tables and graphs showing market information submitted by (x) Mr. Scruton and (y) the Ignition Switch Plaintiffs—though I rely on the superseding Scruton Decl. Supplemental Exhibits (B–1 through F–1). I then draw my own factual conclusions, and substitute my own judgment, based on the evidence presented to me as to the projected yields that GUC Trust Beneficiaries could reasonably be expected to obtain if distributions to them were not stayed.
In re Motors Liquidation Co., 529 B.R. 510 (Bankr.S.D.N.Y.2015).
ECF # 13441.
I find particularly significant the following facts, which bear on the exercise of my discretion on this motion:
As of June 30, 2015, about 32 million units of beneficial interest in the GUC Trust (each, a “GUC Trust Unit ” or “Unit ”) were outstanding, corresponding to about $32 billion in Allowed General Unsecured Claims.
As of June 30, 2015, there was only one remaining Disputed General Unsecured Claim, in an asserted amount of approximately $20 million, which claim is subject to pending objections filed by the GUC Trust. The Ignition Switch Plaintiffs have expressed the intention to file one or more late proofs of claim (including one or more class proofs of claim for amounts as high as $10 billion), but they have not yet done so.
After recent dispositions of New GM stock and warrants received by Old GM incident to the 363 Sale of the bulk of its assets to New GM, the GUC Trust holds approximately $809.9 million, all or substantially all of which has been invested in “Permissible Investments,” as defined in the GUC Trust Agreement, which are very safe investments but which as a result get a very low yield.
Under the Plan and its organizational documents, the GUC Trust needs to reserve approximately $792 million, for a variety of administrative costs and future obligations. But it expects to distribute approximately $135 million of its assets in mid-November 2015, and will do so unless enjoined.
Additionally, the GUC Trust anticipates that it will be in a position to make a further distribution of approximately $109 million to holders of Units at a later time, expected to be a year or more from now. Thus there is a total of $244 million that the GUC Trust has or expects to have available for distribution that the Ignition Switch Plaintiffs, sooner or later, wish to freeze.
However, presumably because the second distribution is anticipated to be so far in the future, the Ignition Switch Plaintiffs did not press what I understood to be a request that I stay the distribution of the entire $244 million, and the resulting argument focused on the first $135 million. I do likewise, and my determination of the appropriate bond is based on the prejudice resulting from a stay of the $135 million alone. In what I believe to be the unlikely event that the Second Circuit has not decided the mootness issues by the time any additional distribution is imminent, the parties will have their respective rights with respect to any follow-up stay request.
Prior to the Stock Sale, to the limited extent the GUC Trust then held cash, it held such cash in certain Permissible Investments as well. The average rate of return for such investments was approximately 0.08% per annum for 2013, 0.15% per annum for 2014, and 0.10% per annum for 2015. The GUC Trust now has quite a bit more cash. It is investing its assets in U.S. Treasury obligations, and expects to continue to do so. The GUC Trust Administrator anticipates that through year end, the average rate of return on these investments will be approximately 0.12% per annum.
The U.S. Treasury obligations in which the GUC Trust intends to invest its funds are Permissible Investments, but there may be other Permissible Investments that draw somewhat higher yields. The Ignition Switch Plaintiffs contend that the GUC Trust has been unduly cautious in choosing to invest its assets solely in U.S. Treasury obligations when there are other Permissible Investments that draw higher yields—which may be as high as “ten times that amount” (i.e., up to about 1.2% per annum). Mr. Scruton understood the reason for investing in U.S. Treasury obligations alone to be the GUC Trust's desire to avoid the risk of being deemed to be an “investment company” under the Investment Company Act of 1940, and thus being subject to the additional regulation and resulting expense of such status. No evidence was submitted to me that the GUC Trust's explanation for investing its assets in U.S. Treasury obligations was pretextual, and I accept the facts on the ground as they are. Thus I find that for as long as a stay is imposed, the GUC Trust can reasonably be expected to earn a very modest 0.12% on the investments—U.S. Treasuries—that it now is holding.
Tr. of Hrg. of 9/24/15 at 9:15–19.
Tr. of Hrg. of 9/22/15 at 114:9–20, 159:21–25.
Each side provided information on yields that might be obtained on alternative investments by GUC Trust Beneficiaries if they received the distributions that the Ignition Switch Plaintiffs wish to enjoin. They were set forth in Scruton. Decl. Supp. Exhs. B–1 through F–1, and Pl. Exhs. B, C, D, and F. The Court has no reason to doubt the accuracy
of the underlying data in any of those exhibits, and accepts that data as accurate. But it forms its own view as to the use of that data—which to avoid repetition, appears only in the discussion below.
Under a feature of the Sale Agreement, called the “Accordion Feature, ” New GM is obligated to provide additional consideration in the form of additional shares of New GM Common Stock if the aggregate amount of Allowed General Unsecured Claims against the Debtors exceeds $35 billion—though it is capped at an incremental $7 billion, i.e., at the level of $42 billion in total Allowed General Unsecured Claims. As of June 30, 2015, the current aggregate value of Allowed General Unsecured Claims is a little less than $32 billion, more than $3 billion below that threshold amount. I thus find that based on the claims filed and allowed or disallowed to date, and on the one disputed claim (for $20 million) that is still pending, there is no reasonable expectation that the amount of General Unsecured Claims would reach $35 billion.
See Sale Agreement Section 3.2(c)(i) as modified by Sale Agreement Second Amendment.
If the aggregate value of Allowed General Unsecured Claims reaches $42 billion, then New GM would be required to contribute the full value available pursuant to the Accordion Feature, which is 30 million shares of New GM Common Stock, worth approximately $921.6 million as of market-close on September 14, 2015.
The Ignition Switch Plaintiffs say their claims are “conservatively worth $10 billion.” (I used a $7 to $10 billion figure earlier in the Motion to Enforce Decision, but then and now, I have been merely reporting accounts of what was being sought). I cannot and do not express a view as to whether they're right. But assuming, for the sake of argument, that claims of that amount were allowed, those claims could push the overall claims amount over the $35 billion level, and, perhaps, to the $42 billion level.
The Ignition Switch Plaintiffs seek to tap the $244 million which the GUC Trust now has on hand and that it wishes to distribute in November and in the future, and the $921.6 million in value that would be allowed under the Accordion Feature if the $42 billion threshold were reached.
Preliminarily, I note that while the Ignition Switch Plaintiffs are correct in their observation that I said, earlier in these proceedings, that if a request for a similar stay had been made in the fall of 2014, I'd have granted that in a heartbeat, that was before two events that I now need to take into account: the showing of prejudice to GUC Trust Unitholders that has been made on this application, and my rulings of April 15, 2015 with the benefit of briefing and judicial consideration at that time. Thus I look at the issues here based on the present record, and certainly can't regard anything I previously said as law of the case.
Determining this motion requires me to decide:
(1) whether any stay of distributions can be granted at all (as against the contention that doing so would amount to an impermissible modification of the Plan);
(2) if so, whether the requirements for a stay have been satisfied;
(3) if so, whether a bond should be imposed as a condition to the requested stay; and
(4) if so, in what amount.
I conclude that under the facts here:
(1) a stay can be granted, assuming that the requirements for obtaining a stay are otherwise met;
(2) the requirements for a stay have been satisfied (or, more precisely, would be satisfied if, but only if, a satisfactory bond were posted);
(3) a bond must indeed be posted, to protect GUC Trust Unitholders from the substantial loss they would suffer by reason of unjustified delay in receiving their GUC Trust distributions; and
(4) the amount of the bond must be set at $10.6 million.
1. Would a Stay Be Violative of the Plan?
As a threshold matter, the GUC Trust contends that I should not even consider the potential grant of a stay delaying distributions—arguing that “[p]ursuant to the Plan and the GUC Trust Agreement, the GUC Trust is not only authorized, it is required to make quarterly distributions to GUC Trust Unitholders to the extent that assets of the GUC Trust are available for distribution and exceed certain thresholds.” On that premise, the GUC Trust further argues that “[a]ny attempt to enjoin a liquidating trustee from making a distribution to creditors as required by a plan and liquidating trust agreement constitutes a proposed plan modification that is governed by Section 1127(b) of the Bankruptcy Code ” and that there has been no compliance with section 1127(b).
GUC Trust Br. (ECF # 13256) at 5.
Id. at 6–7.
While I agree that additional requirements that section 1127(b) would impose plainly have not been satisfied, I otherwise cannot agree.
As the predicate for this contention, the GUC Trust cites a provision of the Plan—its Paragraph 6.2(l )—but does not quote it. That paragraph, in an Article 6.2 captioned “The GUC Trust” and itself labeled “Distribution of GUC Trust Assets,” provides, in relevant part:
Subject to Section 5.2(a) hereof, the GUC Trust Administrator shall distribute quarterly (to the extent there are sufficient assets available for distribution in accordance with the GUC Trust Agreement), beginning on the first Business Day following the Effective Date, or as soon thereafter as is practicable, the appropriate amount of New GM Securities (and other distributions of Cash, if any) to holders of Allowed General Unsecured Claims and/or GUC Trust Units, as applicable.
Plan Paragraph 6.2(l ) (i.e., the lower case letter “l”) (emphasis added).
That provision follows a separate provision of the Plan, its Paragraph 4.3(a) (in an earlier Article 4.3 providing for the treatment of general unsecured claims), providing (along with other unsecured creditor entitlements) for an initial distribution to Old GM unsecured creditors on or about the Plan's Effective Date of New GM stock and warrants, and GUC Trust Units, which would provide Old GM creditors with the bulk of their distributions.
Especially when read in context, Paragraph 6.2(l ) is a timing provision, calling for supplemental distributions to Unitholders (who would be a combination of initial Old GM unsecured creditors and subsequent purchasers of GUC Trust Units) “to the extent there are sufficient assets available for distribution,” and then the “appropriate amount.” But it does not specify how those matters are to be determined, or what the GUC Trust should do when a court determines that assets should not be regarded as available for distribution or the distribution of assets is inappropriate.
That is not to say that interfering with Unitholder distributions isn't serious business. Unitholders have a justified expectation of receiving those distributions in the absence of very good reasons to the contrary. In fact, that very important concern underlies several of my mootness conclusions as discussed back at the time of the Motion to Enforce Decision. But with appropriate protections, a judicial determination that payment now isn't “appropriate”—and a stay making funds temporarily unavailable for distribution—wouldn't be violative of the Plan.
In further support of its contentions, the GUC Trust points to Judge Glenn's decision on a similar stay request in the Borders Books chapter 11 case, In re BGI, Inc., 2012 Bankr.LEXIS 5244, 2012 WL 5392208 (Bankr.S.D.N.Y. Nov. 2, 2012) (“ Borders–Bankruptcy ”), aff'd In re BGI, Inc., 2013 U.S. Dist. LEXIS 77740, 2013 WL 10822966 (S.D.N.Y. May 22, 2013). There Judge Glenn considered plan modification concerns to be among the several reasons he denied a stay to the applicants there. I fully concur with the conclusions and reasoning in Borders–Bankruptcy, but note the obvious: the starting point for any analysis of whether a stay of distributions would be violative of the Plan starts with an analysis of what the Plan actually says. Thus it is unnecessary to fully discuss other distinctions between the facts in Borders–Bankruptcy and here, such as the fact that in Borders–Bankruptcy, the stay was sought by parties whose claims already had been disallowed, a fact not present here.
2. Requirements for a Stay
Though I'm not as persuaded as the Ignition Switch Plaintiffs seem to be that the standards applicable to a Rule 8007(a)(1)(C) injunction motion should be regarded as the same as those to a Rule 8007(a)(1)(A) motion for a stay of an order or judgment itself, I don't need to decide that, as I don't think the difference here matters. The applicable standards on similar requests have been stated slightly differently, but they can be distilled into a requirement of irreparable injury; the degree of prejudice to the party to be enjoined; a showing of a likelihood of success (albeit to somewhat varying degrees); and the public interest.
(a) Irreparable Injury
First, I think there's no doubt that the Ignition Switch Plaintiffs have shown irreparable injury. I don't need to decide, and don't decide, whether the possibility that an appeal would be mooted or further mooted is sufficient by itself to constitute irreparable injury—since here, as a practical matter, it would be impossible or very difficult to get any distributions back. So I consider this factor satisfied, and consider it in a major way.
(b) Prejudice to the Other Side
The second factor, prejudice to the other side, can be looked at in two separate ways. At one level (the level at which the Ignition Switch Plaintiffs prefer to examine it), the requested relief amounts to a preservation of the status quo, which by itself represents no more than modest injury. But as the GUC Trust properly observes, that view ignores an obvious injury of another type. The GUC Trust has shown that the yields that the GUC Trust gets on its invested cash, in the current interest rate environment, are extraordinarily low—only 0.12%. That's quite a bit less than 1%—very close to zero. And investing in instruments that might generate materially higher yields isn't an option, because the GUC Trust is limited to investing in Permitted Investments, and for very good reason. The GUC Trust is holding its reserves as a fiduciary, with the duty to protect the principal for its beneficiaries.
So there's what one might call a species of negative arbitrage here. Holding on to cash equivalents that generate such a low yield is dreadful from a business perspective. Thus I conclude that the GUC Trust—and in particular, its beneficiaries—would be prejudiced too, because it's been shown to my satisfaction that GUC Trust Unitholders could get greater yields once they received their distributions, and they'll suffer material losses by reason of the difference.
Generally, and here as well, I consider prejudice to the party to be enjoined to be weighed quite heavily. But I also recognize that prejudice of this character can at least often be addressed by a bond, so we'll come back to this momentarily.
(c) Likelihood of Success
Then I turn to the requirement for a showing of likelihood of success. How much of a likelihood of success needs to be shown depends somewhat on whether we look at this as we might if this were an application for a stay pending appeal—which it is in one sense, but not in another—or an ordinary request for a TRO or preliminary injunction.
I say in one sense but not in another because the Ignition Switch Plaintiffs are not asking for a stay of the underlying Judgment itself—i.e., of the order or judgment appealed from. Instead they're asking me to enjoin a separate event, the fairness of proceeding with an act which could be affected by the outcome of a now ongoing appeal. As previously noted, that is properly regarded as being covered by Bankruptcy Rule 8007(a)(1)(C) —“an order granting an injunction while an appeal is pending”—rather than (a)(1)(A), “a stay of judgment, order, or decree of the bankruptcy court pending appeal.”
See n.3 above.
I'm not as persuaded as the Ignition Switch Plaintiffs are that the standards—and in particular, the likelihood of success standards—for a stay pending appeal under (a)(1)(A) and for an injunction under (a)(1)(C) are the same. The cases the Ignition Switch Plaintiffs cite all precede the 2014 revisions in the 8000 series of Bankruptcy Rules. Rule 8007, which comes from former Rule 8005, now has, in its subsection (a), separate subparagraphs relating to the different types of relief —and none of the Ignition Switch Plaintiffs' cited cases, including Judge Haight's 1994 decision in Bidermann, says, directly or by implication, that there is no distinction between the Second Circuit's standards for a stay pending appeal and for an injunction pending appeal. Rather, the likelihood of success requirement for a stay pending appeal is “a substantial possibility, although less than a likelihood,” of success on appeal. The comparable requirement for ordinary preliminary injunction analysis is either a likelihood of success or serious issues going to the merits, and a substantial tipping of the hardships in the applicant's favor.
See 10 Collier on Bankruptcy (“ Collier ”) ¶ 8007.RH (16th ed.2015).
See 10 Collier ¶ 8007.05[1].
Bidermann v. RHI Holdings (In re Bidermann), 1994 U.S. Dist. LEXIS 9700, 1994 WL 376090 (S.D.N.Y. July 18, 1994) (Haight, J.).
Hirschfeld v. Board of Elections, 984 F.2d 35, 39 (2d Cir.1993).
See, e.g., Jackson Dairy, Inc. v. H.P. Hood & Sons, Inc., 596 F.2d 70, 72 (2d Cir.1979) (“ Jackson Dairy ”).
Here I don't find either a likelihood of success on the appeal nor a substantial possibility (although less than a likelihood) of success on appeal on the mootness issue. I was well aware of the Accordion Feature when I issued the Motion to Enforce Decision and the resulting Judgment. In fact, I expressly discussed the Accordion Feature in the Motion to Enforce Decision, and again in a later decision before entry of the Judgment when addressing the form of Judgment that would implement the former's rulings. Thus, in the Form of Judgment Decision, I noted that when people invested in GUC Trust Units, they had no reasonable basis for a concern that the accordion feature would be triggered:
See 533 B.R. at 538 & n. 58.
See In re Motors Liquidation Co., 531 B.R. 354 (Bankr.S.D.N.Y.2015) (the “ Form of Judgment Decision ”).
When Old GM creditors received distributions under the Plan, and when Unitholders—even if as aftermarket acquirors of GUC Trust Units—acquired their units, they had a reasonable expectation that the total universe of claims filed against Old GM would not increase. And while they knew that there was an accordion feature, they also knew that claims exposure would result, with exceptions exceedingly difficult to show, only from previously filed claims.
So I don't see it as particularly likely that a reversal would be forthcoming based on a failure on my part to have considered the Accordion Feature. And importantly, the Ignition Switch Plaintiffs are trying to access not just value that might become newly available under the Accordion Feature but also funds already in the GUC Trust that GUC Trust Beneficiaries would reasonably have expected to tap. Even if the Ignition Switch Plaintiffs were right in their contention that they could access incremental value that might come in under the Accordion Feature, they would be particularly unlikely to succeed in arguing on appeal that they should access the funds now in the GUC Trust—which are the funds whose distribution the Ignition Switch Plaintiffs want to block—unless they came up with a different argument, which they haven't made yet if they ever will.
Nevertheless, I find that the Ignition Switch Plaintiffs have shown enough in the way of likelihood of success to meet the requirements for injunctive relief now. A traditional alternate basis for injunctive relief in this Circuit—as a substitute for showing a likelihood of success—has been “sufficiently serious questions going to the merits to make them a fair ground for litigation and a balance of hardships tipping decidedly toward the party requesting the preliminary relief.” The Ignition Switch Plaintiffs' mootness arguments do raise sufficiently serious issues going to the merits, and though I believe, as noted above, that GUC Trust Unitholders would also be prejudiced by a stay, the balance of hardships still tips decidedly in the Ignition Switch Plaintiffs' favor.
Jackson Dairy, 596 F.2d at 72.
(d) Public Interest
Finally, I turn to the public interest. The GUC Trust properly notes what I said in Chemtura : that “the public interest ... recognizes the desirability of implementing the legitimate expectations of creditors ... to get paid.” And in Borders–Bankruptcy, Judge Glenn found that a stay of interim distributions to creditors pending an appeal would not serve the public's interest because “Congress and the courts have stressed the need for parties to be able to rely on the finality of chapter 11 plans and related orders in conducting business and in dealing with the reorganized debtor. But there is also a public interest in protecting the right to appellate review when it can be done without undue prejudice to the side that won below. And I don't see that issuing a stay would impair the public interest, especially if the resulting loss were protected by the requirement of a bond.
In re Chemtura Corp., 2010 Bankr.LEXIS 3988, 2010 WL 4638898 (Bankr.S.D.N.Y. Nov. 8, 2010).
2010 Bankr.LEXIS 3988 at *26, 2010 WL 4638898 at *8.
2012 Bankr.LEXIS 5244 at *20, 2012 WL 5392208 at *6.
(e) Stay Conclusions
Looking at the factors as a whole, and weighing the first two factors most heavily, I believe that if a satisfactory bond were posted, the requirements for a preliminary injunction delaying the distributions would then be satisfied. So I'll issue the requested stay. The real issue is the amount of the bond that I should require to compensate the GUC Trust Unitholders, through the GUC Trust itself, for the delay in receiving their distributions.
3. Bond
On the matter of the bond, there is no dispute as to the underlying legal framework. The purpose of requiring a bond in this context is to indemnify the party prevailing in the original action against loss caused by an unsuccessful attempt to reverse the holding of the bankruptcy court.
See Tribune Media Co. v. Aurelius Capital Mgmt., L.P. (In re Tribune Media Co., ) 799 F.3d 272, 281 (3d Cir.2015) (“ Tribune–Circuit ”).
It is agreed that the posting of a bond is a matter within my discretion. It is also agreed that a bond is not mandatory, and that in theory, I could issue the requested injunction without a bond. But it is also the case that “if the movant seeks imposition of a stay without a bond, the applicant has the burden of demonstrating why a court should deviate from the ordinary full security requirement.”
See, e.g., In re Overmyer, 53 B.R. 952, 955 (Bankr.S.D.N.Y.1985) (Schwartzberg, J.); In re General Motors, Corp., 409 B.R. 24, 30 (Bankr.S.D.N.Y.2009) (Gerber, J.) (“ General Motors–Stay ”).
Triple Net Investments IX, LP v. DJK Residential, LLC (In re DJK Residential, LLC), 2008 U.S. Dist. LEXIS 19801 at *6, 2008 WL 650389 at *2 (S.D.N.Y. March 7, 2008) (Lynch, J.). See also General Motors–Stay, 409 B.R. at 30 ; In re Tribune Co., 477 B.R. 465, n. 11 (Bankr.D.Del.2012) (“ Tribune–Bankruptcy ”); In re W.R. Grace & Co., 475 B.R. 34, 209 (Bankr.D.Del.2012).
Preliminarily, I agree with the GUC Trust that a bond is necessary. There can be no serious dispute that GUC Trust Unitholders will be prejudiced by a delay in distributions when the underlying assets would continue to draw a 0.12% yield while the stay is in place, and those assets would generate much higher yields after receipt by their recipients. Though the prejudice here is not as large as it was in the Adelphia chapter 11 case on my watch (where the estate suffered monetary losses of $2.33 million per day (and $49 million in the aggregate) during a period of an unbonded stay before the district court required a bond), the prejudice still is quite substantial. For reasons discussed below in connection with my computation of the necessary bond that should be imposed, GUC Trust Unitholders would be prejudiced to the extent of millions of dollars if their distributions were enjoined during the pendency of the appeal.
See General Motors–Stay, 409 B.R. at 34 (discussing the Adelphia episode, and explaining why I would require a bond in an amount no less than $7.4 billion before staying the 363 Sale).
I assume that if I made the bond prohibitively high, it could result in the bond not being posted, and thus that the $135 million at issue here (and, though less likely, the $109 million to be distributed farther out in the future) would be distributed before appellate review. But I'm not at all persuaded by the Ignition Switch Plaintiffs' argument (Motion at 15), that a bond requirement “may chill participation in the appeal.” The plaintiffs could and undoubtedly would appeal the mootness ruling anyway, even if the $135 million went out, in an effort to access the remaining $109 million and, more importantly, the $922 million in incremental value that might exist under the Accordion Feature.
So I then must determine how large a bond should be required. The principles incident to setting an appropriate level of bond here are very much like those employed by Judge Carey in Tribune–Bankruptcy —in which he required a $1.5 billion bond in the largely similar, though not identical, context of a request to stay confirmation, in an analysis later stated in Tribune–Circuit to be as “well-considered and as convincing as the alchemy of valuation in bankruptcy can be.”
See Tribune–Circuit, 799 F.3d at 282.
Here, I consider one element in particular of the several types of losses analyzed in Tribune–Bankruptcy —opportunity costs to GUC Trust Unitholders whose distributions would be delayed. After doing so, I consider the $18.4 million for which the GUC Trust argues, and the zero for which the Ignition Switch Plaintiffs argue, to be way too high and low, respectively. I conclude, contrary to the positions taken by each of them, that the opportunity cost should be regarded as $10.6 million—based on an annual estimated yield of 9.23% (compounded) for what I believe should be estimated to be a 10–month delay—resulting in the $10.6 million that I fix as the bond.
The yield that we should assume as the alternative yield that we could expect the universe of GUC Trust Beneficiaries to obtain if their distributions were not stayed is a function of (1) the duration of the time during which the distribution would be delayed if stayed, and (2) the yield on the alternative investments that could be made if the distribution could be made as planned. The latter, in turn, is a function of the particular yields that could be obtained on various types of alternate investments, and, if more than one type of investment were considered, the mix—and consequent weighting—of the investments that GUC Trust Unitholders might make.
(a) Duration
The Ignition Switch Plaintiffs note that (at Judge Furman's direction), they filed a motion for an expedited appeal (which was unopposed), and that as a result, I should assume a stay of a duration less than the 12 months that the GUC Trust assumed. As of the time of this writing, the Second Circuit has not ruled on the motion, and there can be no assurance that it will be granted. But assuming that it is, I note the timing the Ignition Switch Plaintiffs proposed. Their motion contemplated the filing of the last round of briefs on February 22, and oral argument on March 8, 2016 or the earliest possible date thereafter. Thus, if the Circuit were to rule on the day of argument itself, there would still be a delay of approximately four months from the mid-November 2015 date upon which the GUC Trust expects to make its $135 million distribution.
See Motion, Elliott v. General Motors L.L.C., No. 15-2844 (2d Cir., filed Sep. 29, 2015) at 2.
In a demonstrative handed up to the Court, the Ignition Switch Plaintiffs ran numbers assuming stay durations of from 4 to 12 months. But I consider all but the last two of them to be unrealistic, and believe that I should assume a period of 10 months—i.e., assuming a ruling 6 months from the requested date of oral argument. The Circuit will have to deal with a raft of appeals and appellate issues, from parties who are variously appellants with respect to some issues and appellees with respect to others. My opinion had to run 134 pages, by reason of the number and difficulty of the issues to be addressed. The same reasons underlying my certification of the foreseeable appeals for direct review by the Circuit (and, presumably, that caused the Circuit to grant a direct appeal) underscore the importance of the issues to be decided; the appeal will have a significant effect on chapter 11 practice (and section 363 sale practice, in particular) going forward. The parties to the appeal (understandably) sought a relaxation of the page limits for the briefs to be submitted, for those reasons and others. In light of these factors, and others, I believe it is most reasonable to anticipate a Circuit opinion about 6 months after argument—10 months after the date the distribution would otherwise go out.
In his declaration (see Scruton Decl. ¶ 21), GUC Trust expert Scruton noted that it took the Second Circuit 18 months (from appeal filing to decision, and apart from time getting views from the Delaware Supreme Court) to decide the other appeal that I certified for direct view by the Circuit—one involving the mistaken filing of a UCC–3 that caused the security interest in a $1.5 term loan to come to an end. The earlier appeal involved a perfect storm of issues, and this one does too, and after that the similarity ends. The issues here are more difficult, but that was not an expedited appeal. On balance, I believe 6 months from the proposed date for oral argument (which is about 13 months from the 9/9/2015 appeal filing and about 10 months from the date of the contemplated distribution) represents the best estimate of when the Circuit might reasonably be expected to rule.
(b) Alternate Yields
Turning now to the alternate yields, there is little dispute (and in any event I find) that the Unitholders' prospective loss should be measured by comparing the GUC Trust's projected returns of 0.12% on the Permitted Investments in which it expects to hold its cash (e.g., a mix of short-term U.S. Treasuries) against the rate of return that Unitholders could reasonably earn if they were able to invest the anticipated GUC Trust distributions during the life of the stay.
I'm unpersuaded by the Ignition Switch Plaintiffs' contention that I should assume a higher yield by the GUC Trust on the cash it now has, because they disagree with the GUC Trust's determination to avoid making investments creating a risk that it would be deemed to be subject to the Investment Company Act of 1940, like mutual funds. I find the GUC Trust's concern entirely reasonable.
Here, as in Tribune–Bankruptcy, evidence of the yield on alternative investments was offered by the party opposing the stay. In Tribune–Bankruptcy, Judge Carey was provided with a declaration from an investment banker at Lazard, drawing conclusions from yield information taken from a commonly-used index to analyze the high yield—sometimes called “junk bond”—market. Here I was given information of a somewhat similar character, but GUC Trust expert Scruton, recognizing that a meaningful percentage (at least 47%) of the GUC Trust's Unitholders were hedge funds, principally based his opinion on a mix of returns reported by hedge funds, equity investments, fixed income investments (of a grade materially higher than junk bonds), and money market quality yields. Without needing to decide whether the type of yield evidence presented here is superior to that presented to Judge Carey (or vice-versa), or whether “Best Practices” would call for use of a mix of different kinds of investment data (by reason of differing investor goals and practices), I find the use of a mix of different kinds of investments here at least to be appropriate.
See Kurtz Decl., In re Tribune Company, No. 08–13141–KJC, ECF # 12217–2 (Bankr.D.Del.), at 6 & n.6 (basing conclusions on the Merrill Lynch U.S. High Yield Master II Index (“High Yield Index”).
Measured by the Credit Suisse Event Driven Multi–Strategy Hedge Fund Index (“Hedge Fund Index”).
Measured by the S & P 500 Index.
Measured by the Bank of America Merrill Lynch U.S. High Grade Master Index. Note that the concepts of “High Grade” and “High Yield” are largely inconsistent, and that this index is quite different than the one presented to Judge Carey in Tribune–Bankruptcy.
Measured by historic yields on the U.S. Treasury 10–Year note.
I don't need to address that further because here (and I think it is only by coincidence), the High Yield Index “Yield to Worst,” used in Tribune–Bankruptcy, would be 7.17%, and the weighted average mean, for reasons addressed below, would be 7.13% and its weighted Protection Return Rate would be 8.34%—figures in the former case very close, and in the latter case relatively close, to the 7.17% figure.
But different types of investments generate different yields, and those yields can fluctuate over time. The two sides thus debate whether I should go with a historical mean (or “trimmed” mean, discarding outlier years), on the one hand, or a yield high enough to minimize likely prejudice to Unitholders (a “Protection Return Rate ”) (recognizing that use of a mean would result in a 50% likelihood of providing insufficient protection), on the other. They also debate how I should weigh the yields on the alternative investments that GUC Trust Unitholders might make, given the lack of data I have on what particular Unitholders' investment goals and practices might be.
I agree with the GUC Trust that use of some kind of Protection Return Rate—and not just a 10–year mean—is generally necessary and appropriate. By definition, use of a mean will understate results, on a weighted basis, 50% of the time. If, as we all agree (or should), the purpose of a bond is to protect the side that prevailed below, I think a 50% risk of getting it wrong is too high a risk.
Using a median—where half of the results would be higher, and half lower—would present similar conceptual difficulties, with the only difference being less sensitivity to the highs and lows of investment performance in given years.
But by the same token, I believe that particular Protection Rate methodology proposed by the GUC Trust—one based on the weighted average of the third-best year from the last ten years of annual investment returns for each asset class (the “3rd Highest Return ”) is too aggressive. The effect of that methodology, as Mr. Scruton testified at the hearing on this motion, would be to cover approximately 80% of all possible outcomes. That is overkill; I can't agree that a bond “giving effectively 80 percent protection” to Unitholders is required to protect against injury resulting from a stay. For lack of more or better data, I use the trimmed mean calculated by Mr. Scruton (adjusted as described below), with an additional upward adjustment of 1/3 of the difference between the trimmed mean and the rate that would result using the 3rd Highest Return.
Scruton Decl. ¶ 18.
Tr. of Hrg. of 9/22/15 at 135:22–136:3.
Id. at 136:14–21.
Then, with respect to weighting, the evidence showed that yields obtained by hedge funds would typically exceed— sometimes by significant amounts—the yields obtained on equities and, especially, fixed income and money market type investments. Thus the weighting of the different types of investments would make a material difference. It was established that 47% of the Unitholders—actively represented in this case by a single counsel (the “Participating Unitholders ”)—were hedge funds. But while it was suggested that a significant number of the other Unitholders were too, no evidence was offered to back up that assumption, possibly because of the unavailability of such evidence or possibly because of hedge funds' historic aversion to disclosure as to their investments in the cases in which they invest. In any event, in determining weighted yields for the universe of GUC Trust Unitholders—who would make individualized decisions as to whether to invest in hedge fund investments, equities, fixed income, or money market type investments—I estimate the mix based on the 47% of Unitholders of whom we are aware, and (for lack of a better alternative) assume that the remaining 53% would invest in the four types of investments in equal numbers—i.e., with 13.25% (1/4 of the 53%) investing in each.
For example, the Participating Unitholders declined to provide information as to their own actual yields, even though it is quite possible that their actual yields would have been higher than the Hedge Fund Index upon which they rely. I don't penalize them for that (nor do I either require them to provide that information or draw an adverse inference from their failure to do so), but this once again causes me to be reluctant to make assumptions as to their yields where they have not given me a firm evidentiary basis for doing so.
That would result in 60.25% of GUC Trust Unitholders being assumed to invest in hedge fund type investments (47% + 13.25%), and 13.25% in each of the other three kinds.
This type of weighting was the premise of “Scenario 3” in the GUC Trust's “Protection Return Rate Estimates,” Scruton Decl. Exh. D–1, and it is the data appearing in this Scenario, with the “Rates of Return” shown on Exh. D–1, with which I start going forward.
Then we turn to the yields I should assume for each class of investment. In each case, GUC Trust expert Scruton used historic data, trimmed to peel off what some might regard as outlier data—the highest and lowest historical returns. I find that approach appropriate as a general matter, but with an exception—where we have better data upon which to form a view.
See Exh. D–1 n.1.
Historic data is appropriate—and, in fact, necessary, in my view—with respect to hedge fund, equities and fixed income investments, for lack of any better basis upon which to make an intelligent decision as to investment results in the upcoming 10 months. But we know more with respect to money market equivalents. I think I can assume with nearly total certainty that given the present level of money market interest rates, and recent events in the economy and actions by the Federal Reserve Board, that yields on U.S. Treasuries will not climb to the 3.21% trimmed mean—much less the 4.29% Protection Return Rate for which the GUC Trust argues—in the upcoming year. We know (and the Ignition Switch Plaintiffs showed, by their Exh. C, showing the Department of the Treasury's Yield Curve Rates) that a 1–Year Treasury would have a yield of 0.36% and even a 10–Year Treasury would have a yield of only 2.20% (each, as of the day before the evidentiary hearing on this matter). As I consider it reasonable to assume that in this environment, with no more than modest increases in interest rates likely in the next 10 months, there would be at most modest declines in the value of a November 2015–purchased 10–Year Treasury over that period. Therefore, I use 2.2% as the yield on money market equivalents, as contrasted to the 4.29% and 3.21% upon which Scruton relied.
I would do likewise for fixed income investments if the Ignition Switch Plaintiffs had given me information like that in Pl. Exh. C to provide an alternative. But as they did not, I use the historic data the GUC Trust provided.
Accordingly, I choose to work with Scruton Decl. Exh. D–1's “Scenario 3” “Trimmed Mean” data, as modified in two respects: (1) to substitute 2.2% (in place of 3.21%) as the rate of return for the 13.25% of Unitholders who would be assumed to be making money market-type investments; and (2) to increase the assumed yield on each of the four classes of investments by adding 1/3 of the difference between the trimmed mean, as adjusted above, and the 3rd Highest Return to the adjusted trimmed mean to provide a less aggressive (and less inclusive) Protection Return Rate than the one proposed by the GUC Trust, but that is more inclusive than a rate based on the historical trimmed mean. That results in a new “Scenario 4”, adapted from “Scenario 3,” as follows:
[Editor's Note: The preceding image contains the reference for footnote ].”
Determining whether the yield for investment in the Hedge Fund asset class used here should be net or gross of a 2% management and incentive fee requires a difficult and subjective judgment. Mr. Scruton testified that he took a conservative approach in his calculations by netting out the 2% fee (resulting in a lower yield than the alternative approach would have resulted in), and thus I do the same.See Tr. of Hrg. of 9/22/15 at 100:19–101:7.
Thus, instead of the 13.82% Weighted Average Protection Return Rate proposed by the GUC Trust, I consider it more appropriate to use a “Court–Determined Protection Return Rate” of 9.23%. That assumed alternative yield, in the exercise of my discretion—informed by the data discussed above—most fairly protects the GUC Trust in connection with the distribution that would be delayed by reason of the imposition of a stay. (c) Resulting Size of Appropriate Bond
Then I need to determine the size of the bond that would compensate the GUC Trust and its Unitholders for the delay in making that $135 million distribution. Conceptually, that is roughly ($135 million) x (9.23% per year) x (10/12 of a year)–or about $10.38 million. But it is not exactly that, because of the effect of compounding.
The exact result of that computation is $10.38375 million.
Scruton testified upon my questioning under Rule 614 that compounding would result in about a 10% increase (over this time period) as compared to a simple interest computation. But he also provided a table, submitted as another supplemental exhibit to his declaration, setting forth computations for lost opportunity costs as functions of various illustrative Protection Return Rates (in increments of 1.00%) and numbers of months of lost opportunity costs (ranging from 4 months to 18 months), which presumably did take into account compounding. Those Protection Rates used in that table that were closest to the Court–Determined Protection Rate were 9.0% (resulting in a bond, for a 10–month delay, of $10.3 million) and 10.0% (resulting in a bond of 11.5 million). Assuming linear ratios, and then multiplying the lower $10.3 million by the ratio of 9.23%/9.00%, I reach a bond size of approximately $10.6 million, and set the bond in that amount.
Tr. of Hrg. of 9/22/15 at 171:17–172:12.
See Scruton Decl. Exh. F–1.
That may or may not be appropriate given the compounding, but I believe it to be sufficiently close for the purposes of this analysis.
Conclusion
The bond is fixed in the amount of $10.6 million. But distributions by the GUC Trust will be stayed for a period of 14 days from this order to allow the Ignition Switch Plaintiffs time to post the bond, or, if they are so advised, to pursue an appellate modification of this determination.
I've considered, and rejected, the possibility of any lengthier stay. Fed.R.Bankr.P. 8007(a)(1)'s requirement that motions for stays ordinarily should be made first in the bankruptcy court has effectively been satisfied, and any further request should be made only in a higher court.
--------
This order takes effect immediately. But either party may settle a more formal order if it is so advised.
SO ORDERED.