The Fifth Circuit settled one question about Serta's 2020 uptier in December 2024. It was not a permissible open market purchase. What it left open was the question that actually decides money: if the transaction was not permitted, did the participating lenders breach the credit agreement, and if so, what do they owe.

On July 7, 2026, the United States Bankruptcy Court for the Southern District of Texas answered both. The participating lenders breached Section 2.18(c) of the credit agreement, the sacred right of pro rata sharing, by taking payment on their first-lien debt without buying participations in the loans held by the lenders they left behind. Judgment is for the plaintiffs, the excluded lenders and the LCM lenders. The award is $261.13 million, measured as of the closing of the 2020 Transaction, with prejudgment interest running at nine percent per year from that date through the date of the opinion.

This ruling translates the Fifth Circuit's open market purchase holding into a damages number. It does so on the text of a pro rata sharing provision of a kind found in many syndicated credit agreements, and it rests on three determinations that are not specific to Serta's facts: what the word "payment" means, how a court measures the benefit that must be shared, and the consequences of the choice of New York law. Read together, they describe the exposure that sits behind an uptier a court later finds impermissible.

A note on posture. The opinion is a final memorandum of findings of fact and conclusions of law, but the court directed the plaintiffs to submit a proposed form of final judgment that identifies each remaining defendant, the face value of its holdings, its several share of the award, and the interest accrued on that share. Per-defendant numbers are therefore still to be computed, and the decision is subject to appeal.

Serta borrowed $1.95 billion under a First Lien Term Loan Agreement in November 2016. That agreement carried a pro rata sharing provision in Section 2.18(c) and a narrow set of carve-outs, including one for an open market purchase in Section 9.05(g). By 2019 the company carried more than $2 billion in secured debt, its largest retail partner had restructured, and the pandemic had closed more than half of its manufacturing capacity. Serta needed to reduce debt and raise new money, and it ran a process designed to make its lenders compete on discount.

Two groups formed. One, later the excluded lenders, brought a drop-down structure that would move intellectual property into an unrestricted subsidiary and lend new money against it. The other, later the participating lenders, brought an uptier open to a majority that would prime everyone else. Serta accepted the uptier. It closed on June 22, 2020. The participating lenders provided $200 million of new-money first-out loans and exchanged existing first-lien and second-lien debt into an $875 million second-out tranche, with an open-ended basket for future third-out exchanges that would rank ahead of the legacy first-lien debt left with the excluded lenders.

The exchange itself is the fact that decides this case. The participating lenders sold and assigned more than $990 million of first-lien term loans back to Serta and received roughly $734 million of new first-lien second-out debt, a ratio of 74 cents on the dollar. Serta then retired and cancelled the loans it took in. To keep that exchange outside the pro rata sharing requirement, the parties papered it as an open market purchase under Section 9.05(g). The Fifth Circuit later held that it was not one.

Before the transaction, the two sides of the first-lien class were close to evenly matched. The participating lenders held about $992 million, roughly 52.6 percent of the class. The plaintiffs held about $895 million, roughly 47.4 percent. Those percentages carry through the entire damages analysis, because Section 2.18(c) shares value in proportion to holdings across the class.

Six years separate the transaction from the award. The interval is not incidental. It is the base over which mandatory prejudgment interest compounds, and it is the product of a procedural path that ran through state court, a bankruptcy filing, the Fifth Circuit, and a denial of certiorari before returning here for trial.

The participating lenders' position depended on one word. Section 2.18(c) is triggered when a lender "obtains payment" in respect of principal or interest and ends up with a greater proportion than others in its class. If a debt exchange is not a "payment," the provision never fires, and there is nothing to share. So the participating lenders argued that "payment" means cash, and that an exchange of one instrument for another is not a payment at all.

The court read the contract against that argument and rejected it. The credit agreement does not define "payment," so its meaning comes from how the word is used across the document. The participating lenders leaned on the "in Dollars" language in Section 2.18(a), which governs how the borrower pays its lenders through the administrative agent. But Section 2.18(c) governs something different. It governs what happens when a lender receives consideration in any form and from any source, and it carries no "in Dollars" limitation. The text of Section 2.18(c) itself reaches "payment (whether voluntary, involuntary, through the exercise of any right of set-off or otherwise)." A set-off is not cash. The "or otherwise" catch-all is open-ended. A debt exchange sits inside it.

The court's other basis was structural, and it closes the reading the participating lenders proposed. Section 2.18(c) carves out four transaction types from ratable treatment, in Sections 2.22, 2.23, 9.02(c), and 9.05(g). Every one of them is a non-cash exchange of one loan for another. If a debt exchange could never be a "payment" that triggers Section 2.18(c) in the first place, there would be nothing for those four carve-outs to carve out. They would be surplusage, and New York law does not read contracts to make provisions meaningless. The presence of the carve-outs is proof that the drafters understood debt exchanges to be payments. The participating lenders tried to fit their exchange into one of those carve-outs, the open market purchase. That is the whole point. You do not need an exception for something the rule never covered.

The court also declined to import an industry definition. The participating lenders offered a Loan Syndications and Trading Association guide that defines payment as a cash transfer. That guide was never referenced or incorporated into the credit agreement, "payment" is not a technical term of art, and several of the plaintiffs' witnesses testified they had never used the guide or heard of it. New York law does not let industry custom rewrite a term whose meaning is found within the four corners of the contract. The word "payment" carried its ordinary breadth, and the exchange was a payment.

Once the exchange is a payment, the remaining elements of Section 2.18(c) follow quickly, and the participating lenders' own transaction documents supply the proof. The provision requires that the payment be "in respect of" principal or interest on a loan in the class. The phrase appears several hundred times across the credit agreement and consistently means concerning, regarding, or in connection with. The participating lenders received their second-out debt as direct consideration for their first-lien term loans, at a ratio tied precisely to the principal of those loans. The signature pages to the exchange agreement show each participating lender receiving new debt equal to 74 percent of the principal amount of the first-lien loans it surrendered. The connection is not inferred. It is documented.

The court pointed to the exchange agreement's own recitals, which state that Section 2.18(c) "does not apply to any payments made as consideration for the assignment of Existing First Lien Term Loans" under Section 9.05(g). That recital is an admission built into the deal. The parties knew Section 2.18(c) would otherwise apply, which is exactly why they wrote language to escape it. The escape depended on the open market purchase holding up. It did not.

That leaves the last two elements. The participating lenders received a payment on their first-lien term loans. The plaintiffs, holding the same class of loans, received nothing. And the participating lenders never purchased participations in the plaintiffs' loans, which is the specific remedy Section 2.18(c) requires to restore ratable treatment. The court noted that the discount the participating lenders accepted is beside the point. Section 2.18(c) asks whether one lender recovered a greater proportion of its loans than another lender in the same class. The participating lenders recovered a payment. The plaintiffs recovered zero. The proportion is unequal on its face.

With liability established, the participating lenders' remaining arguments were efforts to avoid or shrink the award. The court addressed each and rejected each. Two are worth pausing on before the table, because they show how the Fifth Circuit's mandate constrains what a defendant can still argue on remand.

The ratification defense argued that a majority of lenders amended Section 9.05(g) to confirm the transaction, so the majority ratified it. But ratification presupposes a transaction that was valid to ratify. The Fifth Circuit held the transaction was not an open market purchase, and it held that changing the ratable-sharing protection required unanimous consent precisely so a majority could not bargain away a right belonging to all lenders. The ratification argument is the open market argument in new packaging, and it lost for the same reason. The law-of-the-case defense fared no better. It leaned on the prior judge's statements about the excluded lenders' "objective lack of good faith," but those statements were dicta that could be deleted without disturbing any ruling. Dicta decides nothing, so there was nothing for the excluded lenders to appeal and nothing they forfeited.

The mitigation defense deserves a word, because it is the one most likely to recur. The participating lenders argued the plaintiffs should have sold their positions rather than hold them through a decline. The court found no market to sell into. In the eighteen months after closing, only about $162 million of first-lien term loans traded, against plaintiff holdings of more than $700 million, and dealers were sitting on inventory they could not move. The plaintiffs' fact witnesses testified to weekly and biweekly calls with brokers that never produced a buyer. The court also held that a plaintiff need not sell to mitigate when selling would extinguish the very breach claim it is trying to preserve. For the LCM lenders, whose business depends on the senior-secured-lending model the transaction threatened, that claim was worth more than any price the thin market could offer.

The liability holding decides who pays. The damages holding decides how much, and the method the court used to reach it applies beyond this transaction. The dispute reduced to a single question. When Section 2.18(c) requires that "the benefit of all such payments shall be shared ratably," what is the benefit? The plaintiffs' expert measured it at the face value of the consideration the participating lenders received. The participating lenders' expert measured it at economic value, discounted for the risk that the new paper might not perform. The court took face value.

The mechanics are clean once the benefit is fixed. The participating lenders received $734 million of second-out consideration. Spread ratably across the $1.887 billion first-lien class, that benefit is worth about 38.9 cents per dollar of class principal. Applied to the plaintiffs' $895 million position, the ratable share the participating lenders should have delivered is roughly $348 million. That figure is the baseline for the entire award.

The court rejected two competing methodologies for the same reason: both drifted away from the text. The plaintiffs' first method measured damages at the time Serta emerged from bankruptcy in 2023, three years after the breach, which folded in independent market forces like the pandemic-era mattress downturn and Serta's later refinancing failure. New York law measures contract damages at the time of breach, so the court rejected it. The participating lenders' method measured the change in expected recoveries across the first-lien class from the first to the third quarter of 2020, an attempt to show that everyone's recovery improved and that the plaintiffs therefore lost nothing. The court rejected that too. Section 2.18(c) shares the benefit of the payment, not the benefits of the transaction as a whole. New money, second-lien buybacks, the price at which lenders bought their debt, and the general economics of the deal are all outside what the provision measures.

With the benefit fixed at face value and the measurement date fixed at closing, the court adopted the plaintiffs' time-of-breach model using the 25-cent first-lien price that prevailed on the closing date. The model compares a but-for world, in which the plaintiffs receive the $348 million cash payment and keep the balance of their loans, against the actual world, in which they kept all their loans and received nothing. The table below shows the sensitivity of that calculation to the assumed first-lien price. The court adopted the 25-cent column.

The participating lenders' principal objection was that the result makes no economic sense. It requires them to pay $348 million for a participation worth roughly $87 million at the closing price, a trade no rational party would accept. The court answered that the objection proves the point rather than defeating it. The parties structured a transaction that always carried the risk of not complying with Section 2.18(c). They accepted that risk, litigated it, and nearly won. When the Fifth Circuit held the open market purchase did not qualify, the consequence the provision was written to impose became the consequence they owed. Strict textual analysis decided whether the transaction complied. The same strict text decides the price of getting it wrong.

The damages number is not the exposure. The exposure is the damages number plus six years of interest at a rate the court had no discretion to reduce. The credit agreement is governed by New York law, and under New York law prejudgment interest on a breach of contract claim is mandatory at nine percent per year. It is not a penalty and not a matter of equity. It runs from the earliest ascertainable date the claim existed, which here is a single date, the closing of the transaction on June 22, 2020, and it runs through the date of the opinion.

The participating lenders argued that nine percent over six years on a multi-hundred-million-dollar award is inequitable, that New York's rate is among the highest in the country, and that the plaintiffs dismissed early suits and waited before filing again. The court held that none of that matters, because equitable considerations do not enter a breach of contract case where interest is mandatory by statute. The cases the participating lenders cited all involved federal claims or non-New York law, where no mandatory New York provision applied.

How far this reaches is best seen in the authority the court relied on. In one New York Court of Appeals case, a 1982 breach produced a damages trial thirteen years later, and prejudgment interest ran the full interval across multiple appeals. By the time judgment entered, the interest award of $417,785 exceeded the damages award of $338,521. The court also cited authority holding that a plaintiff's delay in bringing suit is not a controlling factor in computing interest. The lesson for the party that structures the transaction is direct. Choosing New York law for a credit agreement means that a breach found years later carries a mandatory nine percent running from the breach, and time is not on the breaching party's side.

The award is not a single joint number that any one defendant might be forced to satisfy alone. The credit agreement makes the lenders' obligations several and not joint, so each remaining defendant is liable only for its own proportionate share of the total, calculated by its individual first-lien holdings as of the closing date against the total holdings of all participating lenders at that time. A defendant's liability does not grow because other defendants have settled or been dismissed, and there are several who are no longer in the case.

Two adjustments sit on the plaintiffs' side of the ledger. The portion of the award attributable to any excluded or LCM lender that is no longer a plaintiff drops out of the recoverable amount. And one of the plaintiffs, Apollo, has its recovery limited to 50 percent of its holdings, consistent with a ruling on a motion in limine and a May 2023 stipulation and agreed order under which half of the applicable purchases were deemed null and void.

The court directed the plaintiffs to submit a proposed form of final judgment. That judgment must identify each remaining defendant, the face value of its first-lien holdings as of June 22, 2020, its several share of the adjusted damages total, and the prejudgment interest accrued on that share through July 7, 2026. Until that document is entered, the $261.13 million figure is the aggregate before these several allocations, and the individual exposures remain to be computed.

The significance of this opinion is not that an uptier lost. The Fifth Circuit's 2024 decision already held that Serta's structure did not fit the open market purchase exception. What this opinion adds is a method for turning that holding into a damages number, and the method rests on reasoning that is not specific to Serta. If you advise on either side of a liability management exercise, three parts of that reasoning bear on how you price risk.

The first is the "payment" holding. The participating lenders' best textual argument was that a debt exchange is not a payment and never triggers pro rata sharing. That argument is now harder to make against a credit agreement built on the common New York template, because the same carve-outs that appear in most agreements are the proof that debt exchanges are payments. The escape hatch that some structures counted on is closed by the very exceptions those structures rely on. When you evaluate whether an exchange sits inside or outside a sacred right, the presence of non-cash carve-outs cuts against the argument that the sacred right never applied.

The second is the measure of damages. A minority lender's exposure is not the discounted market value of the paper the majority received. It is the face value of the consideration the majority negotiated for itself, shared ratably across the class. That distinction is the difference between a modest number and a large one. In this case it was the difference between an economic-value theory that produced little or nothing and a face-value theory that produced $261 million. When you model the downside of a transaction that might fail the sacred right, the face value of what the participating group takes is the right input, and it is a number the participating group sets itself.

The third is time and law. Nine percent mandatory interest running from the breach turns litigation risk into a growing liability that neither delay nor a favorable early ruling can arrest. Serta's participating lenders won at the bankruptcy court and lost at the Fifth Circuit, and the interest ran the entire time. A choice-of-law clause that reads as boilerplate carries a real and compounding cost when a transaction is later unwound. That cost belongs in the model at signing, not in the post-mortem.

Two cautions keep this in proportion. This is a trial court's decision, applying New York law to one credit agreement, and it is subject to appeal. Another agreement with different sacred-right language, different carve-outs, or a different governing law could come out differently, and the proposed final judgment that fixes each defendant's several share has not yet been entered. But the direction is clear enough to act on. The instruments and the mechanics that made these transactions attractive are the same instruments and mechanics a court now reads strictly against the party that built them. The discipline that goes into structuring an exchange to satisfy a sacred right is the same discipline that will be applied, word for word, when a court decides what the failure costs.