Copyright (c) 2025 the National Conference of Bankruptcy Judges
THE SAFE HARBOR FOR LEVERAGED BUYOUTS IN BANKRUPTCY
by
David Gray Carlson*
Introduction
When I was a lad, Pennzoil Co. v. Texaco, Inc.[1] passed as the biggest litigation of all time (as measured by dollars).[2] But this case has since been matched or maybe surpassed by fraudulent transfer claims in leveraged buyout (LBO) bankruptcy cases.[3] These bankruptcy cases have often wended their way into a safe harbor, wherein, if the payout is laundered through a “financial institution,” LBO defendants who received funds for no reasonably equivalent value are seemingly untouchable. According to § 546(e):
Notwithstanding sections 544, 545, 547, 548(a)(1)(B), and 548(b) of this title, the trustee may not avoid a transfer that is a margin payment, as defined in section 101, 741, or 761 of this title, or a settlement payment, as defined in section 101 or 741 of this title, made by or to (or for the benefit of) a commodity broker, forward contract merchant, stockbroker, financial institution, financial participant, or securities clearing agency, in connection with a securities contract, as defined in section 741(7), commodity contract, as defined in section 761(4), or forward contract, that is made before the commencement of the case, except under section 548(a)(1)(A) of this title.
Application of § 546(e) to shield LBO payouts is thought to be “outrageous”[4] and “absurd.”[5] The Supreme Court ended the outrage in Merit Management Group, LP v. FTI Consulting, Inc.,[6] or so it seemed.[7] But appearances deceive. Effective money laundering through financial institutions still lives. The reason why is a defect in the statutory definition of “financial institution.” The definition defines “financial institution,” basically, as a bank and a customer of a bank.[8] Thus, in Deutsche Bank Trust Co. Americas v. Large Private Beneficial Owners (In re Tribune Co. Fraudulent Conveyance Litigation) (Tribune II),[9] the Second Circuit ruled that the Tribune Co. (a publisher of newspapers and the owner of a baseball team) was in effect a bank, because it was the customer of a bank. Any fraudulent transfer made by the Tribune Co. in the LBO was therefore safe-harbored.
Tribune involved an attempt to recover LBO “settlement payments” from departing shareholders laundered through a financial institution. Seemingly, the bankruptcy trustee (T) can still target the mortgages and security interests that the corporate debtor has issued to the lender that financed the LBO, if these mortgages were intended to hinder the unsecured creditors.[10] Such a lender gave value to the debtor[11]—the loan for which the mortgages were security. But perhaps the lender is not a purchaser in good faith. The Seventh Circuit in Petr v. BMO Harris Bank, N.A.[12] hints that even those security interests are safe-harbored. Section 546(e) does not merely protect settlement payments. After 2006,[13] it also protects transfers (i.e., security interests) to financial institutions made “in connection with” securities contracts. Where the LBO lender is a financial institution, its security interests are safe-harbored.
Putting these ideas together, it will be the rare LBO that is not completely immune from fraudulent critique. Arguably, this may seem to be a good thing, since on the merits LBOs are not typically fraudulent transfers.[14] But still, an LBO might constitute a looting scheme in fraud of creditors. As it stands, § 546(e) threatens to take billions off the claw-back table—which is arguably outrageous and absurd.
In this article, I wish to argue that courts should give up on the language of § 546(e) and instead consult the policy behind it: to protect the institutions that regularly participate in the securities clearance system.[15] This core policy suggests that LBOs should not be safe-harbored: the integrity of the securities clearance mechanism does not depend on it.[16] Tribune establishes the safe harbor for LBOs based on the plain meaning of § 546(e). But Tribune also covertly depends on setting aside the plain meaning. In effect, I argue that two absurdities cancel each other out. Plain meaning cannot govern. Accordingly, we should ignore the words and always consult the policy of § 546(e).
The unacknowledged absurdity that the courts pass over when they safe-harbor LBOs is that § 548(a)(1) of the Bankruptcy Code invites avoidance of obligations as well as transfers. Section 546(e) safe-harbors many (not all) fraudulent transfers when funds are laundered through a financial institution. But § 548(a)(1) also avoids fraudulent obligations. Section 546(e) is silent about avoiding obligations. The plain meaning of § 546(e) is that T may avoid the obligation to settle a securities contract. But to follow this plain meaning completely destroys § 546(e)’s core mission of safe-harboring participants in the securities clearance system. Therefore, § 546(e) is absurd and cannot be enforced according to its plain meaning.
Instead of the plain meaning we must consult the policy of § 546(e). Policy dictates that, where T may not avoid transfers incident to securities contracts, she also may not avoid the obligation incident thereto, though the plain language of § 548(a)(1) says otherwise and § 546(e) poses no obstacle. Section 546(e) is so absurdly drafted that it fails to meet its core purpose. Since the plain meaning is absurd, courts (whether they admit it or not) are in the position of setting aside the plain meaning, and there is nothing left but to consult the policy behind the safe harbor.
This article is divided into nine parts. Part I gives minimal necessary background to understand the logical structure of a fraudulent transfer theory. Part II describes why LBOs might be (but are probably not) fraudulent transfers. Part III describes the history of applying the § 546(e) safe harbor to LBOs. Part IV describes how Merit Management seemingly mined the LBO safe harbor (though Merit Management was not actually an LBO case). Part V describes how Tribune reinstated the harbor by mine-sweeping the Merit Management opinion. It did so by declaring that customers of banks are constructive banks when the real bank is agent to its customer. Part VI examines whether the banks are really agents for their customers in an LBO. It concludes that they usually are. Part VII considers whether a financial institution “benefits” when it takes fees for its role in an LBO. If modest fees are a benefit, then the entire LBO is safe-harbored, as § 546(e) is currently drafted. But this is absurd and so must be rejected in favor of the underlying policy. Part VIII discusses the Seventh Circuit’s expansion of the safe harbor to protect LBO lenders.
Finally, Part IX shows that, under the plain meaning of § 546(e), it is always open for T to avoid the obligations created by an LBO, even though T may not avoid the transfer of property attendant thereto. Because avoiding obligations blows a gaping hole in the vital battlements of the § 546(e) safe harbor,[17] the plain meaning of § 546(e) cannot govern. Courts are ignoring the plain meaning of § 546(e) whether they admit it or not. The contradiction means that courts are left with nothing but consulting the policy behind § 546(e). According to policy, LBO defendants (i.e., the departing shareholders and the LBO lender) should never be safe-harbored.
I. Fraudulent Transfers in Bankruptcy
This section gives a brief summary of what fraudulent transfers are, in and out of bankruptcy proceedings.[18]
The usual fraudulent transfer is one intended by a debtor to delay, defraud, and hinder creditors. Suppose an insolvent debtor (D) owns a gold brick. To keep her creditors from levying the brick, D gives it away to X. Or D sells the brick to X and absconds with the cash that X paid. The brick has been fraudulently transferred. Where X gave no value to D or where X gave value but knew D would abscond, the unsecured creditors of D can levy on X’s brick.
Starting early in the 20th century, American courts tired of litigating D’s state of mind. Instead, it was decreed (by courts and later by statute)[19] that any transfer by an insolvent D for no reasonably equivalent value was fraudulent. This became known as a constructive fraudulent transfer. It should be noted that there is a tremendous overlap between intentional and constructive fraudulent transfers. A gift to X when D is insolvent may have been intended to flummox the creditors of D.[20]
Bankruptcy trustees have always had standing to recover fraudulent transfers on behalf of all the unsecured creditors.[21] In modern times, T can directly recover the gold brick under § 548(a)(1)(A) (intentional fraudulent transfer) or § 548(a)(1)(B) (constructive fraudulent transfer). But § 548(a)(1) has a short lookback period of two years.
To supplement T’s avoidance power under § 548(a)(1), T succeeds to the avoidance rights of existing creditors of D under § 544(b)(1). This is known as the subrogation provision, though the nickname is not exactly accurate. To establish a cause of action under § 544(b)(1), T must locate a “trigger” creditor who, under state law, can actually reach the gold brick.[22] For instance, suppose D gives X a gold brick worth $100. This is avoidable by a creditor of D (C1) who claims $60 from D. Although the nickname “subrogation” suggests that T can only avoid $60 worth of the brick, § 544(b)(1) allows T to have the entire brick, not just $60 of it. The brick goes into the bankruptcy estate where the creditors share and share alike (including C1, who at state law would have received the entire $60). This known as the rule of Moore v. Bay.[23]
State fraudulent transfer law usually has a longer lookback period than § 548(a)(1). Under the Uniform Fraudulent Transfer Act (UFTA), the lookback is at least four years.[24] Therefore, where the transfer of the brick was two years or less before the bankruptcy, T can sue under § 548(a)(1) and also § 544(b)(1) (if T can locate a trigger). If the transfer is between two and four years old (or perhaps older), T must rely on § 544(b)(1) and state law as the theory of avoidance.[25]
II. Redemptions and LBOs
In a stock redemption, a corporation buys back its shares on the open market and then cancels them, leaving the non-redeemers with a greater percentage share of the company’s net worth. A company will redeem when it thinks its shares are under-valued by the market. When these facts are true, the non-redeemers profit when the redeemers accept the market price for their shares.
When a corporate debtor (D Corp.) is insolvent, a stock redemption is a constructive fraudulent transfer. To illustrate, suppose D Corp. has assets worth $100 (including $20 in cash) and debts for $120. On these numbers, D Corp. is insolvent. The shares of D Corp. are worth about zero. If D Corp. were liquidated presently, each dollar of debt merits a dividend of 83 cents. Suppose instead D Corp. redeems some shares for $20. The shareholders have tendered shares worth about zero and have walked away with $20. Now D Corp.’s bankruptcy estate is worth $80, while the credit claims are still $120. Liquidation yields a dividend of 66 cents. The creditors have been hindered by 17 cents on the dollar. Redemptions, when D Corp. is insolvent, are therefore fraudulent transfers.[26]
LBOs are transactions in which a corporate raider R takes over the company while investing minimal R dollars in D shares. Instead, R arranges for D Corp. to borrow on a secured basis. D Corp. uses the loan proceeds to redeem the shares from the shareholders.[27] For example, suppose, before the LBO, D Corp. has $120 of assets against $100 of unsecured debts. The unsecured creditors have an equity cushion of $20. The shares are worth about $20. R buys $1 worth of D Corp. shares on the open market. D Corp. borrows $20 on a secured basis from the LBO lender. D Corp. redeems all the shares (except R’s) for $20. D Corp.’s balance sheet now shows $20 of senior secured LBO debt and $100 of unsecured debt against $120 in assets. Previously, the unsecured creditors had a cushion of $20. Now they have a cushion of zero. R’s idea is that R can reduce costs and increase D Corp.’s value.[28] Suppose R reduces costs by $3, thereby increasing the value of D Corp.’s cash flow from $120 to $123. R has tripled her money. She invested $1 but now has equity worth $3. The creditors’ equity cushion was $20. At the moment of the LBO, it was zero, though it has grown to $3. This is R’s vision. If it comes true, the LBO actually helps creditors (viewed from the point of origin, where the $20 cushion was already lost). But, where the economy crashes, as it did in 2008, D Corp. is often bankrupt.
LBOs are not usually fraudulent transfers. A fraudulent transfer is a deliberate scheme to hinder creditors. Or it is a gift by an insolvent. LBOs greatly increase the risk to the unsecured creditors, but it is not usually intended to hinder them. Fraudulent transfer law does not punish risky leverage strategies.[29] It is usually designed to create a leaner, more profitable D Corp. It is not a fraudulent transfer to decrease the equity cushion for the unsecured creditors. Otherwise, stock dividends would be impossible. Nevertheless, courts might find the LBO to be a fraudulent transfer. According to fraudulent transfer law, a gift to the redeeming shareholders is fraudulent when D Corp. “was engaged in business . . . for which any property remaining with the debtor was an unreasonably small capital.”[30] Since there was always a chance that the economy will crash, a finder of fact could judge that the equity cushion after the LBO was unreasonably small, thus making the stock redemption voidable.[31]
There are two categories of transfers in an LBO. First, the LBO lender who funds the deal has taken mortgages and security interests in D Corp. assets. The lender is a purchaser of its mortgages for value. [32] It has extended the loan to D Corp. If the lender did not think in advance that D Corp. was doomed, the lender is a good faith purchaser for value with a defense against avoiding the mortgages. For instance, if the lender studied the cash flows and judged that it was likely that cash flow was sufficient to cover production costs plus debt service, then D Corp. was structured to survive. The LBO, though risky, was believed by the lender to be viable. The mortgages could not then be challenged because the lender was a good faith transferee for value.[33] Nevertheless, if the LBO lender in bad faith financed the stock redemption knowing it would hinder creditors, T may avoid the obligation of D Corp. to pay and therefore the mortgages that secure the obligation.
The second category of transfers is the cash and other property conveyed to the redeeming shareholders. If the redemption rendered D Corp. insolvent, then T could recover the consideration D Corp. conveyed to the shareholders.
When T files suit against the departing shareholders, § 546(e) enters the scene. If, at the end of the LBO, the shareholders have received a settlement payment laundered through a financial institution, the shareholders are arguably immune from avoidance under § 548(a)(1)(B) or § 544(b)(1). Suppose in the LBO, D Corp. funds the LBO by forwarding sufficient funds to a clearing house or bank. The clearing house or bank then receives the tendered shares and pays out the agreed-upon consideration. The shareholders have received a settlement payment under a securities contract and are arguably eligible for § 546(e) protection.
Once again, § 546(e) provides:
Notwithstanding sections 544 . . . [or] 548(b)(1)(B) . . . of this title, the trustee may not avoid a transfer that is a . . . settlement payment made by or to a . . . financial institution . . . or that is a transfer made by or to (or for the benefit of) a . . . financial institution . . . in connection with a securities contract . . . that is made before commencement of the case, except under section 548(a)(1)(A) . . . .
A lot is going on in this subsection. For starters, T is not barred from claiming that the securities contract is a fraudulent obligation. This is the absurdity of § 546(e)’s plain language, which we shall visit later.[34]
Second, T is not barred from all fraudulent transfer theories against the shareholders. Section 548(a)(1)(A) refers to intentional fraudulent transfers. T may (and actually does)[35] sue the shareholders and even the financial institution on the theory that the redemption was an intentional fraudulent transfer, but this only works if the transfer occurred within the two-year lookback period in § 548(a)(1). For transfers older than that, the trustee is barred from any recovery that is based on § 544(b)(1). Suppose a trigger C1 could avoid an LBO transfer to the shareholders because D Corp. intended to hinder C1’s collection. T is barred by § 546(e) from making this claim. In addition, T is barred from constructive fraudulent transfer claims that C1 might have brought under § 548(a)(1)(B) and § 544(b)(1). Furthermore, under the reasoning of Tribune, even C1 is barred from avoiding a fraudulent LBO under state law, since the Bankruptcy Code preempts the state law of fraudulent transfer.[36]
Section 546(e) requires that the payment to the shareholders must have been made by a financial institution. According to Bankruptcy Code § 101(22)(A), a “financial institution” is a “bank . . . and, when any such [bank] is acting as agent or custodian for a customer . . . in connection with a securities contract (as defined in section 741) such customer . . . .” So, a financial institution is a bank and, perhaps, the customer of a bank.
In the context of an LBO, § 546(e) requires that the payment received by the departing shareholders be “in connection with a securities contract.” Section 546(e) borrows the lengthy definition from § 741(7) from “Subchapter III—Stockbroker Liquidation.” One can bring the LBO under this definition, in that the LBO constitutes “a contract for the purchase . . . of a security . . . .”[37]
Finally, the payment must be a settlement payment (or, after 2006,[38] a transfer by or to a financial institution in connection with a securities contract). Again, with regard to settlement payments, § 546(e) borrows from § 741:
(8) “settlement payment” means a preliminary settlement payment, a partial settlement payment, and interim settlement payment, settlement payment on account, a final settlement payment, or any other similar payment commonly used in the securities trade . . . .[39]
The definition is circular, as courts have complained.[40] It is easy to see, however, that the payout in an LBO closes out or “settles” a securities contract.[41]
III. Section 546(e) As Applied to LBOs
The case that first applied § 546(e)[42] to an LBO was Kaiser Steel Corp. v. Charles Schwab & Co.[43] In Kaiser Steel, the debtor corporation (Kaiser) offered to redeem common shares for cash and preferred shares. The redeemers were required to tender their shares to Bank of America (BA) which received the shares and paid out cash and preferred shares. Among those who tendered was Charles Schwab & Co., acting as broker for various redeemers. Schwab conformed to the tender by ordering its agent, the Depository Trust Co. (DTC),[44] to transfer its security entitlement to BA, so that we had this chain:
Kaiser was soon bankrupt. T (the debtor-in-possession) sued everyone in the chain on the theory that the redemption constituted a fraudulent transfer—a transfer for no reasonably equivalent value at a time when Kaiser was insolvent. Schwab moved to dismiss on the ground of § 546(e).
T argued that § 546(e) should be limited to ordinary course clearance of securities. But the Tenth Circuit found no reference to ordinary course in § 546(e). It found that Schwab was under the protection of § 546(e), which expressly mentions settlement payments made to a stockbroker.[45] Schwab (but not Schwab’s clients) were out of the case. So far, this was unexceptional.
A year later, in Kaiser Steel Corp. v. Pearl Brewing Co. (In re Kaiser Steel Corp),[46] the shareholders sought refuge in the same safe harbor, because they had received settlement payments from Schwab or from BA in connection with a securities contract. T argued that the payments were not settlement payments because they were extraordinary (i.e., part of an LBO), not routine. Also, T argued, only settlement payments between financial institutions were safeguarded.[47] The court found that payments to the shareholders were plainly settlement payments. The shareholders were thus safe-harbored. By consulting the plain meaning of § 546(e), the court safe-harbored the departing LBO shareholders.
This result was extended in Lowenschuss v. Resorts International, Inc. (In re Resorts International, Inc.),[48] which involved the following chain
In Resorts, Lowenschuss was safe-harbored even though the shares and funds did not pass through the securities clearance system.[49]
A further extension occurred in Brandt v. B.A. Capital Co. LP (In re Plassein International Corp.),[50] which involved the privately held shares not traded in public markets.
Here was the chain:
Here the shares were never delivered to an intermediary. But D Corp. instructed its bank to wire payment to the banks in which the shareholders had accounts. Thus, any payment of a securities contract through a bank qualified the transferees for the safe harbor.[51]
In the cases considered so far, departing LBO shareholders won an ill-deserved voyage to the safe harbor. There was a counter-tradition. In Munford v. Valuation Research Corp. (In re Munford, Inc.),[52] the relevant chain was
In explaining why §546(e) did not safe-harbor the shareholders, the court said:
Importantly, a trustee may only avoid a transfer to a “transferee.” Since the bank never acquired a beneficial interest in the funds, it was not a “transferee” in the LBO transaction. Rather the shareholders were the only “transferees” of the funds here. And, of course, section 546(e) offers no protection from the trustee’s avoiding powers to shareholders; rather, section 546(e) protects only commodity brokers, forward contract merchants, stockbrokers, financial institutions, and securities clearing agencies. Accordingly, regardless of whether the payments qualify as settlement payments, section 546(e) is not applicable since the LBO transaction did not involve a transfer to one of the listed protected entitles.[53]
Since this short solution was later adopted by the Supreme Court, it bears some elaboration. The idea is that, when Munford sent funds to Citizens & Southern Trust Co. (CS), it was making two transfers to two transferees. Legal title to the funds was transferred to CS (for the benefit of the shareholders). So, CS was an initial transferee of the funds. It received legal title alone. This was a valueless thing.
Separately, D Corp. transferred the beneficial interest in the funds directly to the shareholders. As to this, the shareholders were initial transferees. Thus, T could recover the beneficial interest directly from the shareholders. So conceived, § 546(e) did not figure in the analysis. Section 546(e) would have been relevant if T were pursuing CS as initial transferee of equitable title to funds. But it did not apply when T was pursuing the shareholders as initial transferee of equitable title.
Munford has an apparent weakness which can nevertheless be explained away. Granted, the shareholders were initial transferees of the equitable interest of the settlement payment. But still, this equitable interest was cashed out when the financial institution wired the funds to the depositary banks of the shareholders. They received a “transfer that is . . . a settlement payment.” The trustee was seeking the return of that payment. Are not the shareholders safe-harbored in spite of the Munford reasoning?
They are not safe-harbored for a subtle reason suggested by the Seventh Circuit in in FTI Consulting, Inc. v. Merit Management Corp., LP.[54] Section 546(e) says that “the trustee may not avoid a transfer that is a . . . settlement payment by . . . a financial institution.” A fraudulent transfer by D Corp. is something that the unsecured creditors of D Corp. can avoid. In Munford, the unsecured creditors of CS cannot avoid CS’s transfer of legal title to the shareholders. That is what it takes to put the settlement payment into the safe harbor. True enough, the ultimate check or wire transfer from CS to the shareholders was a settlement payment. But no one is claiming that the settlement payment by CS was a fraud on the creditors of CS. The claim is that the payment by D Corp. was a fraud on D Corp.’s unsecured creditors.
The Seventh Circuit in Merit Management said this of § 546(e):
It is impossible to say in the abstract what the italicized words, “by or to,” mean here . . . [A] postcard sent through the U.S. Postal Service could be said to have been sent “by” the Postal Service or “by” the sender who filled it out. When a person pays her bills using an electronic bank transfer, the funds could be said to be sent “by” the owner of the account or by the bank. Similarly, a transfer through a financial institution as intermediary could reasonably be interpreted as being “made by or to” the financial institution or “made by or to” the entity ultimately receiving the money.[55]
The Seventh Circuit concluded that, for the safe harbor to apply, what is sent “by” the financial institution must be voidable by the unsecured creditors of the financial institution. “It makes sense to understand the safe harbor as applying to the transfers that are eligible for avoidance in the first place.”[56]
IV. Merit Management and the Overarching Transfer
Courts were thus divided over whether shareholders who tender their shares in LBO are safe-harbored under § 546(e). In Merit Management Group, LP v. FTI Consulting, Inc.,[57] the Supreme Court intervened (unsuccessfully) to resolve the controversy.
The case did not actually involve an LBO. It involved an ordinary contract of sale whereby a buyer agreed to buy, and a seller agreed to sell securities in a privately held company. The fraudulent transfer claim was that D Corp. paid too much for the stock it bought.
Justice Sotomayor begins her opinion with a verbal schema:
This diagram unnecessarily inserts the financer of the purchase (Credit Suisse, (CS)) in the middle of the chain. I think we can ignore the fact that D Corp. obtained bank financing for the purchase. A settlement payment should be interpreted to mean a payment by D Corp., which the creditors of D Corp. claim is voidable. A bank loan to finance the purchase of a security is not a settlement payment; the creditors of the bank are not claiming that the bank loan is a fraud. The bank loan becomes a settlement payment when D Corp. directs it toward the seller of the securities. This is the “subtle” reasoning presented earlier.
The fact that CS’s loan constituted a credit in the deposit account of D Corp. with CS does not win CS a place in the chain. As the depositary bank that received a payment order from its customer, CS received nothing (new) from D Corp., nor did it transfer D Corp. property. Once CS credited the account of D Corp., CS owed D Corp. a debt. (That’s what a deposit account is.)[58] In effect, D Corp. transferred this debt to Citizens Bank (CB) in trust for the benefit of Merit Management. Now indebted to CB and Merit Management, CS transferred its own property (not D Corp. property) in satisfaction of an antecedent debt. What CB received was proceeds of D Corp.’s deposit account. CS was neither transferor nor transferee of D Corp. property but was the thing that D Corp. transferred to CB (legal title) and to Merit Management (equitable title).[59] Thus, we expel CS from the chain.
Here is how I would have drawn the lattice:
Justice Sotomayor ruled “the relevant transfer for purposes of the § 546(e) safe-harbor inquiry is the overarching transfer that the trustee seeks to avoid . . . .”[60] The safe harbor thus applies only to “overarching” transfers. The harbor applies only when D Corp. transfers equitable (not legal) title to CB. In that case, D Corp. overarches to CB, a financial institution, and so the safe harbor applies. But those were not the facts in Merit Management. T was not trying to avoid D Corp. → CB. In Merit Management, the overarching transfer was D Corp. → Merit Management. Merit Management was not within the range of § 546(e).[61]
V. Tribune
If the Supreme Court thought that it was driving the departing LBO shareholders out of the safe harbor, the Second Circuit would soon teach it a rude lesson. In Deutsche Bank Trust Co. Americas v. Large Private Beneficial Owners (In re Tribune Co. Fraudulent Conveyance Litigation) (Tribune II),[62] the Second Circuit held that by enacting § 546(e) of the Bankruptcy Code, Congress intended to shield LBOs from fraudulent transfer scrutiny.[63] And, going further, Congress even intended to preempt the state law of fraudulent transfers, which is not burdened by any such shield. Thus, according to the Second Circuit, not only did Congress intend to shield LBOs from regulation, but it also intended to preempt state law, to the extent state law views LBOs as potential fraudulent transfers.[64] What is to some an outrage and an absurdity was found to be exactly what Congress intended, both in and out of bankruptcy.
In Tribune, a corporate raider (R) took control of Tribune Co. through an LBO. The form of the deal was a share redemption.[65] In the deal, R’s subsidiary (R Sub) invested $250 million in Tribune in exchange for shares issued by Tribune. Tribune agreed to borrow $11 billion from a syndicate in exchange for a security interest and mortgages on Tribune assets.[66] Over $8 billion were used to redeem Tribune shares. Tribune paid the $8 billion to a “financial institution.” The financial institution then used the $8 billion to buy Tribune’s shares on behalf of Tribune. Tribune canceled these shares leaving R the only Tribune shareholder. Less than two years later, Tribune was bankrupt. An unsecured creditors’ committee (CC) was constituted to take charge of the fraudulent transfer litigation.
It seemed at the time—before the Merit Management decision—that the CC could not sue the shareholders under § 548(a)(1)(B) or § 544(b)(1).[67] But the CC could and did sue the shareholders under § 548(a)(1)(A), the intentional fraudulent transfer theory. [68]
As for the constructive fraudulent transfer theories, a metaphysical puzzle manifested itself. T is subrogated to the right of the general creditors (the Cg) to avoid a constructive fraudulent transfer. But T is barred by § 546(e) from asserting any theory based on § 544(b)(1). Since T could not bring these claims, did this mean that the fraudulent transfer right bounced back to the Cg? The CC and the Cg thought so. If so, T (and therefore the CC) may have been barred by § 546(e) from challenging the LBO as a constructive fraudulent transfer, but the Cg could bring state fraudulent transfer challenges. Section 546(e) defanged bankruptcy trustees only. It did not expressly prevent state-law LBO challenges.
To erase doubt, the Cg moved to lift the automatic stay to enable private constructive fraudulent transfer lawsuits. The Delaware bankruptcy court granted the motion and later confirmed a chapter 11 plan that dissolved the CC but created a “litigation trust” (LT) to pursue § 548(a)(1)(A) intentional fraudulent transfer claims against the shareholders. The plan also “expressly allowed [the Cg] to pursue ‘any and all LBO-Related Causes of Action arising under state fraudulent conveyance law.’”[69]
The Cg soon filed private constructive fraud cases against the shareholders around the country. These were consolidated by a multidistrict litigation panel in the Southern District of New York. Thus, the Tribune bankruptcy was in Delaware and New York was the host of state-law fraudulent transfer claims by the Cg against the departing LBO shareholders.
In New York, the shareholders moved to dismiss the fraudulent transfer theories of the Cg. District Judge Richard J. Sullivan granted the motion. He reasoned that since the LT had already sued the shareholders for intentional fraudulent transfer, the LT owned this cause of action and the Cgs’ claim interfered with it. This violated the automatic stay, justifying dismissal.[70]
The Second Circuit affirmed in Tribune I, but not on the ground asserted by Judge Sullivan. The Second Circuit noted that the Delaware bankruptcy court had lifted the automatic stay. So, the stay could not be grounds to dismiss the Cg. The Second Circuit went on to rule that § 546(e) preempted the state law rights of the Cg to challenge the LBO.
Tribune I was handed down shortly before the Supreme Court decided Merit Management. Originally, Judge Ralph K. Winter, in Tribune I, found for safe harbor and preemption of state law. Thereafter, the Cg sought Supreme Court review of Judge Winter’s ruling. While the petition for certiorari was pending, the Supreme Court severed the moorings of Judge Winter’s ruling, or so it seemed. In Merit Management, the Supreme Court ruled that § 544(e) could not be read to immunize the shareholders. The shareholders were seemingly caught in the overarch. In light of the Merit Management holding, the Supreme Court reversed Tribune I and remanded, asking the Second Circuit to review its holding.[71]
In a jointly authored opinion in Tribune II, Judge Christopher F. Droney and Judge Winter found that § 546(e) still protected LBOs from constructive fraudulent transfer claims. “In Tribune II, the Second Circuit eviscerated the Supreme Court’s holding in Merit and turned it on its head.”[72]
In Merit Management, the Supreme Court had insinuated that immunity from fraudulent transfer suits applied only when a “financial institution” was the fraudulent transferor or a fraudulent transferee of an equitable interest in property. Judges Droney and Winter, however, consulted the statutory definition of “financial institution.” According to Bankruptcy Code § 101(22)(A), a “financial institution” is a “bank . . . and, when any such [bank] is acting as agent or custodian for a customer . . . in connection with a securities contract . . . [also] such customer . . . .” Since the payoff to the shareholders was a securities contract and the clearing agent (a bank)[73] was acting for Tribune (a customer), Tribune (a publishing company) was a financial institution and so § 546(e) still provided a defense.[74] In a footnote in the Merit Management opinion, Justice Sotomayor had indicated that the Court was not addressing the “customer” question.[75] Tribune II demonstrated that the footnote would doom the Merit Management analysis to oblivion, at least insofar as LBOs are concerned.
Why does § 101(22)(A) define a financial institution as a bank or, if the bank is agent for a customer, the customer? According to Professor Peter Marchetti, Congress wished to shield a specific practice—securities lending transactions. In these transactions, a non-bank (typically a broker) lends securities (not cash) to cover someone else’s short sales of securities.[76] The broker transmits the securities to its bank. The bank guarantees to its broker customer that the borrower will return the securities.[77] If T could sue the customer for fraudulently receiving a margin payment, the customer, having paid T, could revisit the loss on its bank under the guaranty. This augured a systematic risk on the banking industry. Therefore, it was supposedly necessary to safe-harbor the customer as a means of safe-harboring the bank. One can imagine a desire to protect broker-lenders, although why brokers should be allowed a safe-harbor when it takes calculated lending risks may be questioned. But, where the customer is not a broker, why can’t the banks simply refrain from the guaranty, since it is such an existential threat? Congress would do well to delete the reference to “customer” in the definition of “financial institution” and let the banks amend their guaranty contracts. Brokers would still be harbored by § 546(e), if they are the lenders. Such a statutory reform would vindicate the holding in Merit Management.
Protection of short-sale lending involved fraudulent transfers or voidable preferences made by a short-selling D Corp. to a customer of a financial institution. But the general identity of bank and customer in the definition also means that § 546(e) safe-harbors fraudulent transfers made by the customer to anyone. This is where the drafting error is located. Transfers to the customers should be safe-harbored (supposedly), to protect the financial institution from its own guaranty. Transfers by customers should not have been protected. Unfortunately, there is no way to cabinet the faulty definition to “to” without also applying it to “by.”[78]
VI. Agency and Custodianship
Perhaps we can save Merit Management from oblivion. A financial institution is a customer of a financial institution when the financial institution “is acting as agent or custodian for a customer . . . .”[79] Thus, a bank must be an “agent or custodian” of its customer, and, when it is, the customer is also a financial institution. A custodian is defined as a receiver or similar court-appointed official or a third party empowered to enforce a lien.[80] This is not likely to describe an intermediary’s role in the LBO. The Bankruptcy Code does not define “agent.” Whether Merit Management survives depends upon a painful examination of whether the financial institution in the LBO is the agent of the fraudulent transferor. If the answer is yes, § 546(e) safe-harbors LBOs after all.
In Tribune, the chain was as follows:
In this chain, Computer-Shares Trust (CT) was a financial institution. If CT was agent to Tribune (publisher of newspapers), then Tribune was a financial institution. Therefore, any LBO shareholder taking a fraudulent transfer from Tribune was off the hook so long as the transfer was laundered through a bank.
According to Judges Droney and Winter, CT was a “depositary.” Shareholders directly or indirectly “deposited” security entitlements with CT.
In its role as depositary, [CT] performed multiple services for Tribune. First, [CT] received and held Tribune’s deposit of the aggregate purchase price for the shares. Then [CT] received tendered shares, retain them on Tribune’s behalf, and paid the tendering shareholders. [¶] Given these facts, we conclude that Tribune was [CT]’s “customer” with respect to the LBO payments.[81]
In determining whether CT was an agent, the court saw nothing else to do but apply the common law meaning of “agent.” This it drew from the Restatement (Third) of Agency: “Agency is the fiduciary relationship that arises when one person (a ‘principal’) manifests assent to another person (an ‘agent’) that the agent shall act on the principal’s behalf and subject to the principal’s control.”[82] An agency is created when the principal manifests an intent to grant authority to the agent, the agent accepts, and the principal can at any time revoke the authority.[83]
Most tellingly, the court quoted from Tribune’s Offer to Purchase. “For purposes of the Tender Offer, [Tribune] will be deemed to have accepted payment [sic] . . . shares that are properly tendered and not properly withdrawn only when, as and if we give oral or written notice to [CT] of out acceptance of the shares for payment pursuant to the Tender offer . . . .”[84] It is hard to deny that CT was granted authority, CT consented to be agent, and Tribune could cancel CT’s authority to accept shares.[85] CT was an agent, Tribune was a customer, and Tribune was therefore a financial institution.
Holliday v. K Road Power Management LLC (In re Boston Generating LLC)[86] is a straightforward application of Tribune II marking that Merit Management is all but dead in LBO cases.[87] The case placed redeeming LLC interest holders in the harbor because, as the customer of the depositary bank in the LBO, the fraudulent transferor was a financial institution.
In Boston Generating, Boston Generating LLC (BG) was the operating subsidiary of EBG Holdings LLC (EBG), a Delaware LLC. A syndicate lent billions to BG on the assumption that this sum would be paid to EBG, which in turn would redeem some of the EBG interests. These LLC interests were found to be “securities.”[88] EG was to send the money to the Bank of New York (BNY), which would then pay the tendering interest holders, provided the interests presented were approved in advance by EBG and BG. Lending in the first instance to BG assured the syndicate that BG assets were reachable by the syndicate, in case of default.
Three years later, the economy crashed and both EBG and BG were bankrupt. The trustee in Boston Generating sued the tendering interest holders for fraudulent transfers under New York law. Judge Robert E. Grossman concluded that the payment from BG to EBG was safe-harbored by § 546(e). Both BG and EBG were customers of BNY, and BNY was agent to both entities. Therefore, both BG and EBG were banks, under the Bankruptcy Code definition of “financial institution.” A bank was making the transfer and so the harbor applied.[89] Assuming that an agent could have two principals, Boston Generating simply follows Tribune II. So conceived, BNY was an escrow agent. BNY was authorized to act only when BG and EBG joined to instruct BNY.
One case subsequent to Tribune II refused to shield limited liability company (LLC) interest holders in an LBO. But the LBO was highly unusual and not likely to prove an effective precedent for the standard LBO.
In Greektown Litigation Trust v. Papas (In re Greektown Holdings, LLC),[90] the court found that Merrill Lynch (ML) was neither the agent nor the custodian of D Corp. Therefore, D Corp. was not a financial institution. Accordingly, § 546(e) did not prevent T’s recovery against the LLC interest holders who had sold their shares to the issuer LLC.
Greektown started with the following corporate structure:
The Kewadin holders wanted control of the casino and Papas/Gatzoros agreed to sell. In anticipation of this transaction, Kewadin bought some LLC interests in Monroe. Then Monroe redeemed the Papas/Zatzeros interests for cash to be paid over time. As a result, Kewadin controlled Greektown Casino. Papas/Gatzoros were unsecured creditors of Monroe.
Soon thereafter, the parties decided to alter the arrangement. Papas agreed to take discounted cash in exchange for its right to be paid the purchase price by Monroe. Gatzoros received some cash and would have the right to be paid the rest over time. The transaction therefore constituted a novation of a credit sale of the securities.
To raise the cash to pay Papas/Gratzoros, Monroe and Kewadin chartered a new LLC (Greektown Holdings). Monroe and Kewadin transferred their LLC interests in the Casino to Holdings. Holdings issued promissory notes, which were “purchased” by ML. Or one steeped in naiveté might say, ML lent money and Holdings issued notes. ML paid Holdings by crediting its deposit account with ML. Then (on behalf of Holdings) ML wired funds to the banks of Papas and Gatzoros. When the dust settled, the following corporate structure resulted.
In a pre-Merit Management decision, Judge Walter Shapero found that the proceeds of the ML-Holdings loan (wired by agreement to Papas and Gatzoros) was a settlement payment. This seems somewhat surprising, in that Holdings was simply borrowing and ML was simply lending. Under Article 8 of the Uniform Commercial Code, notes are possibly but not necessarily securities.[91] The notes, however, were securities[92] under Bankruptcy Code § 101(49): “the term ‘security’—(A) includes—(i) note . . . .” As far as the Bankruptcy Code is concerned, lending is just a version of buying securities from the issuer, when the loan is reified into a promissory note.[93] So viewed, ML had made a settlement payment to Holdings. According to Judge Shapero, any time a bank advances funds on a note, repayment is a settlement payment safe-harbored by § 546(e), where the loan proceeds were used to settle a securities contract. This preferential payment to a bank is very far indeed from protecting the integrity of the securities clearing system.
By the end of 2008, all the above entities were bankrupt. T brought fraudulent transfer actions against all transferees. What were the transfers? Here is the chain:
First, there was (1) a transfer of loan proceeds from ML to Holdings. This was deemed to be an eligible payment[94] because it was “in connection with a securities contract.” Holdings used the loan proceeds to buy LLC interests from Monroe.
Transfer (2) was the sale of Monroe’s Casino LLC holdings to Holdings in exchange for cash. So, this cannot be viewed as a constructive fraudulent transfer. A constructive fraudulent transfer requires that Monroe gave to Holdings no reasonably equivalent value. Still, Holdings might have intended a fraudulent transfer, if Holdings intended to flummox the creditors. But where Monroe gave valuable LLC interests to Holdings and Holdings (at Monroe’s request) wired the sales proceeds to creditors (i.e., Papas/Gratzoros), it is hard to see how any creditor of Holdings was flummoxed. Our concern, however, is with the safe harbor, not the underlying merits of the fraudulent transfer claim.[95]
Transfer (3) constituted a payment by Monroe on antecedent debt. Papas and Gratzoros were not at this point redeeming LLC shares from Monroe. That had occurred earlier. When these interests had been tendered earlier on, Papas and Gratzoros became unsecured creditors of Monroe. What occurred at (3) was actually the novation of an earlier securities contract whereby Holdings bought securities on credit. Is novation of a securities contract also a securities contract? If so, then Papas/Gratzoros had received a payment from ML in connection with a securities contract and (prior to Merit Management) was safe-harbored.
Judge Shapero ruled § 546(e) applied because the payment was a “transfer[] made by [a] financial institution . . . in connection with a securities contract.”[96] Said Judge Shapero, “§ 546(e) requires ‘a’ connection and nothing more . . . [A] transfer can be in connection with more than one thing.”[97] The connection was that “Holdings was legally bound to use the Senior Note proceeds to pay [Papas/Gratzoros]. In other words, the ‘connection’ not only existed, but it was also a thoroughly contemplated and mandatory connection.”[98]
In the end, Judge Shapero felt obliged to hold[99] that Papas/Gratzoros were safe-harbored. T appealed from the decision, but before the Sixth Circuit could weigh in, the Supreme Court’s Merit Management opinion was issued. Merit Management contradicted the governing Sixth Circuit decision.[100] The Greektown panel therefore reversed and remanded with instructions to investigate the meaning of Merit Management for the case. Since Judge Shapero had retired, the matter ended up in front of Judge Maria L. Oxholm.
Inspired by Tribune II, Papas/Gatzoros argued that Holdings was a financial institution, because it was the customer of ML. If so, Papas/Gatzoros were safe-harbored. Judge Oxholm disagreed and found that ML was not Holdings’ agent. Since § 546(e) is a defense, Judge Oxholm imposed on Papas/Gatzoros the burden of proving that ML was agent to Holdings:
Accordingly, the Court holds that to prove agency Defendants must establish that (1) Holdings manifested assent to [ML or an ML affiliate] shall act on Holdings’ behalf; (2) subject to Holdings’ control; and (3) [ML or affiliate] manifest or otherwise consent to the fact. Furthermore, for the first requirement, “to act on the principal’s behalf” means to be “a business representative” with the ability “to bring about, modify, affect, accept performance of, or terminate contractual obligations between his principal and third persons.”[101]
Judge Oxholm reviewed various agreements between ML and Holdings (or affiliates of Holdings).[102] She concluded that Papas/Gratzoros had failed to establish the first element. Holdings never manifested assent to ML acting as agent.
The documents examined included an “engagement letter” between Casino and ML, whereby Casino agreed for itself and its affiliates to retain ML as financial advisor and representative. This agreement bound Casino (and other related entities) “to engage [ML] (or one or more of its affiliates as designated by [ML]) as its sole lead administrative agent . . . .”[103]
According to Judge Oxholm, this agreement did not make ML agent to Holdings. The agreement was between Casino and ML. Holdings was not a party to it. Holdings was what contract scholars call a third-party beneficiary of the engagement letter.[104] “Thus, while Holdings may have benefited from this agreement, the agreement does not evidence Holdings’ assent that [ML] act on its behalf or that [ML] be subject to Holdings’ control.”[105] This may be questioned. Casino purported to act for Holdings. If Casino had actual authority to bind Holdings, then Holdings was obliged to and did retain ML as agent. The argument that Holdings never consented to retaining ML is feeble, in that ML did tender services to the corporate family of which Holdings was a part.[106]
Judge Oxholm concluded, “Therefore, [Papas/Gatzoros] failed to prove the first element of agency—that Holdings manifested assent to [ML] that [ML] shall act on Holdings’ behalf.”[107] Because the first element (Holdings’ consent) failed, the second and third element also failed.
Judge Oxholm’s finding that ML was not agent in any part of its relationship with the payment to Papas/Gatzoros might be wrong. But it is a wonder that LBO immunity should turn on disputable details of whether ML owed its “customer” the duties of an agent. Here is a convenient place to quote the final (and, as always, thoughtful) judicial words of now-retired Judge Robert D. Drain. Section 546(e) was the subject of his valedictory address:
As this is my last opinion before retiring from the bench, perhaps I can be indulged in asking, why Congress has put the courts to all this parsing and hair splitting over (a) whether a transaction is one or many and, if many, has the avoidable transaction has been properly identified, or (b) whether there is a qualifying participant that in a proper customer, agent or custodian. After all, at issue here is [an LBO] transaction whereby, after encumbering a privately held company’s assets with privately issued debt, a handful of sophisticated private equity investors took massive dividends that . . . left the pensions plans . . . and hundreds of creditors holding the bag. Only the veracity of that last assertion—that is, whether [D Corp.] was insolvent or rendered insolvent by the dividends—not whether the dividends are safe-harbored, should be at issue. The avoidance of these dividends and the loans that funded them would have no effect on the public securities markets, the ostensible purpose for section 546(e). On the other hand, the transfer avoidance provisions of the Bankruptcy Code are of fundamental importance, “help[ing] implement the core principles of bankruptcy . . . . [G]iven the importance of fraudulent transfer law in bankruptcy cases, Congress should act to restrict to public transactions its current overly broad free pass in section 546(e) that had informed the playbook of private loan and equity participants to loot privately held companies to the detriment of their non-insider creditors with effective impunity. This is no trivial matter.[108]
VII. “Or For the Benefit Of”
Judge Oxholm also faced the puzzle of the language “or for the benefit of,” which was added to § 546(e) in 2006. Thereafter, not only are transfers by or to a financial institution safe-harbored. Transfers to non-institutions are harbored if a financial institution benefited. Since financial institutions always receive fees for their services, LBO securities contracts always benefit some financial institution, or so it could be argued, and the safe-harbor is preserved from Merit Management.
“For whose benefit” originally comes from the voidable preference provision in § 60(b) of the old Bankruptcy Act, repealed in 1978 by the Bankruptcy Code. According to § 60(b):
Any such preference may be avoided by the trustee if the creditor receiving it or to be benefited thereby or his agent acting with reference thereto has, at the time when the transfer is made, reasonable cause to believe that the debtor is insolvent. Where the preference is voidable, the trustee may recover the property or, if it has been converted, its value from any person who has received or converted such property, except a bona fide purchaser form or lienor of the debtor’s transferee for a present fair equivalent value . . .[109]
“Benefited thereby” was added in 1938.[110] It codified a dictum from National Bank v. National Herkimer County Bank.[111] The idea of dunning nontransferees was to sweep in sureties that had guaranteed claims against a bankrupt debtor.
To make this concrete, suppose D borrowed from C. S, a surety, guaranteed it. Commonly S was a relative or insider of D. Suppose, just before bankruptcy, D paid C. This vastly benefits S. Voidable preference law states that, where D was insolvent at the time of the payment, the trustee can “avoid the transfer of an interest of the debtor in property (1) to or for the benefit of a creditor . . . .”[112] This language makes C liable. But it does not exactly make S liable. S is a contingent creditor of D for reimbursement,[113] in case S was compelled to pay C. S benefited when D paid C.[114] But so far T can avoid the transfer that C has received. S has received no property.[115]
The Bankruptcy Act’s voidable preference provision was drafted in vague terms, and courts had to invent a great many legal fictions to make it work. The problem was that old § 60(b) allowed the trustee to recover transfers.[116] But what debtor property had S (the surety) received?
Courts responded that the benefit was somehow a transfer. Since S had converted this fictional debtor property to her own use, D’s bankruptcy trustee could have a money judgment against S. Thus, it was said, in the case of a transfer to C, there were two transfers—one to C and a quite separate one to S.[117] Of course, the trustee could have only one recovery between S and C put together.[118] If the trustee recovered from C, then C’s claim against D (and against S) revived. Under the guaranty, S had to pay C and approach D’s bankruptcy trustee as a mere unsecured creditor, seeking pennies on the dollar.[119] Or the trustee could sue S for the value of what C received. The value was somehow property that D transferred to S. To be noted is that S was not liable for the value of the actual benefit to S. S was liable for the market value of what C received.[120]
In 1978, Congress borrowed “or for the benefit” in § 60(b) and placed it, not only in § 547(b)(1), but in § 550(a)(1).[121] Putting this language in § 550(a)(1) was necessary. Coupled with “the trustee may recover, for the benefit of the estate, the property transferred, or, if the court so orders, the value of such property,” § 550(a) successfully suggests that, even if S has received no property, S can be made to pay the value of property that C received. But “or for the benefit of” in § 550(a) exceeds the domain of voidable preferences under § 547. For the first time, fraudulent transfers under § 548(a)(1) and § 544(b)(1) are covered by those words. As a result, we must figure out what those words mean in the context of fraudulent transfer law.
In 2006, Congress compounded the error by adding “or for the benefit of” to § 546(e). Before 2006, a defendant had to show a payment to or by a financial institution in connection with a securities contract. Under Merit Management’s idea of the overarching transfer, we are to ignore the financial institutions—they only received bare legal title. Merit Management directs our attention to the equitable interest which never went to the financial institutions. After 2006, it became possible to argue that, if transfer to the recipient of the equitable interest benefited the financial institutions, the safe harbor applied after all, despite the Merit Management by-pass.
Justice Sotomayor had addressed the words “or for the benefit of” in the Merit Management opinion:
The primary argument Merit [the fraudulent transfer defendant] advances that is moored in the statutory test concerns the 2006 addition of the parenthetical “(or for the benefit of)” to § 546(e). Merit contends that in adding the phrase “or for the benefit of” to the requirement that a transfer be “made by or to” a protected entity, Congress meant to abrogate the 1998 decision of the Court of Appeals for the Eleventh Circuit in In re Munford, Inc., 98 F.3d 604, 610 (per curiam) . . . . Congress abrogated Munford, Merit reasons, by use of the disjunctive “or,” so that even if a beneficial interest, i.e., a transfer “for the benefit of” a financial institution or other covered entity, is sufficient to trigger safe harbor protection, it is not necessary for the financial institution to have a beneficial interest in the transfer for the safe harbor to apply.[122]
Justice Sotomayor found a simpler explanation of the 2006 amendment. Congress just wanted to match up § 546(e) with § 550(a)(1). Thus, the words did not interfere with the principle of the overarching transfer. This amounts to an instruction that we are to read “for the benefit” out of § 546(e). If so, Merit Management stands for the proposition that the plain meaning of § 546(e) does not govern.
In Greektown, Papas/Gratzoros claimed that ML had greatly benefited from the transaction that got funds into Papas/Gratzoros pockets, ML retained a substantial portion of the notes that Holdings had issued, which carried a 10.75% interest rate. ML received millions in fees. Accordingly, Papas/Gratzoros argued, they were safe-harbored.
Judge Oxholm could have rested with the observation that, per the Supreme Court, the language did not abrogate the principle of the overarching transfer. But she went further. Relying on Reily v. Kapila (In re International Management Associates.),[123] she held the benefit to ML was “unquantifiable.”[124] Basically, the reasoning of International Management was that “for the benefit of” referred to a paradigm case: the benefit a guarantor receives when the assured creditor is voidably paid. Anything else is too far from the paradigm. Judge Oxholm admitted that ML benefited from fees and the like. But such benefits were indirect, incidental, “unquantifiable,” and could be ignored. The holding illustrates that Congress erred in associating “or for the benefit of” with fraudulent transfers. The phrase should be cabined to voidable preference cases. The holding also indicates that the plain meaning of § 546(e) is not to be followed. Just because a financial institution received fees does not mean the LBO payout is safe-harbored, even though the language of § 546(e) “plainly” says otherwise.
VIII. The Safe Harbor for the LBO Lender
So far, the focus has been on safe-harboring the departing shareholders in an LBO. The Seventh Circuit, however, has extended the safe harbor to the LBO lender.
In Petr v. BMO Hariss Bank N.A.,[125] Sun Capital’s subsidiary Intermediate Holding wished to acquire privately held BWGS LLC (BWGS) from its shareholders.
To finance the purchase, Holding borrowed on an unsecured bridge loan basis from BMO, a financial institution. Later, BWGS borrowed from other lenders. BWGS used the proceeds to pay off BMO. BWGS soon went bankrupt. T sued BMO for receiving a fraudulent transfer.
Digressing to the merits, BWGS should be viewed as dividending up the loan proceeds to Holding, who paid its creditor BMO. BMO was therefore a transferee of a transferee of the dividend. BMO was liable only if the dividend to Holding was a fraudulent transfer and BMO had knowledge of this fact. Since Holdings’ only asset was BWGS shares, Holdings received a fraudulent transfer only if BWGS was insolvent at the time of the dividend or made insolvent by it.[126]
Judge Amy St. Eve viewed the transaction differently. Reading T’s complaint in the most favorable light to T, she viewed BMO as the initial transferee from BWGS. “The Transfer thus relieved Intermediate Holding and Sun Capital of their obligations under the bridge loan. BWGS received no value from the transfer.”[127] But so viewed, BWGS did receive value. As a result of this transaction, BWGS would be subrogated to the right of BMO to collect from Holding. If Holding was solvent at the time of subrogation, BWGS received a reasonably equivalent value from BMO. [128]The Seventh Circuit stifled this inquiry by ruling that the (indirect) transfer to BMO was a transfer in connection with a securities contract. Since BMO was a financial institution, it could claim immunity from T’s fraudulent transfer cause of action. This ruling exploits the 2006 amendment to § 546(e). Prior to 2006, an immunized transfer had to be a settlement payment under a securities contract. BMO, however, was not buying securities and so would not have been eligible for the harbor. But after 2006, a financial institution could claim immunity if it received a transfer of some other sort that was “in connection” with a securities contract.
The bankruptcy court had ruled that the bridge loan agreement was not a securities contract. Only the agreement between the shareholders and Holding was. Financing a settlement payment lacked a “sufficient material nexus” to the SH-Holdings agreement.[129] In other words, the payoff of the bridge loan was connected to a securities contract, but not sufficiently so.
Judge St. Eve ruled that the bridge loan was itself a securities contract. According to § 741(7)(a)(v) a securities contract includes “any extension of credit for the clearance or settlement of securities transactions.” BMO had not received a settlement payment. BMO had not sold stock. Its payoff, however, was “in connection with a securities contract,”[130] because financing a securities contract is a securities contract.
The implication of Petr is that the LBO lender’s agreement to finance the LBO is itself a securities contract. If the LBO lender is a financial institution, and if the LBO lender receives security interests on DC’s assets, the LBO lender is entitled to the safe harbor. Petr therefore completes the picture. Not only are the shareholders safe-harbored when their transferee is a customer of a bank, but the LBO lender is harbored as well (when it is a bank or perhaps a customer of a bank).
IX. Avoiding Transfers v. Avoiding Obligations
Everyone forgets that fraudulent transfer law does not merely avoid transfers. It also avoids obligations. Redacting “transfer” out so that “obligation” can shine forth, § 548(a)(1) provides
The trustee may avoid . . . any obligation . . . incurred by the debtor that was . . . incurred on or within 2 years before the date of the filing of the petition, if the debtor . . .
(A) . . . incurred such obligation with actual intent to hinder, delay, or defraud any entity to which the debtor was or becomes, on or after the date that such obligation was incurred, indebted; or
(B)(i) received less than a reasonably equivalent value in exchange for such . . . obligation; and was
(ii) [insolvent or some enumerated proxy therefor].[131]
Section 546(e) prohibits T from avoiding transfers. It says nothing about avoiding obligations. Can T challenging an LBO exploit this oversight and avoid D Corp.’s obligation to pay for shares?
Avoiding a debt is an odd idea. Suppose X owns a gold brick worth $100 and delivers it to insolvent D who promises to pay $150 in the future. The obligation to pay is a valid contract, but the obligation is not for a reasonably equivalent value. Creditors may avoid it. But, armed with a money judgment against D, why would C1 ever seek to avoid D’s obligation? What C1 wants is assets to levy on (such as the brick). Canceling D’s obligation to pay does not help C1 get an asset of D. Suppose, however, before C1 can move, X sues on the debt and gets a money judgment against D. Pursuant to the money judgment, X serves an execution on the sheriff who levies the brick. C1 cannot attack X’s judicial lien directly. X has a money judgment to back it up. But C1 has the motive to “avoid” the obligation upon which the judicial lien is based,[132] provided “avoid” means “transfer to C1.” By avoiding, C1 becomes the assignee of the debt (to the extent of C1’s claim against D). The lien travels with the debt, so C1 now has a judicial lien against the brick. The lien, not itself a fraudulent transfer, is proceeds of the avoided obligation of D to pay for the brick.[133]
In a previous work, I claimed “avoidance” is a misleading term if it suggests erasure. Rather, to “avoid” means “transfer to the plaintiff.” “In actuality, fraudulent transfer law assigns the property of third parties to the defrauded creditors as security for their claims.”[134] This must be the case, because otherwise creditors with no avoidance rights benefit from erasure. This free ride impoverishes the creditor who actually does have avoidance rights as well as X, the fraudulent obligee.[135]
To illustrate the free ride, assume for simplicity that we are in a liquidation proceeding not governed by Moore v. Bay.[136] Probate of an insolvent estate serves as an example. Suppose C1 has a claim for $100 against D’s estate and an avoidance right against X because D has extended a fraudulent obligation to X. The amount of X’s claim against D is $100. Assume C2-4 claim $100 each, but they have no avoidance claim against X. Perhaps they are blocked by a statute of limitations whereas C1 is not. Together, C1-4 and X claim $500 from D. The estate of D owns $375 in unencumbered assets. Without avoidance of the fraudulent obligation, C1-4 and X receive a pro rata dividend of $75. If we simply erased X’s claim, C1-4 each receive $93.75. This means C2-4 have gained $12.75 because C1 has an avoidance right. Erasure allows C2-4 to “free ride” off C1’s avoidance right.
Suppose instead we say that X’s claim is not erased. Instead, it is deemed assigned from X to C1. In that case, C2-4 receive $75—exactly what they would have received if C1 had no avoidance right. Meanwhile, X’s right to a $75 dividend is payable to C1 as X’s assignee. C1 receives $75 in her own right and $25 of the dividend otherwise payable to X. C1 has therefore received $100. Since, under the UFTA, C1 is only entitled to “avoidance of the . . . obligation to the extent necessary to satisfy the creditor’s claim,”[137] there is a $50 surplus to which X is entitled. Thus, X receives $50 on her unsecured claim against D’s estate.
In the end, C2-4 receive $75—exactly what they would have got if C1 had no avoidance right. C1 gets $75 in her own right plus $25 of X’s dividend. X gets the surplus of $50. Thus, voidability of an obligation is not erasure of it. Otherwise C2-4 enjoy a free ride at C1’s expense and X’s expense. Avoidance is a nonrecourse assignment for security from X to C1. In short, X is subordinated to C1, but not to C2-4.[138] X deserves the surplus once C1 is paid out. As to C2-4, they simply were not defrauded n any actionable way.
Does Moore v. Bay [139] (favorably noted in the legislative history to the Bankruptcy Code)[140] negate avoidance-as-transfer and substitute avoidance-as-erasure? It does not. Moore (which is nothing, but an Erie-Thompkins guess as to the content of state law)[141] is consistent with avoidance-as-transfer, with the proviso that the transfer is at X’s expense and in favor of T. T is assigned the contractual obligation and takes the judicial lien as proceeds of the obligation. T then liquidates the judicial lien and distributes the money received according to the priorities set forth in Bankruptcy Code § 726(a).
In Geltzer v. Mooney (In re MacMenamin’s Grill Ltd.),[142] Judge Drain recognized that avoiding an obligation is not the same as avoiding a transfer:
There is clearly a difference between making a transfer and incurring an obligation; otherwise, the relevant statutory provisions would not have used both terms . . . Bankruptcy Code section 101(54) broadly defines “transfer” to include “each mode, direct or indirect, absolute or conditional, voluntary or involuntary, of disposing of or parting with—(i) property or (ii) an interest in property . . . but the definition does include the incurrence of an obligation. A “transfer”—even under the exceedingly broad Bankruptcy Code definition—is ultimately a disposition of property. If avoided, a transfer is preserved under Bankruptcy Code section 551 and recoverable under Bankruptcy Code section 550, while the avoidance of an obligation, which is not mentioned in Bankruptcy Code sections 550 and 551, instead reduces dollar for dollar the claims that the estate must pay; there is nothing to preserve or bring back.[143]
Judge Drain correctly observed that Bankruptcy Code § 550(a) provides for recovering transferred property or its value but says nothing about recovering fraudulent obligations. Section 551 preserves avoided transfers but says nothing about preserving avoided obligations. In enacting §§ 550(a) and 551, Congress obviously forgot about avoided obligations. It would forget again when it enacted § 546(e) in 1982.[144] Yet avoidance of an obligation must mean something. Judge Drain assumed the avoidance means erasure. But our reference to state law suggests that avoidance means transfer. T (as creditor representative) is the assignee of the avoided obligation. On behalf of the other creditors, T asserts the avoided claim against D Corp. and adds the resultant dividend to the bankruptcy estate, where the legitimate creditors share according to their § 726(a) priority.
It is possible from this example to concoct a theory whereby T could seek to avoid a securities contract and therefore be heir to proceeds of that contract. Under this theory, the core mission of § 546(e)—shielding the securities clearance mechanism from fraudulent transfer liability—is severely compromised. T could sue clearing houses for giving or receiving proceeds of voidable obligations. This becomes a decisive reason to reject the theory. A court could easily find that Congress was merely inattentive to the transfer-obligation distinction[145] and would, if it could re-legislate, add the words “or avoid an obligation” to § 546(e). Such a conclusion does not quite concede that the plain meaning movement for interpreting the Bankruptcy Code is dead. Plain meaning is usually coupled with an “unless absurd” proviso.[146] Rejecting the theory about to be rehearsed is simply the absurdity proviso at work.
According to this theory which we have rejected in advance, the trustee may, consistent with § 546(e), avoid a fraudulent obligation which is a securities contract. Obligation avoidance does not mean there is no securities contract. Avoidance means that T is the assignee of that contract and is entitled to the proceeds of the contract. Ownership of the proceeds is not achieved by avoiding a transfer in connection with the securities contract. Ownership is achieved by avoiding the obligation of D Corp. under the securities contract. The avoided contract becomes T’s property. The actual shareholders in the LBO who received the funds in the avoided securities contract are holding these funds in constructive trust for T’s benefit. This argument strikes very hard at the core mission of § 546(e).[147]
In Lehman Bros. Holdings v. JPMorgan Chase Bank, N.A. (In re Lehman Bros. Holdings),[148] Judge James M. Peck rejected the argument. In the case, a CC sued JPMorgan for constructive fraudulent obligations received in the course of JPMorgan’s rendering of clearing services to Lehman. Judge Peck found the relevant contracts between the parties were securities contracts. As to whether the CC could avoid the obligations underlying the transfer of security interests to JPMorgan, Judge Peck remarked:
Because the language of section 546(e) as written includes no express references to the incurrence of obligations, Plaintiffs are correct that the incurrence of obligations is not exempt from avoidance . . . . Plaintiffs are not correct that this notional ability to assert that an obligation is not exempt from avoidance is an acceptable means to whittle away at or undermine the effectiveness of the safe harbors. Despite the linguistic exercise, the safe harbors still protect the transactions between Lehman and JPMC.
The exclusion of “obligations” from the statutory exemption, thus, becomes something of a Pyrrhic victory for [the CC]. The Guarantees [i.e., the securities contracts] are not transfers themselves, yet they are resistant to successful challenge because they connect so directly to transfers that are exempt and beyond reach. A transfer made in connection with a securities contract remains unavoidable regardless of whether the Guarantees could potentially be avoided.[149]
With the proviso that the CC’s position is wrong because it is absurd, Judge Peck’s analysis does not function. The security interests assigned to JPMorgan were not fraudulent transfers if they were pursuant to a valid guaranty. But if the guaranties were fraudulent obligations, the obligation was instantly assigned by state law to the unsecured creditors of Lehman. Since they own JPMorgan’s rights under the guaranty, they also own the proceeds of the guaranty. This follows even if the transfer is left unavoided.
Judge Peck was closer to the truth when he remarked:
The choice of language in [§ 546(e)] may reflect an intentional decision by Congress to differentiate between termination statement and obligations . . . . It is difficult to know what Congress actually meant with any confidence, although given the purposes of section 546(e) to immunize the markets from certain bankruptcy risks and the seemingly boundless definition of the term “transfer,” this may be an example of a word that is supposed to transcend its ordinary meaning to include the incurrence of obligations.[150]
In other words, when Congress uttered § 546(e), it meant “transfers and obligations” when it said “transfers.” Anything else would be absurd.
In MacMenamin’s Grill,[151] Judge Drain considered whether an LBO lender might claim the harbor. Up to now, we have mostly considered whether the departing shareholders are harbored because they received settlement payments. But we have seen that in 2006, Congress extended the safe harbor to include any transfer by or to a financial institution that is merely connected to a securities contract.[152] And we have seen that the Seventh Circuit in Petr v. BMO Harris Bank, N.A.[153] used this principle to shelter the rights of LBO lenders. In this context, Judge Drain affirmed that T could indeed sue the bank on the theory that the LBO loan was a fraudulent obligation. Like Judge Peck, Judge Drain assumed that the mortgage would still be protected, even though the underlying obligation was not.
Applying section 546(e)’s plain terms to exclude the avoidance of obligations would not render the exemption completely meaningless, however, for parties like the Lender—far from it. For example, any payments and any lien that the Lender received in connection with the securities contract would not be avoidable, because they were transfers.[154]
But this misses the point. When the loan agreement funding the LBO is proclaimed a fraudulent obligation, the loan agreement is assigned to the creditors. The mortgage (perfectly valid to secure the loan) must follow the loan.[155] Since the creditors own the loan receivable, they own the mortgages; any paydown received by the lender is held in trust for the creditors to whom the loan agreement had been assigned.
Although CC’s claim in Lehman was grammatically correct but wrong on absurdity grounds, the matter should give pause to the § 546(e) plain-meaning advocates who insist that § 546(e) covers LBOs in privately held stock.[156] Courts holding that § 546(e) covers closely held stock transactions simultaneously invoke plain meaning on that issue but also implicitly assume that allowing the LBO obligation to be avoided is an absurdity. Accordingly, plain meaning of § 546(e) never governs in any of these cases. Nevertheless, Congress had an idea when it enacted the language of § 546(e). We should consult the idea behind Congress’s instruction, not the faulty language in which the idea was expressed
Conclusion
The Supreme Court in Merit Management Group, LP v. FTI Consulting, Inc.,[157] thought it was ending the safe harbor for LBO payouts, but the Second Circuit in Deutsche Bank Trust Co. Americas v. Large Private Beneficial Owners (In re Tribune Co. Fraudulent Conveyance Litigation),[158] completely cut the ground out from under the Supreme Court ruling. The means of doing so was the definition of “financial institution.” Section 546(e) indicates that a bankruptcy trustee may not avoid transfers by or to a financial institution in connection with a securities contract. But the Bankruptcy Code defines “financial institution” to include the customer of a financial institution when the financial institution acts as agent. Thus, any LBO payout by a customer to shareholders is safe-harbored.
In this article I have argued that the plain meaning of § 546(e) is completely absurd. Courts relying on the plain meaning of § 546(e) in LBO cases ignore the plain meaning when they do not permit trustees to avoid LBO obligations instead of LBO transfers. Therefore, courts should give up on trying to read § 546(e) closely. Instead, they should consult the policy behind § 546(e). The policy is to avoid cascading bankruptcies of financial institutions involved in the securities market clearing system. Protecting former shareholders from fraudulent transfer liability in LBO cases, especially when the corporate shares are not publicly traded, is not even remotely necessary to keep securities exchanges humming. Nor is protecting banks who finance LBOs. “[G]ranting a safe harbor to a constructive fraudulent private stock sale has little if anything to do with Congress’ stated purpose in enacting section 546(e): reducing systemic risk to the financial markets.”[159]
***
* Professor of Law, Cardozo Law School.