chevron-down Created with Sketch Beta.

The Business Lawyer

Fall 2023 | Volume 78, Issue 4

Contracting for Default

Wm Robertson Dorsett

Contracting for Default Manustrong

Jump to:


This Article addresses two issues with Credit Default Swaps (CDS). First, though bankruptcy courts have correctly intuited that CDS may dictate creditor incentives, they have not fully appreciated how. Disclosure requirements in Chapter 11 are inconsistent with the actual mechanics of CDS, and therefore are unlikely to capture many of the transactions with which courts should be concerned. Second, though CDS function as synthetic debt instruments for performing credits, the economic payouts of CDS and cash bonds can deviate significantly and unpredictably for non-performing credits. To the extent that market participants deem this breakdown in equivalency a defect, rather than a feature, changes should be made to CDS’ governing language—ISDA’s Credit Derivatives Definitions.

In addressing these issues, this Article aims to provide bankruptcy practitioners, market participants, and academics with the tools to analyze CDS, “engineered” transactions, and the implications for (i) regulatory disclosure requirements inside and outside of bankruptcy and (ii) the governing ISDA Credit Derivatives Definitions. First, I provide a detailed primer on the mechanics of CDS, paying particular attention to the auction process. Second, I analyze illustrative historical transactions, including Europcar, Codere, Hovnanian, iHeart, Supervalu, and Windstream. Third, I analyze regulatory and private contractual responses, including (i) the SEC’s recently proposed Rule 10B-1 and finalized Rule 9j-1, (ii) the 2019 ISDA Amendments, and (iii) Windstream provisions. Fourth, I analyze the relevance of engineered transactions to the “good faith” and “notice and hearing” requirements of the Bankruptcy Code. Fifth, I propose responsive amendments to (i) the Bankruptcy Rules’ disclosure requirement and (ii) ISDA’s Credit Derivatives Definitions.

I. Credit Default Swap Primer

A credit default swap (CDS) can be characterized as an (i) an insurance contract, (ii) a synthetic debt instrument, or (iii) a sui generis credit derivative. The characterization (i) “insurance contract” is an oversimplification, (ii) “synthetic debt instrument” is an aspiration, and (iii) “sui generis credit derivative” is a reality. This deviation in cash-derivative equivalency is a function of CDS mechanics—namely, the limits of private contractual language and the auction process that determines CDS settlement value—which can result in a breakdown of equivalency between cash bonds and CDS derivatives. In particular, this breakdown in equivalency is due to the manner in which (i) the auction settlement is triggered, (ii) the set of debt obligations deliverable at auction is determined, and (iii) the auction settlement price is established. To limit the scope of the article, I restrict my analysis to the terms of North American, single-issuer, corporate CDS. This article does not analyze sovereign CDS or corporate CDS in other jurisdictions—e.g., Europe and Asia—which include Restructuring as a Credit Event.

A. Basic Contractual Terms of CDS

Whether a CDS is insurance, a synthetic debt instrument, or a sui generis credit derivative, it can more accurately be described by its contractual terms. By its terms, a CDS is a bilateral agreement referencing a debt-issuing corporate entity, wherein (i) the swap Buyer pays to the swap Seller a premium fee and, (ii) if a contractually defined Event of Default occurs, Seller must effectively pay Buyer the difference between par and the extant value—or Recovery Value—of the referenced entity’s debt obligation. In the case of physical settlement, Seller pays Buyer par value and Buyer delivers to Seller one of a predetermined set of corporate debt obligations. In the case of auction settlement—the current prevailing standard—Seller pays Buyer the difference between par value and the recovery value, as determined during a Dutch auction process. The auction process also allows for participants to deliver/receive physical bonds at the recovery value. Consequently, auction settlement allows for both (i) cash settlement and (ii) physical settlement—subject to market supply/demand limitations.

1. Counterparties—Buyer (Credit Short) and Seller (Credit Long)

There are two parties to a CDS contract: the CDS buyer and the CDS seller. Buyer has short credit exposure and profits from a deterioration in the creditworthiness of a contractually specified corporate debt issuer, the Reference Entity. Buyer is a Fixed Rate Payer, meaning that, prior to and absent an Event of Default, Buyer pays a fixed premium to Seller, quoted in annual terms and payable quarterly. Buyer may also pay/receive an upfront true-up fee to account for the difference between (i) a fixed standardized coupon rate and (ii) a bilaterally agreed credit spread. Subsequent to an Event of Default, Buyer receives the Auction Settlement Amount—the difference between par value and the price at which the auction settles.

Seller has long credit exposure and profits from an improvement in the creditworthiness of the Reference Entity. Seller is a Floating Rate Payer and, prior to and absent an Event of Default, makes no payment other than the receipt/payment of the upfront true-up fee. Subsequent to an Event of Default, Seller will pay the Auction Settlement Amount—the difference between par value and the price at which the auction settles.

2. Settlement Trigger–Credit Event Determination

The occurrence of a Credit Event triggers Seller’s payment obligations under the CDS contract. When a market participant believes a Credit Event has occurred, thus triggering CDS payouts, a notification must be sent to the geographically relevant International Swaps and Derivatives Association (ISDA) Determinations Committee (“DC”), along with publicly available information as evidence that the Credit Event occurred. The DCs consist of both buy-side and sell-side entities, and they have two business days to meet and determine—by a supermajority—whether a credit event has occurred. ISDA-defined credit events include “Bankruptcy, Failure to Pay, Obligation Acceleration, Obligation Default, Repudiation/Moratorium, Restructuring, or Governmental Intervention, as specified.” Standard North American contracts, however, include only Bankruptcy and Failure to Pay.

3. Credit Event: Bankruptcy

Bankruptcy occurs when an entity (i) is dissolved, (ii) becomes insolvent or unable to pay its debts, (iii) institutes or has instituted against it a legal bankruptcy process, or (iv) has substantially all of its assets taken by a secured creditor. For example, on May 9, 2022, Talen Energy Supply, LLC filed for Chapter 11 in the Southern District of Texas. On May 10, 2022, the DC accepted a request to determine whether a Credit Event—specifically, Bankruptcy—had occurred. On May 11, 2022, the DC confirmed that a Bankruptcy Credit Event had occurred on May 9, 2022.

4. Credit Event: Failure to Pay

Failure to Pay occurs when, subsequent to the expiration of any applicable Grace Period, the Reference Entity fails to make payments aggregating to a predetermined Payment Requirement (by default, USD $ lmm). For example, on May 29, 2020, California Resources Corporation missed certain interest payments and did not make such payments within the applicable five-day grace period. On June 10, 2020, the DC accepted a request to determine whether a Credit Event—specifically, a Failure to Pay—had occurred. On June 11, 2022, the DC determined that a Failure-to-Pay Credit Event had occurred on June 3, 2020.

B. Characterizing Credit Default Swaps

1. CDS as Insurance Contract

In a simplified sense, a CDS is an insurance contract. The CDS buyer can be said to purchase insurance on a specified notional value of a corporate debt obligation, the running cost of which is a fixed percentage of that notional value. For example, the CDS buyer might agree to pay 5 percent per annum to insure $1mm in par notional of a corporate bond. What the CDS buyer is insuring against is the possibility that the corporate bond might be repaid at less than par value—that is, at less than 100 cents on the dollar—due to the corporate debt issuer’s default. In the event that Corporate does default, CDS Buyer will receive a payment from CDS Seller that approximates the difference between the par value and the actual, post-default value of Corporate Bond. For example, if holders of Corporate Bond recover 45 cents for each dollar notional of bonds held, CDS Buyer should receive 100 cents less 45 cents, for a total of 55 cents, for each dollar notional of CDS held.

Note the timing mismatch between when the corporate debt–issuer defaults and when the final recovery rate on the bonds is determined. If the recovery—for the purposes of a CDS—must be determined at the time of default, it may not match the actual recovery that bond holders would receive in a Chapter 7 liquidation or Chapter 11 reorganization. Consequently, this “insurance model” is most accurate where CDS is settled “physically.” By settled “physically,” we mean that upon Corporate’s default, the following exchange happens:

  • CDS Buyer delivers Corporate Bond to CDS Seller. In this example, CDS Buyer delivers a bond then valued at 45 cents on the dollar to CDS Seller.
  • CDS Seller delivers par—100 cents on the dollar—to CDS Buyer.

In this way, if Credit Investor (i) owns Corporate Debt and (ii) buys CDS of equivalent notional, upon an event of default Credit Investor will still receive par value. However, as discussed infra, this insurance model breaks down where, inter alia, (i) multiple bonds are deliverable into a given CDS contract, (ii) there is a mismatch of duration between these bonds and CDS, and (iii) the impracticability of full physical settlement requires an auction process to estimate recovery value.

2. CDS as Synthetic Debt Instrument

In another sense, a CDS is a synthetic debt instrument. The CDS seller steps into the shoes of a corporate debt creditor and the CDS buyer steps into the shoes of the corporate debt issuer, with no initial exchange of principal. A CDS enables a bilateral exchange of credit exposure, in a manner that mirrors the cash-bond market, but provides greater flexibility as to duration, notional value, funding costs, and interest rate exposure. For example, banks may use a CDS to hedge the credit risk they hold against counterparties to whom they have provided financing, whether through a traditional loan or other bilateral agreement that results in counterparty liability. A CDS may also be used to reduce the credit risk component of hybrid securities, such as convertible bonds; to obtain credit exposure without exposure to interest rates; or to obtain credit exposure out to a maturity not met by the existing cash-bond market.

In a simplified world—where, inter alia, (i) actual bond recovery rates are equivalent to CDS recovery rates and (ii) funding and counterparty risks are removed—there is a fundamental relationship between the cost of a CDS (premium) and the credit spread of the Corporate Bond it references. Specifically, in this simplified world, the cost of a CDS equals the credit spread—in excess of the risk-free interest rate—that Corporate would pay Creditor for a bond of equivalent duration.

As discussed infra, this equivalency breaks down when considering the impact of, inter alia, funding costs, deliverability, and CDS auction mechanics. However, a simplified example where equivalency does hold is useful for understanding the ultimate breakdown in equivalency. Consider the following scenario:

  • Corporate has one five-year Corporate Bond outstanding, which was issued, is redeemable, and currently trades at par, with a 5% coupon, payable in quarterly installments.
  • There is one CDS contract available on the Corporate, with a 5y maturity, at a cost of X%. The CDS contract is physically settled.
  • The risk-free interest rate is zero, and therefore Corporate’s 5% coupon also reflects its Credit Spread.
  • There are no funding costs.
  • Creditor purchases $1mm notional of Corporate Bond, and CDS Buyer purchases $1mm notional of CDS.
  • All contracts are held to maturity, with no intermediate trading.

Those assumptions are reproduced in a table below:

  Bond Assumptions CDS Assumptions
Maturity 5y 5y
Issue Price Par (100c) N/A
Redemption Price Par (100c) N/A
Coupon 5% X%
  paid quarterly  
Settlement N/A Physical
Risk Free Interest Rate 0%
Funding Costs 0%

If there is no default, the following will occur:

  • Bond Creditor will (i) pay $1mm to Corporate at T(0); (ii) receive $50k per year, for five years; and (iii) receive $1mm in principal repayment at maturity at T(5); for a net nominal cash receipt of [$1mm*5%*5] = [$50k*5] = $250k.
  • Corporate (Bond Issuer) will (i) receive $1mm from Bond Creditor; (ii) pay $50k per year, for five years; and (iii) pay $1mm in principal repayment at maturity at T(5), for a net nominal cash payment of [$1mm*5%*5] = [$50k*5] = $250k.
  • CDS Seller will receive $1mm*X% per year, for a net nominal cash receipt of [$1mm*X%*5].
  • CDS Buyer will pay $1mm*X% per year, for a net nominal cash payment of [$1mm*X%*5].

If there is a default (“D”) at time T(D), for 0 < T(D) < 5y, the following will occur:

  • Bond Creditor will (i) pay $1mm to Corporate at T(0); (ii) receive $50k (5%) per year, up until T(D); and (iii) receive Recovery Value on Corporate Bond, as determined at the end of a Chapter 7 or Chapter 11 process, for a net nominal cash receipt of $1mm*[R–100% + 5%*T(D)].
  • Corporate (Bond Issuer) will (i) receive $1mm from Bond Creditor at T(0); (ii) pay $50k per year, up until T(D); and (iii) pay Recovery Value on Corporate Bond, as determined at the end of a Chapter 7 or Chapter 11 process, for a net nominal cash payment of (1mm)*[R - 100% + 5%*T(D)].
  • CDS Seller will (i) receive $1mm*X% per year, up until T(D); (ii) pay $1mm to CDS Buyer; (iii) receive Corporate Bond from CDS Buyer, and ultimately receive Recovery Value on Corporate Bond, as determined at the end of a Chapter 7 or 11 process, for a net nominal cash receipt of$1mm*[R–100% + X%*T(D)].
  • CDS Buyer will (i) pay $1mm*X% per year, up until T(D); (ii) receive $1mm from CDS Seller; and (iii) deliver Corporate Bond to CDS Seller, for a net nominal cash payment of ($1mm)*[R–100% + X%*T(D)].

The profit of these positions is reproduced in the table below:

Net Nominal Cash Receipt/(Payment)
No Default
Bond Creditor $1mm*5%*5 CDS Seller $1mmx%*5
Bond Issuer (Corporate) ($1mm*5%*5) CDS Buyer ($1mm*x%*5)
Bond Creditor $1mm*[R-100%+X%*T(D)] CDS Seller $1mm*[R-100%+X%*T(D)]
Bond Issuer (Corporate) ($1mm)*[R-100%+X%*T(D)] CDS Buyer ($1mm)*[R-100%+X%*T(D)]

In our simplified world, the “X%” above—the CDS Credit Spread—must be 5%. Otherwise, there would be an arbitrage opportunity to buy the “cheaper” asset, and sell the “richer” asset, for a risk-free profit that violates no-arbitrage constraints. Consequently, in this simplified world, the cost of CDS must equal the credit spread of the bond on which it insures. As discussed infra, this equivalency breaks down when considering the impact of, inter alia, funding costs, deliverability, tenor mismatch, and CDS auction mechanics.

3. CDS as Sui Generis Credit Derivative

In yet another sense, a CDS is a semi–sui generis credit derivative with a value that derives from—but does not fully replicate—the value of the debt obligations it references. Most notably, the CDS recovery rate—impacted by, inter alia, the manner in which (i) auction settlement is triggered, (ii) the set of debt obligations deliverable at auction is determined, and (iii) the auction clears—can differ significantly from the actual recovery rate of referenced debt obligations in a Chapter 7 liquidation or Chapter 11 reorganization. Additionally, funding costs, counterparty risk, and trading dynamics contribute to deviations from cash bond and derivative equivalency.

The sui generis nature of CDS is most evident in (i) the auction process—including the option of deliverability and two-step clearing mechanism—as described infra in Part I.C and (ii) historical deviations from cash-derivative equivalency, as described infra in Part II.

C. CDS Settlement—The Auction Process

1. Why an Auction?

When first traded, CDS contracts were physically settled, subsequent to the occurrence of a Credit Event. The CDS buyer would deliver the defaulted obligation to the CDS seller, and the CDS seller would deliver par value in cash to the CDS buyer. However, as the CDS market developed, so too did the need for a cash settlement mechanism because an increasing number of CDS participants neither owned nor wished to own the underlying debt obligation. This need for a cash settlement mechanism has led to the standardized CDS auction process, which enables CDS participants to choose either physical or cash settlement without having to ensure that they face an initial counterparty seeking the same method of settlement.

To see this, consider a universe without an auction process. In this universe, CDS Buyer A has entered into a CDS contract with CDS Seller A. If both Buyer A and Seller A want to settle physically, or if both want to settle in cash, there is no settlement issue. However, if Buyer A wants to settle physically and Seller A wants to settle in cash, there is a mismatch. To resolve this mismatch, Buyer A will need to find another counterparty—some Seller B—who also wants to settle physically. Then, Seller A and Seller B will need to agree to novate or otherwise exchange their positions so that Buyer A and Seller B become CDS counterparties. Similarly, Seller A will need to find another counterparty—some Buyer B—who wants to settle in cash. This process becomes increasingly complicated as the universe of CDS participants expands.

The auction process resolves this mismatch by providing a venue where CDS participants can effectively opt for either cash or physical settlement without having to locate a specific counterparty whose settlement preferences match their own. CDS buyers opting for cash settlement will receive cash equivalent to [Par] less [Recovery Value]. CDS buyers opting for full physical settlement will deliver bonds into the auction in notional value equivalent to their CDS and receive cash equivalent to (i) [Par] less [Recovery Value] on the CDS contract and (ii) [Recovery Value] on the physical bond sold at auction, for a total cash receipt of [Par].

Consequently, the auction process enables CDS participants to settle their full cash/derivative positions physically, or in cash, without having to ensure that they face a specific counterparty with matching settlement preferences. It also enables CDS participants to use a CDS either as a true hedge of physical debt obligations or as a derivative play involving no exchange of physical debt obligations. However, note that while physical settlement will always ensure net receipt of par value, unhedged cash settlement is entirely dependent on the price determined at auction. This can lead to volatile and unexpected outcomes, as demonstrated by Europcar, described in Part II.A.

2. Pre-Auction Determinations

a. Determination That an Auction Will Occur

Shortly after the determination of a Credit Event, the DC will determine whether to hold an auction. For North American contracts, where Bankruptcy or Failure to Pay are the potential credit events, an auction will be held so long as there are (i) deliverable obligations and (ii) “300 or more [CDS transactions] for which Auction Settlement is specified as the Settlement Method . . . which would be likely to be covered by one single set of [settlement terms] and to which five or more [sell-side market makers] are parties.” If an auction will be held, the DC will provide the relevant terms governing that auction—the Credit Derivatives Auction Settlement Terms.

In Talen, for example, the Americas DC determined on May 11, 2022, that a credit event—Bankruptcy—had occurred on May 9, 2022. On May 17, 2022, the DC voted to hold an auction on June 7, 2022, and provided the relevant terms that would govern the auction (the “Credit Derivatives Auction Settlement Terms”).

b. Determination of Deliverable Obligations

The DC will also determine which debt obligations are deliverable into the CDS auction. DC members propose an initial list; market participants may propose additional obligations and may oppose potential obligations. For example, in Talen, the DC published a preliminary list of Senior Secured and Unsecured Notes that would be deliverable in the upcoming auction. Subsequently, the DC published a notice that it lacked the necessary documentation for two of the notes included in the preliminary list, which were ultimately removed from the list of deliverables.

For North American CDS, there are few restrictions on what constitutes a deliverable obligation. In general, a deliverable obligation is one that, inter alia, (i) constitutes Borrowed Money, (ii) is of the correct seniority, (iii) matures within 30 years; (iv) is assignable; and (v) “has an Outstanding Principal Balance or Due and Payable Amount that is greater than zero.” Relevant to the Supervalu discussed below, for North American contracts, deliverable obligations include Qualifying Affiliate Guarantees, or guarantees provided by the Reference Entity (parent) for majority-owned affiliate (subsidiary) debt (“Downstream Affiliates”). Consequently, if Parent defaults, the Subsidiary-level obligations guaranteed by Parent would be deliverable into Parent CDS. Additionally, if Subsidiary defaults, and Parent does not honor its guarantee on Subsidiary-level obligations—causing a Credit Event at both Subsidiary and Parent entity—the Subsidiary-level obligations guaranteed by Parent would be deliverable into Parent CDS. However, if Subsidiary defaults, Parent-level debt guaranteed by Subsidiary would not be deliverable into Subsidiary CDS—only Subsidiary-level debt would be deliverable.

The value of the cheapest available deliverable obligation (“cheapest to deliver”) drives the auction process. It is in the best interest of the seller of bonds at auction to deliver the bond cheapest to acquire; it is in the best interest of the buyer of bonds at auction to price her bid in expectation of receipt of the bond cheapest to acquire. Consequently, determination of what does, and does not, constitute a deliverable obligation is of paramount importance to whether CDS valuations deviate significantly from cash-bond valuations.

c. Determination of Auction Participants

The DC will publish a list of Participating Bidders, which consist of DC member dealers—sell-side market makers—and non-DC dealers whose formal Participating Bidder Letter has been approved. These Participating Bidders will participate in step one of the auction process.

The auction process will produce an Auction Final Price applicable to all CDS transactions. Those market participants wishing to settle physically—that is, CDS buyers wishing to sell bonds and CDS sellers wishing to buy bonds—must submit a formal request through one of the Participating Bidders above. This request may not exceed the size of their CDS positions. For example, a holder of $20mm notional CDS may submit a request to sell up to—but no more than—$20mm notional of deliverable obligations. All participants have an opportunity to buy/sell bonds in step two of the auction process, in order to enable settlement of the Net Open Interest after buy/sell requests from step one are matched.

Upon the conclusion of an auction, CDS Seller pays to CDS Buyer the notional value of the CDS multiplied by the percentage difference between the Reference Price and the price determined at auction—the Auction Final Price. The method by which the Auction Final Price is determined—through steps one and two of the auction—is described below.

3. Auction Step One

a. Determination of Initial Market Midpoint

In the first phase of the auction, each Participating Bidder may submit Physical Settlement Requests and must submit an Initial Market Submission. An Initial Market Submission consists of an Initial Market Bid and an Initial Market Offer for an “Initial Market Quotation Amount” (that is, a two-way price for a set notional value). Notably, the price is expressed as a percentage of the Outstanding Principal Balance—usually, but not necessarily, par. The bid and offer cannot cross (i.e., the bid cannot be higher than the offer) and cannot differ by more than the Maximum Initial Market Bid-Offer Spread. In the case of Talen, for example, the Initial Market Quotation Amount was set at USD $2,000,000 and the Maximum Initial Market-Bid-Offer Spread at 2%.

Subsequent to dealers’ submissions, the Initial Market Midpoint (IMM) is then determined. First, to create Matched Markets, the Initial Market Bids are sorted in descending order and the Initial Market Offers in ascending order, with the highest bid matched against the lowest offer. For example, assume the Initial Market Submissions are as follows:

Market IM Bid IM Offer Spread
1 39.50% 41.00% 1.50%
2 40.00% 42.00% 2.00%
3 41.00% 43.00% 2.00%
4 45.00% 47.00% 2.00%
5 32.00% 34.00% 2.00%
6 38.75% 40.00% 1.25%
7 38.00% 39.50% 1.50%
8 41.00% 42.75% 1.75%

The Matched Markets will be the following:

IM Bid Rank IM Offer Rank IM Bid IM Offer Spread
1 8 45.00% 34.00% -11.00%
2 7 41.00% 39.50% -1.50%
2 6 41.00% 40.00% -1.00%
4 5 40.00% 41.00% 1.00%
5 4 39.50% 42.00% 2.50%
6 3 38.75% 42.75% 4.00%
7 2 38.00% 43.00% 5.00%
8 1 32.00% 47.00% 15.00%

The Matched Markets are then divided into Tradeable and Non-Tradeable markets. These terms may be misleading to a non-markets audience: Tradeable markets are those where the matched bid exceeds the matched offer, i.e., crossed markets where an arbitrage opportunity would exist. Conversely, Non-Tradeable markets are those where the matched offer exceeds the matched bid, i.e., normal markets. The Tradeable/Non-Tradeable divide between these sample markets would be as follows:

IM Bid Rank IM Offer Rank IM Bid IM Offer Spread Tradeable/Non-Tradeable
1 8 45.00% 34.00% -11.00% Tradeable
2 7 41.00% 39.50% -1.50% Tradeable
2 6 41.00% 40.00% -1.00% Tradeable
4 5 40.00% 41.00% 1.00% Non-Tradeable
5 4 39.50% 42.00% 2.50% Non-Tradeable
6 3 38.75% 42.75% 4.00% Non-Tradeable
7 2 38.00% 43.00% 5.00% Non-Tradeable
8 1 32.00% 47.00% 15.00% Non-Tradeable

The Tradeable Markets are subsequently discarded, and the participating dealers who submitted them are assessed a monetary penalty. At this point in the auction, only Non-Tradeable Markets remain:

IM Bid IM Offer Spread Tradeable/Non-Tradeable
40.00% 41.00% 1.00% Non-Tradeable
39.50% 42.00% 2.50% Non-Tradeable
38.75% 42.75% 4.00% Non-Tradeable
38.00% 43.00% 5.00% Non-Tradeable
32.00% 47.00% 15.00% Non-Tradeable

The Best Half of the Non-Tradeable markets is then determined, consisting of the half of the remaining Non-Tradeable markets with the tightest spread—that is, the smallest difference between bid and offer. Here, as there are five Non-Tradeable Markets, we take the tightest three spreads; the following Best Half of Non-Tradeable Markets remain:

IM Bid IM Offer Spread Tradeable/Non-Tradeable
40.00% 41.00% 1.00% Non-Tradeable
39.50% 42.00% 2.50% Non-Tradeable
38.75% 42.75% 4.00% Non-Tradeable

The average of each remaining bid and offer—here, six prices in total—is then calculated and rounded to the nearest one-eighth of one percentage point to determine the Initial Market Midpoint. Here, averaging these six numbers (40%, 41%, 39.5%, 42%, 38.75%, 42.75%) yields 40.67%, which, rounded to the nearest eighth of a percentage point, is 40.625%—the Initial Market Midpoint in this example. Importantly, this process of removing crossed markets and tightening bid offers has effectively removed “off-market” low bids and “off-market” high offers.

In the Talen auction, the initial markets made were as follows:

Dealer Bid Offer
Bank of America 58.00 60.00
Barclays 57.00 59.00
BNP Paribas 58.50 60.50
Citigroup 59.25 61.25
Credit Suisse 57.50 59.50
Deutsche Bank 59.50 61.50
Goldman Sachs 58.50 60.50
J.P. Morgan Securities LLC 58.00 60.00
Morgan Stanley 56.50 58.50
Société Générale 57.50 59.50

Upon reordering with bids in descending order and offers in ascending order, two Tradeable Markets were removed:

Matched Bids Matched Offers  
59.50 58.50 Tradeable
59.25 59.00 Tradeable
58.50 59.50 Non-Tradeable
58.50 59.50 Non-Tradeable
58.00 60.00 Non-Tradeable
58.00 60.00 Non-Tradeable
57.50 60.00 Non-Tradeable
57.50 60.50 Non-Tradeable
57.00 61.25 Non-Tradeable
56.50 61.50 Non-Tradeable

The Best Half of these Non-Tradeable Markets was determined as follows:

Non-Tradeable Bids Non-Tradeable Offers Spread  
58.50 59.50 1.00 Best Half
58.50 59.50 1.00 Best Half
58.00 60.00 2.00 Best Half
58.00 60.00 2.00 Best Half
57.50 60.50 3.00  
57.50 60.50 3.00  
57.00 61.25 4.25  
56.50 61.25 5.00  

Finally, the average of these eight remaining bids (four bids, four offers) yielded an Initial Market Midpoint of 59.

b. Determination of Net Open Interest

After the IMM is determined, those buy/sell Physical Settlement Requests which can be matched are, and the net Open Interest is calculated. For example, if there are $100mm in sell orders and $120mm in buy orders, all $100mm in sell orders will be matched and $20mm in net Open Interest to buy will remain. This $100mm of matched orders will trade at the Auction Final Price.

For example, in Talen, the set of Physical Settlement Requests placed through dealers was as follows:

Physical Settlement Requests
Dealer Bid/Offer Size
BNP Paribas Offer 0
Credit Suisse Offer 10.647
Deutsche Bank Offer 0
J.P. Morgan Securities LLC Offer 0
Morgan Stanley Bid 4.136
Bank of America Bid 10
Barclays Bid 5
Citigroup Bid 22
Goldman Sachs Bid 81.831
Societe Generale Bid 2
Total Bids 120.831
Total Offers 14.783
Net Interest to Buy 106.048

As there were $120.831mm in notional buy requests and only $14.783mm in notional sell requests, all sell requests were matched. The remaining net Open Interest to buy was the difference between those requests, $106.048mm.

c. Determination of Adjustment Amount

Once the net Open Interest is determined, dealers who submitted Tradeable (discarded) bids/offers opposite to the direction of the net interest are assessed a penalty—the Adjustment Amount. If the net Open Interest is to buy, for instance, dealers who made Tradeable offers are penalized an amount equal to the difference between their respective Tradeable offers and the Initial Market Midpoint multiplied by the Initial Market Quotation Amount. In Talen, for example, where (i) net Open Interest was to buy, (ii) the Initial Market Midpoint was 59, (iii) Morgan Stanley submitted an offer of 58.5, and (iv) the Initial Market Quotation was $2mm, Morgan Stanley was assessed an adjustment amount of [58.5-59]*[$2mm] = $10,000.

d. Publication of Step One Results

Within thirty minutes of the closing of step one of the auction, the Open Interest, Initial Market Midpoint, and Adjustment Amounts are published. The second part of the auction “typically takes place two to three hours later.”

4. Auction Step Two

a. Customer Limit Order Submission

If the net Open Interest is zero, there is no subsequent bidding period, and the Auction Final Price will be equal to the Initial Market Midpoint calculated above. If the net Open Interest is non-zero, a Subsequent Bidding Period will be held.

During the period between the publication of the Initial Bidding Information and the conclusion of the Subsequent Bidding Period, market participants may submit a Customer Limit Order—a limit order opposite to the side of the net Open Interest. For example, if net Open Interest is to sell, market participants may submit orders to buy specified notional values at specified limit prices.

Note also that the Initial Market Bids or Offers—opposite to the direction of net Open Interest—are carried through to step two of the auction at the price initially bid/offered. Any Initial Market Bid or Offer discarded as a Tradeable Market in determining the IMM, however, will be carried through at the IMM.

Customer Limit Orders are subject to two further restrictions. First, each Customer Limit Order must be “to the best of the customer’s knowledge and belief, [sic] (when aggregated with all other Customer Limit Order Submissions, if any, submitted by such customer to one or more Participating Bidders . . . ) not in excess of the size of the Open Interest.” For example, if the net Open Interest is to sell $10mm notional, a market participant may not submit a bid for—or multiple bids aggregating to—$11mm notional.

Second, each Customer Limit Order will be limited by a one-sided Cap Amount. If the net Open Interest is to sell, a limit bid must be no greater than the [Initial Market Midpoint] plus [Cap Amount]. Conversely, if the net Open Interest is to bid, a limit offer must be no less than the [Initial Market Midpoint] less [Cap Amount]. The Cap Amount is equal to half the Maximum Initial Market Bid-Offer Spread—the maximum spread allowed in step one of the auction. Importantly, the cap is one-sided. If net Open Interest is to bid, limit offers may not be less than the Initial Market Midpoint less Cap Amount but may be as high as par. Conversely, if net Open Interest is to sell, limit bids may not be higher than the Initial Market Midpoint plus Cap Amount but may be as low as zero.

In Talen, for example, (i) the net Open Interest was $106.048mm to buy; (ii) the Initial Market Midpoint was 59%; and (iii) the Maximum Initial Market Bid-Offer Spread was 2%. Consequently, market participants could each submit offers to sell up to $106.048mm in bonds, at offer prices greater than or equal to [59%–(1/2*2%)] = 58%.

The one-sided cap ensures that, whatever the dynamics of step two of the auction, the Auction Final Price reflects the direction of net Open Interest. For example, if the net Open Interest is to buy, the Auction Final Price will be no less than one percentage point below the IMM. Thus, even if there is significant selling interest in step two of the auction, the final price will not be driven down far below the IMM.

However, this one-sided cap amount could also lead to a “squeeze” in step two of the auction. If net Open Interest is not matched by market orders in step two, the Auction Final Price can squeeze all the way up to par or down to zero. For example, if net Open Interest from step one is to buy $100mm, and there are only $50mm of aggregate sell orders in step two, the Auction Final Price will be par. Such a squeeze happened in the Europcar case, as discussed below.

b. Matching Open Interest Against Unmatched Limit Orders

Once the Limit Orders are received, a Dutch auction is run to match the Open Interest bids/offers against the (as yet) Unmatched Limit Order offers/bids, resulting in a single price for all transactions. If the Open Interest is to buy, for example, the Unmatched Limit Order offers will be ranked in ascending order, and matched to the extent possible, creating Matched Limit Orders. Offers/Bids at the highest/lowest price at which there is a possibility for a Matched Order are filled pro rata, if necessary.

If the full Open Interest has been met, the Auction Final Price will be the highest Limit Order Offer/lowest Limit Order Bid at which a Matched Order was created, limited by the Cap Amount, and also limited by par and zero. If there is no Open Interest to begin with, the Auction Final Price will simply be the IMM from step one of the auction. However, if the Open Interest is not filled in step two of the auction, then the Auction Final Price will be set at either 100—if Open Interest is to buy—or 0—if Open Interest to sell. In this way, a lack of liquidity can drive a squeeze in the ultimate settlement price.

Consider the results of step two of the auction in Talen. In Talen, the following were established in step one:

  • Net Open Interest of $106.048mm to buy.
  • IMM of 59.
  • Cap Amount of 1%, implying Auction Final Price in the range of [58, 100].

The Dutch auction results were as follows:

Talen Step 2 Auction Results cont'd (2)
Dealer Offer Size (mm) Cumulative Matched Orders Dealer Offer Size (mm)
Barclays 58.0* 10 10 Credit Suisse 78.5 3
Barclays 58.0* 5 15 Goldman Sachs 79 5
Barclays 58.0* 5 20 Goldman Sachs 80 3
Citigroup 58.0* 4 24 Credit Suisse 80 3
Morgan Stanley** 59.0* 2 26 Credit Suisse 81.5 3
Barclays** 59.0* 2 28 Credit Suisse 83 3
Societe Generale 59.5* 2 30 Goldman Sachs 84 5
Credit Suisse** 59.5* 2 32 Barclays 85 5
Citigroup 60.0* 2 34 Goldman Sachs 85 3
Bank of America 60.0* 2 34 Goldman Sachs 90 10
J.P. Morgan Securities LLC** 60.0* 2 38 Barclays 90 5
BNP Paribas** 60.5* 2 40 Barclays 95 5
Goldman Sachs** 60.5* 2 42 Barclays 100 25
Citigroup 61.0* 2 44 Societe Generale 103 2
Citigroup** 61.25* 2 46 Goldman Sachs 105 5
Deutsche Bank** 61.5* 2 48      
Citigroup 62.0* 3 51      
J.P. Morgan Securities Inc. 62.0* 0.77 51.77      
Citigroup 62.5* 3.5 55.27      
Citigroup 63.0* 3 58.27      
Bank of America 63.25* 2 60.27      
Citigroup 64.0* 3.5 63.77      
Citigroup 64.0* 3 66.77      
Goldman Sachs 64.0* 2 68.77      
Citigroup 65.0* 3 71.77      
Goldman Sachs 65.0* 2 73.77      
Bank  of America 65.5* 2 75.77      
Citigroup 66.0* 3 78.77      
Goldman Sachs 66.0* 1 79.77      
Goldman Sachs 67.0* 1 80.77      
Citigroup 67.5* 4.519 85.289      
Goldman Sachs 68.0* 1 86.289      
Goldman Sachs 69.0* 1 87.289      
J.P. Morgan Securities LLC** Citigroup  Credit Suisse 69.875* 5 92.289      
70.0^ 10 106.048 (pro rata)      
70.0^ 3      
Citigroup 70.0^ 3      
Goldman Sachs 70.0^ 1        
Credit Suisse 71 3        
Goldman Sachs 71 1        
Goldman Sachs 72 1        
Credit Suisse 72.5 3        
Citigroup 74 14        
Goldman Sachs 74 5        
Credit Suisse 74 3        
Goldman Sachs 75 5        
Credit Suisse 75.5 3        
Credit Suisse 77 3        
**denotes Limit Orders carried over from Step 1 initial Market
* denotes filled Limit Orders
^ denotes pro rata, partially filled Limit Orders
Bolded Offer denotes Auction Final Price
Itaclized Markets were Unmatched

As above, while the IMM from step one of the auction was 59, the Auction Final Price from step two of the auction was set 11pts higher at 70. Consequently, CDS positions subject to Auction Settlement would be settled by a final payment from CDS Seller to CDS Buyer of [100-70]% = 30% multiplied by the contracted notional value.

c. Auction Settlement

Matched Buy and Sell requests for physical obligation settlement will be settled at the Auction Final Price. Where the net Open Interest is fully matched, all bids/offers on the side of the net Open Interest will settle at the Auction Final Price against those Limit Order offers/bids that were matched in step two of the auction. Where the net Open Interest is not fully matched, however, all bids/offers on the side of the net Open Interest will settle at 100/0 against all Limit Order offers/bids. The CDS contracts will settle based on the Auction Final Price, with CDS Seller paying CDS Buyer the notional contracted amount multiplied by 100 less the Auction Final Price.

II. Historical Deviations from Cash/Derivative Equivalency

In several historical instances, the potential for a breakdown in equivalency between cash bonds and derivatives has been realized. In particular, in this article, I consider the following transactions: (i) Europcar, (ii) Codere, (iii) Hovnanian, (iv) iHeart, (v) Supervalu (and similar transactions in Neiman Marcus and McClatchy), and (vi) Windstream. While Europcar and Codere were governed by European CDS, the relevant credit events in both cases were either Failure to Pay or Bankruptcy—the two credit events recognized in North American contracts—so they are germane to the analysis.

The breakdown in equivalency can be driven by the manner in which (i) auction settlement is triggered, (ii) the set of debt obligations deliverable at auction is determined, and (iii) the auction clears. Codere, Hovnanian, iHeart, and Windstream implicate (i): the manner in which auction settlement is triggered. In Codere, Hovnanian, and iHeart, the CDS auction was triggered by a deliberate—arguably voluntary—Failure to Pay. In Windstream, the CDS auction was triggered by an involuntary bankruptcy petition against the objections of minority creditors and the debtor. Hovnanian and Supervalu (and similarly, Neiman Marcus and McClatchy) implicate (ii): the set of debt obligations deliverable. Under the alleged facts of Hovnanian, the debtor coupled a deliberate Failure to Pay with the issuance of a low-priced bond that would drive settlement at auction. In Supervalu, an acquired subsidiary of the debtor was made a co-borrower—instead of a guarantor—on the acquisition debt, ensuring that such debt would be deliverable in any auction for CDS referencing the subsidiary. Europcar implicates (iii): the manner in which the auction clears—demonstrating the possibility for an illiquidity-driven squeeze in step two of the auction.

A. Auction Squeeze (Europcar)

In December 2020 the EMEA Determinations Committee determined that a Failure-to-Pay Credit Event had occurred with respect to reference entity Europcar Mobility Group, S.A., due to an October 2020 nonpayment of approximately EUR 9mm in interest on EUR 400mm notional of Senior ’26 notes, uncured beyond the relevant grace period. An auction was held on January 13, 2021.

At auction, the IMM settled at 73, with the lowest bid at 68 and the highest offer at 76.5. The net Open Interest was EUR 43.3mm to buy. Despite EUR 1.1bn in deliverable obligations, only EUR 35.9mm of Limit Order offers were received, at offers ranging from 71 (two points below the IMM) to 87 (fourteen points above it). With unfilled open interest to buy, per the rules of the auction, the Auction Final Price was determined as par, or 100 cents on the dollar. Consequently, the payout for an unhedged CDS buyer was zero. By contrast, Europcar’s final restructuring featured a full equitization of the three deliverable obligations alongside EUR 250mm in bondholder backstopped new-money equity and a backstopped refinancing of a EUR 670mm revolving credit facility.

The result in Europcar demonstrates the risks of entering the CDS auction unhedged—that is, opting for cash settlement without submitting either an initial Physical Settlement Request or Limit Order. While physical settlement ensures receipt of par, cash settlement can lead to volatile and unexpected outcomes. Europcar demonstrates that participants who opt for cash settlement need to still “show up” to the auction.

Consider a CDS buyer opting for physical settlement. CDS Buyer will be looking to sell bonds at auction. If net Open Interest is to buy, CDS Buyer is—axiomatically—guaranteed to sell the bond notional submitted in her initial Physical Settlement Request. If there are insufficient Limit Order offers, the Auction Final Price will settle at par, or 100 cents on the dollar. Consequently, while CDS Buyer will receive 0 cents on the CDS, she will receive 100 cents on the bond sold at par. Similarly, if net Open Interest is to sell and there are insufficient Limit Order bids, CDS Buyer may not be guaranteed to sell her bond, and the Auction Final Price will settle at zero. However, here too, CDS Buyer will receive 100 cents—the settlement value of her CDS position.

The same analysis applies to CDS sellers opting for physical settlement. CDS Seller will be looking to buy bonds at auction. If net Open Interest is to sell, CDS Seller is—axiomatically— guaranteed to buy the bond notional submitted in her initial Physical Settlement Request. If there are insufficient Limit Order bids, the Auction Final Price will settle at 0 cents. Consequently, while CDS Seller will owe 100 cents on the CDS, she will obtain the bonds at 0 cents and receive the ultimate recovery on the bond. Similarly, if net Open Interest is to buy and there are insufficient Limit Order offers, CDS Seller may not be guaranteed to buy her bond, and the Auction Final Price will settle at 100 cents. Here, there is some basis risk, though likely favoring CDS Seller. Seller will owe zero on her CDS position, but she may be unable to purchase bonds and close out a short position. However, this would only disadvantage her to the extent bond prices remain above par subsequent to auction—a technical irregularity unlikely to persist—and she is forced to close out her short position at greater than 100 cents.

CDS participants opting for an outright cash settlement, however, will either receive or pay par less the Auction Final Price. This could be characterized as a defect of the CDS settlement process. Alternatively, it could be characterized as well within the expected set of outcomes. Consider a CDS buyer opting for cash settlement, with no bond holdings. Subsequent to the determination of a Credit Event but prior to the auction, suppose the market for CDS on Company A trades at 40 cents, implying an Auction Final Price of 60 cents. CDS Buyer has the opportunity to settle her position prior to auction but opts instead to wait for the auction. Rationally, Buyer would do so only if she expects the Auction Final Price to settle below 60 cents, thus increasing her recovery.

Suppose further that, after step one of the Auction, CDS Buyer sees there is significant net Open Interest to Buy. Consequently, Buyer is on alert that the Auction Final Price (i) is not capped on the upside and, (ii) if net Open Interest remains unmatched by Limit Orders to sell in step two, could reach par, thus reducing her CDS payout to zero. At this point, it is likely optimal to enter a Limit Order to sell. First, CDS Buyer will benefit from any downward price pressure on the Auction Final Price, and downward price pressure is exerted by offers. Second, CDS Buyer has already expressed a short view on the credit through her CDS position and, moreover, entered the auction expecting the Auction Final Price to settle below 60. This likely implies that Buyer believes the value of the bonds to be significantly less than 60. It follows that—if borrow is available—Buyer would be willing to increase her short position at some price. Perhaps that price is not 60 or 70; it might, however, be 80 or 90. Seen in this light, the result in Europcar might have been prevented by greater vigilance among the outright CDS participants.

However, in a universe with large numbers of participants, one CDS participant cannot unilaterally ensure that the final price in step two hews closely to the IMM in step one. In Europcar, for example, there were EUR 35.9mm of limit order offers in step two of the auction. With net Open Interest to buy of EUR 43.3mm and a concomitant deficiency of only EUR 7.4mm, this might suggest that step two of the auction failed to result in an accurate settlement price. This argument is strengthened by the fact that markets submitted by dealers in step one—to create the IMM—are not mere estimates. Those markets are carried through to step two of the auction and matched against net Open Interest to form irrevocable trades.

As discussed infra, one potential solution to this deviation between IMM in step one and the Auction Final Price in step two would be to make a two-sided—rather than one-sided—cap to the allowed deviation between the IMM and the Auction Final Price.

B. Deliberate Failure to Pay (Codere)

The 2014 ISDA Credit Derivatives Definitions explicitly allow for a Failure-to-Pay Credit Event and subsequent auction to be triggered by a deliberately missed coupon payment negotiated between the Reference Entity and a CDS market participant. Codere provides an example of such a transaction.

In 2013, Blackstone reportedly agreed to roll over existing financing it had extended to Reference Entity Codere, contingent upon Codere’s agreement to deliberately miss a coupon payment. Missing the coupon payment triggered a Failure-to-Pay Credit Event, resulting in an auction that settled at 54.5 and, in turn, a CDS payout of [100–54.5] = 45.5 cents on the dollar.

Negotiation of an Event of Default may appear to violate the spirit of credit derivatives. However, such negotiation appeared to be expressly permitted under the Representations section of the 2014 definitions. Specifically, the definitions provide that CDS participants are permitted to

generally engage in any kind of commercial or investment banking or other business with, the Reference Entity . . . regardless of whether any such action might have an adverse effect on the Reference Entity . . . or the position of the other party to such Credit Derivative Transaction or otherwise (including, without limitation, any action which might constitute or give rise to a Credit Event).

CDS participants are additionally permitted to possess nonpublic information “material in the context of such Credit Derivative Transaction” and are under no obligation to disclose such information to CDS counterparties.

Under the 2019 ISDA amendments, such a deliberately missed payment might not constitute a Failure to Pay, as it could fail to meet the new Credit Deterioration Requirement, which provides that a Failure to Pay has not occurred “if such failure does not directly or indirectly either result from, or result in, a deterioration in the creditworthiness or financial condition of the Reference Entity.” However, as discussed infra, the impact of the Credit Deterioration Requirement may be significantly curtailed by the ISDA’s official guidance on its implementation. Specifically, the guidance continues to allow for debtor-creditor negotiations made contingent upon CDS-related transactions, so long as the negotiation is part of a “bona fide” transaction that increases the likelihood of a successful debt restructuring. The Credit Deterioration Requirement has not been tested extensively. Since the July 2019 implementation of the 2019 amendments, there has been only one Failure-to-Pay Credit Event referencing North American CDS: California Resources Corp. In that case, the Americas Determinations Committee found, without providing significant analysis, that the Credit Deterioration Requirement had been met.

C. Deliberate Failure to Pay + Issuance of Cheapest-to-Deliver Deliverable Obligation (Hovnanian)

1. Hovnanian Background

In December 2017, Hovnanian Enterprises announced (i) an exchange offer for existing notes and (ii) new financing transactions with creditor GSO Partners. Under the terms of (i)—the exchange offer—the following would occur:

  • Creditors would tender up to $185mm principal of 8% Senior ’19 Notes;
  • Creditors would receive a combination of (i) cash; (ii) newly issued 13.5% Senior ’26 Notes; and (iii) newly issued 5% Senior ’40 Notes; and
  • Hovnanian would sell $26mm in tendered 8% Senior ’19 Notes to its wholly owned subsidiary, Sunrise Trail.

Notably, the ’40 Senior Notes offered only a 5% coupon, compared with the 13.5% coupon offered on the ’26 Senior Notes. This would, under any reasonable assumption regarding a yield curve, result in a significantly lower cash price for the ’40 note—likely, below 50% of par.

Under the terms of (ii)—the new financing transactions—GSO Partners would do the following:

  • Provide a Senior Unsecured Term Loan facility to refinance Hovnanian’s existing 7% Senior ’19 notes consisting of:
    • Initial term loans of $132.5mm, and
    • Up to $80mm of delayed draw term loans to redeem or repay notes not participating in the exchange offer;
  • Provide $125mm of senior secured first-priority revolving loans to fund the repayment of Hovnanian’s $75mm senior secured term loan facility; and
  • Purchase $25mm in 10.5% Senior ’24 Notes.

Additionally, as explicitly stated in the accompanying 1934 Act report, Hovnanian covenanted to “not . . . make any interest payments on the [8% notes purchased by Sunrise Trail] prior to their stated maturity.” Hovnanian noted that such “non-payment of interest . . . may result in the occurrence of a ‘credit event’ under certain credit default swap contracts entered into by third-parties, resulting in significant monetary exposure for those entities that sold such credit default swaps.” GSO Partners was purported to be an entity that would have benefited from such nonpayment, with an estimated $333mm in notional CDS protection.

2. Relevance to CDS Contracts

Relevant to the determination of CDS settlement value, through the aforementioned transactions, Hovnanian agreed to (i) create a Credit Event through nonpayment of interest and (ii) create a deliverable obligation that would likely result in a low Auction Final Price and therefore a high recovery rate on CDS.

In January 2018, a third party, Solus Asset Management, filed a related complaint against GSO and Hovnanian. The complaint asserted claims (i) against Hovnanian and GSO for securities fraud and tortious interference with “prospective economic advantage under its CDS contracts,” (ii) against Hovnanian for material misrepresentation in the tender offer, and (iii) against certain Hovnanian executives for tortious interference and liability as controlling persons. Solus further sought an injunction against actions that would create an intentional default.

In March 2018, the DC issued a statement responding to a General Interest Question submitted by a market participant, which sought interpretation of section 11.1(b)(iii) of the ISDA 2014 Definitions. That portion of the Representations section, discussed supra, appears to explicitly allow for CDS market participants and the Reference Entity to negotiate commercial actions adverse to other CDS counterparties, including “any action which might constitute or give rise to a credit event.” The DC noted that market participants expected Hovnanian to miss an interest payment in May 2018, subsequent to the expiration of a thirty-day grace period on an interest payment due May 1, 2018. However, the DC declined to prospectively interpret section 11.1(b)(iii) in the context of “a live bilateral dispute . . . that may be presented to the DC in a matter of months.”

Ultimately, no such dispute was presented to the DC. On May 30, 2018—before the expiration of the grace period—Solus, GSO, and Hovnanian settled. Under the terms of the settlement, Hovnanian agreed to “pay to Sunrise Trail [its wholly owned subsidiary] all amounts due to it as a result of the missed interest payment on the Sunrise Trail 8.0 percent Notes that was originally due on May 1, 2018.” Hovnanian further noted that terms of the GSO financing transactions remained unchanged.

D. Deliberate Failure-to-Pay Debt Held by Related Entity (iHeart)

In Hovnanian, the Americas Determinations Committee did not reach the question of whether a section 4.5 Failure to Pay had occurred. However, the DC had previously taken a literalist approach to determining whether deliberate nonpayment constituted a section 4.5 Failure to Pay. In December 2016, iHeart Communications arranged for a wholly owned subsidiary to purchase a portion of its debt obligations ahead of maturity. At maturity, iHeart repaid all obligations except those held by the subsidiary “to avoid a springing lien in favor of certain creditors . . . which would have been triggered had all of the 2016 Notes been retired.” Adopting a literalist approach, the DC determined that a Failure to Pay had occurred. At auction, a final price of 35.5 was established—perhaps, in light of the deliberate default, an odd result and a high return on CDS.

E. Orphaned CDS (Supervalu)

If a debt-issuing reference entity were to retire all instruments that could be construed as debt obligations, a CDS on that Reference Entity would become worthless. There would be nothing on which to default and, relatedly, no obligation deliverable in an auction. The Reference Entity would be an “empty box,” leaving the CDS orphaned. For an investor who bought CDS (short credit) prior to the retirement of all debt obligations, this would lock in a loss in value equivalent to premiums paid and remove the possibility of either future mark-to-market gains or receipt of payment upon an Event of Default. Conversely, for a seller of CDS, it would lock in a profit equivalent to the premiums received, removing the possibility of either future mark-to-market losses or payment upon an Event of Default. Consequently, the ability to move debt into and out of the “box” representing a CDS Reference Entity is another tool debtors can use to obtain favorable financing terms.

For example, in October 2018, United Natural Foods (“UNF”) undertook the acquisition of grocery chain Supervalu in a transaction underwritten by Goldman Sachs. Under the terms of the acquisition, UNF would pay approximately $2.9bn for the acquisition, delivering $1.3bn in cash to Supervalu shareholders and the balance of $1.6bn to Supervalu creditors, extinguishing most of Supervalu’s debt. The transaction, as originally contemplated, would be financed largely through a $2.15bn term loan. Supervalu would be a guarantor—but, importantly for standard North American CDS contracts, not a co-borrower—of the term loan. This would have the effect of significantly reducing the value of any CDS contract referencing Supervalu—effectively, “orphaning” the CDS contract by removing all potentially deliverable obligations. At the time of the transaction, there was alleged to exist approximately $470mm in notional value of CDS referencing Supervalu.

Ultimately, UNF agreed to add Supervalu as a co-borrower. Subsequent to the transaction, UNF brought a claim for damages against Goldman for (i) two counts of breach of contract, (ii) one count of breach of the implied covenant of good faith and fair dealing, and (iii) one count of fraud. In the complaint, UNF alleged that Goldman repeatedly demanded additional financing concessions, including those implicating CDS. In particular, UNF claimed that Goldman “demanded that UNFI add SUPERVALU as a co-borrower on the Term Loan[,] . . . assuring [UNF] that the change would have a ‘muted impact on UNFI but was meaningful to some select accounts.’” For North American CDS contracts, subsidiary debt guaranteed by the parent is deliverable into parent CDS. However, parent debt guaranteed by the subsidiary is not deliverable into subsidiary CDS. Consequently, adding Supervalu as a co-borrower to the acquisition debt would have the effect of ensuring that deliverable obligations remained at Supervalu, preventing a collapse in the value of Supervalu CDS.

UNF further alleged that these “select accounts” were creditors holding Supervalu CDS, who participated in the term loan as part of a quid pro quo to enhance the value of their CDS. Per the complaint, (i) Goldman received payment in the form of term loan participation and (ii) creditors received the benefit of an increase in the value of Supervalu CDS contracts. In addition to harming Supervalu CDS sellers (credit longs), UNF alleged that retaining Supervalu as co-borrower would leave UNF with “less flexibility in borrowing arrangements for SUPERVALU going forward, in addition to an inability to restructure SUPERVALU’s debt or to sell SUPERVALU outright.”

Stated differently, certain creditors were able to extract value from third-party CDS market participants to both the benefit (vis-à-vis a completed financing transaction) and the detriment (vis-à-vis reduced ability to incur future indebtedness at Supervalu) of the debtor. Similar transactions were purportedly considered with respect to Neiman Marcus (Aurelius Capital Management proposed that Neiman Marcus Group Inc. include language in an exchange offer adding Neiman Marcus Group Ltd. LLC—a separate entity—as co-borrower) and McClatchy (Chatham Asset Management—a McClatchy CDS seller—provided McClatchy with financing on the condition that new financing be incurred at a new entity and the proceeds be used to repurchase certain debt obligations owned by Chatham. However, as McClatchy would provide a guarantee on the new entity financing, the risk of orphaned CDS would be mitigated).

F. Involuntary Bankruptcy Petition (Windstream)

1. Windstream SDNY Lawsuit

In January 2013, Windstream Services, LLC issued $700mm in senior unsecured 6.375% ’23 Notes. Section 4.19 of the Indenture governing the ’23 Notes prohibited Windstream from engaging in a Sale and Leaseback Transaction. Under section 6.01 of the Indenture, a Noteholder holding 25% or more of the ’23 Notes could give written notice of Windstream’s failure to comply with section 4.19; if the failure was not cured within sixty days of notice, an Event of Default would occur. Under section 9.02 of the Indenture, an Event of Default could be waived by a majority of ’23 Note holders.

In April 2015, Windstream finalized a transaction that could be—and ultimately was, by the Southern District of New York—construed as a Sale and Leaseback Transaction. By September 21, 2017, Aurelius Capital held beneficial ownership of more than 25% of the ’23 Notes, satisfying the notice requirement of section 4.19. Moreover, including accrued but unpaid interest, Aurelius held approximately $310.459mm of the notes, which constituted more than half the approximately $585.7mm in ’23 Notes outstanding. This would prevent other creditors from issuing a waiver of the Event of Default under section 9.02 of the Indenture.

On September 21, 2017, Aurelius gave written notice that the 2015 transaction violated section 4.19 of the Indenture. On October 5, 2017, Windstream’s board approved an exchange offer wherein holders of other bond tranches would tender their notes in exchange for new 6.375% ’23 Notes and provide consent to waive the default asserted by Aurelius. The completion of the exchange offer was made contingent upon receipt of the requisite consents. On November 1, 2017, Windstream informed the court that it had received the requisite consents.

Aurelius challenged the validity of the Consent Solicitation and the Exchange Offer. The Southern District of New York ultimately held that (i) the 2015 transaction was a Sale and Leaseback Transaction under the terms of the Indenture, (ii) the new ’23 notes violated the additional–indebtedness restrictions of the Indenture, and therefore (iii) consents related to the exchange offer must be disregarded. Consequently, Windstream failed to obtain the requisite consent to waive the Event of Default. On February 15, 2019, the Southern District of New York held that Aurelius’s Notice of Default “ripened into an Event of Default on December 17, 2017”—after the expiration of the cure period—entitling Aurelius to a money judgment from Windstream in the amount of Notes held plus interest.

2. Windstream Bankruptcy Filing and CDS Auction

The notes had been trading at distressed levels, in the mid 40s to low 50s. Subsequent to the court’s determination, the ’23 Notes traded down further, below 40 cents. Ten days later, on February 25, 2019, Windstream filed for bankruptcy. Post-filing, the notes traded below 20 cents. On February 26, 2019, the Americas Determinations Committee determined that a Bankruptcy Credit Event had occurred and that an auction would be held. It was assumed by some market participants and Windstream itself that Aurelius held a significant notional of CDS protection—possibly exceeding the notional of its cash bond holdings.

On March 11, 2019, the Americas Determinations Committee received a challenge to the initial list of deliverables that challenged the inclusion of (i) the new 6.375% ’23 Notes from the exchange offer and (ii) any of the old 6.375% ’23 Notes specifically held by Aurelius. With regard to (i), the challenge asserted that the new ’23 Notes lacked documentation sufficient to determine the notes’ Outstanding Principal Balance and therefore could not qualify as Deliverable Obligations. With regard to (ii), the challenge asserted that Aurelius’s notes might be subject to equitable subordination by the Bankruptcy Court, resulting in a Prohibited Action that would reduce the Outstanding Principal Balance of those specific notes to zero. This would have the effect of rendering Aurelius’s notes non-Deliverable.

On March 13, 2019, a responsive submission to the DC asserted that removal of the old 6.375% Notes held by Aurelius and new 6.375% ’23 Notes would have the effect of “cutting by more than 56% the unsecured Deliverable Obligations . . . and eliminating altogether the lowest priced Deliverable Obligations (about 30% of the unsecured universe).” To the extent that the universe of Deliverable Obligations is reduced and there are fewer obligations that can be sold into an auction, the likelihood of a squeeze that results in a high Auction Final Price is increased. Consequently, removal of the new and old tranches of ’23 Notes would likely reduce the payout on Windstream CDS contracts. The DC committee ultimately rejected the challenges, and both tranches were deemed deliverable. The auction occurred on April 3, 2019, with an IMM of 26.75 and an Auction Final Price of 29.5.

III. Responses to Historical Deviations from Cash/Derivative Equivalency

A. 2019 ISDA Amendments

1. Summary

In April 2018—after Hovnanian, but before Windstream—the ISDA published a statement addressing engineered defaults. The ISDA noted that to ensure objectivity, predictability, and timeliness in the credit event determination process, the DC was necessarily precluded from subjective assessments or considerations of intent or good faith. However, the ISDA also noted that (i) the occurrence of CDS credit events that did not reflect underlying credit deterioration could negatively impact the overall CDS market and (ii) there was a consequent need to consider amendments to the 2014 definitions.

In July 2019, the ISDA published a list of amendments—the 2019 Narrowly Tailored Credit Event Supplement to the 2014 ISDA Credit Derivatives Definitions—which addressed some of the issues that arose in the transactions just discussed. Specifically, the 2019 Supplement amended the definitions of Outstanding Principal Balance and Failure to Pay while also providing guidance on the interpretation of Failure to Pay.

2. Outstanding Principal Balance

Recall that at auction, price is expressed as a percentage of the Outstanding Principal Balance. While the Outstanding Principal Balance is usually par, that is not necessarily so. The amendments to section 3.8 (“Outstanding Principal Balance”) address the concern that an artificially low Auction Final Price—and consequently high CDS recovery—could be effectuated by delivery of either (i) claims reduced in value by bankruptcy or other applicable law or (ii) original-issue-discount (OID) debt issued at a price significantly below redemption value. This addressed, the concern that Reference-Entity debt issuers could arrange with a creditor for the issuance of below-market-coupon, low-cash-price deliverable obligations, as in Hovnanian.

a. Claims Reduced in Value

The 2019 amendments clarify that the Outstanding Principal Balance will be calculated as if the claim were reduced by applicable bankruptcy or insolvency law, whether or not the Reference Entity is bankrupt. Consequently, if a debt obligation issued at par would be reduced to a claim of 50 cents under applicable law, the Outstanding Principal Balance would represent a 50% reduction of the par redemption value.

Specifically, section 3.8(a)(iii)(B) of the ISDA 2014 Definitions provides that the Outstanding Principal Balance is determined, “in accordance with any applicable laws (insofar as such laws reduce or discount the size of the claim to reflect the original issue price or accrued value of the obligation),” as the “lowest amount of the claim which could be validly asserted . . . if the obligation had become redeemable, been accelerated, terminated, or had otherwise become due and payable.” The 2019 amendments clarify that the “‘applicable laws’ shall include any bankruptcy or insolvency law or other law affecting creditors’ rights to which the relevant obligation is, or may become, subject.”

b. Original Issue Discount Debt

The 2019 amendments also provide that bonds issued at a price less than 95% of the redemption value will have their Outstanding Principal Balance reduced from par, generally pursuant to a straight-line interpolation. Consequently, if a bond were issued at 70 cents, for redemption at par ten years later, the Outstanding Principal Balance five years after issue would be 85 cents.

Specifically, the 2019 amendments provide that if (i) the Outstanding Principal Balance is not reduced pursuant to applicable law under section 3.8(a)(iii)(B), (ii) the issue price was less than 95% of redemption value, and (iii) the obligation does not include contractual provisions specifying the accretive amount that would be due upon an early redemption, then a haircut would be applicable. The applicable haircut would be the lesser of (i) the noncontingent portion of redemption value and (ii) “straight line interpolation between the issue price of the Bond or the amount advanced under the Loan and the principal redemption amount.”

3. Failure to Pay

Most notably in Codere, Hovnanian, and iHeart, the market had observed transactions where a Failure-to-Pay Credit Event was triggered without a corresponding inability of the debtor to pay. The 2019 amendments attempted to address this by adding a Credit Deterioration Requirement to section 4.5 (“Failure to Pay”) of the 2014 definitions. The language is broad: “it shall not constitute a Failure to Pay if such failure does not directly or indirectly either result from, or result in, a deterioration in the creditworthiness or financial condition of the Reference Entity.”

The official guidance on interpreting this language is no less broad; it grants significant discretion to the DCs—a marked departure from the intended objectivity and predictability of the credit event determination process referred to in the ISDA’s April 2018 statement. That language in the 2019 guidance follows:

It should be noted that the financial condition of the Reference Entity at the time it fails to pay is not conclusive as to whether or not such non-payment resulted from, or resulted in, a deterioration in the creditworthiness or financial condition of the Reference Entity. Rather, there must be a causal link between the non-payment and the deterioration in the creditworthiness or financial condition of the Reference Entity. While the expectation is that the Credit Deterioration Requirement would generally be met by way of a non-payment resulting from such deterioration, the Credit Deterioration Requirement may also be met by a non-payment that results in such deterioration. This is intended to permit a Failure to Pay to occur where a technical, administrative or operational non-payment occurs that does not itself result from a deterioration in creditworthiness or financial condition, but the consequences of such non-payment result in a deterioration in the creditworthiness or financial condition of the Reference Entity.

In an attempt to narrow the language, the 2019 amendments provide a non-exhaustive list of Eligible Information that may indicate the Credit Deterioration Requirement is not met. The list includes transactions where

  • nonpayment arises from an arrangement between Reference Entity and a third party “where an essential purpose of the arrangement . . . . is to create a benefit under a Credit Derivative Transaction”;
  • as part of an arrangement, the Reference Entity either (i) issues a bond likely to be cheapest to deliver in the auction or (ii) issues a material amount of additional deliverable obligations;
  • The nonpayment did not result in the acceleration of other debts;
  • The Reference Entity had sufficient liquidity to pay;
  • The nonpayment was cured promptly after the grace period expired; and
  • The nonpayment was only on debt held by affiliates or others unlikely to accelerate.

The 2019 amendments similarly provide a non-exhaustive list of Eligible Information that may indicate the Credit Deterioration Requirement is met. The list includes transactions where

  • the Reference Entity previously announced financial distress or intent to restructure;
  • the Reference Entity hired professional advisers;
  • the nonpayment occurred as part of a judicial proceeding;
  • the nonpayment was on debt held by multiple parties;
  • the nonpayment occurred due to an inability to refinance;
  • the nonpayment occurred on a day clearly scheduled well in advance; and
  • other debt obligations are accelerated, the Reference Entity fails to pay other debt obligations, and/or the Reference Entity files for bankruptcy.

Ostensibly, these factors—though subjective—might have prevented triggering of a Failure to Pay in, Codere, Hovnanian, or iHeart. However, two caveats provided in the interpretive guidance suggest otherwise. Specifically, the guidance suggests that where (i) a deliberate nonpayment is the result of arm’s-length negotiation between the debtor and a significant portion of its creditors by value and (ii) transaction participants could articulate some bona fide reason for the transaction, the deliberate nonpayment will be deemed to result from a deterioration in creditworthiness. Specifically, the guidance notes that

[i]f a Reference Entity enters into a forbearance, standstill or other similar arrangement with its creditors for bona fide commercial reasons related to a deterioration in its creditworthiness or financial condition, this would rarely result in a determination that the relevant non-payment . . . did not . . . result from a deterioration in the creditworthiness or financial condition.

More directly, the guidance states that where the engineered default is part of a bona fide transaction that increases the likelihood of a successful restructuring, a section 4.5 Failure to Pay will be triggered:

[W]ithin the context of a bona fide debt restructuring, if the Reference Entity enters into an arrangement or understanding with such creditors that includes a failure to make a payment with the purpose of causing settlement of such Credit Derivative Transactions so as to increase the likelihood of success of such bona fide restructuring, and in circumstances where without such restructuring the Reference Entity would be likely to enter into bankruptcy or similar proceedings, such arrangement or understanding should generally be considered to have the essential purpose of facilitating such restructuring rather than creating a benefit under a Credit Derivative Transaction as described in paragraph 1.10(a) [Eligible Information indicating that Credit Deterioration Requirement is not satisfied].

Consequently, in two of the pre-2019 Amendments transactions—Codere and Hovnanian—a Failure to Pay might still be triggered under the new definitions. Conversely, in iHeart, where the Failure to Pay was for the Corporate’s benefit—to avoid a springing lien—outside the context of a restructuring, it likely would not be triggered under the 2019 Amendments.

B. Private Disclosure Requirements (Windstream Provisions)

In the wake of Windstream, some debtors have inserted “anti-net short” or “Windstream” provisions in their credit agreements. In short, these provisions aim to limit a lender’s ability to vote if their net economic position in the debtor’s obligations is “short”—that is, where it benefits from a deterioration in creditworthiness. For example, in May 2019, Sirius Computer Solutions included the following language in a credit agreement:

[A]ny Lender (other than (x) any Lender that is a Regulated Bank and (y) any Revolving Lender as of the Closing Date) that, as a result of its interest in any total return swap, total rate of return swap, credit default swap or other derivative contract (other than any [such swap] entered into pursuant to bona fide market making activities), has a net short position with respect to the Loans and/or Commitments shall have no right to vote any of its [loans] and shall be deemed to have voted its interest as a Lender without discretion in the same proportion as the allocation of voting with respect to such matter by Lenders who are not [net-short lenders].

As other commentators have noted, these provisions seem to address a limited landscape of CDS-related issues. The provisions might prevent a creditor from holding CDS protection in excess of its bond notional where that bond notional is sufficient to vote in a manner likely to trigger an Event of Default (e.g., by constituting sufficient notional to issue a notice of default and/or to vote against a waiver of default). However, such provisions would not prevent a creditor from the same use of its bondholder rights if the CDS position were equivalent, nearly equivalent, or merely sizeable in relation to its bond position.

For example, a distressed credit investor willing to sit on her long cash bond investment might nonetheless be incentivized to push the debtor into bankruptcy to realize an immediate cash profit on her CDS position. A basis trader—trading on the divergence in CDS spread and cash bond credit spread—would retain voting rights despite trading activity that may be indifferent to or even incentivized to effectuate a default, depending on her view of likely auction dynamics. A bondholder might be of the view that there is a significant excess of CDS buyers in relation to CDS sellers, that consequently there would be significant net Open Interest to sell bonds at auction, and that a lack of buyers would drive the Auction Final Price lower, and the CDS recovery higher, than reflected in the ultimate recovery on cash bonds.

None of these transactions would be precluded by an anti-net-short provision similar to the Sirius provision above. However, as discussed infra, such activity might implicate good-faith requirements under section 1112 (filing) and section 1126 (voting) of the Bankruptcy Code, as well as the section 364 (financing) requirement for notice and hearing. This implication of the Bankruptcy Code’s requirements militates in favor of broader disclosure requirements.

C. Regulatory Disclosure Requirements Outside Bankruptcy

For CDS and other security-based swaps, there is no current equivalent to the Rule 13d‑3 disclosure requirements for equity holdings. Comparable disclosure requirements may be forthcoming, however, in a finalized version of the SEC’s recently proposed Rule 10B‑1 swap regulations.

In December 2021, the SEC proposed two rules—9j‑1 and 10B‑1—intended, in part, to address engineered CDS transactions through (i) application of anti-manipulation prohibitions (9j‑1) and (ii) heightened disclosure requirements (10B‑1). Proposed Rule 9j‑1 would apply the general anti-manipulation prohibitions of Exchange Act section 10(b) to CDS. Proposed Rule 10B‑1 would require public reporting—comparable to Rule 13d‑3 requirements—of net CDS/debt positions exceeding $150mm notional and gross CDS/debt positions exceeding $300mm. On June 7, 2023, the SEC adopted a finalized version of Rule 9j‑1. As of June 7, 2023, the SEC had yet to adopt a finalized version of Rule 10B‑1.

1. Finalized Rule 9j‑1

Finalized Rule 9j‑1 would apply the general anti-manipulation prohibitions of section 10(b) of the Exchange Act to security-based swaps, which includes single-name corporate CDS. The rule is intended, in part, to address engineered CDS transactions. Specifically, Rule 9j‑1(a)(6) is intended to “prohibit manipulation and attempted manipulation of the price or valuation of any security-based swap, in connection with effecting or attempting to effect a transaction in, or purchasing or selling, or inducing or attempting to induce the purchase or sale of, any security-based swap.” The SEC declined to adopt any bright-line, concrete triggers for Rule 9j‑1(a)(6) applicability, adopting instead a fact-specific inquiry that the SEC asserts is supported by an existing body of anti-manipulation case law. Consequently, the extent to which Rule 9j‑1(a)(6) would have implicated historical examples of cash-derivative deviation is unclear.

Militating toward broad application of Rule 9j‑1 to manufactured credit events, the SEC notes that actions prohibited under Rule 9j‑1(a)(6) “may include, for example, orphaning a CDS, avoiding termination of a CDS for a period of time, or causing the termination of a CDS,” as well as “artificially influenc[ing] the composition of the deliverable obligations in a CDS auction.” Respectively, such fact patterns might implicate transactions similar to Supervalu, Codere, iHeart and Windstream, and Hovnanian.

Militating toward narrow application of Rule 9j‑1, the SEC appears to provide a carve-out for transactions coupled with a legitimate financing need. The SEC notes that

reference entities often rely on financing and other forms of relief to avoid defaulting on their debt . . . [and] final rule [9j‑1] is not intended to discourage lenders and prospective lenders from discussing or providing such financing or relief, even when those persons also hold CDS positions. Rather, the Commission is adopting Rule 9j‑1(a)(6) to account for actions taken outside the ordinary course of a typical lender-borrower relationship (or a prospective lender-borrower relationship).

By extension, one reading of the rule is that so long as there is some articulable, non-CDS-related justification for a transaction, the prohibitions of Rule 9j‑1(a)(6) would not apply. Consequently, Rule 9j‑1 might not reach transactions with fact patterns comparable to those discussed, where CDS participants could have argued legitimate needs for (i) financing (as in Codere, Hovnanian, and Supervalu), (ii) avoidance of a detrimental security assignment (as in iHeart), or (iii) the protections of bankruptcy (as in Windstream). Absent a body of case law specific to engineered CDS transactions, the effective reach of Rule 9j‑1 to CDS will remain unclear.

2. Proposed Rule 10B‑1

a. 10B‑1 Requirements and Applicability

Pursuant to Regulation SBSR—effective as of February 14, 2022—certain CDS transactional data are already subject to public dissemination, on a basis confidential as to market participant and aggregate position. Publicly disseminated data under SBSR includes underlying debt obligation, effective date and termination dates, coupon payments, time of transaction, price, and notional. However, publicly disseminated data excludes the identity of counterparties and aggregate market positions. Rule 10B‑1 purports to change this by providing a public reporting mechanism for CDS and other security-based swaps.

Under Proposed Rule 10B‑1, the SEC would rely on authority pursuant to the 2010 Dodd-Frank Act to require security-based swap reporting that is comparable in form—but distinct in applicability—from Rule 13d‑3 requirements applicable to equity securities. While Dodd-Frank section 766(e) purports to grant the SEC authority to broaden 13d‑3 Beneficial Ownership to encompass security-based swaps, the SEC appears to have relied instead on Dodd-Frank section 763(h) to implement Rule 10B‑1. Section 763(h) broadly authorizes the SEC to require market participants to “report such information as the Commission may prescribe regarding any position . . . in any security-based swap.”

Relevant to CDS, Proposed Rule 10B‑1 would require reporting of any net-long CDS (short credit), short-CDS (long-credit) position with a notional exceeding $150mm, or any gross position exceeding $300mm notional. The net amount would be calculated by netting positions in the underlying debt obligation against the CDS position, while the gross amount would aggregate both positions. Qualifying positions would be reported under a proposed Schedule 10B form and made publicly available. The SEC notes as a benefit of public reporting that,

in the case of manufactured or other opportunistic strategies in the CDS market, . . . market participants and regulators [are provided] with advance notice that a person (or a group of persons) is building up a large CDS position which could create an incentive to vote against their interests as a debt holder, possibly with an intent to harm the company, even if such conduct is not inherently fraudulent.

The finalized version of Rule 10B‑1 remains to be seen, and it may be subject to challenge. I have argued elsewhere that, in its current form, Rule 10B‑1 exceeds the mandate of Dodd-Frank. As explained in Part III.C.2.b (“10B‑1 Enforceability”), the public disclosure of individual position data likely violates the confidentiality mandate of Dodd-Frank section 763(i), and the potential inapplicability of 10B‑1 to offshore transactions may frustrate the proposed rule’s purpose.

b. 10B‑1 Enforceability

The Dodd-Frank authority to mandate public reporting of aggregate, confidential swap data appears uncontroversial. The Dodd-Frank Act was passed to reduce systemic, market-wide risk posed by undisclosed securities leverage, and public disclosure of aggregate, confidential transactional data does not exceed that scope of authority. However, while Proposed Rule 10B‑1 meets many systemic risk-reduction goals, portions of the rule exceed the scope of and/or fall short of the mandate of Dodd-Frank.

First, Rule 10B‑1 calls for the public dissemination of participant identity, which may exceed the scope of Dodd-Frank section 763(h) and contradict the confidentiality mandate of section 763(i). Section 763(h) authorizes the SEC to require swap participants to “report such information as the Commission may prescribe.” Similarly, section 763(i) authorizes the SEC to “make security-based swap transaction and pricing data available to the public in such form and at such times as the Commission determines appropriate to enhance price discovery.” However, section 763(i) also circumscribes the authority to identify individual participants, requiring that any disclosure requirement be “in a manner that does not disclose the business transactions and market positions of any person” and does “not identify the participants.” The legislative history similarly supports a bifurcation of (i) public disclosure of aggregate transaction data and (ii) confidential regulatory disclosure of individual transaction data.

Second, the exemption from public dissemination of certain offshore transactions may prevent full public disclosure of the systemic risk at which Dodd-Frank was aimed. By the SEC’s own admission, offshore exemption might encourage jurisdictional arbitrage and investor re-domiciliation. Third, Rule 10B‑1 does not require reporting entities to disclose the purpose of their transaction and requires only a “brief description” of related third-party arrangements. Limited substantive disclosure may conflict with the SEC’s stated goal of providing market transparency.

D. Regulatory Disclosure Requirements Inside Bankruptcy

1. Disclosure Generally: 11 U.S.C. § 501; Rule 3001; Rule 2019; 11 U.S.C. § 1102

Inside the bankruptcy process, position disclosure requirements are limited. Disclosure consists primarily of (i) 11 U.S.C. § 501 (“Filing of proofs of claims or interests”), (ii) Rule 3001 (“Proof of claim”), (iii) Rule 2019 (ad hoc committee disclosure requirements), and (iv) 11 U.S.C. § 1102 (applicable to official committee information disclosure requirements).

11 U.S.C. § 501 provides that “[a] creditor . . . may file a proof of claim.” Rule 3001 does not require that a proof of claim include information related to holdings of CDS or other security-based swaps. Rule 2019 provides disclosure requirements for unofficial, ad hoc creditor committees. Rule 2019(c) provides that each member of such committees must disclose “the nature and amount of each disclosable economic interest held in relation to the debtor” as of the dates on which (i) the committee was formed and (ii) a “verified statement” disclosing the committee’s existence to the court is made. “Disclosable economic interest” is defined as “any claim, interest, pledge, lien, option, participation, derivative instrument, or any other right or derivative right granting the holder an economic interest that is affected by the value, acquisition, or disposition of a claim or interest.” The comments to Rule 2019 provide that “disclosable economic interest” is sufficiently broad to cover any economic interest that could affect the legal and strategic positions a stakeholder takes in a chapter 9 or chapter 11 case” and “includes, among other types of holdings, short positions, credit default swaps, and total return swaps.”

2. Inadequacy of Rule 2019

a. Timing Mismatch: CDS Settlement Precedes Rule 2019 Verified Statement Filing

There are three important ways in which Rule 2019 may not capture relevant interests. First, Rule 2019 requires only disclosure of “disclosable economic interest” held (i) “as of the date . . . the group or committee was formed” and (ii) “as of the date of the statement” filed with the bankruptcy court. Consequently, if the CDS auction settlement date precedes the creation of an ad hoc committee, there would be no CDS position to disclose in a Rule 2019 Verified Statement. Thus, the disclosure timeframe of Rule 2019 may capture ISDA standard CDS where the auction is triggered by a Bankruptcy Credit Event. However, Rule 2019 would likely not capture ISDA standard CDS where the auction is triggered by a Failure-to-Pay Credit Event.

To see this, consider the case of California Resources Corporation. On June 11, 2020, the DC determined that a Failure-to-Pay Event had occurred with respect to California Resources Corporation. The CDS auction took place on July 7, 2020. One week later, on July 15, 2020, the debtor filed for Chapter 11 in the Bankruptcy Court for the Southern District of Texas. The Verified Statement(s) Pursuant to Bankruptcy Procedure 2019 for the Ad Hoc Term Lender Group and Ad Hoc Crossover Group were not filed until July 17 and July 27, respectively. Exhibit A to the former lists Disclosable Economic Interest “accurate as of close of business July 16, 2020.” Exhibit A to the latter lists Disclosable Economic Interest “accurate as of July 21, 2020.” In neither case were interests held as of July 7, 2020 (the date of the CDS auction) or July 10, 2020 (the date of settlement) disclosed. Consequently, had any of the Ad Hoc members held CDS positions referencing the underlying claims, those positions would not have appeared on the Rule 2019 Verified Statement. The CDS would have ceased to exist by the July 10, 2020, settlement date, and there would have been nothing to disclose on July 16 or 21.

Subsequent to the 2011 amendments to Rule 2019, the Americas DC has confirmed the occurrence of five Failure-to-Pay Credit Events, four of which resulted in an auction. In all four, the earliest Rule 2019 Verified Statement was filed after the auction’s settlement date. Consequently, committee members’ CDS positions—if any—would not have appeared on those Rule 2019 filings. The table below demonstrates the timing mismatches:

Table 13

Table 13

b. Potential Nonenforcement of Rule 2019 to Section 1102 Committee Members

Second, the Rule 2019 requirement to provide “disclosable economic interest” (i) does not apply to creditors who are members of neither an official (section 1102) nor an ad hoc committee, and it (ii) may not apply in practice to section 1102 official committee members.

Prior to the 2011 amendments, only ad hoc committee members needed to disclose their Rule 2019 holdings to the public. The pre-2011 version of Rule 2019 did not apply to official committees formed pursuant to 11 U.S.C. § 1102—the Unsecured Creditors Committee (UCC). Rather, pursuant to 11 U.S.C. § 1102(b)(3), the UCC is to “provide access to information for creditors who—(i) hold claims of the kind represented by that committee; and (ii) are not appointed to the committee.” What specific information is required is not provided in the statute. Professors Hu and Black noted that “[i]n theory, a sophisticated bankruptcy trustee could condition appointment of a creditor to an official committee on disclosure of coupled assets.” However, prior to the Rule 2019 amendments, courts interpreting section 1102 did not establish any requirement to report derivative holdings.

Rule 2019 was revised in 2011 to require inclusion of “disclosable economic interests”—including credit derivatives—in Rule 2019 Verified Statements. The language of the statute—as well as Committee Notes to the 2011 Amendment and guidance from the United States Trustee—all suggest that amended Rule 2019 applies to section 1102 committees. However, a review of several bankruptcy dockets suggests that section 1102 committees only sometimes—not always—file Rule 2019 Verified Statements. Additionally, Rule 2019 remains inapplicable to creditors who are members of neither the section 1102 committee nor an ad hoc committee.

c. Incomplete Definition of “Disclosable Economic Interest”

Third, Rule 2019 could more clearly incorporate CDS. Under a common-sense understanding of the definition, and as confirmed by the Committee Notes on the 2011 Amendment, “disclosable economic interest” includes CDS. However, the applicability of Rule 2019 to CDS could be made explicit by referencing statutorily defined (i) security-based swaps and/or statutorily undefined (ii) credit default swaps, which are nonetheless understood to fall under security-based swaps. While CDS are commonly understood to be “derivative instruments,” security-based swaps and derivative instruments are statutorily distinct, as discussed in the SEC’s 2022 proposed amendments to Rule 13d-3. “Derivative securities” are defined under 17 C.F.R. § 240.16a-1(c) to mean “any option, warrant, convertible security, stock appreciation right, or similar right with an exercise or conversion privilege at a price related to an equity security, or similar securities with a value derived from the value of an equity security.” By contrast, CDS fall under “security-based swaps,” as defined in 15 U.S.C. § 78c(a)(68)(A).

IV. Relevance of CDS to Bankruptcy Code Requirements

A. Summary

The requirements of the Bankruptcy Code militate in favor of disclosure of CDS. Specifically, CDS implicates (i) the “good faith” requirement for filing under section 1112, (ii) the “notice and hearing” requirement for financing and sale transactions under section 364, and (iii) the “good faith” requirement for voting under section 1126(e). Whether these requirements are met is a fact-based inquiry carried out by bankruptcy courts. CDS positions may determine—and have determined in examples discussed above—the incentives and ultimate economic results of both creditors and debtors. Consequently, creditor CDS positions and related transactions (and their profit/loss thereon) are facts relevant to determining, in the event the debtor ends up in bankruptcy court, whether the requirements of sections 1112, 364, and 1126 have been met.

B. Section 1112(b)(1) Filing & Requirement of “Good Faith”

1. Good Faith and Voluntary Petitions

Pursuant to 11 U.S.C. § 1112(b)(1), “the court shall convert a case under [chapter 11] to a case under chapter 7 or dismiss a case under [chapter 11], whichever is in the best interests of creditors and the estate, for cause.” In addition to the expressly enumerated types of “cause” in 11 U.S.C. § 1112(b)(4), bankruptcy courts—in SGL Carbon, Integrated Telecom, and their progeny (including most recently, the LTL cases and In re Aearo)—have found an implicit “good-faith” requirement for Chapter 11 petitions. “The focus of the [good-faith] inquiry is whether the petitioner sought ‘to achieve objectives outside the legitimate scope of the bankruptcy laws’ when filing for protection under Chapter 11.” “To be filed in good faith, a chapter 11 petition must be supported by ‘a valid reorganizational purpose.’” “Because the Code’s text neither sets nor bars explicitly a good-faith requirement, we have grounded it in the ‘equitable nature of bankruptcy’ and ‘the purposes underlying Chapter 11.’” Where bankruptcy neither (i) preserves a going concern, nor (ii) maximizes property available to satisfy creditors, a Chapter 11 petition will be dismissed.

“Determining whether the good faith requirement has been satisfied is a fact-intensive inquiry in which the court must examine the totality of facts and circumstances and determine where a petition falls along the spectrum ranging from the clearly acceptable to the patently abusive.” Good faith will be found where “the petition serves a valid bankruptcy purpose, e.g., by preserving a good concern or maximizing the value of the debtor’s estate.” Conversely, good faith will not be found where “the petition is filed merely to obtain a tactical litigation advantage.” “No single factor is determinative of a lack of good faith in filing a petition.” In In re Primestone, the U.S. District Court for the District of Delaware, identified a series of factors commonly considered—including, relevant to involuntary petitions, whether there is “[n]o pressure from non-moving creditors.” This particular factor would be relevant where—as in Windstream—one creditor holds a percentage of notes outstanding sufficient to both (i) issue a notice of default and (ii) block minority creditors from issuing a waiver of default.

2. Good Faith and Involuntary Petitions

The good-faith requirements for voluntary debtor petitions apply similarly to involuntary creditor petitions filed pursuant to 11 U.S.C. § 303. The Third Circuit, for example, uses a “totality of the circumstances” standard to assess good faith under a section 303 involuntary filing that is equivalent to the standard applied to assessing good faith in a voluntary filing under section 1112(b). Third Circuit courts may consider whether (i) “there was evidence of preferential payments to certain creditors or of dissipation of the debtor’s assts”; (ii) “the petitioning creditors used the filing to obtain a disproportionate advantage for themselves rather than to protect against other creditors doing the same”; or (iii) “the filing was used as a substitute for customary debt-collection procedures.”

3. Application to Manufactured Defaults

A fact-specific inquiry as to whether the alleged actions of Aurelius in Windstream would meet the good-faith requirements of section 303 is beyond the scope of this article. However, a generalized version of that transaction might well constitute a bad-faith filing. Consider the following example.

Creditor A purchases a sufficient amount of Debtor’s obligations to meet the default notice requirements and block any other creditors from waiving default. Creditor A separately enters into a third-party agreement that would pay out in the event of a bankruptcy filing (e.g., CDS). While Debtor’s financial condition is poor, the probabilistic need for an in-court restructuring has not materially changed for some time and is not expected to change. Creditor A is willing to hold Debtor’s obligations, either in anticipation of an ultimate recovery or absent anticipation of a near-term significant decline in value.

Creditor A does not believe that Debtor, and its creditors in aggregate, would benefit from the bankruptcy process. However, Creditor A determines that, as an individual creditor, it could receive immediate and significant cash value—through its third-party agreement—if Debtor files for bankruptcy. Creditor A determines that the cost of bankruptcy to its cash-bond holdings is less than the benefit of the expected payout from its third-party agreement (e.g., CDS). Creditor A files a petition for Debtor’s involuntary bankruptcy.

While the payout on the third-party agreement may not implicate any of the assets of Debtor, the bankruptcy filing required to obtain that payout may have negative effects for both Debtor and other creditors. The legal costs of the bankruptcy process and an increased cost of capital, as well as any limitations imposed on Debtor’s normal operations, may decrease the pool of available assets with which to pay other creditor claims. Under such circumstances, Creditor A’s petition might well, as a petition used “to obtain a disproportionate advantage for themselves rather than to protect against other creditors doing the same,” fail to meet the good-faith requirement of section 303.

C. Section 364 Financing and Requirement for “Notice and Hearing”

1. Section 364 Financing, Generally

Debtor may obtain post-petition financing pursuant to 11 U.S.C. § 364. Only ordinary-course unsecured financing may be obtained without bankruptcy court approval. The incurrence of unsecured debt outside of the ordinary course requires court approval “after notice and hearing.” The term “after notice and a hearing” is broadly defined by the Bankruptcy Code, ostensibly granting the court significant discretion to determine the means of notice and hearing “appropriate in the particular circumstances.” Bankruptcy courts have found the notice requirement to be “founded in fundamental notions of procedural due process”; consequently, “there must be ‘notice reasonably calculated . . . to apprise interested parties of the pendency of the action and afford them an opportunity to present their objections.”

2. Sub-rosa Transactions

“A sub rosa plan is a transaction or agreement that commits such a substantial part of the debtor’s assets that, if the transaction were allowed, the terms of a plan of reorganization would be severely limited, if not completely predetermined.” Sub-rosa concerns might implicate: (i) section 363(b) sales; (ii) Rule 9019 settlements; and (iii) section 364 DIP financing transactions. Factors relevant to determining whether a proposed transaction constitutes a sub-rosa plan include whether it “(1) dictate[s] the terms of any future reorganization plan, (2) disenfranchise[s] creditor voting in the plan process, (3) alter[s] creditors rights, (4) dispose[s] of highly significant assets[,] . . . and (5) release[s] claims against the debtor.”

3. Application to Manufactured Defaults

A fact-specific inquiry as to whether the specific transactions in Codere, Hovnanian, and Supervalu would fail the section 364 notice and hearing requirement and/or constitute a sub-rosa plan of reorganization is beyond the scope of this article. However, the broad patterns of those alleged transactions—in a generalized form—could be characterized as sub-rosa plans. Consider the following example.

Debtor, anticipating potential insolvency and a need for either out-of-court or in-court restructuring, obtains additional or replacement financing, at below-market rates, from Creditor A. Debtor is able to obtain below-market financing, and Creditor A is compensated for the difference between the contracted rate and market rate via an upfront, immediate cash payment made by a third party to Creditor A. This third-party payment is triggered not by any bilateral negotiation between the third party and Creditor A or Debtor but by the unilateral action of Debtor. This unilateral action consists of some combination of (i) failing to pay existing obligations to certain or all creditors, (ii) materially changing the guarantee structure of existing obligations, and (iii) issuing new obligations that impact the value of that third-party payment.

In sum, Debtor receives valuable financing, the cost of which is in part borne by a third party. Creditor A most obviously receives value borne by that same third party. Creditor A may also receive value borne by other creditors, where those creditors (i) are primed by Creditor A’s secured financing or (ii) have their percentage holdings of total unsecured debt diluted by Creditor A’s unsecured financing.

While such priming transactions might normally require arm’s-length negotiation of benefit received for consideration tendered, here, Debtor and Creditor A have found a way to extract some of that value from a third party. Were this below-market financing unavailable, Debtor might have entered insolvency proceedings, where any plan of reorganization would be subject to the protections of the Bankruptcy Code. Additionally, if the below-market financing is ultimately unsuccessful, in that Debtor subsequently files for bankruptcy, the new transactions with Creditor A may have the effect of dictating many of the terms of any reorganization.

D. Section 1126(e) Vote Designation and Requirement of “Good Faith”

1. Vote Designation, Generally

Pursuant to 11 U.S.C. § 1126(e), “[o]n request of a party in interest, and after notice and a hearing, the court may designate any entity whose acceptance or rejection of [a plan of reorganization] was not in good faith, or was not solicited or procured in good faith in accordance with the provisions of this title.” “The inquiry in deciding whether a creditor acted in good faith is whether its vote was cast for the ulterior purpose of securing some advantage to which the creditor was not otherwise entitled.” The Second Circuit has noted that vote designation is employed “sparingly, ‘as the exception, not the rule,’” and “not just any ulterior motive constitutes the sort of improper motive that will support a finding of bad faith.” Courts have held the determination of “good faith” to be a fact-intensive inquiry requiring case-by-case analysis.

The Bankruptcy Court for the Southern District of New York has found “bad faith” sufficient for designation under section 1126€ either “where a claim holder attempts to extract a personal advantage not available to other creditors in the class, or . . . where a creditor acts in furtherance of an ulterior motive, unrelated to its claim or its interests as a creditor.” “Badges of bad faith” justifying disqualification “include efforts to (1) assume control of the debtor; (2) put the debtor out of business or otherwise gain a competitive advantage; (3) destroy the debtor out of pure malice; or (4) obtain benefits available under a private agreement with a third party which depends on the debtor’s failure to reorganize.” In In re DBSD North America, Inc., Judge Robert Gerber of the Southern District of New York noted that where there is unrefuted evidence that creditors have voted in order to benefit a separate short position, he would “designate[ ] their votes in a heartbeat.”

2. Application to Manufactured Defaults

By the time a plan of reorganization reaches the voting stage, a CDS auction may have already occurred. Provided the auction has occurred, there would be no ISDA standard CDS positions outstanding by the time of a vote on a plan of reorganization. This is not to say that all short-credit positions will have necessarily expired or settled. It is conceivable that two counterparties could bilaterally contract for a bespoke payout tethered to the result of the bankruptcy process, outside the ISDA framework. However, ISDA standard CDS will have settled, and therefore any short-credit position attributable to ISDA standard CDS will have terminated. Consequently, a creditor who received an economic payout from its CDS position may nonetheless vote on a plan of organization purely in its interest as a holder of cash debt obligations.

Even where the CDS position has terminated, it is conceivable that a creditor’s vote could nonetheless be impacted by its CDS position. For example, a financing transaction wherein Debtor agrees to trigger a credit event in exchange for below-market financing might also encompass future promises to support a debtor-proposed plan of organization. Less maliciously, it is conceivable that the immediate cash receipt from the CDS auction could enable one creditor to support a plan that calls for higher nominal recovery over a longer time horizon in the face of other creditors’ preferences for a lower nominal recovery over a shorter time horizon. While such an agreement or preference is unlikely to be memorialized in writing, voting behavior coupled with a large CDS position might evidence bad faith, pursuant to section 1126, where the CDS position constitutes an “ulterior motive, unrelated to its claim or its interests as a creditor.”

V. Proposals

A. Proposed Amendments to CDS Disclosure Requirements in Bankruptcy

I am not the first to note that disclosure of CDS positions might provide a partial solution to manufactured transactions, both inside and outside the bankruptcy process. Professors Hu and Black were among the first to propose, in 2008, that “disclosure of coupled assets [e.g., CDS and cash bonds] should become a routine part of bankruptcy proceedings, perhaps with an exception for de minimis hedges or general hedges tied to an asset class rather than a particular company’s debt.” Subsequent research has repeated this call for heightened disclosure for reasons ranging from market efficiency to prevention of engineered defaults to maximization of firm value.

However, current disclosure—Regulations SBSR, Rule 2019, section 1102—remains inadequate. Disclosure pursuant to Regulation SBSR does not identify individual positions. Disclosure under Rule 2019 is required—as a practical matter—only for ad hoc committee members, and only for positions held at the time of claim filing or committee formation. Disclosure under section 1102 does not require disclosure of CDS positions at all. Additionally, proposed regulation—Rule 10B‑1—will take a form as yet uncertain and may exceed the confidentiality mandate under Dodd-Frank.

Consequently, I propose several changes to the disclosure requirements in the Bankruptcy Rules. First, I propose amending the Rule 2019(a)(1) definition of “disclosable economic interest” to expressly include “security-based swaps,” incorporating the definition in 15 U.S.C. § 78c(a)(68)(A), and/or the term “credit-default swaps.” Second, I propose expanding the relevant holding periods in Rule 2019(c)(2)(B) (“disclosable economic interest held . . . as of the date . . . the group or committee was formed”) and 2019(c)(3)(B) (“disclosable economic interest held . . . as of the date of the statement”) to include disclosable economic interests held either (i) at any time an ISDA CDS auction was held or (ii) some period of months prior to and including the petition date. Third, I propose expanding these amended requirements of Rule 2019 to all claim filings, ensuring that all creditors—including official committee members—disclose relevant CDS holdings.

While expanding disclosure requirements might not prevent engineered transactions and/or other deviations from cash-derivative equivalency, it would provide necessary visibility to bankruptcy courts. As Robert Gerber, former United States Bankruptcy Judge for the Southern District of New York, has noted,

when anyone in the case—ad hoc committee or not, or distressed debt investor or not—professes to speak on what is best for the estate . . . and/or to influence the outcome of the case, its private agenda can matter. If it does not want to reveal basic information as to its holdings in the case (which are an important indicator of ‘where it is coming from’ in connection with the position it advocates), it should not be trying to influence the court.

If bankruptcy courts understand the full scope of creditors’ economic interests in the debtor, they would be better positioned to understand creditor incentives and behavior throughout the reorganization process. With this more complete understanding, courts would be better equipped to assess whether the good-faith requirements of 1112 (filing) and 1126 (voting) and the notice and hearing requirements of section 364 (financing) are met.

B. Potential Amendments to ISDA Definitions

Whether the ISDA should amend its existing definitions depends on whether market participants view cash/derivative deviation as a feature of CDS or a defect. I have spoken to market participants who fall into both camps; the position of the ISDA is unclear. Consider the interpretive guidance to the 2019 amendments, which clarifies that engineered transactions entered into for “bona fide commercial reasons” or a “bona fide debt restructuring” would still constitute a Failure to Pay, triggering CDS settlement. Such language, coupled with the optionality of deliverability and the Representations section of the 2014 definitions, suggests that these deviations may be a feature, rather than a defect.

However, to the extent these deviations are considered a defect, certain amendments might narrow the gap between cash and derivative equivalency and reduce the potential for engineered transactions. First, applicability of the Credit Deterioration Requirement could be extended from Failure-to-Pay Credit Events to Bankruptcy Credit Events. This would serve to prevent a payout on CDS in the event of bad-faith filings, whether voluntary or involuntary. Second, the interpretive guidance allowing for “bona fide” engineered transactions could be removed. However, both these proposals would create a significant burden for the DC, particularly where a bankruptcy court will—ultimately—make a judicial determination as to whether a filing or transaction satisfies good-faith requirements. Moreover, while the DC is indemnified pursuant to the ISDA 2014 Definitions, section 11.1I(i), one could imagine a scenario where (i) an abuse of DC discretion in determining whether a Credit Event has occurred is alleged and (ii) the money at stake is sufficient to encourage litigation of the scope of DC indemnification.

Third, to prevent an illiquidity-driven squeeze from driving an excessively high or low recovery at auction, the ISDA could place both a minimum and maximum cap—tethered to the IMM—on the Auction Final Price. Ostensibly, the two-step auction process ensures that dealer-estimated markets in step one reflect the market realities observed in step two. However, as Europcar demonstrates, when the Final Price in step two deviates significantly from the IMM in step one, it may be the case that dealer-estimated markets better reflect longer-term valuations, devoid of auction-specific demand/supply constraints.

The author would like to thank the following individuals for their contributions. First, he would like to thank the traders and investors whose expertise and enthusiasm for credit markets helped inspire the direction of this article: Duane Robinson, Jean-Samuel Hentz, and Mark Drabkin. Second, he would like to thank the professors at Columbia Law School who supported and provided valuable criticism of this and related research: Ronald Mann, Jeffrey Gordon, and Joshua Mitts. Third, he would like to thank his editors: Professors Nancy Rapoport, Gregory Duhl, as well as Diane Babal and the staff of The Business Lawyer.