Credit Default Swaps (CDS), like banks, are mind-blowingly simple and potent credit replication machines. Banks replicate their own credit—conjuring gobs of money out of thin air—with a license and a convoluted web of guarantees from the state. CDS can replicate anyone’s credit, with industry-coordinated standard contracts embedded in a tangle of carve-outs, safe harbors, and exemptions from statutes and regulations. The 2007-2009 financial crisis brought a heap of bad press and new regulatory requirements for CDS dealers and trading infrastructure, but did not fundamentally alter the contractual governance paradigm at the heart of this multi-trillion dollar market.
Opportunistic or downright slimy behavior in a space so carefully shielded from substantive regulation can be hard to diagnose, even when it poses an existential threat to the CDS market and could spill beyond it. Recent reports of bad behavior have prompted high-profile lawsuits, contract reforms, and a crop of law review articles revisiting contractual, statutory, and regulatory ecosystems for CDS.
The range of approaches reflects the multifaceted challenge, but can be confusing. Gina-Gail Fletcher’s analysis in Engineered Credit Default Swaps: Innovative or Manipulative? is thoughtful, comprehensive, and a good place to start.
Loaves, Fishes, and Freedom of Contract
Suppose Creditor Cindy Corp. buys $100 million in bonds issued by Borrower Ben, due a year hence, with a $3 million interest payment due in six months. At the other end of the galaxy, Trader Joe—a stranger to Cindy and Ben—bets that Ben is broke. Another stranger, Dealer Dave, takes the opposite side of the bet. He offers to pay Joe $100 million if Ben does not pay Cindy in full and on time, in exchange for a $3 million fee. If Joe accepts, he and Dave would be in a CDS contract on Ben (the “reference entity”), where Joe is a “protection buyer” and Dave a “protection seller.”
If Ben repays Cindy as promised, Dave earns $3 million, just like Cindy. If Ben defaults at the end of the year, Dave loses $100 million, less $3 million, just like Cindy.
The new CDS contract doubles the galaxy’s exposure to Ben’s credit. Absent regulatory constraints, any two counterparties willing to take opposite sides in a bet on Ben’s credit could do the same as many times as they wish. Losses from a Ben default could keep multiplying in distant corners of the galaxy entirely unconnected to Ben or Cindy.1
CDS terms need not track Ben’s original debt contract precisely. Joe and Dave might have a contract for six months (through Ben’s interest payment), or for half the notional principal amount. Because CDS isolate and transfer credit risk, the list of “credit events” that may trigger Dave’s obligation to pay Joe differs from the “events of default” in Ben’s contract with Cindy. The CDS would also ignore Cindy’s actual response to Ben’s distress: the bet is on Ben’s condition, not on Cindy’s risk management.
For instance, if Ben’s interest payment is a few days late, Cindy might let it go if she knows Ben is good for the money, but Dave would still have to pay Joe. If Ben hits a rough patch, he and Cindy might amend their contract to postpone the maturity date, avoiding default to Cindy—but not Dave’s obligation to pay Joe. If Ben files for bankruptcy protection, it would stay Cindy’s claim on Ben, but not Joe’s claim on Dave.
Joe and Dave inhabit a parallel Ben-credit universe, simplified and abstracted from Ben’s real-world relationship with Cindy. Ben’s business problems can affect Joe, Dave, and other perfect strangers elsewhere in the galaxy in ways that are neither visible nor manageable under the Ben-Cindy contract.
Trader Joe and Dealer Dave do not write their CDS contract from scratch. ISDA, the derivatives industry association, has produced standard terms for CDS and other derivatives since the 1990s, leaving the parties to customize a limited set of relationship- and transaction-specific items. ISDA also works with governments and lawyers around the world to ensure that its contracts are enforceable everywhere. Robust industry-coordinated standardization often gets credit for derivatives market liquidity.
Keeping regulators at bay remains an important part of ISDA’s work. To succeed, it must foster public faith in the industry’s capacity to regulate itself: preventing and mitigating market disruptions, responding to shocks, and adapting contracts and market practice to new circumstances. To that end, ISDA has taken on new functions over time. It has established an auction procedure to calculate payouts, and a limited adjudication mechanism in the form of industry-staffed regional “Determinations Committees,” which decide whether a CDS credit event has occurred.
The result is a peculiar ecosystem. The same mechanisms that help the industry adapt and respond to shocks also risk transmitting hiccups in any given bilateral relationship or seemingly isolated contract interpretation to distant markets and otherwise-unrelated firms.
Knaves in Loopholes
ISDA’s CDS contract interpretation posture tends to the formal end of the spectrum. In most cases, Determinations Committee decisions have stuck to plain meaning and the four corners of the contract. This approach is at least intuitively consistent with the emphasis on market liquidity: importing relational context in an arm’s length wager could make it hard to trade.
On the flip side, a commitment to narrow textual reading creates arbitrage opportunities, or worse, invites behavior that is patently inconsistent with the spirit of the CDS contract and its market function.
Enter Hovnanian, a home building firm, and GSO Capital Partners, a hedge fund affiliate of the Blackstone Group. In 2017, GSO offered to refinance Hovnanian’s debt on below-market terms, provided Hovnanian borrowed just over $1 million from its own subsidiary and briefly defaulted on an interest payment. The late payment was enough to trigger more than $330 million in CDS that GSO had bought from another hedge fund, Solus, and Goldman Sachs, but not enough to harm the builder’s credit otherwise. Under ISDA’s auction rules, the debt to Hovnanian’s subsidiary also helped inflate the payout to GSO. Solus lobbied ISDA, made a public stink about GSO’s slimy ways (it had pulled a similar trick in 2013), sued, and settled in 2018.
In the end, Hovnanian got cheaper funding and GSO might have made a bundle off another hedge fund. In the words of one prominent commentator, the incident might have threatened the “integrity and viability of the CDS market,” but had no serious “real world” (read real economy) consequences. Who cares?
Fletcher’s article does a lovely job wading through the thicket of pros and cons to answer this question. The cons come out ahead precisely because she takes CDS to be valuable as a source of market liquidity. If CDS contracts detach from the underlying credit risk and become bets on the likelihood of collusion between the reference entity and a CDS counterparty (in the manner of insurance fraud, minus insurance regulation), they lose value for CDS and credit market liquidity, price discovery, and credit access.
The article’s typology of “engineered CDS”—miscreant behavior involving CDS counterparties and the reference entity—covers Hovnanian-style “manufactured” defaults driven by protection buyers, as well as protection seller schemes to kick default just beyond the CDS contract term, or to frustrate the point of a CDS contract by moving all debt from the reference entity to its subsidiary. In all three transaction types, the trouble starts in the CDS parallel universe; smoke may drift into the real world, but the harm is uncertain and would be hard to ascertain empirically. 2
To the extent these arrangements are harmful, Fletcher’s doctrinal analysis is especially worthwhile for illustrating the breadth and interdependence of contractual, statutory, and regulatory maneuvers shielding CDS contracts from government intervention. Express contract terms and recent court decisions virtually foreclose arguments about good faith, while the SEC and CFTC jurisprudence make it hard to prove manipulation, particularly where the existence of engineered transactions is widely known and baked into CDS prices. This might explain why the SEC and the CFTC have limited themselves to censorious press statements.
Meanwhile, ISDA has tweaked its contracts again, putting more stock in adjudication and Determinations Committees’ judgment. The Determination Committees now have the authority to distinguish between credit events stemming from credit deterioration and those (engineered ones) detached from it.
I share Fletcher’s skepticism of Determination Committees’s capacity to lean into this new authority; I also suspect it is as good as it gets. Instead of investing in ISDA’s already strained adjudication apparatus, I might prefer to revive good faith3 (the article offers two ways to go about it) and, even less plausibly, a return to public adjudication. Generalist judges might get it wrong sometimes, but the derivatives industry is better equipped than most to compensate for judicial missteps. Shielding it from inexpert scrutiny has not made the public more informed or more sympathetic, and has hardly been good for its own health.
- If Creditor Cindy takes such a bet once, she would not create net new exposure to Ben—all else equal, buying a CDS contract on Ben from Dealer Dave would simply transfer Cindy exposure to Dave. However, nothing prevents Cindy from buying three or ten CDS contracts, which would create net two or nine new bets on Ben.
- Other authors situate the Hovnanian caper in different taxonomies. Henry Hu focuses on information asymmetries on the eve of bankruptcy, and the behavior of “net short” creditors with complex incentives in a thoroughly opaque market (more disclosure is indicated). Ted Janger and Adam Levitin cite the incident in their analysis of economic and governance rights decoupling in and out of bankruptcy. Robert Rasmussen and Michael Simkovic discuss it as an instance of flawed contract drafting and opportunism, upending counterparty expectations despite abiding by the letter of the contract. They propose to pay bounties to non-parties for identifying contract flaws. Their exposition is excellent; however, the challenge of telling contract flaws apart from ordinary contract incompleteness strikes me as formidable.
- Reading Fletcher alongside this recent account of distressed debt shenanigans and pitch for good faith makes me think that reviving the doctrine could be a useful check on financial engineering abuses beyond CDS.