Discharging the Discharge for Value Defense

For those seeking watershed moments in contemporary contract law, the area of corporate debt seems an unlikely target. Though gargantuan in size, debt markets have a storied reputation as a refuge for the risk averse—participants expecting stable payouts, low volatility, and few surprises. Nevertheless, corporate debt contracts are themselves notably lengthy and complex. When parlayed with the immense financial sums at stake, that complexity can become a recipe for calamity. And in late 2020, calamity struck in the form of a nearly $1 billion accidental payoff sent to Revlon Inc.’s distressed creditors—not by Revlon itself but rather by Citibank, the administrative agent for the loan. When several lenders refused to return the cash, Citibank commenced what many reckoned would be a successful (if embarrassing) lawsuit to claw it back. But in a dramatic 2021 opinion, a New York federal court sided with the debtholders, applying an obscure equitable doctrine known as the “Discharge for Value” defense. The lenders could keep their wayward windfall, and Citibank got stuck with a sizeable write-down. The decision is currently on appeal to the Second Circuit; but whatever its ultimate resolution, the case seems destined to feature prominently in contracts classes and textbooks for years to come.

In a new working paper, titled “Discharging the Discharge of Value Defense,” I use the Citibank opinion as a lens to understand how contract law evolves, both inside and outside the courtroom. First, I spotlight several doctrinal and logical irregularities in the District Court’s opinion (wrangling several observations of my own along with those of several other collaborators and commentators). Second, I build on these inconsistencies to advance numerous skeptical “hot-takes” on the outcome from a contractarian perspective, criticizing Citibank using an assortment of contract design desiderata that include information disclosure, relational contracting, and efficient allocation of risk and incentives (even throwing in for good measure an extended metaphorical analysis of optimal Whac-a-Mole strategies).

Third (and most substantially), I kick the tires of my own skepticism as an empirical matter. To the extent that the criticisms offered above have merit (and are not merely armchair theorizing), they generate testable predictions about how sophisticated market participants would react. Significantly, regardless of whether one lauds or loathes the reasoning in Citibank, it is clear that the ruling announces a default rule, and is thus subject to alteration (at some expense) by express contractual provisions. Consequently, if Citibank imposed the disruptions and inefficiencies that I posit as a conceptual matter, then it follows that sophisticated contracting parties would respond to it not by reconstituting their internal clerical protocols, but rather by refashioning their contractual language—narrowing the holding or even negating it writ large. In fact, several professional associations advocated doing just that, and they even suggested model language for credit agreements that purportedly did the trick (provisions that have come to be known as “Revlon blockers”).

Using a hand-collected data set of publicly disclosed debt contracts from January 2020 through the end of July 2021, I isolate the incidence of express contractual provisions related to mistaken payments. This time span allows one to analyze not only the response to the Citibank litigation and opinion, but also the practices that prevailed beforehand. I then use a variety of computational text analysis tools to assess the semantic content and structure of such provisions, and I deploy several standard empirical tools from finance to tease out both the drivers of adoption and market reactions. My analysis yields four key findings:

• First, a small but detectable trickle of Revlon blocker provisions began to take root right after the public disclosure of Citibank’s gaffe, and just as the litigation was heating up. But that trickle swelled to a veritable flood (well into the hundreds per month) almost immediately after the opinion issued in February 2021—a trend that substantially continued thereafter. By contrast, there are virtually no explicit provisions that endorsed the trial court’s interpretation. This pattern is consistent with the arguments that (a) the holding delivered a surprise result; and (b) the surprise was an unpleasant one to many market participants.
• Second, my analysis yields insights about the structure and content of the contractual provisions that adopters embraced. Using a variety of tools from machine learning, I show that—somewhat surprisingly—Revlon blockers do not follow a single “cookie cutter” template, where parties copy and paste identical template language from deal to deal with little variation. While the most prominent model provision also appears to be the modal provision, my analysis suggests that there have also been at least two other clusters (or “families”) of Revlon blockers that market participants have embraced, both of which are distinct from cut-and-paste near-clones of a pre-existing model.
• Third, I show that adoption of blockers has been wide ranging across the most sophisticated firms in the market. Adoption does not seem limited to a single industry, sector, or incorporation jurisdiction. Adoptions do, however, tend to be more concentrated among firms with more at stake: although firm size alone is not dispositive, adoptions are strongly concentrated in companies with larger absolute and relative debt loads and issuers with high relative profitability (as measured by return on assets). Trading premia, in contrast (as measured by Tobin’s Q), are negatively associated with adoption. These findings suggest that adoptions are concentrated among those firms with the largest stakes and with elevated prospects for shareholder-debtholder conflict.
• Finally, and somewhat more preliminarily, I uncover evidence about the relationship between Revlon blocker adoption and market reception. Using an event study approach, I find a positive (but modest) price response to the mean adopter’s first disclosure of a Revlon blocker. In light of the possibility that news of blocker adoption may have leaked prior to its public disclosure, I also consider the effect of the Citibank opinion itself (which seems clearly to have been a surprise). Here, I find discernible positive abnormal returns for adopting issuers (as well as for predicted adopters) in the days following the opinion. While this evidence is admittedly partial and incomplete (g., it does not measure gains in contractual surplus to all parties), it is suggestive that the market on the whole has approved of Revlon blocker adoption.

My findings have both specific and general implications. As to the specific, they constitute probative inputs for the relevant appeals courts to consider in reviewing the Citibank case (and others like it). Given the important (and still unresolved) doctrinal issues at play, it would not be surprising if the Second Circuit were to seek counsel of New York Court of Appeals by certifying certain of these issues to them. But regardless of whether that happens, the strong resistance that the lower court’s opinion has met with among commentators and market participants provides a powerful reality check for assessing parties’ expectations, fairness, commercial reasonableness, and the (ultimate) question whether reversal is warranted.

More generally, my analysis helps to illustrate how legal doctrine, legal theory and empirical evidence can helpfully interact. In so doing, it adds to an emerging theoretical and empirical literature on how contractual practices evolve to reflect, channel, and sometimes push back on legal precedents. Courts and policy makers would do well to pay attention to how their opinions create shock waves that cascade through (and are embraced/resisted by) market practices. Assessing such feedback effects can be enormously useful to inform later course correction efforts as contract doctrines and practices evolve. Attending to such forces can even ease the initial burden on judges themselves: for the task of establishing fair and efficient default rules on the spot is tricky, and judicial actors frequently lack requisite information to make such pronouncements confidently—simply because such information simply does not exist at the time they must render decisions. Field testing new doctrinal innovations may be the best (and sometimes the only) way to assess their relative virtues and vices, providing a lodestar for the course ahead. Ignoring such feedback, in contrast, could prove to be much worse than a $1 billion “fat-finger” mistake.

The complete paper is available for download here.