Post-Crisis Derivatives Regulation: What went Right (and what went Wrong)

The perception that derivatives are exotic and uniquely risky financial instruments has given rise to a confusing, incomplete, and contradictory regulatory patchwork. My new article rethinks how derivatives should be regulated.

Much of the outstanding scholarship discusses derivatives according to industry-derived categories, which do not provide an analytical foundation for regulation. This article strives to build that foundation by de-mystifying derivatives and showing that they can be deconstructed, by their economic functions, into two categories of traditional legal instruments—option contracts and guarantees.

The article shows that, like those traditional instruments, most derivatives are neither exotic nor uniquely risky. The perception that derivatives are uniquely risky because they are ‘bets’ is misleading because virtually all financial instruments are bets. The perception that derivatives are uniquely risky because they are much more volatile than other financial instruments fails to recognize that derivatives counterparties usually can estimate the limits of their potential liability. Indeed, the disclosure of this liability is an accounting requirement, and there are a range of methodologies to estimate potential liability for even the most complex derivatives.

The article argues that derivatives can be especially risky, though, when they function as financial guarantees. Credit-default swaps (‘CDS’) epitomize this category of derivatives.

All guarantees are risky because they ensure future events. Non-financial guarantees, however, typically ensure reasonably predictable events. In contrast, financial guarantees, including CDS, ensure less predictable risks and also are subject to cognitive biases. For example, because they do not actually transfer their property at the time they make a guarantee, financial guarantors may view their risk-taking more abstractly than, say, a lender that advances its own funds to a borrower. This ‘abstraction bias’ causes financial guarantors (and thus guarantors under CDS contracts) to underestimate the risk, even after discounting for the fact that payment on a guarantee is a contingent obligation. Empirical findings confirm that abstraction bias is real and that it can influence even sophisticated financial guarantors.

Furthermore, financial guarantee risk—and thus CDS risk—is magnified when the contract has one or more systemically important counterparties. CDS contracts often have at least one systemically important counterparty. Because systemically important counterparties are highly interconnected, a default by a CDS counterparty can sometimes threaten financial stability.

The absence of an insurable-interest requirement further magnifies CDS risk. This requirement mandates that a person taking out insurance must derive some benefit from the continued existence of the insured person or property. An insurable interest mitigates moral hazard by reducing the incentive for the person taking out the insurance to kill the insured person or destroy the insured property in order to collect on the insurance policy. There currently is no ‘insurable interest’ requirement for CDS contracts.  

A CDS contract that lacks an insurable interest is called a ‘naked’ CDS. Naked CDS contracts can also be used for speculation, which additionally can magnify risk by creating an unlimited multiplier effect. For example, a derivatives speculator might receive fees for guaranteeing the same $100,000 bond under ten different naked CDS contracts.

The article analyzes how CDS contracts that have one or more systemically important counterparties should be regulated, and how the absence of an insurable interest should affect that regulation. This analysis begins by building a normative framework for designing financial regulation. The framework takes into account, among other things, the goal—and the costs and benefits—of correcting market failures. The article then uses that normative framework to critique CDS regulation.

The article argues, for example, that current derivatives regulation may both underregulate and overregulate derivatives. It may underregulate derivatives by inadequately addressing systemically risky CDS contracts. The article contends that regulation should set limits on the CDS credit exposure of systemically important firms, and also explains how that regulation could help to correct abstraction bias and other cognitive biases that can motivate excessive CDS risk-taking.

Current regulation also may overregulate derivatives in two ways. First, it requires that most derivatives be cleared and settled through central counterparties (‘CCPs’), which are well-capitalized entities often associated with derivatives, commodities, or other securities exchanges. The goal of this central clearing requirement is to reduce counterparty risk; by legally substituting its credit for that of the contracting parties, the CCP becomes the primary counterparty on both sides of the derivatives contract. The United States and many other countries follow this regulatory scheme. Although intended to reduce risk, central clearing could inadvertently increase systemic risk by shifting counterparty risk from individual counterparties to the CCP, thereby concentrating the risk.

Second, current regulation imposes minimum margin requirements for non-centrally cleared derivatives. Counterparties must provide securities or other liquid assets as collateral to secure their obligation to settle a derivatives contract. Margin requirements, however, can be a mixed blessing. Although they can help to protect against counterparty default, they can lead to unpredictable liquidity-funding demands if the counterparty lacks sufficient liquid assets to post as margin. Indeed, AIG’s near failure resulted from its inability to satisfy CDS margin requirements.

Steven L. Schwarcz is the Stanley A. Star Distinguished Professor of Law & Business at Duke University School of Law and Senior Fellow, the Centre for International Governance Innovation.