I’ve noticed a misunderstanding in informal discussions over speed bankruptcy, bail-in, debt-for-capital resolutions (however you might want to call it). There is, on the one side, the Luigi Zingales & Oliver Hart plan (Lucian Bebchuck is mentioned as an inspiration for this plan), and on the other the Mark Flannery / Garret Jones plan. There’s been many different versions of both of these plans and many authors I’m not citing, but for the sake of simplification I’m going to talk about Zingales’ plan and Jones’ plan. Those are the versions I have in mind. While the motivations for both of these schemes are to limit externalities of systemically important financial institutions failures on other banks, these propositions are not the same, they are actually quite different.

Zingales proposes contingent-capital debt contracts that would turn to capital after a pre-insolvency ratio is triggered, with the possibility for shareholders to out-buy these debt holders. I’ve blogged on the difficulty of finding robust pre-insolvency triggers before. Jones’ proposal is akin to a variation on current FDIC’s resolution procedures (and extension of their power to non-bank financial institution). Under Jones’ plan, at the initiation of bankruptcy procedures regulators would forfeit shareholder rights and arbitrarily turn some debt contracts into ownership claims overnight. Zingales’ tries to prevent bankruptcies, while Jones’ tries to limit its costs.

The latest French Banking reform proposal, for example, seems inspired by a version of Jones’ proposal. As my colleagues have argued, it is bound to run into Constitutional challenges; without due process and with limited, ex post, judicial review it is reminiscent of expropriation. Moreover, if the goal is to avoid costly firesales perhaps automatic instant conversion should be used with caution; current regulation makes it complicated and costly for most financial institutions to hold equity in their books. It is difficult to foresee the kind of transactions such bankruptcy resolutions would arouse.

In any case, it is not obvious that speed bankruptcy is necessary in the first place. The literature on the costs of large financial institution failures does not support doomsday scenarios of systemic risk. Studies on the US market suggest that banks are always too diversified to be pushed into insolvency by counterparty losses alone. Research on the much feared bank runs suggest that they aren’t really conducive to systemic risk either. In these episode banks go bankrupt only when they were already insolvent prior to the bank runs. Bank failures are clustered not because of contagion, but because banks invest in the same kind of assets. It is a problem of homogeneity rather than contagion. Also, speed bankruptcy, in both kinds, seems to implicitly take for granted that the current use of assets is always more valuable than any other alternatives out there. That is to say, that a form or reorganization is always preferable to liquidation. This is, at the very best, questionable.

That doesn’t mean there aren’t any good reason why you’d want such a scheme, but they’re not necessarily the ones being put forward. One interpretation of the 1991 FDIC Improvement Act experiment, a law that was thought by the profession to have permanently eliminated Too Big to Fail in the U.S., could be that the checks and limits to bailout powers will always be trampled. A credible, economically viable, alternative to bailouts that would render them redundant would definitely have its place.

Advertisements