How Would a New Bankruptcy Regime Help?

A Key Issue In The debate over reform of the international financial system is whether to adopt a new international bankruptcy regime for sovereign debt, and an important aspect of that debate is the lack of consensus about the problems that such a regime would be designed to solve and whether it would solve them. My remarks are focused on a new bankruptcy regime rather than a new court, because the former does not necessarily entail the latter. The authors of the three papers in this symposium all have quite different conceptions of the problems, as do other important protagonists.

Bulow's paper, like his past articles on this subject, is focused on the need to reduce bailouts of developing-country debt by international financial institutions (IFIs) and governments in the industrial countries, so as to achieve more optimal flows of private capital to emerging markets. Sachs' paper sees two different roles for the bankruptcy process: avoiding a creditor grab race for the sovereign debtor's assets, and giving the sovereign debtor a new start by substantially reducing its debt. White's paper mainly addresses two objectives that differ from both Bulow's and Sachs': preventing rogue creditors from upsetting private debt reorderings and making sure defaulting countries have adequate sources of funds, through granting seniority to creditors who lend in the wake of a default. The U.S. Treasury, in the form of John Taylor's statement during the same week as this conference,¹ has yet a different objective. He wants to reduce the uncertainty of the current process by providing a more orderly structure for resolving crises when they occur. Finally, the purpose of the IMF proposal offered by Anne Krueger is more encompassing. In her proposal last November she said the IMF's aim is "to create a catalyst that will encourage debtors and creditors to come together to restructure unsustainable debts in a timely and efficient manner."²

Krueger's latest proposal for a sovereign debt restructuring mechanism (SDRM) contains the following elements: a stay for some fixed duration on creditors' seizure of a troubled debtor's assets, reform of the debtor's economic policies as part of the bankruptcy plan, seniority for creditors who lend after the petition for a stay, and a mechanism enabling a super-majority of creditors to bind other creditors to a plan. How does this proposal relate to the objectives identified by Krueger herself, Taylor, and the others?

The relationship between Krueger's SDRM and reducing bailouts is not clear. One could argue that this objective could be achieved more directly than through adoption of an SDRM, by placing restraints on crisis lending by the IFIs. This could be done in a variety of ways. Under the proposal by Adam Lerrick and Allan Meltzer,³ IMF support would be limited to a commitment to buy the debt of defaulting sovereigns at some discount from the price being offered by the sovereign. Under a proposal offered jointly by the Bank of Canada and the Bank of England,⁴ there would be procedures like those adopted in the United States in 1991 under the Federal Deposit Insurance Corporation Improvement Act for official lending to failed banks. Lending would be permitted only when there is a threat to the stability of the international monetary system and only after approval by a supermajority of the IMF's executive board.

Krueger's proposal for an SDRM addresses the bailout problem more indirectly. The idea is that pressure for IMF lending would be lessened if the SDRM offered a mechanism by which a troubled sovereign debtor and its creditors could reach an agreement to reduce the sovereign's debt. One reason that IMF lending has been so prevalent is that there is no workable mechanism for reducing debt. Indeed, the argument would be that the IMF could not risk the chaos, economic and political, that might ensue in the absence of its lending unless an alternative debt reduction process were in place. Of course, in the short term, this is proving wrong in Argentina. One can just say no, without an SDRM...