Debt models are simple, but touchy, explained Michael Dooley in an IMF Institute—sponsored seminar on September 15 on the IMF and the private sector. The former member of the IMF’s Research Department and current professor of economics at the University of California at Santa Cruz urged staff to reexamine the assumptions that underline IMF debt models and take a fresh look at the nature of sovereign debt. Small changes in assumptions, he said, can make big differences in policy conclusions and prescriptions.
According to Dooley, the IMF fashions its debt model from the world’s lengthy experience with domestic banking crises. The analogy may not be completely apt, he said, challenging staff to delve more deeply into why money is lent internationally, what enforcement mechanisms make this lending possible, and why international debt contracts are so difficult to renegotiate. The answers to those questions, he said, suggest an alternative model and a different approach to private sector involvement in crisis resolution. While clearly more needs to be done to design better incentives to avoid and resolve crises, several innovative aspects of the recent debt restructurings, notably that of Ecuador’s debt, left Dooley somewhat encouraged.
Understanding sovereign debt
All models of sovereign debt, Dooley explained, work backward from an assumed punishment for default. Creditors will not lend unless there are enforcement mechanisms in place to ensure repayment. Traditional debt models posit that the threat of trade sanctions or the loss of access to future credit serves such a function. But Dooley countered that neither of these outcomes has been observed in practice. His model assumed an alternative enforcement mechanism—one that is based on the observation that a sharp decline in output results when debtor countries default. These defaults occur when countries cannot renegotiate contracts quickly, and a collapse of domestic financial intermediation then ensues. This breakdown, which customarily impairs the usefulness of a country’s entire capital stock, provides a powerful incentive for debtor governments to avoid default.
Dooley also observed that in developing countries, private debt is really two instruments—one in which all goes well and the debt will be repaid by the borrower and another in which things go badly and costs are socialized. International debt, in effect, is always sovereign in developing countries. And it is important to understand that debt contracts only write down what happens in good states of the world. In defaults (bad states), what is written has meaning, but not the meaning that the words seem to imply.
Because debtors want to borrow and creditors want to be repaid, both agree to contracts that are intentionally very difficult to renegotiate. Contracts are written to avoid strategic defaults—instances in which debt could be repaid but is not. The IMF and others make a big mistake, Dooley said, when they assume that contracts are designed to address a “bad luck” state in which the debtor genuinely cannot repay. “That would be nice,” he noted, “but it’s not the world we live in.” Contracts are designed to invoke automatic punishment regardless of the reason for nonpayment. Long, hard renegotiations produce output losses—exports often boom, but the domestic financial system collapses (since other creditors are unwilling to enter a situation in which their position in the repayment queue is uncertain). Output then falls, capital flees, and the country crashes.
In bad luck circumstances, the creditors suffer as well as the debtors, but that, Dooley stressed, is what the system is designed to do. Contracts are not, and cannot be, conditional on luck. And that seemingly unfortunate circumstance is compounded by the current system in which it is neither the borrowers nor the lenders who suffer the loss of output (workers in debtor countries will bear the brunt of lost output) or finance the eventual bailout (taxpayers in debtor countries will essentially pick up the tab for this). This is a distortion of the first order, Dooley contended, but fixing it will require a clearer understanding of what is happening and why.
Finding a fix?
Something is indeed wrong, Dooley said, and, as many have suggested, the problem is one of coordination. But in his model this coordination problem is intended rather than accidental. A third party might be able to distinguish between bad luck and strategic defaults, but Dooley doubted that the IMF could fulfill this task. A third party can—as the IMF did in the 1980s during the debt crisis—force banks to roll over debt or provide enough assistance to maintain the status quo and prevent a loss in output. But as the 1980s suggested, waiting does not help. And the major industrial countries do not want the IMF to determine the circumstances in which private lenders should be paid. Dooley concluded that if interference with the private sector’s collection processes was not credibly conditioned on the cause of default (and he believed the IMF would never have the political backing to do this), then intervention of this sort would simply reduce international lending to low levels.
Photo Credits: Denio Zara, Padraic Hughes, Pedro Márquez, and Michael Spilotro for the IMF, pages 353, 354, 360, 361, 362, 363, and 364; Shaun Best for Reuters, page 353; Shawn Thew for AFP, page 358.
According to Dooley, periods of high capital flows are synonymous with instances in which the normal constraints on lending are lifted—either through implicit guarantees (such as the ones industrial countries gave to banks because they wanted oil money recycled) or through access to official credits (such as was the case with eastern Europe and Russia after they joined the IMF). Capital inflows continue until the collateral runs out. Then there is a run on the debtor government. If the debtor government’s resources are sufficient to bail out the private sector, a contract renegotiation is not necessary.
Runs should be distinguished from defaults, Dooley cautioned. Korea experienced a large run, but almost no default (because of the size of its bailout), and output loss, though sharp, was short-lived. Indonesia, by contrast, suffered a large run and a massive default because it did not have enough—still does not have enough—official resources for a bailout. Hard thought needs to be given to the relationship between what is expected and what happens, and it will be important to devise a means to forecast whether runs on reserves can lead to financial disruption. That is why, he added, “currency crises are not a big deal in industrial countries and are a big deal in developing countries.”
The IMF views international financial crises from a different perspective, however, Dooley noted. The IMF—drawing heavily on the conventional wisdom that sees an analogy between domestic bank runs and international financial crises—views liquidity runs and contagion as triggering an overreaction in asset prices and subsequent substantial losses in output. From the IMF’s vantage point, then, these seemingly unanticipated—and largely unexplained—shifts in expectations are the culprit. From this, it follows that crises in confidence are inevitable and require a lender of last resort. Moral hazard results, but it is an unfortunate, though acceptable, cost of producing a necessary public good.
This perspective also argues, he said, that crises can be prevented through more effective regulation and better information and through punishment (haircuts) meted out to those at fault. What this view does not account for, Dooley insisted, is why short-term international debt exists in the first place. Ultimately, explanations of why there is international debt return to issues of enforceability and suggest why it is so hard to improve the system by punishing lenders after the fact.
Dooley noted that experience with the Brady bills underscored the fact that creditors managed to draw value from something worth essentially nothing. The conventional response to this state of affairs is to fault collective action clauses, which seem to place unreasonable bargaining power in the hands of creditors. Breaking the capacity to hold out then seems to be the best way to fix the problem. But Dooley was dubious. In his own model, this simply weakened enforcement mechanisms and ultimately reduced international credit levels.
He was more heartened by several recent restructuring agreements that seemed to reduce the incentive for holdouts by making debt not tendered more susceptible to default. In these instances, majority clauses were used to subordinate the debt of holdouts. More promising still, he said, is an element in the Ecuador restructuring agreement that reduces the contractual value of debt but makes this reduction contingent upon the absence of future defaults (in other words, the reduced debt will be added back in the event of a future default). For Ecuador, this amounted to a lot of debt reduction (40 percent) at very little cost. And better yet, it addressed the incentives question that lies at the heart of much of the debt debates, since the contingent clauses substantially increase everyone’s incentives to avoid future defaults. This clever means of overcoming distortions, Dooley suggested, may offer a market-based means to distinguish between bad luck and strategic defaults.