Comments and Discussion

Ricardo Caballero:

This is a must-read article for anyone interested in collective action clauses for emerging market debt. It provides a good historical background of the issue, follows with a simple but useful organizing framework, and then uses the framework to formulate testable hypotheses, which happen to be confirmed by the data. There are many interesting findings in the paper. Among my favorites are: low-credit countries with many bonds would benefit from a super-CAC that lowers the anticipated complexity of restructuring; individual CACs would not be good substitutes in such a situation; moral hazard is a serious concern for investors purchasing a low-credit country's new bonds with CACs; and such concerns are less important during periods when the EMBI spread is low. On top of this, the paper is refreshingly balanced and moderate in its claims.

While this modesty is a strength, it is also somewhat of a weakness (although I suspect the authors are aware of it but do not mind). Indeed, the paper is so balanced that it is almost impossible not to experience a nagging feeling that the issues being discussed are not of first-order importance. In fact the authors are well aware of this as a possible interpretation of their findings. At one point in the paper they acknowledge that CACs, by themselves, are not likely to be a big thing in the broad and worthy agenda of improving the financial stability of emerging market economies.

Later in the paper the authors go on to mount a mild defense and argue that while CACs may not have been too important in the past, their relative importance is bound to rise in the future as other sources of volatility, in particular the danger of borrowing short term, are unlikely to repeat in the future.

At that point, I finally was given the metric to decide that indeed I think the CACs issue is a second-order one (which does not mean that it should not be addressed.). The authors seem to mean that countries should avoid short-term debt since it is too risky. Further, it seems this view is becoming conventional wisdom. However, the cost for emerging market economies of not being able to borrow short term is an order of magnitude larger than the cost and benefit of having (or not having) CACs. No U.S. corporation would be asked to give up inexpensive short-term borrowing. Moreover, this is only the tip of the iceberg. Emerging market countries also accumulate large amounts of international reserves, build stabilization funds, withdraw into precautionary recessions at the first sight of external troubles, and so on. In sum, they act scared.

What do these countries fear? I believe it is capital flow volatility. All of the above are very expensive self-insurance practices against this volatility.

The theme of default and CACs fits nicely into this insurance perspective. Default can also be thought of as insurance if the country uses it during bad states of the world, as is invariably the case. From this perspective, CACs can be seen as reducing the inefficient waste that arises from restructuring complexity and holdouts in such insurance contracts. In the sovereign context, the problem with this form of insurance, which is exacerbated by CACs, is that too much of the insurance-trigger decision is left to the borrower.

Could one do better? I believe so. One needs to create insurance opportunities with much wider coverage (in terms of countries and scenarios) that are less susceptible to opportunistic behavior by negligent countries.

Default only covers countries undergoing deep crises—highly illiquid and bankrupt economies. An important share of the costs of capital flow volatility are borne by countries that experience deep contractions but do not undergo full-blown crises; even for those countries that do fall into deep crises, many of the costs are incurred well before the run phase of the crisis develops. The anticipation of a more orderly workout if the crisis phase arrives also would eliminate (by backward induction) some of the costs that precede these events. But this benefit is...