Endogenous dollarization, expectations, and equilibrium monetary policy

Emerging market countries have trouble soating, and many that claim to soat do not deliver on such promises. That is a main conclusion of much recent empirical work, starting with the papers of Calvo and Reinhart (2002) and Stein et al (1999). The reason for this would seem to be dollarization of liabilities and balance sheet effects. Calvo (1999 and 2000), Krugman (1999 and 2000), Stein, Hausmann, Gavin, and Pages-Serra (1999), and Aghion, Bachetta and Banerjee (1999), among others, make that case. If debts are denominated in dollars while Þrms depend on local currency revenues (or, more precisely, revenues increase with the relative price of goods produced at home), sharp and unexpected changes in relative prices matter for Þnancial stability. The policy conclusion that emerges from this line of work is that sexible exchange rates can be destabilizing, and therefore emerging market nations would be well advised to design alternative arrangements, including currency boards and dollarization. Such a view has become extremely insuential. But even its most ardent advocates understand that it is only half the story. The claim is that soating is infeasible for a given dollarized debt portfolio. But of course portfolio choices — for instance, what shares of debt to hold in peso and dollar-denominated bonds— depend on the risk-return characteristics of these securities, which in turn depend on the structure of shocks and expected monetary and exchange rate policies. Following standard asset pricing and portfolio choice models, variances and covariances (especially with consumption) should matter. Several authors —Ize and Levy-Yeyati (2003), Ize and Parrado (2003) and Moron and Castro (2003) among others— have recently used this approach, in partial equilibrium, to model endogenous dollarization in emerging markets. ∗Preliminary and incomplete. Comments welcome. Prepared for the San Francisco Fed Emerging Markets Conference, June 2004. We are grateful to the National Science Foundation for Þnancial support. †Email: chang@econ.rutgers.edu ‡Email: andres_velasco@harvard.edu