Why do Central Bankers Intervene in the Foreign Exchange Market? Some New Evidence and Theory

I provide new empirical and theoretical evidence about the effectiveness of sterilized interventions on exchange rates. These new developments are particularly important to understand why central bankers from developing countries tend to intervene during periods of financial distress. In the first half of the paper, I apply a VAR formulation to measure the effects of sterilized interventions on the U.S. bilateral exchange rate. Information from the Exchange Stabilization Fund in the U.S. for the period 1974 -- 2000 is used to identify a shock that is orthogonal to the U.S. money supply and therefore mimics the role of sterilized interventions. According to my identification strategy, a sterilized intervention shock in favor of the U.S. dollar would appreciate it against a trade-weighted currency index by roughly 1 percent. This appreciation is statistically significant, lasts for about 1 year, and is robust to alternative identification strategies. Then, I devote the second part of the paper to rationalize the results from the empirical section by studying sterilized interventions within a two-country general equilibrium model. I find that if trading bonds is costly worldwide and asset markets are incomplete, a domestic government purchase of domestic bonds accompanied by a sale of foreign bonds, a sterilized intervention, appreciates the domestic currency. Accordingly, a calibrated version of the model renders similar results to those from the VAR formulation.