Description

Using daily data on intervention activities by the Central Bank of Chile between 1998 and 2003, this paper studies the impact of these actions on the peso/dollar exchange rate. First, results are obtained from an event study, finding no evidence of a significant effect of interventions on the level and trend of the exchange rate. However, these findings could be biased due to the omission of relevant control variables and the simultaneity between interventions and the exchange rate. Time series estimates are used to deal with these potential problems. Statistical evidence of weak exogeneity between intervention and exchange rate shows that the potential endogeneity problem is not relevant given the purpose of the paper. OLS estimations suggest that the effect of daily interventions are not statistically different from zero in 2001, but that the public announcements of the beginning of an intervention period did have an effect in the level and trend of the exchange rate in that year, as well as in 2002-2003. The announcement effect is nil in 1998, as interventions were carried out with less transparency. The results are robust to alternative specifications and instrumental variable estimations.