A note on the geographical dispersion of emerging market currencies: The role of outliers

Abstract

The geographical dispersion of currencies is often studied in a cross-sectional convergence regression using data from the BIS triennial survey (see Cheung et al. (2019)). Using three different approaches to outlier detection, we demonstrate that a small number of outliers can have a disproportional impact on the empirical results. These outliers are often identified as large trading centers. For the 2019 to 2022 BIS triennial survey, our findings reveal substantially stronger evidence of convergence to the share of all currencies, once the outliers are excluded from the sample.