Are Competitive Banking Systems Really More Stable?

Abstract

This study replicates Schaeck et al. (2009), henceforth SCW. SCW conclude that (i) concentration and competition (as measured by Panzar and Rosse’s H-statistic) represent two separate dimensions of the banking sector, with (ii) greater competition being associated with greater financial stability. Using their data, we are able to exactly reproduce their original results. However, when we use current vintage data for the variables in their dataset, from both the same and alternative data sources, we find that H-statistic fails to attain significance at the 5% level. We obtain this result even though we use the same variables, estimation, and sample as SCW—just with more recent data instead of the earlier (later revised) numbers provided by the data vendors. Additional tests, such as expanding the timeframe from 1980–2005 to 1980–2011, employing more recent data for H-statistic and concentration, or using the z-score as an alternative measure of financial distress, further confirm this result. Further, not only are the estimates statistically insignificant, but they are economically insignificant, with small effect magnitudes. Our paper suggests that competition may not positively contribute to financial stability after all.