Rank–size distributions for banks: A cross-country analysis

Rank–size analysis is a powerful methodological instrument for describing the main features of a unified system by moving from an available sample of its disaggregated elements. We here study the rank–size relationship of banks’ total assets in a sample of 13 advanced economies. We show that a standard power-law function is unable to provide an adequate fit of the data, while a Discrete Generalized Beta Distribution provides a statistically satisfactory fit, being able to capture the behaviour of the distribution at low and high ranks. We then analyse the relationship between the parameters of the Discrete Generalized Beta Distribution estimated in each country and the main features of the country's banking industry. Our results point to a connection with the degree of development of the banking sector and with some regulatory and institutional characteristics, but no relationship with riskiness.