Measuring customers portfolio concentration for rating agencies: evidence from Fitch, Moodys and S&P

Purpose – The aim of the paper is to study the degree of independence of customers' portfolio concentration measure from the pricing policy adopted by rating agencies. Design/methodology/approach – The paper tests different measures of customers value (revenues or profits and customer lifetime value) and different concentration measure (top customer or Herfindahl-Hirschman index) on the customers' portfolio of rating agencies in the time period 1999-2008. Simulating different pricing models, the paper tests the sensitivity of these measures to discounted fees applied to best customers and identifies measures that are more and less sensitive to the discount applied. Findings – Concentration measures that consider all the customers' portfolios and look at both cost and revenues related to the service on a multi-period time horizon (CLV) are less sensitive to the discount policy respect to the others. Research limitations/implications – Results point out some opportunities related to apply more complete approaches defined by marketing science on the financial service industry in order to construct better measures for the economic independence. The paper works only with publicly available data and more details about the fee applied to each customer could increase the significance of the results achieved. Practical implications – The paper contributes to the current debate on the economic independence of rating agencies stressing the opportunity of rethinking the measures on economic independence that are currently considered by supervisory authorities. Social implications – The paper is the first empirical application of standard marketing concepts of customers' concentration measure to the rating industry. Originality/value – The paper studies the pricing policies adopted by ratings agencies.