Data for: Sequence-Based Clustering Applied to Long-Term Credit Risk Assessment

Published: 2 September 2020| Version 1 | DOI: 10.17632/z2njc4yvf4.1
Hyejin Ku, Doobae Jun, Richard Le


The data set consisted of monthly corporate credit ratings from 1986-09-01 to 2018-09-01 for 1899 firms in Korean indices such as the KOSDAQ and KOSPI. Firms in this data set can take any rating from the following set of 22 credit ratings, i.e., {AAA, AA+, AA, AA-, A+, A, A-, BBB+, BBB, BBB-, BB+, BB, BB-, B+, B, B-, CCC+, CCC, CCC-, CC, C, D}. The firms that take the "D" rating are considered to be in default. Some firms were "closed" after some time and are considered to be in default. Firms that were missing credit rating sequences, made for sale, or was merged with another firm were removed from the data set. After pruning the data set, there are 1648 firms remaining in the data set.



Applied Mathematics, Risk Assessment, Computational Finance