Company Default prediction - DLMM Internal Rating Model in R
- Steps followed to implement the DLMM Model in R language
- Step 1 – Converting SPSS formatted data
- Step 2 - One by one empirical analysis of variables
- Step 3 - Cross-tabulation 01STATUS versus Industry Sector Code
- Step 4 - Exploring graphically the probability distribution of a variable
- Step 5 - Testing the normality of the probability distribution of a variable
- Step 6 - Evaluating the good/bad discriminant power of a variable
- Step 7 - Empirical monotonicity of ROE relative to good-bad progression
- Step 8 - Correlation between variable couples
- Step 9 - Analysis of outliers
- Step 10 - Data encoding
- Step 11 - Synoptic table of variable properties
- Step 12 - Linear Discriminant Analysis - Initial approach
- Step 13 - Experimenting with Stepwise Linear Discriminant Analysis
- Step 14 - Gaussian Copula encoding scheme
Steps followed to implement the DLMM Model in R language
Purpose
Step by step implementation in R language of the internal rating models approach presented in: De Laurentis G., Maino R. and Molteni L.,, Developing, Validating and Using Internal Ratings: Methodologies and Case Studies, 2010, John Wiley & Sons, Ltd DOI:10.1002/9780470971901 -> (https://onlinelibrary.wiley.com/doi/book/10.1002/9780470971901)
Method
This implementation follows step by step the content of Chap. 4: Developing a statistical based rating system, pp. 93-236
Step 1 – Converting SPSS formatted data -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step1)
Step 2 – One by one empirical analysis of variables -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step2)
Step 3 – Cross-tabulation 01STATUS versus Industry Sector Code -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step3)
Step 4 – Exploring graphically the probability distribution of a variable -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step4)
Step 5 – Testing the normality of the probability distribution of a variable -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step5)
Step 6 – Evaluating the good/bad discriminant power of a variable -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step6)
Step 7 – Empirical monotonicity of ROE relative to good-bad progression -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step7)
Step 8 – Correlation between variable couples -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step8)
Step 9 – Analysis of outliers -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step9)
Step 10 – Data encoding -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step10)
Step 11 – Synoptic table of variable properties -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step11)
Step 12 – Linear Discriminant Analysis – Initial approach -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step12)
Step 13 – Experimenting with Stepwise Linear Discriminant Analysis -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step13)
Step 14 – Gaussian Copula encoding scheme -> (https://github.com/MoiraCorp/DLMM-IRating-in-R/tree/main/steps/step14)