In this study, numerous statistical models like multiple discriminant analysis, ordinal logit and probit regression analysis, cluster analysis, neural network analysis, and decision tree were used to predict the credit ratings of US financial corporations. It was discovered that principal components analysis and factor analysis provided additional power for developing the prediction models utilizing the above analysis. Because of the data limitations, only multiple discriminant analysis and ordinal logit and probit regression analysis were found to have distinguishing power in predicting credit ratings. Of these models, logit regression analysis performed the best, and can correctly predict 70% of corporations…
Contents: Credit risk analysis for financial corporations
1. Introduction
1.1. Background
1.2. Study Objective
1.3. Research Motivation
1.4. Thesis Organization
2. Literature Review
2.1. Statistical Model
2.2. Mathematical Programming Model
2.3. KMV Model
2.4. Recovery Rate
2.5. Simulation
3. Data Description
3.1. Data Selection
3.2. Variable Selection
3.3. Univariate Test
4. Methodology
4.1. Statistical Model
4.1.1 Data Reduction Technique
4.1.1.1 Principal Components Analysis (PCA)
4.1.1.2 Factor Analysis (FA)
4.1.1.3 Comparison between PCA and FA
4.1.2 Predictive Model
4.1.2.1 Multiple Discriminant Analysis (MDA)
4.1.2.2 Ordinal Logit and Probit Regression Analysis
4.1.2.3 Cluster Analysis (CA)
4.1.2.4 Neural Network Analysis (NNA)
4.1.2.5 Decision Tree
4.1.2.6 Model Comparison
4.2. Mathematical Programming Model
4.3. KMV Model
4.4. Evaluation of KMV Model
4.4.1 Power Curve
4.4.2 Intra-Cohort Analysis
4.4.3 Confidence Interval
4.5. Recovery Rate
4.5.1 Long-term Average Recovery Rate
4.5.2 Correlation between Default Rate and Recovery Rate
4.5.3 Relationship between Probability of Default and Recovery Rate
4.6. Simulation
4.6.1 Non-Parametric Bootstrap
4.6.2 Parametric Bootstrap
4.6.3 Empirical Simulation
5. Empirical Analyses for US Financial Corporations
5.1. Statistical Model
5.1.1 Data Reduction Techniques
5.1.1.1 Principal Components Analysis (PCA)
5.1.1.2 Factor Analysis (FA)
5.1.1.3 Comparison between Two Data Reduction Techniques
5.1.2 Predictive Model
5.1.2.1 Multiple Discriminant Analysis (MDA)
5.1.2.2 Ordinal Logit and Probit Regression Analys is
5.1.2.3 Cluster Analysis (CA)
5.1.2.4 Neural Network Analysis (NNA)
5.1.2.5 Decision Tree
5.2. Mathematical Programming Model
5.3. Comparison between Different Credit Rating Models
5.4. KMV Model
5.4.1 Comparison between Developed Model and S&P Rating
5.4.2 Comparison between Empirical EDF and Theoretical EDF
5.4.3 Evaluation of KMV Model
5.4.3.1 Power Curve
5.4.3.2 Intra-cohort Analysis
5.4.4 Limitations of Developed KMV Models
5.5. Recovery Rate
5.5.1 Long-term Average Recovery Rate
5.5.2 Correlation between Default Rate and Recovery Rate
5.5.3 Relationship between Probability of Default and Recovery Rate
6. Simulation
6.1. Non-Parametric Bootstrap
6.1.1 Credit Rating Models
6.1.2 KMV Models
6.1.3 Recovery Rate
6.2. Parametric Bootstrap
6.2.1 Credit Rating Models
6.2.2 KMV Models
6.2.3 Recovery Rate
6.3. Empirical Simulation
6.3.1 Credit Rating Models
6.3.2 KMV Models
6.3.3 Recovery Rate
7. Application of Developed Models
7.1. Application to Asia Rated Corporations
7.1.1 Data Description
7.1.2 Credit Rating Models
7.1.3 KMV Models
7.1.4 Recovery Rate…
Source: City University of Hong Kong
Download URL 2: Visit Now