Credit Risk Modeling Repository consisting of adhoc analysis and data science notebooks
This is part one of the Credit Risk Modeling Project.
- This repository consists of analysis of mortgage loan applications and their performance.
- This project uses Freddie Mac Singly Family Loan-Level Dataset
- Freddie Mac provides data on each loan for every quarter since 1999 Q1
- A single quarter dataset has 2 components
- Origination Data:
- Every row is an approved loan consisting of details at the time of application
- Performance Data
- This data consists of monthly performance of every loan approved in that quarter
- Note: performance data in 1999 Q1 only signifies the approval of that loan in 1999 Q1, it can consist of monthly performance in 2022
- Origination Data:
- This repo consist of data science and research efforts to design a system that is free of assumptions a biases capable of producing a model that predicts the most likely status of a loan 3 or 6 months in the future.
- Engineering codebase for application of this data science system would be another repository.
- This repo is for reports and analysis to explain features of the data.
This matrix explains how one loan_delinquency_status changes to another in every next loan age
- 0-29 days of delinquency means the monthly loan re-payment is up-to-date
- RA - REO Acquisition is when a borrower defaults and the bank re-posseses the property for sale. This is considered a highly delinquent status
- 30-59 days of delinquency is when a borrower has missed one monthly payment. This is considered a low delinquent status.
- Similary, 60-89, 90-119 so and so forth up until 360 days of delinquency is considered, further delinquent statuses are considered default. Higher delinquent statuses rarely recover (low cure rate) and hence this analysis treats all statuses 360 days and higher the same
├── LICENSE
├── Makefile <- Makefile with commands like `make data` or `make train`
├── README.md <- The top-level README for developers using this project.
├── data
│ ├── external <- Data from third party sources.
│ ├── interim <- Intermediate data that has been transformed.
│ ├── processed <- The final, canonical data sets for modeling.
│ └── raw <- The original, immutable data dump.
│
├── docs <- A default Sphinx project; see sphinx-doc.org for details
│
├── models <- Trained and serialized models, model predictions, or model summaries
│
├── notebooks <- Jupyter notebooks. Naming convention is a number (for ordering),
│ the creator's initials, and a short `-` delimited description, e.g.
│ `1.0-jqp-initial-data-exploration`.
│
├── references <- Data dictionaries, manuals, and all other explanatory materials.
│
├── reports <- Generated analysis as HTML, PDF, LaTeX, etc.
│ └── figures <- Generated graphics and figures to be used in reporting
│
├── requirements.txt <- The requirements file for reproducing the analysis environment, e.g.
│ generated with `pip freeze > requirements.txt`
│
├── setup.py <- makes project pip installable (pip install -e .) so src can be imported
├── src <- Source code for use in this project.
│ ├── __init__.py <- Makes src a Python module
│ │
│ ├── data <- Scripts to download or generate data
│ │ └── make_dataset.py
│ │
│ ├── features <- Scripts to turn raw data into features for modeling
│ │ └── build_features.py
│ │
│ ├── models <- Scripts to train models and then use trained models to make
│ │ │ predictions
│ │ ├── predict_model.py
│ │ └── train_model.py
│ │
│ └── visualization <- Scripts to create exploratory and results oriented visualizations
│ └── visualize.py
│
└── tox.ini <- tox file with settings for running tox; see tox.readthedocs.io
Project based on the cookiecutter data science project template. #cookiecutterdatascience