This code was based on a term long project for Fintech Analytics with Professor Roger M. Stein. The model estimates the probability of default for 16,509 unique US banks between 1993 and 2016. You can view the presentation for results and summaries.
- Winterization transformation of key variables
- Logistic Regression Modeling
- Bayesian Priors Calibration
- Bootstrap Testing
- Walk-forward Testing
- R Studio
- GGplot2: Great tool for data visualization and clean charts
- dplyr: For dealing with data
- lubridate: Dealing with dates and stepping forward or back by quarters
I will need to update with instructions on how to fully use all of the functions but they are well defined in the beginning of the code.
Don't hesitate to reach out. You can find my twitter here for any suggestions! You can also reach out via linkedin
Dataset will be uploaded shortly.