The requirement to predict future lifetime credit losses has led many institutions to use statistical tools to understand what drives loss experience and examine how they can estimate future losses. Two of the most common tools are correlation and regression. Join us for an insightful webinar as we cover the basics of correlation and regression, demonstrate how they're being used in CECL, and provide an overview of advanced analysis tools like machine learning.
Upon completion of this program, participants will be able to:
- Explain correlation and how it can be used to describe relationship between loss and potential predictive variables
- Describe how linear and logistic regression are being used in CECL
- Discuss the various forms of model error risk and the importance of testing and validation
- Describe the role of machine learning in credit loss prediction