The ever-growing variety and scope of economic, financial and market data available, including data from periods of significant financial stress over the last decade, holds great promise for applications of machine learning techniques to credit risk and econometric modeling. However there are a number of special challenges in these areas such as relatively low frequency data (quarterly or annual), the need to model the probability of rare events such as investment grade defaults, and requirements for expert judgment in model calibration. In this talk we consider a number of applications developed at S&P of machine learning methods including probability of default modeling, credit risk ranking, issuance forecasting, and automated review of regulatory filings. We consider the performance of these models and various challenges arising from data limitations as well as other factors impacting model training and performance measurement. We also consider advantages and disadvantages of different approaches in the context of the examples.
Copyright. All Rights Reserved.