by Christophe Mues, Cristián Bravo, Kamesh Korangi
Continue on SelfMadeQuant
Suggested foundations for credit and counterparty risk and financial machine learning. Specialist prerequisites vary by paper.
Separate default probability, loss severity and exposure.
Read conditional default and exposure quantities.
Preserve time order and avoid future information in training.
Establish an interpretable baseline before a flexible model.
Study dependence between defaults.
Connect exposure and counterparty losses to valuation adjustments.
Account for dependence and overlapping observations.
Choose a reproducible strategy, specify the trading rule, and test what remains after costs and estimation uncertainty.