As credit card usage grows, the risk of default has also increased. With focus on new to credit (NtC) and new to bank (NtB) customers to drive adoption, card portfolios are witnessing historically high delinquency & chargeoff levels. Probability of Default (PD) is a statistical metric to assess the likely loss that a portfolio of cards will witness.
Probability of Default (PD) is calculated by taking into account a number of factors, including the borrower's credit score, debt-to-income ratio, and payment history. Approaches to computing Probability of Default include :
- Statistical models are based on historical data and are used to identify factors that are associated with default.
- Machine learning algorithms are more sophisticated and can learn to identify patterns in data that are not easily identifiable by humans.
- Scorecards are a type of statistical model that is used to assign a risk score to each borrower.
New-age systems like Drona Pay operate in real time and leverage datasets related to demographic data, credit history, repayment status and transaction history to predict probability of default. Drona Pay can integrate this into real time decisioning especially as new customers, customers approaching higher levels of credit utilization or customers having enhanced limits from other products or overlimit approvals.
In summary, credit card growth has necessitated the need for robust risk management. Employing early warning systems and predictive techniques based on ML models can help credit card companies navigate these challenges effectively.
Comments