HKMA published the second issue of the Regtech Watch newsletter. This issue focuses on applications of technology in credit risk management. It highlights the key challenges in credit underwriting and risk management processes and shares use cases of banks that apply machine learning models and other technologies to overcome these challenges. Three use cases are summarized covering the use of data-driven income estimation to approve personal loans, machine learning-based methods to streamline corporate loan underwriting, and linguistic and sentiment analysis to identify adverse news about the borrower.
In the course of understanding and examining credit underwriting and risk management processes of banks, HKMA has observed that some banks have adopted new technologies to overcome certain longstanding challenges. The application of technology in credit risk management is gaining popularity across banks in Hong Kong. The newsletter showcase the following use cases:
- Data-driven income estimation to approve personal loans. To enable banks to offer a smoother customer experience, HKMA introduced new guidelines in May 2018 to allow banks to adopt innovative technology to manage credit risks related to personal lending business. Several banks have since adopted data-driven income estimation models to inform credit underwriting and lending decisions. By applying these models to readily available information, banks can take income input into account in their credit underwriting decisions very quickly, resulting in a significantly faster application process. Banks using these models are cognizant of the need to re-train and re-validate the models routinely to ensure their continued robustness.
- Machine learning-based methods to streamline corporate loan underwriting. To improve efficiency of corporate lending decisions, some banks are exploring the use of artificial intelligence and machine learning in credit assessment. One bank in Hong Kong recently launched a pilot roll-out of artificial intelligence engines for financial spreading, including the use of Optical Character Recognition to digitize physical copies of financial statements. Another Hong Kong lender supplemented traditional financial ratios with analyses of more advanced indicators about corporate borrowers, for example, by examining the transaction and cash-flow patterns reflected in bank statements. The indicators will then be used to estimate the probability of default, which ultimately informs loan decisions.
- Linguistic and sentiment analysis to identify adverse news about the borrower. To identify adverse news to ensure that any major adverse developments related to corporate borrowers are identified in a timely manner, banks need to review a large volume of news from different sources on an ongoing basis. To this end, some banks have started to use tools such as Natural Language Processing (NLP) and supervised machine learning to automate news screening. This involves the use of linguistic analysis tools to extract relevant information from the news sources, classify the information into different subjects, and digitize it in a machine-readable format. The technology helps banks achieve more timely and accurate identification of adverse news on borrowers, which in turn improves their ongoing assessment of the borrowers’ creditworthiness.
These use cases also demonstrate the potential limitations of new technology. In machine learning credit models, the probability-of-default predictions obtained via different approaches may vary significantly and are not easy to interpret. In addition, the models currently in use were developed only in recent years, so their performance across a complete credit cycle is yet to be seen. Additionally, banks often depend on third-party developers for expertise in designing and developing technology applications, giving rise to potential risks in governance and accountability. To overcome these limitations, HKMA expects banks to implement programs to recruit, train, and retain employees with suitable skill sets and establish effective mechanisms to supervise the relevant staff members.
Apart from these use cases, HKMA noticed that banks overseas are now applying emerging technology solutions to credit risk management, particularly with regard to small-business loans. Given the limited credit history of small businesses, the industry has developed credit-scoring models based on alternative data—"firmographic information" such as borrowers’ firm size and location and social media presence—to facilitate loan approval. Such alternative data have helped banks to arrive at more accurate assessments of the creditworthiness of small businesses,which might otherwise be opaque. The benefits of credit scoring with alternative data are being increasingly acknowledged internationally, for example, by BIS and FSB.
Keywords: Asia Pacific, Hong Kong, Banking, Regtech, Newsletter, Machine Learning, Credit Risk, Artificial Intelligence, Regtech Watch, Natural Language Processing, Fintech, HKMA
EBA issued a revised list of validation rules with respect to the implementing technical standards on supervisory reporting.
EBA published its response to the call for advice of EC on ways to strengthen the EU legal framework on anti-money laundering and countering the financing of terrorism (AML/CFT).
NGFS published a paper on the overview of environmental risk analysis by financial institutions and an occasional paper on the case studies on environmental risk analysis methodologies.
MAS published the guidelines on individual accountability and conduct at financial institutions.
APRA published final versions of the prudential standard APS 220 on credit quality and the reporting standard ARS 923.2 on repayment deferrals.
SRB published two articles, with one article discussing the framework in place to safeguard financial stability amid crisis and the other article outlining the path to a harmonized and predictable liquidation regime.
FSB hosted a virtual workshop as part of the consultation process for its evaluation of the too-big-to-fail reforms.
ECB updated the list of supervised entities in EU, with the number of significant supervised entities being 115.
OSFI published the key findings of a study on third-party risk management.
FSB is extending the implementation timeline, by one year, for the minimum haircut standards for non-centrally cleared securities financing transactions or SFTs.