CREDIT RISK DECISIONING IN THE AGE OF DIGITALISATION
The level of digitalisation varies among customer segments and credit products of financial services. While it is more matured in consumer credit risk decisioning, there are still opportunities in small and medium enterprise (SME) and corporate credit risk.
In consumer lending, in the last five to 10 years, with the impact of FinTech, neo-banks and challenger banks, which leveraged new technologies and advanced modelling approaches, customer journeys have been transformed. These companies mainly use artificial intelligence (AI), deep learning and machine learning (ML) algorithms to improve the predictive power of credit risk scoring models, such as application scoring or collections role models. For example, a Swedish buy now, pay later (BNPL) company is using ML models to predict customers’ payment behaviour.
Some FinTech lenders use these algorithms to process alternative data, and then use it to enrich credit risk strategies, which is used to determine consumers’ credit eligibility and affordability. A US-based FinTech is using ML to process information on ‘thin file borrowers’, i.e., customers in its credit bureau with less than six months of payment history. The company started by targeting student loans and then grew with other customer segments. Another example is from the largest e-commerce company in the US. Its small business lending is using ML models very successfully. This company has a huge amount of proprietary information on what products are sold on its website, how customers feel about those products and the financial status of the companies which make those products. So, it is using this information in ML models to target companies for small business loans. In 2021, it had lent out roughly $1bn to small businesses with this strategy.
Oct-Dec 2023 Issue
SAS Institute Inc.