Key to every lending transaction, and the decision whether or not to extend credit, is an analysis of the likelihood of repayment. Artificial intelligence is increasingly being deployed in this analysis. Indeed, a number of fintech start-ups are built around this technology, which, through the efficiencies it can bring, has the potential to lower the high barriers to entry that conventionally pervade in the banking and wider financial sector.
As this FT Alphaville article points out, the types and sources of data being processed by AI to measure credit risk are often considered 'non-traditional' by past standards. The use of 'big data' is how these alternative lenders hope to gain a competitive advantage.
The theory that this will bring forward creditworthy borrowers ignored by banks and other 'traditional' lenders is compelling, for quite a number of reasons. However, AI-based tools for assessing credit risk have not yet been tested through the cycle. Whether they are effective in identifying patterns that strengthen rather than weaken lending standards when applied to prospective borrowers has obvious implications for the firms involved and from a macro-prudential regulatory perspective.
We have recently conducted a research project into the risks associated with the development, use and application of AI and other disruptive technologies in various fields, including lending. If you would like access to the research, please register at https://www.macfarlanes.com/what-we-think/artificial-intelligence/.
If banks exclude certain kinds of borrowers, this creates an opportunity for alternative lenders. This is the idea behind a host of new platforms that have emerged since the crisis.