Artificial Intelligence in Commercial Lending – Gimmick or Gamechanger?

Have you ever had to choose between two equally good options? And while in that predicament, have you been fortunate enough to be presented with a third option: a combination of both? Although that rarely happens for us humans, in the ocean, there is a common type of symbiosis between the sea anemone and boxer crab: the crab feeds the anemone while benefiting from the anemone’s tentacles for protection. 

According to Bryan Johnson, CEO of Kernel, “the relationship between human intelligence and artificial intelligence will necessarily be one of symbiosis. The challenge and potential of exploring this co-evolutionary future is the biggest story of the next century.”

The material pouring out of the various media channels make it clear that AI is deemed a gamechanger. But how can we go about separating the hype from the reality of AI? In areas such as commercial lending, which is rooted in traditional client relationship management, AI holds great transformational potential.

The AI opportunity is through digitizing the commercial lending process. Taking on the digital challenge, it introduces an increased and broader variety of data-points, alongside innovative technology. These changes offer firms the potential to fast track their decision-making to give their commercial clients quicker and potentially more accurate answers when analyzing their credit needs. In fact, many businesses, particularly small businesses, are increasingly expecting to engage via digital channels because they are easy to use and response times are nearly instant.

Latest Advances in Artificial Intelligence

Having said that, the opportunity to utilize AI must be carefully matched to the business-led use case and the ROI needs to be well assessed. Current promising advances in AI applications impacting commercial lending include sophisticated machine learning algorithms to evaluate exposures based on latest data and provide confidence in the data quality entering the decision process, for example, models that approximate income or turnover. Other advances such as cognitive computing including Natural Language Processing (NLP) and OCR have been applied to read financial statements, so that financial ratios and information used for decision-making are up to date. 

Artificial Intelligence or Augmented Intelligence

So, are humans to be totally replaced in this brave new world of AI in commercial lending? Well no, it does not make sense to automate fully every single step in a complex commercial lending decision process. Human intervention will still be needed to set objectives, simplify and manage processes, and adjust inputs based on their own experience and insights. This is necessary for risk functions to keep control of the decision-making process.

What’s Next?

What happens next will depend on how the introduction of AI develops in terms of speed of adoption; education and lessons learned will be fed back into the development of these new approaches. For example, markets and regulators demand the inclusion of highly granular scenario analysis to help assess the impact of scenarios—a trade war, climate change, or other sources of stress—during the onboarding process. 

When introduced through well-defined use cases, AI has the potential to mature as a useful tool to help humans achieve greater productivity and higher efficiency in commercial lending. 

By Terisa Roberts

SAS is a Gold Sponsor and Exhibitor this year at RMA’s Annual Risk Management Conference, October 27–29, and will be presenting this topic at one of our conference lunch innovation sessions. To hear more from Terisa on this topic, you can register for the conference here.

Washington, The Week Ahead - November 18-22, 2019

Read More

Focus on RMA: John Baier, Relationship Manager

Read More

Model Validation and Governance Framework

Read More

comments powered by Disqus