Artificial intelligence (AI) is being used to create more accurate and transparent credit decisions than human-based systems and may well usurp them within the next five years, according to a report by Finextra.
Credit risk is one of the biggest challenges that face the banking system. The McKinsey Global Institute estimate that AI and its applications like machine learning could generate more than $250 billion in the banking industry. The solutions go from deciding how much a bank should lend a customer, providing signals to traders about risk, detecting fraud and improving compliance.
In this post, we will look at some of the solutions and case studies surrounding AI, credit and risk management.
Why do banks need AI for credit decisions?
The banking industry has a lot of challenges from credit to operational risk and legacy systems that are very susceptible to fraud and cybercrime. As fintech competition continues to grow, the banks face falling long-term rates and low profitability. With all these challenges, banks are looking at new technology and AI.
As AI can harness large amounts of data, something banks have in abundance, it has become key to their future success. Wells Fargo is one institution banking on innovation and in 2018, head of their innovation group Steve Ellis commented on how important AI is for analyzing behaviors and patterns in credit risk.
However, many banks are relying on traditional credit scoring practices that might consider 20 or so factors, forgetting the huge digital footprint that consumers are leaving. This could mean they are leaving creditworthy customers behind.
A McKinsey report showed that several banks are starting to rely more on machine learning techniques in their credit processes, replacing the aforementioned antiquated methods.
American BankMobile is using AI for better analysis of thin-file Millenials. They are using software developed by Upstart that relies on alternative datasets to help make better decisions. To apply for BankMobile loans, applicants share their highest level of education obtained, the name of their school or university, their graduation year and their area of study. That information is what Upstart’s underwriting software uses to make its credit decision.
The Canadian bank has partnered with world-renowned AI company Layer 6 to develop solutions that offer better scoring and cybersecurity. This investment in smart, self-learning technology helps the bank build systems that can more accurately predict customer needs. For example, it becomes aware of potential life events and predicts how that might impact customer risk and creditworthiness. The objective is an award-winning customer service.
Machine learning models are highly intuitive when it comes to creating algorithms for risk profiling. Algorithms like gradient boosting, random forests and decision trees can find the hidden dependencies in a dataset to gain more accurate predictions. Banks, in turn, can determine how the collected parameters should be weighted to predict whether borrowers will repay their loans back on time.
According to Deltec Bank, Bahamas- “Data signals will define the parameters that affect the power of a scoring model.” For example, different business types might have variable weightings for geography and types of audiences. Machine learning tells banks which data points include the desired signals.
The GiniMachine platform combines advanced machine learning techniques with loan portfolio histories to uncover the full potential of data. Unique scoring models can be built in minutes with an accurate way of assessing risk.
The models use data outside of standard data points and have proven accuracy. This includes social activity, education, occupation, industry and family attributes for the applicant. External factors such as the quality of the applicant’s connections are also included for credit analysis.
To use the platform, it only requires 1,000 previous loans to be imported for the models to start training. From that, everything else can be inferred with new credit risk scoring processes being deployed within only 2 hours. The models can learn from new data and ever-changing compliance or regulatory procedures.
Research by administrative company Intertrust showed that of 500 executives, 14% believe AI has surpassed human decision-making skills when it comes to credit. In fact, Intertrust says that AI systems can overtake human counterparts by as soon as 2024.
The reason for the increased use of AI is where credit checks are moving towards less traditional criteria for assessing risk. Some examples being used are social media activity, retail spending patterns, and even political inclinations.
However, with regulations like General Data Protection Regulation (GDPR) in the EU, there have been calls that using personal data e.g. social media, does push the boundaries of privacy. Even before GDPR, in 2016 Admiral were one of the first firms to try using social media to price car insurance for their customers.
Admiral came under a lot of scrutinies for their campaign as it was considered intrusive. In this instance, asking for consent was not obligatory for Admiral, as would be the case with GDPR today.
As consumers start to regain trust with firms using data appropriately, the benefits of having a credit process that using more factors to assess risk will become apparent. For example, customers will get fairer premiums based on them as an individual rather than looking at only surface-level information or financial history alone.
When it comes to credit scoring, traditional data sources like bureaus are still important as a part of the process. They do give some insight into the creditworthiness of potential borrowers and are important for minimizing the risk of lending. However, on their own, these sources do not cover other noteworthy data points which is where AI is starting to come into its own.
Social media, retail and Google Analytics are a few sources being used by banks to provide better insight into the digital footprint of their customers. Using AI and machine learning in this way is how banks can remain competitive at a time where new fintech are an ever-growing threat.
Disclaimer: The author of this text, Robin Trehan, has an Undergraduate degree in economics, Masters in international business and finance and MBA in electronic business. Trehan is Senior VP at Deltec International www.deltecbank.com. The views, thoughts, and opinions expressed in this text are solely the views of the author, and not necessarily reflecting the views of Deltec International Group, its subsidiaries and/or employees.
About Deltec Bank
Headquartered in The Bahamas, Deltec is an independent financial services group that delivers bespoke solutions to meet clients’ unique needs. The Deltec group of companies includes Deltec Bank & Trust Limited, Deltec Fund Services Limited, and Deltec Investment Advisers Limited, Deltec Securities Ltd. and Long Cay Captive Management.