Banking: The new era of alternative Credit Scoring

Published by FirstAlign

Featured image by Dylan Gillis on Unsplash

Would you believe if I told you that the number of selfies in your cellphone photo gallery could impact you getting a loan? Well, welcome to the new era of Alternative Credit Scoring system!

Lending is a massive industry in the United States. It impacts the economy in many different ways. The loans granted by US commercial banks in August 2020 alone amounts to approximately 14.89 US$ trillion dollars. Any technology that makes even a small improvement in its process would be worth a significant amount of money.

Image by InspiredImages from Pixabay

A loan’s value partly depends on the creditworthiness of the individual or business. The more data one has about the borrower, the better it is. This big data problem makes Artificial Intelligence (AI) perfectly suited. Machine Learning can analyze different data sources such as collateral, predictions of growth, etc. to make better-informed decisions.

Determining creditworthiness without credit

Loan values are based on an individual’s or business’s ability to pay back. To determine the payback ability of a person can be difficult.  In the past, banks looked at FICO scores and income, but now companies have started to go beyond that. They can now access your vast digital data to gather more information.  AI can be beneficial in determining a person’s creditworthiness when he has no traditional credit history.

How AI works?

Two types of AI are used: Supervised and Unsupervised. Supervised AI creates specific rules and sorts data based on rules. The lender’s rule could be collateral value, borrower experience, credit score, etc.

AI can work with thousands of data points and applications, sorting by profitability and default risk in no time. Banks mainly use Supervised AI; for those in hard money lending, Unsupervised AI is more relevant.  In Unsupervised AI, we do not create rules initially but feed massive data sets, letting the AI identify patterns.

Hard lending refers to loans sanctioned based on assets, which means the person seeking a loan has an asset as collateral. Commercial lenders usually deal with hard lending; they do not have any stringent criteria but lend money based on guarantee. On the other hand, soft lending is mainly done by banks. They check the credit history before sanctioning loans; no collateral is needed. 

Real-time applications

Lendo, a startup company, uses Machine Learning to analyze a vast quantity of alternate data to predict a person’s creditworthiness. It looks at the applicant’s entire digital footprint, such as social media usage, internet browsing, geolocation, etc. Lendo claims that their Machine Learning model system has helped approve up to 50 percent more applications.

Another startup called Upstart also uses AI to determine creditworthiness and to automate the loan process fully. They started by focusing on young adults who lacked credit history. Upstart’s Machine Learning model considers education, GPA, field of study, and job history for predicting creditworthiness.

Upstart has also automated the process of loan approvals. As of April 2019, it has operated $5 billion in platform originations, of which 67% are fully automated through the Machine Learning process. One can only expect that Machine Learning Algorithm would improve over time as we move towards automating 99% of the approvals.

Future Challenges

As more and more AI-driven lending models gain visibility, regulatory factors are also likely to grow.  Machine Learning to analyze alternate data can raise issues of privacy, ethical and legal concerns. One particular challenge would be the elimination of bias in the loan approval process.  For example, an algorithm may use discriminatory data, such as zip code, that is positively related to ethnicity.

Conclusion

Even with these concerns, Alternative Credit Scoring systems using Artificial Intelligence are growing significantly. Its benefits are many.

  • Faster loan approval
  • Create profile for first-time loan applicants
  • Ensure business growth by providing more loans to people.
  • Improve data and risk management.

With the economy slowing down due to the impact of Covid-19, many small companies and individuals can find it challenging to get loans sanctioned. However, Alternative Credit Scoring can provide a ray of hope and the necessary boost that we all so much deserve in these times of uncertainty.

References

Published by FirstAlign

Click here to connect with us

Spread the word

Related posts