Artificial Intelligence In Lending – Is It A Magic Bullet?

Home / Mobile / Mobile Banking / Artificial Intelligence In Lending – Is It A Magic Bullet?

Artificial Intelligence In Lending – Is It A Magic Bullet?

robot android men
Reading Time: 5 minutes

Banks and other financial service companies are built on security and trust. They keep your money safe and you trust them to give it back when you ask for it. If you don’t trust your bank, you will take your money out. But if “too many” people do that then there will be a run on the bank. No bank has enough “cash” to cover all they “owe” to their customers. So, for the relationship to work, trust is vital.

 

Trust works in both directions. While some loans are “secured” and the bank can take possession of the goods if the loan isn’t repaid – they don’t want to do that – it is costly, time consuming and often generates bad publicity (especially when repossessing homes). So banks are, to some degree, forced to trust that their customers will repay. However, in reality the banks can neither expect all their customers to pay all of what they owe when they owe it, nor can they stop extending credit to more and more people.

 

So how do they reduce their risk (fear)? Banks use various data points such as public records, sources of income, financial associations, repayment capacity, financial habits and credit history to ascertain credit worthiness. However, as the world continues to move faster, trust is being eroded. Millennials don’t trust big companies, many people don’t trust governments and people increasingly don’t trust “news” (fake news). Fraudsters continue to adopt innovative ways to steal personal identities. Establishing and maintaining trust in financial relationships is becoming a more complex challenge to solve.

 

International Monetary Fund estimates that global debt reached USD 152 trillion in 2015 and that it was growing faster than world GDP. Two thirds of this is private sector debt from companies and households, so the debt cycle is continuing to roll at a much faster pace now. Simultaneously, the numbers of bad loans have also been growing and have reached a point where they could stifle growth or even lead to a prolonged period of stagnation. This trend is appearing in many parts of the world with India being the latest addition to the list. No one seems to have come up with a perfect solution, yet.

 

But is there an answer in sight?

As big data continues to gain traction in many industries and as the focus on effectively leveraging micro-level insights increases, the traditional approach in evaluating bank customers is being transformed. Customers’ desire to use digital channels has ensured that they leave behind growing digital footprints, which can then be easily mapped to identify meaningful patterns and provide the answers. This is one area where Artificial Intelligence (AI) can come to the rescue of the banks and financial service companies.

 

Artificial intelligence has moved beyond its early applications in automation, and has evolved into Machine Learning – a set of algorithms that have the capability to change in response to their own output, or computer programs that automatically improve with experience. These algorithms do not offer a single prediction but rather offer probabilistic outputs – a range of predictions with estimates of uncertainty. Deep Learning is an evolved form of Machine Learning where a computer system is made capable of taking its own decisions by exposing it to a huge amount of data. Deep Learning systems comprise of multiple layers of statistical operations between the input and output data, which have been defined by an algorithm, rather than a person. This makes the actual process of arriving at the output much more complex and rendering it beyond the comprehension of humans, thus not allowing humans to predict the outcome. The concept, used by Google in its voice and image recognition algorithms has found applications in many other aspects, some of which include navigation of self-driving cars using sensors and onboard analytics, predicting the outcome of legal, automated analysis and reporting, and precision medicine genetically tailored to an individual’s genome. What makes Deep Learning interesting is the potential it offers for autonomous decision-making.

 

By processing huge amount of data from an individual’s banking transactions, information from social accounts, earning and spending patterns, friends and family history and churning it through an AI system, it is possible to create a very comprehensive credit profile. In contrast, today’s lender has limited ability to access, consolidate this data and derive actionable analytics from it. With Deep Learning, machines can be taught to make credit decisions on their own after they are exposed to a number of data points including past decisions, credit policies, risk appetite, various rules, regulations, eligibility criteria and complex scenarios. Using their own intelligence and based on an individual’s AI driven credit profile, the intelligent machines can now make credit decisions at speed and accuracy which humans can hardly match. More over they are capable of learning continuously and can adapt themselves to any new situations that may arise based on their deductions. So AI has the capability to fill an existing gap. Similar approaches have helped a number of FinTech companies extend credit to a much wider customer base which was previously not considered creditworthy by the banks.

 

With Machine Learning, AI can definitely bring tremendous efficiency and cost reduction in loan processes. Which bank would not love to have an intelligent machine, which can take perfect loan decisions on its own, 24 hours a day, 7 days a week, in nano-seconds? Also, these machines would be far better at picking up the anomalies in applications, data points and behavioural patterns – leading to the early identification of all kinds of possible frauds. A new study by Juniper Research predicts that unsecured consumer loans that use Artificial Intelligence or Machine Learning technology will jump to 960% over the next five years (2016-2021) and potentially becoming a US$17 billion business for FinTechs by 2021.

 

So can we say that Artificial Intelligence – running on super fast computers, crunching millions of pieces of data in real time and that never gets tired – can never make a mistake? Does this mean that banks can now say goodbye to bad loans? Well… not really….

 

Can these machines really never make a mistake? Will these intelligent systems be able to nurture the two-way ‘trust’ that exists in banking today?  The machines are only as good as the data they are fed, and if that data is wrong, or out of date or incomplete? Imagine a situation where an individual is deemed not creditworthy by the machine based on certain specific actions which took place a long time ago, actions that a human would consider and then discount? On the other hand, someone may actually try to stage manage certain specific actions to boost his scores. Is there a possibility that the machine may inherit some biases based on the personal preferences of the individual/team that was involved in the learning exercise? Will there be transparency in the logic used for these decisions (very difficult in Deep Learning)? How will regulators ensure that organizations have safeguards against any kind of discrimination in the criteria used by these machines? What about the cyber security risks associated with AI? What happens when people feel this widespread usage of data invades into their private lives? How will a threshold be defined to avoid getting into this situation? What happens if regulators impose restrictions on private data usage in AI?

 

At this stage, while it appears that while AI has tremendous potential to solve the trust problem in banking and financial service industry, a lot of questions need to be considered before it gets mass adoption. Trust is a delicate issue and banks would like to trudge extremely carefully while dealing with it. Almost every breakthrough in technology goes through this cycle and this is no reason to cast doubt on the value that AI could unlock for both the banks and customer. While Artificial Intelligence may not be the panacea for now, it does look destined to become one in the future. And that future is not as far away as people often think.

 

Share this Post

About the Author

Nitin Garg, Nitin Garg - Senior Manager, Marketing, Nucleus Software

Nitin Garg is a Senior Manager, marketing at Nucleus Software where he is responsible for managing brand promotion and product marketing initiatives globally. He has more than 14 years of experience in driving marketing, business development, service delivery and technical operations across diverse industries.

Leave a comment

Your email address will not be published. Required fields are marked *

About Nucleus

At Nucleus Software we are committed to providing efficient, modern yet proven software solutions for the global Banking and Financial Service industry. We have been pioneers in developing Retail Banking Software, Corporate Banking Solutions, Transaction Banking, Cash Management and Internet Banking Software since 1986. Our success spreads across more than 50 countries, and we serve our customers globally through our direct and partner operations across US, Europe, Asia-Pacific, Africa and the Middle East. We are known for our world-class expertise and innovation in lending and transaction banking technology. Our two flagship products, built on the latest technology are: FinnOne™ and FinnAxia™.