Underwriting , bad loans - The role of AI

The Other day I decided to pick up a short term auto loan from a local Bank.To my amazement the bank requested a lot of documentation. Followed with the wrath of documentation followed a visit by the loan executive at my residence to establish my identity and authenticity. Then came a couple of queries and having resolved all of them , I ended up as a happy customer. In the news I have been following the list of defaulters to the likes to Modi's and Mallya who seemingly have managed to walk out of the country defrauding billions of dollars of public money. This really got me thinking if I as common man has wrath of documentation, intense scrutiny and 2-3 weeks of wait time, How do these A list of world class citizen manage to dodge the system.

Couple of questions started to pop up

A. What is the nature of the problem we are dealing with?

B. What role do people and processes play?

C. Can AI really help in some form here?

D. Do the government really care for public money?

The list is endless. 

Did not want to play the protagonist to this idea , I decided to find ways to solve the problem

A. What's the current state of the problem?

Banks and financial institutions, with some notable exceptions, are struggling with bad loans. According to India Ratings the average Impaired Asset Ratio - the sum of gross NPAs and restructured advances (a measure of the stress on a lender's balance sheet) stands at 12% of advances and is slated to rise to 12.5%. Notwithstanding the widespread notion that it is only corporate loans that have turned sour, retail loans, with an IAR of 8.8% and small & micro enterprise loans, with an IAR of 8.8% and small & micro enterprise loans, with an IAR of 10.5% have contributed significantly to the asset quality woes of the banking system. 

B. Do banks have enough data to decide a potential bad defaulters?

Over the years bank have managed to gather a lot of data including government bodies. However the part of the problem is a comprehensive AI solution which involves people and processes combined. 

C. Conventional Methods ....

Underwriting today for most banks is solely managed by set of decision makers whose sole into in terms of data is credit scores and nothing else. Technically one could still hack there way with credit score According to RBI, the credit outstanding in the banking system grew from Rs. 27,70,012 cr in 2009 to Rs. 64,99,829 cr in 2015 - a whopping 135% increase in just 6 years. 

D. Can AI really help?

AI can help in to start with uncovering a potential defaulters by studying the customer data. This could start of with have a better decision making capability. That's just solving one part of the problem. There is more to story, the problem with high net worth defaulters usually find ways to hack the system or bribe there way to get the loans, AI can help identify potential employee who work against the interest of organization. AI can bring in new life to compliance and governance.