ICICI Bank Is Betting Big On Artificial Intelligence, Machine Learning – ELE Times

ICICI Bank is adopting Aritificial Intelligence (AI) and Machine Learning (ML) in an enormous method. Over the final yr, these applied sciences have decreased inner prices for the financial institution and helped considerably drive its deposit and lending progress, the banks administration mentioned throughout an earnings name with analysts. The financial institution reported a internet revenue of Rs 4,402 crore within the fourth quarter of FY21, round 260.5% greater in comparison with the identical interval within the earlier yr, in keeping with its monetary outcomes.

The prices and advantages of AI and ML

Several of the main banks are turning to know-how, to focus on prospects with the appropriate product and thru data-based underwriting, they will convey their mortgage supply time to just some minutes. ICICI Bank sees AI and ML as a possibility to accumulate good prospects at a sooner pace, whereas additionally serving to it preserve tabs on how their purchasers’ spending and credit score behaviour on a close to real-time foundation.

Banks can have a strong credit score supply course of and underwriting skills for unsecured merchandise which might be delivered in a pre-approved and straight-through format, mentioned Anup Bagchi, Executive Director, ICICI Bank. “The AI and ML essentially are used to a) figure out patterns and b) most importantly, get feedback of early warning signals back into the credit underwriting process. We can intervene in the micro-services market in terms of yields and fees. These technologies lead to a lower operating cost, lower acquisition cost, lower cost of collection and lower credit cost,” he mentioned.

However, not all merchandise will be scaled utilizing AI and ML instruments that automate the underwriting course of. “AI and ML is not as amenable to the business loans and SME segment since the number of cases are small. However, it provides early warning signals on sectors, which are then fed back into the score card model,” he mentioned.

For different secured lending segments, banks are utilizing AI and ML to develop score-cards that are robotically up to date with close to real-time data. “As the ticket sizes come down and as the portfolio follows the law of large numbers, score cards give a better output than human intervention in credit decisioning. You need to keep updating the score card with the inputs and learning from the portfolio,” Bagchi defined.

On UPI and information

The banks’ administration mentioned that the transformation in funds is a chance to derive wealthy information from transactions. “Payments standalone may or may not make you money. But the moment we get more of a digital footprint we are able to put it into underwriting. It is the behavioural data and velocity of data which is more important than static data,” he mentioned.

The financial institution is constructing an unlimited information lake to improve its buyer behaviour analytics. It will experiment with new makes use of that may enhance product penetration and buyer stickiness, it mentioned.

According to its investor presentation, the financial institution reported a dip in Person-2-Merchant transactions on the Unified Payments Interface (UPI) at Rs 34,100 crore in This autumn FY21 in comparison with Rs 36,250 crore in Q3 FY21. On an annual foundation, UPI P2M transactions has grown by 2.5x from Rs 13,715 crore in Q4FY20.  The banks’ marketshare in UPI P2M volumes stood at 14% in March 2021, it mentioned.

“We think that UPI will put pressure on debit cards because the proposition of UPI is almost as strong as the debit card proposition,” Bagchi mentioned. Between bank cards and others, we don’t suppose the cannibalisation by UPI will likely be that a lot, he added.

As fintechs and startups develop new types of distribution, cost options, and new merchandise like Buy Now Pay Later, the funds trade is about to bear a serious transformation within the coming years there may a number of new retail funds umbrella entities which is able to compete with the National Payments Corporation of India (NPCI). ICICI Bank has reportedly partnered with Axis Bank, Visa, Pine Labs and BillDesk to use for a New Umbrella Entity license.

“We have to keep betting on all payment systems and make it open-architecture and partner with other companies,” Bagchi mentioned.

Digital Banking Highlights

  • 60% of commerce transactions by way of Trade Online platform are digital
  • 87% of transactions on enterprise transactions by way of InstaBiz are digital
  • 56% of fastened deposits sourced digitally
  • 75% of bank cards sourced digitally
  • 90% of non-public loans sourced by way of Insta Banking on iMobile, the banks’ flagship funds and banking app
  • 90% of all financial savings account transactions are digital

Financial Highlights

  • Net Profit: Rs 4,400 crore in This autumn FY21, vs Rs 1,420 crore in This autumn FY20
  • Net Interest Income: Rs 10,430 crore in This autumn FY21 vs Rs 8,930 crore in This autumn FY20
  • Cumulative provisions: Rs 12,010 crore in This autumn FY21
  • Advances: Rs 7.34 lakh crore in FY21 vs Rs 6.45 lakh crore in FY20
  • Deposits: Rs 9.32 lakh crore in FY21 vs Rs 7.7 lakh crore in FY20
  • Gross NPA: 4.96% in FY21 vs 5.53% in FY20
  • Net NPA: 1.14% in FY21 vs 1.41% in FY20


Please enter your comment!
Please enter your name here