India’s banks can do with smart tech solutions to address the challenge of bad loans

Picking up early signs of bad loans could be the key

by Dr. Tarun Sharma

Minister of State for Finance Anurag Thakur recently informed Rajya Sabha that the gross non-performing assets (NPA) of public sector banks had declined to Rs. 6.09 lakh crore in September 2020 from 8.96 lakh crore in March 2018. where the Minister highlighted the record recovery of Rs. 2.54 lakh crore during the last two financial years. However, it has been a challenging period for the public sector banks, which have had to write off bad loans worth Rs. 5.85 lakh crore during the three financial years beginning April 2018.

There have been some positive signs. 11 of the 12 public sector banks, the Minister said, had reported a profit during the quarter ended September 2020. But for the quarter ended December 2020, the loans written off by mid and large public sector banks have registered a sharp increase. That could be a worrying sign for the future.

Writing off the loans follows the guidelines prescribed by the Reserve Bank of India (RBI) and a policy approved by the banks’ boards. While the banks may have made every effort to recover the bad loans before writing them off, proactive measures driven by them could reverse the government-owned banks’ situation.

It is time for banks to consider extensive data and analytics applications to combat the scourge at every level. Since the problem needs urgent attention, widespread technology adoption could help. These solutions could be preventive, ongoing, and curative. If banks can focus on solutions that can help with early warning signals, it could directly impact its asset quality.

Tech solutions for NPA worries

Since the boards of various banks are actively involved in addressing NPA’s challenge, they should consider using technology options. These intelligent solutions could be of extensive assistance beginning at the stage of receiving a request for a loan and following through the entire lifecycle of the loan. Since a loan is not a one-off financial transaction, and the relationship extends over time, the inputs from smart technologies can be critical during the entire loan tenure.

Starting from borrower analysis, individuals and companies for a loan application need vital checks from the internal database and third-party sources like credit bureaus. The bank must integrate its database with multiple third-party data sources to make it comprehensive and reliable.

Building and nursing a comprehensive database that can secure inputs from government, corporate, personal, an ecosystem like e-commerce companies can be a potent deterrent against bad loans. Add the information furnished by the borrower to that database, and the quality of information can get even better. An integrated system that thrives on multiple databases could be a good beginning.

Manual analysis of risk could lead to unintended and preventable lapses. The banks’ data can create industry/segment level scorecards applicable at the appraisal lifecycle. It could aid in data-driven decision making.

The banks need to have a data-first approach to each application to provide inputs for analytics solutions, which can create, deploy and monitor various analytical models using artificial intelligence and machine learning models.

Technologies can also help in completely digitizing eligibility / servicing calculations without any manual intervention. Technology-driven processes can understand what borrowers can service, extend their credit accordingly, and ultimately help the financial institution. It cuts down their exposure to the customer, which can later put them into hardship. Doing this using digital processes allows for error-free, repeatable results. Today with multiple data sources, e.g., bank statement analysis, current liability status availability via bureau, GST reports, etc., it is achievable for almost all retail and MSME segment.

Banks have traditionally focused on asset-based lending. The process of asset liquidation is slow and time taking. In comparison, cash flow-based lending is a very effective way of understanding the actual dynamics of the borrower’s business. It involves looking at the cash flow statements, analyzing debtor-creditor reports, understanding sales, recovery trends, and multiple other similar steps. These analyses are complex, time-consuming, and full of errors if done manually. Hence, technology can provide a great respite in this area by automating computations and fetching relevant financial statements from government sources digitally.

Technology can also prevent over-exposure to a customer or group. When a loan application is received, the bank first needs to verify if the applicant’s limits are in order, at the company’s level, group level, or even the facility or individual level. With numerous borrowers in the bank’s databases having multiple product relationships, the task is near impossible to track manually. Technology solutions can make it possible to track it actively, and if needed, daily.

Each company and its group limits can be made available in a digital format, with a limit tree structure. It will enable the system to prevent over-exposure to any entity.

Identifying early NPA signals

With the availability of multiple data sources and ready analytical tools, technology makes it possible to pick up early signals when any loan is on its way to going bad. Today, various data sources provide information about the economic, industry, segment, and sub-segments trends. When technology processes assess the bank’s portfolio, it can give the banks a pre-emptive view if any part is going through hardship with analytics. Enabling banks to make informed strategic decisions may include cautious/stop lending to such segments or take pre-emptive pre-delinquency measures.

If a retail customer takes a home loan, could undertake revaluation of the asset later to check if the asset price has fallen below the exposure level. The system could then indicate a review of the collateral. For corporate loans, the system must constantly monitor financials and track covenants. If the system enables digital capture of financial statements and covenants, the internal and external users can be alerted accordingly.

The expertise of Nucleus Software

Solutions from Nucleus software have helped financial institutions onboard and manage healthy asset portfolios. These solutions provide a holistic lending environment to lenders by providing end-to-end lending solutions right from loan origination to loan servicing to delinquency management, with underlying lending analytics capabilities.

The solution gets its strength from 8 core engines, including rule, policy, workflow, repayment, allocation, communication, etc., providing the ability to configure and robustness across the lending lifecycle.

FinnOne Neo CAS allows financial institutions to integrate the various acquisition and pre-disbursal processes. It supports the entire acquisition lifecycle from customer walk-in to decision by the underwriter for the disbursement, ensuring high-quality credit.

FinnOne Neo LMS provides end-to-end loan servicing. It enables banks to improve their lending solutions’ agility, transparency, and efficiency, including complete automation of its processes. These include NPA classification, tagging, provisioning, and reporting. The solution also allows for tracking and monitoring various posts disbursal activities ensuring credit quality remains healthy during servicing lifecycle.

FinnOne Neo Collections is a customer-centric, web-based, and workflow-driven solution that allows financial institutions to manage, monitor, and control delinquent loan accounts while automating the loan collections management framework.

With cutting-edge machine learning and artificial intelligence capabilities such as text analytics and neural networks, Nucleus Lending Analytics is an easy-to-build platform that helps you make better credit decisions through accurate predictive models.