Mumbai: Banks are considering asking the Reserve Bank of India (RBI) to ease classification norms for fraud-hit loan accounts, amid a surge in such cases in India.

Banks currently set aside capital worth the full value of their dues from a fraud-hit loan account over a period of four quarters to cover the risk of defaults by the borrower. Lobby group Indian Banks’ Association is considering asking RBI to allow banks to classify only that part of the loan amount as fraud where fraudulent transactions have been detected.

“A forensic audit helps us identify transactions where fraud has taken place. So the question is whether only that portion should be declared as fraud or the entire account?" a senior banker aware of the matter said on condition of anonymity. “If your nose is hurt, do you call your entire body sick? That’s the key question we are grappling with. We are planning to take up the issue among lenders next week."

In the case of Dewan Housing Finance Corp. Ltd (DHFL), lenders have already segregated the bad loan book from the retail loan book as part of a resolution plan. DHFL had been looking to sell its project loan book worth 35,000 crore to Oaktree Capital, which had earlier bought assets worth 3,000 crore from the housing finance company.

“In DHFL’s case, we will be declaring only a part of the loan amount as fraud. Assets of NBFCs (non-banking financial companies) are largely loan book. Why should all our accounts including retail be declared as fraud? Not every loan asset is bad. It’s the wholesale loan book where there are inter-related party transactions, which has issues," said Rajkiran Rai G., managing director and chief executive officer of Union Bank of India. “In case it is declared as fraud, we expect only 40% of the banking exposure worth 40,000 crore to be classified as fraud."

(Graphic: Santosh Kumar Sharma/Mint)
(Graphic: Santosh Kumar Sharma/Mint)

A draft forensic report by accounting firm KPMG discovered that out of funds worth around 27,000 crore borrowed by DHFL from banks for on-lending to homebuyers, around 10,050 crore was invested in mutual funds. The forensic report also found that about 25 group companies to which DHFL had lent a total of 14,000 crore had an average profit of about 1 lakh, raising suspicion that the mortgage lender might have diverted funds.

Forensic experts said classification of fraud based on the loan amount is difficult as the fraud happens at the entity level. It’s the intent of borrower that needs to be looked at, they said, adding that banks have been under-reporting cases of loan fraud.

“There is a growing trend of wilful default cases from a fraud perspective, which is hurting banks. Banks need to look at technologies and early warning signals from a fraud perspective," said Vikram Babbar, partner and financial services lead (forensic and integrity services) at EY. “However, banks are only looking at credit risks—whether money will come back or not. They don’t have proper system and methodologies in place to identify frauds. That should be the focus so that they can avoid provisions in future."

In its Financial Stability Report released in June, RBI said it was reviewing its “master direction" on frauds and considering additional measures for timely recognition of frauds and enforcement action against violations.

“There is a significant time lag between the occurrence of a bank fraud and its detection," said RBI. For instance, the amount involved in frauds that occurred between FY01 and FY18 formed 90.6% of those reported in FY19. The relative share of public sector banks in the overall fraud amount reported in FY19 was in excess of their relative share in the credit, RBI said in its report.

The RBI report also found that frauds related to loans continued to be dominant, constituting 90% of all frauds by value that were disclosed in FY19.

In this category, cash credit and working capital loan frauds dominated in the case of state-run banks, whereas retail term loans (non-housing) were a major contributor to such frauds at private banks.

According to RBI, one of the major areas of non-uniformity in processes pertains to identifying red-flagged accounts.

The red-flagging of accounts based on an indicative list of early warning signals is not uniform across banks, said RBI.