Has the RBI started micromanaging banks? I had slipped in this question at a fireside chat with Reserve Bank of India (RBI) Governor Shaktikanta Das at the Business Standard Banking Summit last December.

He looked surprised: “This is the first time I am hearing that!”

He said the RBI had intensified and deepened its supervision. It’s a continual process: The RBI has changed its method as well as approach. Going beyond numbers, it is looking at the business model of banks and raising the red flag when required.

What’s more, the RBI takes banks into confidence all through. It’s a “consultative approach”.

Some things became clear. Historically, the RBI has always tried to keep crooked bankers at bay by issuing a circular a day. No longer. It is on high alert. Earlier, the regulator would applaud a bank that earned high interest income even from borrowers who belonged to the so-called prime category and whose books did not show any bad loans. Now, it holds the clapping and wonders whether the CEO is a banker or a magician!

What are the objectives of supervision? Protect depositors and customers, keep the financial sector stable, and ensure the flow of credit for economic growth. The RBI has not changed the objectives of supervision, but how it supervises the system. And, the change is dramatic.

Before we focus on the new regime, let’s look back.

Earlier, on-site supervision was the mainstay of the RBI’s supervision. About five years ago, the focus shifted to off-site. Now, there’s a balance. The RBI is increasingly relying on off-site analysis and being proactive by tracking data — daily, weekly, fortnightly, monthly and quarterly. The regulator identifies areas of concern using on-site supervision of banks and other financial intermediaries.

The RBI used to do transaction testing but the scope for this was reduced after it set up a new supervisory framework called SPARC or Supervisory Programme for Assessment of Risk and Capital. SPARC, based on the report of a committee headed by a deputy governor, K C Chakrabarty, shifted to risk-based supervision (RBS). At the core of this is an integrated risk and impact scoring assessment.

The RBS is a rather complex model. In the first phase of its rollout, 28 banks were brought under the framework from FY2013. The RBS involved a comprehensive evaluation of both present and future risks, and identification of incipient issues, and helped the central bank determine its supervisory stance to ensure timely intervention and corrective action.

This marked a radical shift from the earlier CAMELS regime of supervision. CAMELS or capital adequacy, asset quality, management, earnings, liquidity, and system and control, was compliance-based, mostly backward-looking and oriented towards transaction testing.

Under CAMELS, the banks were classified into just two groups — Indian and foreign. The RBS changed that. Banks are now grouped based on their profile, business model as well as size and perceived risks. By 2017, all banks were brought under this.

Historically, the RBI has been tilted towards on-site inspection and light-touch off-site supervision. In the past few years, the trend has been reversed.

Once the focus shifted from on-site to off-site supervision, transaction testing turned more thematic and based on samples. The process runs through the year. The inspections focused mostly on a bank’s headquarters while the large branches, specialised branches, controlling offices, etc., were not touched.

This is how Punjab National Bank’s Brady House branch on Horniman Circle in Mumbai, just about 100 metres from the RBI’s central office on Mint Road, incubated the largest-ever fraud in Indian banking. The branch never came under the regulator’s inspection scanner. Today, wiser with experience, the central bank is changing its inspection and supervision model.

The bedrock of this model is the Central Repository of Information on Large Credits (CRILC), established in June 2014. Before CRILC, data on bad loans was not available at one point. Once CRILC was in place, the banks started supplying data every month for all loans of Rs 5 crore and above. For accounts turning bad, the data supply was weekly.

They also have to classify borrowers as special mention accounts (SMA) of various levels to gauge the probability of such accounts turning bad.

If a borrower has not paid a loan instalment (principal or interest) for more than 90 days, the account turns bad and the bank has to make a provision or set aside money. Until the 90-day mark, it remains standard but stressed.

For instance, in case a borrower is not able to pay one instalment (30 days), it signals incipient stress and the account is called SMA-0. For non-payment until 31 to 60 days, it is SMA-1; and for 61 to 90 days SMA-2.

The RBI and the banks (but not the rating agencies) could access this data. The data gave a comprehensive view of the banking system’s exposure to every large borrower, and how the exposure to the same borrower was classified differently by different banks.

Most importantly, the central bank could see how funds were moved across banks to keep accounts “standard”. CRILC enabled the RBI to do the first asset quality review, popularly known as AQR, of banks in FY2016 to clean up their balance sheets by uncovering heaps of bad loans.

The supervisory model had shifted from CAMELS to RBS just a year before CRILC came into existence. While most banks had weak internal control and risk management, they tailored their internal audit to make the top management feel comfortable. The RBI and CRILC made things transparent.

The central theme of the RBI’s new supervisory regime is the proactive identification of excessive risk built up in the system and timely mitigation to improve the resilience of banks.

Simply put, the regulatory initiative was outcome-based till recently — the RBI used to get into action only when bad loans surfaced. But now, it is playing the role of a sniffer dog – track the symptoms to the root cause.

Earlier, an RBI executive in charge of supervision of a bank would have been garlanded by his boss had the bank been forced to make provision for a huge pile of bad assets that had surfaced because of intense scrutiny. Now, the same executive slips from the good books of his boss for having let the bad assets pile up on his watch. The mantra is: Nip any crisis in the bud.

Also, supervision has been harmonised. Earlier, the RBI focused only on the banks, did some soft-touch regulation for non-banking financial companies and mostly ignored urban cooperative banks. A June 2020 ordinance made the RBI the sole regulator of cooperative banks. Since then, the RBI has been supervising all financial intermediaries with equal zeal.

The architecture of supervision has migrated from being entity- to activity-based.

Incidentally, a January 19, 2018, International Monetary Fund (IMF) report titled “India: Financial Sector Assessment Program”, commended the RBI for the “remarkable progress in strengthening banking supervision”. Noting that the RBI’s supervision and regulation remain strong and have improved in recent years, the IMF lists the implementation of a risk-based supervisory approach as the key achievement.

What Raghuram Rajan started by setting up CRILC and Urjit Patel nurtured, Das is taking forward with speed and precision. The senior-most deputy governor, M K Jain, is driving the change. A former commercial banker, Jain oversees supervision, among other things.

The next stage of CRILC is LEI or legal entity identifier. It’s a 20-digit unique code to identify the parties involved in financial transactions worldwide. To start with, LEI was mandated from December 2019 for all borrowers who gave a lender an exposure of at least Rs 50 crore each.

I understand that an internal RBI study has found that the compliance level for small loans is extremely poor. There are a few banks that have either not reported LEI or furnished incomplete data.

More on this next week.

(This is the first of a two-part series)