How to make India’s bad bank workable

Such an ARC can collate large non-performing assets into a single entity and facilitate their resolution. It can help free up the rest of the financial sector ecosystem to focus on the core business of funding India’s large growth aspirations.

India’s past experiments with bad banks ended as failures because crucial design imperatives were ignored. The new ARC has to be manned by seasoned distressed debt management professionals. The transfer of assets to the ARC has to be at a Goldilocks just-right price—neither too high, nor too low. The entity should be able to enlist the support of key stakeholders in the system—especially the government—while avoiding any moral hazard issues. It should have a pre-defined, limited shelf-life.

To achieve these objectives, in line with the Malaysian experience post the Asian crisis, a time-bound government-owned ARC should be formed. There are ways in which a “fair price" for asset purchases can be determined, such that neither is the resolution process impaired, nor do supernormal profits accrue to any stakeholder.

Beyond cleansing the financial services ecosystem, a well-designed ARC can help revive the real economy by recommending policy steps to address chronic stresses in sectors such as real estate and infrastructure. Finally, a one-time ARC has to necessarily be accompanied by comprehensive governance and management reforms, so that this cycle of lending and taxpayer bail-out does not perpetuate.

The context

The RBI’s Financial Stability Report (FSR) of January 2021 suggests that banking Gross Non-Performing Assets (GNPA) may rise from 7.5% of advances as of September 2020 to between 13.5% of advances and 14.8% of advances by September 2021.

This is over and above the ₹8.8 trillion of assets that have been written off by banks between FY15 and FY20, but where recovery processes may still be underway. In addition, RBI’s stress tests on 200 large Non-Bank Finance Companies (NBFCs) indicate that their GNPAs may rise to between 6.8% to 8.4% of advances.

Put together, across all banks and NBFCs, we may be staring at non-performing assets between ₹26.7 trillion and ₹28.8 trillion by September 2021, or 13.7% to 14.8% of this year’s GDP. These enormous numbers weigh on our financial services ecosystem.

The Insolvency and Bankruptcy Code (IBC) is an excellent piece of legislation. Data from the Insolvency and Bankruptcy Board of India (IBBI) shows that till September 2020, 2,066 cases had been disposed off either through settlement, withdrawal, resolution or liquidation. The 277 resolutions yielded a creditable recovery of 43.5% of claims.

However, 1,942 cases were still pending for resolution, of which 1,442 cases were already more than 270 days old. Given IBC is on standstill as a result of covid-19 till March 2021, and the impending increase in GNPAs, there are many more cases that are waiting to be admitted.

While the IBC infrastructure is well capable of handling steady-state incremental stressed assets, the enormous existing stock of GNPA cries out for a one-time exceptional resolution mechanism – and hence the need for a “bad bank".

While smaller retail and MSME loans are best handled by the lender on record, there is a case for a bad bank to aggregate larger stressed assets, with the objective of taking the larger GNPA out of the financial services ecosystem and bring them onto a specialized entity that is resourced to specifically handle large, distressed debt.

As mentioned earlier, the transfer of GNPA from the lender to the bad bank has to be at a ‘just-right’ price. If the GNPA is transferred at too low a value, the recovery process might deliver supernormal profits to the buyers. This could be especially problematic if the eventual buyers are in the private sector.

If the transfer value is too high, such transfers would merely shift the burden from one entity to another, leaving little space for any realistic resolution. This was one of the many issues with India’s 2004 experiment with the Stressed Asset Stabilization Fund (SASF) to handle a chunk of IDBI’s stressed assets.

Ideally, to preserve capacity and jobs, resolution should be the default outcome from any insolvency process, rather than liquidation.

In many of the large, stressed assets in power, infrastructure, real estate, and construction, any bad bank will likely require appropriate government policy action and support for any sustainable resolution. This does make the case for a deep government involvement in any bad bank to address concerns around any moral hazard.

The bad bank should be designed with the clear mandate to wind down in, say, five years.

The nuts and bolts

Here’s a construct of an Indian Bad Bank ARC that could meet the above objectives and design considerations. This draws heavily from the reasonably successful Malaysian experience with its stressed asset management AMC ‘Danaharta’, post the Asian crisis.

I suggest a government-owned bad bank ARC, capitalized the same way recapitalization of public sector banks has been undertaken by the government—by issuance of special government of India (recap) bonds. To start with, ₹1trillion of capital could be infused, against issuance of special GOI bonds to the same entity. This could support the purchase of ₹2trillion to ₹3 trillion of stressed assets, assuming a 33% to 50% haircut.

The ARC should be headed by an inspiring and independent finance professional of strong repute—such as Deepak Parekh or Aditya Puri. It should similarly have an investment committee and board of unimpeachable repute, experience and credentials. On the ground, it should be manned by professionals with knowledge and experience of distressed asset management in sectors such as infrastructure and real estate.

The ARC could consider all stressed assets in India’s financial sector ecosystem across banks and NBFCs, beyond a threshold size of say ₹1,000 crore per borrowing entity. A database of such assets across lenders could be obtained from RBI’s Central Repository of Information on Large Credits (CRILC).

The ARC could be tasked with bidding for each of the large, stressed assets at a fair and equitable price.

To take a hypothetical example, the ARC might bid to purchase a stressed power sector loan at a 60% haircut, and so offer ₹40 for every ₹100 of loan. A lender can accept the ARC’s bid, in which case the ARC would pay 40% of the face value of the stressed loan in special GOI bonds.

Alternatively, the lender can reject the ARC bid as being too low. While retaining the loan, however, it would be required to ensure steep loan loss provisions, of say 64% of the face value. In effect, net of provisions, the loan would be marked at most at 36% of face value, 10% lower than the ARC bid of 40%.

As a further incentive, lenders would be assured of receiving back 80% of the excess value over the bid price, as recovered by the ARC through the resolution process. In the example above, if the ARC actually recovered 60% via resolution, it would transfer 16% (80% of the additional 20% over the bid price of 40%) back to the original lender.

Note that if the ARC were to recover less than 40%, that loss would be solely to the ARC’s books—there would be no recourse to the original seller of the stressed asset.

Such a mechanism, which broadly mirrors the mechanism used by Danaharta in Malaysia, would help ensure some semblance of a Goldilocks “fair price" in the asset transfer.

The ARC can also offer policy recommendations to address chronic issues in severely-stressed sectors such as power, real estate, construction, telecom, and airlines and shipping. Such policy tweaks would also allow the ARC to bring about timely resolutions, while preserving underlying businesses and jobs.

Government ownership of the ARC—alongside the commitment to transfer back a substantial chunk of the surplus recovery to the original lender—would also reduce the moral hazard associated with effecting policy changes that improve recovery and resolution prospects.

Beyond just cleaning up the financial services ecosystem, therefore, a well-designed ARC could also pave the way for sustainable all-round economic recovery and growth.

Finally, the ARC would have a strict mandate to wind down in five years—this has to be a one-time exercise and cannot be even remotely construed as an invitation to perpetuate the cycle of bad assets creation and resolution.

Life beyond a bad bank

A one-time resolution through an ARC would be incomplete by itself, unless accompanied by serious governance and management reforms to ensure that we do not slip back to repeating the cycle of lend, extend, pretend and finally bailout.

The past few years have exposed deep weakness in every aspect of governance checks and balances in respect of our financial services ecosystem. These governance issues are ownership neutral—they permeate both public sector and private sector entities.

While there is some work underway, much more needs to be done to call to account and reform critical governance pillars such as the conduct and operations of risk management departments in financial institutions, auditors, boards, rating agencies, independent analysts and regulatory supervisors.

Beyond governance issues, Public Sector Banks (PSBs)—accounting for 60% of banking in India—also need special attention. The RBI’s FSR reckons that PSB GNPA could rise to 16.2% of advances as a baseline, more than double the 7.9% of advances of private sector banks.

While it is tempting to beat up PSBs as chronically inefficient, at the core, they do not enjoy a level playing field. For one, they are set up under the SBI or Bank Nationalization Acts rather than under Companies Act. They are controlled by the mandarins of the Department of Financial Services (DFS)—and, hence, subject to intrusion in their day-to-day banking operations, ranging from phone-banking to nudges towards Mudra and shamiana banking.

Their professional decisions are subject to scrutiny by the dreaded 3 Cs —CBI, CVC and CAG—something their private sector counterparts do not have to lose sleep over.

Finally, their middle- and senior- management compensation are capped by government pay scales that are substantially lower than their private sector counterparts. The P.J. Nayak committee (2014) laid out a clear road map for addressing each of these issues and more. The substance of these recommendations needs to be implemented now.

PSBs have to be provided professional autonomy and a level playing field with their private sector counterparts. While social obligations are part of every banking operation in India (consider the requirement for priority sector lending, for instance), the government should address specific social objectives using explicit fiscal instruments and transfers, rather than leaning on commercial organizations and their shareholders for this purpose.

In conclusion, given the severe overhang of non-performing assets in our financial services ecosystem, the finance minister’s intent to launch a bad bank ARC is very welcome. The Malaysian experience provides many pointers on how such an ARC may be successfully designed. Alongside, we will need comprehensive financial services governance and management reforms.

To be sure, this solution can well be tweaked. But the fact remains that for India to achieve its immense economic potential, the status quo is simply not an option.

Ananth Narayan is associate professor at SPJIMR.