Bank recapitalisation: The myth around growth capital

Bank recapitalisation: The myth around growth capital The Leap Blog, November 21, 2017 (with Anjali Sharma).

NPA Ordinance: The impact of secrecy in ordinance making

NPA Ordinance: The impact of secrecy in ordinance making The Leap Blog, June 30, 2017 (with Pratik Datta).

Bank financing of stressed firms

Bank financing of stressed firms Ideas for India, June 7, 2017 (with Anjali Sharma).

The severity of the NPA crisis

Mint, May 30, 2017 (with Harsh Vardhan)

To understand the extent of the non-performing asset (NPA) problem in the Indian banking sector and come up with solutions, it is important to have a good measure of the problem. While the standard measures have their uses, they do not answer the question that everyone seems interested in: How severe is the NPA crisis? We propose a new measure—the ratio of NPAs to bank capital. This measure shows that the crisis is indeed severe, among the worst in India’s history.

Traditionally, two metrics are used to assess the scale of the problem—the ratio of NPAs (gross or net) to gross domestic product (GDP) and the ratio of NPAs to total loans. The ratio of NPAs to GDP measures the potential losses in relation to the size of the economy. This is especially useful in cross-country comparisons, given that countries are at different levels of GDP. The problem with this measure is that it does not indicate whether banks are able to handle the NPAs with their own resources—their capital.

A more commonly used metric is the NPA to loans ratio. This shows the fraction of bank loans that has turned bad. One shortcoming of this measure is that it suggests the problem can be solved through denominator management—growing the loan books of banks to make the NPA ratio smaller. This growing out of the problem approach is not a reliable one. It depends on the overall economic environment and on the demand for credit. It assumes that the source of the NPA problem is external to the banking system and not in the weaknesses of the lending processes. If the demand for credit is slow it becomes difficult for the banks to grow their loan books. Even if the demand for credit were high, it would be difficult for the NPA-ridden banks to grow. Prolonged NPA episodes erode banks’ capital and constrain their ability to grow their loan books.

In sum, the standard measures do not convey the full picture and can be misleading. A high level of NPAs in the banking sector need not necessarily create a crisis situation, if the banks have the capital needed to provision for the losses. Bank capital is also a key factor in resolving the banking crisis. Hence, banks’ ability to withstand NPAs is best measured in relation to capital.

In the graphs we plot the ratios of gross NPAs (GNPA) and net NPAs (NNPA) to bank loans and to bank capital for the last 20 years. We have also added the amount of restructured loans to GNPAs in both the panels from 2007 onward.

To understand the importance of the NPA to capital ratio, we compare the depth of the NPA problem across two banking crisis episodes in India. The banking sector had witnessed an NPA crisis in the mid to late 1990s. The surge in bank NPAs at that time took place in the aftermath of the introduction of robust income recognition and asset classification norms which exposed two decades of bad loans.

The NPA to loans ratio suggests that the current crisis is considerably less severe than that of the late 1990s. However, when NPAs are measured in relation to bank capital, the current crisis looks just as bad. In particular, the deterioration in the balance sheets of banks post-2010 is much sharper when measured using the NPA to capital ratio.

The drop in the NPA to capital ratios in 2016 looks hopeful but this is partly due to the additional capital received by many public sector banks as part of the government’s Indradhanush programme. Also, in light of the revised disclosures of NPA levels by some large private sector banks in the last few weeks, the 2016 ratios are likely to be much worse.

We also compare the growth rates of NPAs, capital and loans across the two crisis episodes. The average annual growth rates of GNPA and NNPA over the five-year period from 1997-2001 were 8.5% and 9.8%, respectively. During this period, bank capital grew 13.14% and bank loans grew 15.87%. The corresponding numbers for the current crisis are much worse. The average GNPA and NNPA growth rates for the period 2011-2015 were 45.9% and 54.9%, respectively. The average growth rate of bank capital for this period was 16.1%. Bank loans grew at 16.2%.

This shows that during the last NPA crisis, bank capital grew at a higher rate than NPAs. While the NPA to loans ratio was higher then, banks were not undercapitalised. They had better ability to withstand the problem. In the current crisis, however, the growth rate of NPAs has been considerably higher than that of bank capital, further underscoring the severity of the crisis. The growth rates of bank loans on the other hand have been similar across both crisis episodes.

The emphasis on the alternative measure of the NPA problem also highlights the importance of capital in resolving the crisis. If the NPA to capital ratio is to be restored to a level that was prevalent during the high growth years of 2003-2007, the capital base has to roughly quadruple. Even if we assume that roughly 50% of the net NPAs will be recovered by the banking sector, the capital base has to double. This is unlikely to happen through retained profits or sale of real estate or other similar strategies.

An attempt to revive the banking sector must include a credible commitment of capital for it to be meaningful. In absence of capital and accompanying structural reforms, any solution will be incomplete and the banking sector may remain in the quagmire for a long time to come.

Banking ordinance opens up Pandora’s box

Mint, May 23, 2017 (with Anjali Sharma)

The recently promulgated Banking Regulation (Amendment) Ordinance is aimed at resolving the non-performing assets (NPA) crisis in the banking sector. It creates an illusion of state action, and does little by way of addressing the real concerns. We highlight some of the problems created by the ordinance.

Resolution of NPAs is a two-stage process. The first stage involves assessing the viability of the debtor’s business. The second stage involves deciding whether the debtor’s company should be restructured or liquidated. Any such resolution, be it restructuring or liquidation, imposes losses on the banks that had lent money to the corporate debtor. The larger the losses, the higher the amount of capital needed by the banks to meet the Reserve Bank of India’s (RBI) guidelines on provisioning requirements. While the government has promulgated the ordinance, it has not made any commitment of additional capital to support the resolution efforts. Capital allocated for the banking sector in the 2017-18 Union budget, or as part of the mid-term capital infusion plan, falls short of what the banks collectively need.

In absence of additional capital, the RBI’s directions to the banks under the ordinance may impede the resolution process. The RBI may have no option but to direct the banks to extend lifelines to unviable companies to defer the problem to a future date. This can happen as part of the new Insolvency and Bankruptcy Code (IBC) or otherwise. As pointed out in the Economic Survey 2016-17, over the last few years, cash flows of the large stressed companies have been declining. Restoring their viability necessitates loan write-offs. For the banks, resolving these cases requires the most capital. Given the lack of capital, banks could be given the regulatory cover under the ordinance to refinance these large corporate debtors. We have already seen this happen under the RBI’s corporate debt restructuring (CDR) mechanism. The ordinance does not change the status quo where good money is thrown after bad and no real resolution of NPAs takes place.

Second, resolving a bank’s NPAs requires resolving the entity to which money has been lent. The ordinance may instead create perverse outcomes. Under IBC, banks as members of the creditors’ committee are required to vote on a resolution plan. If the RBI directs a bank to initiate IBC action against a corporate debtor, it may also have to direct the bank on the decisions in the creditors’ committee. By empowering the RBI to act, the government has taken away any incentive of the banks to act on their own. In India today, different banks are at different levels of capital adequacy. An IBC resolution plan that works for one bank may be inimical to the interests of another. As the banking regulator, RBI is responsible for the health of all banks. So it is possible that the resolution plan that finally gets approved by the banks under RBI’s directions will focus more on the health of the banks as opposed to addressing the insolvency of the corporate debtor. This defeats the purpose of an IBC resolution.

Third, RBI giving directions on the resolution of banks’ NPAs may undermine the IBC process in other ways too. It may thwart the incentives of third parties which would have otherwise been willing to offer their bids or resolution plans in a market-driven process.

Fourth, with the ordinance in place, all eyes are now on the RBI to resolve the NPAs of the banking sector. This could be problematic because the range of actions that the banks can take to address the problem is limited by the shortage of capital. The tools available to RBI are limited. If the RBI intervenes on a case-by-case basis, questions about conflict of interest, regulatory capacity and capability will arise. If RBI intervenes through general rules and conditions, this will be no different from the corporate debt restructuring (CDR) mechanism, the strategic debt restructuring (SDR) scheme, and the scheme for sustainable structuring of stressed assets (S4A) that have failed in the past to resolve the problem. Either way, the ordinance puts RBI’s credibility and reputation as a micro-prudential regulator at stake.

Fifth, non-commercial factors may also be at play when it comes to resolving the large NPA cases. In the absence of clarity on the rationale behind the ordinance, we conjecture that one reason banks have not been initiating resolution proceedings against the large stressed companies is because of pressure from politically connected promoters. The ordinance gives banks the regulatory cover to take resolution-related decisions but it is not clear whether it also gives the required political cover. If the RBI is to now get directly involved in these loan restructuring decisions, or indirectly through committees reporting to it, this would put it in a difficult spot.

Finally, public and private sector banks have non-government shareholders, and non-bank creditors. So do companies that may get referred to the IBC following RBI’s directions to the banks. Any action under the ordinance that adversely affects the interests of these parties may be litigated in court. In a litigation if courts take cognizance of the rights of private shareholders and rule in their favour, this can further weaken the IBC process. Also private parties, domestic and foreign, may view this as state interference in market processes. This may affect their future investment decisions.

The ordinance was presumably brought about because banks on their own could not trigger IBC proceedings against the stressed companies for fear of investigation and prosecution, or due to lack of capital or because of challenges in negotiating with politically connected promoters. The ordinance gives banks the regulatory cover to take resolution decisions, but it is, by design, limited in its capacity to resolve the crisis. It opens up a pandora’s box of new problems. Most importantly it puts the RBI in a difficult spot and makes the IBC vulnerable to potential abuse. It also creates the problem of misaligning creditors’ and debtors’ incentives farther away from an effective resolution.

Understanding the recent Banking Regulation (Amendment) Ordinance, 2017

Understanding the recent Banking Regulation (Amendment) Ordinance, 2017 The Leap Blog, May 8, 2017 (with Pratik Datta).

Challenges in the Transition to the New Insolvency and Bankruptcy Code

The Wire, December 15, 2016 (with Anjali Sharma)

The Insolvency and Bankruptcy Code, 2016 (IBC) replaces a fragmented legal framework and a broken institutional set-up that has been delivering poor outcomes for years for creditors and distressed businesses seeking an exit. Currently, corporate insolvency and debt recovery cases are being dealt with under various laws and forums. In a previous article, we had argued that almost all of these are now eligible to be initiated as new cases under the IBC. However, this will create a massive burden for the nascent institutional infrastructure that is being rushed to operationalise the IBC. In this article, we analyse some of the challenges of transitioning existing cases to the new law.

The IBC offers a time-bound resolution process aimed at maximising the value of a distressed business. This will benefit not just the creditor and debtor companies, but also the overall economy because capital and productive resources will get redeployed relatively quickly. To meet the objectives of timeliness and value maximisation, the IBC proposes a new institutional set-up comprising four critical pillars:

1. A robust and efficient adjudicating authority to hear the cases.
2. A regulated profession of insolvency professionals (IPs) to manage the insolvency and bankruptcy cases.
3. A regulated competitive industry of information utilities (IUs) to reduce information asymmetries in the insolvency resolution process.
4. A regulator – the Insolvency and Bankruptcy Board of India (IBBI) – to perform legislative, executive and quasi-judicial functions with respect to the IPs, and IUs and draft regulations for the resolution procedures under IBC.

Steps are currently underway to set up this institutional infrastructure. The National Company Law Tribunal (NCLT) has been notified as the adjudicating authority for the corporate insolvency and bankruptcy cases. The IBBI has been constituted and has started building capacity. The regulatory framework for IPs and IPAs is in place and some IPs and IPAs have already been registered by the IBBI. It may take some time for new IUs to get registered but even without the IUs in place, the provisions of the IBC related to the Corporate Insolvency Resolution Process (CIRP) and liquidation have been notified. While the former has been operational from December 1, 2016, liquidation provisions will come into force from December 15, 2016. Clearly, the speed of implementation has been prioritised leaving the institutional set-up to develop in parallel to the functioning of the law.

Challenges in implementation

The NCLT will face the biggest challenge in the process of transitioning existing cases to the IBC. The NCLT currently has 11 benches with 16 judicial members and seven technical members among them. Its mandate includes hearing cases earlier dealt with by the Company Law Board (CLB) under the Companies Act 2013, in addition to cases under the IBC. As of March 2015, there were around 4,200 pending CLB cases. All of these will now be transferred to the NCLT. In addition, the CLB receives around 4,000 new cases every year. Now these will have to be dealt with by NCLT. With the IBC provisions on CIRP becoming operational and the provisions on liquidation expected to be notified soon, all the 4,500 winding-up cases pending at the high courts as of March 2015 are also likely to get transferred to the NCLT. Our analysis shows that corporate recovery cases at the debt recovery tribunals (DRTs) and the rehabilitation cases at the Board for Industrial and Financial Reconstruction (BIFR) are eligible to be initiated as new IBC cases.

Therefore, given its limited capacity, how will the NCLT deal with fresh IBC cases, added to which will be the case load from the CLB, the high courts, the BIFR and potentially the DRT? Unless its adjudication capacity is enhanced, the NCLT will fail to hear and dispose cases in a timely manner from the start. For the IBC cases, this could mean the inability of the NCLT to adhere to the 180-day timeline that is the duration of the CIRP. This has earlier been seen in the case of DRTs too. DRTs were intended to dispose of recovery cases in 180 days, but given capacity issues and pendency, often the first hearing for a case takes place after 180 days. The 180-day timeline represents the core design intent of the IBC – rapid resolution of insolvency to maximise recovery. If this is compromised due to capacity constraints, the effectiveness of the IBC will get diluted.

The second concern related to the NCLT is regarding the case law that develops under the IBC. Given that it is a new law, the procedures and common practices under it need to develop independently from the case laws under the pre-IBC regime. Since the first cases that come to the IBC are likely to be the existing ones, the initial case law that develops under IBC will reflect the context of these cases. The behaviour of creditors, debtors, auditors, lawyers, valuers, liquidators are all steeped in the old case laws under the Companies Act 1956, the Sick Industrial Companies Act 1985 (SICA), the Securitisation and Reconstruction of Financial Assets and Enforcement of Security Interest Act, 2002 and the like. This will change only when the IBC gets perfected as a law and its institutional infrastructure reaches its capacity, allowing the NCLT to focus on upholding the design and objective of the IBC over all priors that exist. For this, the IPs, the IUs, the NCLT and the IBBI all need to be properly set up and functioning in a manner envisaged by the IBC.

IPs form the backbone of the IBC. Their role requires a fine balancing act, given that they are in charge of managing the debtor company and are accountable to the committee of creditors and the adjudicating authority for their actions. To ensure that the IPs perform their role without any misfeasance, well-defined entry barriers to the profession must be designed and the IPs must be closely regulated by the IBBI. A qualification examination has been proposed to get registered as IPs. This is modelled after the best practices of other jurisdictions such as Canada and the UK that have a well-functioning industry of IPs.

The IBBI, however, has now licensed the first batch of IPs only on the basis of their professional experience. Given the speed with which the law is being implemented, this seems to be the only option till the infrastructure for conducting the examination is set up. IPs registered in the first round also need to be prepared to handle the complexities of the existing cases. The IBBI needs to have adequate capacity to monitor the IPs and ensure that malpractice and fraud do not seep into the profession in the early days. For this, the IBBI needs to ramp up its human resources and its IT capability and build a far greater level of preparedness than it is at, right now.

The lack of IU infrastructure is going to be another challenge. Under the IBC, a CIRP can be triggered only when a default by the debtor company has taken place. In the IBC design, the IU enables a quicker initiation of cases by providing access to irrefutable and transparent evidence of the default. At present, winding up petitions under the Companies Act, 1956 and cases under SICA take around one to two years to get admitted. The situation is marginally better at the DRTs, where the Bankers Books Evidence Act allows bank books’ to be treated as primary evidence in cases. Even then there are delays in establishing the existence of a debt and default.

In absence of IUs, the IBBI is required to specify the evidence of default that can be used to trigger an IBC case. This can cause inordinate delays especially if the NCLT gets involved in evaluating whether a default has indeed taken place. Under the IBC design, IUs also facilitate the formation of the creditors’ committee within 14 days from the date of registration of a case. All information regarding the creditors’ claims needed by an IP to form the committee can be easily collected from the IUs.

It is therefore likely that in absence of IUs, initiating a case as well as forming the creditors’ committee will take far longer than envisaged in the IBC design. This will make it difficult to meet the 180-day timeline for completing the CIRP, giving rise to two possible outcomes. First, the delay in forming the creditors’ committee will reduce the time available to agree on a resolution plan. If the committee cannot agree on a resolution plan within the specified time limit, the NCLT will order the liquidation of the company. Secondly, the NCLT may exercise its judicial discretion and extend the CIRP beyond the time limit specified in the law. Both these outcomes have negative consequences. The former creates a liquidation bias in CIRP while the latter compromises the fundamental design of time-bound resolution in the IBC.

The manner in which the IBC is currently being implemented seems to focus more on expeditiously operationalising the law rather than effectively implementing it. These concerns, if not addressed suitably, will defeat the purpose of enacting a new insolvency law to improve the recovery rate in order to promote the development of credit markets and entrepreneurship.

Will it work?

The IBC is an important reform for India and its successful implementation depends on meticulous transition planning. The first cases coming to the IBC are likely to be the existing corporate insolvency cases. Four steps are needed to ensure that these do not impact the design and effectiveness of the IBC in an adverse manner.

The capability of the NCLT needs to be developed with adequate project planning. This could mean a special bench designated only for the IBC cases, scaled to the expected IBC case load and trained in dealing with commercial matters including the complexities of the existing cases. The NCLT needs to ensure that for any case coming to the IBC, whatever be its priors, the IBC provisions are followed without any exception. If the NCLT is built like a conventional Indian tribunal, it can develop a multi-year backlog very soon.

It is critical that the winding-up cases that come to NCLT are disposed of separately from the IBC cases and their outcomes in no way impact the IBC case laws. The capacity of the IBBI needs to be brought to scale as quickly as possible so that no malpractices or adverse priors develop in the IP practice and the entry barriers to the profession need to be implemented in accordance with the law. India has a long history of failure in the regulation of professions. The regulatory system for IPs must be designed to avoid creating yet another failed profession. Finally, the IUs need to be operationalised. In their absence, the IBBI needs to specify clear guidelines that enable the timely admission and disposal of the existing cases. This means a rapid creation of regulatory capacity in terms of people, processes and information technology systems.

Adequate institutional capacity is essential to ensure that the IBC does not suffer from the predicament of earlier reform attempts such as the DRTs. Doing all of these needs time and needs proper planning. Rushing through the implementation of the new law may serve to improve India’s ranking in World Bank’s ‘Doing Business’ report but may not result in a de facto improvement of the insolvency resolution framework, thereby defeating the very purpose of the IBC.