In 2016 and 2017, we wrote ‘State of the art’ articles on the Indian bankruptcy reform. This article presents the third in this series, the state of the art as we see it in 2018.

A few anecdotes about IBC at work are high in the public imagination. The takeover of Bhushan Steel by Tata Steel was an unexpected output in nearly all elements of the insolvency. Bhushan Steel is a plant with a capacity of 5 million tonnes of steel per annum which was producing 3 million tonnes per annum under financial distress. Under the new management, it is expected to produce 5 million tonnes of annum. The recovery rate for creditors was an impressive 63 percent. Unsecured bond holders, expecting nothing, received significant cash. Another surprising resolution was that of Sharon Bio Medicine, where the unsecured creditors got close to 98 percent of their dues.

Those that stand out include Binani Cement, Essar Steel, Jaypee Infratech.

There has been a great deal of rumination of these cases, and other such anecdotes in the policy discourse. We try to think systematically about these questions in this article, which distills the insights of the associated literature. It is useful to think within the classification scheme of inputs - outputs - outcomes. In the case of the Indian bankruptcy reform, these three pillars map to:

• Inputs: Laws (both Parliamentary law and subordinate legislation), the institutional infrastructure that is required for the IBC to work, and capabilities of various private persons.
• Outputs: Transactions that go through the system and the performance of all parties that come together to make these transactions happen.
• Outcomes: Recovery rates, the growth of credit and the rise beyond simple secured credit, and the deeper changes in behaviour by private persons who borrow and lend. All these agents re-optimise their strategies based on their perception of the working of the bankruptcy process.

The graph above shows the ‘survivor function’ of cases that are admitted at NCLT on date 0. At date 0, the survivor function takes the value 1, which means that 100 percent of the cases are pending. As time flows forward, some insolvencies are resolved. The cases that are still pending are the ‘survivors’; the survivor function shows the fraction of cases that are not resolved.

The IBC intended that cases would ordinarily be completed in 180 days.

Things are more challenging when it comes to the biggest cases. The dashed red line shows the RBI-12 firms. For these firms, none was resolved in 180 days and at 270 days, the survivor function had reached 0.9.

While 180 or 270 days may appear to be arbitrary cutoffs, they were enshrined in the Act and this should matter.

If, instead, we want to simply apply common sense, and wonder about what gets done in the first year after a case is admitted, the answer is: about half of all cases are resolved. But when it comes to the RBI-12, about 90 percent of the cases are pending, one year after admission.

From 2015 onwards, it has been clear that pulling off the IBC reform is a complex puzzle. By 2018, we are able to compute this survivor function of admitted cases, and we see that there is a problem.

Outputs: Resolution Outputs: Resolution Plan Versus Liquidated

When a firm is resolved, it can be liquidated, or it can transition into a resolution plan. When we think of the outputs of the IBC, it is useful to know the fraction of firms that fall into each of these buckets. Surprisingly, there is a third outcome: cases that are dismissed by the appellate authority. We summarise the facts here:

Outputs: A Few Firsts

First case where interim financing is obtained?

Not known. There is no public data about interim financing.

First vote by a creditors' committee?

Not known. There is no public data about what happens in the IRP.

First case to complete IRP with a super-majority in favour of a restructuring plan?

Synergies Dooray Automative Ltd.

First case where the promoters are rejected?

Innoventive Industries. This is a guess, as the relevant documents about the bankruptcy process are not released.

First case to commence liquidation?

Bhupen Electronic on 7th August 2017.

First case to complete liquidation?

None.
First individual insolvency that commenced?

Provisions for individual insolvency are not yet notified

Outcomes: Important Gains Are Taking Place Outside The Public Eye

The IBC is delivering poor outputs. Particularly with large firms, the institutional arrangements do not deliver rapid resolution. There is increasing gloom surrounding the bankruptcy process for large firms. When enough money is at stake, it is efficient for persons to spend millions of rupees of lawyers per year, and obtain delays through litigation.

Private negotiations are always more efficient and reach better outcomes when compared with the formal process of bankruptcy law. In out-of-court negotiations, debtors and creditors have more flexibility on the structure of the resolution than when under the watchful eye of the IBC adjudicator. For instance, efficient resolution plans will involve various schemes under the Companies Act (Sections 230-234, Companies Act 2013), which are already available to the stressed firm. What the IBC does is to provide the threat of the worst case scenario. Numerous private negotiations take place under the shadow of IBC:

1. Shareholders of many stressed firms would sell off the business, or obtain fresh equity capital, at an early stage of distress, when they know how things are going to work out under IBC.
2. There is anecdotal evidence that firms are behaving better with regard to financial and operational creditors, now that the threat of IBC is in the picture. The IBC helps set the stage for a better renegotiation between the borrower and the lenders.

These gains are among the desired outcomes of the bankruptcy reform. Since they tend to be invisible to the public eye, they tend to generate no news stories. For this reason, the gains for India owing to the bankruptcy reform tend to be understated.

One area where this learning by Indian promoters may be seen in the quick negotiations and settlement of cases after an IBC filing. This is seen in the FRG Insolvency Cases Dataset : Of the 830 cases for which were final orders were passed by the NCLT till 30 November 2017, 137 were dismissed as settled.

We should focus on how the incentives of promoters are changed. For these agents, the important issue is: Will we be harmed if we default?. The IBC is likely faring well on this, even if only a small fraction of large cases are resolved in a year.

In this thinking, it is important to distinguish between the stock of defaulted firms and a decision made from today onwards about default. For the stock, IBC induces fresh thoughts in negotiation for all parties. For the future, borrowers see their prospects after default, and are less likely to default.

Inputs: The Law

Amendments to the law have been energetically pushed out. A first amendment was in January and a second in August of 2018. But the law remains a work in progress in many ways. Gaps between the text of the law and the design principles of the BLRC continue to cause problems in the other elements listed ahead. Further, some of the amendments have become the source of new problems in the smooth functioning of the law. Three major issues concern the inclusion of home buyers as financial creditors, the change in the threshold of cram down from 75 percent to 66 percent, and the 29A amendments ruling out the debtor from providing a resolution plan during the Insolvency Resolution Process.

Further, promoters may have little incentive to preserving value once it becomes clear that a firm is headed towards default. In the extreme scenario, promoters may accelerate cash extraction if they see a world where theft goes unpunished and default leads to a loss of control of the company. At the same time, participants in the IRP are unlikely to compete with the promoter, so keeping the promoter at a distance from the IRP may give confidence to others who are interested in taking over the enterprise. In the BLRC design, it was important to differentiate between promoters who were guilty of theft and fraudulent practices from those who were not.

Conceptual thinking on the exclusion of promoters from the bankruptcy process came out after 29A was enacted (Feibelman and Sane, 2017; Sengupta and Sharma, 2017; Shah, 2017). As argued by these articles, 29A has given unanticipated effects, and an escalation of delays in the bankruptcy process. Private detectives all over India are ferreting out family connections and defaulted businesses of the Indian business elite. This section has become a tool for competitors to hinder the resolution of a distressed company.

Many amendments of the law, such as 29A, are based on practical considerations and lack conceptual coherence or grounding in the empirical evidence.

An example of this lies in the tension between recovery rate in one particular case versus the sanctity of process. We have been too quick to jettison the sanctity of process. While this may seem nice at the level of a certain case, it undermines the incentives of participants to take the IBC process seriously, and to commit long-term resources in their participation (Shah, 2018).

Inputs: The IBC Institutional Infrastructure

The law required four new elements of institutional infrastructure to become operational. Given such a burden of requirement, it is quite a remarkable feat of implementation that within six months of the law being passed, the first cases were being filed under this new law. The adjudication authority, the National Company Law Tribunal, was already in place by the time that the law was passed. So, the front-runner in institutional development was the Insolvency and Bankruptcy Board of India, IBBI, which was up and running on the 1st of October, 2016. After that, came the Insolvency Professional Agencies, IPAs, and the Insolvency Professionals, IPs. The Information Utilities or the IUs, are yet to make their presence felt.

The Regulator And The Subordinate Legislation

The regulator has been one of the four institutions under the law that has performed well in its startup phase. But now that this phase is behind us, there is a need to strengthen institutional capabilities and feedback loops between the regulator and the rest of the insolvency ecosystem (Khurana, 2017). The checks and balances described in Roy et. al. (2018) need to be put into the law to ensure accountability.

IBBI is at the Indian frontier, on the sound working of regulators, by binding itself to a sound regulation-making process. At the same time, a research-driven regulation-making process requires a commensurate statistical system and a research community, both of which are at present lacking.

The IPA Industry And The IP Industry

The IPA industry is still to establish itself as a competitive industry, where the IPA plays a regulatory role. The BLRC had visualised the IPAs would play a regulatory function that is akin to what NSE and BSE do in the brokerage industry, and this has not arisen. The IBBI has recognised the difficulties and is now exploring the SEBI approach to financial market infrastructure institutions. On the other hand, a large IP industry has developed, with IPs often being placed into complex situations that they are poorly prepared for. As an example, Shrivastava, 2018, brings back remarkable tales from the field, where learning by doing is taking place. However, in the absence of IPA capability, IPs remain largely unregulated.

The IPs have often been deferential to promoters. Under the IBC, the IP is supposed to choose whether to conduct a forensic audit, to look back in the recent two years for malfeasance in the form of cash extraction from the business. There is little evidence of IPs having done this. We are able to see one example: The IP of Binani Cements conducted a forensic audit into the affairs of the company and has filed an application before the NCLT alleging diversion of funds by promoters (Lavi, 2018). Since IBBI was constituted, no IPA has taken action against an IP, and only 4 final orders have been passed by its Disciplinary Committee.

The IU Industry

The Information Utilities are an important institutional feature of the law that has yet come into being. Participation in this industry has been hobbled through restrictions in regulations and concerns about legal liability of the IU. The block of Parliamentary law required for the IU industry to work properly has not come about. While the BLRC had envisioned a private, competitive IU industry, there are attempts to create a monopoly IU by the RBI (Regy, 2017). Public sector monopolies tend to work poorly and the attempt to establish a Public Credit Registry can further derail the nascent IU industry.

Adjudication

The block of Parliamentary law required for the NCLT to work properly has not come about. As a consequence, the NCLT has taken shape like a conventional Indian tribunal. The case load has started generating a back log and delays at NCLT, due to capacity constraints. The BLRC vision was for the adjudication to not engage in commercial decisions during the process of insolvency resolution. NCLT, and amendments to the law, are bringing NCLT back into commercial decisions (Datta and Sengupta, 2017; Feibelman, 2018).

The NCLT and other courts have now begun using their adjudication powers to favour resolution as opposed to liquidation; re-bids are being allowed and even resolution is being given another chance.

The NCLT has also taken the view that piecemeal sale of assets is not allowed even though the IBC does not expressly prohibit such a practice. Before the relevant amendments were brought in, the Supreme Court (and even NCLTs) was allowing settlement of cases after admission, something that was not provided for under the IBC as it existed. Each such event creates a flawed jurisprudence, and harms the objectives of the Indian bankruptcy reform.

Three of the four institutions under the IBC took off at an admirable pace. However, under the combination of being tasked to deal with cases not under its domain (such as the defaults to home buyers which would have been more optimally dealt with under the Real Estate (Regulation and Development) Act, 2016) or with large cases in its infancy (such as the large 12 cases that the RBI placed into the IBC in 2017) and being inadequately resourced to deal with this jump in workload, the institutions are predictably beginning to show delays in performance.

Inputs: The Creditors

Our articles in earlier years raised concerns about the incentives of lenders and the efficiency of their choices, given the weaknesses of banking regulation and the dominance of public sector banks. These concerns remain:

Inputs: The Buyers

For high recovery rates, there should be a large base of participants in the IRP from outside the existing creditors and debtor. There have to be interim finance providers, specialists in managing financial stress who can buy loans and debt securities from creditors who have lower appetite in holding those assets, as well as buyers who are specialists in resolving business stress itself. The Indian credit ecosystem has relatively few participants who can play these multiple roles. As a consequence, the existing evidence on transactions in the IRP stage currently present a skewed and stunted picture:

1. The present legal framework constrains the sale of stressed bank loans only to the Asset Reconstruction Companies (ARCs), who in turn have to hold a fixed fraction (15 percent) of these assets in their book. There are a limited number ARCs, holding a limited amount of capital that is available to purchase stressed loans. This increases the cost of a resolution both because of higher delays (going through an ARC will necessarily increase the time taken to participate) and higher costs (a transaction through an intermediary introduces an additional layer of cost, compared to directly transaction with the debt holder).
   
2. A standard approach taken during the IRP is to aggregate a sufficient base of the credit, so as to control and manage the decision of the committee of creditors. There appears to be significant uncertainty in what constitutes the liabilities of the stressed entity, which in turn decides the weights that various creditors hold in the committee of creditors. For example, IDBI Bank which was a creditor in the Binani Cements IRP had their weights in the committee of creditors adjusted upwards at a significantly advanced stage of the bidding process. This creates delays and deters outside participants. There is similar uncertainty in identifying the assets that are available in the stressed entities. The IUs were intended to have solved these problems, but the IUs have not materialised.
   
3. Many financial and non-financial firms have expended resources on developing bids, only to face failure for unreasonable reasons such as a deviation from the process defined in the law. This has generated reduced interest from prospective participants. In numerous recent experiences, the IRP seems to obtain only one or two bids. Many private equity funds or large listed companies have seen the difficulties of the IBC and made a conscious call to avoid IRP participation.
   
4. We are still at the early stages of a process where new financial firms such as private equity funds have successfully completed a stressed asset transaction. After such experience builds up, dedicated stressed asset funds come about and succeed, and then a great wave of capital comes in. This process is still many years away, given the paucity of credible outcomes so far.

An Input That Is Lacking: A Knowledge Rich Environment

A recurrent theme disrupting a smooth path towards easier participation and higher certainty of outcomes is a lack of abundant, commonly available data. Defaulting firms, lenders, IPs, IBBI and NCLT have disparate facts about transactions, and systems for dissemination do not yet exist to bring these all together. The few initiatives on data release tend towards being practitioner oriented: e.g. the user has to supply a case docket number. There are no public systems that release machine readable data-sets that can be used for research.

Where is data for the working of the Code coming from? The most readily and consistently accessible source of such information is the set of orders that are issued by the NCLT at various stages that an admitted case goes through in the process towards resolution. The data-set of cases collated in the FRG IBC dataset is collated after analysing the cases with orders posted at the NCLT website about admission or dismissal into the IRP. The second source of information is a similar collection of orders IBBI website and information derived from these orders in theIBBI newsletters. The data dissemination systems are not perfectly synchronised yet. There are gaps observed between the derived data in the IBBI newsletters compared to the source NCLT orders.

For listed companies, the core principle of securities law is that all information that shapes the thinking of speculative traders should be disclosed rapidly to the market. The existing principles-based SEBI LODR correctly embeds this principle. But today, neither defaulting listed companies nor listed lenders are disclosing default events, at what stage of the resolution process they are at in the IBC, the details of chosen resolution plan, etc. (Shah and Zaveri, 2018). In such an environment, there is little published research on the Indian bankruptcy reform. Initiatives like Chatterjee et. al. (2018), where the authors hand-construct a data-set and analyse it, are valuable but rare.

Decision making by practitioners and policy makers tends to pursue outliers. While such seat of the pants thinking does generate certain self-correcting mechanisms, the pace of learning is inefficient. Sengupta and Sharma, 2018, have built an annotated reading list which pulls together the extant literature. A lot more is required by way of data and research, to strengthen this literature.

The Return Of Nurturing

The Bankruptcy Legislative Reforms Committee (BLRC) is explicit on the purpose of the bankruptcy code: it is to achieve a rapid resolution. Whether this takes the form of debt restructuring, enterprise restructuring, or enterprise liquidation is not important. The key point is to rapidly resolve the enterprise failure that is visible at default, and to release capital.

In recent months, we see a return to the old rhetoric of SICA and BIFR, where the desired outcome was the continued survival of the firm, and that there are persons other than the committee of creditors who should have a say in the commercial decisions that the law put in the hands of this committee. The term `stakeholders' is being used a lot. The judiciary, for instance, is assuming roles in deciding who is eligible to be a resolution applicant, is demanding progress reports during the resolution phase, is deciding whether fresh bids need to be submitted and is even providing a fresh opportunity at revisiting resolution proposals even after the expiry of 270 days.

This is antithetic to the concept of bankruptcy of firms. If such ideas gather momentum, they could potentially feed into the future decisions of judges, regulatory staff or law makers. This would derail the 2016 attempt at Indian bankruptcy reform.

There should be no deviation from this ethos of a market-based economy.

Conclusion

The Indian bankruptcy reforms, unlike in most other countries, have been developed top down rather than bottom up. The bankruptcy process requires an institutional infrastructure: the law, IBBI, subordinate legislation, IPAs, IPs, IUs, organisational capabilities in lenders, a wide pool of buyers.

In two years since the start of the reforms, institutions that have a well-defined role in the process of insolvency resolution have been rapidly developed and deployed to carry out their tasks (the regulator, the adjudicator and the insolvency professional). Those institutions that can be substituted for are yet to get the stimulus and support for development (IPA, IUs).

Into this fragile, relatively unproven institutional infrastructure, when large cases were placed, we engaged in premature load bearing (Shah, 2018). As a consequence, only five of the 40 large cases have emerged from the IBC process.

We are still at the early stages of revelation of the Indian corporate credit crisis. There is a large pool of distressed companies in India (Shah and Sinha, 2018), which will start to trickle into the bankruptcy process, when private negotiations with financial or operational creditors break down.

The bankruptcy reform has yielded valuable gains in the behaviour of borrowers, both towards financial creditors and towards operational creditors.

But a bankruptcy process which reliably goes from a default to a solution within six to nine months remains elusive.

The way forward lies in building data-sets and knowledge, and high quality policy teams, which will carry this work program forward on all the multiple fronts described above, both within the IBBI and outside of it. As with many other parts of Indian economic reform, the centre of the puzzle lies in establishing feedback loops, from evidence to research to policy.

This article was originally published on Ajay Shah’s blog: The Indian bankruptcy reform: The state of the art, 2018, by by Ajay Shah and Susan Thomas, December 22, 2018.

Ajay Shah is a professor at the National Institute of Public Finance and Policy. Susan Thomas is an assistant professor at the Indira Gandhi Institute of Development Research.

The views expressed here are those of the author’s and do not necessarily represent the views of BloombergQuint or its editorial team.