There is an urgency about bankruptcy reforms in India. The credit boom of the mid 2000s gave rise to many failed firms. There are 14,900 non-financial firms in the CMIE database with recent accounting data, and of these, there are 1,039 firms where there is not enough cash (PBIT) to pay interest. Firms under extreme financial stress are a drag upon the economy: they are unlikely to add capital or labour or obtain productivity growth. The exit of such zombie firms will free up capital and labour, and will help improve the financial strength of healthy firms. The economic purpose of the bankruptcy process is to close the circle of life; to recycle this labour and capital into healthy firms.
The numbers above are likely to be an under-estimate. CMIE only tracks the biggest companies. There are a very large number of smaller firms which are likely to be in default. There are other organisational forms used by firms, where we do not have data, and where there will be failed firms. Demonetisation and GST have stressed firms' health. Taken together, tens of thousands of cases are likely to be headed for the bankruptcy process.
When the Insolvency and Bankruptcy Code came into effect on 28 May 2016, we raised questions about the Indian bankruptcy reforms in an overview article (Shah & Thomas, 2016). A year has passed and we revisit those question. What is the state of the play in bankruptcy reforms? What of the new process is working well, and what are the areas of concern?
When the government and RBI decided to put 12 big defaults into the IBC, some felt this demonstrated the capabilities of the new bankruptcy code. When we speak with practitioners, we get a rush of war stories and practical detail. In this article we try to distill the hopes and fears, and try to see the woods for the trees.
Inputs, Outputs, Outcomes
In public policy, it is useful to think in terms of inputs, outputs and outcomes. As an example from the field of education, the inputs are schools and teachers. The outputs are kids who enroll. The outcome is what kids know, as measured through tools like OECD PISA or Pratham's ASER.
For bankruptcy reform, once the Insolvency and Bankruptcy Code was pased, there are:
Laws (both Parliamentary law and subordinate legislation), the institutional infrastructure that is required for the IBC to work, and capabilities of various private persons.
Transactions that go through the system.
Recovery rates, the growth of broader credit markets, and the deeper changes in behaviour by private persons who borrow and lend, who will re-optimise as the bankruptcy process starts working smoothly.
The inputs perspective
Nine inputs are required for the bankrutpcy process to work, as envisaged by the Bankruptcy Law Reforms Committee (BLRC):
An Amendment Act to fix the mistakes in the 2016 law.
The Insolvency and Bankruptcy Board of India (IBBI) has to achieve the scale required for a high performance regulator.
An array of well drafted regulations have to be issued by IBBI, with a feedback loop to feed from practical and statistical experience into a robust regulation-making process to refine the regulations.
A competitive industry of private Information Utilities (IUs) has to arise.
A competitive industry of private Insolvency Professional Agencies (IPAs) has to arise.
A competitive industry of private Insolvency Professionals (IPs) has to arise.
NCLT has to find its feet in dealing with corporate bankruptcy, and DRT has to do similarly for individual bankruptcy.
Financial firms have to develop capacity on how to best to initiate the insolvency resolution process and participate in the process to ensure an optimal restructuring plans collectively.
Strategic investors, distressed asset funds and private equity funds have to gain confidence about expected outcomes, either when making a bid for a going concern or when buying assets in liquidation.
Input 1: IBC Amendment Act
The IBC, 2016, suffers from conceptual errors. There are contradictions in definitions, ambiguous definitions, problems in the establishment of the IBBI, failure to establish sound processes at IBBI, the lack of legal foundations in the institutional infrastructure including insolvency professionals, insolvency professional agencies and information utilities, the lack of clear integration of secured credit (i.e. SARFAESI) into the main bankruptcy process, etc. As an example, the Working Group on Information Utilities, chaired by K. V. R. Murty (MCA, 2017) has a chapter on amendments required in the IBC in respect of information utilities.
The IBC has elementary drafting errors. Some examples of these are discussed in Malhotra & Sengupta, 2016, and Singh & Mishra, 2017. A fuller examination will reveal a larger list of such drafting errors.
The insolvency resolution and liquidation processes are procedural law where drafting errors can lead to litigation. As an example, Sibal & Shah, 2017, analyse an anomaly about antecedent transactions.
Many laws in India undergo a constitutional challenge in their initial days. Some founders/shareholders may go to the courts claiming that the IBC violates basic constitutional rights. Well funded attempts of this nature are likely with the 12 big cases. The Bombay HC dismissed an early challenge to the constitutionality of the IBC, but it did this without substantively deciding on its merits. We believe that, in the design of the IBC, sufficient thought was given towards ensuring due process and fairness to all creditors as well as the debtor. There may be certain drafting flaws in the IBC which the government may need to rapidly solve.
An Amendment Act is required which addresses these problems. In our knowledge, there is no drafting effort that is presently in motion to solve this.
Input 2: IBBI emerging as a high performance regulator
Indian regulators suffer from low State capacity. Capacity in a regulator comes about through five processes: (a) The composition and working of the board; (b) The legislative process; (c) The executive process; (d) The quasi-judicial process; (e) Reporting and accountability. Hygiene in these world-facing processes should have been codified in the IBC, but was not.
Considerable new knowledge has developed in the last decade on how to achieve State capacity by setting up such processes. A good deal of this is discussed in the report of the Working Group on the Establishment of IBBI, chaired by Ravi Narain (MCA, 2016). Many aspects of this report have yet to be brought into IBBI.
The IBBI has been set up. It has an office and a team. It has been shouldering the effort of drafting regulations on a very tight timeline. IBBI has done a particularly good job in some respects, such as the recent unveiling of a mechanism to take feedback from the public in structured documents about the regulations that it has drafted. In this, IBBI is now ahead of SEBI and RBI on good governance practices.
The regulation-making process that has been used by IBBI so far has good features. The organisation structure used at IBBI respects the difficulties associated with setting up a regulator that violates the principle of separation of powers. However, these things are not in the IBC, or in rules under IBC, or in legal instruments issued by the board. There is the danger that good practices may be fall by the wayside at a future date.
IBBI is expected to perform a certain statistical system role, and a research capacity that can use this to strengthen regulatory capacity. The systematic process of using data to improve the working of the bankruptcy process is critical to the bankruptcy reform. Statistical system work at IBBI has, however, not yet commenced.
Thus, in critical aspects, IBBI is going down the route of conventional Indian regulators such as SEBI or RBI. This will reproduce the well known infirmities of conventional Indian regulators and runs counter to what the bankruptcy reform requires.
Input 3: Sound subordinate legislation
IBBI has issued 12 pieces of subordinate legislation. There are flaws in many of these regulations, some of which are described under the other sections in this article, which will hamper the working of the bankruptcy process and of the institutional infrastructure under the Act.
A critical part of the bankruptcy reform is individual insolvency. Advancing on this front will get India away from the recurrent loan waivers which spoil the repayment culture of borrowers, raise the cost of lending to individuals, and harm credit access to individuals as well as small entrepreneurs. This part of the law has not been notified, and IBBI has not released regulations that are required to operationalise the individual insolvency component of the IBC.
Input 4: The IU industry
Under normal circumstances in an insolvency resolution process (IRP), a considerable amount of human effort is required in order to construct the list of creditors and the size of their claims. The BLRC design envisaged that this information would be stored in `information utilities (IUs)', as electronic records of credit contracts in computer systems, authenticated by creditor and debtor, which would thus eliminate delays and costs. In the Indian legal system, disputes about facts are well acknowledged as the source of delays, wastage of judicial time, and payments to lawyers. Irrepudiable records from IUs would eliminate these problems.
So far, no IU has come into operation.
One part of the problem lies in the IU regulations issued by IBBI. Prashant et. al. 2017 point out its anti-competitive features. The licensing requirement for the IUs are overly stringent, particularly for an industry where the costs of implementation are likely to be low because of the constantly decreasing costs of information technology. The regulations ask for capital requirements that are excessive, given the low levels of value at risk in the business. This is a new business with no precedent anywhere in the world. Entrants into this are are taking on the risk of failure of the business model, which must be compensated by sufficient returns to investment. The IU regulations simultaneously prescribe shareholding arrangements that deter enterpreneurs from viewing this as a viable business opportunity.
These barriers are likely to keep the Indian bankruptcy system from achieving a competitive industry of technologically capable IUs to serve multiple types of borrowers and lenders, as was visualised by the BLRC when designing the IBC. There are also other technical flaws in the IU regulations as is pointed out in Prashant et. al, 2017.
Input 5: The IPA industry
Insolvency Professional Agencies were envisaged by the BLRC as a strategy to regulate professionals through the structure of a self-regulatory organisation. However, the IPA regulations issued by IBBI have many features that vitiate this objective. As a consequence, the key players who were envisaged in this industry by the BLRC have been barred from it.
The existing players are generally passive and are not performing the role that is required of IPAs in the bankruptcy reform. In most aspects, the IPAs are going down the route of conventional Indian regulators-of-professions such as ICAI or BCI. This is likely to reproduce the widely acknowledged infirmities of these organisations, and is counter to what the bankruptcy reform had attempted to achieve by way of well regulated insolvency professionals who act in the best interest of the stakeholders in the resolution process. IPAs are expected to create certain supervisory databases. These would be extremely valuable in the process of diagnosing problems of the bankruptcy reform and addressing them. So far this has not come into being.
Input 6: The IP industry
A set of insolvency professionals (IPs) are in place. During the IRP, if required, the IP is expected to put together a temporary management team and temporary financing (under the guidance and approval of the creditors' committee) that will stabilise the firm. These are new roles for the IPs in India, who have yet to develop capabilities to carry out these functions, either within their organisations or through an extended network. It will take time and competition for IPs to develop the teams through which they are able to fully discharge such functions, particularly in complex cases.
One way to jump-start getting such capability would have been to open the industry to foreign insolvency professionals. Their participation, particularly at the early stages of the reforms implementation, would have helped augment capacity and diffuse knowledge. Most Indian IPs are likely to be in a repeated game with promoters; foreign IPs would be particularly valuable in their ability to be harsh with promoters, which would help set the tone for the working of the bankruptcy process. The entry of foreign IPs was, however, blocked by IBBI through the subordinate legislation (Burman & Sengupta, 2016).
A critical factor in dealing with a going concern is lining up interim financing. The IBC and the subordinate legislation fail to clarify the priority of interim financing, in case the firm goes into liquidation. This has hampered the ease of access to interim finance while in the IRP.
IPs are given considerable power in the working of the IRP. There is a need for regulation of the profession in order to deal with various kinds of misbehaviour that can arise. BLRC had developed sophisticated thinking on how the IP and IPA industries should work (Burman & Roy, 2015). A lot of this did not make it into the IBC or the subordinate legislation. The IPAs as constructed today are not performing the roles required of them in regulation of IPs. While IBBI has enforced against IPs on relatively trivial violations, IBBI itself is not equipped to enforce against the real challenge, of malpractice by IPs: it cannot overcome the lack of IPA capacity.
When IPs step into the shoes of the board, and make vital decisions, they are exposed to a new level of legal risk. There is a need for insurance against these risks. This is not yet available in the market.
Input 7: The working of NCLT and DRT
While the design of the IBC has many features that will yield speed under Indian conditions, the working of NCLT/DRT is still a key factor that will determine rapid resolution as part of the Indian bankruptcy reform (Datta and Regy, 2017).
At present, we see many problems, such as inconsistencies in the behaviour of NCLT across locations, a few orders that are wrong, the lack of orders organised as structured documents, low transparency on the web, and delays. These may be a small precursor of the difficulties to come, as the case load has thus far been mild. Most defaults are, as yet, not going to the NCLT as creditors are waiting to see how the IBC works out. If the bankruptcy reform progresses, we will go from a case load of 20 per month to $10\times$ or $100\times$ as much (Damle & Regy, 2017). At this level of load, the unreconstructed NCLT will experience an organisational rout and will become the chokepoint of the bankruptcy reform.
NCLT has gone down the route of conventional courts and tribunals, which has reproduced the well acknowledged infirmities of the judicial process in India. This is not commensurate with what the bankruptcy reform requires. New knowledge needs to be brought to bear on the working of NCLT (Datta & Shah, 2015).
Input 8: The thinking of financial firms
New thinking by banks and insurance companies is required if they are to play their part in the new bankruptcy process. However, their thinking is greatly shaped by regulations. Errors in the present body of banking regulations, and the associated enforcement machinery, have created an incentive to hide bad news, to not initiate the IRP and to vote irrationally in the creditors' committee.
Sound micro-prudential regulation is one which would require that when a default takes place, the lender must rapidly mark down the value of the asset to zero. This loss should go into the Income and Expenditure statement immediately. Once this is done, there is no overhang of the past, and the lender will be rational about recoveries. Whether the asset is sold off, whether IRP is filed, how to vote on the creditors committee: All these decisions will be based on commercial considerations alone. Recoveries in the future would flow back into the Income and Expenditure statement.
These issues have yet to make their way into micro-prudential regulation of banks and insurance companies.
Assuming RBI and IRDA address these mistakes in micro-prudential regulation, we may see a new cycle being established within two years. New defaults should immediately show up as expenditure, and there would be a flow of cash from old defaults where the bankruptcy process has been completed. This is the opposite of the present arrangement, where it is claimed that all loans are profitable other than some old loans which induce losses.
In the case of NAV-based financial firms, such as mutual funds and pension funds, the event of default should influence the marked-to-market (MTM) prices even if the secondary market for the bond is not liquid. Here also, the bias in micro-prudential regulation should be to mark down the prices to near zero values quickly. This will create incentives for these funds to sell off these assets to distressed debt funds, when these transactions would yield a price that exceeds the MTM price that was used internally.
On 4 May 2017, an Ordinance was promulgated that gave RBI powers to push banks to initiate IRP. There is less to it than meets the eye (Datta & Sengupta, 2017). Backseat driving in a few cases, even if done wisely, cannot solve the regulation-induced bad behaviour of banks. There is no substitute for the slow hard work of reforms of bank regulation and supervision.
Input 9: A pool of buyers
The reforms requires the presence of two key groups of buyers. These are strategic players in the same sector, such as Jet Airways for the Kingfisher bankruptcy, and private equity funds (Shah, 2017).
Small firms lack the ability to set up dedicated teams that focus on opportunities coming up in the bankruptcy process. However, there should be teams at the top 2000 companies that watch the bankruptcy process and look to buy up useful things that come along, either in the form of going concerns or the liquidation process. Private equity funds have not yet started looking at this area on a significant scale. A few pioneers will get started. As they reap strong returns, other funds will follow, and new money will be raised to pursue these opportunities.
We could have developed these capabilities through `asset reconstruction companies' from 2002-2016, but through mistakes in the RBI regulations for these (Shah et. al. 2014), that opportunity was wasted.
As long as the IBC and its institutions are unproven, there will be a shortage of buyers, which in turn will lead to very low prices for the stressed assets. At the early stage, buyers will fear legal risk in the IBC, and will shy away from investing in building organisational capital. The critical story in the evolution of insolvency institutions in India is the emergence of thousands of skilled professionals at private equity funds and the 2000 big companies, each surrounded by an ecosystem of lawyers, accountants and consultants, armed with capital, process manuals and authorisations, who are ready to go. We are at the early stages of that journey.
If completed transactions are the output of the bankruptcy process, we are still some way from observing outputs under the IBC. So far, roughly 100 transactions have begun on their journey in the IRP. None has completed it. The intent of the law was that six months after the initiation, the IRP should end with either a successful vote for a restructuring plan or the start of the liquidation process. The threat of value destruction in the liquidation would create a focus in the minds of the creditors committee, and thus avoid the delaying tactics that are seen in India today.
The first case, Innoventive Industries, started on 17 January 2017. Six months from that date is 16 July 2017. The delay with which this first case is completed will be an important first milestone for the bankruptcy reform.
In the coming months, a series of important milestones is anticipated:
First case where interim financing is obtained.
First vote by a creditors' committee.
First case to complete IRP with a super-majority in favour of a restructuring plan.
First case where the IP ejects the promoters from the firm
First case to commence liquidation.
First case to complete liquidation.
First individual insolvency to commence and complete.
An analysis of the orders passed by the NCLT (Chatterjee et. al, 2017 (forthcoming)) shows that some of these milestones will come soon. We will soon be talking in the language of cases per month and Rs.Bln. per month, in the aggregate and across categories of defaulters, which will be the output measures of the bankruptcy reform.
The proximate outcome of the bankruptcy process is the recovery rate. This expresses the NPV of recoveries on the date of default as a ratio to the face value of the debt on the date of default. What would give us a high recovery rate?
The lack of delays in the IBC processes;
The extent to which liquidation is avoided for recent defaults;
Competition on the buy side, i.e. the number of prospective buyers; and
The extent that buyers feel there is predictability and certainty in the IBC processes (i.e. the lack of a lemon model problem).
The value of a firm is greater than the liquidation value only when the firm is a going concern. Financial distress disrupts the smooth working of the firm and damages organisational capital. For this reason, it is optimal that the resolution process must commence a day after the first default. However, for the bulk of the cases which are now being brought to the IBC process, the first default is likely to have taken place a long time ago. The recovery rates obtained there are going to be low. This is not a comment on the bankruptcy reform.
We should focus on new defaults that are brought into the IBC. These are the real test of IBC, where there continues to be organisational capital, and there is an opportunity to obtain higher recovery rates. If the bankruptcy reform is successful, new defaults should result in restructuring plans that achieve the super-majority in the creditors committee, obtain a new management team, avoid liquidation, and achieve high recovery rates. An equally important issue for new defaults is the ability of the market to correctly distinguish between firms that are worth rescuing as going concerns versus those that should be put into liquidation.
In liquidation, good outcomes will be rapid sales through which the recovery rate is enhanced, and predictable flow of cash through the IBC waterfall.
Once a certain recovery rate is consistently obtained --- even if it is a low value --- this will bring a new level of confidence to lenders. The first milestone is to gain confidence that we will consistently get a (say) 20% recovery rate. The second milestone is to get the recovery rate up to better values. As these milestones are achieved, lenders will become more comfortable with a broader range of credit risk and maturity. Once many transactions are completed, we will be able to do statistical analysis of delays and recovery rates, differentiated between firms across size categories (small/medium/large) and date of default (recent/old). That will yield report cards of the Indian bankruptcy reform.
A story about organisational capability. Almost everything described here is a story about the capabilities of organisations: of MCA, IBBI, RBI, SEBI, NCLT, DRT, IPAs, IPs, IPAs, IUs, the biggest 2000 companies, distressed debt and private equity funds. A well functioning bankruptcy process involves all these persons skillfully playing their own part. At first, these organisational capabilities do not exist. Nine months after being setup, IBBI has just five senior officers. We need to assess gaps in organisational capabilities, and undertake steps which foster this capability, both in individual organisations as well as jointly across them all.
There is a coordination problem here: organisation $x$ tends to underinvest in building organisational capacity as it sees that the remainder of the ecosystem lacks the requisite capabilities. This hampers the rate of return to its own investments in building organisational capability. If person $i$ were sure that person $j$ was going to invest in building organisational capital, then the marginal product of investment in organisational capital for person $i$ would be higher.
Too often, new organisations in this field are emulating existing organisations. IBBI is slipping into the mores of SEBI and RBI, NCLT is built like a conventional Indian court, IPAs have become like their parents, the quality of drafting law and regulations is slipping closer to conventional Indian standards. This ends up in an environment of low expectations and low organisational capabilities. Instead, we must create a climate of excellence, and work to build high performance organisations.
The lack of data is shaping up as a big barrier. The BLRC design had envisioned a data-rich environment for bankruptcy reform. This included the NCLT issuing structured orders, IBBI building a statistical system, IPAs building supervisory databases and large-scale data capture at IUs. So far, a small part of this has begun. The fog of war is heightened, and all persons are faring poorly on strategy and tactics. The creation of data, and downstream research based on this, that has been done by projects such as Chatterjee et. al. (2017, forthcoming) is a critical element of the bankruptcy reform process.
From uncertainty to risk, and reduction of risk. As with the establishment of all markets, private persons avoid committing resources that are required for building human capital in the field. This leads to a chicken-and-egg situation: lenders are loath to take distressed firms into IRP as the buyers are missing. The buyers are missing as there are not enough available transactions.
The ex-ante legal risk as seen by buyers is considerable. They know the inputs that are required to make the bankruptcy process work. They worry about the legal challenges that may materialise even after they have supposedly got a transaction. They compensate for these risks by low-balling their bids.
As cases move through the IBC processes, uncertainty about the working of the process is reducing for private persons. In time, uncertainty will be replaced by risk: they will understand the contours of the problem and they will build some priors about the process. In due course, more private persons will gain confidence and show up as buyers, thus yielding the ultimate desired outcome: high recovery rates.
Load and load bearing capacity. Big cases present a bigger load upon the fledgling institutions (Shah, 2017). With big cases, private persons have more to lose, and will use every means, fair or foul, to avoid losses. The high powered legal teams that have come together around the Essar Steel default are consistent with this prediction. Their actions, the precedents that they establish, and the way these events reshape the prior distributions of all the players, may end up harming the Indian bankruptcy reforms. The 12 big cases are hotspots for the bankruptcy reform where many things can go wrong. We are likely to obtain particularly weak recovery rates for these, as a big load is juxtaposed against a weak load-bearing capacity.
Success is not assured. There is universal optimism about the Indian bankruptcy reform. We worry that failure has not been ruled out. If the present effort at bankruptcy reform fails, it will be a tremendous loss of confidence in the eyes of creditors, and there will be sustained cynicism on the part of the private sector about future reform efforts. The Indian bond market is an example of persistent failure of reform, leading to endemic cynicism on the part of private persons. Such cynicism leads to reduced investments by private persons in organisational capital, and thus reduces the probability of success in the future.
The Indian bankruptcy reform is work in progress and there are many areas for concern.
A lot of the policy work in the bankruptcy reform has been approached as business as usual. In the Indian policy process, business as usual results in endemic failure. The success stories of the Indian policy process, like the telecom reforms, the equity market reforms, and the New Pension System, did not come from business as usual. The visible difficulties, after the first year of the bankruptcy reforms, call for higher resourcing and improved organisation for the nine areas of inputs, and a shift away from business as usual.
We will never have a perfect law and perfect regulations and a perfect IBBI at the early stages of a reform. What matters most is the intellectual capacity in discerning incipient problems, diagnosing them swiftly and correctly, and coming out with effective solutions. The private sector is ultimately watching the policy apparatus and waiting for this feedback loop to emerge.
Ajay Shah is a researcher at the National Institute for Public Finance and Policy, and Susan Thomas is a researcher at the Indira Gandhi Institute for Development Research. The authors thank Prasanth Regy, Bhargavi Zaveri and Rajeswari Sengupta for extensive comments and discussion on these issues.