Ever since the Global Financial Crisis, India has been trying to come to grips with its twin balance-sheet (TBS) problem—overleveraged corporates and the bad-loan encumbered banks. But a decisive resolution has proved elusive, and the problem has continued to fester. Perhaps, it is time to consider a different approach—a centralised Public Sector Asset Rehabilitation Agency (PARA) that could take charge of the largest, most difficult cases, and make politically tough decisions to reduce debt.
For some years, it seemed possible to regard TBS as a minor problem, which would largely be resolved as the economy’s recovery took hold. But as time has gone on, the problem has only continued to worsen. Earnings of the stressed companies have kept deteriorating, forcing them to borrow more and more to sustain their operations. Since 2007-08, the debts of the top 10 stressed corporate groups have multiplied five times, to more than R7.5 lakh crore. Even with such large infusions of funds, corporates have still had problems servicing their debts, so much so that by September 2016 no less than 12% of the gross advances of public sector banks had turned non-performing. According to some private sector estimates, these numbers are considerably greater.
This situation is beginning to take a toll on the economy. With balance-sheets under such strain, the private corporate sector has been forced to curb its investments, while banks have been reducing credit in real terms. To sustain growth, these trends will need to be reversed. And the only way to reverse these trends is by fixing the underlying balance-sheet problems.
The question is how to do this. So far, the strategy has been to solve the TBS through a decentralised approach, under which banks have been put in charge of the restructuring decisions. A number of such schemes have been put in place by RBI. Most of the time, this is indeed the best strategy. But in the current circumstances effectiveness has proved elusive, as banks have simply been overwhelmed by the size of the problem confronting them.
Eight steps lead to the conclusion that the time may have arrived to try a centralised approach, the PARA. A detailed case is set out in the new Economic Survey 2016-17, Chapter 4.
It’s not just about banks, it’s a lot about companies. So far, public discussion of the bad loan problem has focused on bank capital, as if the main obstacle to resolving TBS was finding the funds needed by the public sector banks. But securing funding is actually the easiest part, as the cost is small relative to the resources the government commands (no more than 2-3% of GDP in a worst case scenario). Far more problematic is finding a way to resolve the bad debts in the first place.
It is an economic problem, not a morality play. Without doubt, the stench of crony capitalism permeates discussions of the twin balance sheet problem. And it is true that there have been cases where debt repayment problems have been caused by diversion of funds. But a vast bulk of the problem has been caused by unexpected changes in the economic environment: timetables, exchange rates, and growth rate assumptions that have gone badly wrong. A persistent narrative of crony capitalism risks leading to punitive rather than incentive-compatible solutions.
The stressed debt is heavily concentrated in large companies. Concentration creates an opportunity, because TBS could be overcome by solving a relatively small number of cases. But it presents an even bigger challenge, because large cases are inherently difficult to resolve.
Many of these companies are unviable at current levels of debt, requiring debt write-downs in many cases. Unviability varies across sectors and companies. But a rough estimate would be that debt reductions of about 50% will often be needed to restore viability.
Banks are finding it difficult to resolve these cases, despite a proliferation of schemes to help them. Among other issues, they face severe coordination problems, since large debtors have many creditors, with different interests. If public sector banks grant large debt reductions, this could attract the attention of the investigative agencies. But converting debt to equity, taking over the companies, and then selling them at a loss—even in transparent auctions—will be politically difficult, as well.
Other mechanisms haven’t worked—and won’t work. Private Asset Reconstruction Companies (ARCs) haven’t proved any more successful than banks in resolving bad debts of companies, and in any case are too small to handle the large cases. Moreover, the incentives facing the ARC-bank relationship can be inherently distorted: for example, ARCs earn management fees for handling bad debts, even if they don’t actually work them out. The new bankruptcy system is not yet fully in place, and even when it is, some considerable time will be needed before it is ready to handle the large cases.
Delay is costly. Since banks can’t resolve the big cases, they have simply refinanced the debtors, effectively “kicking the problems down the road”. But this is costly for the government, because it means the bad debts keep rising, increasing the ultimate recapitalisation bill for the government and the associated political difficulties. And delay is very costly for the economy.
Progress may require a PARA. Such an approach could eliminate most of the obstacles currently plaguing loan resolution. It could solve the coordination problem, since debts would be centralised in one agency; it could be set up with proper incentives by giving it an explicit mandate to maximise recoveries within a defined time period; and it would separate the loan resolution process from concerns about bank capital. For all these reasons, asset rehabilitation agencies have been adopted by many of the countries facing TBS problems, notably the East Asian crisis cases.
How would a PARA actually work? There are many possible variants, but the broad outlines are clear. It would purchase specified loans (for example, those belonging to large, over-indebted infrastructure and steel firms) from banks and then work them out, depending on professional assessments of the value-maximising strategy.
Once the loans are off the books of the public sector banks, the government would recapitalise them, thereby restoring them to financial health and allowing them to shift their resources—financial and human—back toward the critical task of making new loans. Similarly, once the financial viability of the over-indebted enterprises is restored, they will be able to focus on their operations, rather than their finances. And they will finally be able to consider new investments.
Of course, all of this will come at a price, namely accepting and paying for the losses. But this cost is inevitable. Loans have already been made, losses have already occurred, and because state banks are the major creditors, the bulk of the burden will necessarily fall on the government (though the shareholders in the stressed enterprises will need to lose their equity as well). In other words, the issue for any resolution strategy—PARA or decentralised—is not whether the government should assume any new liability. Rather, it is how to minimise a liability that has already been incurred by resolving the bad loan problem as quickly and effectively as possible. And that is precisely what creation of the PARA would aim to do.
That said, the capital requirements would nonetheless be large. Part of the funding would need to come from government issues of securities. Part could come from the capital markets, if stakes in the public sector banks were sold or the PARA were structured in a way that would encourage the private sector to take up an equity share. A third source of capital could be RBI. The Reserve Bank would (in effect) transfer some of the government securities it is currently holding to public sector banks and PARA. As a result, RBI’s capital would decrease, while that of the banks and PARA would increase. There would be no implications for monetary policy, since no new money would be created.
Creating a PARA is not without its own difficulties and risks; the country’s history is not favourable to public sector endeavours. Yet equally one has to ask how long India should continue with the current decentralised approach, which has still not produced the desired results eight years after the Global Financial Crisis, even as East Asian countries were able to resolve their much larger TBS problems within two years. One reason, of course, was that the East Asian countries were under much more pressure: they were in crisis, whereas India has continued to grow rapidly. But an important reason was that it deployed a centralised strategy, which allowed debt problems to be worked out quickly using the vehicle of public asset rehabilitation companies.
In sum, current efforts have not been successful in addressing the twin balance sheet problem. New solutions must be tried. Perhaps it is time for India to consider a PARA as one such solution.
The authors, Joshua Felman, Rangeet Ghosh, Syed Zubair Noqvi & Arvind Subramanian, worked on this year’s Economic Survey