My last week’s column, “What ails asset reconstruction firms?” evoked strong reactions from various quarters. While some bankers feel I was unnecessarily harsh on banks and did not understand the “spirit” of securitization (whatever that means), a few others are not surprised with my ‘findings’ as the practice of financial incest has been rampant since the inception of the industry. The column was the first of a two-part series on the subject that I had planned. Before writing the concluding part, I have discussed the issues with four senior professionals of the Indian asset reconstruction industry, including S. Khasnobis, former managing director and chief executive of Asset Reconstruction Co. (India) Ltd (Arcil), the country’s oldest and biggest asset reconstruction firm.
In the first quarter of fiscal 2012 ended June, the non-performing assets, or NPAs, of 11 banks with maximum stressed loans rose by Rs15,425 crore. Since fiscal 2010, these banks have added almost Rs83,000 crore worth of bad assets. With interest rates rising and the economy slowing, more and more corporations will default on bank loans. But not too many banks are selling their bad assets to asset reconstruction companies, or ARCs. Why? Many of them are restructuring stressed assets by giving borrowers more time to repay, while a few are disbursing fresh funds to the stressed accounts to pay up the bad loans.
Also Read | Tamal Bandyopadhyay’s earlier columns
ARCs can buy the bad assets from banks by paying cash or offering security receipts (SRs) that get redeemed a few years later. Typically, banks look for higher valuations while accepting SRs against bad assets, and discounts are steeper for cash deals. Overall, banks are not too willing to sell bad assets to ARCs and there are many reasons behind that. When a bank parts with an asset to an ARC, it has to set aside money or make full provision between the book value of the loan and the value at which it is sold to an ARC. This impacts the bank’s profitability. Typically, banks make full provision for a bad asset over three to four years. This means they can postpone the impact on their profits by carrying the bad assets on their books for a few years and sell them to ARCs as a last resort of recovery when no more fresh provisions are required.
Besides, banks enjoy the same powers that an ARC enjoys even though, operationally, they may lack the expertise to recover bad loans. So most banks attempt to recover a bad loan or rehabilitate the stressed borrower on their own and when they fail to do so, sell the bad assets to ARCs.
Illustration by Jayachandran/Mint
Finally, there is always a gap between the value that banks expect from bad assets and the price ARCs are willing to offer. The driving factor here is the cost of funds. While banks discount the future expected realizations from recovery at about 10% a year, ARCs insist on at least 20% discount. This means over three years, while bank wants to recover Rs70 from an asset worth Rs100, ARCs quote a price of around Rs40. Besides, banks need to pay them a management fee that varies between 1% and 2% of the size of the asset in case of sale through SRs. Most banks find such a hefty discount unacceptable; they also do not realize that the longer they delay the transfer of assets to ARCs, the faster is the fall in the final realizable value.
Banks auction bad assets to ARCs, but do not offer any floor price for such auctions. Often, after receiving price quotes from ARCs, they withdraw the assets from auction and start negotiating with the borrower for a settlement, using the highest bid as a floor. The auction of bad loans for many banks is not a part of the process of selling such loans, but a price discovery method to bargain hard with the defaulter for recovery.
While banks are allowed to take profit from the sale of bad assets (if the sale price consideration is higher than book value), in case of a settlement with the defaulter borrower, such profit from the sale of assets to ARCs cannot be booked even if it is paid in cash. They money thus generated needs to be kept as a cushion against future provisions. This is also a disincentive for banks to sell NPAs to ARCs.
Let’s look at the regulatory aspects of the industry. The Reserve Bank of India (RBI) issued the final guidelines for ARCs in April 2003 after the Parliament passed the Securitisation and Reconstruction of Financial Assets and Enforcement of Security Interest Act, 2002, and subsequently several changes were made. The norms allow the acquisition of financial assets both on the books of ARCs as well as under a trust structure outlined in the Act. There are 14 ARCs in India and about 98% of the assets that they have acquired are through the trust structure. This means ARCs do not own the assets but they manage the assets of the trust. In that sense, they cannot have NPAs on their books; but RBI norms insist that when they are not able to recover the bad assets in accordance with the plan envisaged, they need to classify them as NPAs and this means they cannot earn their management fee on such assets.
Incidentally, SRs are subject to declaration of net asset value (NAV) every six months, based on ratings by ratings agencies; and SR investors (including ARCs for their own investments) are required to book losses in case of a drop in NAV.
The capital requirement for ARCs is also an issue on which the industry is divided. Going by RBI norms, an ARC needs to follow 15% capital adequacy till it has Rs100 crore capital. This means that for every Rs100 worth of bad loan bought, they need to have Rs15 capital. This norm is relaxed once an ARC has Rs100 crore capital, but all ARCs must pick up at least a 5% stake in the SRs that they sell against the bad assets.
Many say that ARCs should have more capital and invest more in SRs as they will be more aggressive and diligent in recovery when they have more skin in the game. Often when they buy bad assets and offer SRs, the valuation is too high and ARCs strike deals knowing fully well the SRs will not fetch such a high price at redemption and seller banks will lose. They will not do so if they hold a larger part of the SRs. But there is a strong opposite view too: since ARCs are managing the trusts, why do they need to have hefty capital? Also, why do they need to invest in SRs? After all, they are playing the role of asset management companies (managing bad assets); and the mutual funds that follow the same principle do not require a big capital base and they do not need to make own investments.
ARCs are being created to clean up the banking system and prevent capital infusion in banks (as banks need to set aside money for bad assets, they need more capital when bad assets grow), and if ARCs themselves need hefty capital, the purpose of their creation is not well served.
Instead, RBI needs to broaden the investor base in SRs. Currently, qualified institutional buyers, or QIBs, such as banks and insurance firms, are allowed to subscribe to SRs which are rated instruments. Foreign investment in such instruments is capped at 49% and no individual foreign institutional investor, or FII, can hold more than 10%. Perhaps, the regulator feels that a larger role for foreign investors will encourage them to take over sick Indian firms through the back door. But such apprehensions have no basis as most defaulters are in bad shape and they cannot attract serious investors. Foreign investors should be allowed to play a larger role in investing in SRs floated by the trust and encouraged to get actively involved in the recovery process, the way a private equity fund handholds the promoters of a firm in which it invests. Foreign distressed-debt investors are specialized institutions who like to have a significant stake in a trust or scheme with some control.
Typically, ARCs follow three models of recovery. The first is the asset stripping or the vulture model. In this case, they shut the unit’s functioning and strip the assets to recover money.
The second model is the arbitrage model. In this case, ARCs add very little value; they acquire an asset from a bank at a price and go back to the same borrower and settle at a higher price, creating a spread by virtue of their superior negotiating skills with the borrower. This also explains why banks use ARCs as a price discovery platform and then go back to settle with the borrower themselves using the ARC-quoted price as a floor.
The last model is the revival model that involves the restructuring of financials, processes and the infusion of long-term funds. The mere acquisition of an asset doesn’t revive an account. Apart from arranging funds, ARCs should also be allowed to take equity exposure in a sick company by converting a part of the debt for speedy recovery. While dealing with a listed entity, any such exposure will attract the so-called takeover code of the market regulator that makes an open offer mandatory for any acquisition of a 25% stake or more in a company; for unlisted entities, it can be a contract between the company and ARC.
The Securitisation Act allowed ARCs to change or take over the management and sale, or lease, of the business of the defaulting borrowers; but RBI took seven years to actually empower them, that too in a truncated manner. They can take over the management, but cannot lease the business as yet.
There have been very few cases where the business has been taken over, but the powers to do so act as a threat and make the recovery process relatively easier.
The rogue borrowers always want to oppose any recovery move; there have been thousands of cases in which they have dragged ARCs to court to delay the process.
Technically, the borrowers are required to offer one-fourth of the dues to legally challenge any recovery move, but this norm is not always followed. Besides, ARCs are allowed to acquire only secured NPAs from the Indian banking system, leaving the other debt-holders to proceed under civil court procedure.
There are many other issues that RBI should look into to make the asset reconstruction industry work well, such as allowing the transfer of assets among ARCs and permitting them to offer working capital support to industrial units under revival; but no model will work unless the banks themselves appreciate the importance of selling bad assets and stay away from financial incest.
The concept of securitization has not taken off as banks that sell their bad assets to ARCs often insist that these cannot be mixed to create a pool.
This means ARCs need to hold bad assets of individual banks as separate pools under a trust, and the banks are subscribing to SRs of their own assets. In other words, the banks are simply removing the bad assets from their loan books and bringing them back as good assets through their investment books (that hold SRs). Arcil, I am told, has set up at least 350 trusts, and many of them are seller-specific trusts. This practice might prove to be the proverbial last nail in the coffin of India’s asset reconstruction industry if it dies a premature death.
There is a clear conflict of interest as the banks are holding the dual role of owners as well as beneficiaries—sellers as well as investors in SRs. They are also the major shareholders in some ARCs. One way of reconstructing ARCs could be by capping sponsor banks’ exposures at 10% and keeping the nominees of the sponsors out of the acquisition and resolutions committees of the firms.
The representation of banks as a class of shareholder in the board of ARCs should also be capped. This is not a unique idea as in credit information bureaux, too, no bank is allowed to hold more than a 10% stake. This, along with an expansion in the investor base in SRs will help ARCs securitise bad loans in the true sense of the term. Let the banks focus on their main business of lending, and ARCs on recovery.
And till such time the industry fully understands the nuances of bad asset buys and recovery, no new ARC should be allowed to set shop.
Tamal Bandyopadhyay keeps a close eye on all things banking from his perch as Mint’s deputy managing editor in Mumbai. Your comments are welcome at firstname.lastname@example.org