The RBI's Financial Stability Report estimates the gross non-performing assets of the country's banks to reach 13.5% by September 2021. It comes amidst discussions about a bad bank. The Business Standard has an editorial which dwells on this topic.
Fundamentally a bad bank idea has three parts:
1. Establishment of a bad bank and freeing up of the resultant good banks to resume normal lending activity.
2. Transfer of bad assets from all the banks to the loan book of the good bank.
3. Resolution of the bad assets held by the bad bank and maximise their recovery.
While hardly trivial, the first two steps are nevertheless only administrative exercises. Here too things like valuations while transferring assets can be tricky. The banks will have the incentive to maximise their valuations. But given these are G2G transfers, this is a problem that can be surmounted.
But it is the third step that is where the real challenge lies. And without significant progress on the third step, the first two steps are costly diversions. The net cost on the system (markets and the government) would remain the same, perhaps become even higher.
The question that then needs to be asked is this. What's the binding constraint on the resolution of bad assets? Are the bad assets amenable to the standard resolution and liquidation processes? Is it lack of specialised expertise and adequate management time, that the regular banks could not have done it on their own?
I believe that a bad bank solution's popular appeal comes from its proximate effect of being able to separate the good and bad assets, free up the balance sheets of the regular banks, and therefore restore normalcy in the credit markets. This is illusory and like sweeping things under the carpet or kicking the can down the road. It overlooks the reality of having to resolve the assets transferred to the bad bank at some fair enough price, so as to complete the process.
Most commentators, including eminent academic researchers, are carried away by the proximate steps and first-order effects in their prescriptions to complex public policy challenges. Further, as this BIS paper finds, bad banks also require a complementary set of steps for success
The main finding is that bad banks are effective in cleaning up balance sheets and promoting bank lending only if asset segregation is combined with recapitalisation of the bank's balance sheet. Used in isolation, neither tool will suffice to spur lending and reduce future NPLs. Looking at a wide range of episodes, we find that assets segregation is more effective when (i) asset purchases are funded privately; (ii) smaller shares of the originating bank's assets are segregated; and (iii) asset segregation occurs in countries with more efficient legal systems.
The Indian context may struggle to meet many of these requirements.
The resolution of bad assets has several dimensions. As I have blogged here, the biggest issue in India, unlike the western countries where bad bank ideas have been tried, is that the banking system here is largely publicly owned. This has at least four implications.
One, given their size, the large inevitable haircuts to be taken during the resolution are fiscal costs to be borne by the government. This by itself is a significant issue. Second, the officials in the resolving bad bank face the acute problem of stifling and paralysing external oversight on their actions from investigative and prosecuting agencies and the judiciary. The judgements of all these agencies are most often based on limited understanding of the context and the business activity as well as deeply hindsight biased.
Then there is the issue of resolvability of these assets. As I have blogged earlier, the resolvability of largely public good infrastructure and related physical assets, both operational and at different stages of completion, is qualitatively different when compared to the resolution of standard mortgages, consumer retail and credit card loans, corporate loans, student loans etc. The bad banks in the west were deployed to resolve largely the former category of assets. The former requires a more heterodox approach tor resolution.
Finally, there is the issue of the market's ability to absorb these volumes of bad assets in one flush. The hyper-ventilation by commentators on resolution overlooks the limited depth of local capital and relatively very small foreign capital interested in such assets at fair prices. There would obviously be willing buyers at exorbitant haircuts. Who would not want assets at a steal? But would the political economy and the fiscal burden allow for such resolutions at scale?
In view of all the aforementioned, with bad bank we would not have moved much on the issue of putting to bed the problem of bad loans. We would only have transferred the problem from one finger of the government to another. Besides it would have come at a high cost in terms of the bandwidth it would occupy among all stakeholders concerned, diverting effort within banks from pursuing their main activity for at least six months.
The solution instead is to retain the assets on the bank's balance sheets and resolve them through a heterodox approach as outlined here, here and here that depends on the nature of the impaired assets.
One, continue the ARC and leveraged buyout fund routes. Two, encourage purchases by private (infrastructure funds) and public (like NIIF, IIFCL etc) funds which could either securitize some assets or establish SPVs and manage these assets till they become profitable. Three, strategic acquisitions of some of the infrastructure assets, especially in power, by the better governed public sector units. Four, strategic auction of certain other assets, especially in steel and metals, to reputed private buyers. In at least some of these cases, provisions like back-ended clawback of some share of windfall gains by potential buyers may ease the resistance and apprehensions associated with such sales.
Even this is not going to be easy. But it stands a far greater likelihood of success than the illusion of bad bank.
No comments:
Post a Comment