What is Wrong in the Risk Management Process in Indian Banks?

The hard truth is that every loan has both a borrower and a lender. If the loan is inherently bad, the lender is as much at fault as the borrower.

The literature on managing the “NPA” problems of Indian PSU Banks keeps growing, but the “NPA” problem keeps growing even faster! Somehow, the weak, archaic, and anachronistic conceptual framework under which most commercial bank lending is done in India, and its associated processes, rarely find mention - other than exhortations that banks should improve their risk management systems.

In a manufacturing organisation to have low levels of rejects not only should the technology and design of manufacturing process be in tune with the product being manufactured, but also the quality control process should monitor the level of bad parts being produced. Accordingly, indications of larger than permitted number of rejects suggests that entire manufacturing process needs to be re-tolled and / or fine-tuned.

For banks to have low levels of NPAs, their entire lending process from initial appraisal, to detailed due diligence, to monitoring, to detection of incipient sickness, to recovery has to be under control. The key to this is an appropriate conceptual framework for lending, ie, the technology of lending. Recurrent high NPA levels clearly indicates that debt capacities are being overestimated by the appraisal process and the relevant monitoring parameters are not doing their job. The responsibility for rectifying this rests squarely on the lender.

Along with this there should be low tolerance for NPAs. Banks are highly leveraged entities and start making operating losses if gross NPA levels are more than 4-5%. The ground level situation is that gross NPA levels of PSU Banks was as high as 24.8% for the year ended March 1998, before slowly falling down to a low of 2% by March 2009 before bouncing back to a high of 14.8% by the end of March 2018, suggesting high tolerance to NPAs.

That the conceptual framework for commercial bank lending is a major impediment in making good quality lending is reflected not only in recurring bouts of high NPAs, but inter alia, in constant complaints by industry that credit decision making in banks is slow and cumbersome, getting credit, especially by the MSME sector is extremely difficult, involves humongous amounts of paperwork, the extremely long appraisal notes, the low level of credit as a percentage of GDP, the over-emphasis on availability of tangible collateral security, the preference of making investments in SLR securities than in making loans and advances etc. etc.

Let me try and substantiate my contention.

Modern banking in India evolved out of Calcutta Agency Houses which were trading firms which undertook banking operations for the benefit of their constituents (Tannan’s Banking Law and Practice in India). Naturally, the methodology which developed was geared for lending for trading in commodities. This is a natural outcome, since the evolution of capital markets through the introduction and diffusion of financial innovations is largely dependent on occupational specialisation of financial intermediaries (V V Bhatt EPW May 1987).

The credit risk in such cases was substantially covered by the cash credit system and its associated practices of stock statements, its verification and calculation of drawing power. Moral hazard being eliminated by having goods under lock and key of the banker, and by periodically inspecting stocks to ensure that they confirmed in terms of specified quantity and quality. Adverse selection was controlled by keeping a margin over market value of security. This was easy as long as valuation of the underlying assets was straightforward and simple.

This credit risk management framework starts breaking down if it is extended for lending for industry. First, the dichotomy between working capital and term finance is artificial since most borrowers need both to continue as a going concern, and only a going-concern can service its debt.

Second, in a manufacturing concern it is difficult to identify and easily value the various components of current assets. The more complex the manufacturing process, the more difficult it is to value the underlying security. This system breaks down completely while evaluating the borrowing capacity of borrowers operating in the services sector – where most of the earning assets are intangible. Currently the services sector contributes more than 55% of the country’s GDP while its borrowing levels have been hovering around 30% of bank credit over the last few decades – and most of the exposure is predicated against tangible security.

Third, the holy methods of lending laid down by Tandon, Chore, Nayak Committee et al, and still being followed in various avatars are essentially credit rationing devices. The background for setting up the Tandon Committee (1974) was the need to curb the use of bank credit for hoarding of commodities in short supply. The mandate given to them was for framing guidelines for commercial banks for follow-up and supervision of bank credit for ensuring proper end use of funds.

Similarly, the mandate of the Chore Committee (1979) was to review the operation of the cash credit system of lending, particularly with reference to the gap between the sanctioned credit limits and the extent of their utilisation.

These methods of lending were never designed or meant to be credit risk management tools. The fact that credit rationing implied some kind of credit risk control is only incidental. Debt repayment capacity under methods of lending is predicated on underlying security rather than debt servicing capacity. This leads to the primacy of the Current Ratio, while leaving estimation of leverage and interest coverage ratios or designing processes for entrapment of cash flows, as after-thoughts. The possibility and need for amortisation of working capital borrowings, including the ubiquitous large permanent cash credit component is wholly missing. 

It also leads to much avoidable confusion on the concept of leverage and its calculation. For some it is Total Term Liabilities / Tangible Net Worth (TTL/TNW) – which was used typically by term lenders. For “working capital” bankers, it is generally calculated as Total Borrowings (Term plus Working Capital) / Tangible Net Worth. Other use the concept of Total Outside Borrowings (say including Trade Credit) / TNW.

If leverage is to be considered as a measure of extent of (a) Moral Hazard (the extent of the borrower’s stake in the business as a proportion of total capital employed by the business) and (b) Debt Servicing Capacity, it is the third measure which is more appropriate for banks. In fact, TOL should include contingent financial liabilities (say, Financial Guarantees, Outstanding LCs, Bills discounted etc), ie everything which is taken at more than 0% risk weight in calculation of CRAR.

Fourth, methods of lending, such as MPBF, creates entitlements on the amount of borrowing. What is lost sight of is the kind of risk involved. In this process, the credit exposure comprises both equity as well as debt risk, while the pricing is wholly related to debt risk. It is natural that the overall portfolio is sub-optimally priced - result NPAs. The fact that at times bank finance substitutes equity is acknowledged by various authorities (e.g., Dr. C Rangarajan in the TTK Memorial Lecture (11/11/1997; Dr. Raghuram Rajan, Note to Parliamentary Estimates Committee on Bank NPAs dated 06/09/2018;).

The comfort that entitlements of the amount of borrowing is covered by security (current or fixed assets as the case may be) is misplaced. No bank can run its business on foreclosure of security as the primary means of recovery, especially in India with its slow and cranky legal processes. The transaction costs would be too high, especially since banks survive on high leverage with thin operating margins. The predominant consideration of tangible security in lending, not only limits the volume of good loans which can be made but also low recovery in case the loans go bad.

The primary value the banker brings to the societal table is ability to evaluate and manage all kinds of risks in the lending process. As such, the risk management processes in banks should be able to handle all kinds of risk, including credit risk. This realisation is also conspicuous by its absence. The view that the latest round of high NPAs resulted inter alia from falling commodity prices and exuberance in lending for infrastructure is a reflection of lack of skills in handling such risks. Classifying it as a reason for creation of NPAs from this perspective makes it sound more like an excuse.

An oft heard reason for high levels of NPAs in bank exposure to the infrastructure sector is that such lending created large asset liability mismatches. But then this should have first led banks to breach their asset/liability prudential norms. There is no mention of this ever having happened.

The main factor for non-evolution of viable and cogent lending mechanisms is the complete break-down in governance in PSU banks, essentially due to the extremely corrosive influence of the DFS in their functioning. While the mandarins in DFS exercise large control rents they have effectively no equity stake. The pathological effects of such control have been documented time and again starting with the Note by Prof M Datta Chaudhury and Shri M R Shroff in the report of CFS (Narasimham I), to the Nayak Committee Recommendations, and by Viral Acharya and Raghuram Rajan in their paper, Indian Banks: A Time to Reform, wherein they have strongly recommended for abolition of DFS. The end result of this hegemony of DFS is the slow but steady deterioration in the levels of professionalism of our banks with all its negative consequences, including non-evolution of workable processes to handle credit risk. 

A pernicious aspect of the control by DOFS is reflected in their interference in transfers, postings, and appointments, especially senior level positions.  There is no reason that DOFS had to issue a clarification wide Office Memo dated 13/01/2015 to the effect that, “Each Bank/FI should have their own objective, well laid out transfer and posting rules which should be followed strictly. No exception, should be made in such rules at the behest of any recommendation given by anyone including anybody from the Ministry of Finance.”

The poor governance of PSU banks is reflected not only in the recurrence of high NPA levels but also other aspects of their functioning – witness the steady decline in their market share in all aspects over the last 30 years. To give just one example, the complete neglect of the importance of providing safe, convenient, remunerative deposit services. The opportunity offered by the PMJDY has been all but squandered away.

Most policy prescriptions on managing NPAs seem to be focused only on timely recognition of NPAs and actions to be taken after the loan has gone bad or at the most likely to go bad. There is very little discussion on how to make good loans. After all, prevention is better than cure.

Was it Einstein who said that "It is insanity doing the same thing over and over again, and expecting different results"?