How to build capacity at RBI


Business Standard, 05 March 2018


With the NPA crisis and the PNB crisis, all eyes are on RBI reform. We should not engage in a witch-hunt and we should not fight the last war. The place to focus on is the rules of the game that make RBI. These rules need to be modified so as to create the institutionalised application of mind. State capacity at RBI will come from focus and from good governance procedures.

Micro prudential regulation

Banks make a promise to depositors that money will be returned with interest. The purpose of micro-prudential regulation is to obtain a high probability that this promise will, indeed, be upheld. Bank regulation pushes towards this by writing regulations and by enforcing them. The PNB crisis, and the vastly larger NPA crisis, have demonstrated that RBI's oversight of banks is flawed. With trillions of rupees of public money at stake, there is an urgent need for RBI reforms.

As an example of what banking regulation needs to do, consider the problem of NPAs. Regulations must be written which require banks to rapidly mark down the value of a defaulted loan to zero. These losses should immediately appear in the income and expenditure statement. Later on, recoveries through the bankruptcy code will directly travel into equity capital. These regulations should be accompanied by rules about adequacy of equity capital, that limit the leverage of banks.

Once these regulations are in place, we have to think about compliance. The first level issue is that of a bank that explicitly says to RBI that it is in violation with the rules (either on recognition of NPAs or on adequacy of equity capital). The bank must be forced to bring in more equity capital or reduce deposits.

Far more dangerous is a bank that lies; e.g. a bank that claims to not have bad assets while it actually does. An audit process run by the regulator must test the veracity of claims made by the bank. This would work at two levels. First, the regulator would examine all steps of the audit process that was run by the board of the bank. Second, the regulator would mistrust these claims, and conduct/commission studies on its own. For example, the regulator may take a statistical sample of 100 loan accounts in a bank, and test the extent to which the internal valuation of these assets is consistent with their market value. Using this, it would get independent feedback about what is going on inside the bank, over and beyond the audit process of the bank.

In India, we know that both steps -- writing regulations and conducting examinations -- have had weaknesses. RBI is a remarkable Indian government institution on the question of corruption. Why has a team, with a remarkably strong ethos on the question of corruption, failed on its core objectives? There are two lines of thought in RBI reform which are wrong. We should not engage in a witch-hunt, and we should not fight the last war. The right question to ask in RBI reforms is: What are the processes, and the checks-and-balances, through which RBI will continuously refine the text of the regulations and the process manuals for enforcement, so as to achieve high State capacity?

Regulation-making process

Let's start at the drafting of regulations. At present, RBI staff has arbitrary power in rushing regulations to the RBI website. This is done without input from experts and practitioners, and without oversight by the board. This arbitrary power yields low State capacity: We routinely get the wrong regulations.

What is a sound process for writing regulations? Regulation-making projects should be commissioned by the board. The problem that is sought to be solved should be clearly articulated along with a demonstration that market failure is present. The lowest cost intervention should be designed, and it should be demonstrated that the proposed intervention addresses the problem that is sought to be solved. The regulation should be drafted. These documents should be put up for public comment. The staff must modify the regulation, utilising the comments that were received. This packet should go back to the board for one last debate, and only the board should be able to authorise the release of a new regulation.

This process forces the application of mind. It will force the staff to work harder, to rise to higher levels of intellectual capability, and produce better regulations. This is how we get to State capacity, to an RBI which does not make mistakes on the drafting of regulations.

Executive process

Similar issues apply on the executive branch. At present, RBI staff has arbitrary power in how supervision is done, which generates low State capacity. The path to State capacity lies in having well-designed formal processes. The board must lead the process of designing the enforcement process. Formal process manuals must be drafted, put out for public comment, debated, and then authorised by the board. As an example, the US Federal Reserve website shows us full manuals about how supervision is done, but the RBI website does not. The organisation must be reshaped to live by these manuals.

Quasi-judicial process

When violations are detected, punishments are triggered. At present, RBI staff has arbitrary power in secretively awarding punishments . Arbitrary power yields low State capacity. SEBI has a better arrangement. The accused must get a hearing in front of a neutral party, and a reasoned order should come out on the website, which awards punishment. The accused can appeal the order at SAT. The rigour of this process will force the executive branch to rise to higher levels of capability, as was the case at SEBI when SAT was created.

Focus

Finally, there is the question of focus. RBI suffers from a sprawling mandate. Limited capacity is being frittered across too many objectives. The scarcest resource is the time and attention of the board and the top management. Focus will yield enhanced State capacity. RBI's objectives should be narrowed down to just two things: sound money (i.e. inflation targeting) and sound banking.


Back up to Ajay Shah's 2018 media page
Back up to Ajay Shah's home page