Raghuram Rajan’s Message To India: Fix Structural Issues In Banking, Don’t Take Shortcuts

Raghuram Rajan: Both. At the end of that licensing process when we set out the terms for longer term on-tap licensing we had indicated from the RBI that, perhaps, it would be wise not to let full-fledged corporates that is, industrial houses whose primary business was industry or non-finance activities, into banking. We had left some room for houses that were primarily financial companies to enter banking. So, it’s not as if all industrial houses are prohibited, it’s just those that are primarily outside of the financial business.

We looked at the experience in other countries as well as our own. The temptation for industrial houses to lend to themselves, what is called self-lending, is significant and, of course, this happens more in bad times when the industrial house is in trouble. That’s really a time when the bank should be exercising tremendous caution on who it’s lending to and how much it’s lending. If you’re owned by a borrower, it’s very hard to make that distinction. Previous episodes of allowing this—in the U.S. this is called banking and commerce, but across countries where there’s been banking and commerce, there’s been pretty severe crises. That does not mean that there are no good industrial houses that will stay independent but the conflicts of interest are significant.

So that’s one reason why this has been prohibited in many countries and also in India. The second reason is that it creates an undue concentration of economic power. People put their money in banks, they think it’s safe. If you own a bank and you need the money, well, it’s very easy to get the money. As an industrial house you can grow significantly based on owning a bank and that makes you enormously powerful—economically and also to some extent politically. This is one other reason why banking and commerce has been frowned upon in other countries. Let’s keep the financial side separate from the industrial side, that allows for more competition on either side and eliminates the conflicts of interest.

So, for sound reasons the RBI historically has not permitted it. There was a short episode when the first banking licenses were given out, when in fact the rules had been changed. Based on the experience during that time, we decided there were plenty of good possibilities outside the industrial house sector for new banking licenses and there was no need for the amendment so we changed the rule back to what it was before.

Well, typically, it hasn’t been allowed in the United States. There are stray instances of small entities owning a bank, sometimes at arm’s length, but in general you haven’t had the mega conglomerates which combine industrial houses along with strong banks. In places like Japan, we had that we had the Zaibatsu, which were basically industrial combines around the house bank, and there was a belief that it was the strong Zaibatsus which led Japan into World War II.

The concentration of economic power also meant a concentration of political power and it led to dictatorial tendencies that drove Japan into war. In fact, post-World War, the U.S. administration focused on breaking up the Zaibatsus. Similarly, in Germany, the post-World War II, U.S. administration broke up the large combines and arguably that helped the German democracy post-World War II. So, this is an old debate, self-lending is part of the issue and it’s important in India where the ability to regulate and supervise is still a work in progress.

Apart from the financial risks and stability issues, is the economic concentration issues. In India as we have some industrial houses gaining tremendous economic power, it’s an issue which has to constantly be examined and it’s one reason why we had this rule on the table.

You’ve seen the difficulty in India in determining connectedness. Who is connected to whom? We know, in the past, of entities supposedly bringing in equity into a venture but, in fact that equity has been borrowed by somebody else within the group.

So, when things are still quite murky, when we have only modest confidence in accounting, when we are working on making sure supervision is stronger and we have minute by minute evidence of where the money is flowing....When we don’t have any of this really in place, simply laws will not make a difference.

You can exhort to all you want and criminalise some of these transactions but they still take place. The reality is that, until we have much greater confidence, and even then, what is the reason you need this, right? If you want more talent in banking from the private sector, you can always create more banks but they can be opened by professionals, they can be opened by financial houses, they can be opened by industrial houses that are largely in finance and there are a number of those.

There’s plenty of opportunity you haven’t exploited yet. So, the question is, why do you need to get the license now? Who is there knocking on the door at this point, who says, I want a license? Are you doing it only for a few or is it because this is in the larger interests of the country over the longer term?

Yes, but I would argue that if you look at credit growth, it’s fallen tremendously amongst the public sector banks. So, if you want to fix that why not look at your public sector banks and ask what’s gone wrong? What is it that can be remedied here in order to increase credit growth? Why do you have to bring in a whole new set of banks to get that growth?

The problem sometimes lies in thinking that there are shortcuts to the longer run issues of how do I improve credit growth, how do I improve credit information, how do I improve credit recovery? We’ve had this reform of the IBC. Now the IBC is suspended and even while it was working it was slowly increasing the delays in recovering the money. So, I mean, think about the framework and start fixing the framework. That’s tough. That requires more work. It’s not a shortcut. Shortcuts could get you into even bigger problems than you are in today.

It is true, India’s credit to GDP is very low but India’s NPA to credit is very high. How with such low credit to GDP, do you have such high NPAs? Ask that question before you ask a conflicted set of bank owners to come in and possibly add to the load of NPAs.

I would say shortcuts are the bane of our system. Let’s take the long path to figure out what’s going wrong and fix it and that requires hard work and that work can be done. We’ve done it before. We need to do it now.

The whole point of nationalisation was in some sense to expand the access to credit. We can debate till the cows come home whether that was successful or not, but we do have to ask was that a fantastic system beforehand. In the pre-independence days when you had the management houses and so on, we had fairly close networks within which credit circulated and that didn’t do wonders for everybody else.

I am willing to believe there are some corporates who are going to be the next Ant Financials and create a tremendous amount of new lending etc. But that’s when we also have to ask, are these the same corporates who are huge in the country already and can we afford the kind of concentration of economic and political power that means? I mean, I think that we need to ask all these questions before we say this is the one answer.

If we want to do a good job on privatising the public sector firms, Viral [Acharya] and I wrote a piece on this, and our point is really, you have to improve governance. There is no shortcut to that. You can’t give it to somebody else and say I’ll give you a full fledged bank with all the access to funds and you fix it. Because it could be fixed in very bad ways, especially if you have a conflicted owner. Give it to somebody who already has huge debts and they will find a way to make those debts paid down or borrow from the banks to pay down those debts and even expand.

So, our controls on some of these large industrial houses and what they do is limited and, in that scenario, this is playing with fire.

I think you have to be careful. Some of the old private banks, which are very narrowly focused in terms of the holdings and often are focused on specific communities, there’s a long history of those governance problems. Some of them have made the transition into much more modern banks and have done very well. Banks like Federal Bank and so on. So, I think it is possible to transform them but they’re really quite small and therefore not a huge issue. Over time, the ownership has changed and some of them have improved.

On the large private banks especially the widely held private banks, yes, there have been some very well publicised incidents of mis-governance in a couple of them but if you look at the size of the NPAs, the size of the problems in those banks, they are relatively limited compared to the size of the problems of the public sector banks. So, I think in terms of magnitude of the mis-lending or the poor lending, the public sector banks are much higher.

It’s not just the infrastructure sector. If you look at this small and medium sector lending, that’s where a lot of problems are starting to emerge and the size of the NPAs in that lending [needs to be watched]. Some of these loans are quite recent and when you make recent loans it takes time for them to go bad. Now the size of the NPAs in MSME lending is increasing significantly in public sector banks. So, I would say the number one governance problem is in the public sector banks.

Clearly, we have to improve governance across the board but I think it would be wrong to characterise the size of the governance problems in the private banks as equivalent to the size of the problems in the public sector banks.

When we thought of payments banks, these license were meant as a way to get payments and inclusion done by potentially industrial houses, including some telecom houses but with no access to credit. Everything that they collected had to be put either in deposits at the scheduled banks or in government securities. So that was a way of ensuring they would be safe and they would focus on payments. If they wanted to make loans they could tie up with a bank and actually some of them applied for alliances jointly with banks. I haven’t kept track of how that is progressing. So that was the idea behind payments banks.

Now sometime later in 2018 or 2019 they said, okay payment banks, after five years we will consider you for a banking license. However, I think that meant only for payments banks that were run by financial companies. The idea was okay we’ve seen how you handled money. We have more confidence, now we think you can possibly start lending. Both payments banks and small finance banks were thought of as ‘proving grounds’ for entities to learn the business of banking to show they were good and after they showed they were good, they could migrate to a banking license if they were thought fit and proper at that time.

So, all this was only for financial sector companies not industrial houses. The assumption that industrial houses are now eligible if they hold a payment bank license, I think is wrong. They would still be prohibited if they are industrial houses from migrating from a payment bank to a full-fledged universal bank or a small finance bank. However, with this amendment, if it is approved by the RBI, then there will be nothing standing in the way of getting a payment bank license, spending three years, then applying for a universal bank license.

NBFC conversion has the same problem. The whole advantage of banks comes from offering deposits because the public thinks deposits are safe no matter who offers them. Finance companies, NBFCs can’t offer deposits, not the short-term deposits that banks can. So, they are distinguished from the banks. So, industrial houses can operate NBFCs but when it comes to deposit taking, we should put a barrier because otherwise it becomes too easy to raise the money and then you become too-big-to-fail in the eyes of the country.

Very few large banks have been allowed to fail and that allows you essentially to own a money printing machine and entrusting industrial houses with that would be problematic for the most part.

That’s precisely why Viral and I asked -- why now?

You’ve just had two accidents, should we focus on learning the lessons from that. For IL&FS, what is it that we’re missing on large NBFCs? How do we improve the whole process? To my mind, there’s blame enough to spread on IL&FS including who the directors were, who the auditors were and, of course, what the regulators and supervisors were doing in monitoring the progress of IL&FS.

Yes Bank is another issue. I mean we have for some time been looking very closely at Yes Bank and trying to understand what it is that they were doing to show such profits while showing such low loan losses. Over time we understood what they were doing and started bringing them to book but, of course, it has come at some cost.

We need to understand this, we have some processes in place in trying to improve the flow of information, automating some of the data transfers from the banks to the RBI, but also trying to understand better what it implies.

So, I would say, let’s get all that straightened out. Let’s make compliance much more effective for the existing banks and, maybe, at some point, we will be able to deal with conflicted banks, banks that start with a huge conflict of interest but because of very stringent regulation and supervision, don’t succumb to them.

We’re not there yet and so to do this now, it seems to me, would be really to invite trouble. Why not fix the problems in the system already, rather than invent new problems?

The holding company if you’re only a finance company you don’t need that holding company structure. I mean, you don’t need a holding company on top of one bank. There are some of those rationalisations that make absolute sense. I guess the promoter holding we’ve gone back and forth on should it be 15% or 26%. I know some people want it to be 26%. You could have a long debate about that and there have been long debates. So, I think the main issue is really this one and I would ask the question, precisely, why now?